The following changes since commit 7e7eb9f852a46b51a71ae9d82590b2e4d28827ee: Merge remote-tracking branch 'remotes/armbru/tags/pull-qapi-2021-01-28' into staging (2021-01-28 22:43:18 +0000) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20210129 for you to fetch changes up to 11749122e1a86866591306d43603d2795a3dea1a: hw/arm/stellaris: Remove board-creation reset of STELLARIS_SYS (2021-01-29 10:47:29 +0000) ---------------------------------------------------------------- target-arm queue: * Implement ID_PFR2 * Conditionalize DBGDIDR * rename xlnx-zcu102.canbusN properties * provide powerdown/reset mechanism for secure firmware on 'virt' board * hw/misc: Fix arith overflow in NPCM7XX PWM module * target/arm: Replace magic value by MMU_DATA_LOAD definition * configure: fix preadv errors on Catalina macOS with new XCode * Various configure and other cleanups in preparation for iOS support * hvf: Add hypervisor entitlement to output binaries (needed for Big Sur) * Implement pvpanic-pci device * Convert the CMSDK timer devices to the Clock framework ---------------------------------------------------------------- Alexander Graf (1): hvf: Add hypervisor entitlement to output binaries Hao Wu (1): hw/misc: Fix arith overflow in NPCM7XX PWM module Joelle van Dyne (7): configure: cross-compiling with empty cross_prefix osdep: build with non-working system() function darwin: remove redundant dependency declaration darwin: fix cross-compiling for Darwin configure: cross compile should use x86_64 cpu_family darwin: detect CoreAudio for build darwin: remove 64-bit build detection on 32-bit OS Maxim Uvarov (3): hw: gpio: implement gpio-pwr driver for qemu reset/poweroff arm-virt: refactor gpios creation arm-virt: add secure pl061 for reset/power down Mihai Carabas (4): hw/misc/pvpanic: split-out generic and bus dependent code hw/misc/pvpanic: add PCI interface support pvpanic : update pvpanic spec document tests/qtest: add a test case for pvpanic-pci Paolo Bonzini (1): arm: rename xlnx-zcu102.canbusN properties Peter Maydell (26): configure: Move preadv check to meson.build ptimer: Add new ptimer_set_period_from_clock() function clock: Add new clock_has_source() function tests: Add a simple test of the CMSDK APB timer tests: Add a simple test of the CMSDK APB watchdog tests: Add a simple test of the CMSDK APB dual timer hw/timer/cmsdk-apb-timer: Rename CMSDKAPBTIMER struct to CMSDKAPBTimer hw/timer/cmsdk-apb-timer: Add Clock input hw/timer/cmsdk-apb-dualtimer: Add Clock input hw/watchdog/cmsdk-apb-watchdog: Add Clock input hw/arm/armsse: Rename "MAINCLK" property to "MAINCLK_FRQ" hw/arm/armsse: Wire up clocks hw/arm/mps2: Inline CMSDK_APB_TIMER creation hw/arm/mps2: Create and connect SYSCLK Clock hw/arm/mps2-tz: Create and connect ARMSSE Clocks hw/arm/musca: Create and connect ARMSSE Clocks hw/arm/stellaris: Convert SSYS to QOM device hw/arm/stellaris: Create Clock input for watchdog hw/timer/cmsdk-apb-timer: Convert to use Clock input hw/timer/cmsdk-apb-dualtimer: Convert to use Clock input hw/watchdog/cmsdk-apb-watchdog: Convert to use Clock input tests/qtest/cmsdk-apb-watchdog-test: Test clock changes hw/arm/armsse: Use Clock to set system_clock_scale arm: Don't set freq properties on CMSDK timer, dualtimer, watchdog, ARMSSE arm: Remove frq properties on CMSDK timer, dualtimer, watchdog, ARMSSE hw/arm/stellaris: Remove board-creation reset of STELLARIS_SYS Philippe Mathieu-Daudé (1): target/arm: Replace magic value by MMU_DATA_LOAD definition Richard Henderson (2): target/arm: Implement ID_PFR2 target/arm: Conditionalize DBGDIDR docs/devel/clocks.rst | 16 +++ docs/specs/pci-ids.txt | 1 + docs/specs/pvpanic.txt | 13 ++- docs/system/arm/virt.rst | 2 + configure | 78 ++++++++------ meson.build | 34 ++++++- include/hw/arm/armsse.h | 14 ++- include/hw/arm/virt.h | 2 + include/hw/clock.h | 15 +++ include/hw/misc/pvpanic.h | 24 ++++- include/hw/pci/pci.h | 1 + include/hw/ptimer.h | 22 ++++ include/hw/timer/cmsdk-apb-dualtimer.h | 5 +- include/hw/timer/cmsdk-apb-timer.h | 34 ++----- include/hw/watchdog/cmsdk-apb-watchdog.h | 5 +- include/qemu/osdep.h | 12 +++ include/qemu/typedefs.h | 1 + target/arm/cpu.h | 1 + hw/arm/armsse.c | 48 ++++++--- hw/arm/mps2-tz.c | 14 ++- hw/arm/mps2.c | 28 ++++- hw/arm/musca.c | 13 ++- hw/arm/stellaris.c | 170 +++++++++++++++++++++++-------- hw/arm/virt.c | 111 ++++++++++++++++---- hw/arm/xlnx-zcu102.c | 4 +- hw/core/ptimer.c | 34 +++++++ hw/gpio/gpio_pwr.c | 70 +++++++++++++ hw/misc/npcm7xx_pwm.c | 23 ++++- hw/misc/pvpanic-isa.c | 94 +++++++++++++++++ hw/misc/pvpanic-pci.c | 94 +++++++++++++++++ hw/misc/pvpanic.c | 85 ++-------------- hw/timer/cmsdk-apb-dualtimer.c | 53 +++++++--- hw/timer/cmsdk-apb-timer.c | 55 +++++----- hw/watchdog/cmsdk-apb-watchdog.c | 29 ++++-- target/arm/helper.c | 27 +++-- target/arm/kvm64.c | 2 + tests/qtest/cmsdk-apb-dualtimer-test.c | 130 +++++++++++++++++++++++ tests/qtest/cmsdk-apb-timer-test.c | 75 ++++++++++++++ tests/qtest/cmsdk-apb-watchdog-test.c | 131 ++++++++++++++++++++++++ tests/qtest/npcm7xx_pwm-test.c | 4 +- tests/qtest/pvpanic-pci-test.c | 94 +++++++++++++++++ tests/qtest/xlnx-can-test.c | 30 +++--- MAINTAINERS | 3 + accel/hvf/entitlements.plist | 8 ++ hw/arm/Kconfig | 1 + hw/gpio/Kconfig | 3 + hw/gpio/meson.build | 1 + hw/i386/Kconfig | 2 +- hw/misc/Kconfig | 12 ++- hw/misc/meson.build | 4 +- scripts/entitlement.sh | 13 +++ tests/qtest/meson.build | 6 +- 52 files changed, 1432 insertions(+), 319 deletions(-) create mode 100644 hw/gpio/gpio_pwr.c create mode 100644 hw/misc/pvpanic-isa.c create mode 100644 hw/misc/pvpanic-pci.c create mode 100644 tests/qtest/cmsdk-apb-dualtimer-test.c create mode 100644 tests/qtest/cmsdk-apb-timer-test.c create mode 100644 tests/qtest/cmsdk-apb-watchdog-test.c create mode 100644 tests/qtest/pvpanic-pci-test.c create mode 100644 accel/hvf/entitlements.plist create mode 100755 scripts/entitlement.sh From: Richard Henderson <richard.henderson@linaro.org> This was defined at some point before ARMv8.4, and will shortly be used by new processor descriptions. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210120204400.1056582-1-richard.henderson@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.h | 1 + target/arm/helper.c | 4 ++-- target/arm/kvm64.c | 2 ++ 3 files changed, 5 insertions(+), 2 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 @@ struct ARMCPU { uint32_t id_mmfr4; uint32_t id_pfr0; uint32_t id_pfr1; + uint32_t id_pfr2; uint32_t mvfr0; uint32_t mvfr1; uint32_t mvfr2; 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 register_cp_regs_for_features(ARMCPU *cpu) .access = PL1_R, .type = ARM_CP_CONST, .accessfn = access_aa64_tid3, .resetvalue = 0 }, - { .name = "MVFR4_EL1_RESERVED", .state = ARM_CP_STATE_AA64, + { .name = "ID_PFR2", .state = ARM_CP_STATE_BOTH, .opc0 = 3, .opc1 = 0, .crn = 0, .crm = 3, .opc2 = 4, .access = PL1_R, .type = ARM_CP_CONST, .accessfn = access_aa64_tid3, - .resetvalue = 0 }, + .resetvalue = cpu->isar.id_pfr2 }, { .name = "MVFR5_EL1_RESERVED", .state = ARM_CP_STATE_AA64, .opc0 = 3, .opc1 = 0, .crn = 0, .crm = 3, .opc2 = 5, .access = PL1_R, .type = ARM_CP_CONST, diff --git a/target/arm/kvm64.c b/target/arm/kvm64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/kvm64.c +++ b/target/arm/kvm64.c @@ -XXX,XX +XXX,XX @@ bool kvm_arm_get_host_cpu_features(ARMHostCPUFeatures *ahcf) ARM64_SYS_REG(3, 0, 0, 1, 0)); err |= read_sys_reg32(fdarray[2], &ahcf->isar.id_pfr1, ARM64_SYS_REG(3, 0, 0, 1, 1)); + err |= read_sys_reg32(fdarray[2], &ahcf->isar.id_pfr2, + ARM64_SYS_REG(3, 0, 0, 3, 4)); err |= read_sys_reg32(fdarray[2], &ahcf->isar.id_dfr0, ARM64_SYS_REG(3, 0, 0, 1, 2)); err |= read_sys_reg32(fdarray[2], &ahcf->isar.id_mmfr0, -- 2.20.1 From: Richard Henderson <richard.henderson@linaro.org> Only define the register if it exists for the cpu. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20210120031656.737646-1-richard.henderson@linaro.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/helper.c | 21 +++++++++++++++------ 1 file changed, 15 insertions(+), 6 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 @@ static void define_debug_regs(ARMCPU *cpu) */ int i; int wrps, brps, ctx_cmps; - ARMCPRegInfo dbgdidr = { - .name = "DBGDIDR", .cp = 14, .crn = 0, .crm = 0, .opc1 = 0, .opc2 = 0, - .access = PL0_R, .accessfn = access_tda, - .type = ARM_CP_CONST, .resetvalue = cpu->isar.dbgdidr, - }; + + /* + * The Arm ARM says DBGDIDR is optional and deprecated if EL1 cannot + * use AArch32. Given that bit 15 is RES1, if the value is 0 then + * the register must not exist for this cpu. + */ + if (cpu->isar.dbgdidr != 0) { + ARMCPRegInfo dbgdidr = { + .name = "DBGDIDR", .cp = 14, .crn = 0, .crm = 0, + .opc1 = 0, .opc2 = 0, + .access = PL0_R, .accessfn = access_tda, + .type = ARM_CP_CONST, .resetvalue = cpu->isar.dbgdidr, + }; + define_one_arm_cp_reg(cpu, &dbgdidr); + } /* Note that all these register fields hold "number of Xs minus 1". */ brps = arm_num_brps(cpu); @@ -XXX,XX +XXX,XX @@ static void define_debug_regs(ARMCPU *cpu) assert(ctx_cmps <= brps); - define_one_arm_cp_reg(cpu, &dbgdidr); define_arm_cp_regs(cpu, debug_cp_reginfo); if (arm_feature(&cpu->env, ARM_FEATURE_LPAE)) { -- 2.20.1 From: Paolo Bonzini <pbonzini@redhat.com> The properties to attach a CANBUS object to the xlnx-zcu102 machine have a period in them. We want to use periods in properties for compound QAPI types, and besides the "xlnx-zcu102." prefix is both unnecessary and different from any other machine property name. Remove it. Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Message-id: 20210118162537.779542-1-pbonzini@redhat.com Reviewed-by: Vikram Garhwal <fnu.vikram@xilinx.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/xlnx-zcu102.c | 4 ++-- tests/qtest/xlnx-can-test.c | 30 +++++++++++++++--------------- 2 files changed, 17 insertions(+), 17 deletions(-) diff --git a/hw/arm/xlnx-zcu102.c b/hw/arm/xlnx-zcu102.