target-arm queue for softfreeze: this is quite big as I was on holiday last week, so this is all just sneaking in under the wire. I particularly wanted to get Philippe's patches in before freeze as that sort of code-movement patchset is painful to have to rebase. thanks -- PMM The following changes since commit ae9108f8f0746ce64d02afb1a216153a50926132: Merge remote-tracking branch 'remotes/vivier2/tags/linux-user-for-4.1-pull-request' into staging (2019-07-01 15:55:40 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20190701 for you to fetch changes up to 787a7e76c2e93a48c47b324fea592c9910a70483: target/arm: Declare some M-profile functions publicly (2019-07-01 17:29:01 +0100) ---------------------------------------------------------------- target-arm queue: * hw/arm/boot: fix direct kernel boot with initrd * hw/arm/msf2-som: Exit when the cpu is not the expected one * i.mx7: fix bugs in PCI controller needed to boot recent kernels * aspeed: add RTC device * aspeed: fix some timer device bugs * aspeed: add swift-bmc board * aspeed: vic: Add support for legacy register interface * aspeed: add aspeed-xdma device * Add new sbsa-ref board for aarch64 * target/arm: code refactoring in preparation for support of compilation with TCG disabled ---------------------------------------------------------------- Adriana Kobylak (1): aspeed: Add support for the swift-bmc board Andrew Jeffery (3): aspeed/timer: Status register contains reload for stopped timer aspeed/timer: Fix match calculations aspeed: vic: Add support for legacy register interface Andrew Jones (1): hw/arm/boot: fix direct kernel boot with initrd Andrey Smirnov (5): i.mx7d: Add no-op/unimplemented APBH DMA module i.mx7d: Add no-op/unimplemented PCIE PHY IP block pci: designware: Update MSI mapping unconditionally pci: designware: Update MSI mapping when MSI address changes i.mx7d: pci: Update PCI IRQ mapping to match HW Christian Svensson (1): aspeed/timer: Ensure positive muldiv delta Cédric Le Goater (7): aspeed: add a per SoC mapping for the interrupt space aspeed: add a per SoC mapping for the memory space aspeed: introduce a configurable number of CPU per machine aspeed: add support for multiple NICs aspeed: remove the "ram" link aspeed: add a RAM memory region container aspeed/smc: add a 'sdram_base' property Eddie James (1): hw/misc/aspeed_xdma: New device Hongbo Zhang (2): hw/arm: Add arm SBSA reference machine, skeleton part hw/arm: Add arm SBSA reference machine, devices part Jan Kiszka (1): hw/arm/virt: Add support for Cortex-A7 Joel Stanley (4): hw: timer: Add ASPEED RTC device hw/arm/aspeed: Add RTC to SoC aspeed/timer: Fix behaviour running Linux aspeed: Link SCU to the watchdog Philippe Mathieu-Daudé (19): hw/arm/msf2-som: Exit when the cpu is not the expected one target/arm: Makefile cleanup (Aarch64) target/arm: Makefile cleanup (ARM) target/arm: Makefile cleanup (KVM) target/arm: Makefile cleanup (softmmu) target/arm: Add copyright boilerplate target/arm/helper: Remove unused include target/arm: Fix multiline comment syntax target/arm: Fix coding style issues target/arm: Move CPU state dumping routines to cpu.c target/arm: Declare get_phys_addr() function publicly target/arm: Move TLB related routines to tlb_helper.c target/arm/vfp_helper: Move code around target/arm/vfp_helper: Extract vfp_set_fpscr_to_host() target/arm/vfp_helper: Extract vfp_set_fpscr_from_host() target/arm/vfp_helper: Restrict the SoftFloat use to TCG target/arm: Restrict PSCI to TCG target/arm: Declare arm_log_exception() function publicly target/arm: Declare some M-profile functions publicly Samuel Ortiz (1): target/arm: Move the DC ZVA helper into op_helper hw/arm/Makefile.objs | 1 + hw/misc/Makefile.objs | 1 + hw/timer/Makefile.objs | 2 +- target/arm/Makefile.objs | 24 +- include/hw/arm/aspeed_soc.h | 53 ++- include/hw/arm/fsl-imx7.h | 14 +- include/hw/misc/aspeed_xdma.h | 30 ++ include/hw/ssi/aspeed_smc.h | 3 + include/hw/timer/aspeed_rtc.h | 31 ++ include/hw/watchdog/wdt_aspeed.h | 1 + target/arm/cpu.h | 2 - target/arm/internals.h | 69 ++- target/arm/translate.h | 5 - hw/arm/aspeed.c | 76 +++- hw/arm/aspeed_soc.c | 262 +++++++++--- hw/arm/boot.c | 3 +- hw/arm/fsl-imx7.c | 11 + hw/arm/msf2-som.c | 1 + hw/arm/sbsa-ref.c | 806 ++++++++++++++++++++++++++++++++++++ hw/arm/virt.c | 1 + hw/intc/aspeed_vic.c | 105 +++-- hw/misc/aspeed_xdma.c | 165 ++++++++ hw/pci-host/designware.c | 18 +- hw/ssi/aspeed_smc.c | 1 + hw/timer/aspeed_rtc.c | 180 ++++++++ hw/timer/aspeed_timer.c | 76 ++-- hw/watchdog/wdt_aspeed.c | 20 + target/arm/cpu.c | 232 ++++++++++- target/arm/helper.c | 498 +++++++++------------- target/arm/op_helper.c | 262 ++++++------ target/arm/tlb_helper.c | 200 +++++++++ target/arm/translate-a64.c | 128 ------ target/arm/translate.c | 91 +--- target/arm/vfp_helper.c | 199 +++++---- MAINTAINERS | 8 + default-configs/aarch64-softmmu.mak | 1 + hw/arm/Kconfig | 14 + hw/misc/trace-events | 3 + hw/timer/trace-events | 4 + 39 files changed, 2675 insertions(+), 926 deletions(-) create mode 100644 include/hw/misc/aspeed_xdma.h create mode 100644 include/hw/timer/aspeed_rtc.h create mode 100644 hw/arm/sbsa-ref.c create mode 100644 hw/misc/aspeed_xdma.c create mode 100644 hw/timer/aspeed_rtc.c create mode 100644 target/arm/tlb_helper.c From: Andrew Jones <drjones@redhat.com> Fix the condition used to check whether the initrd fits into RAM; in some cases if an initrd was also passed on the command line we would get an error stating that it was too big to fit into RAM after the kernel. Despite the error the loader continued anyway, though, so also add an exit(1) when the initrd is actually too big. Fixes: 852dc64d665f ("hw/arm/boot: Diagnose layouts that put initrd or DTB off the end of RAM") Signed-off-by: Andrew Jones <drjones@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190618125844.4863-1-drjones@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/boot.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/hw/arm/boot.c b/hw/arm/boot.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/boot.c +++ b/hw/arm/boot.c @@ -XXX,XX +XXX,XX @@ static void arm_setup_direct_kernel_boot(ARMCPU *cpu, info->initrd_filename); exit(1); } - if (info->initrd_start + initrd_size > info->ram_size) { + if (info->initrd_start + initrd_size > ram_end) { error_report("could not load initrd '%s': " "too big to fit into RAM after the kernel", info->initrd_filename); + exit(1); } } else { initrd_size = 0; -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> This machine correctly defines its default_cpu_type to cortex-m3 and report an error if the user requested another cpu_type, however it does not exit, and this can confuse users trying to use another core: $ qemu-system-arm -M emcraft-sf2 -cpu cortex-m4 -kernel test-m4.elf qemu-system-arm: This board can only be used with CPU cortex-m3-arm-cpu [output related to M3 core ...] The CPU is indeed a M3 core: (qemu) info qom-tree /machine (emcraft-sf2-machine) /unattached (container) /device[0] (msf2-soc) /armv7m (armv7m) /cpu (cortex-m3-arm-cpu) Add the missing exit() call to return to the shell. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Eduardo Habkost <ehabkost@redhat.com> Message-id: 20190617160136.29930-1-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/msf2-som.c | 1 + 1 file changed, 1 insertion(+) diff --git a/hw/arm/msf2-som.c b/hw/arm/msf2-som.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/msf2-som.c +++ b/hw/arm/msf2-som.c @@ -XXX,XX +XXX,XX @@ static void emcraft_sf2_s2s010_init(MachineState *machine) if (strcmp(machine->cpu_type, mc->default_cpu_type) != 0) { error_report("This board can only be used with CPU %s", mc->default_cpu_type); + exit(1); } memory_region_init_ram(ddr, NULL, "ddr-ram", DDR_SIZE, -- 2.20.1 From: Jan Kiszka <jan.kiszka@siemens.com> Allow cortex-a7 to be used with the virt board; it supports the v7VE features and there is no reason to deny this type. Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: fc5404f7-4d1d-c28f-6e48-d8799c82acc0@web.de Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/virt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/hw/arm/virt.c b/hw/arm/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/virt.c +++ b/hw/arm/virt.c @@ -XXX,XX +XXX,XX @@ static const int a15irqmap[] = { }; static const char *valid_cpus[] = { + ARM_CPU_TYPE_NAME("cortex-a7"), ARM_CPU_TYPE_NAME("cortex-a15"), ARM_CPU_TYPE_NAME("cortex-a53"), ARM_CPU_TYPE_NAME("cortex-a57"), -- 2.20.1 From: Andrey Smirnov <andrew.smirnov@gmail.com> Instantiate no-op APBH DMA module. Needed to boot latest Linux kernel. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: qemu-devel@nongnu.org Cc: qemu-arm@nongnu.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/fsl-imx7.h | 3 +++ hw/arm/fsl-imx7.c | 6 ++++++ 2 files changed, 9 insertions(+) diff --git a/include/hw/arm/fsl-imx7.h b/include/hw/arm/fsl-imx7.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/fsl-imx7.h +++ b/include/hw/arm/fsl-imx7.h @@ -XXX,XX +XXX,XX @@ enum FslIMX7MemoryMap { FSL_IMX7_PCIE_REG_SIZE = 16 * 1024, FSL_IMX7_GPR_ADDR = 0x30340000, + + FSL_IMX7_DMA_APBH_ADDR = 0x33000000, + FSL_IMX7_DMA_APBH_SIZE = 0x2000, }; enum FslIMX7IRQs { diff --git a/hw/arm/fsl-imx7.c b/hw/arm/fsl-imx7.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/fsl-imx7.c +++ b/hw/arm/fsl-imx7.c @@ -XXX,XX +XXX,XX @@ static void fsl_imx7_realize(DeviceState *dev, Error **errp) */ create_unimplemented_device("lcdif", FSL_IMX7_LCDIF_ADDR, FSL_IMX7_LCDIF_SIZE); + + /* + * DMA APBH + */ + create_unimplemented_device("dma-apbh", FSL_IMX7_DMA_APBH_ADDR, + FSL_IMX7_DMA_APBH_SIZE); } static void fsl_imx7_class_init(ObjectClass *oc, void *data) -- 2.20.1 From: Andrey Smirnov <andrew.smirnov@gmail.com> Add no-op/unimplemented PCIE PHY IP block. Needed by new kernels to use PCIE. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: qemu-devel@nongnu.org Cc: qemu-arm@nongnu.org Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/fsl-imx7.h | 3 +++ hw/arm/fsl-imx7.c | 5 +++++ 2 files changed, 8 insertions(+) diff --git a/include/hw/arm/fsl-imx7.h b/include/hw/arm/fsl-imx7.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/fsl-imx7.h +++ b/include/hw/arm/fsl-imx7.h @@ -XXX,XX +XXX,XX @@ enum FslIMX7MemoryMap { FSL_IMX7_ADC2_ADDR = 0x30620000, FSL_IMX7_ADCn_SIZE = 0x1000, + FSL_IMX7_PCIE_PHY_ADDR = 0x306D0000, + FSL_IMX7_PCIE_PHY_SIZE = 0x10000, + FSL_IMX7_GPC_ADDR = 0x303A0000, FSL_IMX7_I2C1_ADDR = 0x30A20000, diff --git a/hw/arm/fsl-imx7.c b/hw/arm/fsl-imx7.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/fsl-imx7.c +++ b/hw/arm/fsl-imx7.c @@ -XXX,XX +XXX,XX @@ static void fsl_imx7_realize(DeviceState *dev, Error **errp) */ create_unimplemented_device("dma-apbh", FSL_IMX7_DMA_APBH_ADDR, FSL_IMX7_DMA_APBH_SIZE); + /* + * PCIe PHY + */ + create_unimplemented_device("pcie-phy", FSL_IMX7_PCIE_PHY_ADDR, + FSL_IMX7_PCIE_PHY_SIZE); } static void fsl_imx7_class_init(ObjectClass *oc, void *data) -- 2.20.1 From: Andrey Smirnov <andrew.smirnov@gmail.com> Expression to calculate update_msi_mapping in code handling writes to DESIGNWARE_PCIE_MSI_INTR0_ENABLE is missing an ! operator and should be: !!root->msi.intr[0].enable ^ !!val; so that MSI mapping is updated when enabled transitions from either "none" -> "any" or "any" -> "none". Since that register shouldn't be written to very often, change the code to update MSI mapping unconditionally instead of trying to fix the update_msi_mapping logic. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: qemu-devel@nongnu.org Cc: qemu-arm@nongnu.org Acked-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/pci-host/designware.c | 10 ++-------- 1 file changed, 2 insertions(+), 8 deletions(-) diff --git a/hw/pci-host/designware.c b/hw/pci-host/designware.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci-host/designware.c +++ b/hw/pci-host/designware.c @@ -XXX,XX +XXX,XX @@ static void designware_pcie_root_config_write(PCIDevice *d, uint32_t address, root->msi.base |= (uint64_t)val << 32; break; - case DESIGNWARE_PCIE_MSI_INTR0_ENABLE: { - const bool update_msi_mapping = !root->msi.intr[0].enable ^ !!val; - + case DESIGNWARE_PCIE_MSI_INTR0_ENABLE: root->msi.intr[0].enable = val; - - if (update_msi_mapping) { - designware_pcie_root_update_msi_mapping(root); - } + designware_pcie_root_update_msi_mapping(root); break; - } case DESIGNWARE_PCIE_MSI_INTR0_MASK: root->msi.intr[0].mask = val; -- 2.20.1 From: Andrey Smirnov <andrew.smirnov@gmail.com> MSI mapping needs to be update when MSI address changes, so add the code to do so. Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: qemu-devel@nongnu.org Cc: qemu-arm@nongnu.org Acked-by: Michael S. Tsirkin <mst@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/pci-host/designware.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/hw/pci-host/designware.c b/hw/pci-host/designware.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci-host/designware.c +++ b/hw/pci-host/designware.c @@ -XXX,XX +XXX,XX @@ static void designware_pcie_root_config_write(PCIDevice *d, uint32_t address, case DESIGNWARE_PCIE_MSI_ADDR_LO: root->msi.base &= 0xFFFFFFFF00000000ULL; root->msi.base |= val; + designware_pcie_root_update_msi_mapping(root); break; case DESIGNWARE_PCIE_MSI_ADDR_HI: root->msi.base &= 0x00000000FFFFFFFFULL; root->msi.base |= (uint64_t)val << 32; + designware_pcie_root_update_msi_mapping(root); break; case DESIGNWARE_PCIE_MSI_INTR0_ENABLE: -- 2.20.1 From: Andrey Smirnov <andrew.smirnov@gmail.com> Datasheet for i.MX7 is incorrect and i.MX7's PCI IRQ mapping matches that of i.MX6: * INTD/MSI 122 * INTC 123 * INTB 124 * INTA 125 Fix all of the relevant code to reflect that fact. Needed by latest Linux kernels. (Reference: Linux kernel commit 538d6e9d597584e80 from an NXP employee confirming that the datasheet is incorrect and with a report of a test against hardware.) Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com> Cc: Peter Maydell <peter.maydell@linaro.org> Cc: Michael S. Tsirkin <mst@redhat.com> Cc: qemu-devel@nongnu.org Cc: qemu-arm@nongnu.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> [PMM: added ref to kernel commit confirming the datasheet error] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/fsl-imx7.h | 8 ++++---- hw/pci-host/designware.c | 6 ++++-- 2 files changed, 8 insertions(+), 6 deletions(-) diff --git a/include/hw/arm/fsl-imx7.h b/include/hw/arm/fsl-imx7.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/fsl-imx7.h +++ b/include/hw/arm/fsl-imx7.h @@ -XXX,XX +XXX,XX @@ enum FslIMX7IRQs { FSL_IMX7_USB2_IRQ = 42, FSL_IMX7_USB3_IRQ = 40, - FSL_IMX7_PCI_INTA_IRQ = 122, - FSL_IMX7_PCI_INTB_IRQ = 123, - FSL_IMX7_PCI_INTC_IRQ = 124, - FSL_IMX7_PCI_INTD_IRQ = 125, + FSL_IMX7_PCI_INTA_IRQ = 125, + FSL_IMX7_PCI_INTB_IRQ = 124, + FSL_IMX7_PCI_INTC_IRQ = 123, + FSL_IMX7_PCI_INTD_IRQ = 122, FSL_IMX7_UART7_IRQ = 126, diff --git a/hw/pci-host/designware.c b/hw/pci-host/designware.c index XXXXXXX..XXXXXXX 100644 --- a/hw/pci-host/designware.c +++ b/hw/pci-host/designware.c @@ -XXX,XX +XXX,XX @@ #define DESIGNWARE_PCIE_ATU_DEVFN(x) (((x) >> 16) & 0xff) #define DESIGNWARE_PCIE_ATU_UPPER_TARGET 0x91C +#define DESIGNWARE_PCIE_IRQ_MSI 3 + static DesignwarePCIEHost * designware_pcie_root_to_host(DesignwarePCIERoot *root) { @@ -XXX,XX +XXX,XX @@ static void designware_pcie_root_msi_write(void *opaque, hwaddr addr, root->msi.intr[0].status |= BIT(val) & root->msi.intr[0].enable; if (root->msi.intr[0].status & ~root->msi.intr[0].mask) { - qemu_set_irq(host->pci.irqs[0], 1); + qemu_set_irq(host->pci.irqs[DESIGNWARE_PCIE_IRQ_MSI], 1); } } @@ -XXX,XX +XXX,XX @@ static void designware_pcie_root_config_write(PCIDevice *d, uint32_t address, case DESIGNWARE_PCIE_MSI_INTR0_STATUS: root->msi.intr[0].status ^= val; if (!root->msi.intr[0].status) { - qemu_set_irq(host->pci.irqs[0], 0); + qemu_set_irq(host->pci.irqs[DESIGNWARE_PCIE_IRQ_MSI], 0); } break; -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> This will simplify the definition of new SoCs, like the AST2600 which should use a different CPU and a different IRQ number layout. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-2-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 36 +++++++++++++++++++++++ hw/arm/aspeed_soc.c | 57 +++++++++++++++++++++++++++++++------ 2 files changed, 85 insertions(+), 8 deletions(-) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCInfo { const char *fmc_typename; const char **spi_typename; int wdts_num; + const int *irqmap; } AspeedSoCInfo; typedef struct AspeedSoCClass { @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCClass { #define ASPEED_SOC_GET_CLASS(obj) \ OBJECT_GET_CLASS(AspeedSoCClass, (obj), TYPE_ASPEED_SOC) +enum { + ASPEED_IOMEM, + ASPEED_UART1, + ASPEED_UART2, + ASPEED_UART3, + ASPEED_UART4, + ASPEED_UART5, + ASPEED_VUART, + ASPEED_FMC, + ASPEED_SPI1, + ASPEED_SPI2, + ASPEED_VIC, + ASPEED_SDMC, + ASPEED_SCU, + ASPEED_ADC, + ASPEED_SRAM, + ASPEED_GPIO, + ASPEED_RTC, + ASPEED_TIMER1, + ASPEED_TIMER2, + ASPEED_TIMER3, + ASPEED_TIMER4, + ASPEED_TIMER5, + ASPEED_TIMER6, + ASPEED_TIMER7, + ASPEED_TIMER8, + ASPEED_WDT, + ASPEED_PWM, + ASPEED_LPC, + ASPEED_IBT, + ASPEED_I2C, + ASPEED_ETH1, + ASPEED_ETH2, +}; + #endif /* ASPEED_SOC_H */ diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ #define ASPEED_SOC_ETH1_BASE 0x1E660000 #define ASPEED_SOC_ETH2_BASE 0x1E680000 -static const int uart_irqs[] = { 9, 32, 33, 34, 10 }; -static const int timer_irqs[] = { 16, 17, 18, 35, 36, 37, 38, 39, }; +static const int aspeed_soc_ast2400_irqmap[] = { + [ASPEED_UART1] = 9, + [ASPEED_UART2] = 32, + [ASPEED_UART3] = 33, + [ASPEED_UART4] = 34, + [ASPEED_UART5] = 10, + [ASPEED_VUART] = 8, + [ASPEED_FMC] = 19, + [ASPEED_SDMC] = 0, + [ASPEED_SCU] = 21, + [ASPEED_ADC] = 31, + [ASPEED_GPIO] = 20, + [ASPEED_RTC] = 22, + [ASPEED_TIMER1] = 16, + [ASPEED_TIMER2] = 17, + [ASPEED_TIMER3] = 18, + [ASPEED_TIMER4] = 35, + [ASPEED_TIMER5] = 36, + [ASPEED_TIMER6] = 37, + [ASPEED_TIMER7] = 38, + [ASPEED_TIMER8] = 39, + [ASPEED_WDT] = 27, + [ASPEED_PWM] = 28, + [ASPEED_LPC] = 8, + [ASPEED_IBT] = 8, /* LPC */ + [ASPEED_I2C] = 12, + [ASPEED_ETH1] = 2, + [ASPEED_ETH2] = 3, +}; #define AST2400_SDRAM_BASE 0x40000000 #define AST2500_SDRAM_BASE 0x80000000 +/* AST2500 uses the same IRQs as the AST2400 */ +#define aspeed_soc_ast2500_irqmap aspeed_soc_ast2400_irqmap + static const hwaddr aspeed_soc_ast2400_spi_bases[] = { ASPEED_SOC_SPI_BASE }; static const char *aspeed_soc_ast2400_typenames[] = { "aspeed.smc.spi" }; @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, + .irqmap = aspeed_soc_ast2400_irqmap, }, { .name = "ast2400-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, + .irqmap = aspeed_soc_ast2400_irqmap, }, { .name = "ast2400", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, + .irqmap = aspeed_soc_ast2400_irqmap, }, { .name = "ast2500-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm1176"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .fmc_typename = "aspeed.smc.ast2500-fmc", .spi_typename = aspeed_soc_ast2500_typenames, .wdts_num = 3, + .irqmap = aspeed_soc_ast2500_irqmap, }, }; +static qemu_irq aspeed_soc_get_irq(AspeedSoCState *s, int ctrl) +{ + AspeedSoCClass *sc = ASPEED_SOC_GET_CLASS(s); + + return qdev_get_gpio_in(DEVICE(&s->vic), sc->info->irqmap[ctrl]); +} + static void aspeed_soc_init(Object *obj) { AspeedSoCState *s = ASPEED_SOC(obj); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) return; } sysbus_mmio_map(SYS_BUS_DEVICE(&s->timerctrl), 0, ASPEED_SOC_TIMER_BASE); - for (i = 0; i < ARRAY_SIZE(timer_irqs); i++) { - qemu_irq irq = qdev_get_gpio_in(DEVICE(&s->vic), timer_irqs[i]); + for (i = 0; i < ASPEED_TIMER_NR_TIMERS; i++) { + qemu_irq irq = aspeed_soc_get_irq(s, ASPEED_TIMER1 + i); sysbus_connect_irq(SYS_BUS_DEVICE(&s->timerctrl), i, irq); } /* UART - attach an 8250 to the IO space as our UART5 */ if (serial_hd(0)) { - qemu_irq uart5 = qdev_get_gpio_in(DEVICE(&s->vic), uart_irqs[4]); + qemu_irq uart5 = aspeed_soc_get_irq(s, ASPEED_UART5); serial_mm_init(get_system_memory(), ASPEED_SOC_IOMEM_BASE + ASPEED_SOC_UART_5_BASE, 2, uart5, 38400, serial_hd(0), DEVICE_LITTLE_ENDIAN); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) } sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c), 0, ASPEED_SOC_I2C_BASE); sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c), 0, - qdev_get_gpio_in(DEVICE(&s->vic), 12)); + aspeed_soc_get_irq(s, ASPEED_I2C)); /* FMC, The number of CS is set at the board level */ object_property_set_bool(OBJECT(&s->fmc), true, "realized", &err); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) sysbus_mmio_map(SYS_BUS_DEVICE(&s->fmc), 1, s->fmc.ctrl->flash_window_base); sysbus_connect_irq(SYS_BUS_DEVICE(&s->fmc), 0, - qdev_get_gpio_in(DEVICE(&s->vic), 19)); + aspeed_soc_get_irq(s, ASPEED_FMC)); /* SPI */ for (i = 0; i < sc->info->spis_num; i++) { @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) } sysbus_mmio_map(SYS_BUS_DEVICE(&s->ftgmac100), 0, ASPEED_SOC_ETH1_BASE); sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100), 0, - qdev_get_gpio_in(DEVICE(&s->vic), 2)); + aspeed_soc_get_irq(s, ASPEED_ETH1)); } static void aspeed_soc_class_init(ObjectClass *oc, void *data) -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> This will simplify the definition of new SoCs, like the AST2600 which should use a slightly different address space and have a different set of controllers. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-3-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 4 +- hw/arm/aspeed.c | 8 +-- hw/arm/aspeed_soc.c | 117 ++++++++++++++++++++++-------------- 3 files changed, 78 insertions(+), 51 deletions(-) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCInfo { const char *name; const char *cpu_type; uint32_t silicon_rev; - hwaddr sdram_base; uint64_t sram_size; int spis_num; - const hwaddr *spi_bases; const char *fmc_typename; const char **spi_typename; int wdts_num; const int *irqmap; + const hwaddr *memmap; } AspeedSoCInfo; typedef struct AspeedSoCClass { @@ -XXX,XX +XXX,XX @@ enum { ASPEED_I2C, ASPEED_ETH1, ASPEED_ETH2, + ASPEED_SDRAM, }; #endif /* ASPEED_SOC_H */ diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, &error_abort); memory_region_allocate_system_memory(&bmc->ram, NULL, "ram", ram_size); - memory_region_add_subregion(get_system_memory(), sc->info->sdram_base, - &bmc->ram); + memory_region_add_subregion(get_system_memory(), + sc->info->memmap[ASPEED_SDRAM], &bmc->ram); object_property_add_const_link(OBJECT(&bmc->soc), "ram", OBJECT(&bmc->ram), &error_abort); @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, memory_region_init_io(&bmc->max_ram, NULL, &max_ram_ops, NULL, "max_ram", max_ram_size - ram_size); memory_region_add_subregion(get_system_memory(), - sc->info->sdram_base + ram_size, + sc->info->memmap[ASPEED_SDRAM] + ram_size, &bmc->max_ram); aspeed_board_init_flashes(&bmc->soc.fmc, cfg->fmc_model, &error_abort); @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, aspeed_board_binfo.initrd_filename = machine->initrd_filename; aspeed_board_binfo.kernel_cmdline = machine->kernel_cmdline; aspeed_board_binfo.ram_size = ram_size; - aspeed_board_binfo.loader_start = sc->info->sdram_base; + aspeed_board_binfo.loader_start = sc->info->memmap[ASPEED_SDRAM]; if (cfg->i2c_init) { cfg->i2c_init(bmc); diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ #include "hw/i2c/aspeed_i2c.h" #include "net/net.h" -#define ASPEED_SOC_UART_5_BASE 0x00184000 #define ASPEED_SOC_IOMEM_SIZE 0x00200000 -#define ASPEED_SOC_IOMEM_BASE 0x1E600000 -#define ASPEED_SOC_FMC_BASE 0x1E620000 -#define ASPEED_SOC_SPI_BASE 0x1E630000 -#define ASPEED_SOC_SPI2_BASE 0x1E631000 -#define ASPEED_SOC_VIC_BASE 0x1E6C0000 -#define ASPEED_SOC_SDMC_BASE 0x1E6E0000 -#define ASPEED_SOC_SCU_BASE 0x1E6E2000 -#define ASPEED_SOC_SRAM_BASE 0x1E720000 -#define ASPEED_SOC_TIMER_BASE 0x1E782000 -#define ASPEED_SOC_WDT_BASE 0x1E785000 -#define ASPEED_SOC_I2C_BASE 0x1E78A000 -#define ASPEED_SOC_ETH1_BASE 0x1E660000 -#define ASPEED_SOC_ETH2_BASE 0x1E680000 + +static const hwaddr aspeed_soc_ast2400_memmap[] = { + [ASPEED_IOMEM] = 0x1E600000, + [ASPEED_FMC] = 0x1E620000, + [ASPEED_SPI1] = 0x1E630000, + [ASPEED_VIC] = 0x1E6C0000, + [ASPEED_SDMC] = 0x1E6E0000, + [ASPEED_SCU] = 0x1E6E2000, + [ASPEED_ADC] = 0x1E6E9000, + [ASPEED_SRAM] = 0x1E720000, + [ASPEED_GPIO] = 0x1E780000, + [ASPEED_RTC] = 0x1E781000, + [ASPEED_TIMER1] = 0x1E782000, + [ASPEED_WDT] = 0x1E785000, + [ASPEED_PWM] = 0x1E786000, + [ASPEED_LPC] = 0x1E789000, + [ASPEED_IBT] = 0x1E789140, + [ASPEED_I2C] = 0x1E78A000, + [ASPEED_ETH1] = 0x1E660000, + [ASPEED_ETH2] = 0x1E680000, + [ASPEED_UART1] = 0x1E783000, + [ASPEED_UART5] = 0x1E784000, + [ASPEED_VUART] = 0x1E787000, + [ASPEED_SDRAM] = 0x40000000, +}; + +static const hwaddr aspeed_soc_ast2500_memmap[] = { + [ASPEED_IOMEM] = 0x1E600000, + [ASPEED_FMC] = 0x1E620000, + [ASPEED_SPI1] = 0x1E630000, + [ASPEED_SPI2] = 0x1E631000, + [ASPEED_VIC] = 0x1E6C0000, + [ASPEED_SDMC] = 0x1E6E0000, + [ASPEED_SCU] = 0x1E6E2000, + [ASPEED_ADC] = 0x1E6E9000, + [ASPEED_SRAM] = 0x1E720000, + [ASPEED_GPIO] = 0x1E780000, + [ASPEED_RTC] = 0x1E781000, + [ASPEED_TIMER1] = 0x1E782000, + [ASPEED_WDT] = 0x1E785000, + [ASPEED_PWM] = 0x1E786000, + [ASPEED_LPC] = 0x1E789000, + [ASPEED_IBT] = 0x1E789140, + [ASPEED_I2C] = 0x1E78A000, + [ASPEED_ETH1] = 0x1E660000, + [ASPEED_ETH2] = 0x1E680000, + [ASPEED_UART1] = 0x1E783000, + [ASPEED_UART5] = 0x1E784000, + [ASPEED_VUART] = 0x1E787000, + [ASPEED_SDRAM] = 0x80000000, +}; static const int aspeed_soc_ast2400_irqmap[] = { [ASPEED_UART1] = 9, @@ -XXX,XX +XXX,XX @@ static const int aspeed_soc_ast2400_irqmap[] = { [ASPEED_ETH2] = 3, }; -#define AST2400_SDRAM_BASE 0x40000000 -#define AST2500_SDRAM_BASE 0x80000000 - -/* AST2500 uses the same IRQs as the AST2400 */ #define aspeed_soc_ast2500_irqmap aspeed_soc_ast2400_irqmap -static const hwaddr aspeed_soc_ast2400_spi_bases[] = { ASPEED_SOC_SPI_BASE }; static const char *aspeed_soc_ast2400_typenames[] = { "aspeed.smc.spi" }; - -static const hwaddr aspeed_soc_ast2500_spi_bases[] = { ASPEED_SOC_SPI_BASE, - ASPEED_SOC_SPI2_BASE}; static const char *aspeed_soc_ast2500_typenames[] = { "aspeed.smc.ast2500-spi1", "aspeed.smc.ast2500-spi2" }; @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .name = "ast2400-a0", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), .silicon_rev = AST2400_A0_SILICON_REV, - .sdram_base = AST2400_SDRAM_BASE, .sram_size = 0x8000, .spis_num = 1, - .spi_bases = aspeed_soc_ast2400_spi_bases, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, + .memmap = aspeed_soc_ast2400_memmap, }, { .name = "ast2400-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), .silicon_rev = AST2400_A1_SILICON_REV, - .sdram_base = AST2400_SDRAM_BASE, .sram_size = 0x8000, .spis_num = 1, - .spi_bases = aspeed_soc_ast2400_spi_bases, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, + .memmap = aspeed_soc_ast2400_memmap, }, { .name = "ast2400", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), .silicon_rev = AST2400_A0_SILICON_REV, - .sdram_base = AST2400_SDRAM_BASE, .sram_size = 0x8000, .spis_num = 1, - .spi_bases = aspeed_soc_ast2400_spi_bases, .fmc_typename = "aspeed.smc.fmc", .spi_typename = aspeed_soc_ast2400_typenames, .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, + .memmap = aspeed_soc_ast2400_memmap, }, { .name = "ast2500-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm1176"), .silicon_rev = AST2500_A1_SILICON_REV, - .sdram_base = AST2500_SDRAM_BASE, .sram_size = 0x9000, .spis_num = 2, - .spi_bases = aspeed_soc_ast2500_spi_bases, .fmc_typename = "aspeed.smc.ast2500-fmc", .spi_typename = aspeed_soc_ast2500_typenames, .wdts_num = 3, .irqmap = aspeed_soc_ast2500_irqmap, + .memmap = aspeed_soc_ast2500_memmap, }, }; @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) Error *err = NULL, *local_err = NULL; /* IO space */ - create_unimplemented_device("aspeed_soc.io", - ASPEED_SOC_IOMEM_BASE, ASPEED_SOC_IOMEM_SIZE); + create_unimplemented_device("aspeed_soc.io", sc->info->memmap[ASPEED_IOMEM], + ASPEED_SOC_IOMEM_SIZE); /* CPU */ object_property_set_bool(OBJECT(&s->cpu), true, "realized", &err); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - memory_region_add_subregion(get_system_memory(), ASPEED_SOC_SRAM_BASE, - &s->sram); + memory_region_add_subregion(get_system_memory(), + sc->info->memmap[ASPEED_SRAM], &s->sram); /* SCU */ object_property_set_bool(OBJECT(&s->scu), true, "realized", &err); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->scu), 0, ASPEED_SOC_SCU_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->scu), 0, sc->info->memmap[ASPEED_SCU]); /* VIC */ object_property_set_bool(OBJECT(&s->vic), true, "realized", &err); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->vic), 0, ASPEED_SOC_VIC_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->vic), 0, sc->info->memmap[ASPEED_VIC]); sysbus_connect_irq(SYS_BUS_DEVICE(&s->vic), 0, qdev_get_gpio_in(DEVICE(&s->cpu), ARM_CPU_IRQ)); sysbus_connect_irq(SYS_BUS_DEVICE(&s->vic), 1, @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->timerctrl), 0, ASPEED_SOC_TIMER_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->timerctrl), 0, + sc->info->memmap[ASPEED_TIMER1]); for (i = 0; i < ASPEED_TIMER_NR_TIMERS; i++) { qemu_irq irq = aspeed_soc_get_irq(s, ASPEED_TIMER1 + i); sysbus_connect_irq(SYS_BUS_DEVICE(&s->timerctrl), i, irq); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) /* UART - attach an 8250 to the IO space as our UART5 */ if (serial_hd(0)) { qemu_irq uart5 = aspeed_soc_get_irq(s, ASPEED_UART5); - serial_mm_init(get_system_memory(), - ASPEED_SOC_IOMEM_BASE + ASPEED_SOC_UART_5_BASE, 2, + serial_mm_init(get_system_memory(), sc->info->memmap[ASPEED_UART5], 2, uart5, 38400, serial_hd(0), DEVICE_LITTLE_ENDIAN); } @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c), 0, ASPEED_SOC_I2C_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->i2c), 0, sc->info->memmap[ASPEED_I2C]); sysbus_connect_irq(SYS_BUS_DEVICE(&s->i2c), 0, aspeed_soc_get_irq(s, ASPEED_I2C)); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->fmc), 0, ASPEED_SOC_FMC_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->fmc), 0, sc->info->memmap[ASPEED_FMC]); sysbus_mmio_map(SYS_BUS_DEVICE(&s->fmc), 1, s->fmc.ctrl->flash_window_base); sysbus_connect_irq(SYS_BUS_DEVICE(&s->fmc), 0, @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->spi[i]), 0, sc->info->spi_bases[i]); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->spi[i]), 0, + sc->info->memmap[ASPEED_SPI1 + i]); sysbus_mmio_map(SYS_BUS_DEVICE(&s->spi[i]), 1, s->spi[i].ctrl->flash_window_base); } @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->sdmc), 0, ASPEED_SOC_SDMC_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->sdmc), 0, sc->info->memmap[ASPEED_SDMC]); /* Watch dog */ for (i = 0; i < sc->info->wdts_num; i++) { @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) return; } sysbus_mmio_map(SYS_BUS_DEVICE(&s->wdt[i]), 0, - ASPEED_SOC_WDT_BASE + i * 0x20); + sc->info->memmap[ASPEED_WDT] + i * 0x20); } /* Net */ @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) error_propagate(errp, err); return; } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->ftgmac100), 0, ASPEED_SOC_ETH1_BASE); + sysbus_mmio_map(SYS_BUS_DEVICE(&s->ftgmac100), 0, + sc->info->memmap[ASPEED_ETH1]); sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100), 0, aspeed_soc_get_irq(s, ASPEED_ETH1)); } -- 2.20.1 From: Joel Stanley <joel@jms.id.au> The RTC is modeled to provide time and date functionality. It is initialised at zero to match the hardware. There is no modelling of the alarm functionality, which includes the IRQ line. As there is no guest code to exercise this function that is acceptable for now. Signed-off-by: Joel Stanley <joel@jms.id.au> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20190618165311.27066-4-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/Makefile.objs | 2 +- include/hw/timer/aspeed_rtc.h | 31 ++++++ hw/timer/aspeed_rtc.c | 180 ++++++++++++++++++++++++++++++++++ hw/timer/trace-events | 4 + 4 files changed, 216 insertions(+), 1 deletion(-) create mode 100644 include/hw/timer/aspeed_rtc.h create mode 100644 hw/timer/aspeed_rtc.c diff --git a/hw/timer/Makefile.objs b/hw/timer/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/Makefile.objs +++ b/hw/timer/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_MC146818RTC) += mc146818rtc.o obj-$(CONFIG_ALLWINNER_A10_PIT) += allwinner-a10-pit.o common-obj-$(CONFIG_STM32F2XX_TIMER) += stm32f2xx_timer.o -common-obj-$(CONFIG_ASPEED_SOC) += aspeed_timer.o +common-obj-$(CONFIG_ASPEED_SOC) += aspeed_timer.o aspeed_rtc.o common-obj-$(CONFIG_SUN4V_RTC) += sun4v-rtc.o common-obj-$(CONFIG_CMSDK_APB_TIMER) += cmsdk-apb-timer.o diff --git a/include/hw/timer/aspeed_rtc.h b/include/hw/timer/aspeed_rtc.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/timer/aspeed_rtc.h @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED Real Time Clock + * Joel Stanley <joel@jms.id.au> + * + * Copyright 2019 IBM Corp + * SPDX-License-Identifier: GPL-2.0-or-later + */ +#ifndef ASPEED_RTC_H +#define ASPEED_RTC_H + +#include <stdint.h> + +#include "hw/hw.h" +#include "hw/irq.h" +#include "hw/sysbus.h" + +typedef struct AspeedRtcState { + SysBusDevice parent_obj; + + MemoryRegion iomem; + qemu_irq irq; + + uint32_t reg[0x18]; + int offset; + +} AspeedRtcState; + +#define TYPE_ASPEED_RTC "aspeed.rtc" +#define ASPEED_RTC(obj) OBJECT_CHECK(AspeedRtcState, (obj), TYPE_ASPEED_RTC) + +#endif /* ASPEED_RTC_H */ diff --git a/hw/timer/aspeed_rtc.c b/hw/timer/aspeed_rtc.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/timer/aspeed_rtc.c @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED Real Time Clock + * Joel Stanley <joel@jms.id.au> + * + * Copyright 2019 IBM Corp + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "qemu-common.h" +#include "hw/timer/aspeed_rtc.h" +#include "qemu/log.h" +#include "qemu/timer.h" + +#include "trace.h" + +#define COUNTER1 (0x00 / 4) +#define COUNTER2 (0x04 / 4) +#define ALARM (0x08 / 4) +#define CONTROL (0x10 / 4) +#define ALARM_STATUS (0x14 / 4) + +#define RTC_UNLOCKED BIT(1) +#define RTC_ENABLED BIT(0) + +static void aspeed_rtc_calc_offset(AspeedRtcState *rtc) +{ + struct tm tm; + uint32_t year, cent; + uint32_t reg1 = rtc->reg[COUNTER1]; + uint32_t reg2 = rtc->reg[COUNTER2]; + + tm.tm_mday = (reg1 >> 24) & 0x1f; + tm.tm_hour = (reg1 >> 16) & 0x1f; + tm.tm_min = (reg1 >> 8) & 0x3f; + tm.tm_sec = (reg1 >> 0) & 0x3f; + + cent = (reg2 >> 16) & 0x1f; + year = (reg2 >> 8) & 0x7f; + tm.tm_mon = ((reg2 >> 0) & 0x0f) - 1; + tm.tm_year = year + (cent * 100) - 1900; + + rtc->offset = qemu_timedate_diff(&tm); +} + +static uint32_t aspeed_rtc_get_counter(AspeedRtcState *rtc, int r) +{ + uint32_t year, cent; + struct tm now; + + qemu_get_timedate(&now, rtc->offset); + + switch (r) { + case COUNTER1: + return (now.tm_mday << 24) | (now.tm_hour << 16) | + (now.tm_min << 8) | now.tm_sec; + case COUNTER2: + cent = (now.tm_year + 1900) / 100; + year = now.tm_year % 100; + return ((cent & 0x1f) << 16) | ((year & 0x7f) << 8) | + ((now.tm_mon + 1) & 0xf); + default: + g_assert_not_reached(); + } +} + +static uint64_t aspeed_rtc_read(void *opaque, hwaddr addr, + unsigned size) +{ + AspeedRtcState *rtc = opaque; + uint64_t val; + uint32_t r = addr >> 2; + + switch (r) { + case COUNTER1: + case COUNTER2: + if (rtc->reg[CONTROL] & RTC_ENABLED) { + rtc->reg[r] = aspeed_rtc_get_counter(rtc, r); + } + /* fall through */ + case CONTROL: + val = rtc->reg[r]; + break; + case ALARM: + case ALARM_STATUS: + default: + qemu_log_mask(LOG_UNIMP, "%s: 0x%" HWADDR_PRIx "\n", __func__, addr); + return 0; + } + + trace_aspeed_rtc_read(addr, val); + + return val; +} + +static void aspeed_rtc_write(void *opaque, hwaddr addr, + uint64_t val, unsigned size) +{ + AspeedRtcState *rtc = opaque; + uint32_t r = addr >> 2; + + switch (r) { + case COUNTER1: + case COUNTER2: + if (!(rtc->reg[CONTROL] & RTC_UNLOCKED)) { + break; + } + /* fall through */ + case CONTROL: + rtc->reg[r] = val; + aspeed_rtc_calc_offset(rtc); + break; + case ALARM: + case ALARM_STATUS: + default: + qemu_log_mask(LOG_UNIMP, "%s: 0x%" HWADDR_PRIx "\n", __func__, addr); + break; + } + trace_aspeed_rtc_write(addr, val); +} + +static void aspeed_rtc_reset(DeviceState *d) +{ + AspeedRtcState *rtc = ASPEED_RTC(d); + + rtc->offset = 0; + memset(rtc->reg, 0, sizeof(rtc->reg)); +} + +static const MemoryRegionOps aspeed_rtc_ops = { + .read = aspeed_rtc_read, + .write = aspeed_rtc_write, + .endianness = DEVICE_NATIVE_ENDIAN, +}; + +static const VMStateDescription vmstate_aspeed_rtc = { + .name = TYPE_ASPEED_RTC, + .version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(reg, AspeedRtcState, 0x18), + VMSTATE_INT32(offset, AspeedRtcState), + VMSTATE_INT32(offset, AspeedRtcState), + VMSTATE_END_OF_LIST() + } +}; + +static void aspeed_rtc_realize(DeviceState *dev, Error **errp) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); + AspeedRtcState *s = ASPEED_RTC(dev); + + sysbus_init_irq(sbd, &s->irq); + + memory_region_init_io(&s->iomem, OBJECT(s), &aspeed_rtc_ops, s, + "aspeed-rtc", 0x18ULL); + sysbus_init_mmio(sbd, &s->iomem); +} + +static void aspeed_rtc_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = aspeed_rtc_realize; + dc->vmsd = &vmstate_aspeed_rtc; + dc->reset = aspeed_rtc_reset; +} + +static const TypeInfo aspeed_rtc_info = { + .name = TYPE_ASPEED_RTC, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(AspeedRtcState), + .class_init = aspeed_rtc_class_init, +}; + +static void aspeed_rtc_register_types(void) +{ + type_register_static(&aspeed_rtc_info); +} + +type_init(aspeed_rtc_register_types) diff --git a/hw/timer/trace-events b/hw/timer/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/trace-events +++ b/hw/timer/trace-events @@ -XXX,XX +XXX,XX @@ cmsdk_apb_dualtimer_read(uint64_t offset, uint64_t data, unsigned size) "CMSDK A cmsdk_apb_dualtimer_write(uint64_t offset, uint64_t data, unsigned size) "CMSDK APB dualtimer write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %u" cmsdk_apb_dualtimer_reset(void) "CMSDK APB dualtimer: reset" +# hw/timer/aspeed-rtc.c +aspeed_rtc_read(uint64_t addr, uint64_t value) "addr 0x%02" PRIx64 " value 0x%08" PRIx64 +aspeed_rtc_write(uint64_t addr, uint64_t value) "addr 0x%02" PRIx64 " value 0x%08" PRIx64 + # sun4v-rtc.c sun4v_rtc_read(uint64_t addr, uint64_t value) "read: addr 0x%" PRIx64 " value 0x%" PRIx64 sun4v_rtc_write(uint64_t addr, uint64_t value) "write: addr 0x%" PRIx64 " value 0x%" PRIx64 -- 2.20.1 From: Joel Stanley <joel@jms.id.au> All systems have an RTC. The IRQ is hooked up but the model does not use it at this stage. There is no guest code that uses it, so this limitation is acceptable. Signed-off-by: Joel Stanley <joel@jms.id.au> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20190618165311.27066-5-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 2 ++ hw/arm/aspeed_soc.c | 13 +++++++++++++ 2 files changed, 15 insertions(+) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ #include "hw/misc/aspeed_scu.h" #include "hw/misc/aspeed_sdmc.h" #include "hw/timer/aspeed_timer.h" +#include "hw/timer/aspeed_rtc.h" #include "hw/i2c/aspeed_i2c.h" #include "hw/ssi/aspeed_smc.h" #include "hw/watchdog/wdt_aspeed.h" @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCState { ARMCPU cpu; MemoryRegion sram; AspeedVICState vic; + AspeedRtcState rtc; AspeedTimerCtrlState timerctrl; AspeedI2CState i2c; AspeedSCUState scu; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sysbus_init_child_obj(obj, "vic", OBJECT(&s->vic), sizeof(s->vic), TYPE_ASPEED_VIC); + sysbus_init_child_obj(obj, "rtc", OBJECT(&s->rtc), sizeof(s->rtc), + TYPE_ASPEED_RTC); + sysbus_init_child_obj(obj, "timerctrl", OBJECT(&s->timerctrl), sizeof(s->timerctrl), TYPE_ASPEED_TIMER); object_property_add_const_link(OBJECT(&s->timerctrl), "scu", @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->vic), 1, qdev_get_gpio_in(DEVICE(&s->cpu), ARM_CPU_FIQ)); + /* RTC */ + object_property_set_bool(OBJECT(&s->rtc), true, "realized", &err); + if (err) { + error_propagate(errp, err); + return; + } + sysbus_mmio_map(SYS_BUS_DEVICE(&s->rtc), 0, sc->info->memmap[ASPEED_RTC]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->rtc), 0, + aspeed_soc_get_irq(s, ASPEED_RTC)); + /* Timer */ object_property_set_bool(OBJECT(&s->timerctrl), true, "realized", &err); if (err) { -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The current models of the Aspeed SoCs only have one CPU but future ones will support SMP. Introduce a new num_cpus field at the SoC class level to define the number of available CPUs per SoC and also introduce a 'num-cpus' property to activate the CPUs configured for the machine. The max_cpus limit of the machine should depend on the SoC definition but, unfortunately, these values are not available when the machine class is initialized. This is the reason why we add a check on num_cpus in the AspeedSoC realize handler. SMP support will be activated when models for such SoCs are implemented. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-6-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 5 ++++- hw/arm/aspeed.c | 7 +++++-- hw/arm/aspeed_soc.c | 33 +++++++++++++++++++++++++++------ 3 files changed, 36 insertions(+), 9 deletions(-) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ #define ASPEED_SPIS_NUM 2 #define ASPEED_WDTS_NUM 3 +#define ASPEED_CPUS_NUM 2 typedef struct AspeedSoCState { /*< private >*/ DeviceState parent; /*< public >*/ - ARMCPU cpu; + ARMCPU cpu[ASPEED_CPUS_NUM]; + uint32_t num_cpus; MemoryRegion sram; AspeedVICState vic; AspeedRtcState rtc; @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCInfo { int wdts_num; const int *irqmap; const hwaddr *memmap; + uint32_t num_cpus; } AspeedSoCInfo; typedef struct AspeedSoCClass { diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ #include "hw/misc/tmp105.h" #include "qemu/log.h" #include "sysemu/block-backend.h" +#include "sysemu/sysemu.h" #include "hw/loader.h" #include "qemu/error-report.h" #include "qemu/units.h" static struct arm_boot_info aspeed_board_binfo = { .board_id = -1, /* device-tree-only board */ - .nb_cpus = 1, }; struct AspeedBoardState { @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, &error_abort); object_property_set_int(OBJECT(&bmc->soc), cfg->num_cs, "num-cs", &error_abort); + object_property_set_int(OBJECT(&bmc->soc), smp_cpus, "num-cpus", + &error_abort); if (machine->kernel_filename) { /* * When booting with a -kernel command line there is no u-boot @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, aspeed_board_binfo.kernel_cmdline = machine->kernel_cmdline; aspeed_board_binfo.ram_size = ram_size; aspeed_board_binfo.loader_start = sc->info->memmap[ASPEED_SDRAM]; + aspeed_board_binfo.nb_cpus = bmc->soc.num_cpus; if (cfg->i2c_init) { cfg->i2c_init(bmc); @@ -XXX,XX +XXX,XX @@ static void aspeed_machine_class_init(ObjectClass *oc, void *data) mc->desc = board->desc; mc->init = aspeed_machine_init; - mc->max_cpus = 1; + mc->max_cpus = ASPEED_CPUS_NUM; mc->no_sdcard = 1; mc->no_floppy = 1; mc->no_cdrom = 1; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ #include "hw/char/serial.h" #include "qemu/log.h" #include "qemu/module.h" +#include "qemu/error-report.h" #include "hw/i2c/aspeed_i2c.h" #include "net/net.h" @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, + .num_cpus = 1, }, { .name = "ast2400-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, + .num_cpus = 1, }, { .name = "ast2400", .cpu_type = ARM_CPU_TYPE_NAME("arm926"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .wdts_num = 2, .irqmap = aspeed_soc_ast2400_irqmap, .memmap = aspeed_soc_ast2400_memmap, + .num_cpus = 1, }, { .name = "ast2500-a1", .cpu_type = ARM_CPU_TYPE_NAME("arm1176"), @@ -XXX,XX +XXX,XX @@ static const AspeedSoCInfo aspeed_socs[] = { .wdts_num = 3, .irqmap = aspeed_soc_ast2500_irqmap, .memmap = aspeed_soc_ast2500_memmap, + .num_cpus = 1, }, }; @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) AspeedSoCClass *sc = ASPEED_SOC_GET_CLASS(s); int i; - object_initialize_child(obj, "cpu", OBJECT(&s->cpu), sizeof(s->cpu), - sc->info->cpu_type, &error_abort, NULL); + for (i = 0; i < sc->info->num_cpus; i++) { + object_initialize_child(obj, "cpu[*]", OBJECT(&s->cpu[i]), + sizeof(s->cpu[i]), sc->info->cpu_type, + &error_abort, NULL); + } sysbus_init_child_obj(obj, "scu", OBJECT(&s->scu), sizeof(s->scu), TYPE_ASPEED_SCU); @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) create_unimplemented_device("aspeed_soc.io", sc->info->memmap[ASPEED_IOMEM], ASPEED_SOC_IOMEM_SIZE); + if (s->num_cpus > sc->info->num_cpus) { + warn_report("%s: invalid number of CPUs %d, using default %d", + sc->info->name, s->num_cpus, sc->info->num_cpus); + s->num_cpus = sc->info->num_cpus; + } + /* CPU */ - object_property_set_bool(OBJECT(&s->cpu), true, "realized", &err); - if (err) { - error_propagate(errp, err); - return; + for (i = 0; i < s->num_cpus; i++) { + object_property_set_bool(OBJECT(&s->cpu[i]), true, "realized", &err); + if (err) { + error_propagate(errp, err); + return; + } } /* SRAM */ @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100), 0, aspeed_soc_get_irq(s, ASPEED_ETH1)); } +static Property aspeed_soc_properties[] = { + DEFINE_PROP_UINT32("num-cpus", AspeedSoCState, num_cpus, 0), + DEFINE_PROP_END_OF_LIST(), +}; static void aspeed_soc_class_init(ObjectClass *oc, void *data) { @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_class_init(ObjectClass *oc, void *data) dc->realize = aspeed_soc_realize; /* Reason: Uses serial_hds and nd_table in realize() directly */ dc->user_creatable = false; + dc->props = aspeed_soc_properties; } static const TypeInfo aspeed_soc_type_info = { -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The Aspeed SoCs have two MACs. Extend the Aspeed model to support a second NIC. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-7-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/arm/aspeed_soc.h | 3 ++- hw/arm/aspeed_soc.c | 33 +++++++++++++++++++-------------- 2 files changed, 21 insertions(+), 15 deletions(-) diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ #define ASPEED_SPIS_NUM 2 #define ASPEED_WDTS_NUM 3 #define ASPEED_CPUS_NUM 2 +#define ASPEED_MACS_NUM 2 typedef struct AspeedSoCState { /*< private >*/ @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCState { AspeedSMCState spi[ASPEED_SPIS_NUM]; AspeedSDMCState sdmc; AspeedWDTState wdt[ASPEED_WDTS_NUM]; - FTGMAC100State ftgmac100; + FTGMAC100State ftgmac100[ASPEED_MACS_NUM]; } AspeedSoCState; #define TYPE_ASPEED_SOC "aspeed-soc" diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sc->info->silicon_rev); } - sysbus_init_child_obj(obj, "ftgmac100", OBJECT(&s->ftgmac100), - sizeof(s->ftgmac100), TYPE_FTGMAC100); + for (i = 0; i < ASPEED_MACS_NUM; i++) { + sysbus_init_child_obj(obj, "ftgmac100[*]", OBJECT(&s->ftgmac100[i]), + sizeof(s->ftgmac100[i]), TYPE_FTGMAC100); + } } static void aspeed_soc_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) } /* Net */ - qdev_set_nic_properties(DEVICE(&s->ftgmac100), &nd_table[0]); - object_property_set_bool(OBJECT(&s->ftgmac100), true, "aspeed", &err); - object_property_set_bool(OBJECT(&s->ftgmac100), true, "realized", - &local_err); - error_propagate(&err, local_err); - if (err) { - error_propagate(errp, err); - return; + for (i = 0; i < nb_nics; i++) { + qdev_set_nic_properties(DEVICE(&s->ftgmac100[i]), &nd_table[i]); + object_property_set_bool(OBJECT(&s->ftgmac100[i]), true, "aspeed", + &err); + object_property_set_bool(OBJECT(&s->ftgmac100[i]), true, "realized", + &local_err); + error_propagate(&err, local_err); + if (err) { + error_propagate(errp, err); + return; + } + sysbus_mmio_map(SYS_BUS_DEVICE(&s->ftgmac100[i]), 0, + sc->info->memmap[ASPEED_ETH1 + i]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100[i]), 0, + aspeed_soc_get_irq(s, ASPEED_ETH1 + i)); } - sysbus_mmio_map(SYS_BUS_DEVICE(&s->ftgmac100), 0, - sc->info->memmap[ASPEED_ETH1]); - sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100), 0, - aspeed_soc_get_irq(s, ASPEED_ETH1)); } static Property aspeed_soc_properties[] = { DEFINE_PROP_UINT32("num-cpus", AspeedSoCState, num_cpus, 0), -- 2.20.1 From: Joel Stanley <joel@jms.id.au> The Linux kernel driver was updated in commit 4451d3f59f2a ("clocksource/drivers/fttmr010: Fix set_next_event handler) to fix an issue observed on hardware: > RELOAD register is loaded into COUNT register when the aspeed timer > is enabled, which means the next event may be delayed because timer > interrupt won't be generated until <0xFFFFFFFF - current_count + > cycles>. When running under Qemu, the system appeared "laggy". The guest is now scheduling timer events too regularly, starving the host of CPU time. This patch modifies the timer model to attempt to schedule the timer expiry as the guest requests, but if we have missed the deadline we re interrupt and try again, which allows the guest to catch up. Provides expected behaviour with old and new guest code. Fixes: c04bd47db6b9 ("hw/timer: Add ASPEED timer device model") Signed-off-by: Joel Stanley <joel@jms.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190618165311.27066-8-clg@kaod.org [clg: - merged a fix from Andrew Jeffery <andrew@aj.id.au> "Fire interrupt on failure to meet deadline" https://lists.ozlabs.org/pipermail/openbmc/2019-January/014641.html - adapted commit log - checkpatch fixes ] Signed-off-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/aspeed_timer.c | 57 ++++++++++++++++++++++------------------- 1 file changed, 30 insertions(+), 27 deletions(-) diff --git a/hw/timer/aspeed_timer.c b/hw/timer/aspeed_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/aspeed_timer.c +++ b/hw/timer/aspeed_timer.c @@ -XXX,XX +XXX,XX @@ static inline uint64_t calculate_time(struct AspeedTimer *t, uint32_t ticks) static uint64_t calculate_next(struct AspeedTimer *t) { - uint64_t next = 0; - uint32_t rate = calculate_rate(t); + uint64_t now = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + uint64_t next; - while (!next) { - /* We don't know the relationship between the values in the match - * registers, so sort using MAX/MIN/zero. We sort in that order as the - * timer counts down to zero. */ - uint64_t seq[] = { - calculate_time(t, MAX(t->match[0], t->match[1])), - calculate_time(t, MIN(t->match[0], t->match[1])), - calculate_time(t, 0), - }; - uint64_t reload_ns; - uint64_t now = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + /* + * We don't know the relationship between the values in the match + * registers, so sort using MAX/MIN/zero. We sort in that order as + * the timer counts down to zero. + */ - if (now < seq[0]) { - next = seq[0]; - } else if (now < seq[1]) { - next = seq[1]; - } else if (now < seq[2]) { - next = seq[2]; - } else if (t->reload) { - reload_ns = muldiv64(t->reload, NANOSECONDS_PER_SECOND, rate); - t->start = now - ((now - t->start) % reload_ns); - } else { - /* no reload value, return 0 */ - break; - } + next = calculate_time(t, MAX(t->match[0], t->match[1])); + if (now < next) { + return next; } - return next; + next = calculate_time(t, MIN(t->match[0], t->match[1])); + if (now < next) { + return next; + } + + next = calculate_time(t, 0); + if (now < next) { + return next; + } + + /* We've missed all deadlines, fire interrupt and try again */ + timer_del(&t->timer); + + if (timer_overflow_interrupt(t)) { + t->level = !t->level; + qemu_set_irq(t->irq, t->level); + } + + t->start = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + return calculate_time(t, MAX(MAX(t->match[0], t->match[1]), 0)); } static void aspeed_timer_mod(AspeedTimer *t) -- 2.20.1 From: Andrew Jeffery <andrew@aj.id.au> From the datasheet: This register stores the current status of counter #N. When timer enable bit TMC30[N * b] is disabled, the reload register will be loaded into this counter. When timer bit TMC30[N * b] is set, the counter will start to decrement. CPU can update this register value when enable bit is set. Signed-off-by: Andrew Jeffery <andrew@aj.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-9-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/aspeed_timer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/hw/timer/aspeed_timer.c b/hw/timer/aspeed_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/aspeed_timer.c +++ b/hw/timer/aspeed_timer.c @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_timer_get_value(AspeedTimer *t, int reg) switch (reg) { case TIMER_REG_STATUS: - value = calculate_ticks(t, qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL)); + if (timer_enabled(t)) { + value = calculate_ticks(t, qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL)); + } else { + value = t->reload; + } break; case TIMER_REG_RELOAD: value = t->reload; -- 2.20.1 From: Andrew Jeffery <andrew@aj.id.au> If the match value exceeds reload then we don't want to include it in calculations for the next event. Signed-off-by: Andrew Jeffery <andrew@aj.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190618165311.27066-10-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/aspeed_timer.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/hw/timer/aspeed_timer.c b/hw/timer/aspeed_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/aspeed_timer.c +++ b/hw/timer/aspeed_timer.c @@ -XXX,XX +XXX,XX @@ static inline uint64_t calculate_time(struct AspeedTimer *t, uint32_t ticks) return t->start + delta_ns; } +static inline uint32_t calculate_match(struct AspeedTimer *t, int i) +{ + return t->match[i] < t->reload ? t->match[i] : 0; +} + static uint64_t calculate_next(struct AspeedTimer *t) { uint64_t now = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); @@ -XXX,XX +XXX,XX @@ static uint64_t calculate_next(struct AspeedTimer *t) * the timer counts down to zero. */ - next = calculate_time(t, MAX(t->match[0], t->match[1])); + next = calculate_time(t, MAX(calculate_match(t, 0), calculate_match(t, 1))); if (now < next) { return next; } - next = calculate_time(t, MIN(t->match[0], t->match[1])); + next = calculate_time(t, MIN(calculate_match(t, 0), calculate_match(t, 1))); if (now < next) { return next; } @@ -XXX,XX +XXX,XX @@ static uint64_t calculate_next(struct AspeedTimer *t) qemu_set_irq(t->irq, t->level); } + next = MAX(MAX(calculate_match(t, 0), calculate_match(t, 1)), 0); t->start = qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); - return calculate_time(t, MAX(MAX(t->match[0], t->match[1]), 0)); + + return calculate_time(t, next); } static void aspeed_timer_mod(AspeedTimer *t) -- 2.20.1 From: Christian Svensson <bluecmd@google.com> If the host decrements the counter register that results in a negative delta. This is then passed to muldiv64 which only handles unsigned numbers resulting in bogus results. This fix ensures the delta being operated on is positive. Test case: kexec a kernel using aspeed_timer and it will freeze on the second bootup when the kernel initializes the timer. With this patch that no longer happens and the timer appears to run OK. Signed-off-by: Christian Svensson <bluecmd@google.com> Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Reviewed-by: Andrew Jeffery <andrew@aj.id.au> Message-id: 20190618165311.27066-12-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/timer/aspeed_timer.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/hw/timer/aspeed_timer.c b/hw/timer/aspeed_timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/aspeed_timer.c +++ b/hw/timer/aspeed_timer.c @@ -XXX,XX +XXX,XX @@ static void aspeed_timer_set_value(AspeedTimerCtrlState *s, int timer, int reg, int64_t delta = (int64_t) value - (int64_t) calculate_ticks(t, now); uint32_t rate = calculate_rate(t); - t->start += muldiv64(delta, NANOSECONDS_PER_SECOND, rate); + if (delta >= 0) { + t->start += muldiv64(delta, NANOSECONDS_PER_SECOND, rate); + } else { + t->start -= muldiv64(-delta, NANOSECONDS_PER_SECOND, rate); + } aspeed_timer_mod(t); } break; -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> It has never been used as far as I can tell from the git history. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-13-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/aspeed.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, memory_region_allocate_system_memory(&bmc->ram, NULL, "ram", ram_size); memory_region_add_subregion(get_system_memory(), sc->info->memmap[ASPEED_SDRAM], &bmc->ram); - object_property_add_const_link(OBJECT(&bmc->soc), "ram", OBJECT(&bmc->ram), - &error_abort); max_ram_size = object_property_get_uint(OBJECT(&bmc->soc), "max-ram-size", &error_abort); -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The RAM memory region is defined after the SoC is realized when the SDMC controller has checked that the defined RAM size for the machine is correct. This is problematic for controller models requiring a link on the RAM region, for DMA support in the SMC controller for instance. Introduce a container memory region for the RAM that we can link into the controllers early, before the SoC is realized. It will be populated with the RAM region after the checks have be done. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-14-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/aspeed.c | 13 +++++++++---- 1 file changed, 9 insertions(+), 4 deletions(-) diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ static struct arm_boot_info aspeed_board_binfo = { struct AspeedBoardState { AspeedSoCState soc; + MemoryRegion ram_container; MemoryRegion ram; MemoryRegion max_ram; }; @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, ram_addr_t max_ram_size; bmc = g_new0(AspeedBoardState, 1); + + memory_region_init(&bmc->ram_container, NULL, "aspeed-ram-container", + UINT32_MAX); + object_initialize_child(OBJECT(machine), "soc", &bmc->soc, (sizeof(bmc->soc)), cfg->soc_name, &error_abort, NULL); @@ -XXX,XX +XXX,XX @@ static void aspeed_board_init(MachineState *machine, &error_abort); memory_region_allocate_system_memory(&bmc->ram, NULL, "ram", ram_size); + memory_region_add_subregion(&bmc->ram_container, 0, &bmc->ram); memory_region_add_subregion(get_system_memory(), - sc->info->memmap[ASPEED_SDRAM], &bmc->ram); + sc->info->memmap[ASPEED_SDRAM], + &bmc->ram_container); max_ram_size = object_property_get_uint(OBJECT(&bmc->soc), "max-ram-size", &error_abort); memory_region_init_io(&bmc->max_ram, NULL, &max_ram_ops, NULL, "max_ram", max_ram_size - ram_size); - memory_region_add_subregion(get_system_memory(), - sc->info->memmap[ASPEED_SDRAM] + ram_size, - &bmc->max_ram); + memory_region_add_subregion(&bmc->ram_container, ram_size, &bmc->max_ram); aspeed_board_init_flashes(&bmc->soc.fmc, cfg->fmc_model, &error_abort); aspeed_board_init_flashes(&bmc->soc.spi[0], cfg->spi_model, &error_abort); -- 2.20.1 From: Cédric Le Goater <clg@kaod.org> The DRAM address of a DMA transaction depends on the DRAM base address of the SoC. Inform the SMC controller model with this value. Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190618165311.27066-15-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/ssi/aspeed_smc.h | 3 +++ hw/arm/aspeed_soc.c | 6 ++++++ hw/ssi/aspeed_smc.c | 1 + 3 files changed, 10 insertions(+) diff --git a/include/hw/ssi/aspeed_smc.h b/include/hw/ssi/aspeed_smc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ssi/aspeed_smc.h +++ b/include/hw/ssi/aspeed_smc.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSMCState { uint8_t r_timings; uint8_t conf_enable_w0; + /* for DMA support */ + uint64_t sdram_base; + AspeedSMCFlash *flashes; uint8_t snoop_index; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) aspeed_soc_get_irq(s, ASPEED_I2C)); /* FMC, The number of CS is set at the board level */ + object_property_set_int(OBJECT(&s->fmc), sc->info->memmap[ASPEED_SDRAM], + "sdram-base", &err); + if (err) { + error_propagate(errp, err); + return; + } object_property_set_bool(OBJECT(&s->fmc), true, "realized", &err); if (err) { error_propagate(errp, err); diff --git a/hw/ssi/aspeed_smc.c b/hw/ssi/aspeed_smc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/aspeed_smc.c +++ b/hw/ssi/aspeed_smc.c @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_aspeed_smc = { static Property aspeed_smc_properties[] = { DEFINE_PROP_UINT32("num-cs", AspeedSMCState, num_cs, 1), + DEFINE_PROP_UINT64("sdram-base", AspeedSMCState, sdram_base, 0), DEFINE_PROP_END_OF_LIST(), }; -- 2.20.1 From: Adriana Kobylak <anoo@us.ibm.com> The Swift board is an OpenPOWER system hosting POWER processors. Add support for their BMC including the I2C devices as found on HW. Signed-off-by: Adriana Kobylak <anoo@us.ibm.com> Reviewed-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-20-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/aspeed.c | 50 +++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 50 insertions(+) diff --git a/hw/arm/aspeed.c b/hw/arm/aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed.c +++ b/hw/arm/aspeed.c @@ -XXX,XX +XXX,XX @@ struct AspeedBoardState { SCU_AST2500_HW_STRAP_ACPI_ENABLE | \ SCU_HW_STRAP_SPI_MODE(SCU_HW_STRAP_SPI_MASTER)) +/* Swift hardware value: 0xF11AD206 */ +#define SWIFT_BMC_HW_STRAP1 ( \ + AST2500_HW_STRAP1_DEFAULTS | \ + SCU_AST2500_HW_STRAP_SPI_AUTOFETCH_ENABLE | \ + SCU_AST2500_HW_STRAP_GPIO_STRAP_ENABLE | \ + SCU_AST2500_HW_STRAP_UART_DEBUG | \ + SCU_AST2500_HW_STRAP_DDR4_ENABLE | \ + SCU_H_PLL_BYPASS_EN | \ + SCU_AST2500_HW_STRAP_ACPI_ENABLE | \ + SCU_HW_STRAP_SPI_MODE(SCU_HW_STRAP_SPI_MASTER)) + /* Witherspoon hardware value: 0xF10AD216 (but use romulus definition) */ #define WITHERSPOON_BMC_HW_STRAP1 ROMULUS_BMC_HW_STRAP1 @@ -XXX,XX +XXX,XX @@ static void romulus_bmc_i2c_init(AspeedBoardState *bmc) i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 11), "ds1338", 0x32); } +static void swift_bmc_i2c_init(AspeedBoardState *bmc) +{ + AspeedSoCState *soc = &bmc->soc; + + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 3), "pca9552", 0x60); + + /* The swift board expects a TMP275 but a TMP105 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 7), "tmp105", 0x48); + /* The swift board expects a pca9551 but a pca9552 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 7), "pca9552", 0x60); + + /* The swift board expects an Epson RX8900 RTC but a ds1338 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 8), "ds1338", 0x32); + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 8), "pca9552", 0x60); + + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 9), "tmp423", 0x4c); + /* The swift board expects a pca9539 but a pca9552 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 9), "pca9552", 0x74); + + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 10), "tmp423", 0x4c); + /* The swift board expects a pca9539 but a pca9552 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 10), "pca9552", + 0x74); + + /* The swift board expects a TMP275 but a TMP105 is compatible */ + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 12), "tmp105", 0x48); + i2c_create_slave(aspeed_i2c_get_bus(DEVICE(&soc->i2c), 12), "tmp105", 0x4a); +} + static void witherspoon_bmc_i2c_init(AspeedBoardState *bmc) { AspeedSoCState *soc = &bmc->soc; @@ -XXX,XX +XXX,XX @@ static const AspeedBoardConfig aspeed_boards[] = { .num_cs = 2, .i2c_init = romulus_bmc_i2c_init, .ram = 512 * MiB, + }, { + .name = MACHINE_TYPE_NAME("swift-bmc"), + .desc = "OpenPOWER Swift BMC (ARM1176)", + .soc_name = "ast2500-a1", + .hw_strap1 = SWIFT_BMC_HW_STRAP1, + .fmc_model = "mx66l1g45g", + .spi_model = "mx66l1g45g", + .num_cs = 2, + .i2c_init = swift_bmc_i2c_init, + .ram = 512 * MiB, }, { .name = MACHINE_TYPE_NAME("witherspoon-bmc"), .desc = "OpenPOWER Witherspoon BMC (ARM1176)", -- 2.20.1 From: Eddie James <eajames@linux.ibm.com> The XDMA engine embedded in the Aspeed SOCs performs PCI DMA operations between the SOC (acting as a BMC) and a host processor in a server. The XDMA engine exists on the AST2400, AST2500, and AST2600 SOCs, so enable it for all of those. Add trace events on the important register writes in the XDMA engine. Signed-off-by: Eddie James <eajames@linux.ibm.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Cédric Le Goater <clg@kaod.org> Message-id: 20190618165311.27066-21-clg@kaod.org [clg: - changed title ] Signed-off-by: Cédric Le Goater <clg@kaod.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/Makefile.objs | 1 + include/hw/arm/aspeed_soc.h | 3 + include/hw/misc/aspeed_xdma.h | 30 +++++++ hw/arm/aspeed_soc.c | 17 ++++ hw/misc/aspeed_xdma.c | 165 ++++++++++++++++++++++++++++++++++ hw/misc/trace-events | 3 + 6 files changed, 219 insertions(+) create mode 100644 include/hw/misc/aspeed_xdma.h create mode 100644 hw/misc/aspeed_xdma.c diff --git a/hw/misc/Makefile.objs b/hw/misc/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Makefile.objs +++ b/hw/misc/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_ARMSSE_MHU) += armsse-mhu.o obj-$(CONFIG_PVPANIC) += pvpanic.o obj-$(CONFIG_AUX) += auxbus.o +obj-$(CONFIG_ASPEED_SOC) += aspeed_xdma.o obj-$(CONFIG_ASPEED_SOC) += aspeed_scu.o aspeed_sdmc.o obj-$(CONFIG_MSF2) += msf2-sysreg.o obj-$(CONFIG_NRF51_SOC) += nrf51_rng.o diff --git a/include/hw/arm/aspeed_soc.h b/include/hw/arm/aspeed_soc.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/aspeed_soc.h +++ b/include/hw/arm/aspeed_soc.h @@ -XXX,XX +XXX,XX @@ #include "hw/intc/aspeed_vic.h" #include "hw/misc/aspeed_scu.h" #include "hw/misc/aspeed_sdmc.h" +#include "hw/misc/aspeed_xdma.h" #include "hw/timer/aspeed_timer.h" #include "hw/timer/aspeed_rtc.h" #include "hw/i2c/aspeed_i2c.h" @@ -XXX,XX +XXX,XX @@ typedef struct AspeedSoCState { AspeedTimerCtrlState timerctrl; AspeedI2CState i2c; AspeedSCUState scu; + AspeedXDMAState xdma; AspeedSMCState fmc; AspeedSMCState spi[ASPEED_SPIS_NUM]; AspeedSDMCState sdmc; @@ -XXX,XX +XXX,XX @@ enum { ASPEED_ETH1, ASPEED_ETH2, ASPEED_SDRAM, + ASPEED_XDMA, }; #endif /* ASPEED_SOC_H */ diff --git a/include/hw/misc/aspeed_xdma.h b/include/hw/misc/aspeed_xdma.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/misc/aspeed_xdma.h @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED XDMA Controller + * Eddie James <eajames@linux.ibm.com> + * + * Copyright (C) 2019 IBM Corp. + * SPDX-License-Identifer: GPL-2.0-or-later + */ + +#ifndef ASPEED_XDMA_H +#define ASPEED_XDMA_H + +#include "hw/sysbus.h" + +#define TYPE_ASPEED_XDMA "aspeed.xdma" +#define ASPEED_XDMA(obj) OBJECT_CHECK(AspeedXDMAState, (obj), TYPE_ASPEED_XDMA) + +#define ASPEED_XDMA_NUM_REGS (ASPEED_XDMA_REG_SIZE / sizeof(uint32_t)) +#define ASPEED_XDMA_REG_SIZE 0x7C + +typedef struct AspeedXDMAState { + SysBusDevice parent; + + MemoryRegion iomem; + qemu_irq irq; + + char bmc_cmdq_readp_set; + uint32_t regs[ASPEED_XDMA_NUM_REGS]; +} AspeedXDMAState; + +#endif /* ASPEED_XDMA_H */ diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static const hwaddr aspeed_soc_ast2400_memmap[] = { [ASPEED_VIC] = 0x1E6C0000, [ASPEED_SDMC] = 0x1E6E0000, [ASPEED_SCU] = 0x1E6E2000, + [ASPEED_XDMA] = 0x1E6E7000, [ASPEED_ADC] = 0x1E6E9000, [ASPEED_SRAM] = 0x1E720000, [ASPEED_GPIO] = 0x1E780000, @@ -XXX,XX +XXX,XX @@ static const hwaddr aspeed_soc_ast2500_memmap[] = { [ASPEED_VIC] = 0x1E6C0000, [ASPEED_SDMC] = 0x1E6E0000, [ASPEED_SCU] = 0x1E6E2000, + [ASPEED_XDMA] = 0x1E6E7000, [ASPEED_ADC] = 0x1E6E9000, [ASPEED_SRAM] = 0x1E720000, [ASPEED_GPIO] = 0x1E780000, @@ -XXX,XX +XXX,XX @@ static const int aspeed_soc_ast2400_irqmap[] = { [ASPEED_I2C] = 12, [ASPEED_ETH1] = 2, [ASPEED_ETH2] = 3, + [ASPEED_XDMA] = 6, }; #define aspeed_soc_ast2500_irqmap aspeed_soc_ast2400_irqmap @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sysbus_init_child_obj(obj, "ftgmac100[*]", OBJECT(&s->ftgmac100[i]), sizeof(s->ftgmac100[i]), TYPE_FTGMAC100); } + + sysbus_init_child_obj(obj, "xdma", OBJECT(&s->xdma), sizeof(s->xdma), + TYPE_ASPEED_XDMA); } static void aspeed_soc_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_realize(DeviceState *dev, Error **errp) sysbus_connect_irq(SYS_BUS_DEVICE(&s->ftgmac100[i]), 0, aspeed_soc_get_irq(s, ASPEED_ETH1 + i)); } + + /* XDMA */ + object_property_set_bool(OBJECT(&s->xdma), true, "realized", &err); + if (err) { + error_propagate(errp, err); + return; + } + sysbus_mmio_map(SYS_BUS_DEVICE(&s->xdma), 0, + sc->info->memmap[ASPEED_XDMA]); + sysbus_connect_irq(SYS_BUS_DEVICE(&s->xdma), 0, + aspeed_soc_get_irq(s, ASPEED_XDMA)); } static Property aspeed_soc_properties[] = { DEFINE_PROP_UINT32("num-cpus", AspeedSoCState, num_cpus, 0), diff --git a/hw/misc/aspeed_xdma.c b/hw/misc/aspeed_xdma.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/aspeed_xdma.c @@ -XXX,XX +XXX,XX @@ +/* + * ASPEED XDMA Controller + * Eddie James <eajames@linux.ibm.com> + * + * Copyright (C) 2019 IBM Corp + * SPDX-License-Identifer: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "qemu/log.h" +#include "qemu/error-report.h" +#include "hw/misc/aspeed_xdma.h" +#include "qapi/error.h" + +#include "trace.h" + +#define XDMA_BMC_CMDQ_ADDR 0x10 +#define XDMA_BMC_CMDQ_ENDP 0x14 +#define XDMA_BMC_CMDQ_WRP 0x18 +#define XDMA_BMC_CMDQ_W_MASK 0x0003FFFF +#define XDMA_BMC_CMDQ_RDP 0x1C +#define XDMA_BMC_CMDQ_RDP_MAGIC 0xEE882266 +#define XDMA_IRQ_ENG_CTRL 0x20 +#define XDMA_IRQ_ENG_CTRL_US_COMP BIT(4) +#define XDMA_IRQ_ENG_CTRL_DS_COMP BIT(5) +#define XDMA_IRQ_ENG_CTRL_W_MASK 0xBFEFF07F +#define XDMA_IRQ_ENG_STAT 0x24 +#define XDMA_IRQ_ENG_STAT_US_COMP BIT(4) +#define XDMA_IRQ_ENG_STAT_DS_COMP BIT(5) +#define XDMA_IRQ_ENG_STAT_RESET 0xF8000000 +#define XDMA_MEM_SIZE 0x1000 + +#define TO_REG(addr) ((addr) / sizeof(uint32_t)) + +static uint64_t aspeed_xdma_read(void *opaque, hwaddr addr, unsigned int size) +{ + uint32_t val = 0; + AspeedXDMAState *xdma = opaque; + + if (addr < ASPEED_XDMA_REG_SIZE) { + val = xdma->regs[TO_REG(addr)]; + } + + return (uint64_t)val; +} + +static void aspeed_xdma_write(void *opaque, hwaddr addr, uint64_t val, + unsigned int size) +{ + unsigned int idx; + uint32_t val32 = (uint32_t)val; + AspeedXDMAState *xdma = opaque; + + if (addr >= ASPEED_XDMA_REG_SIZE) { + return; + } + + switch (addr) { + case XDMA_BMC_CMDQ_ENDP: + xdma->regs[TO_REG(addr)] = val32 & XDMA_BMC_CMDQ_W_MASK; + break; + case XDMA_BMC_CMDQ_WRP: + idx = TO_REG(addr); + xdma->regs[idx] = val32 & XDMA_BMC_CMDQ_W_MASK; + xdma->regs[TO_REG(XDMA_BMC_CMDQ_RDP)] = xdma->regs[idx]; + + trace_aspeed_xdma_write(addr, val); + + if (xdma->bmc_cmdq_readp_set) { + xdma->bmc_cmdq_readp_set = 0; + } else { + xdma->regs[TO_REG(XDMA_IRQ_ENG_STAT)] |= + XDMA_IRQ_ENG_STAT_US_COMP | XDMA_IRQ_ENG_STAT_DS_COMP; + + if (xdma->regs[TO_REG(XDMA_IRQ_ENG_CTRL)] & + (XDMA_IRQ_ENG_CTRL_US_COMP | XDMA_IRQ_ENG_CTRL_DS_COMP)) + qemu_irq_raise(xdma->irq); + } + break; + case XDMA_BMC_CMDQ_RDP: + trace_aspeed_xdma_write(addr, val); + + if (val32 == XDMA_BMC_CMDQ_RDP_MAGIC) { + xdma->bmc_cmdq_readp_set = 1; + } + break; + case XDMA_IRQ_ENG_CTRL: + xdma->regs[TO_REG(addr)] = val32 & XDMA_IRQ_ENG_CTRL_W_MASK; + break; + case XDMA_IRQ_ENG_STAT: + trace_aspeed_xdma_write(addr, val); + + idx = TO_REG(addr); + if (val32 & (XDMA_IRQ_ENG_STAT_US_COMP | XDMA_IRQ_ENG_STAT_DS_COMP)) { + xdma->regs[idx] &= + ~(XDMA_IRQ_ENG_STAT_US_COMP | XDMA_IRQ_ENG_STAT_DS_COMP); + qemu_irq_lower(xdma->irq); + } + break; + default: + xdma->regs[TO_REG(addr)] = val32; + break; + } +} + +static const MemoryRegionOps aspeed_xdma_ops = { + .read = aspeed_xdma_read, + .write = aspeed_xdma_write, + .endianness = DEVICE_NATIVE_ENDIAN, + .valid.min_access_size = 4, + .valid.max_access_size = 4, +}; + +static void aspeed_xdma_realize(DeviceState *dev, Error **errp) +{ + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); + AspeedXDMAState *xdma = ASPEED_XDMA(dev); + + sysbus_init_irq(sbd, &xdma->irq); + memory_region_init_io(&xdma->iomem, OBJECT(xdma), &aspeed_xdma_ops, xdma, + TYPE_ASPEED_XDMA, XDMA_MEM_SIZE); + sysbus_init_mmio(sbd, &xdma->iomem); +} + +static void aspeed_xdma_reset(DeviceState *dev) +{ + AspeedXDMAState *xdma = ASPEED_XDMA(dev); + + xdma->bmc_cmdq_readp_set = 0; + memset(xdma->regs, 0, ASPEED_XDMA_REG_SIZE); + xdma->regs[TO_REG(XDMA_IRQ_ENG_STAT)] = XDMA_IRQ_ENG_STAT_RESET; + + qemu_irq_lower(xdma->irq); +} + +static const VMStateDescription aspeed_xdma_vmstate = { + .name = TYPE_ASPEED_XDMA, + .version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_UINT32_ARRAY(regs, AspeedXDMAState, ASPEED_XDMA_NUM_REGS), + VMSTATE_END_OF_LIST(), + }, +}; + +static void aspeed_xdma_class_init(ObjectClass *classp, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(classp); + + dc->realize = aspeed_xdma_realize; + dc->reset = aspeed_xdma_reset; + dc->vmsd = &aspeed_xdma_vmstate; +} + +static const TypeInfo aspeed_xdma_info = { + .name = TYPE_ASPEED_XDMA, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(AspeedXDMAState), + .class_init = aspeed_xdma_class_init, +}; + +static void aspeed_xdma_register_type(void) +{ + type_register_static(&aspeed_xdma_info); +} +type_init(aspeed_xdma_register_type); diff --git a/hw/misc/trace-events b/hw/misc/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/trace-events +++ b/hw/misc/trace-events @@ -XXX,XX +XXX,XX @@ armsse_cpuid_write(uint64_t offset, uint64_t data, unsigned size) "SSE-200 CPU_I # armsse-mhu.c armsse_mhu_read(uint64_t offset, uint64_t data, unsigned size) "SSE-200 MHU read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %u" armsse_mhu_write(uint64_t offset, uint64_t data, unsigned size) "SSE-200 MHU write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %u" + +# aspeed_xdma.c +aspeed_xdma_write(uint64_t offset, uint64_t data) "XDMA write: offset 0x%" PRIx64 " data 0x%" PRIx64 -- 2.20.1 From: Andrew Jeffery <andrew@aj.id.au> The legacy interface only supported up to 32 IRQs, which became restrictive around the AST2400 generation. QEMU support for the SoCs started with the AST2400 along with an effort to reimplement and upstream drivers for Linux, so up until this point the consumers of the QEMU ASPEED support only required the 64 IRQ register interface. In an effort to support older BMC firmware, add support for the 32 IRQ interface. Signed-off-by: Andrew Jeffery <andrew@aj.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Joel Stanley <joel@jms.id.au> Message-id: 20190618165311.27066-22-clg@kaod.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/intc/aspeed_vic.c | 105 ++++++++++++++++++++++++++----------------- 1 file changed, 63 insertions(+), 42 deletions(-) diff --git a/hw/intc/aspeed_vic.c b/hw/intc/aspeed_vic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/intc/aspeed_vic.c +++ b/hw/intc/aspeed_vic.c @@ -XXX,XX +XXX,XX @@ static void aspeed_vic_set_irq(void *opaque, int irq, int level) static uint64_t aspeed_vic_read(void *opaque, hwaddr offset, unsigned size) { - uint64_t val; - const bool high = !!(offset & 0x4); - hwaddr n_offset = (offset & ~0x4); AspeedVICState *s = (AspeedVICState *)opaque; + hwaddr n_offset; + uint64_t val; + bool high; if (offset < AVIC_NEW_BASE_OFFSET) { - qemu_log_mask(LOG_UNIMP, "%s: Ignoring read from legacy registers " - "at 0x%" HWADDR_PRIx "[%u]\n", __func__, offset, size); - return 0; + high = false; + n_offset = offset; + } else { + high = !!(offset & 0x4); + n_offset = (offset & ~0x4); } - n_offset -= AVIC_NEW_BASE_OFFSET; - switch (n_offset) { - case 0x0: /* IRQ Status */ + case 0x80: /* IRQ Status */ + case 0x00: val = s->raw & ~s->select & s->enable; break; - case 0x08: /* FIQ Status */ + case 0x88: /* FIQ Status */ + case 0x04: val = s->raw & s->select & s->enable; break; - case 0x10: /* Raw Interrupt Status */ + case 0x90: /* Raw Interrupt Status */ + case 0x08: val = s->raw; break; - case 0x18: /* Interrupt Selection */ + case 0x98: /* Interrupt Selection */ + case 0x0c: val = s->select; break; - case 0x20: /* Interrupt Enable */ + case 0xa0: /* Interrupt Enable */ + case 0x10: val = s->enable; break; - case 0x30: /* Software Interrupt */ + case 0xb0: /* Software Interrupt */ + case 0x18: val = s->trigger; break; - case 0x40: /* Interrupt Sensitivity */ + case 0xc0: /* Interrupt Sensitivity */ + case 0x24: val = s->sense; break; - case 0x48: /* Interrupt Both Edge Trigger Control */ + case 0xc8: /* Interrupt Both Edge Trigger Control */ + case 0x28: val = s->dual_edge; break; - case 0x50: /* Interrupt Event */ + case 0xd0: /* Interrupt Event */ + case 0x2c: val = s->event; break; - case 0x60: /* Edge Triggered Interrupt Status */ + case 0xe0: /* Edge Triggered Interrupt Status */ val = s->raw & ~s->sense; break; /* Illegal */ - case 0x28: /* Interrupt Enable Clear */ - case 0x38: /* Software Interrupt Clear */ - case 0x58: /* Edge Triggered Interrupt Clear */ + case 0xa8: /* Interrupt Enable Clear */ + case 0xb8: /* Software Interrupt Clear */ + case 0xd8: /* Edge Triggered Interrupt Clear */ qemu_log_mask(LOG_GUEST_ERROR, "%s: Read of write-only register with offset 0x%" HWADDR_PRIx "\n", __func__, offset); @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_vic_read(void *opaque, hwaddr offset, unsigned size) } if (high) { val = extract64(val, 32, 19); + } else { + val = extract64(val, 0, 32); } trace_aspeed_vic_read(offset, size, val); return val; @@ -XXX,XX +XXX,XX @@ static uint64_t aspeed_vic_read(void *opaque, hwaddr offset, unsigned size) static void aspeed_vic_write(void *opaque, hwaddr offset, uint64_t data, unsigned size) { - const bool high = !!(offset & 0x4); - hwaddr n_offset = (offset & ~0x4); AspeedVICState *s = (AspeedVICState *)opaque; + hwaddr n_offset; + bool high; if (offset < AVIC_NEW_BASE_OFFSET) { - qemu_log_mask(LOG_UNIMP, - "%s: Ignoring write to legacy registers at 0x%" - HWADDR_PRIx "[%u] <- 0x%" PRIx64 "\n", __func__, offset, - size, data); - return; + high = false; + n_offset = offset; + } else { + high = !!(offset & 0x4); + n_offset = (offset & ~0x4); } - n_offset -= AVIC_NEW_BASE_OFFSET; trace_aspeed_vic_write(offset, size, data); /* Given we have members using separate enable/clear registers, deposit64() @@ -XXX,XX +XXX,XX @@ static void aspeed_vic_write(void *opaque, hwaddr offset, uint64_t data, } switch (n_offset) { - case 0x18: /* Interrupt Selection */ + case 0x98: /* Interrupt Selection */ + case 0x0c: /* Register has deposit64() semantics - overwrite requested 32 bits */ if (high) { s->select &= AVIC_L_MASK; @@ -XXX,XX +XXX,XX @@ static void aspeed_vic_write(void *opaque, hwaddr offset, uint64_t data, } s->select |= data; break; - case 0x20: /* Interrupt Enable */ + case 0xa0: /* Interrupt Enable */ + case 0x10: s->enable |= data; break; - case 0x28: /* Interrupt Enable Clear */ + case 0xa8: /* Interrupt Enable Clear */ + case 0x14: s->enable &= ~data; break; - case 0x30: /* Software Interrupt */ + case 0xb0: /* Software Interrupt */ + case 0x18: qemu_log_mask(LOG_UNIMP, "%s: Software interrupts unavailable. " "IRQs requested: 0x%016" PRIx64 "\n", __func__, data); break; - case 0x38: /* Software Interrupt Clear */ + case 0xb8: /* Software Interrupt Clear */ + case 0x1c: qemu_log_mask(LOG_UNIMP, "%s: Software interrupts unavailable. " "IRQs to be cleared: 0x%016" PRIx64 "\n", __func__, data); break; - case 0x50: /* Interrupt Event */ + case 0xd0: /* Interrupt Event */ /* Register has deposit64() semantics - overwrite the top four valid * IRQ bits, as only the top four IRQs (GPIOs) can change their event * type */ @@ -XXX,XX +XXX,XX @@ static void aspeed_vic_write(void *opaque, hwaddr offset, uint64_t data, "Ignoring invalid write to interrupt event register"); } break; - case 0x58: /* Edge Triggered Interrupt Clear */ + case 0xd8: /* Edge Triggered Interrupt Clear */ + case 0x38: s->raw &= ~(data & ~s->sense); break; - case 0x00: /* IRQ Status */ - case 0x08: /* FIQ Status */ - case 0x10: /* Raw Interrupt Status */ - case 0x40: /* Interrupt Sensitivity */ - case 0x48: /* Interrupt Both Edge Trigger Control */ - case 0x60: /* Edge Triggered Interrupt Status */ + case 0x80: /* IRQ Status */ + case 0x00: + case 0x88: /* FIQ Status */ + case 0x04: + case 0x90: /* Raw Interrupt Status */ + case 0x08: + case 0xc0: /* Interrupt Sensitivity */ + case 0x24: + case 0xc8: /* Interrupt Both Edge Trigger Control */ + case 0x28: + case 0xe0: /* Edge Triggered Interrupt Status */ qemu_log_mask(LOG_GUEST_ERROR, "%s: Write of read-only register with offset 0x%" HWADDR_PRIx "\n", __func__, offset); -- 2.20.1 From: Joel Stanley <joel@jms.id.au> The ast2500 uses the watchdog to reset the SDRAM controller. This operation is usually performed by u-boot's memory training procedure, and it is enabled by setting a bit in the SCU and then causing the watchdog to expire. Therefore, we need the watchdog to be able to access the SCU's register space. This causes the watchdog to not perform a system reset when the bit is set. In the future it could perform a reset of the SDMC model. Signed-off-by: Joel Stanley <joel@jms.id.au> Signed-off-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Cédric Le Goater <clg@kaod.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190621065242.32535-1-joel@jms.id.au Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/watchdog/wdt_aspeed.h | 1 + hw/arm/aspeed_soc.c | 2 ++ hw/watchdog/wdt_aspeed.c | 20 ++++++++++++++++++++ 3 files changed, 23 insertions(+) diff --git a/include/hw/watchdog/wdt_aspeed.h b/include/hw/watchdog/wdt_aspeed.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/watchdog/wdt_aspeed.h +++ b/include/hw/watchdog/wdt_aspeed.h @@ -XXX,XX +XXX,XX @@ typedef struct AspeedWDTState { MemoryRegion iomem; uint32_t regs[ASPEED_WDT_REGS_MAX]; + AspeedSCUState *scu; uint32_t pclk_freq; uint32_t silicon_rev; uint32_t ext_pulse_width_mask; diff --git a/hw/arm/aspeed_soc.c b/hw/arm/aspeed_soc.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/aspeed_soc.c +++ b/hw/arm/aspeed_soc.c @@ -XXX,XX +XXX,XX @@ static void aspeed_soc_init(Object *obj) sizeof(s->wdt[i]), TYPE_ASPEED_WDT); qdev_prop_set_uint32(DEVICE(&s->wdt[i]), "silicon-rev", sc->info->silicon_rev); + object_property_add_const_link(OBJECT(&s->wdt[i]), "scu", + OBJECT(&s->scu), &error_abort); } for (i = 0; i < ASPEED_MACS_NUM; i++) { diff --git a/hw/watchdog/wdt_aspeed.c b/hw/watchdog/wdt_aspeed.c index XXXXXXX..XXXXXXX 100644 --- a/hw/watchdog/wdt_aspeed.c +++ b/hw/watchdog/wdt_aspeed.c @@ -XXX,XX +XXX,XX @@ #define WDT_RESTART_MAGIC 0x4755 +#define SCU_RESET_CONTROL1 (0x04 / 4) +#define SCU_RESET_SDRAM BIT(0) + static bool aspeed_wdt_is_enabled(const AspeedWDTState *s) { return s->regs[WDT_CTRL] & WDT_CTRL_ENABLE; @@ -XXX,XX +XXX,XX @@ static void aspeed_wdt_timer_expired(void *dev) { AspeedWDTState *s = ASPEED_WDT(dev); + /* Do not reset on SDRAM controller reset */ + if (s->scu->regs[SCU_RESET_CONTROL1] & SCU_RESET_SDRAM) { + timer_del(s->timer); + s->regs[WDT_CTRL] = 0; + return; + } + qemu_log_mask(CPU_LOG_RESET, "Watchdog timer expired.\n"); watchdog_perform_action(); timer_del(s->timer); @@ -XXX,XX +XXX,XX @@ static void aspeed_wdt_realize(DeviceState *dev, Error **errp) { SysBusDevice *sbd = SYS_BUS_DEVICE(dev); AspeedWDTState *s = ASPEED_WDT(dev); + Error *err = NULL; + Object *obj; + + obj = object_property_get_link(OBJECT(dev), "scu", &err); + if (!obj) { + error_propagate(errp, err); + error_prepend(errp, "required link 'scu' not found: "); + return; + } + s->scu = ASPEED_SCU(obj); if (!is_supported_silicon_rev(s->silicon_rev)) { error_setg(errp, "Unknown silicon revision: 0x%" PRIx32, -- 2.20.1 From: Hongbo Zhang <hongbo.zhang@linaro.org> For AArch64, the existing "virt" machine is primarily meant to run on KVM and execute virtualization workloads, but we need an environment as faithful as possible to physical hardware, for supporting firmware and OS development for physical Aarch64 machines. This patch introduces new machine type 'sbsa-ref' with main features: - Based on 'virt' machine type. - A new memory map. - CPU type cortex-a57. - EL2 and EL3 are enabled. - GIC version 3. - System bus AHCI controller. - System bus EHCI controller. - CDROM and hard disc on AHCI bus. - E1000E ethernet card on PCIE bus. - VGA display adaptor on PCIE bus. - No virtio devices. - No fw_cfg device. - No ACPI table supplied. - Only minimal device tree nodes. Arm Trusted Firmware and UEFI porting to this are done accordingly, and the firmware should supply ACPI tables to the guest OS. The minimal device tree nodes supplied by QEMU for this platform are only to pass the dynamic info reflecting command line input to firmware, not for loading the guest OS. To make the review easier, this task is split into two patches, the fundamental skeleton part and the peripheral devices part; this patch is the first part. Signed-off-by: Hongbo Zhang <hongbo.zhang@linaro.org> Message-id: 1561890034-15921-2-git-send-email-hongbo.zhang@linaro.org [PMM: commit message tweaks; moved some bits between patch 1 and 2 to ensure patch 1 builds cleanly; removed unneeded lines from Kconfig stanza; only provide board for qemu-system-aarch64, not qemu-system-arm; added MAINTAINERS entry] Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/Makefile.objs | 1 + hw/arm/sbsa-ref.c | 271 ++++++++++++++++++++++++++++ MAINTAINERS | 8 + default-configs/aarch64-softmmu.mak | 1 + hw/arm/Kconfig | 14 ++ 5 files changed, 295 insertions(+) create mode 100644 hw/arm/sbsa-ref.c diff --git a/hw/arm/Makefile.objs b/hw/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Makefile.objs +++ b/hw/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_SPITZ) += spitz.o obj-$(CONFIG_TOSA) += tosa.o obj-$(CONFIG_Z2) += z2.o obj-$(CONFIG_REALVIEW) += realview.o +obj-$(CONFIG_SBSA_REF) += sbsa-ref.o obj-$(CONFIG_STELLARIS) += stellaris.o obj-$(CONFIG_COLLIE) += collie.o obj-$(CONFIG_VERSATILE) += versatilepb.o diff --git a/hw/arm/sbsa-ref.c b/hw/arm/sbsa-ref.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/arm/sbsa-ref.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM SBSA Reference Platform emulation + * + * Copyright (c) 2018 Linaro Limited + * Written by Hongbo Zhang <hongbo.zhang@linaro.org> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2 or later, as published by the Free Software Foundation. + * + * This program is distributed in the hope it will be useful, but WITHOUT + * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or + * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for + * more details. + * + * You should have received a copy of the GNU General Public License along with + * this program. If not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu/error-report.h" +#include "qemu/units.h" +#include "sysemu/numa.h" +#include "sysemu/sysemu.h" +#include "exec/address-spaces.h" +#include "exec/hwaddr.h" +#include "kvm_arm.h" +#include "hw/arm/boot.h" +#include "hw/boards.h" +#include "hw/intc/arm_gicv3_common.h" + +#define RAMLIMIT_GB 8192 +#define RAMLIMIT_BYTES (RAMLIMIT_GB * GiB) + +enum { + SBSA_FLASH, + SBSA_MEM, + SBSA_CPUPERIPHS, + SBSA_GIC_DIST, + SBSA_GIC_REDIST, + SBSA_SMMU, + SBSA_UART, + SBSA_RTC, + SBSA_PCIE, + SBSA_PCIE_MMIO, + SBSA_PCIE_MMIO_HIGH, + SBSA_PCIE_PIO, + SBSA_PCIE_ECAM, + SBSA_GPIO, + SBSA_SECURE_UART, + SBSA_SECURE_UART_MM, + SBSA_SECURE_MEM, + SBSA_AHCI, + SBSA_EHCI, +}; + +typedef struct MemMapEntry { + hwaddr base; + hwaddr size; +} MemMapEntry; + +typedef struct { + MachineState parent; + struct arm_boot_info bootinfo; + int smp_cpus; + void *fdt; + int fdt_size; + int psci_conduit; +} SBSAMachineState; + +#define TYPE_SBSA_MACHINE MACHINE_TYPE_NAME("sbsa-ref") +#define SBSA_MACHINE(obj) \ + OBJECT_CHECK(SBSAMachineState, (obj), TYPE_SBSA_MACHINE) + +static const MemMapEntry sbsa_ref_memmap[] = { + /* 512M boot ROM */ + [SBSA_FLASH] = { 0, 0x20000000 }, + /* 512M secure memory */ + [SBSA_SECURE_MEM] = { 0x20000000, 0x20000000 }, + /* Space reserved for CPU peripheral devices */ + [SBSA_CPUPERIPHS] = { 0x40000000, 0x00040000 }, + [SBSA_GIC_DIST] = { 0x40060000, 0x00010000 }, + [SBSA_GIC_REDIST] = { 0x40080000, 0x04000000 }, + [SBSA_UART] = { 0x60000000, 0x00001000 }, + [SBSA_RTC] = { 0x60010000, 0x00001000 }, + [SBSA_GPIO] = { 0x60020000, 0x00001000 }, + [SBSA_SECURE_UART] = { 0x60030000, 0x00001000 }, + [SBSA_SECURE_UART_MM] = { 0x60040000, 0x00001000 }, + [SBSA_SMMU] = { 0x60050000, 0x00020000 }, + /* Space here reserved for more SMMUs */ + [SBSA_AHCI] = { 0x60100000, 0x00010000 }, + [SBSA_EHCI] = { 0x60110000, 0x00010000 }, + /* Space here reserved for other devices */ + [SBSA_PCIE_PIO] = { 0x7fff0000, 0x00010000 }, + /* 32-bit address PCIE MMIO space */ + [SBSA_PCIE_MMIO] = { 0x80000000, 0x70000000 }, + /* 256M PCIE ECAM space */ + [SBSA_PCIE_ECAM] = { 0xf0000000, 0x10000000 }, + /* ~1TB PCIE MMIO space (4GB to 1024GB boundary) */ + [SBSA_PCIE_MMIO_HIGH] = { 0x100000000ULL, 0xFF00000000ULL }, + [SBSA_MEM] = { 0x10000000000ULL, RAMLIMIT_BYTES }, +}; + +static void sbsa_ref_init(MachineState *machine) +{ + SBSAMachineState *sms = SBSA_MACHINE(machine); + MachineClass *mc = MACHINE_GET_CLASS(machine); + MemoryRegion *sysmem = get_system_memory(); + MemoryRegion *secure_sysmem = NULL; + MemoryRegion *ram = g_new(MemoryRegion, 1); + const CPUArchIdList *possible_cpus; + int n, sbsa_max_cpus; + + if (strcmp(machine->cpu_type, ARM_CPU_TYPE_NAME("cortex-a57"))) { + error_report("sbsa-ref: CPU type other than the built-in " + "cortex-a57 not supported"); + exit(1); + } + + if (kvm_enabled()) { + error_report("sbsa-ref: KVM is not supported for this machine"); + exit(1); + } + + /* + * This machine has EL3 enabled, external firmware should supply PSCI + * implementation, so the QEMU's internal PSCI is disabled. + */ + sms->psci_conduit = QEMU_PSCI_CONDUIT_DISABLED; + + sbsa_max_cpus = sbsa_ref_memmap[SBSA_GIC_REDIST].size / GICV3_REDIST_SIZE; + + if (max_cpus > sbsa_max_cpus) { + error_report("Number of SMP CPUs requested (%d) exceeds max CPUs " + "supported by machine 'sbsa-ref' (%d)", + max_cpus, sbsa_max_cpus); + exit(1); + } + + sms->smp_cpus = smp_cpus; + + if (machine->ram_size > sbsa_ref_memmap[SBSA_MEM].size) { + error_report("sbsa-ref: cannot model more than %dGB RAM", RAMLIMIT_GB); + exit(1); + } + + possible_cpus = mc->possible_cpu_arch_ids(machine); + for (n = 0; n < possible_cpus->len; n++) { + Object *cpuobj; + CPUState *cs; + + if (n >= smp_cpus) { + break; + } + + cpuobj = object_new(possible_cpus->cpus[n].type); + object_property_set_int(cpuobj, possible_cpus->cpus[n].arch_id, + "mp-affinity", NULL); + + cs = CPU(cpuobj); + cs->cpu_index = n; + + numa_cpu_pre_plug(&possible_cpus->cpus[cs->cpu_index], DEVICE(cpuobj), + &error_fatal); + + if (object_property_find(cpuobj, "reset-cbar", NULL)) { + object_property_set_int(cpuobj, + sbsa_ref_memmap[SBSA_CPUPERIPHS].base, + "reset-cbar", &error_abort); + } + + object_property_set_link(cpuobj, OBJECT(sysmem), "memory", + &error_abort); + + object_property_set_link(cpuobj, OBJECT(secure_sysmem), + "secure-memory", &error_abort); + + object_property_set_bool(cpuobj, true, "realized", &error_fatal); + object_unref(cpuobj); + } + + memory_region_allocate_system_memory(ram, NULL, "sbsa-ref.ram", + machine->ram_size); + memory_region_add_subregion(sysmem, sbsa_ref_memmap[SBSA_MEM].base, ram); + + sms->bootinfo.ram_size = machine->ram_size; + sms->bootinfo.kernel_filename = machine->kernel_filename; + sms->bootinfo.nb_cpus = smp_cpus; + sms->bootinfo.board_id = -1; + sms->bootinfo.loader_start = sbsa_ref_memmap[SBSA_MEM].base; + arm_load_kernel(ARM_CPU(first_cpu), &sms->bootinfo); +} + +static uint64_t sbsa_ref_cpu_mp_affinity(SBSAMachineState *sms, int idx) +{ + uint8_t clustersz = ARM_DEFAULT_CPUS_PER_CLUSTER; + return arm_cpu_mp_affinity(idx, clustersz); +} + +static const CPUArchIdList *sbsa_ref_possible_cpu_arch_ids(MachineState *ms) +{ + SBSAMachineState *sms = SBSA_MACHINE(ms); + int n; + + if (ms->possible_cpus) { + assert(ms->possible_cpus->len == max_cpus); + return ms->possible_cpus; + } + + ms->possible_cpus = g_malloc0(sizeof(CPUArchIdList) + + sizeof(CPUArchId) * max_cpus); + ms->possible_cpus->len = max_cpus; + for (n = 0; n < ms->possible_cpus->len; n++) { + ms->possible_cpus->cpus[n].type = ms->cpu_type; + ms->possible_cpus->cpus[n].arch_id = + sbsa_ref_cpu_mp_affinity(sms, n); + ms->possible_cpus->cpus[n].props.has_thread_id = true; + ms->possible_cpus->cpus[n].props.thread_id = n; + } + return ms->possible_cpus; +} + +static CpuInstanceProperties +sbsa_ref_cpu_index_to_props(MachineState *ms, unsigned cpu_index) +{ + MachineClass *mc = MACHINE_GET_CLASS(ms); + const CPUArchIdList *possible_cpus = mc->possible_cpu_arch_ids(ms); + + assert(cpu_index < possible_cpus->len); + return possible_cpus->cpus[cpu_index].props; +} + +static int64_t +sbsa_ref_get_default_cpu_node_id(const MachineState *ms, int idx) +{ + return idx % nb_numa_nodes; +} + +static void sbsa_ref_class_init(ObjectClass *oc, void *data) +{ + MachineClass *mc = MACHINE_CLASS(oc); + + mc->init = sbsa_ref_init; + mc->desc = "QEMU 'SBSA Reference' ARM Virtual Machine"; + mc->default_cpu_type = ARM_CPU_TYPE_NAME("cortex-a57"); + mc->max_cpus = 512; + mc->pci_allow_0_address = true; + mc->minimum_page_bits = 12; + mc->block_default_type = IF_IDE; + mc->no_cdrom = 1; + mc->default_ram_size = 1 * GiB; + mc->default_cpus = 4; + mc->possible_cpu_arch_ids = sbsa_ref_possible_cpu_arch_ids; + mc->cpu_index_to_instance_props = sbsa_ref_cpu_index_to_props; + mc->get_default_cpu_node_id = sbsa_ref_get_default_cpu_node_id; +} + +static const TypeInfo sbsa_ref_info = { + .name = TYPE_SBSA_MACHINE, + .parent = TYPE_MACHINE, + .class_init = sbsa_ref_class_init, + .instance_size = sizeof(SBSAMachineState), +}; + +static void sbsa_ref_machine_init(void) +{ + type_register_static(&sbsa_ref_info); +} + +type_init(sbsa_ref_machine_init); diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/hw/arm/fsl-imx6.h F: include/hw/misc/imx6_*.h F: include/hw/ssi/imx_spi.h +SBSA-REF +M: Radoslaw Biernacki <radoslaw.biernacki@linaro.org> +M: Peter Maydell <peter.maydell@linaro.org> +R: Leif Lindholm <leif.lindholm@linaro.org> +L: qemu-arm@nongnu.org +S: Maintained +F: hw/arm/sbsa-ref.c + Sharp SL-5500 (Collie) PDA M: Peter Maydell <peter.maydell@linaro.org> L: qemu-arm@nongnu.org diff --git a/default-configs/aarch64-softmmu.mak b/default-configs/aarch64-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/default-configs/aarch64-softmmu.mak +++ b/default-configs/aarch64-softmmu.mak @@ -XXX,XX +XXX,XX @@ include arm-softmmu.mak CONFIG_XLNX_ZYNQMP_ARM=y CONFIG_XLNX_VERSAL=y +CONFIG_SBSA_REF=y diff --git a/hw/arm/Kconfig b/hw/arm/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/Kconfig +++ b/hw/arm/Kconfig @@ -XXX,XX +XXX,XX @@ config REALVIEW select DS1338 # I2C RTC+NVRAM select USB_OHCI +config SBSA_REF + bool + imply PCI_DEVICES + select AHCI + select ARM_SMMUV3 + select GPIO_KEY + select PCI_EXPRESS + select PCI_EXPRESS_GENERIC_BRIDGE + select PFLASH_CFI01 + select PL011 # UART + select PL031 # RTC + select PL061 # GPIO + select USB_EHCI_SYSBUS + config SABRELITE bool select FSL_IMX6 -- 2.20.1 From: Hongbo Zhang <hongbo.zhang@linaro.org> Following the previous patch, this patch adds peripheral devices to the newly introduced SBSA-ref machine. Signed-off-by: Hongbo Zhang <hongbo.zhang@linaro.org> Message-id: 1561890034-15921-3-git-send-email-hongbo.zhang@linaro.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/sbsa-ref.c | 535 ++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 535 insertions(+) diff --git a/hw/arm/sbsa-ref.c b/hw/arm/sbsa-ref.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/sbsa-ref.c +++ b/hw/arm/sbsa-ref.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu-common.h" #include "qapi/error.h" #include "qemu/error-report.h" #include "qemu/units.h" +#include "sysemu/device_tree.h" #include "sysemu/numa.h" #include "sysemu/sysemu.h" #include "exec/address-spaces.h" #include "exec/hwaddr.h" #include "kvm_arm.h" #include "hw/arm/boot.h" +#include "hw/block/flash.h" #include "hw/boards.h" +#include "hw/ide/internal.h" +#include "hw/ide/ahci_internal.h" #include "hw/intc/arm_gicv3_common.h" +#include "hw/loader.h" +#include "hw/pci-host/gpex.h" +#include "hw/usb.h" +#include "net/net.h" #define RAMLIMIT_GB 8192 #define RAMLIMIT_BYTES (RAMLIMIT_GB * GiB) +#define NUM_IRQS 256 +#define NUM_SMMU_IRQS 4 +#define NUM_SATA_PORTS 6 + +#define VIRTUAL_PMU_IRQ 7 +#define ARCH_GIC_MAINT_IRQ 9 +#define ARCH_TIMER_VIRT_IRQ 11 +#define ARCH_TIMER_S_EL1_IRQ 13 +#define ARCH_TIMER_NS_EL1_IRQ 14 +#define ARCH_TIMER_NS_EL2_IRQ 10 + enum { SBSA_FLASH, SBSA_MEM, @@ -XXX,XX +XXX,XX @@ typedef struct { void *fdt; int fdt_size; int psci_conduit; + PFlashCFI01 *flash[2]; } SBSAMachineState; #define TYPE_SBSA_MACHINE MACHINE_TYPE_NAME("sbsa-ref") @@ -XXX,XX +XXX,XX @@ static const MemMapEntry sbsa_ref_memmap[] = { [SBSA_MEM] = { 0x10000000000ULL, RAMLIMIT_BYTES }, }; +static const int sbsa_ref_irqmap[] = { + [SBSA_UART] = 1, + [SBSA_RTC] = 2, + [SBSA_PCIE] = 3, /* ... to 6 */ + [SBSA_GPIO] = 7, + [SBSA_SECURE_UART] = 8, + [SBSA_SECURE_UART_MM] = 9, + [SBSA_AHCI] = 10, + [SBSA_EHCI] = 11, +}; + +/* + * Firmware on this machine only uses ACPI table to load OS, these limited + * device tree nodes are just to let firmware know the info which varies from + * command line parameters, so it is not necessary to be fully compatible + * with the kernel CPU and NUMA binding rules. + */ +static void create_fdt(SBSAMachineState *sms) +{ + void *fdt = create_device_tree(&sms->fdt_size); + const MachineState *ms = MACHINE(sms); + int cpu; + + if (!fdt) { + error_report("create_device_tree() failed"); + exit(1); + } + + sms->fdt = fdt; + + qemu_fdt_setprop_string(fdt, "/", "compatible", "linux,sbsa-ref"); + qemu_fdt_setprop_cell(fdt, "/", "#address-cells", 0x2); + qemu_fdt_setprop_cell(fdt, "/", "#size-cells", 0x2); + + if (have_numa_distance) { + int size = nb_numa_nodes * nb_numa_nodes * 3 * sizeof(uint32_t); + uint32_t *matrix = g_malloc0(size); + int idx, i, j; + + for (i = 0; i < nb_numa_nodes; i++) { + for (j = 0; j < nb_numa_nodes; j++) { + idx = (i * nb_numa_nodes + j) * 3; + matrix[idx + 0] = cpu_to_be32(i); + matrix[idx + 1] = cpu_to_be32(j); + matrix[idx + 2] = cpu_to_be32(numa_info[i].distance[j]); + } + } + + qemu_fdt_add_subnode(fdt, "/distance-map"); + qemu_fdt_setprop(fdt, "/distance-map", "distance-matrix", + matrix, size); + g_free(matrix); + } + + qemu_fdt_add_subnode(sms->fdt, "/cpus"); + + for (cpu = sms->smp_cpus - 1; cpu >= 0; cpu--) { + char *nodename = g_strdup_printf("/cpus/cpu@%d", cpu); + ARMCPU *armcpu = ARM_CPU(qemu_get_cpu(cpu)); + CPUState *cs = CPU(armcpu); + + qemu_fdt_add_subnode(sms->fdt, nodename); + + if (ms->possible_cpus->cpus[cs->cpu_index].props.has_node_id) { + qemu_fdt_setprop_cell(sms->fdt, nodename, "numa-node-id", + ms->possible_cpus->cpus[cs->cpu_index].props.node_id); + } + + g_free(nodename); + } +} + +#define SBSA_FLASH_SECTOR_SIZE (256 * KiB) + +static PFlashCFI01 *sbsa_flash_create1(SBSAMachineState *sms, + const char *name, + const char *alias_prop_name) +{ + /* + * Create a single flash device. We use the same parameters as + * the flash devices on the Versatile Express board. + */ + DeviceState *dev = qdev_create(NULL, TYPE_PFLASH_CFI01); + + qdev_prop_set_uint64(dev, "sector-length", SBSA_FLASH_SECTOR_SIZE); + qdev_prop_set_uint8(dev, "width", 4); + qdev_prop_set_uint8(dev, "device-width", 2); + qdev_prop_set_bit(dev, "big-endian", false); + qdev_prop_set_uint16(dev, "id0", 0x89); + qdev_prop_set_uint16(dev, "id1", 0x18); + qdev_prop_set_uint16(dev, "id2", 0x00); + qdev_prop_set_uint16(dev, "id3", 0x00); + qdev_prop_set_string(dev, "name", name); + object_property_add_child(OBJECT(sms), name, OBJECT(dev), + &error_abort); + object_property_add_alias(OBJECT(sms), alias_prop_name, + OBJECT(dev), "drive", &error_abort); + return PFLASH_CFI01(dev); +} + +static void sbsa_flash_create(SBSAMachineState *sms) +{ + sms->flash[0] = sbsa_flash_create1(sms, "sbsa.flash0", "pflash0"); + sms->flash[1] = sbsa_flash_create1(sms, "sbsa.flash1", "pflash1"); +} + +static void sbsa_flash_map1(PFlashCFI01 *flash, + hwaddr base, hwaddr size, + MemoryRegion *sysmem) +{ + DeviceState *dev = DEVICE(flash); + + assert(size % SBSA_FLASH_SECTOR_SIZE == 0); + assert(size / SBSA_FLASH_SECTOR_SIZE <= UINT32_MAX); + qdev_prop_set_uint32(dev, "num-blocks", size / SBSA_FLASH_SECTOR_SIZE); + qdev_init_nofail(dev); + + memory_region_add_subregion(sysmem, base, + sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), + 0)); +} + +static void sbsa_flash_map(SBSAMachineState *sms, + MemoryRegion *sysmem, + MemoryRegion *secure_sysmem) +{ + /* + * Map two flash devices to fill the SBSA_FLASH space in the memmap. + * sysmem is the system memory space. secure_sysmem is the secure view + * of the system, and the first flash device should be made visible only + * there. The second flash device is visible to both secure and nonsecure. + * If sysmem == secure_sysmem this means there is no separate Secure + * address space and both flash devices are generally visible. + */ + hwaddr flashsize = sbsa_ref_memmap[SBSA_FLASH].size / 2; + hwaddr flashbase = sbsa_ref_memmap[SBSA_FLASH].base; + + sbsa_flash_map1(sms->flash[0], flashbase, flashsize, + secure_sysmem); + sbsa_flash_map1(sms->flash[1], flashbase + flashsize, flashsize, + sysmem); +} + +static bool sbsa_firmware_init(SBSAMachineState *sms, + MemoryRegion *sysmem, + MemoryRegion *secure_sysmem) +{ + int i; + BlockBackend *pflash_blk0; + + /* Map legacy -drive if=pflash to machine properties */ + for (i = 0; i < ARRAY_SIZE(sms->flash); i++) { + pflash_cfi01_legacy_drive(sms->flash[i], + drive_get(IF_PFLASH, 0, i)); + } + + sbsa_flash_map(sms, sysmem, secure_sysmem); + + pflash_blk0 = pflash_cfi01_get_blk(sms->flash[0]); + + if (bios_name) { + char *fname; + MemoryRegion *mr; + int image_size; + + if (pflash_blk0) { + error_report("The contents of the first flash device may be " + "specified with -bios or with -drive if=pflash... " + "but you cannot use both options at once"); + exit(1); + } + + /* Fall back to -bios */ + + fname = qemu_find_file(QEMU_FILE_TYPE_BIOS, bios_name); + if (!fname) { + error_report("Could not find ROM image '%s'", bios_name); + exit(1); + } + mr = sysbus_mmio_get_region(SYS_BUS_DEVICE(sms->flash[0]), 0); + image_size = load_image_mr(fname, mr); + g_free(fname); + if (image_size < 0) { + error_report("Could not load ROM image '%s'", bios_name); + exit(1); + } + } + + return pflash_blk0 || bios_name; +} + +static void create_secure_ram(SBSAMachineState *sms, + MemoryRegion *secure_sysmem) +{ + MemoryRegion *secram = g_new(MemoryRegion, 1); + hwaddr base = sbsa_ref_memmap[SBSA_SECURE_MEM].base; + hwaddr size = sbsa_ref_memmap[SBSA_SECURE_MEM].size; + + memory_region_init_ram(secram, NULL, "sbsa-ref.secure-ram", size, + &error_fatal); + memory_region_add_subregion(secure_sysmem, base, secram); +} + +static void create_gic(SBSAMachineState *sms, qemu_irq *pic) +{ + DeviceState *gicdev; + SysBusDevice *gicbusdev; + const char *gictype; + uint32_t redist0_capacity, redist0_count; + int i; + + gictype = gicv3_class_name(); + + gicdev = qdev_create(NULL, gictype); + qdev_prop_set_uint32(gicdev, "revision", 3); + qdev_prop_set_uint32(gicdev, "num-cpu", smp_cpus); + /* + * Note that the num-irq property counts both internal and external + * interrupts; there are always 32 of the former (mandated by GIC spec). + */ + qdev_prop_set_uint32(gicdev, "num-irq", NUM_IRQS + 32); + qdev_prop_set_bit(gicdev, "has-security-extensions", true); + + redist0_capacity = + sbsa_ref_memmap[SBSA_GIC_REDIST].size / GICV3_REDIST_SIZE; + redist0_count = MIN(smp_cpus, redist0_capacity); + + qdev_prop_set_uint32(gicdev, "len-redist-region-count", 1); + qdev_prop_set_uint32(gicdev, "redist-region-count[0]", redist0_count); + + qdev_init_nofail(gicdev); + gicbusdev = SYS_BUS_DEVICE(gicdev); + sysbus_mmio_map(gicbusdev, 0, sbsa_ref_memmap[SBSA_GIC_DIST].base); + sysbus_mmio_map(gicbusdev, 1, sbsa_ref_memmap[SBSA_GIC_REDIST].base); + + /* + * Wire the outputs from each CPU's generic timer and the GICv3 + * maintenance interrupt signal to the appropriate GIC PPI inputs, + * and the GIC's IRQ/FIQ/VIRQ/VFIQ interrupt outputs to the CPU's inputs. + */ + for (i = 0; i < smp_cpus; i++) { + DeviceState *cpudev = DEVICE(qemu_get_cpu(i)); + int ppibase = NUM_IRQS + i * GIC_INTERNAL + GIC_NR_SGIS; + int irq; + /* + * Mapping from the output timer irq lines from the CPU to the + * GIC PPI inputs used for this board. + */ + const int timer_irq[] = { + [GTIMER_PHYS] = ARCH_TIMER_NS_EL1_IRQ, + [GTIMER_VIRT] = ARCH_TIMER_VIRT_IRQ, + [GTIMER_HYP] = ARCH_TIMER_NS_EL2_IRQ, + [GTIMER_SEC] = ARCH_TIMER_S_EL1_IRQ, + }; + + for (irq = 0; irq < ARRAY_SIZE(timer_irq); irq++) { + qdev_connect_gpio_out(cpudev, irq, + qdev_get_gpio_in(gicdev, + ppibase + timer_irq[irq])); + } + + qdev_connect_gpio_out_named(cpudev, "gicv3-maintenance-interrupt", 0, + qdev_get_gpio_in(gicdev, ppibase + + ARCH_GIC_MAINT_IRQ)); + qdev_connect_gpio_out_named(cpudev, "pmu-interrupt", 0, + qdev_get_gpio_in(gicdev, ppibase + + VIRTUAL_PMU_IRQ)); + + sysbus_connect_irq(gicbusdev, i, qdev_get_gpio_in(cpudev, ARM_CPU_IRQ)); + sysbus_connect_irq(gicbusdev, i + smp_cpus, + qdev_get_gpio_in(cpudev, ARM_CPU_FIQ)); + sysbus_connect_irq(gicbusdev, i + 2 * smp_cpus, + qdev_get_gpio_in(cpudev, ARM_CPU_VIRQ)); + sysbus_connect_irq(gicbusdev, i + 3 * smp_cpus, + qdev_get_gpio_in(cpudev, ARM_CPU_VFIQ)); + } + + for (i = 0; i < NUM_IRQS; i++) { + pic[i] = qdev_get_gpio_in(gicdev, i); + } +} + +static void create_uart(const SBSAMachineState *sms, qemu_irq *pic, int uart, + MemoryRegion *mem, Chardev *chr) +{ + hwaddr base = sbsa_ref_memmap[uart].base; + int irq = sbsa_ref_irqmap[uart]; + DeviceState *dev = qdev_create(NULL, "pl011"); + SysBusDevice *s = SYS_BUS_DEVICE(dev); + + qdev_prop_set_chr(dev, "chardev", chr); + qdev_init_nofail(dev); + memory_region_add_subregion(mem, base, + sysbus_mmio_get_region(s, 0)); + sysbus_connect_irq(s, 0, pic[irq]); +} + +static void create_rtc(const SBSAMachineState *sms, qemu_irq *pic) +{ + hwaddr base = sbsa_ref_memmap[SBSA_RTC].base; + int irq = sbsa_ref_irqmap[SBSA_RTC]; + + sysbus_create_simple("pl031", base, pic[irq]); +} + +static DeviceState *gpio_key_dev; +static void sbsa_ref_powerdown_req(Notifier *n, void *opaque) +{ + /* use gpio Pin 3 for power button event */ + qemu_set_irq(qdev_get_gpio_in(gpio_key_dev, 0), 1); +} + +static Notifier sbsa_ref_powerdown_notifier = { + .notify = sbsa_ref_powerdown_req +}; + +static void create_gpio(const SBSAMachineState *sms, qemu_irq *pic) +{ + DeviceState *pl061_dev; + hwaddr base = sbsa_ref_memmap[SBSA_GPIO].base; + int irq = sbsa_ref_irqmap[SBSA_GPIO]; + + pl061_dev = sysbus_create_simple("pl061", base, pic[irq]); + + gpio_key_dev = sysbus_create_simple("gpio-key", -1, + qdev_get_gpio_in(pl061_dev, 3)); + + /* connect powerdown request */ + qemu_register_powerdown_notifier(&sbsa_ref_powerdown_notifier); +} + +static void create_ahci(const SBSAMachineState *sms, qemu_irq *pic) +{ + hwaddr base = sbsa_ref_memmap[SBSA_AHCI].base; + int irq = sbsa_ref_irqmap[SBSA_AHCI]; + DeviceState *dev; + DriveInfo *hd[NUM_SATA_PORTS]; + SysbusAHCIState *sysahci; + AHCIState *ahci; + int i; + + dev = qdev_create(NULL, "sysbus-ahci"); + qdev_prop_set_uint32(dev, "num-ports", NUM_SATA_PORTS); + qdev_init_nofail(dev); + sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, base); + sysbus_connect_irq(SYS_BUS_DEVICE(dev), 0, pic[irq]); + + sysahci = SYSBUS_AHCI(dev); + ahci = &sysahci->ahci; + ide_drive_get(hd, ARRAY_SIZE(hd)); + for (i = 0; i < ahci->ports; i++) { + if (hd[i] == NULL) { + continue; + } + ide_create_drive(&ahci->dev[i].port, 0, hd[i]); + } +} + +static void create_ehci(const SBSAMachineState *sms, qemu_irq *pic) +{ + hwaddr base = sbsa_ref_memmap[SBSA_EHCI].base; + int irq = sbsa_ref_irqmap[SBSA_EHCI]; + + sysbus_create_simple("platform-ehci-usb", base, pic[irq]); +} + +static void create_smmu(const SBSAMachineState *sms, qemu_irq *pic, + PCIBus *bus) +{ + hwaddr base = sbsa_ref_memmap[SBSA_SMMU].base; + int irq = sbsa_ref_irqmap[SBSA_SMMU]; + DeviceState *dev; + int i; + + dev = qdev_create(NULL, "arm-smmuv3"); + + object_property_set_link(OBJECT(dev), OBJECT(bus), "primary-bus", + &error_abort); + qdev_init_nofail(dev); + sysbus_mmio_map(SYS_BUS_DEVICE(dev), 0, base); + for (i = 0; i < NUM_SMMU_IRQS; i++) { + sysbus_connect_irq(SYS_BUS_DEVICE(dev), i, pic[irq + i]); + } +} + +static void create_pcie(SBSAMachineState *sms, qemu_irq *pic) +{ + hwaddr base_ecam = sbsa_ref_memmap[SBSA_PCIE_ECAM].base; + hwaddr size_ecam = sbsa_ref_memmap[SBSA_PCIE_ECAM].size; + hwaddr base_mmio = sbsa_ref_memmap[SBSA_PCIE_MMIO].base; + hwaddr size_mmio = sbsa_ref_memmap[SBSA_PCIE_MMIO].size; + hwaddr base_mmio_high = sbsa_ref_memmap[SBSA_PCIE_MMIO_HIGH].base; + hwaddr size_mmio_high = sbsa_ref_memmap[SBSA_PCIE_MMIO_HIGH].size; + hwaddr base_pio = sbsa_ref_memmap[SBSA_PCIE_PIO].base; + int irq = sbsa_ref_irqmap[SBSA_PCIE]; + MemoryRegion *mmio_alias, *mmio_alias_high, *mmio_reg; + MemoryRegion *ecam_alias, *ecam_reg; + DeviceState *dev; + PCIHostState *pci; + int i; + + dev = qdev_create(NULL, TYPE_GPEX_HOST); + qdev_init_nofail(dev); + + /* Map ECAM space */ + ecam_alias = g_new0(MemoryRegion, 1); + ecam_reg = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 0); + memory_region_init_alias(ecam_alias, OBJECT(dev), "pcie-ecam", + ecam_reg, 0, size_ecam); + memory_region_add_subregion(get_system_memory(), base_ecam, ecam_alias); + + /* Map the MMIO space */ + mmio_alias = g_new0(MemoryRegion, 1); + mmio_reg = sysbus_mmio_get_region(SYS_BUS_DEVICE(dev), 1); + memory_region_init_alias(mmio_alias, OBJECT(dev), "pcie-mmio", + mmio_reg, base_mmio, size_mmio); + memory_region_add_subregion(get_system_memory(), base_mmio, mmio_alias); + + /* Map the MMIO_HIGH space */ + mmio_alias_high = g_new0(MemoryRegion, 1); + memory_region_init_alias(mmio_alias_high, OBJECT(dev), "pcie-mmio-high", + mmio_reg, base_mmio_high, size_mmio_high); + memory_region_add_subregion(get_system_memory(), base_mmio_high, + mmio_alias_high); + + /* Map IO port space */ + sysbus_mmio_map(SYS_BUS_DEVICE(dev), 2, base_pio); + + for (i = 0; i < GPEX_NUM_IRQS; i++) { + sysbus_connect_irq(SYS_BUS_DEVICE(dev), i, pic[irq + i]); + gpex_set_irq_num(GPEX_HOST(dev), i, irq + i); + } + + pci = PCI_HOST_BRIDGE(dev); + if (pci->bus) { + for (i = 0; i < nb_nics; i++) { + NICInfo *nd = &nd_table[i]; + + if (!nd->model) { + nd->model = g_strdup("e1000e"); + } + + pci_nic_init_nofail(nd, pci->bus, nd->model, NULL); + } + } + + pci_create_simple(pci->bus, -1, "VGA"); + + create_smmu(sms, pic, pci->bus); +} + +static void *sbsa_ref_dtb(const struct arm_boot_info *binfo, int *fdt_size) +{ + const SBSAMachineState *board = container_of(binfo, SBSAMachineState, + bootinfo); + + *fdt_size = board->fdt_size; + return board->fdt; +} + static void sbsa_ref_init(MachineState *machine) { SBSAMachineState *sms = SBSA_MACHINE(machine); @@ -XXX,XX +XXX,XX @@ static void sbsa_ref_init(MachineState *machine) MemoryRegion *sysmem = get_system_memory(); MemoryRegion *secure_sysmem = NULL; MemoryRegion *ram = g_new(MemoryRegion, 1); + bool firmware_loaded; const CPUArchIdList *possible_cpus; int n, sbsa_max_cpus; + qemu_irq pic[NUM_IRQS]; if (strcmp(machine->cpu_type, ARM_CPU_TYPE_NAME("cortex-a57"))) { error_report("sbsa-ref: CPU type other than the built-in " @@ -XXX,XX +XXX,XX @@ static void sbsa_ref_init(MachineState *machine) exit(1); } + /* + * The Secure view of the world is the same as the NonSecure, + * but with a few extra devices. Create it as a container region + * containing the system memory at low priority; any secure-only + * devices go in at higher priority and take precedence. + */ + secure_sysmem = g_new(MemoryRegion, 1); + memory_region_init(secure_sysmem, OBJECT(machine), "secure-memory", + UINT64_MAX); + memory_region_add_subregion_overlap(secure_sysmem, 0, sysmem, -1); + + firmware_loaded = sbsa_firmware_init(sms, sysmem, + secure_sysmem ?: sysmem); + + if (machine->kernel_filename && firmware_loaded) { + error_report("sbsa-ref: No fw_cfg device on this machine, " + "so -kernel option is not supported when firmware loaded, " + "please load OS from hard disk instead"); + exit(1); + } + /* * This machine has EL3 enabled, external firmware should supply PSCI * implementation, so the QEMU's internal PSCI is disabled. @@ -XXX,XX +XXX,XX @@ static void sbsa_ref_init(MachineState *machine) machine->ram_size); memory_region_add_subregion(sysmem, sbsa_ref_memmap[SBSA_MEM].base, ram); + create_fdt(sms); + + create_secure_ram(sms, secure_sysmem); + + create_gic(sms, pic); + + create_uart(sms, pic, SBSA_UART, sysmem, serial_hd(0)); + create_uart(sms, pic, SBSA_SECURE_UART, secure_sysmem, serial_hd(1)); + /* Second secure UART for RAS and MM from EL0 */ + create_uart(sms, pic, SBSA_SECURE_UART_MM, secure_sysmem, serial_hd(2)); + + create_rtc(sms, pic); + + create_gpio(sms, pic); + + create_ahci(sms, pic); + + create_ehci(sms, pic); + + create_pcie(sms, pic); + sms->bootinfo.ram_size = machine->ram_size; sms->bootinfo.kernel_filename = machine->kernel_filename; sms->bootinfo.nb_cpus = smp_cpus; sms->bootinfo.board_id = -1; sms->bootinfo.loader_start = sbsa_ref_memmap[SBSA_MEM].base; + sms->bootinfo.get_dtb = sbsa_ref_dtb; + sms->bootinfo.firmware_loaded = firmware_loaded; arm_load_kernel(ARM_CPU(first_cpu), &sms->bootinfo); } @@ -XXX,XX +XXX,XX @@ sbsa_ref_get_default_cpu_node_id(const MachineState *ms, int idx) return idx % nb_numa_nodes; } +static void sbsa_ref_instance_init(Object *obj) +{ + SBSAMachineState *sms = SBSA_MACHINE(obj); + + sbsa_flash_create(sms); +} + static void sbsa_ref_class_init(ObjectClass *oc, void *data) { MachineClass *mc = MACHINE_CLASS(oc); @@ -XXX,XX +XXX,XX @@ static void sbsa_ref_class_init(ObjectClass *oc, void *data) static const TypeInfo sbsa_ref_info = { .name = TYPE_SBSA_MACHINE, .parent = TYPE_MACHINE, + .instance_init = sbsa_ref_instance_init, .class_init = sbsa_ref_class_init, .instance_size = sizeof(SBSAMachineState), }; -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Group Aarch64 rules together, TCG related ones at the bottom. This will help when restricting TCG-only objects. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-2-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/Makefile.objs | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(call lnot,$(CONFIG_KVM)) += kvm-stub.o obj-y += translate.o op_helper.o helper.o cpu.o obj-y += neon_helper.o iwmmxt_helper.o vec_helper.o vfp_helper.o obj-y += gdbstub.o -obj-$(TARGET_AARCH64) += cpu64.o translate-a64.o helper-a64.o gdbstub64.o -obj-$(TARGET_AARCH64) += pauth_helper.o +obj-$(TARGET_AARCH64) += cpu64.o gdbstub64.o obj-y += crypto_helper.o obj-$(CONFIG_SOFTMMU) += arm-powerctl.o @@ -XXX,XX +XXX,XX @@ target/arm/translate-sve.o: target/arm/decode-sve.inc.c target/arm/translate.o: target/arm/decode-vfp.inc.c target/arm/translate.o: target/arm/decode-vfp-uncond.inc.c +obj-$(TARGET_AARCH64) += translate-a64.o helper-a64.o obj-$(TARGET_AARCH64) += translate-sve.o sve_helper.o +obj-$(TARGET_AARCH64) += pauth_helper.o -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Group ARM objects together, TCG related ones at the bottom. This will help when restricting TCG-only objects. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-3-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/Makefile.objs | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-$(CONFIG_KVM) += kvm.o obj-$(call land,$(CONFIG_KVM),$(call lnot,$(TARGET_AARCH64))) += kvm32.o obj-$(call land,$(CONFIG_KVM),$(TARGET_AARCH64)) += kvm64.o obj-$(call lnot,$(CONFIG_KVM)) += kvm-stub.o -obj-y += translate.o op_helper.o helper.o cpu.o -obj-y += neon_helper.o iwmmxt_helper.o vec_helper.o vfp_helper.o -obj-y += gdbstub.o +obj-y += helper.o vfp_helper.o +obj-y += cpu.o gdbstub.o obj-$(TARGET_AARCH64) += cpu64.o gdbstub64.o -obj-y += crypto_helper.o obj-$(CONFIG_SOFTMMU) += arm-powerctl.o DECODETREE = $(SRC_PATH)/scripts/decodetree.py @@ -XXX,XX +XXX,XX @@ target/arm/translate-sve.o: target/arm/decode-sve.inc.c target/arm/translate.o: target/arm/decode-vfp.inc.c target/arm/translate.o: target/arm/decode-vfp-uncond.inc.c +obj-y += translate.o op_helper.o +obj-y += crypto_helper.o +obj-y += iwmmxt_helper.o vec_helper.o neon_helper.o + obj-$(TARGET_AARCH64) += translate-a64.o helper-a64.o obj-$(TARGET_AARCH64) += translate-sve.o sve_helper.o obj-$(TARGET_AARCH64) += pauth_helper.o -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Group KVM rules together. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-4-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/Makefile.objs | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-) diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-y += arm-semi.o obj-$(CONFIG_SOFTMMU) += machine.o psci.o arch_dump.o monitor.o -obj-$(CONFIG_KVM) += kvm.o -obj-$(call land,$(CONFIG_KVM),$(call lnot,$(TARGET_AARCH64))) += kvm32.o -obj-$(call land,$(CONFIG_KVM),$(TARGET_AARCH64)) += kvm64.o -obj-$(call lnot,$(CONFIG_KVM)) += kvm-stub.o obj-y += helper.o vfp_helper.o obj-y += cpu.o gdbstub.o obj-$(TARGET_AARCH64) += cpu64.o gdbstub64.o obj-$(CONFIG_SOFTMMU) += arm-powerctl.o +obj-$(CONFIG_KVM) += kvm.o +obj-$(call land,$(CONFIG_KVM),$(call lnot,$(TARGET_AARCH64))) += kvm32.o +obj-$(call land,$(CONFIG_KVM),$(TARGET_AARCH64)) += kvm64.o +obj-$(call lnot,$(CONFIG_KVM)) += kvm-stub.o + DECODETREE = $(SRC_PATH)/scripts/decodetree.py target/arm/decode-sve.inc.c: $(SRC_PATH)/target/arm/sve.decode $(DECODETREE) -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Group SOFTMMU objects together. Since PSCI is TCG specific, keep it separate. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-5-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/Makefile.objs | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ obj-y += arm-semi.o -obj-$(CONFIG_SOFTMMU) += machine.o psci.o arch_dump.o monitor.o obj-y += helper.o vfp_helper.o obj-y += cpu.o gdbstub.o obj-$(TARGET_AARCH64) += cpu64.o gdbstub64.o + +obj-$(CONFIG_SOFTMMU) += machine.o arch_dump.o monitor.o obj-$(CONFIG_SOFTMMU) += arm-powerctl.o obj-$(CONFIG_KVM) += kvm.o @@ -XXX,XX +XXX,XX @@ obj-y += translate.o op_helper.o obj-y += crypto_helper.o obj-y += iwmmxt_helper.o vec_helper.o neon_helper.o +obj-$(CONFIG_SOFTMMU) += psci.o + obj-$(TARGET_AARCH64) += translate-a64.o helper-a64.o obj-$(TARGET_AARCH64) += translate-sve.o sve_helper.o obj-$(TARGET_AARCH64) += pauth_helper.o -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Robert Bradford <robert.bradford@intel.com> Reviewed-by: Samuel Ortiz <sameo@linux.intel.com> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-6-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 7 +++++++ 1 file changed, 7 insertions(+) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM generic helpers. + * + * This code is licensed under the GNU GPL v2 or later. + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ #include "qemu/osdep.h" #include "qemu/units.h" #include "target/arm/idau.h" -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-7-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ #include "exec/gdbstub.h" #include "exec/helper-proto.h" #include "qemu/host-utils.h" -#include "sysemu/arch_init.h" #include "sysemu/sysemu.h" #include "qemu/bitops.h" #include "qemu/crc32c.h" @@ -XXX,XX +XXX,XX @@ #include "hw/semihosting/semihost.h" #include "sysemu/cpus.h" #include "sysemu/kvm.h" -#include "fpu/softfloat.h" #include "qemu/range.h" #include "qapi/qapi-commands-target.h" #include "qapi/error.h" -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Since commit 8c06fbdf36b checkpatch.pl enforce a new multiline comment syntax. Since we'll move this code around, fix its style first. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-8-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 237 ++++++++++++++++++++++++++-------------- target/arm/op_helper.c | 54 ++++++--- target/arm/vfp_helper.c | 3 +- 3 files changed, 196 insertions(+), 98 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_vlldm)(CPUARMState *env, uint32_t fptr) uint32_t HELPER(v7m_tt)(CPUARMState *env, uint32_t addr, uint32_t op) { - /* The TT instructions can be used by unprivileged code, but in + /* + * The TT instructions can be used by unprivileged code, but in * user-only emulation we don't have the MPU. * Luckily since we know we are NonSecure unprivileged (and that in * turn means that the A flag wasn't specified), all the bits in the @@ -XXX,XX +XXX,XX @@ static bool v7m_stack_write(ARMCPU *cpu, uint32_t addr, uint32_t value, return true; pend_fault: - /* By pending the exception at this point we are making + /* + * By pending the exception at this point we are making * the IMPDEF choice "overridden exceptions pended" (see the * MergeExcInfo() pseudocode). The other choice would be to not * pend them now and then make a choice about which to throw away @@ -XXX,XX +XXX,XX @@ static bool v7m_stack_read(ARMCPU *cpu, uint32_t *dest, uint32_t addr, return true; pend_fault: - /* By pending the exception at this point we are making + /* + * By pending the exception at this point we are making * the IMPDEF choice "overridden exceptions pended" (see the * MergeExcInfo() pseudocode). The other choice would be to not * pend them now and then make a choice about which to throw away @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_preserve_fp_state)(CPUARMState *env) */ } -/* Write to v7M CONTROL.SPSEL bit for the specified security bank. +/* + * Write to v7M CONTROL.SPSEL bit for the specified security bank. * This may change the current stack pointer between Main and Process * stack pointers if it is done for the CONTROL register for the current * security state. @@ -XXX,XX +XXX,XX @@ static void write_v7m_control_spsel_for_secstate(CPUARMState *env, } } -/* Write to v7M CONTROL.SPSEL bit. This may change the current +/* + * Write to v7M CONTROL.SPSEL bit. This may change the current * stack pointer between Main and Process stack pointers. */ static void write_v7m_control_spsel(CPUARMState *env, bool new_spsel) @@ -XXX,XX +XXX,XX @@ static void write_v7m_control_spsel(CPUARMState *env, bool new_spsel) void write_v7m_exception(CPUARMState *env, uint32_t new_exc) { - /* Write a new value to v7m.exception, thus transitioning into or out + /* + * Write a new value to v7m.exception, thus transitioning into or out * of Handler mode; this may result in a change of active stack pointer. */ bool new_is_psp, old_is_psp = v7m_using_psp(env); @@ -XXX,XX +XXX,XX @@ static void switch_v7m_security_state(CPUARMState *env, bool new_secstate) return; } - /* All the banked state is accessed by looking at env->v7m.secure + /* + * All the banked state is accessed by looking at env->v7m.secure * except for the stack pointer; rearrange the SP appropriately. */ new_ss_msp = env->v7m.other_ss_msp; @@ -XXX,XX +XXX,XX @@ static void switch_v7m_security_state(CPUARMState *env, bool new_secstate) void HELPER(v7m_bxns)(CPUARMState *env, uint32_t dest) { - /* Handle v7M BXNS: + /* + * Handle v7M BXNS: * - if the return value is a magic value, do exception return (like BX) * - otherwise bit 0 of the return value is the target security state */ @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_bxns)(CPUARMState *env, uint32_t dest) } if (dest >= min_magic) { - /* This is an exception return magic value; put it where + /* + * This is an exception return magic value; put it where * do_v7m_exception_exit() expects and raise EXCEPTION_EXIT. * Note that if we ever add gen_ss_advance() singlestep support to * M profile this should count as an "instruction execution complete" @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_bxns)(CPUARMState *env, uint32_t dest) void HELPER(v7m_blxns)(CPUARMState *env, uint32_t dest) { - /* Handle v7M BLXNS: + /* + * Handle v7M BLXNS: * - bit 0 of the destination address is the target security state */ @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_blxns)(CPUARMState *env, uint32_t dest) assert(env->v7m.secure); if (dest & 1) { - /* target is Secure, so this is just a normal BLX, + /* + * Target is Secure, so this is just a normal BLX, * except that the low bit doesn't indicate Thumb/not. */ env->regs[14] = nextinst; @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_blxns)(CPUARMState *env, uint32_t dest) env->regs[13] = sp; env->regs[14] = 0xfeffffff; if (arm_v7m_is_handler_mode(env)) { - /* Write a dummy value to IPSR, to avoid leaking the current secure + /* + * Write a dummy value to IPSR, to avoid leaking the current secure * exception number to non-secure code. This is guaranteed not * to cause write_v7m_exception() to actually change stacks. */ @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_blxns)(CPUARMState *env, uint32_t dest) static uint32_t *get_v7m_sp_ptr(CPUARMState *env, bool secure, bool threadmode, bool spsel) { - /* Return a pointer to the location where we currently store the + /* + * Return a pointer to the location where we currently store the * stack pointer for the requested security state and thread mode. * This pointer will become invalid if the CPU state is updated * such that the stack pointers are switched around (eg changing @@ -XXX,XX +XXX,XX @@ static bool arm_v7m_load_vector(ARMCPU *cpu, int exc, bool targets_secure, mmu_idx = arm_v7m_mmu_idx_for_secstate_and_priv(env, targets_secure, true); - /* We don't do a get_phys_addr() here because the rules for vector + /* + * We don't do a get_phys_addr() here because the rules for vector * loads are special: they always use the default memory map, and * the default memory map permits reads from all addresses. * Since there's no easy way to pass through to pmsav8_mpu_lookup() @@ -XXX,XX +XXX,XX @@ static bool arm_v7m_load_vector(ARMCPU *cpu, int exc, bool targets_secure, return true; load_fail: - /* All vector table fetch fails are reported as HardFault, with + /* + * All vector table fetch fails are reported as HardFault, with * HFSR.VECTTBL and .FORCED set. (FORCED is set because * technically the underlying exception is a MemManage or BusFault * that is escalated to HardFault.) This is a terminal exception, @@ -XXX,XX +XXX,XX @@ static uint32_t v7m_integrity_sig(CPUARMState *env, uint32_t lr) static bool v7m_push_callee_stack(ARMCPU *cpu, uint32_t lr, bool dotailchain, bool ignore_faults) { - /* For v8M, push the callee-saves register part of the stack frame. + /* + * For v8M, push the callee-saves register part of the stack frame. * Compare the v8M pseudocode PushCalleeStack(). * In the tailchaining case this may not be the current stack. */ @@ -XXX,XX +XXX,XX @@ static bool v7m_push_callee_stack(ARMCPU *cpu, uint32_t lr, bool dotailchain, return true; } - /* Write as much of the stack frame as we can. A write failure may + /* + * Write as much of the stack frame as we can. A write failure may * cause us to pend a derived exception. */ sig = v7m_integrity_sig(env, lr); @@ -XXX,XX +XXX,XX @@ static bool v7m_push_callee_stack(ARMCPU *cpu, uint32_t lr, bool dotailchain, static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, bool ignore_stackfaults) { - /* Do the "take the exception" parts of exception entry, + /* + * Do the "take the exception" parts of exception entry, * but not the pushing of state to the stack. This is * similar to the pseudocode ExceptionTaken() function. */ @@ -XXX,XX +XXX,XX @@ static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, if (arm_feature(env, ARM_FEATURE_V8)) { if (arm_feature(env, ARM_FEATURE_M_SECURITY) && (lr & R_V7M_EXCRET_S_MASK)) { - /* The background code (the owner of the registers in the + /* + * The background code (the owner of the registers in the * exception frame) is Secure. This means it may either already * have or now needs to push callee-saves registers. */ if (targets_secure) { if (dotailchain && !