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/xlnx-zcu102.c +++ b/hw/arm/xlnx-zcu102.c @@ -XXX,XX +XXX,XX @@ static void xlnx_zcu102_machine_instance_init(Object *obj) s->secure = false; /* Default to virt (EL2) being disabled */ s->virt = false; - object_property_add_link(obj, "xlnx-zcu102.canbus0", TYPE_CAN_BUS, + object_property_add_link(obj, "canbus0", TYPE_CAN_BUS, (Object **)&s->canbus[0], object_property_allow_set_link, 0); - object_property_add_link(obj, "xlnx-zcu102.canbus1", TYPE_CAN_BUS, + object_property_add_link(obj, "canbus1", TYPE_CAN_BUS, (Object **)&s->canbus[1], object_property_allow_set_link, 0); diff --git a/tests/qtest/xlnx-can-test.c b/tests/qtest/xlnx-can-test.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/xlnx-can-test.c +++ b/tests/qtest/xlnx-can-test.c @@ -XXX,XX +XXX,XX @@ static void test_can_bus(void) uint8_t can_timestamp = 1; QTestState *qts = qtest_init("-machine xlnx-zcu102" - " -object can-bus,id=canbus0" - " -machine xlnx-zcu102.canbus0=canbus0" - " -machine xlnx-zcu102.canbus1=canbus0" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" ); /* Configure the CAN0 and CAN1. */ @@ -XXX,XX +XXX,XX @@ static void test_can_loopback(void) uint32_t status = 0; QTestState *qts = qtest_init("-machine xlnx-zcu102" - " -object can-bus,id=canbus0" - " -machine xlnx-zcu102.canbus0=canbus0" - " -machine xlnx-zcu102.canbus1=canbus0" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" ); /* Configure the CAN0 in loopback mode. */ @@ -XXX,XX +XXX,XX @@ static void test_can_filter(void) uint8_t can_timestamp = 1; QTestState *qts = qtest_init("-machine xlnx-zcu102" - " -object can-bus,id=canbus0" - " -machine xlnx-zcu102.canbus0=canbus0" - " -machine xlnx-zcu102.canbus1=canbus0" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" ); /* Configure the CAN0 and CAN1. */ @@ -XXX,XX +XXX,XX @@ static void test_can_sleepmode(void) uint8_t can_timestamp = 1; QTestState *qts = qtest_init("-machine xlnx-zcu102" - " -object can-bus,id=canbus0" - " -machine xlnx-zcu102.canbus0=canbus0" - " -machine xlnx-zcu102.canbus1=canbus0" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" ); /* Configure the CAN0. */ @@ -XXX,XX +XXX,XX @@ static void test_can_snoopmode(void) uint8_t can_timestamp = 1; QTestState *qts = qtest_init("-machine xlnx-zcu102" - " -object can-bus,id=canbus0" - " -machine xlnx-zcu102.canbus0=canbus0" - " -machine xlnx-zcu102.canbus1=canbus0" + " -object can-bus,id=canbus" + " -machine canbus0=canbus" + " -machine canbus1=canbus" ); /* Configure the CAN0. */ -- 2.20.1 From: Maxim Uvarov <maxim.uvarov@linaro.org> Implement gpio-pwr driver to allow reboot and poweroff machine. This is simple driver with just 2 gpios lines. Current use case is to reboot and poweroff virt machine in secure mode. Secure pl066 gpio chip is needed for that. Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org> Reviewed-by: Hao Wu <wuhaotsh@google.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/gpio/gpio_pwr.c | 70 +++++++++++++++++++++++++++++++++++++++++++++ hw/gpio/Kconfig | 3 ++ hw/gpio/meson.build | 1 + 3 files changed, 74 insertions(+) create mode 100644 hw/gpio/gpio_pwr.c diff --git a/hw/gpio/gpio_pwr.c b/hw/gpio/gpio_pwr.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/gpio/gpio_pwr.c @@ -XXX,XX +XXX,XX @@ +/* + * GPIO qemu power controller + * + * Copyright (c) 2020 Linaro Limited + * + * Author: Maxim Uvarov <maxim.uvarov@linaro.org> + * + * Virtual gpio driver which can be used on top of pl061 + * to reboot and shutdown qemu virtual machine. One of use + * case is gpio driver for secure world application (ARM + * Trusted Firmware.). + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +/* + * QEMU interface: + * two named input GPIO lines: + * 'reset' : when asserted, trigger system reset + * 'shutdown' : when asserted, trigger system shutdown + */ + +#include "qemu/osdep.h" +#include "hw/sysbus.h" +#include "sysemu/runstate.h" + +#define TYPE_GPIOPWR "gpio-pwr" +OBJECT_DECLARE_SIMPLE_TYPE(GPIO_PWR_State, GPIOPWR) + +struct GPIO_PWR_State { + SysBusDevice parent_obj; +}; + +static void gpio_pwr_reset(void *opaque, int n, int level) +{ + if (level) { + qemu_system_reset_request(SHUTDOWN_CAUSE_GUEST_RESET); + } +} + +static void gpio_pwr_shutdown(void *opaque, int n, int level) +{ + if (level) { + qemu_system_reset_request(SHUTDOWN_CAUSE_GUEST_SHUTDOWN); + } +} + +static void gpio_pwr_init(Object *obj) +{ + DeviceState *dev = DEVICE(obj); + + qdev_init_gpio_in_named(dev, gpio_pwr_reset, "reset", 1); + qdev_init_gpio_in_named(dev, gpio_pwr_shutdown, "shutdown", 1); +} + +static const TypeInfo gpio_pwr_info = { + .name = TYPE_GPIOPWR, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(GPIO_PWR_State), + .instance_init = gpio_pwr_init, +}; + +static void gpio_pwr_register_types(void) +{ + type_register_static(&gpio_pwr_info); +} + +type_init(gpio_pwr_register_types) diff --git a/hw/gpio/Kconfig b/hw/gpio/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/Kconfig +++ b/hw/gpio/Kconfig @@ -XXX,XX +XXX,XX @@ config PL061 config GPIO_KEY bool +config GPIO_PWR + bool + config SIFIVE_GPIO bool diff --git a/hw/gpio/meson.build b/hw/gpio/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/gpio/meson.build +++ b/hw/gpio/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_E500', if_true: files('mpc8xxx.c')) softmmu_ss.add(when: 'CONFIG_GPIO_KEY', if_true: files('gpio_key.c')) +softmmu_ss.add(when: 'CONFIG_GPIO_PWR', if_true: files('gpio_pwr.c')) softmmu_ss.add(when: 'CONFIG_MAX7310', if_true: files('max7310.c')) softmmu_ss.add(when: 'CONFIG_PL061', if_true: files('pl061.c')) softmmu_ss.add(when: 'CONFIG_PUV3', if_true: files('puv3_gpio.c')) -- 2.20.1 From: Maxim Uvarov <maxim.uvarov@linaro.org> No functional change. Just refactor code to better support secure and normal world gpios. Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org> Reviewed-by: Andrew Jones <drjones@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/virt.c | 57 ++++++++++++++++++++++++++++++++------------------- 1 file changed, 36 insertions(+), 21 deletions(-) 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 void virt_powerdown_req(Notifier *n, void *opaque) } } -static void create_gpio(const VirtMachineState *vms) +static void create_gpio_keys(const VirtMachineState *vms, + DeviceState *pl061_dev, + uint32_t phandle) +{ + gpio_key_dev = sysbus_create_simple("gpio-key", -1, + qdev_get_gpio_in(pl061_dev, 3)); + + qemu_fdt_add_subnode(vms->fdt, "/gpio-keys"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-keys", "compatible", "gpio-keys"); + qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys", "#size-cells", 0); + qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys", "#address-cells", 1); + + qemu_fdt_add_subnode(vms->fdt, "/gpio-keys/poweroff"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-keys/poweroff", + "label", "GPIO Key Poweroff"); + qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys/poweroff", "linux,code", + KEY_POWER); + qemu_fdt_setprop_cells(vms->fdt, "/gpio-keys/poweroff", + "gpios", phandle, 3, 0); +} + +static void create_gpio_devices(const VirtMachineState *vms, int gpio, + MemoryRegion *mem) { char *nodename; DeviceState *pl061_dev; - hwaddr base = vms->memmap[VIRT_GPIO].base; - hwaddr size = vms->memmap[VIRT_GPIO].size; - int irq = vms->irqmap[VIRT_GPIO]; + hwaddr base = vms->memmap[gpio].base; + hwaddr size = vms->memmap[gpio].size; + int irq = vms->irqmap[gpio]; const char compat[] = "arm,pl061\0arm,primecell"; + SysBusDevice *s; - pl061_dev = sysbus_create_simple("pl061", base, - qdev_get_gpio_in(vms->gic, irq)); + pl061_dev = qdev_new("pl061"); + s = SYS_BUS_DEVICE(pl061_dev); + sysbus_realize_and_unref(s, &error_fatal); + memory_region_add_subregion(mem, base, sysbus_mmio_get_region(s, 0)); + sysbus_connect_irq(s, 0, qdev_get_gpio_in(vms->gic, irq)); uint32_t phandle = qemu_fdt_alloc_phandle(vms->fdt); nodename = g_strdup_printf("/pl061@%" PRIx64, base); @@ -XXX,XX +XXX,XX @@ static void create_gpio(const VirtMachineState *vms) qemu_fdt_setprop_string(vms->fdt, nodename, "clock-names", "apb_pclk"); qemu_fdt_setprop_cell(vms->fdt, nodename, "phandle", phandle); - gpio_key_dev = sysbus_create_simple("gpio-key", -1, - qdev_get_gpio_in(pl061_dev, 3)); - qemu_fdt_add_subnode(vms->fdt, "/gpio-keys"); - qemu_fdt_setprop_string(vms->fdt, "/gpio-keys", "compatible", "gpio-keys"); - qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys", "#size-cells", 0); - qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys", "#address-cells", 1); - - qemu_fdt_add_subnode(vms->fdt, "/gpio-keys/poweroff"); - qemu_fdt_setprop_string(vms->fdt, "/gpio-keys/poweroff", - "label", "GPIO Key Poweroff"); - qemu_fdt_setprop_cell(vms->fdt, "/gpio-keys/poweroff", "linux,code", - KEY_POWER); - qemu_fdt_setprop_cells(vms->fdt, "/gpio-keys/poweroff", - "gpios", phandle, 3, 0); g_free(nodename); + + /* Child gpio devices */ + create_gpio_keys(vms, pl061_dev, phandle); } static void create_virtio_devices(const VirtMachineState *vms) @@ -XXX,XX +XXX,XX @@ static void machvirt_init(MachineState *machine) if (has_ged && aarch64 && firmware_loaded && virt_is_acpi_enabled(vms)) { vms->acpi_dev = create_acpi_ged(vms); } else { - create_gpio(vms); + create_gpio_devices(vms, VIRT_GPIO, sysmem); } /* connect powerdown request */ -- 2.20.1 From: Maxim Uvarov <maxim.uvarov@linaro.org> Add secure pl061 for reset/power down machine from the secure world (Arm Trusted Firmware). Connect it with gpio-pwr driver. Signed-off-by: Maxim Uvarov <maxim.uvarov@linaro.org> Reviewed-by: Andrew Jones <drjones@redhat.com> [PMM: Added mention of the new device to the documentation] Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/system/arm/virt.rst | 2 ++ include/hw/arm/virt.h | 2 ++ hw/arm/virt.c | 56 +++++++++++++++++++++++++++++++++++++++- hw/arm/Kconfig | 1 + 4 files changed, 60 insertions(+), 1 deletion(-) diff --git a/docs/system/arm/virt.rst b/docs/system/arm/virt.