(lr & R_V7M_EXCRET_ES_MASK)) { - /* We took an exception from Secure to NonSecure + /* + * We took an exception from Secure to NonSecure * (which means the callee-saved registers got stacked) * and are now tailchaining to a Secure exception. * Clear DCRS so eventual return from this Secure @@ -XXX,XX +XXX,XX @@ static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, lr &= ~R_V7M_EXCRET_DCRS_MASK; } } else { - /* We're going to a non-secure exception; push the + /* + * We're going to a non-secure exception; push the * callee-saves registers to the stack now, if they're * not already saved. */ @@ -XXX,XX +XXX,XX @@ static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, lr |= R_V7M_EXCRET_SPSEL_MASK; } - /* Clear registers if necessary to prevent non-secure exception + /* + * Clear registers if necessary to prevent non-secure exception * code being able to see register values from secure code. * Where register values become architecturally UNKNOWN we leave * them with their previous values. */ if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { if (!targets_secure) { - /* Always clear the caller-saved registers (they have been + /* + * Always clear the caller-saved registers (they have been * pushed to the stack earlier in v7m_push_stack()). * Clear callee-saved registers if the background code is * Secure (in which case these regs were saved in @@ -XXX,XX +XXX,XX @@ static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, } if (push_failed && !ignore_stackfaults) { - /* Derived exception on callee-saves register stacking: + /* + * Derived exception on callee-saves register stacking: * we might now want to take a different exception which * targets a different security state, so try again from the top. */ @@ -XXX,XX +XXX,XX @@ static void v7m_exception_taken(ARMCPU *cpu, uint32_t lr, bool dotailchain, return; } - /* Now we've done everything that might cause a derived exception + /* + * Now we've done everything that might cause a derived exception * we can go ahead and activate whichever exception we're going to * take (which might now be the derived exception). */ @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_vlldm)(CPUARMState *env, uint32_t fptr) static bool v7m_push_stack(ARMCPU *cpu) { - /* Do the "set up stack frame" part of exception entry, + /* + * Do the "set up stack frame" part of exception entry, * similar to pseudocode PushStack(). * Return true if we generate a derived exception (and so * should ignore further stack faults trying to process @@ -XXX,XX +XXX,XX @@ static bool v7m_push_stack(ARMCPU *cpu) } } - /* Write as much of the stack frame as we can. If we fail a stack + /* + * Write as much of the stack frame as we can. If we fail a stack * write this will result in a derived exception being pended * (which may be taken in preference to the one we started with * if it has higher priority). @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) bool ftype; bool restore_s16_s31; - /* If we're not in Handler mode then jumps to magic exception-exit + /* + * If we're not in Handler mode then jumps to magic exception-exit * addresses don't have magic behaviour. However for the v8M * security extensions the magic secure-function-return has to * work in thread mode too, so to avoid doing an extra check in @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) return; } - /* In the spec pseudocode ExceptionReturn() is called directly + /* + * In the spec pseudocode ExceptionReturn() is called directly * from BXWritePC() and gets the full target PC value including * bit zero. In QEMU's implementation we treat it as a normal * jump-to-register (which is then caught later on), and so split @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) } if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { - /* EXC_RETURN.ES validation check (R_SMFL). We must do this before + /* + * EXC_RETURN.ES validation check (R_SMFL). We must do this before * we pick which FAULTMASK to clear. */ if (!env->v7m.secure && @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) } if (env->v7m.exception != ARMV7M_EXCP_NMI) { - /* Auto-clear FAULTMASK on return from other than NMI. + /* + * Auto-clear FAULTMASK on return from other than NMI. * If the security extension is implemented then this only * happens if the raw execution priority is >= 0; the * value of the ES bit in the exception return value indicates @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) /* still an irq active now */ break; case 1: - /* we returned to base exception level, no nesting. + /* + * We returned to base exception level, no nesting. * (In the pseudocode this is written using "NestedActivation != 1" * where we have 'rettobase == false'.) */ @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) if (arm_feature(env, ARM_FEATURE_V8)) { if (!arm_feature(env, ARM_FEATURE_M_SECURITY)) { - /* UNPREDICTABLE if S == 1 or DCRS == 0 or ES == 1 (R_XLCP); + /* + * UNPREDICTABLE if S == 1 or DCRS == 0 or ES == 1 (R_XLCP); * we choose to take the UsageFault. */ if ((excret & R_V7M_EXCRET_S_MASK) || @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) break; case 13: /* Return to Thread using Process stack */ case 9: /* Return to Thread using Main stack */ - /* We only need to check NONBASETHRDENA for v7M, because in + /* + * We only need to check NONBASETHRDENA for v7M, because in * v8M this bit does not exist (it is RES1). */ if (!rettobase && @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) } if (ufault) { - /* Bad exception return: instead of popping the exception + /* + * Bad exception return: instead of popping the exception * stack, directly take a usage fault on the current stack. */ env->v7m.cfsr[env->v7m.secure] |= R_V7M_CFSR_INVPC_MASK; @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) switch_v7m_security_state(env, return_to_secure); { - /* The stack pointer we should be reading the exception frame from + /* + * The stack pointer we should be reading the exception frame from * depends on bits in the magic exception return type value (and * for v8M isn't necessarily the stack pointer we will eventually * end up resuming execution with). Get a pointer to the location @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) v7m_stack_read(cpu, &xpsr, frameptr + 0x1c, mmu_idx); if (!pop_ok) { - /* v7m_stack_read() pended a fault, so take it (as a tail + /* + * v7m_stack_read() pended a fault, so take it (as a tail * chained exception on the same stack frame) */ qemu_log_mask(CPU_LOG_INT, "...derived exception on unstacking\n"); @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) return; } - /* Returning from an exception with a PC with bit 0 set is defined + /* + * Returning from an exception with a PC with bit 0 set is defined * behaviour on v8M (bit 0 is ignored), but for v7M it was specified * to be UNPREDICTABLE. In practice actual v7M hardware seems to ignore * the lsbit, and there are several RTOSes out there which incorrectly @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) } if (arm_feature(env, ARM_FEATURE_V8)) { - /* For v8M we have to check whether the xPSR exception field + /* + * For v8M we have to check whether the xPSR exception field * matches the EXCRET value for return to handler/thread * before we commit to changing the SP and xPSR. */ bool will_be_handler = (xpsr & XPSR_EXCP) != 0; if (return_to_handler != will_be_handler) { - /* Take an INVPC UsageFault on the current stack. + /* + * Take an INVPC UsageFault on the current stack. * By this point we will have switched to the security state * for the background state, so this UsageFault will target * that state. @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) frameptr += 0x40; } } - /* Undo stack alignment (the SPREALIGN bit indicates that the original + /* + * Undo stack alignment (the SPREALIGN bit indicates that the original * pre-exception SP was not 8-aligned and we added a padding word to * align it, so we undo this by ORing in the bit that increases it * from the current 8-aligned value to the 8-unaligned value. (Adding 4 @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) V7M_CONTROL, SFPA, sfpa); } - /* The restored xPSR exception field will be zero if we're + /* + * The restored xPSR exception field will be zero if we're * resuming in Thread mode. If that doesn't match what the * exception return excret specified then this is a UsageFault. * v7M requires we make this check here; v8M did it earlier. */ if (return_to_handler != arm_v7m_is_handler_mode(env)) { - /* Take an INVPC UsageFault by pushing the stack again; + /* + * Take an INVPC UsageFault by pushing the stack again; * we know we're v7M so this is never a Secure UsageFault. */ bool ignore_stackfaults; @@ -XXX,XX +XXX,XX @@ static void do_v7m_exception_exit(ARMCPU *cpu) static bool do_v7m_function_return(ARMCPU *cpu) { - /* v8M security extensions magic function return. + /* + * v8M security extensions magic function return. * We may either: * (1) throw an exception (longjump) * (2) return true if we successfully handled the function return @@ -XXX,XX +XXX,XX @@ static bool do_v7m_function_return(ARMCPU *cpu) frame_sp_p = get_v7m_sp_ptr(env, true, threadmode, spsel); frameptr = *frame_sp_p; - /* These loads may throw an exception (for MPU faults). We want to + /* + * These loads may throw an exception (for MPU faults). We want to * do them as secure, so work out what MMU index that is. */ mmu_idx = arm_v7m_mmu_idx_for_secstate(env, true); @@ -XXX,XX +XXX,XX @@ static void arm_log_exception(int idx) static bool v7m_read_half_insn(ARMCPU *cpu, ARMMMUIdx mmu_idx, uint32_t addr, uint16_t *insn) { - /* Load a 16-bit portion of a v7M instruction, returning true on success, + /* + * Load a 16-bit portion of a v7M instruction, returning true on success, * or false on failure (in which case we will have pended the appropriate * exception). * We need to do the instruction fetch's MPU and SAU checks @@ -XXX,XX +XXX,XX @@ static bool v7m_read_half_insn(ARMCPU *cpu, ARMMMUIdx mmu_idx, v8m_security_lookup(env, addr, MMU_INST_FETCH, mmu_idx, &sattrs); if (!sattrs.nsc || sattrs.ns) { - /* This must be the second half of the insn, and it straddles a + /* + * This must be the second half of the insn, and it straddles a * region boundary with the second half not being S&NSC. */ env->v7m.sfsr |= R_V7M_SFSR_INVEP_MASK; @@ -XXX,XX +XXX,XX @@ static bool v7m_read_half_insn(ARMCPU *cpu, ARMMMUIdx mmu_idx, static bool v7m_handle_execute_nsc(ARMCPU *cpu) { - /* Check whether this attempt to execute code in a Secure & NS-Callable + /* + * Check whether this attempt to execute code in a Secure & NS-Callable * memory region is for an SG instruction; if so, then emulate the * effect of the SG instruction and return true. Otherwise pend * the correct kind of exception and return false. @@ -XXX,XX +XXX,XX @@ static bool v7m_handle_execute_nsc(ARMCPU *cpu) ARMMMUIdx mmu_idx; uint16_t insn; - /* We should never get here unless get_phys_addr_pmsav8() caused + /* + * We should never get here unless get_phys_addr_pmsav8() caused * an exception for NS executing in S&NSC memory. */ assert(!env->v7m.secure); @@ -XXX,XX +XXX,XX @@ static bool v7m_handle_execute_nsc(ARMCPU *cpu) } if (insn != 0xe97f) { - /* Not an SG instruction first half (we choose the IMPDEF + /* + * Not an SG instruction first half (we choose the IMPDEF * early-SG-check option). */ goto gen_invep; @@ -XXX,XX +XXX,XX @@ static bool v7m_handle_execute_nsc(ARMCPU *cpu) } if (insn != 0xe97f) { - /* Not an SG instruction second half (yes, both halves of the SG + /* + * Not an SG instruction second half (yes, both halves of the SG * insn have the same hex value) */ goto gen_invep; } - /* OK, we have confirmed that we really have an SG instruction. + /* + * OK, we have confirmed that we really have an SG instruction. * We know we're NS in S memory so don't need to repeat those checks. */ qemu_log_mask(CPU_LOG_INT, "...really an SG instruction at 0x%08" PRIx32 @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) arm_log_exception(cs->exception_index); - /* For exceptions we just mark as pending on the NVIC, and let that - handle it. */ + /* + * For exceptions we just mark as pending on the NVIC, and let that + * handle it. + */ switch (cs->exception_index) { case EXCP_UDEF: armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_USAGE, env->v7m.secure); @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) break; case EXCP_PREFETCH_ABORT: case EXCP_DATA_ABORT: - /* Note that for M profile we don't have a guest facing FSR, but + /* + * Note that for M profile we don't have a guest facing FSR, but * the env->exception.fsr will be populated by the code that * raises the fault, in the A profile short-descriptor format. */ switch (env->exception.fsr & 0xf) { case M_FAKE_FSR_NSC_EXEC: - /* Exception generated when we try to execute code at an address + /* + * Exception generated when we try to execute code at an address * which is marked as Secure & Non-Secure Callable and the CPU * is in the Non-Secure state. The only instruction which can * be executed like this is SG (and that only if both halves of @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) } break; case M_FAKE_FSR_SFAULT: - /* Various flavours of SecureFault for attempts to execute or + /* + * Various flavours of SecureFault for attempts to execute or * access data in the wrong security state. */ switch (cs->exception_index) { @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) armv7m_nvic_set_pending(env->nvic, ARMV7M_EXCP_BUS, false); break; default: - /* All other FSR values are either MPU faults or "can't happen + /* + * All other FSR values are either MPU faults or "can't happen * for M profile" cases. */ switch (cs->exception_index) { @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) if (arm_feature(env, ARM_FEATURE_V8)) { lr = R_V7M_EXCRET_RES1_MASK | R_V7M_EXCRET_DCRS_MASK; - /* The S bit indicates whether we should return to Secure + /* + * The S bit indicates whether we should return to Secure * or NonSecure (ie our current state). * The ES bit indicates whether we're taking this exception * to Secure or NonSecure (ie our target state). We set it @@ -XXX,XX +XXX,XX @@ void arm_v7m_cpu_do_interrupt(CPUState *cs) v7m_exception_taken(cpu, lr, false, ignore_stackfaults); } -/* Function used to synchronize QEMU's AArch64 register set with AArch32 +/* + * Function used to synchronize QEMU's AArch64 register set with AArch32 * register set. This is necessary when switching between AArch32 and AArch64 * execution state. */ @@ -XXX,XX +XXX,XX @@ void aarch64_sync_32_to_64(CPUARMState *env) env->xregs[i] = env->regs[i]; } - /* Unless we are in FIQ mode, x8-x12 come from the user registers r8-r12. + /* + * Unless we are in FIQ mode, x8-x12 come from the user registers r8-r12. * Otherwise, they come from the banked user regs. */ if (mode == ARM_CPU_MODE_FIQ) { @@ -XXX,XX +XXX,XX @@ void aarch64_sync_32_to_64(CPUARMState *env) } } - /* Registers x13-x23 are the various mode SP and FP registers. Registers + /* + * Registers x13-x23 are the various mode SP and FP registers. Registers * r13 and r14 are only copied if we are in that mode, otherwise we copy * from the mode banked register. */ @@ -XXX,XX +XXX,XX @@ void aarch64_sync_32_to_64(CPUARMState *env) env->xregs[23] = env->banked_r13[bank_number(ARM_CPU_MODE_UND)]; } - /* Registers x24-x30 are mapped to r8-r14 in FIQ mode. If we are in FIQ + /* + * Registers x24-x30 are mapped to r8-r14 in FIQ mode. If we are in FIQ * mode, then we can copy from r8-r14. Otherwise, we copy from the * FIQ bank for r8-r14. */ @@ -XXX,XX +XXX,XX @@ void aarch64_sync_32_to_64(CPUARMState *env) env->pc = env->regs[15]; } -/* Function used to synchronize QEMU's AArch32 register set with AArch64 +/* + * Function used to synchronize QEMU's AArch32 register set with AArch64 * register set. This is necessary when switching between AArch32 and AArch64 * execution state. */ @@ -XXX,XX +XXX,XX @@ void aarch64_sync_64_to_32(CPUARMState *env) env->regs[i] = env->xregs[i]; } - /* Unless we are in FIQ mode, r8-r12 come from the user registers x8-x12. + /* + * Unless we are in FIQ mode, r8-r12 come from the user registers x8-x12. * Otherwise, we copy x8-x12 into the banked user regs. */ if (mode == ARM_CPU_MODE_FIQ) { @@ -XXX,XX +XXX,XX @@ void aarch64_sync_64_to_32(CPUARMState *env) } } - /* Registers r13 & r14 depend on the current mode. + /* + * Registers r13 & r14 depend on the current mode. * If we are in a given mode, we copy the corresponding x registers to r13 * and r14. Otherwise, we copy the x register to the banked r13 and r14 * for the mode. @@ -XXX,XX +XXX,XX @@ void aarch64_sync_64_to_32(CPUARMState *env) } else { env->banked_r13[bank_number(ARM_CPU_MODE_USR)] = env->xregs[13]; - /* HYP is an exception in that it does not have its own banked r14 but + /* + * HYP is an exception in that it does not have its own banked r14 but * shares the USR r14 */ if (mode == ARM_CPU_MODE_HYP) { @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_mrs)(CPUARMState *env, uint32_t reg) return value; } case 0x94: /* CONTROL_NS */ - /* We have to handle this here because unprivileged Secure code + /* + * We have to handle this here because unprivileged Secure code * can read the NS CONTROL register. */ if (!env->v7m.secure) { @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_mrs)(CPUARMState *env, uint32_t reg) return env->v7m.faultmask[M_REG_NS]; case 0x98: /* SP_NS */ { - /* This gives the non-secure SP selected based on whether we're + /* + * This gives the non-secure SP selected based on whether we're * currently in handler mode or not, using the NS CONTROL.SPSEL. */ bool spsel = env->v7m.control[M_REG_NS] & R_V7M_CONTROL_SPSEL_MASK; @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_mrs)(CPUARMState *env, uint32_t reg) void HELPER(v7m_msr)(CPUARMState *env, uint32_t maskreg, uint32_t val) { - /* We're passed bits [11..0] of the instruction; extract + /* + * We're passed bits [11..0] of the instruction; extract * SYSm and the mask bits. * Invalid combinations of SYSm and mask are UNPREDICTABLE; * we choose to treat them as if the mask bits were valid. @@ -XXX,XX +XXX,XX @@ void HELPER(v7m_msr)(CPUARMState *env, uint32_t maskreg, uint32_t val) return; case 0x98: /* SP_NS */ { - /* This gives the non-secure SP selected based on whether we're + /* + * This gives the non-secure SP selected based on whether we're * currently in handler mode or not, using the NS CONTROL.SPSEL. */ bool spsel = env->v7m.control[M_REG_NS] & R_V7M_CONTROL_SPSEL_MASK; @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_tt)(CPUARMState *env, uint32_t addr, uint32_t op) bool targetsec = env->v7m.secure; bool is_subpage; - /* Work out what the security state and privilege level we're + /* + * Work out what the security state and privilege level we're * interested in is... */ if (alt) { @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_tt)(CPUARMState *env, uint32_t addr, uint32_t op) /* ...and then figure out which MMU index this is */ mmu_idx = arm_v7m_mmu_idx_for_secstate_and_priv(env, targetsec, targetpriv); - /* We know that the MPU and SAU don't care about the access type + /* + * We know that the MPU and SAU don't care about the access type * for our purposes beyond that we don't want to claim to be * an insn fetch, so we arbitrarily call this a read. */ - /* MPU region info only available for privileged or if + /* + * MPU region info only available for privileged or if * inspecting the other MPU state. */ if (arm_current_el(env) != 0 || alt) { @@ -XXX,XX +XXX,XX @@ bool arm_cpu_tlb_fill(CPUState *cs, vaddr address, int size, void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) { - /* Implement DC ZVA, which zeroes a fixed-length block of memory. + /* + * Implement DC ZVA, which zeroes a fixed-length block of memory. * Note that we do not implement the (architecturally mandated) * alignment fault for attempts to use this on Device memory * (which matches the usual QEMU behaviour of not implementing either @@ -XXX,XX +XXX,XX @@ void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) #ifndef CONFIG_USER_ONLY { - /* Slightly awkwardly, QEMU's TARGET_PAGE_SIZE may be less than + /* + * Slightly awkwardly, QEMU's TARGET_PAGE_SIZE may be less than * the block size so we might have to do more than one TLB lookup. * We know that in fact for any v8 CPU the page size is at least 4K * and the block size must be 2K or less, but TARGET_PAGE_SIZE is only @@ -XXX,XX +XXX,XX @@ void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) } } if (i == maxidx) { - /* If it's all in the TLB it's fair game for just writing to; + /* + * If it's all in the TLB it's fair game for just writing to; * we know we don't need to update dirty status, etc. */ for (i = 0; i < maxidx - 1; i++) { @@ -XXX,XX +XXX,XX @@ void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) memset(hostaddr[i], 0, blocklen - (i * TARGET_PAGE_SIZE)); return; } - /* OK, try a store and see if we can populate the tlb. This + /* + * OK, try a store and see if we can populate the tlb. This * might cause an exception if the memory isn't writable, * in which case we will longjmp out of here. We must for * this purpose use the actual register value passed to us @@ -XXX,XX +XXX,XX @@ void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) } } - /* Slow path (probably attempt to do this to an I/O device or + /* + * Slow path (probably attempt to do this to an I/O device or * similar, or clearing of a block of code we have translations * cached for). Just do a series of byte writes as the architecture * demands. It's not worth trying to use a cpu_physical_memory_map(), diff --git a/target/arm/op_helper.c b/target/arm/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/op_helper.c +++ b/target/arm/op_helper.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t merge_syn_data_abort(uint32_t template_syn, { uint32_t syn; - /* ISV is only set for data aborts routed to EL2 and + /* + * ISV is only set for data aborts routed to EL2 and * never for stage-1 page table walks faulting on stage 2. * * Furthermore, ISV is only set for certain kinds of load/stores. @@ -XXX,XX +XXX,XX @@ static inline uint32_t merge_syn_data_abort(uint32_t template_syn, syn = syn_data_abort_no_iss(same_el, ea, 0, s1ptw, is_write, fsc); } else { - /* Fields: IL, ISV, SAS, SSE, SRT, SF and AR come from the template + /* + * Fields: IL, ISV, SAS, SSE, SRT, SF and AR come from the template * syndrome created at translation time. * Now we create the runtime syndrome with the remaining fields. */ @@ -XXX,XX +XXX,XX @@ void arm_deliver_fault(ARMCPU *cpu, vaddr addr, MMUAccessType access_type, if (target_el == 2 || arm_el_is_aa64(env, target_el) || arm_s1_regime_using_lpae_format(env, arm_mmu_idx)) { - /* LPAE format fault status register : bottom 6 bits are + /* + * LPAE format fault status register : bottom 6 bits are * status code in the same form as needed for syndrome */ fsr = arm_fi_to_lfsc(fi); fsc = extract32(fsr, 0, 6); } else { fsr = arm_fi_to_sfsc(fi); - /* Short format FSR : this fault will never actually be reported + /* + * Short format FSR : this fault will never actually be reported * to an EL that uses a syndrome register. Use a (currently) * reserved FSR code in case the constructed syndrome does leak * into the guest somehow. @@ -XXX,XX +XXX,XX @@ void arm_cpu_do_unaligned_access(CPUState *cs, vaddr vaddr, arm_deliver_fault(cpu, vaddr, access_type, mmu_idx, &fi); } -/* arm_cpu_do_transaction_failed: handle a memory system error response +/* + * arm_cpu_do_transaction_failed: handle a memory system error response * (eg "no device/memory present at address") by raising an external abort * exception */ @@ -XXX,XX +XXX,XX @@ static bool linked_bp_matches(ARMCPU *cpu, int lbn) int bt; uint32_t contextidr; - /* Links to unimplemented or non-context aware breakpoints are + /* + * Links to unimplemented or non-context aware breakpoints are * CONSTRAINED UNPREDICTABLE: either behave as if disabled, or * as if linked to an UNKNOWN context-aware breakpoint (in which * case DBGWCR<n>_EL1.LBN must indicate that breakpoint). @@ -XXX,XX +XXX,XX @@ static bool linked_bp_matches(ARMCPU *cpu, int lbn) bt = extract64(bcr, 20, 4); - /* We match the whole register even if this is AArch32 using the + /* + * We match the whole register even if this is AArch32 using the * short descriptor format (in which case it holds both PROCID and ASID), * since we don't implement the optional v7 context ID masking. */ @@ -XXX,XX +XXX,XX @@ static bool linked_bp_matches(ARMCPU *cpu, int lbn) case 9: /* linked VMID match (reserved if no EL2) */ case 11: /* linked context ID and VMID match (reserved if no EL2) */ default: - /* Links to Unlinked context breakpoints must generate no + /* + * Links to Unlinked context breakpoints must generate no * events; we choose to do the same for reserved values too. */ return false; @@ -XXX,XX +XXX,XX @@ static bool bp_wp_matches(ARMCPU *cpu, int n, bool is_wp) CPUARMState *env = &cpu->env; uint64_t cr; int pac, hmc, ssc, wt, lbn; - /* Note that for watchpoints the check is against the CPU security + /* + * Note that for watchpoints the check is against the CPU security * state, not the S/NS attribute on the offending data access. */ bool is_secure = arm_is_secure(env); @@ -XXX,XX +XXX,XX @@ static bool bp_wp_matches(ARMCPU *cpu, int n, bool is_wp) } cr = env->cp15.dbgwcr[n]; if (wp->hitattrs.user) { - /* The LDRT/STRT/LDT/STT "unprivileged access" instructions should + /* + * The LDRT/STRT/LDT/STT "unprivileged access" instructions should * match watchpoints as if they were accesses done at EL0, even if * the CPU is at EL1 or higher. */ @@ -XXX,XX +XXX,XX @@ static bool bp_wp_matches(ARMCPU *cpu, int n, bool is_wp) } cr = env->cp15.dbgbcr[n]; } - /* The WATCHPOINT_HIT flag guarantees us that the watchpoint is + /* + * The WATCHPOINT_HIT flag guarantees us that the watchpoint is * enabled and that the address and access type match; for breakpoints * we know the address matched; check the remaining fields, including * linked breakpoints. We rely on WCR and BCR having the same layout @@ -XXX,XX +XXX,XX @@ static bool check_watchpoints(ARMCPU *cpu) CPUARMState *env = &cpu->env; int n; - /* If watchpoints are disabled globally or we can't take debug + /* + * If watchpoints are disabled globally or we can't take debug * exceptions here then watchpoint firings are ignored. */ if (extract32(env->cp15.mdscr_el1, 15, 1) == 0 @@ -XXX,XX +XXX,XX @@ static bool check_breakpoints(ARMCPU *cpu) CPUARMState *env = &cpu->env; int n; - /* If breakpoints are disabled globally or we can't take debug + /* + * If breakpoints are disabled globally or we can't take debug * exceptions here then breakpoint firings are ignored. */ if (extract32(env->cp15.mdscr_el1, 15, 1) == 0 @@ -XXX,XX +XXX,XX @@ void HELPER(check_breakpoints)(CPUARMState *env) bool arm_debug_check_watchpoint(CPUState *cs, CPUWatchpoint *wp) { - /* Called by core code when a CPU watchpoint fires; need to check if this + /* + * Called by core code when a CPU watchpoint fires; need to check if this * is also an architectural watchpoint match. */ ARMCPU *cpu = ARM_CPU(cs); @@ -XXX,XX +XXX,XX @@ vaddr arm_adjust_watchpoint_address(CPUState *cs, vaddr addr, int len) ARMCPU *cpu = ARM_CPU(cs); CPUARMState *env = &cpu->env; - /* In BE32 system mode, target memory is stored byteswapped (on a + /* + * In BE32 system mode, target memory is stored byteswapped (on a * little-endian host system), and by the time we reach here (via an * opcode helper) the addresses of subword accesses have been adjusted * to account for that, which means that watchpoints will not match. @@ -XXX,XX +XXX,XX @@ vaddr arm_adjust_watchpoint_address(CPUState *cs, vaddr addr, int len) void arm_debug_excp_handler(CPUState *cs) { - /* Called by core code when a watchpoint or breakpoint fires; + /* + * Called by core code when a watchpoint or breakpoint fires; * need to check which one and raise the appropriate exception. */ ARMCPU *cpu = ARM_CPU(cs); @@ -XXX,XX +XXX,XX @@ void arm_debug_excp_handler(CPUState *cs) uint64_t pc = is_a64(env) ? env->pc : env->regs[15]; bool same_el = (arm_debug_target_el(env) == arm_current_el(env)); - /* (1) GDB breakpoints should be handled first. + /* + * (1) GDB breakpoints should be handled first. * (2) Do not raise a CPU exception if no CPU breakpoint has fired, * since singlestep is also done by generating a debug internal * exception. @@ -XXX,XX +XXX,XX @@ void arm_debug_excp_handler(CPUState *cs) } env->exception.fsr = arm_debug_exception_fsr(env); - /* FAR is UNKNOWN: clear vaddress to avoid potentially exposing + /* + * FAR is UNKNOWN: clear vaddress to avoid potentially exposing * values to the guest that it shouldn't be able to see at its * exception/security level. */ diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(vfp_set_fpscr)(CPUARMState *env, uint32_t val) set_default_nan_mode(dnan_enabled, &env->vfp.fp_status_f16); } - /* The exception flags are ORed together when we read fpscr so we + /* + * The exception flags are ORed together when we read fpscr so we * only need to preserve the current state in one of our * float_status values. */ -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Since we'll move this code around, fix its style first. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-9-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.c | 11 ++++++----- target/arm/vfp_helper.c | 36 ++++++++++++++++++++++++------------ 2 files changed, 30 insertions(+), 17 deletions(-) diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) loaded_base = 0; loaded_var = NULL; n = 0; - for(i=0;i<16;i++) { + for (i = 0; i < 16; i++) { if (insn & (1 << i)) n++; } @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } } j = 0; - for(i=0;i<16;i++) { + for (i = 0; i < 16; i++) { if (insn & (1 << i)) { if (is_load) { /* load */ @@ -XXX,XX +XXX,XX @@ void arm_cpu_dump_state(CPUState *cs, FILE *f, int flags) return; } - for(i=0;i<16;i++) { + for (i = 0; i < 16; i++) { qemu_fprintf(f, "R%02d=%08x", i, env->regs[i]); - if ((i % 4) == 3) + if ((i % 4) == 3) { qemu_fprintf(f, "\n"); - else + } else { qemu_fprintf(f, " "); + } } if (arm_feature(env, ARM_FEATURE_M)) { diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_from_host(int host_bits) { int target_bits = 0; - if (host_bits & float_flag_invalid) + if (host_bits & float_flag_invalid) { target_bits |= 1; - if (host_bits & float_flag_divbyzero) + } + if (host_bits & float_flag_divbyzero) { target_bits |= 2; - if (host_bits & float_flag_overflow) + } + if (host_bits & float_flag_overflow) { target_bits |= 4; - if (host_bits & (float_flag_underflow | float_flag_output_denormal)) + } + if (host_bits & (float_flag_underflow | float_flag_output_denormal)) { target_bits |= 8; - if (host_bits & float_flag_inexact) + } + if (host_bits & float_flag_inexact) { target_bits |= 0x10; - if (host_bits & float_flag_input_denormal) + } + if (host_bits & float_flag_input_denormal) { target_bits |= 0x80; + } return target_bits; } @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_to_host(int target_bits) { int host_bits = 0; - if (target_bits & 1) + if (target_bits & 1) { host_bits |= float_flag_invalid; - if (target_bits & 2) + } + if (target_bits & 2) { host_bits |= float_flag_divbyzero; - if (target_bits & 4) + } + if (target_bits & 4) { host_bits |= float_flag_overflow; - if (target_bits & 8) + } + if (target_bits & 8) { host_bits |= float_flag_underflow; - if (target_bits & 0x10) + } + if (target_bits & 0x10) { host_bits |= float_flag_inexact; - if (target_bits & 0x80) + } + if (target_bits & 0x80) { host_bits |= float_flag_input_denormal; + } return host_bits; } -- 2.20.1 From: Samuel Ortiz <sameo@linux.intel.com> Those helpers are a software implementation of the ARM v8 memory zeroing op code. They should be moved to the op helper file, which is going to eventually be built only when TCG is enabled. Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Robert Bradford <robert.bradford@intel.com> Signed-off-by: Samuel Ortiz <sameo@linux.intel.com> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-10-philmd@redhat.com [PMD: Rebased] Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 92 ----------------------------------------- target/arm/op_helper.c | 93 ++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 93 insertions(+), 92 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ bool arm_cpu_tlb_fill(CPUState *cs, vaddr address, int size, #endif } -void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) -{ - /* - * Implement DC ZVA, which zeroes a fixed-length block of memory. - * Note that we do not implement the (architecturally mandated) - * alignment fault for attempts to use this on Device memory - * (which matches the usual QEMU behaviour of not implementing either - * alignment faults or any memory attribute handling). - */ - - ARMCPU *cpu = env_archcpu(env); - uint64_t blocklen = 4 << cpu->dcz_blocksize; - uint64_t vaddr = vaddr_in & ~(blocklen - 1); - -#ifndef CONFIG_USER_ONLY - { - /* - * Slightly awkwardly, QEMU's TARGET_PAGE_SIZE may be less than - * the block size so we might have to do more than one TLB lookup. - * We know that in fact for any v8 CPU the page size is at least 4K - * and the block size must be 2K or less, but TARGET_PAGE_SIZE is only - * 1K as an artefact of legacy v5 subpage support being present in the - * same QEMU executable. So in practice the hostaddr[] array has - * two entries, given the current setting of TARGET_PAGE_BITS_MIN. - */ - int maxidx = DIV_ROUND_UP(blocklen, TARGET_PAGE_SIZE); - void *hostaddr[DIV_ROUND_UP(2 * KiB, 1 << TARGET_PAGE_BITS_MIN)]; - int try, i; - unsigned mmu_idx = cpu_mmu_index(env, false); - TCGMemOpIdx oi = make_memop_idx(MO_UB, mmu_idx); - - assert(maxidx <= ARRAY_SIZE(hostaddr)); - - for (try = 0; try < 2; try++) { - - for (i = 0; i < maxidx; i++) { - hostaddr[i] = tlb_vaddr_to_host(env, - vaddr + TARGET_PAGE_SIZE * i, - 1, mmu_idx); - if (!hostaddr[i]) { - break; - } - } - if (i == maxidx) { - /* - * If it's all in the TLB it's fair game for just writing to; - * we know we don't need to update dirty status, etc. - */ - for (i = 0; i < maxidx - 1; i++) { - memset(hostaddr[i], 0, TARGET_PAGE_SIZE); - } - memset(hostaddr[i], 0, blocklen - (i * TARGET_PAGE_SIZE)); - return; - } - /* - * OK, try a store and see if we can populate the tlb. This - * might cause an exception if the memory isn't writable, - * in which case we will longjmp out of here. We must for - * this purpose use the actual register value passed to us - * so that we get the fault address right. - */ - helper_ret_stb_mmu(env, vaddr_in, 0, oi, GETPC()); - /* Now we can populate the other TLB entries, if any */ - for (i = 0; i < maxidx; i++) { - uint64_t va = vaddr + TARGET_PAGE_SIZE * i; - if (va != (vaddr_in & TARGET_PAGE_MASK)) { - helper_ret_stb_mmu(env, va, 0, oi, GETPC()); - } - } - } - - /* - * Slow path (probably attempt to do this to an I/O device or - * similar, or clearing of a block of code we have translations - * cached for). Just do a series of byte writes as the architecture - * demands. It's not worth trying to use a cpu_physical_memory_map(), - * memset(), unmap() sequence here because: - * + we'd need to account for the blocksize being larger than a page - * + the direct-RAM access case is almost always going to be dealt - * with in the fastpath code above, so there's no speed benefit - * + we would have to deal with the map returning NULL because the - * bounce buffer was in use - */ - for (i = 0; i < blocklen; i++) { - helper_ret_stb_mmu(env, vaddr + i, 0, oi, GETPC()); - } - } -#else - memset(g2h(vaddr), 0, blocklen); -#endif -} - /* Note that signed overflow is undefined in C. The following routines are careful to use unsigned types where modulo arithmetic is required. Failure to do so _will_ break on newer gcc. */ diff --git a/target/arm/op_helper.c b/target/arm/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/op_helper.c +++ b/target/arm/op_helper.c @@ -XXX,XX +XXX,XX @@ * License along with this library; if not, see <http://www.gnu.org/licenses/>. */ #include "qemu/osdep.h" +#include "qemu/units.h" #include "qemu/log.h" #include "qemu/main-loop.h" #include "cpu.h" @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(ror_cc)(CPUARMState *env, uint32_t x, uint32_t i) return ((uint32_t)x >> shift) | (x << (32 - shift)); } } + +void HELPER(dc_zva)(CPUARMState *env, uint64_t vaddr_in) +{ + /* + * Implement DC ZVA, which zeroes a fixed-length block of memory. + * Note that we do not implement the (architecturally mandated) + * alignment fault for attempts to use this on Device memory + * (which matches the usual QEMU behaviour of not implementing either + * alignment faults or any memory attribute handling). + */ + + ARMCPU *cpu = env_archcpu(env); + uint64_t blocklen = 4 << cpu->dcz_blocksize; + uint64_t vaddr = vaddr_in & ~(blocklen - 1); + +#ifndef CONFIG_USER_ONLY + { + /* + * Slightly awkwardly, QEMU's TARGET_PAGE_SIZE may be less than + * the block size so we might have to do more than one TLB lookup. + * We know that in fact for any v8 CPU the page size is at least 4K + * and the block size must be 2K or less, but TARGET_PAGE_SIZE is only + * 1K as an artefact of legacy v5 subpage support being present in the + * same QEMU executable. So in practice the hostaddr[] array has + * two entries, given the current setting of TARGET_PAGE_BITS_MIN. + */ + int maxidx = DIV_ROUND_UP(blocklen, TARGET_PAGE_SIZE); + void *hostaddr[DIV_ROUND_UP(2 * KiB, 1 << TARGET_PAGE_BITS_MIN)]; + int try, i; + unsigned mmu_idx = cpu_mmu_index(env, false); + TCGMemOpIdx oi = make_memop_idx(MO_UB, mmu_idx); + + assert(maxidx <= ARRAY_SIZE(hostaddr)); + + for (try = 0; try < 2; try++) { + + for (i = 0; i < maxidx; i++) { + hostaddr[i] = tlb_vaddr_to_host(env, + vaddr + TARGET_PAGE_SIZE * i, + 1, mmu_idx); + if (!hostaddr[i]) { + break; + } + } + if (i == maxidx) { + /* + * If it's all in the TLB it's fair game for just writing to; + * we know we don't need to update dirty status, etc. + */ + for (i = 0; i < maxidx - 1; i++) { + memset(hostaddr[i], 0, TARGET_PAGE_SIZE); + } + memset(hostaddr[i], 0, blocklen - (i * TARGET_PAGE_SIZE)); + return; + } + /* + * OK, try a store and see if we can populate the tlb. This + * might cause an exception if the memory isn't writable, + * in which case we will longjmp out of here. We must for + * this purpose use the actual register value passed to us + * so that we get the fault address right. + */ + helper_ret_stb_mmu(env, vaddr_in, 0, oi, GETPC()); + /* Now we can populate the other TLB entries, if any */ + for (i = 0; i < maxidx; i++) { + uint64_t va = vaddr + TARGET_PAGE_SIZE * i; + if (va != (vaddr_in & TARGET_PAGE_MASK)) { + helper_ret_stb_mmu(env, va, 0, oi, GETPC()); + } + } + } + + /* + * Slow path (probably attempt to do this to an I/O device or + * similar, or clearing of a block of code we have translations + * cached for). Just do a series of byte writes as the architecture + * demands. It's not worth trying to use a cpu_physical_memory_map(), + * memset(), unmap() sequence here because: + * + we'd need to account for the blocksize being larger than a page + * + the direct-RAM access case is almost always going to be dealt + * with in the fastpath code above, so there's no speed benefit + * + we would have to deal with the map returning NULL because the + * bounce buffer was in use + */ + for (i = 0; i < blocklen; i++) { + helper_ret_stb_mmu(env, vaddr + i, 0, oi, GETPC()); + } + } +#else + memset(g2h(vaddr), 0, blocklen); +#endif +} -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Suggested-by: Samuel Ortiz <sameo@linux.intel.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-11-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 2 - target/arm/translate.h | 5 - target/arm/cpu.c | 226 +++++++++++++++++++++++++++++++++++++ target/arm/translate-a64.c | 128 --------------------- target/arm/translate.c | 88 --------------- 5 files changed, 226 insertions(+), 223 deletions(-) diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ void arm_cpu_do_interrupt(CPUState *cpu); void arm_v7m_cpu_do_interrupt(CPUState *cpu); bool arm_cpu_exec_interrupt(CPUState *cpu, int int_req); -void arm_cpu_dump_state(CPUState *cs, FILE *f, int flags); - hwaddr arm_cpu_get_phys_page_attrs_debug(CPUState *cpu, vaddr addr, MemTxAttrs *attrs); diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ static inline void disas_set_insn_syndrome(DisasContext *s, uint32_t syn) #ifdef TARGET_AARCH64 void a64_translate_init(void); void gen_a64_set_pc_im(uint64_t val); -void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags); extern const TranslatorOps aarch64_translator_ops; #else static inline void a64_translate_init(void) @@ -XXX,XX +XXX,XX @@ static inline void a64_translate_init(void) static inline void gen_a64_set_pc_im(uint64_t val) { } - -static inline void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) -{ -} #endif void arm_test_cc(DisasCompare *cmp, int cc); diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/qemu-print.h" #include "qemu-common.h" #include "target/arm/idau.h" #include "qemu/module.h" @@ -XXX,XX +XXX,XX @@ static void arm_disas_set_info(CPUState *cpu, disassemble_info *info) #endif } +#ifdef TARGET_AARCH64 + +static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) +{ + ARMCPU *cpu = ARM_CPU(cs); + CPUARMState *env = &cpu->env; + uint32_t psr = pstate_read(env); + int i; + int el = arm_current_el(env); + const char *ns_status; + + qemu_fprintf(f, " PC=%016" PRIx64 " ", env->pc); + for (i = 0; i < 32; i++) { + if (i == 31) { + qemu_fprintf(f, " SP=%016" PRIx64 "\n", env->xregs[i]); + } else { + qemu_fprintf(f, "X%02d=%016" PRIx64 "%s", i, env->xregs[i], + (i + 2) % 3 ? " " : "\n"); + } + } + + if (arm_feature(env, ARM_FEATURE_EL3) && el != 3) { + ns_status = env->cp15.scr_el3 & SCR_NS ? "NS " : "S "; + } else { + ns_status = ""; + } + qemu_fprintf(f, "PSTATE=%08x %c%c%c%c %sEL%d%c", + psr, + psr & PSTATE_N ? 'N' : '-', + psr & PSTATE_Z ? 'Z' : '-', + psr & PSTATE_C ? 'C' : '-', + psr & PSTATE_V ? 'V' : '-', + ns_status, + el, + psr & PSTATE_SP ? 'h' : 't'); + + if (cpu_isar_feature(aa64_bti, cpu)) { + qemu_fprintf(f, " BTYPE=%d", (psr & PSTATE_BTYPE) >> 10); + } + if (!(flags & CPU_DUMP_FPU)) { + qemu_fprintf(f, "\n"); + return; + } + if (fp_exception_el(env, el) != 0) { + qemu_fprintf(f, " FPU disabled\n"); + return; + } + qemu_fprintf(f, " FPCR=%08x FPSR=%08x\n", + vfp_get_fpcr(env), vfp_get_fpsr(env)); + + if (cpu_isar_feature(aa64_sve, cpu) && sve_exception_el(env, el) == 0) { + int j, zcr_len = sve_zcr_len_for_el(env, el); + + for (i = 0; i <= FFR_PRED_NUM; i++) { + bool eol; + if (i == FFR_PRED_NUM) { + qemu_fprintf(f, "FFR="); + /* It's last, so end the line. */ + eol = true; + } else { + qemu_fprintf(f, "P%02d=", i); + switch (zcr_len) { + case 0: + eol = i % 8 == 7; + break; + case 1: + eol = i % 6 == 5; + break; + case 2: + case 3: + eol = i % 3 == 2; + break; + default: + /* More than one quadword per predicate. */ + eol = true; + break; + } + } + for (j = zcr_len / 4; j >= 0; j--) { + int digits; + if (j * 4 + 4 <= zcr_len + 1) { + digits = 16; + } else { + digits = (zcr_len % 4 + 1) * 4; + } + qemu_fprintf(f, "%0*" PRIx64 "%s", digits, + env->vfp.pregs[i].p[j], + j ? ":" : eol ? "\n" : " "); + } + } + + for (i = 0; i < 32; i++) { + if (zcr_len == 0) { + qemu_fprintf(f, "Z%02d=%016" PRIx64 ":%016" PRIx64 "%s", + i, env->vfp.zregs[i].d[1], + env->vfp.zregs[i].d[0], i & 1 ? "\n" : " "); + } else if (zcr_len == 1) { + qemu_fprintf(f, "Z%02d=%016" PRIx64 ":%016" PRIx64 + ":%016" PRIx64 ":%016" PRIx64 "\n", + i, env->vfp.zregs[i].d[3], env->vfp.zregs[i].d[2], + env->vfp.zregs[i].d[1], env->vfp.zregs[i].d[0]); + } else { + for (j = zcr_len; j >= 0; j--) { + bool odd = (zcr_len - j) % 2 != 0; + if (j == zcr_len) { + qemu_fprintf(f, "Z%02d[%x-%x]=", i, j, j - 1); + } else if (!odd) { + if (j > 0) { + qemu_fprintf(f, " [%x-%x]=", j, j - 1); + } else { + qemu_fprintf(f, " [%x]=", j); + } + } + qemu_fprintf(f, "%016" PRIx64 ":%016" PRIx64 "%s", + env->vfp.zregs[i].d[j * 2 + 1], + env->vfp.zregs[i].d[j * 2], + odd || j == 0 ? "\n" : ":"); + } + } + } + } else { + for (i = 0; i < 32; i++) { + uint64_t *q = aa64_vfp_qreg(env, i); + qemu_fprintf(f, "Q%02d=%016" PRIx64 ":%016" PRIx64 "%s", + i, q[1], q[0], (i & 1 ? "\n" : " ")); + } + } +} + +#else + +static inline void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) +{ + g_assert_not_reached(); +} + +#endif + +static void arm_cpu_dump_state(CPUState *cs, FILE *f, int flags) +{ + ARMCPU *cpu = ARM_CPU(cs); + CPUARMState *env = &cpu->env; + int i; + + if (is_a64(env)) { + aarch64_cpu_dump_state(cs, f, flags); + return; + } + + for (i = 0; i < 16; i++) { + qemu_fprintf(f, "R%02d=%08x", i, env->regs[i]); + if ((i % 4) == 3) { + qemu_fprintf(f, "\n"); + } else { + qemu_fprintf(f, " "); + } + } + + if (arm_feature(env, ARM_FEATURE_M)) { + uint32_t xpsr = xpsr_read(env); + const char *mode; + const char *ns_status = ""; + + if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { + ns_status = env->v7m.secure ? "S " : "NS "; + } + + if (xpsr & XPSR_EXCP) { + mode = "handler"; + } else { + if (env->v7m.control[env->v7m.secure] & R_V7M_CONTROL_NPRIV_MASK) { + mode = "unpriv-thread"; + } else { + mode = "priv-thread"; + } + } + + qemu_fprintf(f, "XPSR=%08x %c%c%c%c %c %s%s\n", + xpsr, + xpsr & XPSR_N ? 'N' : '-', + xpsr & XPSR_Z ? 'Z' : '-', + xpsr & XPSR_C ? 'C' : '-', + xpsr & XPSR_V ? 'V' : '-', + xpsr & XPSR_T ? 'T' : 'A', + ns_status, + mode); + } else { + uint32_t psr = cpsr_read(env); + const char *ns_status = ""; + + if (arm_feature(env, ARM_FEATURE_EL3) && + (psr & CPSR_M) != ARM_CPU_MODE_MON) { + ns_status = env->cp15.scr_el3 & SCR_NS ? "NS " : "S "; + } + + qemu_fprintf(f, "PSR=%08x %c%c%c%c %c %s%s%d\n", + psr, + psr & CPSR_N ? 'N' : '-', + psr & CPSR_Z ? 'Z' : '-', + psr & CPSR_C ? 'C' : '-', + psr & CPSR_V ? 'V' : '-', + psr & CPSR_T ? 'T' : 'A', + ns_status, + aarch32_mode_name(psr), (psr & 0x10) ? 32 : 26); + } + + if (flags & CPU_DUMP_FPU) { + int numvfpregs = 0; + if (arm_feature(env, ARM_FEATURE_VFP)) { + numvfpregs += 16; + } + if (arm_feature(env, ARM_FEATURE_VFP3)) { + numvfpregs += 16; + } + for (i = 0; i < numvfpregs; i++) { + uint64_t v = *aa32_vfp_dreg(env, i); + qemu_fprintf(f, "s%02d=%08x s%02d=%08x d%02d=%016" PRIx64 "\n", + i * 2, (uint32_t)v, + i * 2 + 1, (uint32_t)(v >> 32), + i, v); + } + qemu_fprintf(f, "FPSCR: %08x\n", vfp_get_fpscr(env)); + } +} + uint64_t arm_cpu_mp_affinity(int idx, uint8_t clustersz) { uint32_t Aff1 = idx / clustersz; diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ #include "translate.h" #include "internals.h" #include "qemu/host-utils.h" -#include "qemu/qemu-print.h" #include "hw/semihosting/semihost.h" #include "exec/gen-icount.h" @@ -XXX,XX +XXX,XX @@ static void set_btype(DisasContext *s, int val) s->btype = -1; } -void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) -{ - ARMCPU *cpu = ARM_CPU(cs); - CPUARMState *env = &cpu->env; - uint32_t psr = pstate_read(env); - int i; - int el = arm_current_el(env); - const char *ns_status; - - qemu_fprintf(f, " PC=%016" PRIx64 " ", env->pc); - for (i = 0; i < 32; i++) { - if (i == 31) { - qemu_fprintf(f, " SP=%016" PRIx64 "\n", env->xregs[i]); - } else { - qemu_fprintf(f, "X%02d=%016" PRIx64 "%s", i, env->xregs[i], - (i + 2) % 3 ? " " : "\n"); - } - } - - if (arm_feature(env, ARM_FEATURE_EL3) && el != 3) { - ns_status = env->cp15.scr_el3 & SCR_NS ? "NS " : "S "; - } else { - ns_status = ""; - } - qemu_fprintf(f, "PSTATE=%08x %c%c%c%c %sEL%d%c", - psr, - psr & PSTATE_N ? 'N' : '-', - psr & PSTATE_Z ? 'Z' : '-', - psr & PSTATE_C ? 'C' : '-', - psr & PSTATE_V ? 'V' : '-', - ns_status, - el, - psr & PSTATE_SP ? 'h' : 't'); - - if (cpu_isar_feature(aa64_bti, cpu)) { - qemu_fprintf(f, " BTYPE=%d", (psr & PSTATE_BTYPE) >> 10); - } - if (!(flags & CPU_DUMP_FPU)) { - qemu_fprintf(f, "\n"); - return; - } - if (fp_exception_el(env, el) != 0) { - qemu_fprintf(f, " FPU disabled\n"); - return; - } - qemu_fprintf(f, " FPCR=%08x FPSR=%08x\n", - vfp_get_fpcr(env), vfp_get_fpsr(env)); - - if (cpu_isar_feature(aa64_sve, cpu) && sve_exception_el(env, el) == 0) { - int j, zcr_len = sve_zcr_len_for_el(env, el); - - for (i = 0; i <= FFR_PRED_NUM; i++) { - bool eol; - if (i == FFR_PRED_NUM) { - qemu_fprintf(f, "FFR="); - /* It's last, so end the line. */ - eol = true; - } else { - qemu_fprintf(f, "P%02d=", i); - switch (zcr_len) { - case 0: - eol = i % 8 == 7; - break; - case 1: - eol = i % 6 == 5; - break; - case 2: - case 3: - eol = i % 3 == 2; - break; - default: - /* More than one quadword per predicate. */ - eol = true; - break; - } - } - for (j = zcr_len / 4; j >= 0; j--) { - int digits; - if (j * 4 + 4 <= zcr_len + 1) { - digits = 16; - } else { - digits = (zcr_len % 4 + 1) * 4; - } - qemu_fprintf(f, "%0*" PRIx64 "%s", digits, - env->vfp.pregs[i].p[j], - j ? ":" : eol ? "\n" : " "); - } - } - - for (i = 0; i < 32; i++) { - if (zcr_len == 0) { - qemu_fprintf(f, "Z%02d=%016" PRIx64 ":%016" PRIx64 "%s", - i, env->vfp.zregs[i].d[1], - env->vfp.zregs[i].d[0], i & 1 ? "\n" : " "); - } else if (zcr_len == 1) { - qemu_fprintf(f, "Z%02d=%016" PRIx64 ":%016" PRIx64 - ":%016" PRIx64 ":%016" PRIx64 "\n", - i, env->vfp.zregs[i].d[3], env->vfp.zregs[i].d[2], - env->vfp.zregs[i].d[1], env->vfp.zregs[i].d[0]); - } else { - for (j = zcr_len; j >= 0; j--) { - bool odd = (zcr_len - j) % 2 != 0; - if (j == zcr_len) { - qemu_fprintf(f, "Z%02d[%x-%x]=", i, j, j - 1); - } else if (!odd) { - if (j > 0) { - qemu_fprintf(f, " [%x-%x]=", j, j - 1); - } else { - qemu_fprintf(f, " [%x]=", j); - } - } - qemu_fprintf(f, "%016" PRIx64 ":%016" PRIx64 "%s", - env->vfp.zregs[i].d[j * 2 + 1], - env->vfp.zregs[i].d[j * 2], - odd || j == 0 ? "\n" : ":"); - } - } - } - } else { - for (i = 0; i < 32; i++) { - uint64_t *q = aa64_vfp_qreg(env, i); - qemu_fprintf(f, "Q%02d=%016" PRIx64 ":%016" PRIx64 "%s", - i, q[1], q[0], (i & 1 ? "\n" : " ")); - } - } -} - void gen_a64_set_pc_im(uint64_t val) { tcg_gen_movi_i64(cpu_pc, val); diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ #include "tcg-op-gvec.h" #include "qemu/log.h" #include "qemu/bitops.h" -#include "qemu/qemu-print.h" #include "arm_ldst.h" #include "hw/semihosting/semihost.h" @@ -XXX,XX +XXX,XX @@ void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) translator_loop(ops, &dc.base, cpu, tb, max_insns); } -void arm_cpu_dump_state(CPUState *cs, FILE *f, int flags) -{ - ARMCPU *cpu = ARM_CPU(cs); - CPUARMState *env = &cpu->env; - int i; - - if (is_a64(env)) { - aarch64_cpu_dump_state(cs, f, flags); - return; - } - - for (i = 0; i < 16; i++) { - qemu_fprintf(f, "R%02d=%08x", i, env->regs[i]); - if ((i % 4) == 3) { - qemu_fprintf(f, "\n"); - } else { - qemu_fprintf(f, " "); - } - } - - if (arm_feature(env, ARM_FEATURE_M)) { - uint32_t xpsr = xpsr_read(env); - const char *mode; - const char *ns_status = ""; - - if (arm_feature(env, ARM_FEATURE_M_SECURITY)) { - ns_status = env->v7m.secure ? "S " : "NS "; - } - - if (xpsr & XPSR_EXCP) { - mode = "handler"; - } else { - if (env->v7m.control[env->v7m.secure] & R_V7M_CONTROL_NPRIV_MASK) { - mode = "unpriv-thread"; - } else { - mode = "priv-thread"; - } - } - - qemu_fprintf(f, "XPSR=%08x %c%c%c%c %c %s%s\n", - xpsr, - xpsr & XPSR_N ? 'N' : '-', - xpsr & XPSR_Z ? 'Z' : '-', - xpsr & XPSR_C ? 'C' : '-', - xpsr & XPSR_V ? 'V' : '-', - xpsr & XPSR_T ? 'T' : 'A', - ns_status, - mode); - } else { - uint32_t psr = cpsr_read(env); - const char *ns_status = ""; - - if (arm_feature(env, ARM_FEATURE_EL3) && - (psr & CPSR_M) != ARM_CPU_MODE_MON) { - ns_status = env->cp15.scr_el3 & SCR_NS ? "NS " : "S "; - } - - qemu_fprintf(f, "PSR=%08x %c%c%c%c %c %s%s%d\n", - psr, - psr & CPSR_N ? 'N' : '-', - psr & CPSR_Z ? 'Z' : '-', - psr & CPSR_C ? 'C' : '-', - psr & CPSR_V ? 'V' : '-', - psr & CPSR_T ? 'T' : 'A', - ns_status, - aarch32_mode_name(psr), (psr & 0x10) ? 32 : 26); - } - - if (flags & CPU_DUMP_FPU) { - int numvfpregs = 0; - if (arm_feature(env, ARM_FEATURE_VFP)) { - numvfpregs += 16; - } - if (arm_feature(env, ARM_FEATURE_VFP3)) { - numvfpregs += 16; - } - for (i = 0; i < numvfpregs; i++) { - uint64_t v = *aa32_vfp_dreg(env, i); - qemu_fprintf(f, "s%02d=%08x s%02d=%08x d%02d=%016" PRIx64 "\n", - i * 2, (uint32_t)v, - i * 2 + 1, (uint32_t)(v >> 32), - i, v); - } - qemu_fprintf(f, "FPSCR: %08x\n", vfp_get_fpscr(env)); - } -} - void restore_state_to_opc(CPUARMState *env, TranslationBlock *tb, target_ulong *data) { -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> In the next commit we will split the TLB related routines of this file, and this function will also be called in the new file. Declare it in the "internals.h" header. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-12-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 16 ++++++++++++++++ target/arm/helper.c | 21 +++++---------------- 2 files changed, 21 insertions(+), 16 deletions(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ static inline int exception_target_el(CPUARMState *env) return target_el; } +#ifndef CONFIG_USER_ONLY + +/* Cacheability and shareability attributes for a memory access */ +typedef struct ARMCacheAttrs { + unsigned int attrs:8; /* as in the MAIR register encoding */ + unsigned int shareability:2; /* as in the SH field of the VMSAv8-64 PTEs */ +} ARMCacheAttrs; + +bool get_phys_addr(CPUARMState *env, target_ulong address, + MMUAccessType access_type, ARMMMUIdx mmu_idx, + hwaddr *phys_ptr, MemTxAttrs *attrs, int *prot, + target_ulong *page_size, + ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); + +#endif /* !CONFIG_USER_ONLY */ + #endif diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ #define ARM_CPU_FREQ 1000000000 /* FIXME: 1 GHz, should be configurable */ #ifndef CONFIG_USER_ONLY -/* Cacheability and shareability attributes for a memory access */ -typedef struct ARMCacheAttrs { - unsigned int attrs:8; /* as in the MAIR register encoding */ - unsigned int shareability:2; /* as in the SH field of the VMSAv8-64 PTEs */ -} ARMCacheAttrs; - -static bool get_phys_addr(CPUARMState *env, target_ulong address, - MMUAccessType access_type, ARMMMUIdx mmu_idx, - hwaddr *phys_ptr, MemTxAttrs *attrs, int *prot, - target_ulong *page_size, - ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, MMUAccessType access_type, ARMMMUIdx mmu_idx, @@ -XXX,XX +XXX,XX @@ static ARMCacheAttrs combine_cacheattrs(ARMCacheAttrs s1, ARMCacheAttrs s2) * @fi: set to fault info if the translation fails * @cacheattrs: (if non-NULL) set to the cacheability/shareability attributes */ -static bool get_phys_addr(CPUARMState *env, target_ulong address, - MMUAccessType access_type, ARMMMUIdx mmu_idx, - hwaddr *phys_ptr, MemTxAttrs *attrs, int *prot, - target_ulong *page_size, - ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs) +bool get_phys_addr(CPUARMState *env, target_ulong address, + MMUAccessType access_type, ARMMMUIdx mmu_idx, + hwaddr *phys_ptr, MemTxAttrs *attrs, int *prot, + target_ulong *page_size, + ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs) { if (mmu_idx == ARMMMUIdx_S12NSE0 || mmu_idx == ARMMMUIdx_S12NSE1) { /* Call ourselves recursively to do the stage 1 and then stage 2 -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> These routines are TCG specific. The arm_deliver_fault() function is only used within the new helper. Make it static. Suggested-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-13-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/Makefile.objs | 1 + target/arm/internals.h | 3 - target/arm/cpu.c | 6 +- target/arm/helper.c | 53 ----------- target/arm/op_helper.c | 135 -------------------------- target/arm/tlb_helper.c | 200 +++++++++++++++++++++++++++++++++++++++ 6 files changed, 205 insertions(+), 193 deletions(-) create mode 100644 target/arm/tlb_helper.c diff --git a/target/arm/Makefile.objs b/target/arm/Makefile.objs index XXXXXXX..XXXXXXX 100644 --- a/target/arm/Makefile.objs +++ b/target/arm/Makefile.objs @@ -XXX,XX +XXX,XX @@ target/arm/translate-sve.o: target/arm/decode-sve.inc.c target/arm/translate.o: target/arm/decode-vfp.inc.c target/arm/translate.o: target/arm/decode-vfp-uncond.inc.c +obj-y += tlb_helper.o obj-y += translate.o op_helper.o obj-y += crypto_helper.o obj-y += iwmmxt_helper.o vec_helper.o neon_helper.o diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ bool arm_cpu_tlb_fill(CPUState *cs, vaddr address, int size, MMUAccessType access_type, int mmu_idx, bool probe, uintptr_t retaddr); -void arm_deliver_fault(ARMCPU *cpu, vaddr addr, MMUAccessType access_type, - int mmu_idx, ARMMMUFaultInfo *fi) QEMU_NORETURN; - /* Return true if the stage 1 translation regime is using LPAE format page * tables */ bool arm_s1_regime_using_lpae_format(CPUARMState *env, ARMMMUIdx mmu_idx); diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_class_init(ObjectClass *oc, void *data) cc->gdb_write_register = arm_cpu_gdb_write_register; #ifndef CONFIG_USER_ONLY cc->do_interrupt = arm_cpu_do_interrupt; - cc->do_unaligned_access = arm_cpu_do_unaligned_access; - cc->do_transaction_failed = arm_cpu_do_transaction_failed; cc->get_phys_page_attrs_debug = arm_cpu_get_phys_page_attrs_debug; cc->asidx_from_attrs = arm_asidx_from_attrs; cc->vmsd = &vmstate_arm_cpu; @@ -XXX,XX +XXX,XX @@ static void arm_cpu_class_init(ObjectClass *oc, void *data) #ifdef CONFIG_TCG cc->tcg_initialize = arm_translate_init; cc->tlb_fill = arm_cpu_tlb_fill; +#if !defined(CONFIG_USER_ONLY) + cc->do_unaligned_access = arm_cpu_do_unaligned_access; + cc->do_transaction_failed = arm_cpu_do_transaction_failed; +#endif /* CONFIG_TCG && !CONFIG_USER_ONLY */ #endif } diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(v7m_tt)(CPUARMState *env, uint32_t addr, uint32_t op) #endif -bool arm_cpu_tlb_fill(CPUState *cs, vaddr address, int size, - MMUAccessType access_type, int mmu_idx, - bool probe, uintptr_t retaddr) -{ - ARMCPU *cpu = ARM_CPU(cs); - -#ifdef CONFIG_USER_ONLY - cpu->env.exception.vaddress = address; - if (access_type == MMU_INST_FETCH) { - cs->exception_index = EXCP_PREFETCH_ABORT; - } else { - cs->exception_index = EXCP_DATA_ABORT; - } - cpu_loop_exit_restore(cs, retaddr); -#else - hwaddr phys_addr; - target_ulong page_size; - int prot, ret; - MemTxAttrs attrs = {}; - ARMMMUFaultInfo fi = {}; - - /* - * Walk the page table and (if the mapping exists) add the page - * to the TLB. On success, return true. Otherwise, if probing, - * return false. Otherwise populate fsr with ARM DFSR/IFSR fault - * register format, and signal the fault. - */ - ret = get_phys_addr(&cpu->env, address, access_type, - core_to_arm_mmu_idx(&cpu->env, mmu_idx), - &phys_addr, &attrs, &prot, &page_size, &fi, NULL); - if (likely(!ret)) { - /* - * Map a single [sub]page. Regions smaller than our declared - * target page size are handled specially, so for those we - * pass in the exact addresses. - */ - if (page_size >= TARGET_PAGE_SIZE) { - phys_addr &= TARGET_PAGE_MASK; - address &= TARGET_PAGE_MASK; - } - tlb_set_page_with_attrs(cs, address, phys_addr, attrs, - prot, mmu_idx, page_size); - return true; - } else if (probe) { - return false; - } else { - /* now we have a real cpu fault */ - cpu_restore_state(cs, retaddr, true); - arm_deliver_fault(cpu, address, access_type, mmu_idx, &fi); - } -#endif -} - /* Note that signed overflow is undefined in C. The following routines are careful to use unsigned types where modulo arithmetic is required. Failure to do so _will_ break on newer gcc. */ diff --git a/target/arm/op_helper.c b/target/arm/op_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/op_helper.c +++ b/target/arm/op_helper.c @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(neon_tbl)(uint32_t ireg, uint32_t def, void *vn, return val; } -#if !defined(CONFIG_USER_ONLY) - -static inline uint32_t merge_syn_data_abort(uint32_t template_syn, - unsigned int target_el, - bool same_el, bool ea, - bool s1ptw, bool is_write, - int fsc) -{ - uint32_t syn; - - /* - * ISV is only set for data aborts routed to EL2 and - * never for stage-1 page table walks faulting on stage 2. - * - * Furthermore, ISV is only set for certain kinds of load/stores. - * If the template syndrome does not have ISV set, we should leave - * it cleared. - * - * See ARMv8 specs, D7-1974: - * ISS encoding for an exception from a Data Abort, the - * ISV field. - */ - if (!(template_syn & ARM_EL_ISV) || target_el != 2 || s1ptw) { - syn = syn_data_abort_no_iss(same_el, - ea, 0, s1ptw, is_write, fsc); - } else { - /* - * Fields: IL, ISV, SAS, SSE, SRT, SF and AR come from the template - * syndrome created at translation time. - * Now we create the runtime syndrome with the remaining fields. - */ - syn = syn_data_abort_with_iss(same_el, - 0, 0, 0, 0, 0, - ea, 0, s1ptw, is_write, fsc, - false); - /* Merge the runtime syndrome with the template syndrome. */ - syn |= template_syn; - } - return syn; -} - -void arm_deliver_fault(ARMCPU *cpu, vaddr addr, MMUAccessType access_type, - int mmu_idx, ARMMMUFaultInfo *fi) -{ - CPUARMState *env = &cpu->env; - int target_el; - bool same_el; - uint32_t syn, exc, fsr, fsc; - ARMMMUIdx arm_mmu_idx = core_to_arm_mmu_idx(env, mmu_idx); - - target_el = exception_target_el(env); - if (fi->stage2) { - target_el = 2; - env->cp15.hpfar_el2 = extract64(fi->s2addr, 12, 47) << 4; - } - same_el = (arm_current_el(env) == target_el); - - if (target_el == 2 || arm_el_is_aa64(env, target_el) || - arm_s1_regime_using_lpae_format(env, arm_mmu_idx)) { - /* - * LPAE format fault status register : bottom 6 bits are - * status code in the same form as needed for syndrome - */ - fsr = arm_fi_to_lfsc(fi); - fsc = extract32(fsr, 0, 6); - } else { - fsr = arm_fi_to_sfsc(fi); - /* - * Short format FSR : this fault will never actually be reported - * to an EL that uses a syndrome register. Use a (currently) - * reserved FSR code in case the constructed syndrome does leak - * into the guest somehow. - */ - fsc = 0x3f; - } - - if (access_type == MMU_INST_FETCH) { - syn = syn_insn_abort(same_el, fi->ea, fi->s1ptw, fsc); - exc = EXCP_PREFETCH_ABORT; - } else { - syn = merge_syn_data_abort(env->exception.syndrome, target_el, - same_el, fi->ea, fi->s1ptw, - access_type == MMU_DATA_STORE, - fsc); - if (access_type == MMU_DATA_STORE - && arm_feature(env, ARM_FEATURE_V6)) { - fsr |= (1 << 11); - } - exc = EXCP_DATA_ABORT; - } - - env->exception.vaddress = addr; - env->exception.fsr = fsr; - raise_exception(env, exc, syn, target_el); -} - -/* Raise a data fault alignment exception for the specified virtual address */ -void arm_cpu_do_unaligned_access(CPUState *cs, vaddr vaddr, - MMUAccessType access_type, - int mmu_idx, uintptr_t retaddr) -{ - ARMCPU *cpu = ARM_CPU(cs); - ARMMMUFaultInfo fi = {}; - - /* now we have a real cpu fault */ - cpu_restore_state(cs, retaddr, true); - - fi.type = ARMFault_Alignment; - arm_deliver_fault(cpu, vaddr, access_type, mmu_idx, &fi); -} - -/* - * arm_cpu_do_transaction_failed: handle a memory system error response - * (eg "no device/memory present at address") by raising an external abort - * exception - */ -void arm_cpu_do_transaction_failed(CPUState *cs, hwaddr physaddr, - vaddr addr, unsigned size, - MMUAccessType access_type, - int mmu_idx, MemTxAttrs attrs, - MemTxResult response, uintptr_t retaddr) -{ - ARMCPU *cpu = ARM_CPU(cs); - ARMMMUFaultInfo fi = {}; - - /* now we have a real cpu fault */ - cpu_restore_state(cs, retaddr, true); - - fi.ea = arm_extabort_type(response); - fi.type = ARMFault_SyncExternal; - arm_deliver_fault(cpu, addr, access_type, mmu_idx, &fi); -} - -#endif /* !defined(CONFIG_USER_ONLY) */ - void HELPER(v8m_stackcheck)(CPUARMState *env, uint32_t newvalue) { /* diff --git a/target/arm/tlb_helper.c b/target/arm/tlb_helper.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/tlb_helper.c @@ -XXX,XX +XXX,XX @@ +/* + * ARM TLB (Translation lookaside buffer) helpers. + * + * This code is licensed under the GNU GPL v2 or later. + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ +#include "qemu/osdep.h" +#include "cpu.h" +#include "internals.h" +#include "exec/exec-all.h" + +#if !defined(CONFIG_USER_ONLY) + +static inline uint32_t merge_syn_data_abort(uint32_t template_syn, + unsigned int target_el, + bool same_el, bool ea, + bool s1ptw, bool is_write, + int fsc) +{ + uint32_t syn; + + /* + * ISV is only set for data aborts routed to EL2 and + * never for stage-1 page table walks faulting on stage 2. + * + * Furthermore, ISV is only set for certain kinds of load/stores. + * If the template syndrome does not have ISV set, we should leave + * it cleared. + * + * See ARMv8 specs, D7-1974: + * ISS encoding for an exception from a Data Abort, the + * ISV field. + */ + if (!(template_syn & ARM_EL_ISV) || target_el != 2 || s1ptw) { + syn = syn_data_abort_no_iss(same_el, + ea, 0, s1ptw, is_write, fsc); + } else { + /* + * Fields: IL, ISV, SAS, SSE, SRT, SF and AR come from the template + * syndrome created at translation time. + * Now we create the runtime syndrome with the remaining fields. + */ + syn = syn_data_abort_with_iss(same_el, + 0, 0, 0, 0, 0, + ea, 0, s1ptw, is_write, fsc, + false); + /* Merge the runtime syndrome with the template syndrome. */ + syn |= template_syn; + } + return syn; +} + +static void QEMU_NORETURN arm_deliver_fault(ARMCPU *cpu, vaddr addr, + MMUAccessType access_type, + int mmu_idx, ARMMMUFaultInfo *fi) +{ + CPUARMState *env = &cpu->env; + int target_el; + bool same_el; + uint32_t syn, exc, fsr, fsc; + ARMMMUIdx arm_mmu_idx = core_to_arm_mmu_idx(env, mmu_idx); + + target_el = exception_target_el(env); + if (fi->stage2) { + target_el = 2; + env->cp15.hpfar_el2 = extract64(fi->s2addr, 12, 47) << 4; + } + same_el = (arm_current_el(env) == target_el); + + if (target_el == 2 || arm_el_is_aa64(env, target_el) || + arm_s1_regime_using_lpae_format(env, arm_mmu_idx)) { + /* + * LPAE format fault status register : bottom 6 bits are + * status code in the same form as needed for syndrome + */ + fsr = arm_fi_to_lfsc(fi); + fsc = extract32(fsr, 0, 6); + } else { + fsr = arm_fi_to_sfsc(fi); + /* + * Short format FSR : this fault will never actually be reported + * to an EL that uses a syndrome register. Use a (currently) + * reserved FSR code in case the constructed syndrome does leak + * into the guest somehow. + */ + fsc = 0x3f; + } + + if (access_type == MMU_INST_FETCH) { + syn = syn_insn_abort(same_el, fi->ea, fi->s1ptw, fsc); + exc = EXCP_PREFETCH_ABORT; + } else { + syn = merge_syn_data_abort(env->exception.syndrome, target_el, + same_el, fi->ea, fi->s1ptw, + access_type == MMU_DATA_STORE, + fsc); + if (access_type == MMU_DATA_STORE + && arm_feature(env, ARM_FEATURE_V6)) { + fsr |= (1 << 11); + } + exc = EXCP_DATA_ABORT; + } + + env->exception.vaddress = addr; + env->exception.fsr = fsr; + raise_exception(env, exc, syn, target_el); +} + +/* Raise a data fault alignment exception for the specified virtual address */ +void arm_cpu_do_unaligned_access(CPUState *cs, vaddr vaddr, + MMUAccessType access_type, + int mmu_idx, uintptr_t retaddr) +{ + ARMCPU *cpu = ARM_CPU(cs); + ARMMMUFaultInfo fi = {}; + + /* now we have a real cpu fault */ + cpu_restore_state(cs, retaddr, true); + + fi.type = ARMFault_Alignment; + arm_deliver_fault(cpu, vaddr, access_type, mmu_idx, &fi); +} + +/* + * arm_cpu_do_transaction_failed: handle a memory system error response + * (eg "no device/memory present at address") by raising an external abort + * exception + */ +void arm_cpu_do_transaction_failed(CPUState *cs, hwaddr physaddr, + vaddr addr, unsigned size, + MMUAccessType access_type, + int mmu_idx, MemTxAttrs attrs, + MemTxResult response, uintptr_t retaddr) +{ + ARMCPU *cpu = ARM_CPU(cs); + ARMMMUFaultInfo fi = {}; + + /* now we have a real cpu fault */ + cpu_restore_state(cs, retaddr, true); + + fi.ea = arm_extabort_type(response); + fi.type = ARMFault_SyncExternal; + arm_deliver_fault(cpu, addr, access_type, mmu_idx, &fi); +} + +#endif /* !defined(CONFIG_USER_ONLY) */ + +bool arm_cpu_tlb_fill(CPUState *cs, vaddr address, int size, + MMUAccessType access_type, int mmu_idx, + bool probe, uintptr_t retaddr) +{ + ARMCPU *cpu = ARM_CPU(cs); + +#ifdef CONFIG_USER_ONLY + cpu->env.exception.vaddress = address; + if (access_type == MMU_INST_FETCH) { + cs->exception_index = EXCP_PREFETCH_ABORT; + } else { + cs->exception_index = EXCP_DATA_ABORT; + } + cpu_loop_exit_restore(cs, retaddr); +#else + hwaddr phys_addr; + target_ulong page_size; + int prot, ret; + MemTxAttrs attrs = {}; + ARMMMUFaultInfo fi = {}; + + /* + * Walk the page table and (if the mapping exists) add the page + * to the TLB. On success, return true. Otherwise, if probing, + * return false. Otherwise populate fsr with ARM DFSR/IFSR fault + * register format, and signal the fault. + */ + ret = get_phys_addr(&cpu->env, address, access_type, + core_to_arm_mmu_idx(&cpu->env, mmu_idx), + &phys_addr, &attrs, &prot, &page_size, &fi, NULL); + if (likely(!ret)) { + /* + * Map a single [sub]page. Regions smaller than our declared + * target page size are handled specially, so for those we + * pass in the exact addresses. + */ + if (page_size >= TARGET_PAGE_SIZE) { + phys_addr &= TARGET_PAGE_MASK; + address &= TARGET_PAGE_MASK; + } + tlb_set_page_with_attrs(cs, address, phys_addr, attrs, + prot, mmu_idx, page_size); + return true; + } else if (probe) { + return false; + } else { + /* now we have a real cpu fault */ + cpu_restore_state(cs, retaddr, true); + arm_deliver_fault(cpu, address, access_type, mmu_idx, &fi); + } +#endif +} -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> To ease the review of the next commit, move the vfp_exceptbits_to_host() function directly after vfp_exceptbits_from_host(). Amusingly the diff shows we are moving vfp_get_fpscr(). Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-15-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/vfp_helper.c | 52 ++++++++++++++++++++--------------------- 1 file changed, 26 insertions(+), 26 deletions(-) diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_from_host(int host_bits) return target_bits; } -uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) -{ - uint32_t i, fpscr; - - fpscr = env->vfp.xregs[ARM_VFP_FPSCR] - | (env->vfp.vec_len << 16) - | (env->vfp.vec_stride << 20); - - i = get_float_exception_flags(&env->vfp.fp_status); - i |= get_float_exception_flags(&env->vfp.standard_fp_status); - /* FZ16 does not generate an input denormal exception. */ - i |= (get_float_exception_flags(&env->vfp.fp_status_f16) - & ~float_flag_input_denormal); - fpscr |= vfp_exceptbits_from_host(i); - - i = env->vfp.qc[0] | env->vfp.qc[1] | env->vfp.qc[2] | env->vfp.qc[3]; - fpscr |= i ? FPCR_QC : 0; - - return fpscr; -} - -uint32_t vfp_get_fpscr(CPUARMState *env) -{ - return HELPER(vfp_get_fpscr)(env); -} - /* Convert vfp exception flags to target form. */ static inline int vfp_exceptbits_to_host(int target_bits) { @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_to_host(int target_bits) return host_bits; } +uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) +{ + uint32_t i, fpscr; + + fpscr = env->vfp.xregs[ARM_VFP_FPSCR] + | (env->vfp.vec_len << 16) + | (env->vfp.vec_stride << 20); + + i = get_float_exception_flags(&env->vfp.fp_status); + i |= get_float_exception_flags(&env->vfp.standard_fp_status); + /* FZ16 does not generate an input denormal exception. */ + i |= (get_float_exception_flags(&env->vfp.fp_status_f16) + & ~float_flag_input_denormal); + fpscr |= vfp_exceptbits_from_host(i); + + i = env->vfp.qc[0] | env->vfp.qc[1] | env->vfp.qc[2] | env->vfp.qc[3]; + fpscr |= i ? FPCR_QC : 0; + + return fpscr; +} + +uint32_t vfp_get_fpscr(CPUARMState *env) +{ + return HELPER(vfp_get_fpscr)(env); +} + void HELPER(vfp_set_fpscr)(CPUARMState *env, uint32_t val) { int i; -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> The vfp_set_fpscr() helper contains code specific to the host floating point implementation (here the SoftFloat library). Extract this code to vfp_set_fpscr_to_host(). Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-16-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/vfp_helper.c | 127 +++++++++++++++++++++------------------- 1 file changed, 66 insertions(+), 61 deletions(-) diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_to_host(int target_bits) return host_bits; } -uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) -{ - uint32_t i, fpscr; - - fpscr = env->vfp.xregs[ARM_VFP_FPSCR] - | (env->vfp.vec_len << 16) - | (env->vfp.vec_stride << 20); - - i = get_float_exception_flags(&env->vfp.fp_status); - i |= get_float_exception_flags(&env->vfp.standard_fp_status); - /* FZ16 does not generate an input denormal exception. */ - i |= (get_float_exception_flags(&env->vfp.fp_status_f16) - & ~float_flag_input_denormal); - fpscr |= vfp_exceptbits_from_host(i); - - i = env->vfp.qc[0] | env->vfp.qc[1] | env->vfp.qc[2] | env->vfp.qc[3]; - fpscr |= i ? FPCR_QC : 0; - - return fpscr; -} - -uint32_t vfp_get_fpscr(CPUARMState *env) -{ - return HELPER(vfp_get_fpscr)(env); -} - -void HELPER(vfp_set_fpscr)(CPUARMState *env, uint32_t val) +static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val) { int i; uint32_t changed = env->vfp.xregs[ARM_VFP_FPSCR]; - /* When ARMv8.2-FP16 is not supported, FZ16 is RES0. */ - if (!cpu_isar_feature(aa64_fp16, env_archcpu(env))) { - val &= ~FPCR_FZ16; - } - - if (arm_feature(env, ARM_FEATURE_M)) { - /* - * M profile FPSCR is RES0 for the QC, STRIDE, FZ16, LEN bits - * and also for the trapped-exception-handling bits IxE. - */ - val &= 0xf7c0009f; - } - - /* - * We don't implement trapped exception handling, so the - * trap enable bits, IDE|IXE|UFE|OFE|DZE|IOE are all RAZ/WI (not RES0!) - * - * If we exclude the exception flags, IOC|DZC|OFC|UFC|IXC|IDC - * (which are stored in fp_status), and the other RES0 bits - * in between, then we clear all of the low 16 bits. - */ - env->vfp.xregs[ARM_VFP_FPSCR] = val & 0xf7c80000; - env->vfp.vec_len = (val >> 16) & 7; - env->vfp.vec_stride = (val >> 20) & 3; - - /* - * The bit we set within fpscr_q is arbitrary; the register as a - * whole being zero/non-zero is what counts. - */ - env->vfp.qc[0] = val & FPCR_QC; - env->vfp.qc[1] = 0; - env->vfp.qc[2] = 0; - env->vfp.qc[3] = 0; - changed ^= val; if (changed & (3 << 22)) { i = (val >> 22) & 3; @@ -XXX,XX +XXX,XX @@ void HELPER(vfp_set_fpscr)(CPUARMState *env, uint32_t val) set_float_exception_flags(0, &env->vfp.standard_fp_status); } +uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) +{ + uint32_t i, fpscr; + + fpscr = env->vfp.xregs[ARM_VFP_FPSCR] + | (env->vfp.vec_len << 16) + | (env->vfp.vec_stride << 20); + + i = get_float_exception_flags(&env->vfp.fp_status); + i |= get_float_exception_flags(&env->vfp.standard_fp_status); + /* FZ16 does not generate an input denormal exception. */ + i |= (get_float_exception_flags(&env->vfp.fp_status_f16) + & ~float_flag_input_denormal); + fpscr |= vfp_exceptbits_from_host(i); + + i = env->vfp.qc[0] | env->vfp.qc[1] | env->vfp.qc[2] | env->vfp.qc[3]; + fpscr |= i ? FPCR_QC : 0; + + return fpscr; +} + +uint32_t vfp_get_fpscr(CPUARMState *env) +{ + return HELPER(vfp_get_fpscr)(env); +} + +void HELPER(vfp_set_fpscr)(CPUARMState *env, uint32_t val) +{ + /* When ARMv8.2-FP16 is not supported, FZ16 is RES0. */ + if (!cpu_isar_feature(aa64_fp16, env_archcpu(env))) { + val &= ~FPCR_FZ16; + } + + if (arm_feature(env, ARM_FEATURE_M)) { + /* + * M profile FPSCR is RES0 for the QC, STRIDE, FZ16, LEN bits + * and also for the trapped-exception-handling bits IxE. + */ + val &= 0xf7c0009f; + } + + /* + * We don't implement trapped exception handling, so the + * trap enable bits, IDE|IXE|UFE|OFE|DZE|IOE are all RAZ/WI (not RES0!) + * + * If we exclude the exception flags, IOC|DZC|OFC|UFC|IXC|IDC + * (which are stored in fp_status), and the other RES0 bits + * in between, then we clear all of the low 16 bits. + */ + env->vfp.xregs[ARM_VFP_FPSCR] = val & 0xf7c80000; + env->vfp.vec_len = (val >> 16) & 7; + env->vfp.vec_stride = (val >> 20) & 3; + + /* + * The bit we set within fpscr_q is arbitrary; the register as a + * whole being zero/non-zero is what counts. + */ + env->vfp.qc[0] = val & FPCR_QC; + env->vfp.qc[1] = 0; + env->vfp.qc[2] = 0; + env->vfp.qc[3] = 0; + + vfp_set_fpscr_to_host(env, val); +} + void vfp_set_fpscr(CPUARMState *env, uint32_t val) { HELPER(vfp_set_fpscr)(env, val); -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> The vfp_set_fpscr() helper contains code specific to the host floating point implementation (here the SoftFloat library). Extract this code to vfp_set_fpscr_from_host(). Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-17-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/vfp_helper.c | 19 +++++++++++++------ 1 file changed, 13 insertions(+), 6 deletions(-) diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ static inline int vfp_exceptbits_to_host(int target_bits) return host_bits; } +static uint32_t vfp_get_fpscr_from_host(CPUARMState *env) +{ + uint32_t i; + + i = get_float_exception_flags(&env->vfp.fp_status); + i |= get_float_exception_flags(&env->vfp.standard_fp_status); + /* FZ16 does not generate an input denormal exception. */ + i |= (get_float_exception_flags(&env->vfp.fp_status_f16) + & ~float_flag_input_denormal); + return vfp_exceptbits_from_host(i); +} + static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val) { int i; @@ -XXX,XX +XXX,XX @@ uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) | (env->vfp.vec_len << 16) | (env->vfp.vec_stride << 20); - i = get_float_exception_flags(&env->vfp.fp_status); - i |= get_float_exception_flags(&env->vfp.standard_fp_status); - /* FZ16 does not generate an input denormal exception. */ - i |= (get_float_exception_flags(&env->vfp.fp_status_f16) - & ~float_flag_input_denormal); - fpscr |= vfp_exceptbits_from_host(i); + fpscr |= vfp_get_fpscr_from_host(env); i = env->vfp.qc[0] | env->vfp.qc[1] | env->vfp.qc[2] | env->vfp.qc[3]; fpscr |= i ? FPCR_QC : 0; -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> This code is specific to the SoftFloat floating-point implementation, which is only used by TCG. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-18-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/vfp_helper.c | 26 +++++++++++++++++++++++--- 1 file changed, 23 insertions(+), 3 deletions(-) diff --git a/target/arm/vfp_helper.c b/target/arm/vfp_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vfp_helper.c +++ b/target/arm/vfp_helper.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "qemu/log.h" #include "cpu.h" #include "exec/helper-proto.h" -#include "fpu/softfloat.h" #include "internals.h" - +#ifdef CONFIG_TCG +#include "qemu/log.h" +#include "fpu/softfloat.h" +#endif /* VFP support. We follow the convention used for VFP instructions: Single precision routines have a "s" suffix, double precision a "d" suffix. */ +#ifdef CONFIG_TCG + /* Convert host exception flags to vfp form. */ static inline int vfp_exceptbits_from_host(int host_bits) { @@ -XXX,XX +XXX,XX @@ static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val) set_float_exception_flags(0, &env->vfp.standard_fp_status); } +#else + +static uint32_t vfp_get_fpscr_from_host(CPUARMState *env) +{ + return 0; +} + +static void vfp_set_fpscr_to_host(CPUARMState *env, uint32_t val) +{ +} + +#endif + uint32_t HELPER(vfp_get_fpscr)(CPUARMState *env) { uint32_t i, fpscr; @@ -XXX,XX +XXX,XX @@ void vfp_set_fpscr(CPUARMState *env, uint32_t val) HELPER(vfp_set_fpscr)(env, val); } +#ifdef CONFIG_TCG + #define VFP_HELPER(name, p) HELPER(glue(glue(vfp_,name),p)) #define VFP_BINOP(name) \ @@ -XXX,XX +XXX,XX @@ float64 HELPER(frint64_d)(float64 f, void *fpst) { return frint_d(f, fpst, 64); } + +#endif -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Under KVM, the kernel gets the HVC call and handle the PSCI requests. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-20-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ vaddr arm_adjust_watchpoint_address(CPUState *cs, vaddr addr, int len); /* Callback function for when a watchpoint or breakpoint triggers. */ void arm_debug_excp_handler(CPUState *cs); -#ifdef CONFIG_USER_ONLY +#if defined(CONFIG_USER_ONLY) || !defined(CONFIG_TCG) static inline bool arm_is_psci_call(ARMCPU *cpu, int excp_type) { return false; } +static inline void arm_handle_psci_call(ARMCPU *cpu) +{ + g_assert_not_reached(); +} #else /* Return true if the r0/x0 value indicates that this SMC/HVC is a PSCI call. */ bool arm_is_psci_call(ARMCPU *cpu, int excp_type); -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> In few commits we will split the M-profile functions from this file, and this function will also be called in the new file. Declare it in the "internals.h" header. Since it is in the middle of a block of M profile functions, move it previous to this block to ease the later refactor. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-21-philmd@redhat.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 2 ++ target/arm/helper.c | 76 +++++++++++++++++++++--------------------- 2 files changed, 40 insertions(+), 38 deletions(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ bool get_phys_addr(CPUARMState *env, target_ulong address, target_ulong *page_size, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); +void arm_log_exception(int idx); + #endif /* !CONFIG_USER_ONLY */ #endif diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ uint32_t arm_phys_excp_target_el(CPUState *cs, uint32_t excp_idx, return target_el; } +void arm_log_exception(int idx) +{ + if (qemu_loglevel_mask(CPU_LOG_INT)) { + const char *exc = NULL; + static const char * const excnames[] = { + [EXCP_UDEF] = "Undefined Instruction", + [EXCP_SWI] = "SVC", + [EXCP_PREFETCH_ABORT] = "Prefetch Abort", + [EXCP_DATA_ABORT] = "Data Abort", + [EXCP_IRQ] = "IRQ", + [EXCP_FIQ] = "FIQ", + [EXCP_BKPT] = "Breakpoint", + [EXCP_EXCEPTION_EXIT] = "QEMU v7M exception exit", + [EXCP_KERNEL_TRAP] = "QEMU intercept of kernel commpage", + [EXCP_HVC] = "Hypervisor Call", + [EXCP_HYP_TRAP] = "Hypervisor Trap", + [EXCP_SMC] = "Secure Monitor Call", + [EXCP_VIRQ] = "Virtual IRQ", + [EXCP_VFIQ] = "Virtual FIQ", + [EXCP_SEMIHOST] = "Semihosting call", + [EXCP_NOCP] = "v7M NOCP UsageFault", + [EXCP_INVSTATE] = "v7M INVSTATE UsageFault", + [EXCP_STKOF] = "v8M STKOF UsageFault", + [EXCP_LAZYFP] = "v7M exception during lazy FP stacking", + [EXCP_LSERR] = "v8M LSERR UsageFault", + [EXCP_UNALIGNED] = "v7M UNALIGNED UsageFault", + }; + + if (idx >= 0 && idx < ARRAY_SIZE(excnames)) { + exc = excnames[idx]; + } + if (!exc) { + exc = "unknown"; + } + qemu_log_mask(CPU_LOG_INT, "Taking exception %d [%s]\n", idx, exc); + } +} + /* * Return true if the v7M CPACR permits access to the FPU for the specified * security state and privilege level. @@ -XXX,XX +XXX,XX @@ static bool do_v7m_function_return(ARMCPU *cpu) return true; } -static void arm_log_exception(int idx) -{ - if (qemu_loglevel_mask(CPU_LOG_INT)) { - const char *exc = NULL; - static const char * const excnames[] = { - [EXCP_UDEF] = "Undefined Instruction", - [EXCP_SWI] = "SVC", - [EXCP_PREFETCH_ABORT] = "Prefetch Abort", - [EXCP_DATA_ABORT] = "Data Abort", - [EXCP_IRQ] = "IRQ", - [EXCP_FIQ] = "FIQ", - [EXCP_BKPT] = "Breakpoint", - [EXCP_EXCEPTION_EXIT] = "QEMU v7M exception exit", - [EXCP_KERNEL_TRAP] = "QEMU intercept of kernel commpage", - [EXCP_HVC] = "Hypervisor Call", - [EXCP_HYP_TRAP] = "Hypervisor Trap", - [EXCP_SMC] = "Secure Monitor Call", - [EXCP_VIRQ] = "Virtual IRQ", - [EXCP_VFIQ] = "Virtual FIQ", - [EXCP_SEMIHOST] = "Semihosting call", - [EXCP_NOCP] = "v7M NOCP UsageFault", - [EXCP_INVSTATE] = "v7M INVSTATE UsageFault", - [EXCP_STKOF] = "v8M STKOF UsageFault", - [EXCP_LAZYFP] = "v7M exception during lazy FP stacking", - [EXCP_LSERR] = "v8M LSERR UsageFault", - [EXCP_UNALIGNED] = "v7M UNALIGNED UsageFault", - }; - - if (idx >= 0 && idx < ARRAY_SIZE(excnames)) { - exc = excnames[idx]; - } - if (!exc) { - exc = "unknown"; - } - qemu_log_mask(CPU_LOG_INT, "Taking exception %d [%s]\n", idx, exc); - } -} - static bool v7m_read_half_insn(ARMCPU *cpu, ARMMMUIdx mmu_idx, uint32_t addr, uint16_t *insn) { -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> In the next commit we will split the M-profile functions from this file. Some function will be called out of helper.c. Declare them in the "internals.h" header. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20190701132516.26392-22-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/internals.h | 42 ++++++++++++++++++++++++++++++++++++++++++ target/arm/helper.c | 38 ++------------------------------------ 2 files changed, 44 insertions(+), 36 deletions(-) diff --git a/target/arm/internals.h b/target/arm/internals.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/internals.h +++ b/target/arm/internals.h @@ -XXX,XX +XXX,XX @@ static inline uint32_t v7m_sp_limit(CPUARMState *env) } } +/** + * v7m_cpacr_pass: + * Return true if the v7M CPACR permits access to the FPU for the specified + * security state and privilege level. + */ +static inline bool v7m_cpacr_pass(CPUARMState *env, + bool is_secure, bool is_priv) +{ + switch (extract32(env->v7m.cpacr[is_secure], 20, 2)) { + case 0: + case 2: /* UNPREDICTABLE: we treat like 0 */ + return false; + case 1: + return is_priv; + case 3: + return true; + default: + g_assert_not_reached(); + } +} + /** * aarch32_mode_name(): Return name of the AArch32 CPU mode * @psr: Program Status Register indicating CPU mode @@ -XXX,XX +XXX,XX @@ static inline int exception_target_el(CPUARMState *env) #ifndef CONFIG_USER_ONLY +/* Security attributes for an address, as returned by v8m_security_lookup. */ +typedef struct V8M_SAttributes { + bool subpage; /* true if these attrs don't cover the whole TARGET_PAGE */ + bool ns; + bool nsc; + uint8_t sregion; + bool srvalid; + uint8_t iregion; + bool irvalid; +} V8M_SAttributes; + +void v8m_security_lookup(CPUARMState *env, uint32_t address, + MMUAccessType access_type, ARMMMUIdx mmu_idx, + V8M_SAttributes *sattrs); + +bool pmsav8_mpu_lookup(CPUARMState *env, uint32_t address, + MMUAccessType access_type, ARMMMUIdx mmu_idx, + hwaddr *phys_ptr, MemTxAttrs *txattrs, + int *prot, bool *is_subpage, + ARMMMUFaultInfo *fi, uint32_t *mregion); + /* Cacheability and shareability attributes for a memory access */ typedef struct ARMCacheAttrs { unsigned int attrs:8; /* as in the MAIR register encoding */ diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ static bool get_phys_addr_lpae(CPUARMState *env, target_ulong address, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, target_ulong *page_size_ptr, ARMMMUFaultInfo *fi, ARMCacheAttrs *cacheattrs); - -/* Security attributes for an address, as returned by v8m_security_lookup. */ -typedef struct V8M_SAttributes { - bool subpage; /* true if these attrs don't cover the whole TARGET_PAGE */ - bool ns; - bool nsc; - uint8_t sregion; - bool srvalid; - uint8_t iregion; - bool irvalid; -} V8M_SAttributes; - -static void v8m_security_lookup(CPUARMState *env, uint32_t address, - MMUAccessType access_type, ARMMMUIdx mmu_idx, - V8M_SAttributes *sattrs); #endif static void switch_mode(CPUARMState *env, int mode); @@ -XXX,XX +XXX,XX @@ void arm_log_exception(int idx) } } -/* - * Return true if the v7M CPACR permits access to the FPU for the specified - * security state and privilege level. - */ -static bool v7m_cpacr_pass(CPUARMState *env, bool is_secure, bool is_priv) -{ - switch (extract32(env->v7m.cpacr[is_secure], 20, 2)) { - case 0: - case 2: /* UNPREDICTABLE: we treat like 0 */ - return false; - case 1: - return is_priv; - case 3: - return true; - default: - g_assert_not_reached(); - } -} - /* * What kind of stack write are we doing? This affects how exceptions * generated during the stacking are treated. @@ -XXX,XX +XXX,XX @@ static bool v8m_is_sau_exempt(CPUARMState *env, (address >= 0xe00ff000 && address <= 0xe00fffff); } -static void v8m_security_lookup(CPUARMState *env, uint32_t address, +void v8m_security_lookup(CPUARMState *env, uint32_t address, MMUAccessType access_type, ARMMMUIdx mmu_idx, V8M_SAttributes *sattrs) { @@ -XXX,XX +XXX,XX @@ static void v8m_security_lookup(CPUARMState *env, uint32_t address, } } -static bool pmsav8_mpu_lookup(CPUARMState *env, uint32_t address, +bool pmsav8_mpu_lookup(CPUARMState *env, uint32_t address, MMUAccessType access_type, ARMMMUIdx mmu_idx, hwaddr *phys_ptr, MemTxAttrs *txattrs, int *prot, bool *is_subpage, -- 2.20.1 I don't have anything else queued up at the moment, so this is just Richard's SME patches. -- PMM The following changes since commit 63b38f6c85acd312c2cab68554abf33adf4ee2b3: Merge tag 'pull-target-arm-20220707' of https://git.linaro.org/people/pmaydell/qemu-arm into staging (2022-07-08 06:17:11 +0530) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20220711 for you to fetch changes up to f9982ceaf26df27d15547a3a7990a95019e9e3a8: linux-user/aarch64: Add SME related hwcap entries (2022-07-11 13:43:52 +0100) ---------------------------------------------------------------- target-arm: * Implement SME emulation, for both system and linux-user ---------------------------------------------------------------- Richard Henderson (45): target/arm: Handle SME in aarch64_cpu_dump_state target/arm: Add infrastructure for disas_sme target/arm: Trap non-streaming usage when Streaming SVE is active target/arm: Mark ADR as non-streaming target/arm: Mark RDFFR, WRFFR, SETFFR as non-streaming target/arm: Mark BDEP, BEXT, BGRP, COMPACT, FEXPA, FTSSEL as non-streaming target/arm: Mark PMULL, FMMLA as non-streaming target/arm: Mark FTSMUL, FTMAD, FADDA as non-streaming target/arm: Mark SMMLA, UMMLA, USMMLA as non-streaming target/arm: Mark string/histo/crypto as non-streaming target/arm: Mark gather/scatter load/store as non-streaming target/arm: Mark gather prefetch as non-streaming target/arm: Mark LDFF1 and LDNF1 as non-streaming target/arm: Mark LD1RO as non-streaming target/arm: Add SME enablement checks target/arm: Handle SME in sve_access_check target/arm: Implement SME RDSVL, ADDSVL, ADDSPL target/arm: Implement SME ZERO target/arm: Implement SME MOVA target/arm: Implement SME LD1, ST1 target/arm: Export unpredicated ld/st from translate-sve.c target/arm: Implement SME LDR, STR target/arm: Implement SME ADDHA, ADDVA target/arm: Implement FMOPA, FMOPS (non-widening) target/arm: Implement BFMOPA, BFMOPS target/arm: Implement FMOPA, FMOPS (widening) target/arm: Implement SME integer outer product target/arm: Implement PSEL target/arm: Implement REVD target/arm: Implement SCLAMP, UCLAMP target/arm: Reset streaming sve state on exception boundaries target/arm: Enable SME for -cpu max linux-user/aarch64: Clear tpidr2_el0 if CLONE_SETTLS linux-user/aarch64: Reset PSTATE.SM on syscalls linux-user/aarch64: Add SM bit to SVE signal context linux-user/aarch64: Tidy target_restore_sigframe error return linux-user/aarch64: Do not allow duplicate or short sve records linux-user/aarch64: Verify extra record lock succeeded linux-user/aarch64: Move sve record checks into restore linux-user/aarch64: Implement SME signal handling linux-user: Rename sve prctls linux-user/aarch64: Implement PR_SME_GET_VL, PR_SME_SET_VL target/arm: Only set ZEN in reset if SVE present target/arm: Enable SME for user-only linux-user/aarch64: Add SME related hwcap entries docs/system/arm/emulation.rst | 4 + linux-user/aarch64/target_cpu.h | 5 +- linux-user/aarch64/target_prctl.h | 62 +- target/arm/cpu.h | 7 + target/arm/helper-sme.h | 126 ++++ target/arm/helper-sve.h | 4 + target/arm/helper.h | 18 + target/arm/translate-a64.h | 45 ++ target/arm/translate.h | 16 + target/arm/sme-fa64.decode | 60 ++ target/arm/sme.decode | 88 +++ target/arm/sve.decode | 41 +- linux-user/aarch64/cpu_loop.c | 9 + linux-user/aarch64/signal.c | 243 ++++++-- linux-user/elfload.c | 20 + linux-user/syscall.c | 28 +- target/arm/cpu.c | 35 +- target/arm/cpu64.c | 11 + target/arm/helper.c | 56 +- target/arm/sme_helper.c | 1140 +++++++++++++++++++++++++++++++++++++ target/arm/sve_helper.c | 28 + target/arm/translate-a64.c | 103 +++- target/arm/translate-sme.c | 373 ++++++++++++ target/arm/translate-sve.c | 393 ++++++++++--- target/arm/translate-vfp.c | 12 + target/arm/translate.c | 2 + target/arm/vec_helper.c | 24 + target/arm/meson.build | 3 + 28 files changed, 2821 insertions(+), 135 deletions(-) create mode 100644 target/arm/sme-fa64.decode create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c From: Richard Henderson <richard.henderson@linaro.org> Dump SVCR, plus use the correct access check for Streaming Mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-2-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 17 ++++++++++++++++- 1 file changed, 16 insertions(+), 1 deletion(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) int i; int el = arm_current_el(env); const char *ns_status; + bool sve; qemu_fprintf(f, " PC=%016" PRIx64 " ", env->pc); for (i = 0; i < 32; i++) { @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) el, psr & PSTATE_SP ? 