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/system/arm/virt.rst +++ b/docs/system/arm/virt.rst @@ -XXX,XX +XXX,XX @@ The virt board supports: - Secure-World-only devices if the CPU has TrustZone: - A second PL011 UART + - A second PL061 GPIO controller, with GPIO lines for triggering + a system reset or system poweroff - A secure flash memory - 16MB of secure RAM diff --git a/include/hw/arm/virt.h b/include/hw/arm/virt.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/virt.h +++ b/include/hw/arm/virt.h @@ -XXX,XX +XXX,XX @@ enum { VIRT_GPIO, VIRT_SECURE_UART, VIRT_SECURE_MEM, + VIRT_SECURE_GPIO, VIRT_PCDIMM_ACPI, VIRT_ACPI_GED, VIRT_NVDIMM_ACPI, @@ -XXX,XX +XXX,XX @@ struct VirtMachineClass { bool kvm_no_adjvtime; bool no_kvm_steal_time; bool acpi_expose_flash; + bool no_secure_gpio; }; struct VirtMachineState { 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 MemMapEntry base_memmap[] = { [VIRT_ACPI_GED] = { 0x09080000, ACPI_GED_EVT_SEL_LEN }, [VIRT_NVDIMM_ACPI] = { 0x09090000, NVDIMM_ACPI_IO_LEN}, [VIRT_PVTIME] = { 0x090a0000, 0x00010000 }, + [VIRT_SECURE_GPIO] = { 0x090b0000, 0x00001000 }, [VIRT_MMIO] = { 0x0a000000, 0x00000200 }, /* ...repeating for a total of NUM_VIRTIO_TRANSPORTS, each of that size */ [VIRT_PLATFORM_BUS] = { 0x0c000000, 0x02000000 }, @@ -XXX,XX +XXX,XX @@ static void create_gpio_keys(const VirtMachineState *vms, "gpios", phandle, 3, 0); } +#define SECURE_GPIO_POWEROFF 0 +#define SECURE_GPIO_RESET 1 + +static void create_secure_gpio_pwr(const VirtMachineState *vms, + DeviceState *pl061_dev, + uint32_t phandle) +{ + DeviceState *gpio_pwr_dev; + + /* gpio-pwr */ + gpio_pwr_dev = sysbus_create_simple("gpio-pwr", -1, NULL); + + /* connect secure pl061 to gpio-pwr */ + qdev_connect_gpio_out(pl061_dev, SECURE_GPIO_RESET, + qdev_get_gpio_in_named(gpio_pwr_dev, "reset", 0)); + qdev_connect_gpio_out(pl061_dev, SECURE_GPIO_POWEROFF, + qdev_get_gpio_in_named(gpio_pwr_dev, "shutdown", 0)); + + qemu_fdt_add_subnode(vms->fdt, "/gpio-poweroff"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-poweroff", "compatible", + "gpio-poweroff"); + qemu_fdt_setprop_cells(vms->fdt, "/gpio-poweroff", + "gpios", phandle, SECURE_GPIO_POWEROFF, 0); + qemu_fdt_setprop_string(vms->fdt, "/gpio-poweroff", "status", "disabled"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-poweroff", "secure-status", + "okay"); + + qemu_fdt_add_subnode(vms->fdt, "/gpio-restart"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-restart", "compatible", + "gpio-restart"); + qemu_fdt_setprop_cells(vms->fdt, "/gpio-restart", + "gpios", phandle, SECURE_GPIO_RESET, 0); + qemu_fdt_setprop_string(vms->fdt, "/gpio-restart", "status", "disabled"); + qemu_fdt_setprop_string(vms->fdt, "/gpio-restart", "secure-status", + "okay"); +} + static void create_gpio_devices(const VirtMachineState *vms, int gpio, MemoryRegion *mem) { @@ -XXX,XX +XXX,XX @@ static void create_gpio_devices(const VirtMachineState *vms, int gpio, qemu_fdt_setprop_string(vms->fdt, nodename, "clock-names", "apb_pclk"); qemu_fdt_setprop_cell(vms->fdt, nodename, "phandle", phandle); + if (gpio != VIRT_GPIO) { + /* Mark as not usable by the normal world */ + qemu_fdt_setprop_string(vms->fdt, nodename, "status", "disabled"); + qemu_fdt_setprop_string(vms->fdt, nodename, "secure-status", "okay"); + } g_free(nodename); /* Child gpio devices */ - create_gpio_keys(vms, pl061_dev, phandle); + if (gpio == VIRT_GPIO) { + create_gpio_keys(vms, pl061_dev, phandle); + } else { + create_secure_gpio_pwr(vms, pl061_dev, phandle); + } } static void create_virtio_devices(const VirtMachineState *vms) @@ -XXX,XX +XXX,XX @@ static void machvirt_init(MachineState *machine) create_gpio_devices(vms, VIRT_GPIO, sysmem); } + if (vms->secure && !vmc->no_secure_gpio) { + create_gpio_devices(vms, VIRT_SECURE_GPIO, secure_sysmem); + } + /* connect powerdown request */ vms->powerdown_notifier.notify = virt_powerdown_req; qemu_register_powerdown_notifier(&vms->powerdown_notifier); @@ -XXX,XX +XXX,XX @@ DEFINE_VIRT_MACHINE_AS_LATEST(6, 0) static void virt_machine_5_2_options(MachineClass *mc) { + VirtMachineClass *vmc = VIRT_MACHINE_CLASS(OBJECT_CLASS(mc)); + virt_machine_6_0_options(mc); compat_props_add(mc->compat_props, hw_compat_5_2, hw_compat_5_2_len); + vmc->no_secure_gpio = true; } DEFINE_VIRT_MACHINE(5, 2) 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 ARM_VIRT select PL011 # UART select PL031 # RTC select PL061 # GPIO + select GPIO_PWR select PLATFORM_BUS select SMBIOS select VIRTIO_MMIO -- 2.20.1 From: Hao Wu <wuhaotsh@google.com> Fix potential overflow problem when calculating pwm_duty. 1. Ensure p->cmr and p->cnr to be from [0,65535], according to the hardware specification. 2. Changed duty to uint32_t. However, since MAX_DUTY * (p->cmr+1) can excceed UINT32_MAX, we convert them to uint64_t in computation and converted them back to uint32_t. (duty is guaranteed to be <= MAX_DUTY so it won't overflow.) Fixes: CID 1442342 Suggested-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Doug Evans <dje@google.com> Signed-off-by: Hao Wu <wuhaotsh@google.com> Message-id: 20210127011142.2122790-1-wuhaotsh@google.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/misc/npcm7xx_pwm.c | 23 +++++++++++++++++++---- tests/qtest/npcm7xx_pwm-test.c | 4 ++-- 2 files changed, 21 insertions(+), 6 deletions(-) diff --git a/hw/misc/npcm7xx_pwm.c b/hw/misc/npcm7xx_pwm.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/npcm7xx_pwm.c +++ b/hw/misc/npcm7xx_pwm.c @@ -XXX,XX +XXX,XX @@ REG32(NPCM7XX_PWM_PWDR3, 0x50); #define NPCM7XX_CH_INV BIT(2) #define NPCM7XX_CH_MOD BIT(3) +#define NPCM7XX_MAX_CMR 65535 +#define NPCM7XX_MAX_CNR 65535 + /* Offset of each PWM channel's prescaler in the PPR register. */ static const int npcm7xx_ppr_base[] = { 0, 0, 8, 8 }; /* Offset of each PWM channel's clock selector in the CSR register. */ @@ -XXX,XX +XXX,XX @@ static uint32_t npcm7xx_pwm_calculate_freq(NPCM7xxPWM *p) static uint32_t npcm7xx_pwm_calculate_duty(NPCM7xxPWM *p) { - uint64_t duty; + uint32_t duty; if (p->running) { if (p->cnr == 0) { @@ -XXX,XX +XXX,XX @@ static uint32_t npcm7xx_pwm_calculate_duty(NPCM7xxPWM *p) } else if (p->cmr >= p->cnr) { duty = NPCM7XX_PWM_MAX_DUTY; } else { - duty = NPCM7XX_PWM_MAX_DUTY * (p->cmr + 1) / (p->cnr + 1); + duty = (uint64_t)NPCM7XX_PWM_MAX_DUTY * (p->cmr + 1) / (p->cnr + 1); } } else { duty = 0; @@ -XXX,XX +XXX,XX @@ static void npcm7xx_pwm_write(void *opaque, hwaddr offset, case A_NPCM7XX_PWM_CNR2: case A_NPCM7XX_PWM_CNR3: p = &s->pwm[npcm7xx_cnr_index(offset)]; - p->cnr = value; + if (value > NPCM7XX_MAX_CNR) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid cnr value: %u", __func__, value); + p->cnr = NPCM7XX_MAX_CNR; + } else { + p->cnr = value; + } npcm7xx_pwm_update_output(p); break; @@ -XXX,XX +XXX,XX @@ static void npcm7xx_pwm_write(void *opaque, hwaddr offset, case A_NPCM7XX_PWM_CMR2: case A_NPCM7XX_PWM_CMR3: p = &s->pwm[npcm7xx_cmr_index(offset)]; - p->cmr = value; + if (value > NPCM7XX_MAX_CMR) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: invalid cmr value: %u", __func__, value); + p->cmr = NPCM7XX_MAX_CMR; + } else { + p->cmr = value; + } npcm7xx_pwm_update_output(p); break; diff --git a/tests/qtest/npcm7xx_pwm-test.c b/tests/qtest/npcm7xx_pwm-test.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/npcm7xx_pwm-test.c +++ b/tests/qtest/npcm7xx_pwm-test.c @@ -XXX,XX +XXX,XX @@ static uint64_t pwm_compute_freq(QTestState *qts, uint32_t ppr, uint32_t csr, static uint64_t pwm_compute_duty(uint32_t cnr, uint32_t cmr, bool inverted) { - uint64_t duty; + uint32_t duty; if (cnr == 0) { /* PWM is stopped. */ @@ -XXX,XX +XXX,XX @@ static uint64_t pwm_compute_duty(uint32_t cnr, uint32_t cmr, bool inverted) } else if (cmr >= cnr) { duty = MAX_DUTY; } else { - duty = MAX_DUTY * (cmr + 1) / (cnr + 1); + duty = (uint64_t)MAX_DUTY * (cmr + 1) / (cnr + 1); } if (inverted) { -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> cpu_get_phys_page_debug() uses 'DATA LOAD' MMU access type. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210127232822.3530782-1-f4bug@amsat.org 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, 1 insertion(+), 1 deletion(-) 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 @@ hwaddr arm_cpu_get_phys_page_attrs_debug(CPUState *cs, vaddr addr, *attrs = (MemTxAttrs) {}; - ret = get_phys_addr(env, addr, 0, mmu_idx, &phys_addr, + ret = get_phys_addr(env, addr, MMU_DATA_LOAD, mmu_idx, &phys_addr, attrs, &prot, &page_size, &fi, &cacheattrs); if (ret) { -- 2.20.1 Move the preadv availability check to meson.build. This is what we want to be doing for host-OS-feature-checks anyway, but it also fixes a problem with building for macOS with the most recent XCode SDK on a Catalina host. On that configuration, 'preadv()' is provided as a weak symbol, so that programs can be built with optional support for it and make a runtime availability check to see whether the preadv() they have is a working one or one which they must not call because it will runtime-assert. QEMU's configure test passes (unless you're building with --enable-werror) because the test program using preadv() compiles, but then QEMU crashes at runtime when preadv() is called, with errors like: dyld: lazy symbol binding failed: Symbol not found: _preadv Referenced from: /Users/pm215/src/qemu/./build/x86/tests/test-replication Expected in: /usr/lib/libSystem.B.dylib dyld: Symbol not found: _preadv Referenced from: /Users/pm215/src/qemu/./build/x86/tests/test-replication Expected in: /usr/lib/libSystem.B.dylib Meson's own function availability check has a special case for macOS which adds '-Wl,-no_weak_imports' to the compiler flags, which forces the test to require the real function, not the macOS-version-too-old stub. So this commit fixes the bug where macOS builds on Catalina currently require --disable-werror. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20210126155846.17109-1-peter.maydell@linaro.org --- configure | 16 ---------------- meson.build | 4 +++- 2 files changed, 3 insertions(+), 17 deletions(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ if compile_prog "" "" ; then iovec=yes fi -########################################## -# preadv probe -cat > $TMPC <<EOF -#include <sys/types.h> -#include <sys/uio.h> -#include <unistd.h> -int main(void) { return preadv(0, 0, 0, 0); } -EOF -preadv=no -if compile_prog "" "" ; then - preadv=yes -fi - ########################################## # fdt probe @@ -XXX,XX +XXX,XX @@ fi if test "$iovec" = "yes" ; then echo "CONFIG_IOVEC=y" >> $config_host_mak fi -if test "$preadv" = "yes" ; then - echo "CONFIG_PREADV=y" >> $config_host_mak -fi if test "$membarrier" = "yes" ; then echo "CONFIG_MEMBARRIER=y" >> $config_host_mak fi diff --git a/meson.build b/meson.build index XXXXXXX..XXXXXXX 100644 --- a/meson.build +++ b/meson.build @@ -XXX,XX +XXX,XX @@ config_host_data.set('HAVE_PTY_H', cc.has_header('pty.h')) config_host_data.set('HAVE_SYS_IOCCOM_H', cc.has_header('sys/ioccom.h')) config_host_data.set('HAVE_SYS_KCOV_H', cc.has_header('sys/kcov.h')) +config_host_data.set('CONFIG_PREADV', cc.has_function('preadv', prefix: '#include <sys/uio.h>')) + ignored = ['CONFIG_QEMU_INTERP_PREFIX'] # actually per-target arrays = ['CONFIG_AUDIO_DRIVERS', 'CONFIG_BDRV_RW_WHITELIST', 'CONFIG_BDRV_RO_WHITELIST'] strings = ['HOST_DSOSUF', 'CONFIG_IASL'] @@ -XXX,XX +XXX,XX @@ summary_info += {'PIE': get_option('b_pie')} summary_info += {'static build': config_host.has_key('CONFIG_STATIC')} summary_info += {'malloc trim support': has_malloc_trim} summary_info += {'membarrier': config_host.has_key('CONFIG_MEMBARRIER')} -summary_info += {'preadv support': config_host.has_key('CONFIG_PREADV')} +summary_info += {'preadv support': config_host_data.get('CONFIG_PREADV')} summary_info += {'fdatasync': config_host.has_key('CONFIG_FDATASYNC')} summary_info += {'madvise': config_host.has_key('CONFIG_MADVISE')} summary_info += {'posix_madvise': config_host.has_key('CONFIG_POSIX_MADVISE')} -- 2.20.1 From: Joelle van Dyne <j@getutm.app> The iOS toolchain does not use the host prefix naming convention. So we need to enable cross-compile options while allowing the PREFIX to be blank. Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Joelle van Dyne <j@getutm.app> Message-id: 20210126012457.39046-3-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ cpu="" iasl="iasl" interp_prefix="/usr/gnemul/qemu-%M" static="no" +cross_compile="no" cross_prefix="" audio_drv_list="" block_drv_rw_whitelist="" @@ -XXX,XX +XXX,XX @@ for opt do optarg=$(expr "x$opt" : 'x[^=]*=\(.*\)') case "$opt" in --cross-prefix=*) cross_prefix="$optarg" + cross_compile="yes" ;; --cc=*) CC="$optarg" ;; @@ -XXX,XX +XXX,XX @@ $(echo Deprecated targets: $deprecated_targets_list | \ --target-list-exclude=LIST exclude a set of targets from the default target-list Advanced options (experts only): - --cross-prefix=PREFIX use PREFIX for compile tools [$cross_prefix] + --cross-prefix=PREFIX use PREFIX for compile tools, PREFIX can be blank [$cross_prefix] --cc=CC use C compiler CC [$cc] --iasl=IASL use ACPI compiler IASL [$iasl] --host-cc=CC use C compiler CC [$host_cc] for code run at @@ -XXX,XX +XXX,XX @@ if has $sdl2_config; then fi echo "strip = [$(meson_quote $strip)]" >> $cross echo "windres = [$(meson_quote $windres)]" >> $cross -if test -n "$cross_prefix"; then +if test "$cross_compile" = "yes"; then cross_arg="--cross-file config-meson.cross" echo "[host_machine]" >> $cross if test "$mingw32" = "yes" ; then -- 2.20.1 From: Joelle van Dyne <j@getutm.app> Build without error on hosts without a working system(). If system() is called, return -1 with ENOSYS. Signed-off-by: Joelle van Dyne <j@getutm.app> Message-id: 20210126012457.39046-6-j@getutm.app Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- meson.build | 1 + include/qemu/osdep.h | 12 ++++++++++++ 2 files changed, 13 insertions(+) diff --git a/meson.build b/meson.build index XXXXXXX..XXXXXXX 100644 --- a/meson.build +++ b/meson.build @@ -XXX,XX +XXX,XX @@ config_host_data.set('HAVE_DRM_H', cc.has_header('libdrm/drm.h')) config_host_data.set('HAVE_PTY_H', cc.has_header('pty.h')) config_host_data.set('HAVE_SYS_IOCCOM_H', cc.has_header('sys/ioccom.h')) config_host_data.set('HAVE_SYS_KCOV_H', cc.has_header('sys/kcov.h')) +config_host_data.set('HAVE_SYSTEM_FUNCTION', cc.has_function('system', prefix: '#include <stdlib.h>')) config_host_data.set('CONFIG_PREADV', cc.has_function('preadv', prefix: '#include <sys/uio.h>')) diff --git a/include/qemu/osdep.h b/include/qemu/osdep.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/osdep.h +++ b/include/qemu/osdep.h @@ -XXX,XX +XXX,XX @@ static inline void qemu_thread_jit_write(void) {} static inline void qemu_thread_jit_execute(void) {} #endif +/** + * Platforms which do not support system() return ENOSYS + */ +#ifndef HAVE_SYSTEM_FUNCTION +#define system platform_does_not_support_system +static inline int platform_does_not_support_system(const char *command) +{ + errno = ENOSYS; + return -1; +} +#endif /* !HAVE_SYSTEM_FUNCTION */ + #endif -- 2.20.1 From: Joelle van Dyne <j@getutm.app> Meson will find CoreFoundation, IOKit, and Cocoa as needed. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Joelle van Dyne <j@getutm.app> Message-id: 20210126012457.39046-7-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 1 - 1 file changed, 1 deletion(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ Darwin) fi audio_drv_list="coreaudio try-sdl" audio_possible_drivers="coreaudio sdl" - QEMU_LDFLAGS="-framework CoreFoundation -framework IOKit $QEMU_LDFLAGS" # Disable attempts to use ObjectiveC features in os/object.h since they # won't work when we're compiling with gcc as a C compiler. QEMU_CFLAGS="-DOS_OBJECT_USE_OBJC=0 $QEMU_CFLAGS" -- 2.20.1 From: Joelle van Dyne <j@getutm.app> Add objc to the Meson cross file as well as detection of Darwin. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Joelle van Dyne <j@getutm.app> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210126012457.39046-8-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ echo "cpp_link_args = [${LDFLAGS:+$(meson_quote $LDFLAGS)}]" >> $cross echo "[binaries]" >> $cross echo "c = [$(meson_quote $cc)]" >> $cross test -n "$cxx" && echo "cpp = [$(meson_quote $cxx)]" >> $cross +test -n "$objcc" && echo "objc = [$(meson_quote $objcc)]" >> $cross echo "ar = [$(meson_quote $ar)]" >> $cross echo "nm = [$(meson_quote $nm)]" >> $cross echo "pkgconfig = [$(meson_quote $pkg_config_exe)]" >> $cross @@ -XXX,XX +XXX,XX @@ if test "$cross_compile" = "yes"; then if test "$linux" = "yes" ; then echo "system = 'linux'" >> $cross fi + if test "$darwin" = "yes" ; then + echo "system = 'darwin'" >> $cross + fi case "$ARCH" in i386|x86_64) echo "cpu_family = 'x86'" >> $cross -- 2.20.1 From: Joelle van Dyne <j@getutm.app> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Joelle van Dyne <j@getutm.app> Message-id: 20210126012457.39046-9-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ if test "$cross_compile" = "yes"; then echo "system = 'darwin'" >> $cross fi case "$ARCH" in - i386|x86_64) + i386) echo "cpu_family = 'x86'" >> $cross ;; + x86_64) + echo "cpu_family = 'x86_64'" >> $cross + ;; ppc64le) echo "cpu_family = 'ppc64'" >> $cross ;; -- 2.20.1 From: Joelle van Dyne <j@getutm.app> On iOS there is no CoreAudio, so we should not assume Darwin always has it. Signed-off-by: Joelle van Dyne <j@getutm.app> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210126012457.39046-11-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 35 +++++++++++++++++++++++++++++++++-- 1 file changed, 33 insertions(+), 2 deletions(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ fdt="auto" netmap="no" sdl="auto" sdl_image="auto" +coreaudio="auto" virtiofsd="auto" virtfs="auto" libudev="auto" @@ -XXX,XX +XXX,XX @@ Darwin) QEMU_CFLAGS="-arch x86_64 $QEMU_CFLAGS" QEMU_LDFLAGS="-arch x86_64 $QEMU_LDFLAGS" fi - audio_drv_list="coreaudio try-sdl" + audio_drv_list="try-coreaudio try-sdl" audio_possible_drivers="coreaudio sdl" # Disable attempts to use ObjectiveC features in os/object.h since they # won't work when we're compiling with gcc as a C compiler. @@ -XXX,XX +XXX,XX @@ EOF fi fi +########################################## +# detect CoreAudio +if test "$coreaudio" != "no" ; then + coreaudio_libs="-framework CoreAudio" + cat > $TMPC << EOF +#include <CoreAudio/CoreAudio.h> +int main(void) +{ + return (int)AudioGetCurrentHostTime(); +} +EOF + if compile_prog "" "$coreaudio_libs" ; then + coreaudio=yes + else + coreaudio=no + fi +fi + ########################################## # Sound support libraries probe @@ -XXX,XX +XXX,XX @@ for drv in $audio_drv_list; do fi ;; - coreaudio) + coreaudio | try-coreaudio) + if test "$coreaudio" = "no"; then + if test "$drv" = "try-coreaudio"; then + audio_drv_list=$(echo "$audio_drv_list" | sed -e 's/try-coreaudio//') + else + error_exit "$drv check failed" \ + "Make sure to have the $drv is available." + fi + else coreaudio_libs="-framework CoreAudio" + if test "$drv" = "try-coreaudio"; then + audio_drv_list=$(echo "$audio_drv_list" | sed -e 's/try-coreaudio/coreaudio/') + fi + fi ;; dsound) -- 2.20.1 From: Joelle van Dyne <j@getutm.app> A workaround added in early days of 64-bit OSX forced x86_64 if the host machine had 64-bit support. This creates issues when cross- compiling for ARM64. Additionally, the user can always use --cpu=* to manually set the host CPU and therefore this workaround should be removed. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Joelle van Dyne <j@getutm.app> Message-id: 20210126012457.39046-12-j@getutm.app Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- configure | 11 ----------- 1 file changed, 11 deletions(-) diff --git a/configure b/configure index XXXXXXX..XXXXXXX 100755 --- a/configure +++ b/configure @@ -XXX,XX +XXX,XX @@ fi # the correct CPU with the --cpu option. case $targetos in Darwin) - # on Leopard most of the system is 32-bit, so we have to ask the kernel if we can - # run 64-bit userspace code. - # If the user didn't specify a CPU explicitly and the kernel says this is - # 64 bit hw, then assume x86_64. Otherwise fall through to the usual detection code. - if test -z "$cpu" && test "$(sysctl -n hw.optional.x86_64)" = "1"; then - cpu="x86_64" - fi HOST_DSOSUF=".dylib" ;; SunOS) @@ -XXX,XX +XXX,XX @@ OpenBSD) Darwin) bsd="yes" darwin="yes" - if [ "$cpu" = "x86_64" ] ; then - QEMU_CFLAGS="-arch x86_64 $QEMU_CFLAGS" - QEMU_LDFLAGS="-arch x86_64 $QEMU_LDFLAGS" - fi audio_drv_list="try-coreaudio try-sdl" audio_possible_drivers="coreaudio sdl" # Disable attempts to use ObjectiveC features in os/object.h since they -- 2.20.1 From: Alexander Graf <agraf@csgraf.de> In macOS 11, QEMU only gets access to Hypervisor.framework if it has the respective entitlement. Add an entitlement template and automatically self sign and apply the entitlement in the build. Signed-off-by: Alexander Graf <agraf@csgraf.de> Reviewed-by: Roman Bolshakov <r.bolshakov@yadro.com> Tested-by: Roman Bolshakov <r.bolshakov@yadro.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- meson.build | 29 +++++++++++++++++++++++++---- accel/hvf/entitlements.plist | 8 ++++++++ scripts/entitlement.sh | 13 +++++++++++++ 3 files changed, 46 insertions(+), 4 deletions(-) create mode 100644 accel/hvf/entitlements.plist create mode 100755 scripts/entitlement.sh diff --git a/meson.build b/meson.build index XXXXXXX..XXXXXXX 100644 --- a/meson.build +++ b/meson.build @@ -XXX,XX +XXX,XX @@ foreach target : target_dirs }] endif foreach exe: execs - emulators += {exe['name']: - executable(exe['name'], exe['sources'], - install: true, + exe_name = exe['name'] + exe_sign = 'CONFIG_HVF' in config_target + if exe_sign + exe_name += '-unsigned' + endif + + emulator = executable(exe_name, exe['sources'], + install: not exe_sign, c_args: c_args, dependencies: arch_deps + deps + exe['dependencies'], objects: lib.extract_all_objects(recursive: true), @@ -XXX,XX +XXX,XX @@ foreach target : target_dirs link_depends: [block_syms, qemu_syms] + exe.get('link_depends', []), link_args: link_args, gui_app: exe['gui']) - } + + if exe_sign + emulators += {exe['name'] : custom_target(exe['name'], + install: true, + install_dir: get_option('bindir'), + depends: emulator, + output: exe['name'], + command: [ + meson.current_source_dir() / 'scripts/entitlement.sh', + meson.current_build_dir() / exe_name, + meson.current_build_dir() / exe['name'], + meson.current_source_dir() / 'accel/hvf/entitlements.plist' + ]) + } + else + emulators += {exe['name']: emulator} + endif if 'CONFIG_TRACE_SYSTEMTAP' in config_host foreach stp: [ diff --git a/accel/hvf/entitlements.plist b/accel/hvf/entitlements.plist new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/accel/hvf/entitlements.plist @@ -XXX,XX +XXX,XX @@ +<?xml version="1.0" encoding="UTF-8"?> +<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd"> +<plist version="1.0"> +<dict> + <key>com.apple.security.hypervisor</key> + <true/> +</dict> +</plist> diff --git a/scripts/entitlement.sh b/scripts/entitlement.sh new file mode 100755 index XXXXXXX..XXXXXXX --- /dev/null +++ b/scripts/entitlement.sh @@ -XXX,XX +XXX,XX @@ +#!/bin/sh -e +# +# Helper script for the build process to apply entitlements + +SRC="$1" +DST="$2" +ENTITLEMENT="$3" + +trap 'rm "$DST.tmp"' exit +cp -af "$SRC" "$DST.tmp" +codesign --entitlements "$ENTITLEMENT" --force -s - "$DST.tmp" +mv "$DST.tmp" "$DST" +trap '' exit -- 2.20.1 From: Mihai Carabas <mihai.carabas@oracle.com> To ease the PCI device addition in next patches, split the code as follows: - generic code (read/write/setup) is being kept in pvpanic.c - ISA dependent code moved to pvpanic-isa.c Also, rename: - ISA_PVPANIC_DEVICE -> PVPANIC_ISA_DEVICE. - TYPE_PVPANIC -> TYPE_PVPANIC_ISA. - MemoryRegion io -> mr. - pvpanic_ioport_* in pvpanic_*. Update the build system with the new files and config structure. Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/misc/pvpanic.h | 23 +++++++++- hw/misc/pvpanic-isa.c | 94 +++++++++++++++++++++++++++++++++++++++ hw/misc/pvpanic.c | 85 +++-------------------------------- hw/i386/Kconfig | 2 +- hw/misc/Kconfig | 6 ++- hw/misc/meson.build | 3 +- tests/qtest/meson.build | 2 +- 7 files changed, 130 insertions(+), 85 deletions(-) create mode 100644 hw/misc/pvpanic-isa.c diff --git a/include/hw/misc/pvpanic.h b/include/hw/misc/pvpanic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/pvpanic.h +++ b/include/hw/misc/pvpanic.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" -#define TYPE_PVPANIC "pvpanic" +#define TYPE_PVPANIC_ISA_DEVICE "pvpanic" #define PVPANIC_IOPORT_PROP "ioport" +/* The bit of supported pv event, TODO: include uapi header and remove this */ +#define PVPANIC_F_PANICKED 0 +#define PVPANIC_F_CRASHLOADED 1 + +/* The pv event value */ +#define PVPANIC_PANICKED (1 << PVPANIC_F_PANICKED) +#define PVPANIC_CRASHLOADED (1 << PVPANIC_F_CRASHLOADED) + +/* + * PVPanicState for any device type + */ +typedef struct PVPanicState PVPanicState; +struct PVPanicState { + MemoryRegion mr; + uint8_t events; +}; + +void pvpanic_setup_io(PVPanicState *s, DeviceState *dev, unsigned size); + static inline uint16_t pvpanic_port(void) { - Object *o = object_resolve_path_type("", TYPE_PVPANIC, NULL); + Object *o = object_resolve_path_type("", TYPE_PVPANIC_ISA_DEVICE, NULL); if (!o) { return 0; } diff --git a/hw/misc/pvpanic-isa.c b/hw/misc/pvpanic-isa.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/pvpanic-isa.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU simulated pvpanic device. + * + * Copyright Fujitsu, Corp. 2013 + * + * Authors: + * Wen Congyang <wency@cn.fujitsu.com> + * Hu Tao <hutao@cn.fujitsu.com> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "sysemu/runstate.h" + +#include "hw/nvram/fw_cfg.h" +#include "hw/qdev-properties.h" +#include "hw/misc/pvpanic.h" +#include "qom/object.h" +#include "hw/isa/isa.h" + +OBJECT_DECLARE_SIMPLE_TYPE(PVPanicISAState, PVPANIC_ISA_DEVICE) + +/* + * PVPanicISAState for ISA device and + * use ioport. + */ +struct PVPanicISAState { + ISADevice parent_obj; + + uint16_t ioport; + PVPanicState pvpanic; +}; + +static void pvpanic_isa_initfn(Object *obj) +{ + PVPanicISAState *s = PVPANIC_ISA_DEVICE(obj); + + pvpanic_setup_io(&s->pvpanic, DEVICE(s), 1); +} + +static void pvpanic_isa_realizefn(DeviceState *dev, Error **errp) +{ + ISADevice *d = ISA_DEVICE(dev); + PVPanicISAState *s = PVPANIC_ISA_DEVICE(dev); + PVPanicState *ps = &s->pvpanic; + FWCfgState *fw_cfg = fw_cfg_find(); + uint16_t *pvpanic_port; + + if (!fw_cfg) { + return; + } + + pvpanic_port = g_malloc(sizeof(*pvpanic_port)); + *pvpanic_port = cpu_to_le16(s->ioport); + fw_cfg_add_file(fw_cfg, "etc/pvpanic-port", pvpanic_port, + sizeof(*pvpanic_port)); + + isa_register_ioport(d, &ps->mr, s->ioport); +} + +static Property pvpanic_isa_properties[] = { + DEFINE_PROP_UINT16(PVPANIC_IOPORT_PROP, PVPanicISAState, ioport, 0x505), + DEFINE_PROP_UINT8("events", PVPanicISAState, pvpanic.events, PVPANIC_PANICKED | PVPANIC_CRASHLOADED), + DEFINE_PROP_END_OF_LIST(), +}; + +static void pvpanic_isa_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + + dc->realize = pvpanic_isa_realizefn; + device_class_set_props(dc, pvpanic_isa_properties); + set_bit(DEVICE_CATEGORY_MISC, dc->categories); +} + +static TypeInfo pvpanic_isa_info = { + .name = TYPE_PVPANIC_ISA_DEVICE, + .parent = TYPE_ISA_DEVICE, + .instance_size = sizeof(PVPanicISAState), + .instance_init = pvpanic_isa_initfn, + .class_init = pvpanic_isa_class_init, +}; + +static void pvpanic_register_types(void) +{ + type_register_static(&pvpanic_isa_info); +} + +type_init(pvpanic_register_types) diff --git a/hw/misc/pvpanic.c b/hw/misc/pvpanic.c index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/pvpanic.c +++ b/hw/misc/pvpanic.c @@ -XXX,XX +XXX,XX @@ #include "hw/misc/pvpanic.h" #include "qom/object.h" -/* The bit of supported pv event, TODO: include uapi header and remove this */ -#define PVPANIC_F_PANICKED 0 -#define PVPANIC_F_CRASHLOADED 1 - -/* The pv event value */ -#define PVPANIC_PANICKED (1 << PVPANIC_F_PANICKED) -#define PVPANIC_CRASHLOADED (1 << PVPANIC_F_CRASHLOADED) - -typedef struct PVPanicState PVPanicState; -DECLARE_INSTANCE_CHECKER(PVPanicState, ISA_PVPANIC_DEVICE, - TYPE_PVPANIC) - static void handle_event(int event) { static bool logged; @@ -XXX,XX +XXX,XX @@ static void handle_event(int event) } } -#include "hw/isa/isa.h" - -struct PVPanicState { - ISADevice parent_obj; - - MemoryRegion io; - uint16_t ioport; - uint8_t events; -}; - /* return supported events on read */ -static uint64_t pvpanic_ioport_read(void *opaque, hwaddr addr, unsigned size) +static uint64_t pvpanic_read(void *opaque, hwaddr addr, unsigned size) { PVPanicState *pvp = opaque; return pvp->events; } -static void pvpanic_ioport_write(void *opaque, hwaddr addr, uint64_t val, +static void pvpanic_write(void *opaque, hwaddr addr, uint64_t val, unsigned size) { handle_event(val); } static const MemoryRegionOps pvpanic_ops = { - .read = pvpanic_ioport_read, - .write = pvpanic_ioport_write, + .read = pvpanic_read, + .write = pvpanic_write, .impl = { .min_access_size = 1, .max_access_size = 1, }, }; -static void pvpanic_isa_initfn(Object *obj) +void pvpanic_setup_io(PVPanicState *s, DeviceState *dev, unsigned size) { - PVPanicState *s = ISA_PVPANIC_DEVICE(obj); - - memory_region_init_io(&s->io, OBJECT(s), &pvpanic_ops, s, "pvpanic", 1); + memory_region_init_io(&s->mr, OBJECT(dev), &pvpanic_ops, s, "pvpanic", size); } - -static void pvpanic_isa_realizefn(DeviceState *dev, Error **errp) -{ - ISADevice *d = ISA_DEVICE(dev); - PVPanicState *s = ISA_PVPANIC_DEVICE(dev); - FWCfgState *fw_cfg = fw_cfg_find(); - uint16_t *pvpanic_port; - - if (!fw_cfg) { - return; - } - - pvpanic_port = g_malloc(sizeof(*pvpanic_port)); - *pvpanic_port = cpu_to_le16(s->ioport); - fw_cfg_add_file(fw_cfg, "etc/pvpanic-port", pvpanic_port, - sizeof(*pvpanic_port)); - - isa_register_ioport(d, &s->io, s->ioport); -} - -static Property pvpanic_isa_properties[] = { - DEFINE_PROP_UINT16(PVPANIC_IOPORT_PROP, PVPanicState, ioport, 0x505), - DEFINE_PROP_UINT8("events", PVPanicState, events, PVPANIC_PANICKED | PVPANIC_CRASHLOADED), - DEFINE_PROP_END_OF_LIST(), -}; - -static void pvpanic_isa_class_init(ObjectClass *klass, void *data) -{ - DeviceClass *dc = DEVICE_CLASS(klass); - - dc->realize = pvpanic_isa_realizefn; - device_class_set_props(dc, pvpanic_isa_properties); - set_bit(DEVICE_CATEGORY_MISC, dc->categories); -} - -static TypeInfo pvpanic_isa_info = { - .name = TYPE_PVPANIC, - .parent = TYPE_ISA_DEVICE, - .instance_size = sizeof(PVPanicState), - .instance_init = pvpanic_isa_initfn, - .class_init = pvpanic_isa_class_init, -}; - -static void pvpanic_register_types(void) -{ - type_register_static(&pvpanic_isa_info); -} - -type_init(pvpanic_register_types) diff --git a/hw/i386/Kconfig b/hw/i386/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/i386/Kconfig +++ b/hw/i386/Kconfig @@ -XXX,XX +XXX,XX @@ config PC imply ISA_DEBUG imply PARALLEL imply PCI_DEVICES - imply PVPANIC + imply PVPANIC_ISA imply QXL imply SEV imply SGA diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config IOTKIT_SYSCTL config IOTKIT_SYSINFO bool -config PVPANIC +config PVPANIC_COMMON + bool + +config PVPANIC_ISA bool depends on ISA_BUS + select PVPANIC_COMMON config AUX bool diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_EMC141X', if_true: files('emc141x.c')) softmmu_ss.add(when: 'CONFIG_UNIMP', if_true: files('unimp.c')) softmmu_ss.add(when: 'CONFIG_EMPTY_SLOT', if_true: files('empty_slot.c')) softmmu_ss.add(when: 'CONFIG_LED', if_true: files('led.c')) +softmmu_ss.add(when: 'CONFIG_PVPANIC_COMMON', if_true: files('pvpanic.c')) # ARM devices softmmu_ss.add(when: 'CONFIG_PL310', if_true: files('arm_l2x0.c')) @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_IOTKIT_SYSINFO', if_true: files('iotkit-sysinfo.c') softmmu_ss.add(when: 'CONFIG_ARMSSE_CPUID', if_true: files('armsse-cpuid.c')) softmmu_ss.add(when: 'CONFIG_ARMSSE_MHU', if_true: files('armsse-mhu.c')) -softmmu_ss.add(when: 'CONFIG_PVPANIC', if_true: files('pvpanic.c')) +softmmu_ss.add(when: 'CONFIG_PVPANIC_ISA', if_true: files('pvpanic-isa.c')) softmmu_ss.add(when: 'CONFIG_AUX', if_true: files('auxbus.c')) softmmu_ss.add(when: 'CONFIG_ASPEED_SOC', if_true: files('aspeed_scu.c', 'aspeed_sdmc.c', 'aspeed_xdma.c')) softmmu_ss.add(when: 'CONFIG_MSF2', if_true: files('msf2-sysreg.c')) diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_i386 = \ (config_host.has_key('CONFIG_LINUX') and \ config_all_devices.has_key('CONFIG_ISA_IPMI_BT') ? ['ipmi-bt-test'] : []) + \ (config_all_devices.has_key('CONFIG_WDT_IB700') ? ['wdt_ib700-test'] : []) + \ - (config_all_devices.has_key('CONFIG_PVPANIC') ? ['pvpanic-test'] : []) + \ + (config_all_devices.has_key('CONFIG_PVPANIC_ISA') ? ['pvpanic-test'] : []) + \ (config_all_devices.has_key('CONFIG_HDA') ? ['intel-hda-test'] : []) + \ (config_all_devices.has_key('CONFIG_I82801B11') ? ['i82801b11-test'] : []) + \ (config_all_devices.has_key('CONFIG_IOH3420') ? ['ioh3420-test'] : []) + \ -- 2.20.1 From: Mihai Carabas <mihai.carabas@oracle.com> Add PCI interface support for PVPANIC device. Create a new file pvpanic-pci.c where the PCI specific routines reside and update the build system with the new files and config structure. Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Reviewed-by: Gerd Hoffmann <kraxel@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/specs/pci-ids.txt | 1 + include/hw/misc/pvpanic.h | 1 + include/hw/pci/pci.h | 1 + hw/misc/pvpanic-pci.c | 94 +++++++++++++++++++++++++++++++++++++++ hw/misc/Kconfig | 6 +++ hw/misc/meson.build | 1 + 6 files changed, 104 insertions(+) create mode 100644 hw/misc/pvpanic-pci.c diff --git a/docs/specs/pci-ids.txt b/docs/specs/pci-ids.txt index XXXXXXX..XXXXXXX 100644 --- a/docs/specs/pci-ids.txt +++ b/docs/specs/pci-ids.txt @@ -XXX,XX +XXX,XX @@ PCI devices (other than virtio): 1b36:000d PCI xhci usb host adapter 1b36:000f mdpy (mdev sample device), linux/samples/vfio-mdev/mdpy.c 1b36:0010 PCIe NVMe device (-device nvme) +1b36:0011 PCI PVPanic device (-device pvpanic-pci) All these devices are documented in docs/specs. diff --git a/include/hw/misc/pvpanic.h b/include/hw/misc/pvpanic.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/misc/pvpanic.h +++ b/include/hw/misc/pvpanic.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" #define TYPE_PVPANIC_ISA_DEVICE "pvpanic" +#define TYPE_PVPANIC_PCI_DEVICE "pvpanic-pci" #define PVPANIC_IOPORT_PROP "ioport" diff --git a/include/hw/pci/pci.h b/include/hw/pci/pci.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/pci/pci.h +++ b/include/hw/pci/pci.h @@ -XXX,XX +XXX,XX @@ extern bool pci_available; #define PCI_DEVICE_ID_REDHAT_PCIE_BRIDGE 0x000e #define PCI_DEVICE_ID_REDHAT_MDPY 0x000f #define PCI_DEVICE_ID_REDHAT_NVME 0x0010 +#define PCI_DEVICE_ID_REDHAT_PVPANIC 0x0011 #define PCI_DEVICE_ID_REDHAT_QXL 0x0100 #define FMT_PCIBUS PRIx64 diff --git a/hw/misc/pvpanic-pci.c b/hw/misc/pvpanic-pci.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/hw/misc/pvpanic-pci.c @@ -XXX,XX +XXX,XX @@ +/* + * QEMU simulated PCI pvpanic device. + * + * Copyright (C) 2020 Oracle + * + * Authors: + * Mihai Carabas <mihai.carabas@oracle.com> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" +#include "qemu/log.h" +#include "qemu/module.h" +#include "sysemu/runstate.h" + +#include "hw/nvram/fw_cfg.h" +#include "hw/qdev-properties.h" +#include "migration/vmstate.h" +#include "hw/misc/pvpanic.h" +#include "qom/object.h" +#include "hw/pci/pci.h" + +OBJECT_DECLARE_SIMPLE_TYPE(PVPanicPCIState, PVPANIC_PCI_DEVICE) + +/* + * PVPanicPCIState for PCI device + */ +typedef struct PVPanicPCIState { + PCIDevice dev; + PVPanicState pvpanic; +} PVPanicPCIState; + +static const VMStateDescription vmstate_pvpanic_pci = { + .name = "pvpanic-pci", + .version_id = 1, + .minimum_version_id = 1, + .fields = (VMStateField[]) { + VMSTATE_PCI_DEVICE(dev, PVPanicPCIState), + VMSTATE_END_OF_LIST() + } +}; + +static void pvpanic_pci_realizefn(PCIDevice *dev, Error **errp) +{ + PVPanicPCIState *s = PVPANIC_PCI_DEVICE(dev); + PVPanicState *ps = &s->pvpanic; + + pvpanic_setup_io(&s->pvpanic, DEVICE(s), 2); + + pci_register_bar(dev, 0, PCI_BASE_ADDRESS_SPACE_MEMORY, &ps->mr); +} + +static Property pvpanic_pci_properties[] = { + DEFINE_PROP_UINT8("events", PVPanicPCIState, pvpanic.events, PVPANIC_PANICKED | PVPANIC_CRASHLOADED), + DEFINE_PROP_END_OF_LIST(), +}; + +static void pvpanic_pci_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + PCIDeviceClass *pc = PCI_DEVICE_CLASS(klass); + + device_class_set_props(dc, pvpanic_pci_properties); + + pc->realize = pvpanic_pci_realizefn; + pc->vendor_id = PCI_VENDOR_ID_REDHAT; + pc->device_id = PCI_DEVICE_ID_REDHAT_PVPANIC; + pc->revision = 1; + pc->class_id = PCI_CLASS_SYSTEM_OTHER; + dc->vmsd = &vmstate_pvpanic_pci; + + set_bit(DEVICE_CATEGORY_MISC, dc->categories); +} + +static TypeInfo pvpanic_pci_info = { + .name = TYPE_PVPANIC_PCI_DEVICE, + .parent = TYPE_PCI_DEVICE, + .instance_size = sizeof(PVPanicPCIState), + .class_init = pvpanic_pci_class_init, + .interfaces = (InterfaceInfo[]) { + { INTERFACE_CONVENTIONAL_PCI_DEVICE }, + { } + } +}; + +static void pvpanic_register_types(void) +{ + type_register_static(&pvpanic_pci_info); +} + +type_init(pvpanic_register_types); diff --git a/hw/misc/Kconfig b/hw/misc/Kconfig index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/Kconfig +++ b/hw/misc/Kconfig @@ -XXX,XX +XXX,XX @@ config IOTKIT_SYSINFO config PVPANIC_COMMON bool +config PVPANIC_PCI + bool + default y if PCI_DEVICES + depends on PCI + select PVPANIC_COMMON + config PVPANIC_ISA bool depends on ISA_BUS diff --git a/hw/misc/meson.build b/hw/misc/meson.build index XXXXXXX..XXXXXXX 100644 --- a/hw/misc/meson.build +++ b/hw/misc/meson.build @@ -XXX,XX +XXX,XX @@ softmmu_ss.add(when: 'CONFIG_ARMSSE_CPUID', if_true: files('armsse-cpuid.c')) softmmu_ss.add(when: 'CONFIG_ARMSSE_MHU', if_true: files('armsse-mhu.c')) softmmu_ss.add(when: 'CONFIG_PVPANIC_ISA', if_true: files('pvpanic-isa.c')) +softmmu_ss.add(when: 'CONFIG_PVPANIC_PCI', if_true: files('pvpanic-pci.c')) softmmu_ss.add(when: 'CONFIG_AUX', if_true: files('auxbus.c')) softmmu_ss.add(when: 'CONFIG_ASPEED_SOC', if_true: files('aspeed_scu.c', 'aspeed_sdmc.c', 'aspeed_xdma.c')) softmmu_ss.add(when: 'CONFIG_MSF2', if_true: files('msf2-sysreg.c')) -- 2.20.1 From: Mihai Carabas <mihai.carabas@oracle.com> Add pvpanic PCI device support details in docs/specs/pvpanic.txt. Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- docs/specs/pvpanic.txt | 13 ++++++++++++- 1 file changed, 12 insertions(+), 1 deletion(-) diff --git a/docs/specs/pvpanic.txt b/docs/specs/pvpanic.txt index XXXXXXX..XXXXXXX 100644 --- a/docs/specs/pvpanic.txt +++ b/docs/specs/pvpanic.txt @@ -XXX,XX +XXX,XX @@ PVPANIC DEVICE ============== -pvpanic device is a simulated ISA device, through which a guest panic +pvpanic device is a simulated device, through which a guest panic event is sent to qemu, and a QMP event is generated. This allows management apps (e.g. libvirt) to be notified and respond to the event. @@ -XXX,XX +XXX,XX @@ The management app has the option of waiting for GUEST_PANICKED events, and/or polling for guest-panicked RunState, to learn when the pvpanic device has fired a panic event. +The pvpanic device can be implemented as an ISA device (using IOPORT) or as a +PCI device. + ISA Interface ------------- @@ -XXX,XX +XXX,XX @@ bit 1: a guest panic has happened and will be handled by the guest; the host should record it or report it, but should not affect the execution of the guest. +PCI Interface +------------- + +The PCI interface is similar to the ISA interface except that it uses an MMIO +address space provided by its BAR0, 1 byte long. Any machine with a PCI bus +can enable a pvpanic device by adding '-device pvpanic-pci' to the command +line. + ACPI Interface -------------- -- 2.20.1 From: Mihai Carabas <mihai.carabas@oracle.com> Add a test case for pvpanic-pci device. The scenario is the same as pvpanic ISA device, but is using the PCI bus. Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Acked-by: Thomas Huth <thuth@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Mihai Carabas <mihai.carabas@oracle.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/qtest/pvpanic-pci-test.c | 94 ++++++++++++++++++++++++++++++++++ tests/qtest/meson.build | 1 + 2 files changed, 95 insertions(+) create mode 100644 tests/qtest/pvpanic-pci-test.c diff --git a/tests/qtest/pvpanic-pci-test.c b/tests/qtest/pvpanic-pci-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/pvpanic-pci-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTest testcase for PV Panic PCI device + * + * Copyright (C) 2020 Oracle + * + * Authors: + * Mihai Carabas <mihai.carabas@oracle.com> + * + * This work is licensed under the terms of the GNU GPL, version 2 or later. + * See the COPYING file in the top-level directory. + * + */ + +#include "qemu/osdep.h" +#include "libqos/libqtest.h" +#include "qapi/qmp/qdict.h" +#include "libqos/pci.h" +#include "libqos/pci-pc.h" +#include "hw/pci/pci_regs.h" + +static void test_panic_nopause(void) +{ + uint8_t val; + QDict *response, *data; + QTestState *qts; + QPCIBus *pcibus; + QPCIDevice *dev; + QPCIBar bar; + + qts = qtest_init("-device pvpanic-pci,addr=04.0 -action panic=none"); + pcibus = qpci_new_pc(qts, NULL); + dev = qpci_device_find(pcibus, QPCI_DEVFN(0x4, 0x0)); + qpci_device_enable(dev); + bar = qpci_iomap(dev, 0, NULL); + + qpci_memread(dev, bar, 0, &val, sizeof(val)); + g_assert_cmpuint(val, ==, 3); + + val = 1; + qpci_memwrite(dev, bar, 0, &val, sizeof(val)); + + response = qtest_qmp_eventwait_ref(qts, "GUEST_PANICKED"); + g_assert(qdict_haskey(response, "data")); + data = qdict_get_qdict(response, "data"); + g_assert(qdict_haskey(data, "action")); + g_assert_cmpstr(qdict_get_str(data, "action"), ==, "run"); + qobject_unref(response); + + qtest_quit(qts); +} + +static void test_panic(void) +{ + uint8_t val; + QDict *response, *data; + QTestState *qts; + QPCIBus *pcibus; + QPCIDevice *dev; + QPCIBar bar; + + qts = qtest_init("-device pvpanic-pci,addr=04.0 -action panic=pause"); + pcibus = qpci_new_pc(qts, NULL); + dev = qpci_device_find(pcibus, QPCI_DEVFN(0x4, 0x0)); + qpci_device_enable(dev); + bar = qpci_iomap(dev, 0, NULL); + + qpci_memread(dev, bar, 0, &val, sizeof(val)); + g_assert_cmpuint(val, ==, 3); + + val = 1; + qpci_memwrite(dev, bar, 0, &val, sizeof(val)); + + response = qtest_qmp_eventwait_ref(qts, "GUEST_PANICKED"); + g_assert(qdict_haskey(response, "data")); + data = qdict_get_qdict(response, "data"); + g_assert(qdict_haskey(data, "action")); + g_assert_cmpstr(qdict_get_str(data, "action"), ==, "pause"); + qobject_unref(response); + + qtest_quit(qts); +} + +int main(int argc, char **argv) +{ + int ret; + + g_test_init(&argc, &argv, NULL); + qtest_add_func("/pvpanic-pci/panic", test_panic); + qtest_add_func("/pvpanic-pci/panic-nopause", test_panic_nopause); + + ret = g_test_run(); + + return ret; +} diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_i386 = \ config_all_devices.has_key('CONFIG_ISA_IPMI_BT') ? ['ipmi-bt-test'] : []) + \ (config_all_devices.has_key('CONFIG_WDT_IB700') ? ['wdt_ib700-test'] : []) + \ (config_all_devices.has_key('CONFIG_PVPANIC_ISA') ? ['pvpanic-test'] : []) + \ + (config_all_devices.has_key('CONFIG_PVPANIC_PCI') ? ['pvpanic-pci-test'] : []) + \ (config_all_devices.has_key('CONFIG_HDA') ? ['intel-hda-test'] : []) + \ (config_all_devices.has_key('CONFIG_I82801B11') ? ['i82801b11-test'] : []) + \ (config_all_devices.has_key('CONFIG_IOH3420') ? ['ioh3420-test'] : []) + \ -- 2.20.1 The ptimer API currently provides two methods for setting the period: ptimer_set_period(), which takes a period in nanoseconds, and ptimer_set_freq(), which takes a frequency in Hz. Neither of these lines up nicely with the Clock API, because although both the Clock and the ptimer track the frequency using a representation of whole and fractional nanoseconds, conversion via either period-in-ns or frequency-in-Hz will introduce a rounding error. Add a new function ptimer_set_period_from_clock() which takes the Clock object directly to avoid the rounding issues. This includes a facility for the user to specify that there is a frequency divider between the Clock proper and the timer, as some timer devices like the CMSDK APB dualtimer need this. To avoid having to drag in clock.h from ptimer.h we add the Clock type to typedefs.h. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-2-peter.maydell@linaro.org Message-id: 20210121190622.22000-2-peter.maydell@linaro.org --- include/hw/ptimer.h | 22 ++++++++++++++++++++++ include/qemu/typedefs.h | 1 + hw/core/ptimer.c | 34 ++++++++++++++++++++++++++++++++++ 3 files changed, 57 insertions(+) diff --git a/include/hw/ptimer.h b/include/hw/ptimer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/ptimer.h +++ b/include/hw/ptimer.h @@ -XXX,XX +XXX,XX @@ void ptimer_transaction_commit(ptimer_state *s); */ void ptimer_set_period(ptimer_state *s, int64_t period); +/** + * ptimer_set_period_from_clock - Set counter increment from a Clock + * @s: ptimer to configure + * @clk: pointer to Clock object to take period from + * @divisor: value to scale the clock frequency down by + * + * If the ptimer is being driven from a Clock, this is the preferred + * way to tell the ptimer about the period, because it avoids any + * possible rounding errors that might happen if the internal + * representation of the Clock period was converted to either a period + * in ns or a frequency in Hz. + * + * If the ptimer should run at the same frequency as the clock, + * pass 1 as the @divisor; if the ptimer should run at half the + * frequency, pass 2, and so on. + * + * This function will assert if it is called outside a + * ptimer_transaction_begin/commit block. + */ +void ptimer_set_period_from_clock(ptimer_state *s, const Clock *clock, + unsigned int divisor); + /** * ptimer_set_freq - Set counter frequency in Hz * @s: ptimer to configure diff --git a/include/qemu/typedefs.h b/include/qemu/typedefs.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/typedefs.h +++ b/include/qemu/typedefs.h @@ -XXX,XX +XXX,XX @@ typedef struct BlockDriverState BlockDriverState; typedef struct BusClass BusClass; typedef struct BusState BusState; typedef struct Chardev Chardev; +typedef struct Clock Clock; typedef struct CompatProperty CompatProperty; typedef struct CoMutex CoMutex; typedef struct CPUAddressSpace CPUAddressSpace; diff --git a/hw/core/ptimer.c b/hw/core/ptimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/ptimer.c +++ b/hw/core/ptimer.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/qtest.h" #include "block/aio.h" #include "sysemu/cpus.h" +#include "hw/clock.h" #define DELTA_ADJUST 1 #define DELTA_NO_ADJUST -1 @@ -XXX,XX +XXX,XX @@ void ptimer_set_period(ptimer_state *s, int64_t period) } } +/* Set counter increment interval from a Clock */ +void ptimer_set_period_from_clock(ptimer_state *s, const Clock *clk, + unsigned int divisor) +{ + /* + * The raw clock period is a 64-bit value in units of 2^-32 ns; + * put another way it's a 32.32 fixed-point ns value. Our internal + * representation of the period is 64.32 fixed point ns, so + * the conversion is simple. + */ + uint64_t raw_period = clock_get(clk); + uint64_t period_frac; + + assert(s->in_transaction); + s->delta = ptimer_get_count(s); + s->period = extract64(raw_period, 32, 32); + period_frac = extract64(raw_period, 0, 32); + /* + * divisor specifies a possible frequency divisor between the + * clock and the timer, so it is a multiplier on the period. + * We do the multiply after splitting the raw period out into + * period and frac to avoid having to do a 32*64->96 multiply. + */ + s->period *= divisor; + period_frac *= divisor; + s->period += extract64(period_frac, 32, 32); + s->period_frac = (uint32_t)period_frac; + + if (s->enabled) { + s->need_reload = true; + } +} + /* Set counter frequency in Hz. */ void ptimer_set_freq(ptimer_state *s, uint32_t freq) { -- 2.20.1 Add a function for checking whether a clock has a source. This is useful for devices which have input clocks that must be wired up by the board as it allows them to fail in realize rather than ploughing on with a zero-period clock. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-3-peter.maydell@linaro.org Message-id: 20210121190622.22000-3-peter.maydell@linaro.org --- docs/devel/clocks.rst | 16 ++++++++++++++++ include/hw/clock.h | 15 +++++++++++++++ 2 files changed, 31 insertions(+) diff --git a/docs/devel/clocks.rst b/docs/devel/clocks.