'h' : 't'); + if (cpu_isar_feature(aa64_sme, cpu)) { + qemu_fprintf(f, " SVCR=%08" PRIx64 " %c%c", + env->svcr, + (FIELD_EX64(env->svcr, SVCR, ZA) ? 'Z' : '-'), + (FIELD_EX64(env->svcr, SVCR, SM) ? 'S' : '-')); + } if (cpu_isar_feature(aa64_bti, cpu)) { qemu_fprintf(f, " BTYPE=%d", (psr & PSTATE_BTYPE) >> 10); } @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_dump_state(CPUState *cs, FILE *f, int flags) qemu_fprintf(f, " FPCR=%08x FPSR=%08x\n", vfp_get_fpcr(env), vfp_get_fpsr(env)); - if (cpu_isar_feature(aa64_sve, cpu) && sve_exception_el(env, el) == 0) { + if (cpu_isar_feature(aa64_sme, cpu) && FIELD_EX64(env->svcr, SVCR, SM)) { + sve = sme_exception_el(env, el) == 0; + } else if (cpu_isar_feature(aa64_sve, cpu)) { + sve = sve_exception_el(env, el) == 0; + } else { + sve = false; + } + + if (sve) { int j, zcr_len = sve_vqm1_for_el(env, el); for (i = 0; i <= FFR_PRED_NUM; i++) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This includes the build rules for the decoder, and the new file for translation, but excludes any instructions. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-3-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 1 + target/arm/sme.decode | 20 ++++++++++++++++++++ target/arm/translate-a64.c | 7 ++++++- target/arm/translate-sme.c | 35 +++++++++++++++++++++++++++++++++++ target/arm/meson.build | 2 ++ 5 files changed, 64 insertions(+), 1 deletion(-) create mode 100644 target/arm/sme.decode create mode 100644 target/arm/translate-sme.c diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) } bool disas_sve(DisasContext *, uint32_t); +bool disas_sme(DisasContext *, uint32_t); void gen_gvec_rax1(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz); diff --git a/target/arm/sme.decode b/target/arm/sme.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME instruction descriptions +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } switch (extract32(insn, 25, 4)) { - case 0x0: case 0x1: case 0x3: /* UNALLOCATED */ + case 0x0: + if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { + unallocated_encoding(s); + } + break; + case 0x1: case 0x3: /* UNALLOCATED */ unallocated_encoding(s); break; case 0x2: diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ +/* + * AArch64 SME translation + * + * Copyright (c) 2022 Linaro, Ltd + * + * This library is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * This library is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with this library; if not, see <http://www.gnu.org/licenses/>. + */ + +#include "qemu/osdep.h" +#include "cpu.h" +#include "tcg/tcg-op.h" +#include "tcg/tcg-op-gvec.h" +#include "tcg/tcg-gvec-desc.h" +#include "translate.h" +#include "exec/helper-gen.h" +#include "translate-a64.h" +#include "fpu/softfloat.h" + + +/* + * Include the generated decoder. + */ + +#include "decode-sme.c.inc" diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), + decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), @@ -XXX,XX +XXX,XX @@ arm_ss.add(when: 'TARGET_AARCH64', if_true: files( 'sme_helper.c', 'translate-a64.c', 'translate-sve.c', + 'translate-sme.c', )) arm_softmmu_ss = ss.source_set() -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This new behaviour is in the ARM pseudocode function AArch64.CheckFPAdvSIMDEnabled, which applies to AArch32 via AArch32.CheckAdvSIMDOrFPEnabled when the EL to which the trap would be delivered is in AArch64 mode. Given that ARMv9 drops support for AArch32 outside EL0, the trap EL detection ought to be trivially true, but the pseudocode still contains a number of conditions, and QEMU has not yet committed to dropping A32 support for EL[12] when v9 features are present. Since the computation of SME_TRAP_NONSTREAMING is necessarily different for the two modes, we might as well preserve bits within TBFLAG_ANY and allocate separate bits within TBFLAG_A32 and TBFLAG_A64 instead. Note that DDI0616A.a has typos for bits [22:21] of LD1RO in the table of instructions illegal in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-4-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 7 +++ target/arm/translate.h | 4 ++ target/arm/sme-fa64.decode | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/helper.c | 41 +++++++++++++++++ target/arm/translate-a64.c | 40 ++++++++++++++++- target/arm/translate-vfp.c | 12 +++++ target/arm/translate.c | 2 + target/arm/meson.build | 1 + 8 files changed, 195 insertions(+), 2 deletions(-) create mode 100644 target/arm/sme-fa64.decode diff --git a/target/arm/cpu.h b/target/arm/cpu.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.h +++ b/target/arm/cpu.h @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A32, HSTR_ACTIVE, 9, 1) * the same thing as the current security state of the processor! */ FIELD(TBFLAG_A32, NS, 10, 1) +/* + * Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. + * This requires an SME trap from AArch32 mode when using NEON. + */ +FIELD(TBFLAG_A32, SME_TRAP_NONSTREAMING, 11, 1) /* * Bit usage when in AArch32 state, for M-profile only. @@ -XXX,XX +XXX,XX @@ FIELD(TBFLAG_A64, SMEEXC_EL, 20, 2) FIELD(TBFLAG_A64, PSTATE_SM, 22, 1) FIELD(TBFLAG_A64, PSTATE_ZA, 23, 1) FIELD(TBFLAG_A64, SVL, 24, 4) +/* Indicates that SME Streaming mode is active, and SMCR_ELx.FA64 is not. */ +FIELD(TBFLAG_A64, SME_TRAP_NONSTREAMING, 28, 1) /* * Helpers for using the above. diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { bool pstate_sm; /* True if PSTATE.ZA is set. */ bool pstate_za; + /* True if non-streaming insns should raise an SME Streaming exception. */ + bool sme_trap_nonstreaming; + /* True if the current instruction is non-streaming. */ + bool is_nonstreaming; /* True if MVE insns are definitely not predicated by VPR or LTPSIZE */ bool mve_no_pred; /* diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ +# AArch64 SME allowed instruction decoding +# +# Copyright (c) 2022 Linaro, Ltd +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. + +# +# This file is processed by scripts/decodetree.py +# + +# These patterns are taken from Appendix E1.1 of DDI0616 A.a, +# Arm Architecture Reference Manual Supplement, +# The Scalable Matrix Extension (SME), for Armv9-A + +{ + [ + OK 0-00 1110 0000 0001 0010 11-- ---- ---- # SMOV W|Xd,Vn.B[0] + OK 0-00 1110 0000 0010 0010 11-- ---- ---- # SMOV W|Xd,Vn.H[0] + OK 0100 1110 0000 0100 0010 11-- ---- ---- # SMOV Xd,Vn.S[0] + OK 0000 1110 0000 0001 0011 11-- ---- ---- # UMOV Wd,Vn.B[0] + OK 0000 1110 0000 0010 0011 11-- ---- ---- # UMOV Wd,Vn.H[0] + OK 0000 1110 0000 0100 0011 11-- ---- ---- # UMOV Wd,Vn.S[0] + OK 0100 1110 0000 1000 0011 11-- ---- ---- # UMOV Xd,Vn.D[0] + ] + FAIL 0--0 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD vector operations +} + +{ + [ + OK 0101 1110 --1- ---- 11-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar) + OK 0101 1110 -10- ---- 00-1 11-- ---- ---- # FMULX/FRECPS/FRSQRTS (scalar, FP16) + OK 01-1 1110 1-10 0001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar) + OK 01-1 1110 1111 1001 11-1 10-- ---- ---- # FRECPE/FRSQRTE/FRECPX (scalar, FP16) + ] + FAIL 01-1 111- ---- ---- ---- ---- ---- ---- # Advanced SIMD single-element operations +} + +FAIL 0-00 110- ---- ---- ---- ---- ---- ---- # Advanced SIMD structure load/store +FAIL 1100 1110 ---- ---- ---- ---- ---- ---- # Advanced SIMD cryptography extensions +FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS + +# These are the "avoidance of doubt" final table of Illegal Advanced SIMD instructions +# We don't actually need to include these, as the default is OK. +# -001 111- ---- ---- ---- ---- ---- ---- # Scalar floating-point operations +# --10 110- ---- ---- ---- ---- ---- ---- # Load/store pair of FP registers +# --01 1100 ---- ---- ---- ---- ---- ---- # Load FP register (PC-relative literal) +# --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) +# --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) +# --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) + +FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR +FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA +FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT +FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS +FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR +FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP +FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) +FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA +FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL +FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD +FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA +FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA +FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions +FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) +FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) +FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) +FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) +FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) +FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) +FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) +FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) +FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) +FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) +FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) +FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch +FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) +FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) +FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) +FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ int sme_exception_el(CPUARMState *env, int el) return 0; } +/* This corresponds to the ARM pseudocode function IsFullA64Enabled(). */ +static bool sme_fa64(CPUARMState *env, int el) +{ + if (!cpu_isar_feature(aa64_sme_fa64, env_archcpu(env))) { + return false; + } + + if (el <= 1 && !el_is_in_host(env, el)) { + if (!FIELD_EX64(env->vfp.smcr_el[1], SMCR, FA64)) { + return false; + } + } + if (el <= 2 && arm_is_el2_enabled(env)) { + if (!FIELD_EX64(env->vfp.smcr_el[2], SMCR, FA64)) { + return false; + } + } + if (arm_feature(env, ARM_FEATURE_EL3)) { + if (!FIELD_EX64(env->vfp.smcr_el[3], SMCR, FA64)) { + return false; + } + } + + return true; +} + /* * Given that SVE is enabled, return the vector length for EL. */ @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a32(CPUARMState *env, int fp_el, DP_TBFLAG_ANY(flags, PSTATE__IL, 1); } + /* + * The SME exception we are testing for is raised via + * AArch64.CheckFPAdvSIMDEnabled(), as called from + * AArch32.CheckAdvSIMDOrFPEnabled(). + */ + if (el == 0 + && FIELD_EX64(env->svcr, SVCR, SM) + && (!arm_is_el2_enabled(env) + || (arm_el_is_aa64(env, 2) && !(env->cp15.hcr_el2 & HCR_TGE))) + && arm_el_is_aa64(env, 1) + && !sme_fa64(env, el)) { + DP_TBFLAG_A32(flags, SME_TRAP_NONSTREAMING, 1); + } + return rebuild_hflags_common_32(env, fp_el, mmu_idx, flags); } @@ -XXX,XX +XXX,XX @@ static CPUARMTBFlags rebuild_hflags_a64(CPUARMState *env, int el, int fp_el, } if (FIELD_EX64(env->svcr, SVCR, SM)) { DP_TBFLAG_A64(flags, PSTATE_SM, 1); + DP_TBFLAG_A64(flags, SME_TRAP_NONSTREAMING, !sme_fa64(env, el)); } DP_TBFLAG_A64(flags, PSTATE_ZA, FIELD_EX64(env->svcr, SVCR, ZA)); } diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static void do_vec_ld(DisasContext *s, int destidx, int element, * unallocated-encoding checks (otherwise the syndrome information * for the resulting exception will be incorrect). */ -static bool fp_access_check(DisasContext *s) +static bool fp_access_check_only(DisasContext *s) { if (s->fp_excp_el) { assert(!s->fp_access_checked); @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } +static bool fp_access_check(DisasContext *s) +{ + if (!fp_access_check_only(s)) { + return false; + } + if (s->sme_trap_nonstreaming && s->is_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, false)); + return false; + } + return true; +} + /* Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. */ @@ -XXX,XX +XXX,XX @@ static void handle_sys(DisasContext *s, uint32_t insn, bool isread, default: g_assert_not_reached(); } - if ((ri->type & ARM_CP_FPU) && !fp_access_check(s)) { + if ((ri->type & ARM_CP_FPU) && !fp_access_check_only(s)) { return; } else if ((ri->type & ARM_CP_SVE) && !sve_access_check(s)) { return; @@ -XXX,XX +XXX,XX @@ static void disas_data_proc_simd_fp(DisasContext *s, uint32_t insn) } } +/* + * Include the generated SME FA64 decoder. + */ + +#include "decode-sme-fa64.c.inc" + +static bool trans_OK(DisasContext *s, arg_OK *a) +{ + return true; +} + +static bool trans_FAIL(DisasContext *s, arg_OK *a) +{ + s->is_nonstreaming = true; + return true; +} + /** * is_guarded_page: * @env: The cpu environment @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_init_disas_context(DisasContextBase *dcbase, dc->mte_active[1] = EX_TBFLAG_A64(tb_flags, MTE0_ACTIVE); dc->pstate_sm = EX_TBFLAG_A64(tb_flags, PSTATE_SM); dc->pstate_za = EX_TBFLAG_A64(tb_flags, PSTATE_ZA); + dc->sme_trap_nonstreaming = EX_TBFLAG_A64(tb_flags, SME_TRAP_NONSTREAMING); dc->vec_len = 0; dc->vec_stride = 0; dc->cp_regs = arm_cpu->cp_regs; @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } } + s->is_nonstreaming = false; + if (s->sme_trap_nonstreaming) { + disas_sme_fa64(s, insn); + } + switch (extract32(insn, 25, 4)) { case 0x0: if (!extract32(insn, 31, 1) || !disas_sme(s, insn)) { diff --git a/target/arm/translate-vfp.c b/target/arm/translate-vfp.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.c +++ b/target/arm/translate-vfp.c @@ -XXX,XX +XXX,XX @@ static bool vfp_access_check_a(DisasContext *s, bool ignore_vfp_enabled) return false; } + /* + * Note that rebuild_hflags_a32 has already accounted for being in EL0 + * and the higher EL in A64 mode, etc. Unlike A64 mode, there do not + * appear to be any insns which touch VFP which are allowed. + */ + if (s->sme_trap_nonstreaming) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_Streaming, + s->base.pc_next - s->pc_curr == 2)); + return false; + } + if (!s->vfp_enabled && !ignore_vfp_enabled) { assert(!arm_dc_feature(s, ARM_FEATURE_M)); unallocated_encoding(s); diff --git a/target/arm/translate.c b/target/arm/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.c +++ b/target/arm/translate.c @@ -XXX,XX +XXX,XX @@ static void arm_tr_init_disas_context(DisasContextBase *dcbase, CPUState *cs) dc->vec_len = EX_TBFLAG_A32(tb_flags, VECLEN); dc->vec_stride = EX_TBFLAG_A32(tb_flags, VECSTRIDE); } + dc->sme_trap_nonstreaming = + EX_TBFLAG_A32(tb_flags, SME_TRAP_NONSTREAMING); } dc->cp_regs = cpu->cp_regs; dc->features = env->features; diff --git a/target/arm/meson.build b/target/arm/meson.build index XXXXXXX..XXXXXXX 100644 --- a/target/arm/meson.build +++ b/target/arm/meson.build @@ -XXX,XX +XXX,XX @@ gen = [ decodetree.process('sve.decode', extra_args: '--decode=disas_sve'), decodetree.process('sme.decode', extra_args: '--decode=disas_sme'), + decodetree.process('sme-fa64.decode', extra_args: '--static-decode=disas_sme_fa64'), decodetree.process('neon-shared.decode', extra_args: '--decode=disas_neon_shared'), decodetree.process('neon-dp.decode', extra_args: '--decode=disas_neon_dp'), decodetree.process('neon-ls.decode', extra_args: '--decode=disas_neon_ls'), -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark ADR as a non-streaming instruction, which should trap if full a64 support is not enabled in streaming mode. Removing entries from sme-fa64.decode is an easy way to see what remains to be done. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-5-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate.h | 7 +++++++ target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 8 ++++---- 3 files changed, 11 insertions(+), 5 deletions(-) diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ uint64_t asimd_imm_const(uint32_t imm, int cmode, int op); static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ { return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); } +#define TRANS_FEAT_NONSTREAMING(NAME, FEAT, FUNC, ...) \ + static bool trans_##NAME(DisasContext *s, arg_##NAME *a) \ + { \ + s->is_nonstreaming = true; \ + return dc_isar_feature(FEAT, s) && FUNC(s, __VA_ARGS__); \ + } + #endif /* TARGET_ARM_TRANSLATE_H */ diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1010 ---- ---- ---- # ADR FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_adr(DisasContext *s, arg_rrri *a, gen_helper_gvec_3 *fn) return gen_gvec_ool_zzz(s, fn, a->rd, a->rn, a->rm, a->imm); } -TRANS_FEAT(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) -TRANS_FEAT(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) -TRANS_FEAT(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) -TRANS_FEAT(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) +TRANS_FEAT_NONSTREAMING(ADR_p32, aa64_sve, do_adr, a, gen_helper_sve_adr_p32) +TRANS_FEAT_NONSTREAMING(ADR_p64, aa64_sve, do_adr, a, gen_helper_sve_adr_p64) +TRANS_FEAT_NONSTREAMING(ADR_s32, aa64_sve, do_adr, a, gen_helper_sve_adr_s32) +TRANS_FEAT_NONSTREAMING(ADR_u32, aa64_sve, do_adr, a, gen_helper_sve_adr_u32) /* *** SVE Integer Misc - Unpredicated Group -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-6-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 9 ++++++--- 2 files changed, 6 insertions(+), 5 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0010 0101 --01 100- 1111 000- ---0 ---- # RDFFR, RDFFRS -FAIL 0010 0101 --10 1--- 1001 ---- ---- ---- # WRFFR, SETFFR FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_predset(DisasContext *s, int esz, int rd, int pat, bool setflag) TRANS_FEAT(PTRUE, aa64_sve, do_predset, a->esz, a->rd, a->pat, a->s) /* Note pat == 31 is #all, to set all elements. */ -TRANS_FEAT(SETFFR, aa64_sve, do_predset, 0, FFR_PRED_NUM, 31, false) +TRANS_FEAT_NONSTREAMING(SETFFR, aa64_sve, + do_predset, 0, FFR_PRED_NUM, 31, false) /* Note pat == 32 is #unimp, to set no elements. */ TRANS_FEAT(PFALSE, aa64_sve, do_predset, 0, a->rd, 32, false) @@ -XXX,XX +XXX,XX @@ static bool trans_RDFFR_p(DisasContext *s, arg_RDFFR_p *a) .rd = a->rd, .pg = a->pg, .s = a->s, .rn = FFR_PRED_NUM, .rm = FFR_PRED_NUM, }; + + s->is_nonstreaming = true; return trans_AND_pppp(s, &alt_a); } -TRANS_FEAT(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) -TRANS_FEAT(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) +TRANS_FEAT_NONSTREAMING(RDFFR, aa64_sve, do_mov_p, a->rd, FFR_PRED_NUM) +TRANS_FEAT_NONSTREAMING(WRFFR, aa64_sve, do_mov_p, FFR_PRED_NUM, a->rn) static bool do_pfirst_pnext(DisasContext *s, arg_rr_esz *a, void (*gen_fn)(TCGv_i32, TCGv_ptr, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-7-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 22 ++++++++++++---------- 2 files changed, 12 insertions(+), 13 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0000 0100 --1- ---- 1011 -0-- ---- ---- # FTSSEL, FEXPA -FAIL 0000 0101 --10 0001 100- ---- ---- ---- # COMPACT -FAIL 0100 0101 --0- ---- 1011 ---- ---- ---- # BDEP, BEXT, BGRP FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_2 * const fexpa_fns[4] = { NULL, gen_helper_sve_fexpa_h, gen_helper_sve_fexpa_s, gen_helper_sve_fexpa_d, }; -TRANS_FEAT(FEXPA, aa64_sve, gen_gvec_ool_zz, - fexpa_fns[a->esz], a->rd, a->rn, 0) +TRANS_FEAT_NONSTREAMING(FEXPA, aa64_sve, gen_gvec_ool_zz, + fexpa_fns[a->esz], a->rd, a->rn, 0) static gen_helper_gvec_3 * const ftssel_fns[4] = { NULL, gen_helper_sve_ftssel_h, gen_helper_sve_ftssel_s, gen_helper_sve_ftssel_d, }; -TRANS_FEAT(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, ftssel_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(FTSSEL, aa64_sve, gen_gvec_ool_arg_zzz, + ftssel_fns[a->esz], a, 0) /* *** SVE Predicate Logical Operations Group @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(TRN2_q, aa64_sve_f64mm, gen_gvec_ool_arg_zzz, static gen_helper_gvec_3 * const compact_fns[4] = { NULL, NULL, gen_helper_sve_compact_s, gen_helper_sve_compact_d }; -TRANS_FEAT(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, compact_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(COMPACT, aa64_sve, gen_gvec_ool_arg_zpz, + compact_fns[a->esz], a, 0) /* Call the helper that computes the ARM LastActiveElement pseudocode * function, scaled by the element size. This includes the not found @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3 * const bext_fns[4] = { gen_helper_sve2_bext_b, gen_helper_sve2_bext_h, gen_helper_sve2_bext_s, gen_helper_sve2_bext_d, }; -TRANS_FEAT(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bext_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BEXT, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bext_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bdep_fns[4] = { gen_helper_sve2_bdep_b, gen_helper_sve2_bdep_h, gen_helper_sve2_bdep_s, gen_helper_sve2_bdep_d, }; -TRANS_FEAT(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bdep_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BDEP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bdep_fns[a->esz], a, 0) static gen_helper_gvec_3 * const bgrp_fns[4] = { gen_helper_sve2_bgrp_b, gen_helper_sve2_bgrp_h, gen_helper_sve2_bgrp_s, gen_helper_sve2_bgrp_d, }; -TRANS_FEAT(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, - bgrp_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(BGRP, aa64_sve2_bitperm, gen_gvec_ool_arg_zzz, + bgrp_fns[a->esz], a, 0) static gen_helper_gvec_3 * const cadd_fns[4] = { gen_helper_sve2_cadd_b, gen_helper_sve2_cadd_h, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-8-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 24 +++++++++++++++--------- 2 files changed, 15 insertions(+), 11 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 000- ---- 0110 1--- ---- ---- # PMULLB, PMULLT (128b result) -FAIL 0110 0100 --1- ---- 1110 01-- ---- ---- # FMMLA, BFMMLA FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_trans_pmull(DisasContext *s, arg_rrr_esz *a, bool sel) gen_helper_gvec_pmull_q, gen_helper_sve2_pmull_h, NULL, gen_helper_sve2_pmull_d, }; - if (a->esz == 0 - ? !dc_isar_feature(aa64_sve2_pmull128, s) - : !dc_isar_feature(aa64_sve, s)) { + + if (a->esz == 0) { + if (!dc_isar_feature(aa64_sve2_pmull128, s)) { + return false; + } + s->is_nonstreaming = true; + } else if (!dc_isar_feature(aa64_sve, s)) { return false; } return gen_gvec_ool_arg_zzz(s, fns[a->esz], a, sel); @@ -XXX,XX +XXX,XX @@ DO_ZPZZ_FP(FMINP, aa64_sve2, sve2_fminp_zpzz) * SVE Integer Multiply-Add (unpredicated) */ -TRANS_FEAT(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_s, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) -TRANS_FEAT(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, gen_helper_fmmla_d, - a->rd, a->rn, a->rm, a->ra, 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_s, aa64_sve_f32mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_s, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FMMLA_d, aa64_sve_f64mm, gen_gvec_fpst_zzzz, + gen_helper_fmmla_d, a->rd, a->rn, a->rm, a->ra, + 0, FPST_FPCR) static gen_helper_gvec_4 * const sqdmlal_zzzw_fns[] = { NULL, gen_helper_sve2_sqdmlal_zzzw_h, @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, TRANS_FEAT(BFDOT_zzxz, aa64_sve_bf16, gen_gvec_ool_arg_zzxz, gen_helper_gvec_bfdot_idx, a) -TRANS_FEAT(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_bfmmla, a, 0) +TRANS_FEAT_NONSTREAMING(BFMMLA, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_bfmmla, a, 0) static bool do_BFMLAL_zzzw(DisasContext *s, arg_rrrr_esz *a, bool sel) { -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-9-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 15 +++++++++++---- 2 files changed, 11 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0110 0101 --0- ---- 0000 11-- ---- ---- # FTSMUL -FAIL 0110 0101 --01 0--- 100- ---- ---- ---- # FTMAD -FAIL 0110 0101 --01 1--- 001- ---- ---- ---- # FADDA FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static gen_helper_gvec_3_ptr * const ftmad_fns[4] = { NULL, gen_helper_sve_ftmad_h, gen_helper_sve_ftmad_s, gen_helper_sve_ftmad_d, }; -TRANS_FEAT(FTMAD, aa64_sve, gen_gvec_fpst_zzz, - ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, - a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) +TRANS_FEAT_NONSTREAMING(FTMAD, aa64_sve, gen_gvec_fpst_zzz, + ftmad_fns[a->esz], a->rd, a->rn, a->rm, a->imm, + a->esz == MO_16 ? FPST_FPCR_F16 : FPST_FPCR) /* *** SVE Floating Point Accumulating Reduction Group @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) if (a->esz == 0 || !dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_FADDA(DisasContext *s, arg_rprr_esz *a) DO_FP3(FADD_zzz, fadd) DO_FP3(FSUB_zzz, fsub) DO_FP3(FMUL_zzz, fmul) -DO_FP3(FTSMUL, ftsmul) DO_FP3(FRECPS, recps) DO_FP3(FRSQRTS, rsqrts) #undef DO_FP3 +static gen_helper_gvec_3_ptr * const ftsmul_fns[4] = { + NULL, gen_helper_gvec_ftsmul_h, + gen_helper_gvec_ftsmul_s, gen_helper_gvec_ftsmul_d +}; +TRANS_FEAT_NONSTREAMING(FTSMUL, aa64_sve, gen_gvec_fpst_arg_zzz, + ftsmul_fns[a->esz], a, 0) + /* *** SVE Floating Point Arithmetic - Predicated Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-10-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 12 ++++++------ 2 files changed, 6 insertions(+), 7 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --0- ---- 1001 10-- ---- ---- # SMMLA, UMMLA, USMMLA FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMLALT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, false, true) TRANS_FEAT(FMLSLB_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, false) TRANS_FEAT(FMLSLT_zzxw, aa64_sve2, do_FMLAL_zzxw, a, true, true) -TRANS_FEAT(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_smmla_b, a, 0) -TRANS_FEAT(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_usmmla_b, a, 0) -TRANS_FEAT(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, - gen_helper_gvec_ummla_b, a, 0) +TRANS_FEAT_NONSTREAMING(SMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_smmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(USMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_usmmla_b, a, 0) +TRANS_FEAT_NONSTREAMING(UMMLA, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, + gen_helper_gvec_ummla_b, a, 0) TRANS_FEAT(BFDOT_zzzz, aa64_sve_bf16, gen_gvec_ool_arg_zzzz, gen_helper_gvec_bfdot, a, 0) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-11-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 1 - target/arm/translate-sve.c | 35 ++++++++++++++++++----------------- 2 files changed, 18 insertions(+), 18 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 0100 0101 --1- ---- 1--- ---- ---- ---- # SVE2 string/histo/crypto instructions FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ DO_SVE2_ZZZ_NARROW(RSUBHNT, rsubhnt) static gen_helper_gvec_flags_4 * const match_fns[4] = { gen_helper_sve2_match_ppzz_b, gen_helper_sve2_match_ppzz_h, NULL, NULL }; -TRANS_FEAT(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(MATCH, aa64_sve2, do_ppzz_flags, a, match_fns[a->esz]) static gen_helper_gvec_flags_4 * const nmatch_fns[4] = { gen_helper_sve2_nmatch_ppzz_b, gen_helper_sve2_nmatch_ppzz_h, NULL, NULL }; -TRANS_FEAT(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) +TRANS_FEAT_NONSTREAMING(NMATCH, aa64_sve2, do_ppzz_flags, a, nmatch_fns[a->esz]) static gen_helper_gvec_4 * const histcnt_fns[4] = { NULL, NULL, gen_helper_sve2_histcnt_s, gen_helper_sve2_histcnt_d }; -TRANS_FEAT(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, - histcnt_fns[a->esz], a, 0) +TRANS_FEAT_NONSTREAMING(HISTCNT, aa64_sve2, gen_gvec_ool_arg_zpzz, + histcnt_fns[a->esz], a, 0) -TRANS_FEAT(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, - a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) +TRANS_FEAT_NONSTREAMING(HISTSEG, aa64_sve2, gen_gvec_ool_arg_zzz, + a->esz == 0 ? gen_helper_sve2_histseg : NULL, a, 0) DO_ZPZZ_FP(FADDP, aa64_sve2, sve2_faddp_zpzz) DO_ZPZZ_FP(FMAXNMP, aa64_sve2, sve2_fmaxnmp_zpzz) @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(SQRDCMLAH_zzzz, aa64_sve2, gen_gvec_ool_zzzz, TRANS_FEAT(USDOT_zzzz, aa64_sve_i8mm, gen_gvec_ool_arg_zzzz, a->esz == 2 ? gen_helper_gvec_usdot_b : NULL, a, 0) -TRANS_FEAT(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, - gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) +TRANS_FEAT_NONSTREAMING(AESMC, aa64_sve2_aes, gen_gvec_ool_zz, + gen_helper_crypto_aesmc, a->rd, a->rd, a->decrypt) -TRANS_FEAT(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, false) -TRANS_FEAT(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, - gen_helper_crypto_aese, a, true) +TRANS_FEAT_NONSTREAMING(AESE, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, false) +TRANS_FEAT_NONSTREAMING(AESD, aa64_sve2_aes, gen_gvec_ool_arg_zzz, + gen_helper_crypto_aese, a, true) -TRANS_FEAT(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4e, a, 0) -TRANS_FEAT(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, - gen_helper_crypto_sm4ekey, a, 0) +TRANS_FEAT_NONSTREAMING(SM4E, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4e, a, 0) +TRANS_FEAT_NONSTREAMING(SM4EKEY, aa64_sve2_sm4, gen_gvec_ool_arg_zzz, + gen_helper_crypto_sm4ekey, a, 0) -TRANS_FEAT(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, gen_gvec_rax1, a) +TRANS_FEAT_NONSTREAMING(RAX1, aa64_sve2_sha3, gen_gvec_fn_arg_zzz, + gen_gvec_rax1, a) TRANS_FEAT(FCVTNT_sh, aa64_sve2, gen_gvec_fpst_arg_zpz, gen_helper_sve2_fcvtnt_sh, a, 0, FPST_FPCR) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-12-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 9 --------- target/arm/translate-sve.c | 6 ++++++ 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 10-- ---- ---- ---- # SVE2 32-bit gather NT load (vector+scalar) FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) -FAIL 1000 010- -01- ---- 1--- ---- ---- ---- # SVE 32-bit gather load (vector+imm) -FAIL 1000 0100 0-0- ---- 0--- ---- ---- ---- # SVE 32-bit gather load byte (scalar+vector) -FAIL 1000 0100 1--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load half (scalar+vector) -FAIL 1000 0101 0--- ---- 0--- ---- ---- ---- # SVE 32-bit gather load word (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch -FAIL 1110 010- -00- ---- 001- ---- ---- ---- # SVE2 64-bit scatter NT store (vector+scalar) -FAIL 1110 010- -10- ---- 001- ---- ---- ---- # SVE2 32-bit scatter NT store (vector+scalar) -FAIL 1110 010- ---- ---- 1-0- ---- ---- ---- # SVE scatter store (scalar+32-bit vector) -FAIL 1110 010- ---- ---- 101- ---- ---- ---- # SVE scatter store (misc) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LD1_zpiz(DisasContext *s, arg_LD1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDNT1_zprz(DisasContext *s, arg_LD1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ST1_zpiz(DisasContext *s, arg_ST1_zpiz *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STNT1_zprz(DisasContext *s, arg_ST1_zprz *a) if (!dc_isar_feature(aa64_sve2, s)) { return false; } + s->is_nonstreaming = true; if (!sve_access_check(s)) { return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. In this case, introduce PRF_ns (prefetch non-streaming) to handle the checks. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-13-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/sve.decode | 10 +++++----- target/arm/translate-sve.c | 11 +++++++++++ 3 files changed, 16 insertions(+), 8 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1000 010- -00- ---- 111- ---- ---- ---- # SVE 32-bit gather prefetch (vector+imm) -FAIL 1000 0100 0-1- ---- 0--- ---- ---- ---- # SVE 32-bit gather prefetch (scalar+vector) FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) -FAIL 1100 010- ---- ---- ---- ---- ---- ---- # SVE 64-bit gather load/prefetch diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ LD1RO_zpri 1010010 .. 01 0.... 001 ... ..... ..... \ @rpri_load_msz nreg=0 # SVE 32-bit gather prefetch (scalar plus 32-bit scaled offsets) -PRF 1000010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1000010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 32-bit gather prefetch (vector plus immediate) -PRF 1000010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1000010 -- 00 ----- 111 --- ----- 0 ---- # SVE contiguous prefetch (scalar plus immediate) PRF 1000010 11 1- ----- 0-- --- ----- 0 ---- @@ -XXX,XX +XXX,XX @@ LD1_zpiz 1100010 .. 01 ..... 1.. ... ..... ..... \ @rpri_g_load esz=3 # SVE 64-bit gather prefetch (scalar plus 64-bit scaled offsets) -PRF 1100010 00 11 ----- 1-- --- ----- 0 ---- +PRF_ns 1100010 00 11 ----- 1-- --- ----- 0 ---- # SVE 64-bit gather prefetch (scalar plus unpacked 32-bit scaled offsets) -PRF 1100010 00 -1 ----- 0-- --- ----- 0 ---- +PRF_ns 1100010 00 -1 ----- 0-- --- ----- 0 ---- # SVE 64-bit gather prefetch (vector plus immediate) -PRF 1100010 -- 00 ----- 111 --- ----- 0 ---- +PRF_ns 1100010 -- 00 ----- 111 --- ----- 0 ---- ### SVE Memory Store Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PRF_rr(DisasContext *s, arg_PRF_rr *a) return true; } +static bool trans_PRF_ns(DisasContext *s, arg_PRF_ns *a) +{ + if (!dc_isar_feature(aa64_sve, s)) { + return false; + } + /* Prefetch is a nop within QEMU. */ + s->is_nonstreaming = true; + (void)sve_access_check(s); + return true; +} + /* * Move Prefix * -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-14-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 2 -- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) -FAIL 1010 010- ---- ---- 011- ---- ---- ---- # SVE contiguous FF load (scalar+scalar) -FAIL 1010 010- ---1 ---- 101- ---- ---- ---- # SVE contiguous NF load (scalar+imm) FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDFF1_zprr(DisasContext *s, arg_rprr_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LDNF1_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { int vsz = vec_full_reg_size(s); int elements = vsz >> dtype_esz[a->dtype]; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Mark these as a non-streaming instructions, which should trap if full a64 support is not enabled in streaming mode. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-15-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme-fa64.decode | 3 --- target/arm/translate-sve.c | 2 ++ 2 files changed, 2 insertions(+), 3 deletions(-) diff --git a/target/arm/sme-fa64.decode b/target/arm/sme-fa64.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme-fa64.decode +++ b/target/arm/sme-fa64.decode @@ -XXX,XX +XXX,XX @@ FAIL 0001 1110 0111 1110 0000 00-- ---- ---- # FJCVTZS # --11 1100 --0- ---- ---- ---- ---- ---- # Load/store FP register (unscaled imm) # --11 1100 --1- ---- ---- ---- ---- --10 # Load/store FP register (register offset) # --11 1101 ---- ---- ---- ---- ---- ---- # Load/store FP register (scaled imm) - -FAIL 1010 010- -01- ---- 000- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+scalar) -FAIL 1010 010- -010 ---- 001- ---- ---- ---- # SVE load & replicate 32 bytes (scalar+imm) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zprr(DisasContext *s, arg_rprr_load *a) if (a->rm == 31) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), dtype_msz(a->dtype)); @@ -XXX,XX +XXX,XX @@ static bool trans_LD1RO_zpri(DisasContext *s, arg_rpri_load *a) if (!dc_isar_feature(aa64_sve_f64mm, s)) { return false; } + s->is_nonstreaming = true; if (sve_access_check(s)) { TCGv_i64 addr = new_tmp_a64(s); tcg_gen_addi_i64(addr, cpu_reg_sp(s, a->rn), a->imm * 32); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These functions will be used to verify that the cpu is in the correct state for a given instruction. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-16-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 21 +++++++++++++++++++++ target/arm/translate-a64.c | 34 ++++++++++++++++++++++++++++++++++ 2 files changed, 55 insertions(+) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void write_fp_dreg(DisasContext *s, int reg, TCGv_i64 v); bool logic_imm_decode_wmask(uint64_t *result, unsigned int immn, unsigned int imms, unsigned int immr); bool sve_access_check(DisasContext *s); +bool sme_enabled_check(DisasContext *s); +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned); + +/* This function corresponds to CheckStreamingSVEEnabled. */ +static inline bool sme_sm_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK); +} + +/* This function corresponds to CheckSMEAndZAEnabled. */ +static inline bool sme_za_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_ZA_MASK); +} + +/* Note that this function corresponds to CheckStreamingSVEAndZAEnabled. */ +static inline bool sme_smza_enabled_check(DisasContext *s) +{ + return sme_enabled_check_with_svcr(s, R_SVCR_SM_MASK | R_SVCR_ZA_MASK); +} + TCGv_i64 clean_data_tbi(DisasContext *s, TCGv_i64 addr); TCGv_i64 gen_mte_check1(DisasContext *s, TCGv_i64 addr, bool is_write, bool tag_checked, int log2_size); diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool sme_access_check(DisasContext *s) return true; } +/* This function corresponds to CheckSMEEnabled. */ +bool sme_enabled_check(DisasContext *s) +{ + /* + * Note that unlike sve_excp_el, we have not constrained sme_excp_el + * to be zero when fp_excp_el has priority. This is because we need + * sme_excp_el by itself for cpregs access checks. + */ + if (!s->fp_excp_el || s->sme_excp_el < s->fp_excp_el) { + s->fp_access_checked = true; + return sme_access_check(s); + } + return fp_access_check_only(s); +} + +/* Common subroutine for CheckSMEAnd*Enabled. */ +bool sme_enabled_check_with_svcr(DisasContext *s, unsigned req) +{ + if (!sme_enabled_check(s)) { + return false; + } + if (FIELD_EX64(req, SVCR, SM) && !s->pstate_sm) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_NotStreaming, false)); + return false; + } + if (FIELD_EX64(req, SVCR, ZA) && !s->pstate_za) { + gen_exception_insn(s, s->pc_curr, EXCP_UDEF, + syn_smetrap(SME_ET_InactiveZA, false)); + return false; + } + return true; +} + /* * This utility function is for doing register extension with an * optional shift. You will likely want to pass a temporary for the -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> The pseudocode for CheckSVEEnabled gains a check for Streaming SVE mode, and for SME present but SVE absent. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-17-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.c | 22 ++++++++++++++++------ 1 file changed, 16 insertions(+), 6 deletions(-) diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.c +++ b/target/arm/translate-a64.c @@ -XXX,XX +XXX,XX @@ static bool fp_access_check(DisasContext *s) return true; } -/* Check that SVE access is enabled. If it is, return true. +/* + * Check that SVE access is enabled. If it is, return true. * If not, emit code to generate an appropriate exception and return false. + * This function corresponds to CheckSVEEnabled(). */ bool sve_access_check(DisasContext *s) { - if (s->sve_excp_el) { - assert(!s->sve_access_checked); - s->sve_access_checked = true; - + if (s->pstate_sm || !dc_isar_feature(aa64_sve, s)) { + assert(dc_isar_feature(aa64_sme, s)); + if (!sme_sm_enabled_check(s)) { + goto fail_exit; + } + } else if (s->sve_excp_el) { gen_exception_insn_el(s, s->pc_curr, EXCP_UDEF, syn_sve_access_trap(), s->sve_excp_el); - return false; + goto fail_exit; } s->sve_access_checked = true; return fp_access_check(s); + + fail_exit: + /* Assert that we only raise one exception per instruction. */ + assert(!s->sve_access_checked); + s->sve_access_checked = true; + return false; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These SME instructions are nominally within the SVE decode space, so we add them to sve.decode and translate-sve.c. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-18-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 12 ++++++++++++ target/arm/sve.decode | 5 ++++- target/arm/translate-sve.c | 38 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 54 insertions(+), 1 deletion(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int vec_full_reg_size(DisasContext *s) return s->vl; } +/* Return the byte size of the vector register, SVL / 8. */ +static inline int streaming_vec_reg_size(DisasContext *s) +{ + return s->svl; +} + /* * Return the offset info CPUARMState of the predicate vector register Pn. * Note for this purpose, FFR is P16. @@ -XXX,XX +XXX,XX @@ static inline int pred_full_reg_size(DisasContext *s) return s->vl >> 3; } +/* Return the byte size of the predicate register, SVL / 64. */ +static inline int streaming_pred_reg_size(DisasContext *s) +{ + return s->svl >> 3; +} + /* * Round up the size of a register to a size allowed by * the tcg vector infrastructure. Any operation which uses this diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ INDEX_ri 00000100 esz:2 1 imm:s5 010001 rn:5 rd:5 # SVE index generation (register start, register increment) INDEX_rr 00000100 .. 1 ..... 010011 ..... ..... @rd_rn_rm -### SVE Stack Allocation Group +### SVE / Streaming SVE Stack Allocation Group # SVE stack frame adjustment ADDVL 00000100 001 ..... 01010 ...... ..... @rd_rn_i6 +ADDSVL 00000100 001 ..... 01011 ...... ..... @rd_rn_i6 ADDPL 00000100 011 ..... 01010 ...... ..... @rd_rn_i6 +ADDSPL 00000100 011 ..... 01011 ...... ..... @rd_rn_i6 # SVE stack frame size RDVL 00000100 101 11111 01010 imm:s6 rd:5 +RDSVL 00000100 101 11111 01011 imm:s6 rd:5 ### SVE Bitwise Shift - Unpredicated Group diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_ADDVL(DisasContext *s, arg_ADDVL *a) return true; } +static bool trans_ADDSVL(DisasContext *s, arg_ADDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ADDPL(DisasContext *s, arg_ADDPL *a) return true; } +static bool trans_ADDSPL(DisasContext *s, arg_ADDSPL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 rd = cpu_reg_sp(s, a->rd); + TCGv_i64 rn = cpu_reg_sp(s, a->rn); + tcg_gen_addi_i64(rd, rn, a->imm * streaming_pred_reg_size(s)); + } + return true; +} + static bool trans_RDVL(DisasContext *s, arg_RDVL *a) { if (!dc_isar_feature(aa64_sve, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_RDVL(DisasContext *s, arg_RDVL *a) return true; } +static bool trans_RDSVL(DisasContext *s, arg_RDSVL *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_enabled_check(s)) { + TCGv_i64 reg = cpu_reg(s, a->rd); + tcg_gen_movi_i64(reg, a->imm * streaming_vec_reg_size(s)); + } + return true; +} + /* *** SVE Compute Vector Address Group */ -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-19-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 4 ++++ target/arm/sme_helper.c | 25 +++++++++++++++++++++++++ target/arm/translate-sme.c | 13 +++++++++++++ 4 files changed, 44 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) + +DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ # # This file is processed by scripts/decodetree.py # + +### SME Misc + +ZERO 11000000 00 001 00000000000 imm:8 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void helper_set_pstate_za(CPUARMState *env, uint32_t i) memset(env->zarray, 0, sizeof(env->zarray)); } } + +void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) +{ + uint32_t i; + + /* + * Special case clearing the entire ZA space. + * This falls into the CONSTRAINED UNPREDICTABLE zeroing of any + * parts of the ZA storage outside of SVL. + */ + if (imm == 0xff) { + memset(env->zarray, 0, sizeof(env->zarray)); + return; + } + + /* + * Recall that ZAnH.D[m] is spread across ZA[n+8*m], + * so each row is discontiguous within ZA[]. + */ + for (i = 0; i < svl; i++) { + if (imm & (1 << (i % 8))) { + memset(&env->zarray[i], 0, svl); + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ */ #include "decode-sme.c.inc" + + +static bool trans_ZERO(DisasContext *s, arg_ZERO *a) +{ + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (sme_za_enabled_check(s)) { + gen_helper_sme_zero(cpu_env, tcg_constant_i32(a->imm), + tcg_constant_i32(streaming_vec_reg_size(s))); + } + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for implementing moves to/from horizontal tile slices, but we need new ones for moves to/from vertical tile slices. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-20-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 12 +++ target/arm/helper-sve.h | 2 + target/arm/translate-a64.h | 8 ++ target/arm/translate.h | 5 ++ target/arm/sme.decode | 15 ++++ target/arm/sme_helper.c | 151 ++++++++++++++++++++++++++++++++++++- target/arm/sve_helper.c | 12 +++ target/arm/translate-sme.c | 127 +++++++++++++++++++++++++++++++ 8 files changed, 331 insertions(+), 1 deletion(-) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_2(set_pstate_sm, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_2(set_pstate_za, TCG_CALL_NO_RWG, void, env, i32) DEF_HELPER_FLAGS_3(sme_zero, TCG_CALL_NO_RWG, void, env, i32, i32) + +/* Move to/from vertical array slices, i.e. columns, so 'c'. */ +DEF_HELPER_FLAGS_4(sme_mova_cz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sve_sel_zpzz_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve_sel_zpzz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sve_sel_zpzz_q, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sve2_addp_zpzz_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ static inline int pred_gvec_reg_size(DisasContext *s) return size_for_gvec(pred_full_reg_size(s)); } +/* Return a newly allocated pointer to the predicate register. */ +static inline TCGv_ptr pred_full_reg_ptr(DisasContext *s, int regno) +{ + TCGv_ptr ret = tcg_temp_new_ptr(); + tcg_gen_addi_ptr(ret, cpu_env, pred_full_reg_offset(s, regno)); + return ret; +} + bool disas_sve(DisasContext *, uint32_t); bool disas_sme(DisasContext *, uint32_t); diff --git a/target/arm/translate.h b/target/arm/translate.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate.h +++ b/target/arm/translate.h @@ -XXX,XX +XXX,XX @@ static inline int plus_2(DisasContext *s, int x) return x + 2; } +static inline int plus_12(DisasContext *s, int x) +{ + return x + 12; +} + static inline int times_2(DisasContext *s, int x) { return x * 2; diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ### SME Misc ZERO 11000000 00 001 00000000000 imm:8 + +### SME Move into/from Array + +%mova_rs 13:2 !function=plus_12 +&mova esz rs pg zr za_imm v:bool to_vec:bool + +MOVA 11000000 esz:2 00000 0 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs +MOVA 11000000 11 00000 1 v:1 .. pg:3 zr:5 0 za_imm:4 \ + &mova to_vec=0 rs=%mova_rs esz=4 + +MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs +MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ + &mova to_vec=1 rs=%mova_rs esz=4 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" -#include "internals.h" +#include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "qemu/int128.h" +#include "vec_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void helper_sme_zero(CPUARMState *env, uint32_t imm, uint32_t svl) } } } + + +/* + * When considering the ZA storage as an array of elements of + * type T, the index within that array of the Nth element of + * a vertical slice of a tile can be calculated like this, + * regardless of the size of type T. This is because the tiles + * are interleaved, so if type T is size N bytes then row 1 of + * the tile is N rows away from row 0. The division by N to + * convert a byte offset into an array index and the multiplication + * by N to convert from vslice-index-within-the-tile to + * the index within the ZA storage cancel out. + */ +#define tile_vslice_index(i) ((i) * sizeof(ARMVectorReg)) + +/* + * When doing byte arithmetic on the ZA storage, the element + * byteoff bytes away in a tile vertical slice is always this + * many bytes away in the ZA storage, regardless of the + * size of the tile element, assuming that byteoff is a multiple + * of the element size. Again this is because of the interleaving + * of the tiles. For instance if we have 1 byte per element then + * each row of the ZA storage has one byte of the vslice data, + * and (counting from 0) byte 8 goes in row 8 of the storage + * at offset (8 * row-size-in-bytes). + * If we have 8 bytes per element then each row of the ZA storage + * has 8 bytes of the data, but there are 8 interleaved tiles and + * so byte 8 of the data goes into row 1 of the tile, + * which is again row 8 of the storage, so the offset is still + * (8 * row-size-in-bytes). Similarly for other element sizes. + */ +#define tile_vslice_offset(byteoff) ((byteoff) * sizeof(ARMVectorReg)) + + +/* + * Move Zreg vector to ZArray column. + */ +#define DO_MOVA_C(NAME, TYPE, H) \ +void HELPER(NAME)(void *za, void *vn, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(za + tile_vslice_offset(i)) = *(TYPE *)(vn + H(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_C(sme_mova_cz_b, uint8_t, H1) +DO_MOVA_C(sme_mova_cz_h, uint16_t, H1_2) +DO_MOVA_C(sme_mova_cz_s, uint32_t, H1_4) + +void HELPER(sme_mova_cz_d)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *n = vn; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +void HELPER(sme_mova_cz_q)(void *za, void *vn, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *n = vn; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++) { + if (pg[H2(i)] & 1) { + a[tile_vslice_index(i)] = n[i]; + } + } +} + +#undef DO_MOVA_C + +/* + * Move ZArray column to Zreg vector. + */ +#define DO_MOVA_Z(NAME, TYPE, H) \ +void HELPER(NAME)(void *vd, void *za, void *vg, uint32_t desc) \ +{ \ + int i, oprsz = simd_oprsz(desc); \ + for (i = 0; i < oprsz; ) { \ + uint16_t pg = *(uint16_t *)(vg + H1_2(i >> 3)); \ + do { \ + if (pg & 1) { \ + *(TYPE *)(vd + H(i)) = *(TYPE *)(za + tile_vslice_offset(i)); \ + } \ + i += sizeof(TYPE); \ + pg >>= sizeof(TYPE); \ + } while (i & 15); \ + } \ +} + +DO_MOVA_Z(sme_mova_zc_b, uint8_t, H1) +DO_MOVA_Z(sme_mova_zc_h, uint16_t, H1_2) +DO_MOVA_Z(sme_mova_zc_s, uint32_t, H1_4) + +void HELPER(sme_mova_zc_d)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 8; + uint8_t *pg = vg; + uint64_t *d = vd; + uint64_t *a = za; + + for (i = 0; i < oprsz; i++) { + if (pg[H1(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) +{ + int i, oprsz = simd_oprsz(desc) / 16; + uint16_t *pg = vg; + Int128 *d = vd; + Int128 *a = za; + + /* + * Int128 is used here simply to copy 16 bytes, and to simplify + * the address arithmetic. + */ + for (i = 0; i < oprsz; i++, za += sizeof(ARMVectorReg)) { + if (pg[H2(i)] & 1) { + d[i] = a[tile_vslice_index(i)]; + } + } +} + +#undef DO_MOVA_Z diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sve_sel_zpzz_d)(void *vd, void *vn, void *vm, } } +void HELPER(sve_sel_zpzz_q)(void *vd, void *vn, void *vm, + void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 16; + Int128 *d = vd, *n = vn, *m = vm; + uint16_t *pg = vg; + + for (i = 0; i < opr_sz; i += 1) { + d[i] = (pg[H2(i)] & 1 ? n : m)[i]; + } +} + /* Two operand comparison controlled by a predicate. * ??? It is very tempting to want to be able to expand this inline * with x86 instructions, e.g. diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ #include "decode-sme.c.inc" +/* + * Resolve tile.size[index] to a host pointer, where tile and index + * are always decoded together, dependent on the element size. + */ +static TCGv_ptr get_tile_rowcol(DisasContext *s, int esz, int rs, + int tile_index, bool vertical) +{ + int tile = tile_index >> (4 - esz); + int index = esz == MO_128 ? 0 : extract32(tile_index, 0, 4 - esz); + int pos, len, offset; + TCGv_i32 tmp; + TCGv_ptr addr; + + /* Compute the final index, which is Rs+imm. */ + tmp = tcg_temp_new_i32(); + tcg_gen_trunc_tl_i32(tmp, cpu_reg(s, rs)); + tcg_gen_addi_i32(tmp, tmp, index); + + /* Prepare a power-of-two modulo via extraction of @len bits. */ + len = ctz32(streaming_vec_reg_size(s)) - esz; + + if (vertical) { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * size + * = (index % (svl >> esz)) << esz + * Perform the power-of-two modulo via extraction of the low @len bits. + * Perform the multiply by shifting left by @pos bits. + * Perform these operations simultaneously via deposit into zero. + */ + pos = esz; + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* + * For big-endian, adjust the indexed column byte offset within + * the uint64_t host words that make up env->zarray[]. + */ + if (HOST_BIG_ENDIAN && esz < MO_64) { + tcg_gen_xori_i32(tmp, tmp, 8 - (1 << esz)); + } + } else { + /* + * Compute the byte offset of the index within the tile: + * (index % (svl / size)) * (size * sizeof(row)) + * = (index % (svl >> esz)) << (esz + log2(sizeof(row))) + */ + pos = esz + ctz32(sizeof(ARMVectorReg)); + tcg_gen_deposit_z_i32(tmp, tmp, pos, len); + + /* Row slices are always aligned and need no endian adjustment. */ + } + + /* The tile byte offset within env->zarray is the row. */ + offset = tile * sizeof(ARMVectorReg); + + /* Include the byte offset of zarray to make this relative to env. */ + offset += offsetof(CPUARMState, zarray); + tcg_gen_addi_i32(tmp, tmp, offset); + + /* Add the byte offset to env to produce the final pointer. */ + addr = tcg_temp_new_ptr(); + tcg_gen_ext_i32_ptr(addr, tmp); + tcg_temp_free_i32(tmp); + tcg_gen_add_ptr(addr, addr, cpu_env); + + return addr; +} + static bool trans_ZERO(DisasContext *s, arg_ZERO *a) { if (!dc_isar_feature(aa64_sme, s)) { @@ -XXX,XX +XXX,XX @@ static bool trans_ZERO(DisasContext *s, arg_ZERO *a) } return true; } + +static bool trans_MOVA(DisasContext *s, arg_MOVA *a) +{ + static gen_helper_gvec_4 * const h_fns[5] = { + gen_helper_sve_sel_zpzz_b, gen_helper_sve_sel_zpzz_h, + gen_helper_sve_sel_zpzz_s, gen_helper_sve_sel_zpzz_d, + gen_helper_sve_sel_zpzz_q + }; + static gen_helper_gvec_3 * const cz_fns[5] = { + gen_helper_sme_mova_cz_b, gen_helper_sme_mova_cz_h, + gen_helper_sme_mova_cz_s, gen_helper_sme_mova_cz_d, + gen_helper_sme_mova_cz_q, + }; + static gen_helper_gvec_3 * const zc_fns[5] = { + gen_helper_sme_mova_zc_b, gen_helper_sme_mova_zc_h, + gen_helper_sme_mova_zc_s, gen_helper_sme_mova_zc_d, + gen_helper_sme_mova_zc_q, + }; + + TCGv_ptr t_za, t_zr, t_pg; + TCGv_i32 t_desc; + int svl; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_zr = vec_full_reg_ptr(s, a->zr); + t_pg = pred_full_reg_ptr(s, a->pg); + + svl = streaming_vec_reg_size(s); + t_desc = tcg_constant_i32(simd_desc(svl, svl, 0)); + + if (a->v) { + /* Vertical slice -- use sme mova helpers. */ + if (a->to_vec) { + zc_fns[a->esz](t_zr, t_za, t_pg, t_desc); + } else { + cz_fns[a->esz](t_za, t_zr, t_pg, t_desc); + } + } else { + /* Horizontal slice -- reuse sve sel helpers. */ + if (a->to_vec) { + h_fns[a->esz](t_zr, t_za, t_zr, t_pg, t_desc); + } else { + h_fns[a->esz](t_za, t_zr, t_za, t_pg, t_desc); + } + } + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_zr); + tcg_temp_free_ptr(t_pg); + + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We cannot reuse the SVE functions for LD[1-4] and ST[1-4], because those functions accept only a Zreg register number. For SME, we want to pass a pointer into ZA storage. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-21-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 82 +++++ target/arm/sme.decode | 9 + target/arm/sme_helper.c | 595 +++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 70 +++++ 4 files changed, 756 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sme_mova_cz_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_cz_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sme_mova_zc_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(sme_ld1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_ld1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_ld1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1b_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1b_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1h_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1h_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1s_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1s_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1d_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1d_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_st1q_be_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) +DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ MOVA 11000000 esz:2 00001 0 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs MOVA 11000000 11 00001 1 v:1 .. pg:3 0 za_imm:4 zr:5 \ &mova to_vec=1 rs=%mova_rs esz=4 + +### SME Memory + +&ldst esz rs pg rn rm za_imm v:bool st:bool + +LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst rs=%mova_rs +LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ + &ldst esz=4 rs=%mova_rs diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "cpu.h" +#include "internals.h" #include "tcg/tcg-gvec-desc.h" #include "exec/helper-proto.h" +#include "exec/cpu_ldst.h" +#include "exec/exec-all.h" #include "qemu/int128.h" #include "vec_internal.h" +#include "sve_ldst_internal.h" /* ResetSVEState */ void arm_reset_sve_state(CPUARMState *env) @@ -XXX,XX +XXX,XX @@ void HELPER(sme_mova_zc_q)(void *vd, void *za, void *vg, uint32_t desc) } #undef DO_MOVA_Z + +/* + * Clear elements in a tile slice comprising len bytes. + */ + +typedef void ClearFn(void *ptr, size_t off, size_t len); + +static void clear_horizontal(void *ptr, size_t off, size_t len) +{ + memset(ptr + off, 0, len); +} + +static void clear_vertical_b(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; ++i) { + *(uint8_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_h(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 2) { + *(uint16_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_s(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 4) { + *(uint32_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_d(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 8) { + *(uint64_t *)(vptr + tile_vslice_offset(i + off)) = 0; + } +} + +static void clear_vertical_q(void *vptr, size_t off, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memset(vptr + tile_vslice_offset(i + off), 0, 16); + } +} + +/* + * Copy elements from an array into a tile slice comprising len bytes. + */ + +typedef void CopyFn(void *dst, const void *src, size_t len); + +static void copy_horizontal(void *dst, const void *src, size_t len) +{ + memcpy(dst, src, len); +} + +static void copy_vertical_b(void *vdst, const void *vsrc, size_t len) +{ + const uint8_t *src = vsrc; + uint8_t *dst = vdst; + size_t i; + + for (i = 0; i < len; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_h(void *vdst, const void *vsrc, size_t len) +{ + const uint16_t *src = vsrc; + uint16_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 2; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_s(void *vdst, const void *vsrc, size_t len) +{ + const uint32_t *src = vsrc; + uint32_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 4; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_d(void *vdst, const void *vsrc, size_t len) +{ + const uint64_t *src = vsrc; + uint64_t *dst = vdst; + size_t i; + + for (i = 0; i < len / 8; ++i) { + dst[tile_vslice_index(i)] = src[i]; + } +} + +static void copy_vertical_q(void *vdst, const void *vsrc, size_t len) +{ + for (size_t i = 0; i < len; i += 16) { + memcpy(vdst + tile_vslice_offset(i), vsrc + i, 16); + } +} + +/* + * Host and TLB primitives for vertical tile slice addressing. + */ + +#define DO_LD(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = HOST(host); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = TLB(env, useronly_clean_ptr(addr), ra); \ + *(TYPE *)(za + tile_vslice_offset(off)) = val; \ +} + +#define DO_ST(NAME, TYPE, HOST, TLB) \ +static inline void sme_##NAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + HOST(host, val); \ +} \ +static inline void sme_##NAME##_v_tlb(CPUARMState *env, void *za, \ + intptr_t off, target_ulong addr, uintptr_t ra) \ +{ \ + TYPE val = *(TYPE *)(za + tile_vslice_offset(off)); \ + TLB(env, useronly_clean_ptr(addr), val, ra); \ +} + +/* + * The ARMVectorReg elements are stored in host-endian 64-bit units. + * For 128-bit quantities, the sequence defined by the Elem[] pseudocode + * corresponds to storing the two 64-bit pieces in little-endian order. + */ +#define DO_LDQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t val0 = HOST(host), val1 = HOST(host + 8); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t val0 = TLB(env, useronly_clean_ptr(addr), ra); \ + uint64_t val1 = TLB(env, useronly_clean_ptr(addr + 8), ra); \ + uint64_t *ptr = za + off; \ + ptr[0] = BE ? val1 : val0, ptr[1] = BE ? val0 : val1; \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +#define DO_STQ(HNAME, VNAME, BE, HOST, TLB) \ +static inline void HNAME##_host(void *za, intptr_t off, void *host) \ +{ \ + uint64_t *ptr = za + off; \ + HOST(host, ptr[BE]); \ + HOST(host + 1, ptr[!BE]); \ +} \ +static inline void VNAME##_v_host(void *za, intptr_t off, void *host) \ +{ \ + HNAME##_host(za, tile_vslice_offset(off), host); \ +} \ +static inline void HNAME##_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + uint64_t *ptr = za + off; \ + TLB(env, useronly_clean_ptr(addr), ptr[BE], ra); \ + TLB(env, useronly_clean_ptr(addr + 8), ptr[!BE], ra); \ +} \ +static inline void VNAME##_v_tlb(CPUARMState *env, void *za, intptr_t off, \ + target_ulong addr, uintptr_t ra) \ +{ \ + HNAME##_tlb(env, za, tile_vslice_offset(off), addr, ra); \ +} + +DO_LD(ld1b, uint8_t, ldub_p, cpu_ldub_data_ra) +DO_LD(ld1h_be, uint16_t, lduw_be_p, cpu_lduw_be_data_ra) +DO_LD(ld1h_le, uint16_t, lduw_le_p, cpu_lduw_le_data_ra) +DO_LD(ld1s_be, uint32_t, ldl_be_p, cpu_ldl_be_data_ra) +DO_LD(ld1s_le, uint32_t, ldl_le_p, cpu_ldl_le_data_ra) +DO_LD(ld1d_be, uint64_t, ldq_be_p, cpu_ldq_be_data_ra) +DO_LD(ld1d_le, uint64_t, ldq_le_p, cpu_ldq_le_data_ra) + +DO_LDQ(sve_ld1qq_be, sme_ld1q_be, 1, ldq_be_p, cpu_ldq_be_data_ra) +DO_LDQ(sve_ld1qq_le, sme_ld1q_le, 0, ldq_le_p, cpu_ldq_le_data_ra) + +DO_ST(st1b, uint8_t, stb_p, cpu_stb_data_ra) +DO_ST(st1h_be, uint16_t, stw_be_p, cpu_stw_be_data_ra) +DO_ST(st1h_le, uint16_t, stw_le_p, cpu_stw_le_data_ra) +DO_ST(st1s_be, uint32_t, stl_be_p, cpu_stl_be_data_ra) +DO_ST(st1s_le, uint32_t, stl_le_p, cpu_stl_le_data_ra) +DO_ST(st1d_be, uint64_t, stq_be_p, cpu_stq_be_data_ra) +DO_ST(st1d_le, uint64_t, stq_le_p, cpu_stq_le_data_ra) + +DO_STQ(sve_st1qq_be, sme_st1q_be, 1, stq_be_p, cpu_stq_be_data_ra) +DO_STQ(sve_st1qq_le, sme_st1q_le, 0, stq_le_p, cpu_stq_le_data_ra) + +#undef DO_LD +#undef DO_ST +#undef DO_LDQ +#undef DO_STQ + +/* + * Common helper for all contiguous predicated loads. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_ld1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no load occurs. */ + clr_fn(za, 0, reg_max); + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_LOAD, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_READ, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. Perform the load + * into scratch memory to preserve register state until the end. + */ + ARMVectorReg scratch = { }; + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, &scratch, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + + cpy_fn(za, &scratch, reg_max); + return; +#endif + } + + /* The entire operation is in RAM, on valid pages. */ + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + if (!vertical) { + memset(za, 0, reg_max); + } else if (reg_off) { + clr_fn(za, 0, reg_off); + } + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } else if (vertical) { + clr_fn(za, reg_off, esize); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_ld1_mte(CPUARMState *env, void *za, uint64_t *vg, + target_ulong addr, uint32_t desc, uintptr_t ra, + const int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn, + ClearFn *clr_fn, + CopyFn *cpy_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_ld1(env, za, vg, addr, desc, ra, esz, mtedesc, vertical, + host_fn, tlb_fn, clr_fn, cpy_fn); +} + +#define DO_LD(L, END, ESZ) \ +void HELPER(sme_ld1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} \ +void HELPER(sme_ld1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_ld1##L##L##END##_host, sve_ld1##L##L##END##_tlb, \ + clear_horizontal, copy_horizontal); \ +} \ +void HELPER(sme_ld1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_ld1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_ld1##L##END##_v_host, sme_ld1##L##END##_v_tlb, \ + clear_vertical_##L, copy_vertical_##L); \ +} + +DO_LD(b, , MO_8) +DO_LD(h, _be, MO_16) +DO_LD(h, _le, MO_16) +DO_LD(s, _be, MO_32) +DO_LD(s, _le, MO_32) +DO_LD(d, _be, MO_64) +DO_LD(d, _le, MO_64) +DO_LD(q, _be, MO_128) +DO_LD(q, _le, MO_128) + +#undef DO_LD + +/* + * Common helper for all contiguous predicated stores. + */ + +static inline QEMU_ALWAYS_INLINE +void sme_st1(CPUARMState *env, void *za, uint64_t *vg, + const target_ulong addr, uint32_t desc, const uintptr_t ra, + const int esz, uint32_t mtedesc, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + const intptr_t reg_max = simd_oprsz(desc); + const intptr_t esize = 1 << esz; + intptr_t reg_off, reg_last; + SVEContLdSt info; + void *host; + int flags; + + /* Find the active elements. */ + if (!sve_cont_ldst_elements(&info, addr, vg, reg_max, esz, esize)) { + /* The entire predicate was false; no store occurs. */ + return; + } + + /* Probe the page(s). Exit with exception for any invalid page. */ + sve_cont_ldst_pages(&info, FAULT_ALL, env, addr, MMU_DATA_STORE, ra); + + /* Handle watchpoints for all active elements. */ + sve_cont_ldst_watchpoints(&info, env, vg, addr, esize, esize, + BP_MEM_WRITE, ra); + + /* + * Handle mte checks for all active elements. + * Since TBI must be set for MTE, !mtedesc => !mte_active. + */ + if (mtedesc) { + sve_cont_ldst_mte_check(&info, env, vg, addr, esize, esize, + mtedesc, ra); + } + + flags = info.page[0].flags | info.page[1].flags; + if (unlikely(flags != 0)) { +#ifdef CONFIG_USER_ONLY + g_assert_not_reached(); +#else + /* + * At least one page includes MMIO. + * Any bus operation can fail with cpu_transaction_failed, + * which for ARM will raise SyncExternal. We cannot avoid + * this fault and will leave with the store incomplete. + */ + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[1]; + if (reg_last < 0) { + reg_last = info.reg_off_split; + if (reg_last < 0) { + reg_last = info.reg_off_last[0]; + } + } + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + reg_off += esize; + } while (reg_off & 63); + } while (reg_off <= reg_last); + return; +#endif + } + + reg_off = info.reg_off_first[0]; + reg_last = info.reg_off_last[0]; + host = info.page[0].host; + + while (reg_off <= reg_last) { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off <= reg_last && (reg_off & 63)); + } + + /* + * Use the slow path to manage the cross-page misalignment. + * But we know this is RAM and cannot trap. + */ + reg_off = info.reg_off_split; + if (unlikely(reg_off >= 0)) { + tlb_fn(env, za, reg_off, addr + reg_off, ra); + } + + reg_off = info.reg_off_first[1]; + if (unlikely(reg_off >= 0)) { + reg_last = info.reg_off_last[1]; + host = info.page[1].host; + + do { + uint64_t pg = vg[reg_off >> 6]; + do { + if ((pg >> (reg_off & 63)) & 1) { + host_fn(za, reg_off, host + reg_off); + } + reg_off += 1 << esz; + } while (reg_off & 63); + } while (reg_off <= reg_last); + } +} + +static inline QEMU_ALWAYS_INLINE +void sme_st1_mte(CPUARMState *env, void *za, uint64_t *vg, target_ulong addr, + uint32_t desc, uintptr_t ra, int esz, bool vertical, + sve_ldst1_host_fn *host_fn, + sve_ldst1_tlb_fn *tlb_fn) +{ + uint32_t mtedesc = desc >> (SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + int bit55 = extract64(addr, 55, 1); + + /* Remove mtedesc from the normal sve descriptor. */ + desc = extract32(desc, 0, SIMD_DATA_SHIFT + SVE_MTEDESC_SHIFT); + + /* Perform gross MTE suppression early. */ + if (!tbi_check(desc, bit55) || + tcma_check(desc, bit55, allocation_tag_from_addr(addr))) { + mtedesc = 0; + } + + sme_st1(env, za, vg, addr, desc, ra, esz, mtedesc, + vertical, host_fn, tlb_fn); +} + +#define DO_ST(L, END, ESZ) \ +void HELPER(sme_st1##L##END##_h)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1(env, za, vg, addr, desc, GETPC(), ESZ, 0, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} \ +void HELPER(sme_st1##L##END##_h_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, false, \ + sve_st1##L##L##END##_host, sve_st1##L##L##END##_tlb); \ +} \ +void HELPER(sme_st1##L##END##_v_mte)(CPUARMState *env, void *za, void *vg, \ + target_ulong addr, uint32_t desc) \ +{ \ + sme_st1_mte(env, za, vg, addr, desc, GETPC(), ESZ, true, \ + sme_st1##L##END##_v_host, sme_st1##L##END##_v_tlb); \ +} + +DO_ST(b, , MO_8) +DO_ST(h, _be, MO_16) +DO_ST(h, _le, MO_16) +DO_ST(s, _be, MO_32) +DO_ST(s, _le, MO_32) +DO_ST(d, _be, MO_64) +DO_ST(d, _le, MO_64) +DO_ST(q, _be, MO_128) +DO_ST(q, _le, MO_128) + +#undef DO_ST diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_MOVA(DisasContext *s, arg_MOVA *a) return true; } + +static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) +{ + typedef void GenLdSt1(TCGv_env, TCGv_ptr, TCGv_ptr, TCGv, TCGv_i32); + + /* + * Indexed by [esz][be][v][mte][st], which is (except for load/store) + * also the order in which the elements appear in the function names, + * and so how we must concatenate the pieces. + */ + +#define FN_LS(F) { gen_helper_sme_ld1##F, gen_helper_sme_st1##F } +#define FN_MTE(F) { FN_LS(F), FN_LS(F##_mte) } +#define FN_HV(F) { FN_MTE(F##_h), FN_MTE(F##_v) } +#define FN_END(L, B) { FN_HV(L), FN_HV(B) } + + static GenLdSt1 * const fns[5][2][2][2][2] = { + FN_END(b, b), + FN_END(h_le, h_be), + FN_END(s_le, s_be), + FN_END(d_le, d_be), + FN_END(q_le, q_be), + }; + +#undef FN_LS +#undef FN_MTE +#undef FN_HV +#undef FN_END + + TCGv_ptr t_za, t_pg; + TCGv_i64 addr; + int svl, desc = 0; + bool be = s->be_data == MO_BE; + bool mte = s->mte_active[0]; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sme_smza_enabled_check(s)) { + return true; + } + + t_za = get_tile_rowcol(s, a->esz, a->rs, a->za_imm, a->v); + t_pg = pred_full_reg_ptr(s, a->pg); + addr = tcg_temp_new_i64(); + + tcg_gen_shli_i64(addr, cpu_reg(s, a->rm), a->esz); + tcg_gen_add_i64(addr, addr, cpu_reg_sp(s, a->rn)); + + if (mte) { + desc = FIELD_DP32(desc, MTEDESC, MIDX, get_mem_index(s)); + desc = FIELD_DP32(desc, MTEDESC, TBI, s->tbid); + desc = FIELD_DP32(desc, MTEDESC, TCMA, s->tcma); + desc = FIELD_DP32(desc, MTEDESC, WRITE, a->st); + desc = FIELD_DP32(desc, MTEDESC, SIZEM1, (1 << a->esz) - 1); + desc <<= SVE_MTEDESC_SHIFT; + } else { + addr = clean_data_tbi(s, addr); + } + svl = streaming_vec_reg_size(s); + desc = simd_desc(svl, svl, desc); + + fns[a->esz][be][a->v][mte][a->st](cpu_env, t_za, t_pg, addr, + tcg_constant_i32(desc)); + + tcg_temp_free_ptr(t_za); + tcg_temp_free_ptr(t_pg); + tcg_temp_free_i64(addr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add a TCGv_ptr base argument, which will be cpu_env for SVE. We will reuse this for SME save and restore array insns. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-22-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/translate-a64.h | 3 +++ target/arm/translate-sve.c | 48 ++++++++++++++++++++++++++++---------- 2 files changed, 39 insertions(+), 12 deletions(-) diff --git a/target/arm/translate-a64.h b/target/arm/translate-a64.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-a64.h +++ b/target/arm/translate-a64.h @@ -XXX,XX +XXX,XX @@ void gen_gvec_xar(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, int64_t shift, uint32_t opr_sz, uint32_t max_sz); +void gen_sve_ldr(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); +void gen_sve_str(DisasContext *s, TCGv_ptr, int vofs, int len, int rn, int imm); + #endif /* TARGET_ARM_TRANSLATE_A64_H */ diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(UCVTF_dd, aa64_sve, gen_gvec_fpst_arg_zpz, * The load should begin at the address Rn + IMM. */ -static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_ldr(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { tcg_gen_qemu_ld_i64(t0, clean_addr, midx, MO_LEUQ); - tcg_gen_st_i64(t0, cpu_env, vofs + i); + tcg_gen_st_i64(t0, base, vofs + i); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } tcg_temp_free_i64(t0); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_addi_i64(clean_addr, clean_addr, 8); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_addi_ptr(i, i, 8); tcg_gen_st_i64(t0, tp, vofs); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* @@ -XXX,XX +XXX,XX @@ static void do_ldr(DisasContext *s, uint32_t vofs, int len, int rn, int imm) default: g_assert_not_reached(); } - tcg_gen_st_i64(t0, cpu_env, vofs + len_align); + tcg_gen_st_i64(t0, base, vofs + len_align); tcg_temp_free_i64(t0); } } /* Similarly for stores. */ -static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) +void gen_sve_str(DisasContext *s, TCGv_ptr base, int vofs, + int len, int rn, int imm) { int len_align = QEMU_ALIGN_DOWN(len, 8); int len_remain = len % 8; @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) t0 = tcg_temp_new_i64(); for (i = 0; i < len_align; i += 8) { - tcg_gen_ld_i64(t0, cpu_env, vofs + i); + tcg_gen_ld_i64(t0, base, vofs + i); tcg_gen_qemu_st_i64(t0, clean_addr, midx, MO_LEUQ); tcg_gen_addi_i64(clean_addr, clean_addr, 8); } @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) clean_addr = new_tmp_a64_local(s); tcg_gen_mov_i64(clean_addr, t0); + if (base != cpu_env) { + TCGv_ptr b = tcg_temp_local_new_ptr(); + tcg_gen_mov_ptr(b, base); + base = b; + } + gen_set_label(loop); t0 = tcg_temp_new_i64(); tp = tcg_temp_new_ptr(); - tcg_gen_add_ptr(tp, cpu_env, i); + tcg_gen_add_ptr(tp, base, i); tcg_gen_ld_i64(t0, tp, vofs); tcg_gen_addi_ptr(i, i, 8); tcg_temp_free_ptr(tp); @@ -XXX,XX +XXX,XX @@ static void do_str(DisasContext *s, uint32_t vofs, int len, int rn, int imm) tcg_gen_brcondi_ptr(TCG_COND_LTU, i, len_align, loop); tcg_temp_free_ptr(i); + + if (base != cpu_env) { + tcg_temp_free_ptr(base); + assert(len_remain == 0); + } } /* Predicate register stores can be any multiple of 2. */ if (len_remain) { t0 = tcg_temp_new_i64(); - tcg_gen_ld_i64(t0, cpu_env, vofs + len_align); + tcg_gen_ld_i64(t0, base, vofs + len_align); switch (len_remain) { case 2: @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LDR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_ldr(s, off, size, a->rn, a->imm * size); + gen_sve_ldr(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_zri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = vec_full_reg_size(s); int off = vec_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_STR_pri(DisasContext *s, arg_rri *a) if (sve_access_check(s)) { int size = pred_full_reg_size(s); int off = pred_full_reg_offset(s, a->rd); - do_str(s, off, size, a->rn, a->imm * size); + gen_sve_str(s, cpu_env, off, size, a->rn, a->imm * size); } return true; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can reuse the SVE functions for LDR and STR, passing in the base of the ZA vector and a zero offset. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-23-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sme.decode | 7 +++++++ target/arm/translate-sme.c | 24 ++++++++++++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 0 esz:2 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst rs=%mova_rs LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ &ldst esz=4 rs=%mova_rs + +&ldstr rv rn imm +@ldstr ....... ... . ...... .. ... rn:5 . imm:4 \ + &ldstr rv=%mova_rs + +LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr +STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool trans_LDST1(DisasContext *s, arg_LDST1 *a) tcg_temp_free_i64(addr); return true; } + +typedef void GenLdStR(DisasContext *, TCGv_ptr, int, int, int, int); + +static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) +{ + int svl = streaming_vec_reg_size(s); + int imm = a->imm; + TCGv_ptr base; + + if (!sme_za_enabled_check(s)) { + return true; + } + + /* ZA[n] equates to ZA0H.B[n]. */ + base = get_tile_rowcol(s, MO_8, a->rv, imm, false); + + fn(s, base, 0, svl, a->rn, imm * svl); + + tcg_temp_free_ptr(base); + return true; +} + +TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) +TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-24-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 11 +++++ target/arm/sme_helper.c | 90 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 31 +++++++++++++ 4 files changed, 137 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_st1q_be_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i DEF_HELPER_FLAGS_5(sme_st1q_le_h_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_be_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) DEF_HELPER_FLAGS_5(sme_st1q_le_v_mte, TCG_CALL_NO_WG, void, env, ptr, ptr, tl, i32) + +DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ LDST1 1110000 111 st:1 rm:5 v:1 .. pg:3 rn:5 0 za_imm:4 \ LDR 1110000 100 0 000000 .. 000 ..... 0 .... @ldstr STR 1110000 100 1 000000 .. 000 ..... 0 .... @ldstr + +### SME Add Vector to Array + +&adda zad zn pm pn +@adda_32 ........ .. ..... . pm:3 pn:3 zn:5 ... zad:2 &adda +@adda_64 ........ .. ..... . pm:3 pn:3 zn:5 .. zad:3 &adda + +ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 +ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 +ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 +ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ DO_ST(q, _be, MO_128) DO_ST(q, _le, MO_128) #undef DO_ST + +void HELPER(sme_addha_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn[H4(col)]; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addha_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn[col]; + } + } + } + } +} + +void HELPER(sme_addva_s)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 4; + uint64_t *pn = vpn, *pm = vpm; + uint32_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ) { + uint64_t pa = pn[row >> 4]; + do { + if (pa & 1) { + uint32_t zn_row = zn[H4(row)]; + for (col = 0; col < oprsz; ) { + uint64_t pb = pm[col >> 4]; + do { + if (pb & 1) { + zda[tile_vslice_index(row) + H4(col)] += zn_row; + } + pb >>= 4; + } while (++col & 15); + } + } + pa >>= 4; + } while (++row & 15); + } +} + +void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint8_t *pn = vpn, *pm = vpm; + uint64_t *zda = vzda, *zn = vzn; + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t zn_row = zn[row]; + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + zda[tile_vslice_index(row) + col] += zn_row; + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_ldst_r(DisasContext *s, arg_ldstr *a, GenLdStR *fn) TRANS_FEAT(LDR, aa64_sme, do_ldst_r, a, gen_sve_ldr) TRANS_FEAT(STR, aa64_sme, do_ldst_r, a, gen_sve_str) + +static bool do_adda(DisasContext *s, arg_adda *a, MemOp esz, + gen_helper_gvec_4 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, 0); + TCGv_ptr za, zn, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + +TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) +TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) +TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) +TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-25-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 5 +++ target/arm/sme.decode | 9 +++++ target/arm/sme_helper.c | 69 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 32 ++++++++++++++++++ 4 files changed, 115 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addha_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDHA_s 11000000 10 01000 0 ... ... ..... 000 .. @adda_32 ADDVA_s 11000000 10 01000 1 ... ... ..... 000 .. @adda_32 ADDHA_d 11000000 11 01000 0 ... ... ..... 00 ... @adda_64 ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 + +### SME Outer Product + +&op zad zn zm pm pn sub:bool +@op_32 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 .. zad:2 &op +@op_64 ........ ... zm:5 pm:3 pn:3 zn:5 sub:1 . zad:3 &op + +FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ #include "exec/cpu_ldst.h" #include "exec/exec-all.h" #include "qemu/int128.h" +#include "fpu/softfloat.h" #include "vec_internal.h" #include "sve_ldst_internal.h" @@ -XXX,XX +XXX,XX @@ void HELPER(sme_addva_d)(void *vzda, void *vzn, void *vpn, } } } + +void HELPER(sme_fmopa_s)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) << 31; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. + */ + fpst = *(float_status *)vst; + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ) { + uint16_t pa = pn[H2(row >> 4)]; + do { + if (pa & 1) { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)) ^ neg; + + for (col = 0; col < oprsz; ) { + uint16_t pb = pm[H2(col >> 4)]; + do { + if (pb & 1) { + uint32_t *a = vza_row + H1_4(col); + uint32_t *m = vzm + H1_4(col); + *a = float32_muladd(n, *m, *a, 0, vst); + } + col += 4; + pb >>= 4; + } while (col & 15); + } + } + row += 4; + pa >>= 4; + } while (row & 15); + } +} + +void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + uint64_t neg = (uint64_t)simd_data(desc) << 63; + uint64_t *za = vza, *zn = vzn, *zm = vzm; + uint8_t *pn = vpn, *pm = vpm; + float_status fpst = *(float_status *)vst; + + set_default_nan_mode(true, &fpst); + + for (row = 0; row < oprsz; ++row) { + if (pn[H1(row)] & 1) { + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row] ^ neg; + + for (col = 0; col < oprsz; ++col) { + if (pm[H1(col)] & 1) { + uint64_t *a = &za_row[col]; + *a = float64_muladd(n, zm[col], *a, 0, &fpst); + } + } + } + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDHA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addha_s) TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) + +static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5_ptr *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm, fpst; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + fpst = fpstatus_ptr(FPST_FPCR); + + fn(za, zn, zm, pn, pm, fpst, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + tcg_temp_free_ptr(fpst); + return true; +} + +TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) +TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-26-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 2 ++ target/arm/sme_helper.c | 56 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 30 ++++++++++++++++++++ 4 files changed, 90 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ ADDVA_d 11000000 11 01000 1 ... ... ..... 00 ... @adda_64 FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 + +BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_fmopa_d)(void *vza, void *vzn, void *vzm, void *vpn, } } } + +/* + * Alter PAIR as needed for controlling predicates being false, + * and for NEG on an enabled row element. + */ +static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) +{ + /* + * The pseudocode uses a conditional negate after the conditional zero. + * It is simpler here to unconditionally negate before conditional zero. + */ + pair ^= neg; + if (!(pg & 1)) { + pair &= 0xffff0000u; + } + if (!(pg & 4)) { + pair &= 0x0000ffffu; + } + return pair; +} + +void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = bfdotadd(*a, n, m); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(ADDVA_s, aa64_sme, do_adda, a, MO_32, gen_helper_sme_addva_s) TRANS_FEAT(ADDHA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addha_d) TRANS_FEAT(ADDVA_d, aa64_sme_i16i64, do_adda, a, MO_64, gen_helper_sme_addva_d) +static bool do_outprod(DisasContext *s, arg_op *a, MemOp esz, + gen_helper_gvec_5 *fn) +{ + int svl = streaming_vec_reg_size(s); + uint32_t desc = simd_desc(svl, svl, a->sub); + TCGv_ptr za, zn, zm, pn, pm; + + if (!sme_smza_enabled_check(s)) { + return true; + } + + /* Sum XZR+zad to find ZAd. */ + za = get_tile_rowcol(s, esz, 31, a->zad, false); + zn = vec_full_reg_ptr(s, a->zn); + zm = vec_full_reg_ptr(s, a->zm); + pn = pred_full_reg_ptr(s, a->pn); + pm = pred_full_reg_ptr(s, a->pm); + + fn(za, zn, zm, pn, pm, tcg_constant_i32(desc)); + + tcg_temp_free_ptr(za); + tcg_temp_free_ptr(zn); + tcg_temp_free_ptr(pn); + tcg_temp_free_ptr(pm); + return true; +} + static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, gen_helper_gvec_5_ptr *fn) { @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) + +/* TODO: FEAT_EBF16 */ +TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-27-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 2 ++ target/arm/sme.decode | 1 + target/arm/sme_helper.c | 74 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 1 + 4 files changed, 78 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_5(sme_addva_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addha_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_5(sme_addva_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_7(sme_fmopa_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_s 10000000 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 +FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ static inline uint32_t f16mop_adj_pair(uint32_t pair, uint32_t pg, uint32_t neg) return pair; } +static float32 f16_dotadd(float32 sum, uint32_t e1, uint32_t e2, + float_status *s_std, float_status *s_odd) +{ + float64 e1r = float16_to_float64(e1 & 0xffff, true, s_std); + float64 e1c = float16_to_float64(e1 >> 16, true, s_std); + float64 e2r = float16_to_float64(e2 & 0xffff, true, s_std); + float64 e2c = float16_to_float64(e2 >> 16, true, s_std); + float64 t64; + float32 t32; + + /* + * The ARM pseudocode function FPDot performs both multiplies + * and the add with a single rounding operation. Emulate this + * by performing the first multiply in round-to-odd, then doing + * the second multiply as fused multiply-add, and rounding to + * float32 all in one step. + */ + t64 = float64_mul(e1r, e2r, s_odd); + t64 = float64r32_muladd(e1c, e2c, t64, 0, s_std); + + /* This conversion is exact, because we've already rounded. */ + t32 = float64_to_float32(t64, s_std); + + /* The final accumulation step is not fused. */ + return float32_add(sum, t32, s_std); +} + +void HELPER(sme_fmopa_h)(void *vza, void *vzn, void *vzm, void *vpn, + void *vpm, void *vst, uint32_t desc) +{ + intptr_t row, col, oprsz = simd_maxsz(desc); + uint32_t neg = simd_data(desc) * 0x80008000u; + uint16_t *pn = vpn, *pm = vpm; + float_status fpst_odd, fpst_std; + + /* + * Make a copy of float_status because this operation does not + * update the cumulative fp exception status. It also produces + * default nans. Make a second copy with round-to-odd -- see above. + */ + fpst_std = *(float_status *)vst; + set_default_nan_mode(true, &fpst_std); + fpst_odd = fpst_std; + set_float_rounding_mode(float_round_to_odd, &fpst_odd); + + for (row = 0; row < oprsz; ) { + uint16_t prow = pn[H2(row >> 4)]; + do { + void *vza_row = vza + tile_vslice_offset(row); + uint32_t n = *(uint32_t *)(vzn + H1_4(row)); + + n = f16mop_adj_pair(n, prow, neg); + + for (col = 0; col < oprsz; ) { + uint16_t pcol = pm[H2(col >> 4)]; + do { + if (prow & pcol & 0b0101) { + uint32_t *a = vza_row + H1_4(col); + uint32_t m = *(uint32_t *)(vzm + H1_4(col)); + + m = f16mop_adj_pair(m, pcol, 0); + *a = f16_dotadd(*a, n, m, &fpst_std, &fpst_odd); + + col += 4; + pcol >>= 4; + } + } while (col & 15); + } + row += 4; + prow >>= 4; + } while (row & 15); + } +} + void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, void *vpm, uint32_t desc) { diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ static bool do_outprod_fpst(DisasContext *s, arg_op *a, MemOp esz, return true; } +TRANS_FEAT(FMOPA_h, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_h) TRANS_FEAT(FMOPA_s, aa64_sme, do_outprod_fpst, a, MO_32, gen_helper_sme_fmopa_s) TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_fmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is SMOPA, SUMOPA, USMOPA_s, UMOPA, for both Int8 and Int16. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-28-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sme.h | 16 ++++++++ target/arm/sme.decode | 10 +++++ target/arm/sme_helper.c | 82 ++++++++++++++++++++++++++++++++++++++ target/arm/translate-sme.c | 10 +++++ 4 files changed, 118 insertions(+) diff --git a/target/arm/helper-sme.h b/target/arm/helper-sme.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sme.h +++ b/target/arm/helper-sme.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_7(sme_fmopa_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_6(sme_bfmopa, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_smopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_umopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_sumopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_6(sme_usmopa_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, ptr, i32) diff --git a/target/arm/sme.decode b/target/arm/sme.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme.decode +++ b/target/arm/sme.decode @@ -XXX,XX +XXX,XX @@ FMOPA_d 10000000 110 ..... ... ... ..... . 0 ... @op_64 BFMOPA 10000001 100 ..... ... ... ..... . 00 .. @op_32 FMOPA_h 10000001 101 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_s 1010000 0 10 0 ..... ... ... ..... . 00 .. @op_32 +SUMOPA_s 1010000 0 10 1 ..... ... ... ..... . 00 .. @op_32 +USMOPA_s 1010000 1 10 0 ..... ... ... ..... . 00 .. @op_32 +UMOPA_s 1010000 1 10 1 ..... ... ... ..... . 00 .. @op_32 + +SMOPA_d 1010000 0 11 0 ..... ... ... ..... . 0 ... @op_64 +SUMOPA_d 1010000 0 11 1 ..... ... ... ..... . 0 ... @op_64 +USMOPA_d 1010000 1 11 0 ..... ... ... ..... . 0 ... @op_64 +UMOPA_d 1010000 1 11 1 ..... ... ... ..... . 0 ... @op_64 diff --git a/target/arm/sme_helper.c b/target/arm/sme_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sme_helper.c +++ b/target/arm/sme_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(sme_bfmopa)(void *vza, void *vzn, void *vzm, void *vpn, } while (row & 15); } } + +typedef uint64_t IMOPFn(uint64_t, uint64_t, uint64_t, uint8_t, bool); + +static inline void do_imopa(uint64_t *za, uint64_t *zn, uint64_t *zm, + uint8_t *pn, uint8_t *pm, + uint32_t desc, IMOPFn *fn) +{ + intptr_t row, col, oprsz = simd_oprsz(desc) / 8; + bool neg = simd_data(desc); + + for (row = 0; row < oprsz; ++row) { + uint8_t pa = pn[H1(row)]; + uint64_t *za_row = &za[tile_vslice_index(row)]; + uint64_t n = zn[row]; + + for (col = 0; col < oprsz; ++col) { + uint8_t pb = pm[H1(col)]; + uint64_t *a = &za_row[col]; + + *a = fn(n, zm[col], *a, pa & pb, neg); + } + } +} + +#define DEF_IMOP_32(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint32_t sum0 = 0, sum1 = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_b(p); \ + sum0 += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum0 += (NTYPE)(n >> 8) * (MTYPE)(m >> 8); \ + sum0 += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum0 += (NTYPE)(n >> 24) * (MTYPE)(m >> 24); \ + sum1 += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum1 += (NTYPE)(n >> 40) * (MTYPE)(m >> 40); \ + sum1 += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + sum1 += (NTYPE)(n >> 56) * (MTYPE)(m >> 56); \ + if (neg) { \ + sum0 = (uint32_t)a - sum0, sum1 = (uint32_t)(a >> 32) - sum1; \ + } else { \ + sum0 = (uint32_t)a + sum0, sum1 = (uint32_t)(a >> 32) + sum1; \ + } \ + return ((uint64_t)sum1 << 32) | sum0; \ +} + +#define DEF_IMOP_64(NAME, NTYPE, MTYPE) \ +static uint64_t NAME(uint64_t n, uint64_t m, uint64_t a, uint8_t p, bool neg) \ +{ \ + uint64_t sum = 0; \ + /* Apply P to N as a mask, making the inactive elements 0. */ \ + n &= expand_pred_h(p); \ + sum += (NTYPE)(n >> 0) * (MTYPE)(m >> 0); \ + sum += (NTYPE)(n >> 16) * (MTYPE)(m >> 16); \ + sum += (NTYPE)(n >> 32) * (MTYPE)(m >> 32); \ + sum += (NTYPE)(n >> 48) * (MTYPE)(m >> 48); \ + return neg ? a - sum : a + sum; \ +} + +DEF_IMOP_32(smopa_s, int8_t, int8_t) +DEF_IMOP_32(umopa_s, uint8_t, uint8_t) +DEF_IMOP_32(sumopa_s, int8_t, uint8_t) +DEF_IMOP_32(usmopa_s, uint8_t, int8_t) + +DEF_IMOP_64(smopa_d, int16_t, int16_t) +DEF_IMOP_64(umopa_d, uint16_t, uint16_t) +DEF_IMOP_64(sumopa_d, int16_t, uint16_t) +DEF_IMOP_64(usmopa_d, uint16_t, int16_t) + +#define DEF_IMOPH(NAME) \ + void HELPER(sme_##NAME)(void *vza, void *vzn, void *vzm, void *vpn, \ + void *vpm, uint32_t desc) \ + { do_imopa(vza, vzn, vzm, vpn, vpm, desc, NAME); } + +DEF_IMOPH(smopa_s) +DEF_IMOPH(umopa_s) +DEF_IMOPH(sumopa_s) +DEF_IMOPH(usmopa_s) +DEF_IMOPH(smopa_d) +DEF_IMOPH(umopa_d) +DEF_IMOPH(sumopa_d) +DEF_IMOPH(usmopa_d) diff --git a/target/arm/translate-sme.c b/target/arm/translate-sme.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sme.c +++ b/target/arm/translate-sme.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(FMOPA_d, aa64_sme_f64f64, do_outprod_fpst, a, MO_64, gen_helper_sme_f /* TODO: FEAT_EBF16 */ TRANS_FEAT(BFMOPA, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_bfmopa) + +TRANS_FEAT(SMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_smopa_s) +TRANS_FEAT(UMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_umopa_s) +TRANS_FEAT(SUMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_sumopa_s) +TRANS_FEAT(USMOPA_s, aa64_sme, do_outprod, a, MO_32, gen_helper_sme_usmopa_s) + +TRANS_FEAT(SMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_smopa_d) +TRANS_FEAT(UMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_umopa_d) +TRANS_FEAT(SUMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_sumopa_d) +TRANS_FEAT(USMOPA_d, aa64_sme_i16i64, do_outprod, a, MO_64, gen_helper_sme_usmopa_d) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-29-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/sve.decode | 20 +++++++++++++ target/arm/translate-sve.c | 57 ++++++++++++++++++++++++++++++++++++++ 2 files changed, 77 insertions(+) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ BFMLALT_zzxw 01100100 11 1 ..... 0100.1 ..... ..... @rrxr_3a esz=2 ### SVE2 floating-point bfloat16 dot-product (indexed) BFDOT_zzxz 01100100 01 1 ..... 010000 ..... ..... @rrxr_2 esz=2 + +### SVE broadcast predicate element + +&psel esz pd pn pm rv imm +%psel_rv 16:2 !function=plus_12 +%psel_imm_b 22:2 19:2 +%psel_imm_h 22:2 20:1 +%psel_imm_s 22:2 +%psel_imm_d 23:1 +@psel ........ .. . ... .. .. pn:4 . pm:4 . pd:4 \ + &psel rv=%psel_rv + +PSEL 00100101 .. 1 ..1 .. 01 .... 0 .... 0 .... \ + @psel esz=0 imm=%psel_imm_b +PSEL 00100101 .. 1 .10 .. 01 .... 0 .... 0 .... \ + @psel esz=1 imm=%psel_imm_h +PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ + @psel esz=2 imm=%psel_imm_s +PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ + @psel esz=3 imm=%psel_imm_d diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool do_BFMLAL_zzxw(DisasContext *s, arg_rrxr_esz *a, bool sel) TRANS_FEAT(BFMLALB_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, false) TRANS_FEAT(BFMLALT_zzxw, aa64_sve_bf16, do_BFMLAL_zzxw, a, true) + +static bool trans_PSEL(DisasContext *s, arg_psel *a) +{ + int vl = vec_full_reg_size(s); + int pl = pred_gvec_reg_size(s); + int elements = vl >> a->esz; + TCGv_i64 tmp, didx, dbit; + TCGv_ptr ptr; + + if (!dc_isar_feature(aa64_sme, s)) { + return false; + } + if (!sve_access_check(s)) { + return true; + } + + tmp = tcg_temp_new_i64(); + dbit = tcg_temp_new_i64(); + didx = tcg_temp_new_i64(); + ptr = tcg_temp_new_ptr(); + + /* Compute the predicate element. */ + tcg_gen_addi_i64(tmp, cpu_reg(s, a->rv), a->imm); + if (is_power_of_2(elements)) { + tcg_gen_andi_i64(tmp, tmp, elements - 1); + } else { + tcg_gen_remu_i64(tmp, tmp, tcg_constant_i64(elements)); + } + + /* Extract the predicate byte and bit indices. */ + tcg_gen_shli_i64(tmp, tmp, a->esz); + tcg_gen_andi_i64(dbit, tmp, 7); + tcg_gen_shri_i64(didx, tmp, 3); + if (HOST_BIG_ENDIAN) { + tcg_gen_xori_i64(didx, didx, 7); + } + + /* Load the predicate word. */ + tcg_gen_trunc_i64_ptr(ptr, didx); + tcg_gen_add_ptr(ptr, ptr, cpu_env); + tcg_gen_ld8u_i64(tmp, ptr, pred_full_reg_offset(s, a->pm)); + + /* Extract the predicate bit and replicate to MO_64. */ + tcg_gen_shr_i64(tmp, tmp, dbit); + tcg_gen_andi_i64(tmp, tmp, 1); + tcg_gen_neg_i64(tmp, tmp); + + /* Apply to either copy the source, or write zeros. */ + tcg_gen_gvec_ands(MO_64, pred_full_reg_offset(s, a->pd), + pred_full_reg_offset(s, a->pn), tmp, pl, pl); + + tcg_temp_free_i64(tmp); + tcg_temp_free_i64(dbit); + tcg_temp_free_i64(didx); + tcg_temp_free_ptr(ptr); + return true; +} -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-30-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper-sve.h | 2 ++ target/arm/sve.decode | 1 + target/arm/sve_helper.c | 16 ++++++++++++++++ target/arm/translate-sve.c | 2 ++ 4 files changed, 21 insertions(+) diff --git a/target/arm/helper-sve.h b/target/arm/helper-sve.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper-sve.h +++ b/target/arm/helper-sve.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_4(sve_revh_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_revw_d, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_4(sme_revd_q, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) + DEF_HELPER_FLAGS_4(sve_rbit_b, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_h, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) DEF_HELPER_FLAGS_4(sve_rbit_s, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, i32) diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ REVB 00000101 .. 1001 00 100 ... ..... ..... @rd_pg_rn REVH 00000101 .. 1001 01 100 ... ..... ..... @rd_pg_rn REVW 00000101 .. 1001 10 100 ... ..... ..... @rd_pg_rn RBIT 00000101 .. 1001 11 100 ... ..... ..... @rd_pg_rn +REVD 00000101 00 1011 10 100 ... ..... ..... @rd_pg_rn_e0 # SVE vector splice (predicated, destructive) SPLICE 00000101 .. 101 100 100 ... ..... ..... @rdn_pg_rm diff --git a/target/arm/sve_helper.c b/target/arm/sve_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve_helper.c +++ b/target/arm/sve_helper.c @@ -XXX,XX +XXX,XX @@ DO_ZPZ_D(sve_revh_d, uint64_t, hswap64) DO_ZPZ_D(sve_revw_d, uint64_t, wswap64) +void HELPER(sme_revd_q)(void *vd, void *vn, void *vg, uint32_t desc) +{ + intptr_t i, opr_sz = simd_oprsz(desc) / 8; + uint64_t *d = vd, *n = vn; + uint8_t *pg = vg; + + for (i = 0; i < opr_sz; i += 2) { + if (pg[H1(i)] & 1) { + uint64_t n0 = n[i + 0]; + uint64_t n1 = n[i + 1]; + d[i + 0] = n1; + d[i + 1] = n0; + } + } +} + DO_ZPZ(sve_rbit_b, uint8_t, H1, revbit8) DO_ZPZ(sve_rbit_h, uint16_t, H1_2, revbit16) DO_ZPZ(sve_rbit_s, uint32_t, H1_4, revbit32) diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ TRANS_FEAT(REVH, aa64_sve, gen_gvec_ool_arg_zpz, revh_fns[a->esz], a, 0) TRANS_FEAT(REVW, aa64_sve, gen_gvec_ool_arg_zpz, a->esz == 3 ? gen_helper_sve_revw_d : NULL, a, 0) +TRANS_FEAT(REVD, aa64_sme, gen_gvec_ool_arg_zpz, gen_helper_sme_revd_q, a, 0) + TRANS_FEAT(SPLICE, aa64_sve, gen_gvec_ool_arg_zpzz, gen_helper_sve_splice, a, a->esz) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> This is an SVE instruction that operates using the SVE vector length but that it is present only if SME is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-31-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.h | 18 +++++++ target/arm/sve.decode | 5 ++ target/arm/translate-sve.c | 102 +++++++++++++++++++++++++++++++++++++ target/arm/vec_helper.c | 24 +++++++++ 4 files changed, 149 insertions(+) diff --git a/target/arm/helper.h b/target/arm/helper.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.h +++ b/target/arm/helper.h @@ -XXX,XX +XXX,XX @@ DEF_HELPER_FLAGS_6(gvec_bfmlal, TCG_CALL_NO_RWG, DEF_HELPER_FLAGS_6(gvec_bfmlal_idx, TCG_CALL_NO_RWG, void, ptr, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_sclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + +DEF_HELPER_FLAGS_5(gvec_uclamp_b, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_h, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_s, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) +DEF_HELPER_FLAGS_5(gvec_uclamp_d, TCG_CALL_NO_RWG, + void, ptr, ptr, ptr, ptr, i32) + #ifdef TARGET_AARCH64 #include "helper-a64.h" #include "helper-sve.h" diff --git a/target/arm/sve.decode b/target/arm/sve.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/sve.decode +++ b/target/arm/sve.decode @@ -XXX,XX +XXX,XX @@ PSEL 00100101 .. 1 100 .. 01 .... 0 .... 0 .... \ @psel esz=2 imm=%psel_imm_s PSEL 00100101 .1 1 000 .. 01 .... 0 .... 0 .... \ @psel esz=3 imm=%psel_imm_d + +### SVE clamp + +SCLAMP 01000100 .. 0 ..... 110000 ..... ..... @rda_rn_rm +UCLAMP 01000100 .. 0 ..... 110001 ..... ..... @rda_rn_rm diff --git a/target/arm/translate-sve.c b/target/arm/translate-sve.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-sve.c +++ b/target/arm/translate-sve.c @@ -XXX,XX +XXX,XX @@ static bool trans_PSEL(DisasContext *s, arg_psel *a) tcg_temp_free_ptr(ptr); return true; } + +static void gen_sclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_smax_i32(d, a, n); + tcg_gen_smin_i32(d, d, m); +} + +static void gen_sclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_smax_i64(d, a, n); + tcg_gen_smin_i64(d, d, m); +} + +static void gen_sclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_smax_vec(vece, d, a, n); + tcg_gen_smin_vec(vece, d, d, m); +} + +static void gen_sclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_smin_vec, INDEX_op_smax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_sclamp_i32, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_sclamp_i64, + .fniv = gen_sclamp_vec, + .fno = gen_helper_gvec_sclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(SCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_sclamp, a) + +static void gen_uclamp_i32(TCGv_i32 d, TCGv_i32 n, TCGv_i32 m, TCGv_i32 a) +{ + tcg_gen_umax_i32(d, a, n); + tcg_gen_umin_i32(d, d, m); +} + +static void gen_uclamp_i64(TCGv_i64 d, TCGv_i64 n, TCGv_i64 m, TCGv_i64 a) +{ + tcg_gen_umax_i64(d, a, n); + tcg_gen_umin_i64(d, d, m); +} + +static void gen_uclamp_vec(unsigned vece, TCGv_vec d, TCGv_vec n, + TCGv_vec m, TCGv_vec a) +{ + tcg_gen_umax_vec(vece, d, a, n); + tcg_gen_umin_vec(vece, d, d, m); +} + +static void gen_uclamp(unsigned vece, uint32_t d, uint32_t n, uint32_t m, + uint32_t a, uint32_t oprsz, uint32_t maxsz) +{ + static const TCGOpcode vecop[] = { + INDEX_op_umin_vec, INDEX_op_umax_vec, 0 + }; + static const GVecGen4 ops[4] = { + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_b, + .opt_opc = vecop, + .vece = MO_8 }, + { .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_h, + .opt_opc = vecop, + .vece = MO_16 }, + { .fni4 = gen_uclamp_i32, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_s, + .opt_opc = vecop, + .vece = MO_32 }, + { .fni8 = gen_uclamp_i64, + .fniv = gen_uclamp_vec, + .fno = gen_helper_gvec_uclamp_d, + .opt_opc = vecop, + .vece = MO_64, + .prefer_i64 = TCG_TARGET_REG_BITS == 64 } + }; + tcg_gen_gvec_4(d, n, m, a, oprsz, maxsz, &ops[vece]); +} + +TRANS_FEAT(UCLAMP, aa64_sme, gen_gvec_fn_arg_zzzz, gen_uclamp, a) diff --git a/target/arm/vec_helper.c b/target/arm/vec_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/vec_helper.c +++ b/target/arm/vec_helper.c @@ -XXX,XX +XXX,XX @@ void HELPER(gvec_bfmlal_idx)(void *vd, void *vn, void *vm, } clear_tail(d, opr_sz, simd_maxsz(desc)); } + +#define DO_CLAMP(NAME, TYPE) \ +void HELPER(NAME)(void *d, void *n, void *m, void *a, uint32_t desc) \ +{ \ + intptr_t i, opr_sz = simd_oprsz(desc); \ + for (i = 0; i < opr_sz; i += sizeof(TYPE)) { \ + TYPE aa = *(TYPE *)(a + i); \ + TYPE nn = *(TYPE *)(n + i); \ + TYPE mm = *(TYPE *)(m + i); \ + TYPE dd = MIN(MAX(aa, nn), mm); \ + *(TYPE *)(d + i) = dd; \ + } \ + clear_tail(d, opr_sz, simd_maxsz(desc)); \ +} + +DO_CLAMP(gvec_sclamp_b, int8_t) +DO_CLAMP(gvec_sclamp_h, int16_t) +DO_CLAMP(gvec_sclamp_s, int32_t) +DO_CLAMP(gvec_sclamp_d, int64_t) + +DO_CLAMP(gvec_uclamp_b, uint8_t) +DO_CLAMP(gvec_uclamp_h, uint16_t) +DO_CLAMP(gvec_uclamp_s, uint32_t) +DO_CLAMP(gvec_uclamp_d, uint64_t) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> We can handle both exception entry and exception return by hooking into aarch64_sve_change_el. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-32-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 15 +++++++++++++-- 1 file changed, 13 insertions(+), 2 deletions(-) diff --git a/target/arm/helper.c b/target/arm/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/helper.c +++ b/target/arm/helper.c @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, return; } + old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; + new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; + + /* + * Both AArch64.TakeException and AArch64.ExceptionReturn + * invoke ResetSVEState when taking an exception from, or + * returning to, AArch32 state when PSTATE.SM is enabled. + */ + if (old_a64 != new_a64 && FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + return; + } + /* * DDI0584A.d sec 3.2: "If SVE instructions are disabled or trapped * at ELx, or not available because the EL is in AArch32 state, then @@ -XXX,XX +XXX,XX @@ void aarch64_sve_change_el(CPUARMState *env, int old_el, * we already have the correct register contents when encountering the * vq0->vq0 transition between EL0->EL1. */ - old_a64 = old_el ? arm_el_is_aa64(env, old_el) : el0_a64; old_len = (old_a64 && !sve_exception_el(env, old_el) ? sve_vqm1_for_el(env, old_el) : 0); - new_a64 = new_el ? arm_el_is_aa64(env, new_el) : el0_a64; new_len = (new_a64 && !sve_exception_el(env, new_el) ? sve_vqm1_for_el(env, new_el) : 0); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Note that SME remains effectively disabled for user-only, because we do not yet set CPACR_EL1.SMEN. This needs to wait until the kernel ABI is implemented. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-33-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/emulation.rst | 4 ++++ target/arm/cpu64.c | 11 +++++++++++ 2 files changed, 15 insertions(+) diff --git a/docs/system/arm/emulation.rst b/docs/system/arm/emulation.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/emulation.rst +++ b/docs/system/arm/emulation.rst @@ -XXX,XX +XXX,XX @@ the following architecture extensions: - FEAT_SHA512 (Advanced SIMD SHA512 instructions) - FEAT_SM3 (Advanced SIMD SM3 instructions) - FEAT_SM4 (Advanced SIMD SM4 instructions) +- FEAT_SME (Scalable Matrix Extension) +- FEAT_SME_FA64 (Full A64 instruction set in Streaming SVE mode) +- FEAT_SME_F64F64 (Double-precision floating-point outer product instructions) +- FEAT_SME_I16I64 (16-bit to 64-bit integer widening outer product instructions) - FEAT_SPECRES (Speculation restriction instructions) - FEAT_SSBS (Speculative Store Bypass Safe) - FEAT_TLBIOS (TLB invalidate instructions in Outer Shareable domain) diff --git a/target/arm/cpu64.c b/target/arm/cpu64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu64.c +++ b/target/arm/cpu64.c @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) */ t = FIELD_DP64(t, ID_AA64PFR1, MTE, 3); /* FEAT_MTE3 */ t = FIELD_DP64(t, ID_AA64PFR1, RAS_FRAC, 0); /* FEAT_RASv1p1 + FEAT_DoubleFault */ + t = FIELD_DP64(t, ID_AA64PFR1, SME, 1); /* FEAT_SME */ t = FIELD_DP64(t, ID_AA64PFR1, CSV2_FRAC, 0); /* FEAT_CSV2_2 */ cpu->isar.id_aa64pfr1 = t; @@ -XXX,XX +XXX,XX @@ static void aarch64_max_initfn(Object *obj) t = FIELD_DP64(t, ID_AA64DFR0, PMUVER, 5); /* FEAT_PMUv3p4 */ cpu->isar.id_aa64dfr0 = t; + t = cpu->isar.id_aa64smfr0; + t = FIELD_DP64(t, ID_AA64SMFR0, F32F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, B16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F16F32, 1); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, I8I32, 0xf); /* FEAT_SME */ + t = FIELD_DP64(t, ID_AA64SMFR0, F64F64, 1); /* FEAT_SME_F64F64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, I16I64, 0xf); /* FEAT_SME_I16I64 */ + t = FIELD_DP64(t, ID_AA64SMFR0, FA64, 1); /* FEAT_SME_FA64 */ + cpu->isar.id_aa64smfr0 = t; + /* Replicate the same data to the 32-bit id registers. */ aa32_max_features(cpu); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-34-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_cpu.h | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/target_cpu.h b/linux-user/aarch64/target_cpu.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_cpu.h +++ b/linux-user/aarch64/target_cpu.h @@ -XXX,XX +XXX,XX @@ static inline void cpu_clone_regs_parent(CPUARMState *env, unsigned flags) static inline void cpu_set_tls(CPUARMState *env, target_ulong newtls) { - /* Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is + /* + * Note that AArch64 Linux keeps the TLS pointer in TPIDR; this is * different from AArch32 Linux, which uses TPIDRRO. */ env->cp15.tpidr_el[0] = newtls; + /* TPIDR2_EL0 is cleared with CLONE_SETTLS. */ + env->cp15.tpidr2_el0 = 0; } static inline abi_ulong get_sp_from_cpustate(CPUARMState *state) -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-35-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/cpu_loop.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/linux-user/aarch64/cpu_loop.c b/linux-user/aarch64/cpu_loop.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/cpu_loop.c +++ b/linux-user/aarch64/cpu_loop.c @@ -XXX,XX +XXX,XX @@ void cpu_loop(CPUARMState *env) switch (trapnr) { case EXCP_SWI: + /* + * On syscall, PSTATE.ZA is preserved, along with the ZA matrix. + * PSTATE.SM is cleared, per SMSTOP, which does ResetSVEState. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + env->svcr = FIELD_DP64(env->svcr, SVCR, SM, 0); + arm_rebuild_hflags(env); + arm_reset_sve_state(env); + } ret = do_syscall(env, env->xregs[8], env->xregs[0], -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Make sure to zero the currently reserved fields. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-36-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 9 ++++++++- 1 file changed, 8 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_extra_context { struct target_sve_context { struct target_aarch64_ctx head; uint16_t vl; - uint16_t reserved[3]; + uint16_t flags; + uint16_t reserved[2]; /* The actual SVE data immediately follows. It is laid out * according to TARGET_SVE_SIG_{Z,P}REG_OFFSET, based off of * the original struct pointer. @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_CONTEXT_SIZE(VQ) \ (TARGET_SVE_SIG_PREG_OFFSET(VQ, 17)) +#define TARGET_SVE_SIG_FLAG_SM 1 + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, { int i, j; + memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); __put_user(size, &sve->head.size); __put_user(vq * TARGET_SVE_VQ_BYTES, &sve->vl); + if (FIELD_EX64(env->svcr, SVCR, SM)) { + __put_user(TARGET_SVE_SIG_FLAG_SM, &sve->flags); + } /* Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian store -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Fold the return value setting into the goto, so each point of failure need not do both. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-37-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 26 +++++++++++--------------- 1 file changed, 11 insertions(+), 15 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - bool err = false; int vq = 0, sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, switch (magic) { case 0: if (size != 0) { - err = true; - goto exit; + goto err; } if (used_extra) { ctx = NULL; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, case TARGET_FPSIMD_MAGIC: if (fpsimd || size != sizeof(struct target_fpsimd_context)) { - err = true; - goto exit; + goto err; } fpsimd = (struct target_fpsimd_context *)ctx; break; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; } } - err = true; - goto exit; + goto err; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { - err = true; - goto exit; + goto err; } __get_user(extra_datap, &((struct target_extra_context *)ctx)->datap); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, /* Unknown record -- we certainly didn't generate it. * Did we in fact get out of sync? */ - err = true; - goto exit; + goto err; } ctx = (void *)ctx + size; } @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (fpsimd) { target_restore_fpsimd_record(env, fpsimd); } else { - err = true; + goto err; } /* SVE data, if present, overwrites FPSIMD data. */ if (sve) { target_restore_sve_record(env, sve, vq); } - - exit: unlock_user(extra, extra_datap, 0); - return err; + return 0; + + err: + unlock_user(extra, extra_datap, 0); + return 1; } static abi_ulong get_sigframe(struct target_sigaction *ka, -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> In parse_user_sigframe, the kernel rejects duplicate sve records, or records that are smaller than the header. We were silently allowing these cases to pass, dropping the record. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-38-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, break; case TARGET_SVE_MAGIC: + if (sve || size < sizeof(struct target_sve_context)) { + goto err; + } if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { vq = sve_vq(env); sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (!sve && size == sve_size) { + if (size == sve_size) { sve = (struct target_sve_context *)ctx; break; } -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-39-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, __get_user(extra_size, &((struct target_extra_context *)ctx)->size); extra = lock_user(VERIFY_READ, extra_datap, extra_size, 0); + if (!extra) { + return 1; + } break; default: -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Move the checks out of the parsing loop and into the restore function. This more closely mirrors the code structure in the kernel, and is slightly clearer. Reject rather than silently skip incorrect VL and SVE record sizes, bringing our checks in to line with those the kernel does. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-40-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 51 +++++++++++++++++++++++++------------ 1 file changed, 35 insertions(+), 16 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, } } -static void target_restore_sve_record(CPUARMState *env, - struct target_sve_context *sve, int vq) +static bool target_restore_sve_record(CPUARMState *env, + struct target_sve_context *sve, + int size) { - int i, j; + int i, j, vl, vq; - /* Note that SVE regs are stored as a byte stream, with each byte element + if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + return false; + } + + __get_user(vl, &sve->vl); + vq = sve_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.SM. */ + if (size <= sizeof(*sve)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_SVE_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + /* + * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load * of our 64-bit hunks. */ @@ -XXX,XX +XXX,XX @@ static void target_restore_sve_record(CPUARMState *env, } } } + return true; } static int target_restore_sigframe(CPUARMState *env, @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, struct target_sve_context *sve = NULL; uint64_t extra_datap = 0; bool used_extra = false; - int vq = 0, sve_size = 0; + int sve_size = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, if (sve || size < sizeof(struct target_sve_context)) { goto err; } - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); - if (size == sve_size) { - sve = (struct target_sve_context *)ctx; - break; - } - } - goto err; + sve = (struct target_sve_context *)ctx; + sve_size = size; + break; case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve) { - target_restore_sve_record(env, sve, vq); + if (sve && !target_restore_sve_record(env, sve, sve_size)) { + goto err; } unlock_user(extra, extra_datap, 0); return 0; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Set the SM bit in the SVE record on signal delivery, create the ZA record. Restore SM and ZA state according to the records present on return. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-41-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/signal.c | 167 +++++++++++++++++++++++++++++++++--- 1 file changed, 154 insertions(+), 13 deletions(-) diff --git a/linux-user/aarch64/signal.c b/linux-user/aarch64/signal.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/signal.c +++ b/linux-user/aarch64/signal.c @@ -XXX,XX +XXX,XX @@ struct target_sve_context { #define TARGET_SVE_SIG_FLAG_SM 1 +#define TARGET_ZA_MAGIC 0x54366345 + +struct target_za_context { + struct target_aarch64_ctx head; + uint16_t vl; + uint16_t reserved[3]; + /* The actual ZA data immediately follows. */ +}; + +#define TARGET_ZA_SIG_REGS_OFFSET \ + QEMU_ALIGN_UP(sizeof(struct target_za_context), TARGET_SVE_VQ_BYTES) +#define TARGET_ZA_SIG_ZAV_OFFSET(VQ, N) \ + (TARGET_ZA_SIG_REGS_OFFSET + (VQ) * TARGET_SVE_VQ_BYTES * (N)) +#define TARGET_ZA_SIG_CONTEXT_SIZE(VQ) \ + TARGET_ZA_SIG_ZAV_OFFSET(VQ, VQ * TARGET_SVE_VQ_BYTES) + struct target_rt_sigframe { struct target_siginfo info; struct target_ucontext uc; @@ -XXX,XX +XXX,XX @@ static void target_setup_end_record(struct target_aarch64_ctx *end) } static void target_setup_sve_record(struct target_sve_context *sve, - CPUARMState *env, int vq, int size) + CPUARMState *env, int size) { - int i, j; + int i, j, vq = sve_vq(env); memset(sve, 0, sizeof(*sve)); __put_user(TARGET_SVE_MAGIC, &sve->head.magic); @@ -XXX,XX +XXX,XX @@ static void target_setup_sve_record(struct target_sve_context *sve, } } +static void target_setup_za_record(struct target_za_context *za, + CPUARMState *env, int size) +{ + int vq = sme_vq(env); + int vl = vq * TARGET_SVE_VQ_BYTES; + int i, j; + + memset(za, 0, sizeof(*za)); + __put_user(TARGET_ZA_MAGIC, &za->head.magic); + __put_user(size, &za->head.size); + __put_user(vl, &za->vl); + + if (size == TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return; + } + assert(size == TARGET_ZA_SIG_CONTEXT_SIZE(vq)); + + /* + * Note that ZA vectors are stored as a byte stream, + * with each byte element at a subsequent address. + */ + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __put_user_e(env->zarray[i].d[j], z + j, le); + } + } +} + static void target_restore_general_frame(CPUARMState *env, struct target_rt_sigframe *sf) { @@ -XXX,XX +XXX,XX @@ static void target_restore_fpsimd_record(CPUARMState *env, static bool target_restore_sve_record(CPUARMState *env, struct target_sve_context *sve, - int size) + int size, int *svcr) { - int i, j, vl, vq; + int i, j, vl, vq, flags; + bool sm; - if (!cpu_isar_feature(aa64_sve, env_archcpu(env))) { + __get_user(vl, &sve->vl); + __get_user(flags, &sve->flags); + + sm = flags & TARGET_SVE_SIG_FLAG_SM; + + /* The cpu must support Streaming or Non-streaming SVE. */ + if (sm + ? !cpu_isar_feature(aa64_sme, env_archcpu(env)) + : !cpu_isar_feature(aa64_sve, env_archcpu(env))) { return false; } - __get_user(vl, &sve->vl); - vq = sve_vq(env); + /* + * Note that we cannot use sve_vq() because that depends on the + * current setting of PSTATE.SM, not the state to be restored. + */ + vq = sve_vqm1_for_el_sm(env, 0, sm) + 1; /* Reject mismatched VL. */ if (vl != vq * TARGET_SVE_VQ_BYTES) { @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return false; } + *svcr = FIELD_DP64(*svcr, SVCR, SM, sm); + /* * Note that SVE regs are stored as a byte stream, with each byte element * at a subsequent address. This corresponds to a little-endian load @@ -XXX,XX +XXX,XX @@ static bool target_restore_sve_record(CPUARMState *env, return true; } +static bool target_restore_za_record(CPUARMState *env, + struct target_za_context *za, + int size, int *svcr) +{ + int i, j, vl, vq; + + if (!cpu_isar_feature(aa64_sme, env_archcpu(env))) { + return false; + } + + __get_user(vl, &za->vl); + vq = sme_vq(env); + + /* Reject mismatched VL. */ + if (vl != vq * TARGET_SVE_VQ_BYTES) { + return false; + } + + /* Accept empty record -- used to clear PSTATE.ZA. */ + if (size <= TARGET_ZA_SIG_CONTEXT_SIZE(0)) { + return true; + } + + /* Reject non-empty but incomplete record. */ + if (size < TARGET_ZA_SIG_CONTEXT_SIZE(vq)) { + return false; + } + + *svcr = FIELD_DP64(*svcr, SVCR, ZA, 1); + + for (i = 0; i < vl; ++i) { + uint64_t *z = (void *)za + TARGET_ZA_SIG_ZAV_OFFSET(vq, i); + for (j = 0; j < vq * 2; ++j) { + __get_user_e(env->zarray[i].d[j], z + j, le); + } + } + return true; +} + static int target_restore_sigframe(CPUARMState *env, struct target_rt_sigframe *sf) { struct target_aarch64_ctx *ctx, *extra = NULL; struct target_fpsimd_context *fpsimd = NULL; struct target_sve_context *sve = NULL; + struct target_za_context *za = NULL; uint64_t extra_datap = 0; bool used_extra = false; int sve_size = 0; + int za_size = 0; + int svcr = 0; target_restore_general_frame(env, sf); @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, sve_size = size; break; + case TARGET_ZA_MAGIC: + if (za || size < sizeof(struct target_za_context)) { + goto err; + } + za = (struct target_za_context *)ctx; + za_size = size; + break; + case TARGET_EXTRA_MAGIC: if (extra || size != sizeof(struct target_extra_context)) { goto err; @@ -XXX,XX +XXX,XX @@ static int target_restore_sigframe(CPUARMState *env, } /* SVE data, if present, overwrites FPSIMD data. */ - if (sve && !target_restore_sve_record(env, sve, sve_size)) { + if (sve && !target_restore_sve_record(env, sve, sve_size, &svcr)) { goto err; } + if (za && !target_restore_za_record(env, za, za_size, &svcr)) { + goto err; + } + if (env->svcr != svcr) { + env->svcr = svcr; + arm_rebuild_hflags(env); + } unlock_user(extra, extra_datap, 0); return 0; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, .total_size = offsetof(struct target_rt_sigframe, uc.tuc_mcontext.__reserved), }; - int fpsimd_ofs, fr_ofs, sve_ofs = 0, vq = 0, sve_size = 0; + int fpsimd_ofs, fr_ofs, sve_ofs = 0, za_ofs = 0; + int sve_size = 0, za_size = 0; struct target_rt_sigframe *frame; struct target_rt_frame_record *fr; abi_ulong frame_addr, return_addr; @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, &layout); /* SVE state needs saving only if it exists. */ - if (cpu_isar_feature(aa64_sve, env_archcpu(env))) { - vq = sve_vq(env); - sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(vq), 16); + if (cpu_isar_feature(aa64_sve, env_archcpu(env)) || + cpu_isar_feature(aa64_sme, env_archcpu(env))) { + sve_size = QEMU_ALIGN_UP(TARGET_SVE_SIG_CONTEXT_SIZE(sve_vq(env)), 16); sve_ofs = alloc_sigframe_space(sve_size, &layout); } + if (cpu_isar_feature(aa64_sme, env_archcpu(env))) { + /* ZA state needs saving only if it is enabled. */ + if (FIELD_EX64(env->svcr, SVCR, ZA)) { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(sme_vq(env)); + } else { + za_size = TARGET_ZA_SIG_CONTEXT_SIZE(0); + } + za_ofs = alloc_sigframe_space(za_size, &layout); + } if (layout.extra_ofs) { /* Reserve space for the extra end marker. The standard end marker @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, target_setup_end_record((void *)frame + layout.extra_end_ofs); } if (sve_ofs) { - target_setup_sve_record((void *)frame + sve_ofs, env, vq, sve_size); + target_setup_sve_record((void *)frame + sve_ofs, env, sve_size); + } + if (za_ofs) { + target_setup_za_record((void *)frame + za_ofs, env, za_size); } /* Set up the stack frame for unwinding. */ @@ -XXX,XX +XXX,XX @@ static void target_setup_frame(int usig, struct target_sigaction *ka, env->btype = 2; } + /* + * Invoke the signal handler with both SM and ZA disabled. + * When clearing SM, ResetSVEState, per SMSTOP. + */ + if (FIELD_EX64(env->svcr, SVCR, SM)) { + arm_reset_sve_state(env); + } + if (env->svcr) { + env->svcr = 0; + arm_rebuild_hflags(env); + } + if (info) { tswap_siginfo(&frame->info, info); env->xregs[1] = frame_addr + offsetof(struct target_rt_sigframe, info); -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Add "sve" to the sve prctl functions, to distinguish them from the coming "sme" prctls with similar names. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-42-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 8 ++++---- linux-user/syscall.c | 12 ++++++------ 2 files changed, 10 insertions(+), 10 deletions(-) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ #ifndef AARCH64_TARGET_PRCTL_H #define AARCH64_TARGET_PRCTL_H -static abi_long do_prctl_get_vl(CPUArchState *env) +static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_get_vl(CPUArchState *env) } return -TARGET_EINVAL; } -#define do_prctl_get_vl do_prctl_get_vl +#define do_prctl_sve_get_vl do_prctl_sve_get_vl -static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) +static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) { /* * We cannot support either PR_SVE_SET_VL_ONEXEC or PR_SVE_VL_INHERIT. @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_set_vl(CPUArchState *env, abi_long arg2) } return -TARGET_EINVAL; } -#define do_prctl_set_vl do_prctl_set_vl +#define do_prctl_sve_set_vl do_prctl_sve_set_vl static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_fp_mode #define do_prctl_set_fp_mode do_prctl_inval1 #endif -#ifndef do_prctl_get_vl -#define do_prctl_get_vl do_prctl_inval0 +#ifndef do_prctl_sve_get_vl +#define do_prctl_sve_get_vl do_prctl_inval0 #endif -#ifndef do_prctl_set_vl -#define do_prctl_set_vl do_prctl_inval1 +#ifndef do_prctl_sve_set_vl +#define do_prctl_sve_set_vl do_prctl_inval1 #endif #ifndef do_prctl_reset_keys #define do_prctl_reset_keys do_prctl_inval1 @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, case PR_SET_FP_MODE: return do_prctl_set_fp_mode(env, arg2); case PR_SVE_GET_VL: - return do_prctl_get_vl(env); + return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: - return do_prctl_set_vl(env, arg2); + return do_prctl_sve_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> These prctl set the Streaming SVE vector length, which may be completely different from the Normal SVE vector length. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-43-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/aarch64/target_prctl.h | 54 +++++++++++++++++++++++++++++++ linux-user/syscall.c | 16 +++++++++ 2 files changed, 70 insertions(+) diff --git a/linux-user/aarch64/target_prctl.h b/linux-user/aarch64/target_prctl.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_prctl.h +++ b/linux-user/aarch64/target_prctl.h @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_get_vl(CPUArchState *env) { ARMCPU *cpu = env_archcpu(env); if (cpu_isar_feature(aa64_sve, cpu)) { + /* PSTATE.SM is always unset on syscall entry. */ return sve_vq(env) * 16; } return -TARGET_EINVAL; @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { uint32_t vq, old_vq; + /* PSTATE.SM is always unset on syscall entry. */ old_vq = sve_vq(env); /* @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_sve_set_vl(CPUArchState *env, abi_long arg2) } #define do_prctl_sve_set_vl do_prctl_sve_set_vl +static abi_long do_prctl_sme_get_vl(CPUArchState *env) +{ + ARMCPU *cpu = env_archcpu(env); + if (cpu_isar_feature(aa64_sme, cpu)) { + return sme_vq(env) * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_get_vl do_prctl_sme_get_vl + +static abi_long do_prctl_sme_set_vl(CPUArchState *env, abi_long arg2) +{ + /* + * We cannot support either PR_SME_SET_VL_ONEXEC or PR_SME_VL_INHERIT. + * Note the kernel definition of sve_vl_valid allows for VQ=512, + * i.e. VL=8192, even though the architectural maximum is VQ=16. + */ + if (cpu_isar_feature(aa64_sme, env_archcpu(env)) + && arg2 >= 0 && arg2 <= 512 * 16 && !(arg2 & 15)) { + int vq, old_vq; + + old_vq = sme_vq(env); + + /* + * Bound the value of vq, so that we know that it fits into + * the 4-bit field in SMCR_EL1. Because PSTATE.SM is cleared + * on syscall entry, we are not modifying the current SVE + * vector length. + */ + vq = MAX(arg2 / 16, 1); + vq = MIN(vq, 16); + env->vfp.smcr_el[1] = + FIELD_DP64(env->vfp.smcr_el[1], SMCR, LEN, vq - 1); + + /* Delay rebuilding hflags until we know if ZA must change. */ + vq = sve_vqm1_for_el_sm(env, 0, true) + 1; + + if (vq != old_vq) { + /* + * PSTATE.ZA state is cleared on any change to SVL. + * We need not call arm_rebuild_hflags because PSTATE.SM was + * cleared on syscall entry, so this hasn't changed VL. + */ + env->svcr = FIELD_DP64(env->svcr, SVCR, ZA, 0); + arm_rebuild_hflags(env); + } + return vq * 16; + } + return -TARGET_EINVAL; +} +#define do_prctl_sme_set_vl do_prctl_sme_set_vl + static abi_long do_prctl_reset_keys(CPUArchState *env, abi_long arg2) { ARMCPU *cpu = env_archcpu(env); diff --git a/linux-user/syscall.c b/linux-user/syscall.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/syscall.c +++ b/linux-user/syscall.c @@ -XXX,XX +XXX,XX @@ abi_long do_arch_prctl(CPUX86State *env, int code, abi_ulong addr) #ifndef PR_SET_SYSCALL_USER_DISPATCH # define PR_SET_SYSCALL_USER_DISPATCH 59 #endif +#ifndef PR_SME_SET_VL +# define PR_SME_SET_VL 63 +# define PR_SME_GET_VL 64 +# define PR_SME_VL_LEN_MASK 0xffff +# define PR_SME_VL_INHERIT (1 << 17) +#endif #include "target_prctl.h" @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl_inval1(CPUArchState *env, abi_long arg2) #ifndef do_prctl_set_unalign #define do_prctl_set_unalign do_prctl_inval1 #endif +#ifndef do_prctl_sme_get_vl +#define do_prctl_sme_get_vl do_prctl_inval0 +#endif +#ifndef do_prctl_sme_set_vl +#define do_prctl_sme_set_vl do_prctl_inval1 +#endif static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, abi_long arg3, abi_long arg4, abi_long arg5) @@ -XXX,XX +XXX,XX @@ static abi_long do_prctl(CPUArchState *env, abi_long option, abi_long arg2, return do_prctl_sve_get_vl(env); case PR_SVE_SET_VL: return do_prctl_sve_set_vl(env, arg2); + case PR_SME_GET_VL: + return do_prctl_sme_get_vl(env); + case PR_SME_SET_VL: + return do_prctl_sme_set_vl(env, arg2); case PR_PAC_RESET_KEYS: if (arg3 || arg4 || arg5) { return -TARGET_EINVAL; -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> There's no reason to set CPACR_EL1.ZEN if SVE disabled. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-44-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 7 +++---- 1 file changed, 3 insertions(+), 4 deletions(-) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) /* and to the FP/Neon instructions */ env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, CPACR_EL1, FPEN, 3); - /* and to the SVE instructions */ - env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, - CPACR_EL1, ZEN, 3); - /* with reasonable vector length */ + /* and to the SVE instructions, with default vector length */ if (cpu_isar_feature(aa64_sve, cpu)) { + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } /* -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Enable SME, TPIDR2_EL0, and FA64 if supported by the cpu. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-45-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 11 +++++++++++ 1 file changed, 11 insertions(+) diff --git a/target/arm/cpu.c b/target/arm/cpu.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/cpu.c +++ b/target/arm/cpu.c @@ -XXX,XX +XXX,XX @@ static void arm_cpu_reset(DeviceState *dev) CPACR_EL1, ZEN, 3); env->vfp.zcr_el[1] = cpu->sve_default_vq - 1; } + /* and for SME instructions, with default vector length, and TPIDR2 */ + if (cpu_isar_feature(aa64_sme, cpu)) { + env->cp15.sctlr_el[1] |= SCTLR_EnTP2; + env->cp15.cpacr_el1 = FIELD_DP64(env->cp15.cpacr_el1, + CPACR_EL1, SMEN, 3); + env->vfp.smcr_el[1] = cpu->sme_default_vq - 1; + if (cpu_isar_feature(aa64_sme_fa64, cpu)) { + env->vfp.smcr_el[1] = FIELD_DP64(env->vfp.smcr_el[1], + SMCR, FA64, 1); + } + } /* * Enable 48-bit address space (TODO: take reserved_va into account). * Enable TBI0 but not TBI1. -- 2.25.1 From: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20220708151540.18136-46-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- linux-user/elfload.c | 20 ++++++++++++++++++++ 1 file changed, 20 insertions(+) diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ enum { ARM_HWCAP2_A64_RNG = 1 << 16, ARM_HWCAP2_A64_BTI = 1 << 17, ARM_HWCAP2_A64_MTE = 1 << 18, + ARM_HWCAP2_A64_ECV = 1 << 19, + ARM_HWCAP2_A64_AFP = 1 << 20, + ARM_HWCAP2_A64_RPRES = 1 << 21, + ARM_HWCAP2_A64_MTE3 = 1 << 22, + ARM_HWCAP2_A64_SME = 1 << 23, + ARM_HWCAP2_A64_SME_I16I64 = 1 << 24, + ARM_HWCAP2_A64_SME_F64F64 = 1 << 25, + ARM_HWCAP2_A64_SME_I8I32 = 1 << 26, + ARM_HWCAP2_A64_SME_F16F32 = 1 << 27, + ARM_HWCAP2_A64_SME_B16F32 = 1 << 28, + ARM_HWCAP2_A64_SME_F32F32 = 1 << 29, + ARM_HWCAP2_A64_SME_FA64 = 1 << 30, }; #define ELF_HWCAP get_elf_hwcap() @@ -XXX,XX +XXX,XX @@ static uint32_t get_elf_hwcap2(void) GET_FEATURE_ID(aa64_rndr, ARM_HWCAP2_A64_RNG); GET_FEATURE_ID(aa64_bti, ARM_HWCAP2_A64_BTI); GET_FEATURE_ID(aa64_mte, ARM_HWCAP2_A64_MTE); + GET_FEATURE_ID(aa64_sme, (ARM_HWCAP2_A64_SME | + ARM_HWCAP2_A64_SME_F32F32 | + ARM_HWCAP2_A64_SME_B16F32 | + ARM_HWCAP2_A64_SME_F16F32 | + ARM_HWCAP2_A64_SME_I8I32)); + GET_FEATURE_ID(aa64_sme_f64f64, ARM_HWCAP2_A64_SME_F64F64); + GET_FEATURE_ID(aa64_sme_i16i64, ARM_HWCAP2_A64_SME_I16I64); + GET_FEATURE_ID(aa64_sme_fa64, ARM_HWCAP2_A64_SME_FA64); return hwcaps; } -- 2.25.1