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/devel/clocks.rst +++ b/docs/devel/clocks.rst @@ -XXX,XX +XXX,XX @@ object during device instance init. For example: /* set initial value to 10ns / 100MHz */ clock_set_ns(clk, 10); +To enforce that the clock is wired up by the board code, you can +call ``clock_has_source()`` in your device's realize method: + +.. code-block:: c + + if (!clock_has_source(s->clk)) { + error_setg(errp, "MyDevice: clk input must be connected"); + return; + } + +Note that this only checks that the clock has been wired up; it is +still possible that the output clock connected to it is disabled +or has not yet been configured, in which case the period will be +zero. You should use the clock callback to find out when the clock +period changes. + Fetching clock frequency/period ------------------------------- diff --git a/include/hw/clock.h b/include/hw/clock.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/clock.h +++ b/include/hw/clock.h @@ -XXX,XX +XXX,XX @@ void clock_clear_callback(Clock *clk); */ void clock_set_source(Clock *clk, Clock *src); +/** + * clock_has_source: + * @clk: the clock + * + * Returns true if the clock has a source clock connected to it. + * This is useful for devices which have input clocks which must + * be connected by the board/SoC code which creates them. The + * device code can use this to check in its realize method that + * the clock has been connected. + */ +static inline bool clock_has_source(const Clock *clk) +{ + return clk->source != NULL; +} + /** * clock_set: * @clk: the clock to initialize. -- 2.20.1 Add a simple test of the CMSDK APB timer, since we're about to do some refactoring of how it is clocked. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-4-peter.maydell@linaro.org Message-id: 20210121190622.22000-4-peter.maydell@linaro.org --- tests/qtest/cmsdk-apb-timer-test.c | 75 ++++++++++++++++++++++++++++++ MAINTAINERS | 1 + tests/qtest/meson.build | 1 + 3 files changed, 77 insertions(+) create mode 100644 tests/qtest/cmsdk-apb-timer-test.c diff --git a/tests/qtest/cmsdk-apb-timer-test.c b/tests/qtest/cmsdk-apb-timer-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/cmsdk-apb-timer-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTest testcase for the CMSDK APB timer device + * + * Copyright (c) 2021 Linaro Limited + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program 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 General Public License + * for more details. + */ + +#include "qemu/osdep.h" +#include "libqtest-single.h" + +/* IoTKit/ARMSSE-200 timer0; driven at 25MHz in mps2-an385, so 40ns per tick */ +#define TIMER_BASE 0x40000000 + +#define CTRL 0 +#define VALUE 4 +#define RELOAD 8 +#define INTSTATUS 0xc + +static void test_timer(void) +{ + g_assert_true(readl(TIMER_BASE + INTSTATUS) == 0); + + /* Start timer: will fire after 40 * 1000 == 40000 ns */ + writel(TIMER_BASE + RELOAD, 1000); + writel(TIMER_BASE + CTRL, 9); + + /* Step to just past the 500th tick and check VALUE */ + clock_step(40 * 500 + 1); + g_assert_cmpuint(readl(TIMER_BASE + INTSTATUS), ==, 0); + g_assert_cmpuint(readl(TIMER_BASE + VALUE), ==, 500); + + /* Just past the 1000th tick: timer should have fired */ + clock_step(40 * 500); + g_assert_cmpuint(readl(TIMER_BASE + INTSTATUS), ==, 1); + g_assert_cmpuint(readl(TIMER_BASE + VALUE), ==, 0); + + /* VALUE reloads at the following tick */ + clock_step(40); + g_assert_cmpuint(readl(TIMER_BASE + VALUE), ==, 1000); + + /* Check write-1-to-clear behaviour of INTSTATUS */ + writel(TIMER_BASE + INTSTATUS, 0); + g_assert_cmpuint(readl(TIMER_BASE + INTSTATUS), ==, 1); + writel(TIMER_BASE + INTSTATUS, 1); + g_assert_cmpuint(readl(TIMER_BASE + INTSTATUS), ==, 0); + + /* Turn off the timer */ + writel(TIMER_BASE + CTRL, 0); +} + +int main(int argc, char **argv) +{ + int r; + + g_test_init(&argc, &argv, NULL); + + qtest_start("-machine mps2-an385"); + + qtest_add_func("/cmsdk-apb-timer/timer", test_timer); + + r = g_test_run(); + + qtest_end(); + + return r; +} diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/hw/rtc/pl031.h F: include/hw/arm/primecell.h F: hw/timer/cmsdk-apb-timer.c F: include/hw/timer/cmsdk-apb-timer.h +F: tests/qtest/cmsdk-apb-timer-test.c F: hw/timer/cmsdk-apb-dualtimer.c F: include/hw/timer/cmsdk-apb-dualtimer.h F: hw/char/cmsdk-apb-uart.c diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_npcm7xx = \ 'npcm7xx_timer-test', 'npcm7xx_watchdog_timer-test'] qtests_arm = \ + (config_all_devices.has_key('CONFIG_CMSDK_APB_TIMER') ? ['cmsdk-apb-timer-test'] : []) + \ (config_all_devices.has_key('CONFIG_PFLASH_CFI02') ? ['pflash-cfi02-test'] : []) + \ (config_all_devices.has_key('CONFIG_NPCM7XX') ? qtests_npcm7xx : []) + \ ['arm-cpu-features', -- 2.20.1 Add a simple test of the CMSDK watchdog, since we're about to do some refactoring of how it is clocked. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-5-peter.maydell@linaro.org Message-id: 20210121190622.22000-5-peter.maydell@linaro.org Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> --- tests/qtest/cmsdk-apb-watchdog-test.c | 79 +++++++++++++++++++++++++++ MAINTAINERS | 1 + tests/qtest/meson.build | 1 + 3 files changed, 81 insertions(+) create mode 100644 tests/qtest/cmsdk-apb-watchdog-test.c diff --git a/tests/qtest/cmsdk-apb-watchdog-test.c b/tests/qtest/cmsdk-apb-watchdog-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/cmsdk-apb-watchdog-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTest testcase for the CMSDK APB watchdog device + * + * Copyright (c) 2021 Linaro Limited + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program 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 General Public License + * for more details. + */ + +#include "qemu/osdep.h" +#include "libqtest-single.h" + +/* + * lm3s811evb watchdog; at board startup this runs at 200MHz / 16 == 12.5MHz, + * which is 80ns per tick. + */ +#define WDOG_BASE 0x40000000 + +#define WDOGLOAD 0 +#define WDOGVALUE 4 +#define WDOGCONTROL 8 +#define WDOGINTCLR 0xc +#define WDOGRIS 0x10 +#define WDOGMIS 0x14 +#define WDOGLOCK 0xc00 + +static void test_watchdog(void) +{ + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); + + writel(WDOG_BASE + WDOGCONTROL, 1); + writel(WDOG_BASE + WDOGLOAD, 1000); + + /* Step to just past the 500th tick */ + clock_step(500 * 80 + 1); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 500); + + /* Just past the 1000th tick: timer should have fired */ + clock_step(500 * 80); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 1); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 0); + + /* VALUE reloads at following tick */ + clock_step(80); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 1000); + + /* Writing any value to WDOGINTCLR clears the interrupt and reloads */ + clock_step(500 * 80); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 500); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 1); + writel(WDOG_BASE + WDOGINTCLR, 0); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 1000); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); +} + +int main(int argc, char **argv) +{ + int r; + + g_test_init(&argc, &argv, NULL); + + qtest_start("-machine lm3s811evb"); + + qtest_add_func("/cmsdk-apb-watchdog/watchdog", test_watchdog); + + r = g_test_run(); + + qtest_end(); + + return r; +} diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: hw/char/cmsdk-apb-uart.c F: include/hw/char/cmsdk-apb-uart.h F: hw/watchdog/cmsdk-apb-watchdog.c F: include/hw/watchdog/cmsdk-apb-watchdog.h +F: tests/qtest/cmsdk-apb-watchdog-test.c F: hw/misc/tz-ppc.c F: include/hw/misc/tz-ppc.h F: hw/misc/tz-mpc.c diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_npcm7xx = \ 'npcm7xx_watchdog_timer-test'] qtests_arm = \ (config_all_devices.has_key('CONFIG_CMSDK_APB_TIMER') ? ['cmsdk-apb-timer-test'] : []) + \ + (config_all_devices.has_key('CONFIG_CMSDK_APB_WATCHDOG') ? ['cmsdk-apb-watchdog-test'] : []) + \ (config_all_devices.has_key('CONFIG_PFLASH_CFI02') ? ['pflash-cfi02-test'] : []) + \ (config_all_devices.has_key('CONFIG_NPCM7XX') ? qtests_npcm7xx : []) + \ ['arm-cpu-features', -- 2.20.1 Add a simple test of the CMSDK dual timer, since we're about to do some refactoring of how it is clocked. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Message-id: 20210128114145.20536-6-peter.maydell@linaro.org Message-id: 20210121190622.22000-6-peter.maydell@linaro.org --- tests/qtest/cmsdk-apb-dualtimer-test.c | 130 +++++++++++++++++++++++++ MAINTAINERS | 1 + tests/qtest/meson.build | 1 + 3 files changed, 132 insertions(+) create mode 100644 tests/qtest/cmsdk-apb-dualtimer-test.c diff --git a/tests/qtest/cmsdk-apb-dualtimer-test.c b/tests/qtest/cmsdk-apb-dualtimer-test.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/qtest/cmsdk-apb-dualtimer-test.c @@ -XXX,XX +XXX,XX @@ +/* + * QTest testcase for the CMSDK APB dualtimer device + * + * Copyright (c) 2021 Linaro Limited + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License as published by the + * Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program 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 General Public License + * for more details. + */ + +#include "qemu/osdep.h" +#include "libqtest-single.h" + +/* IoTKit/ARMSSE dualtimer; driven at 25MHz in mps2-an385, so 40ns per tick */ +#define TIMER_BASE 0x40002000 + +#define TIMER1LOAD 0 +#define TIMER1VALUE 4 +#define TIMER1CONTROL 8 +#define TIMER1INTCLR 0xc +#define TIMER1RIS 0x10 +#define TIMER1MIS 0x14 +#define TIMER1BGLOAD 0x18 + +#define TIMER2LOAD 0x20 +#define TIMER2VALUE 0x24 +#define TIMER2CONTROL 0x28 +#define TIMER2INTCLR 0x2c +#define TIMER2RIS 0x30 +#define TIMER2MIS 0x34 +#define TIMER2BGLOAD 0x38 + +#define CTRL_ENABLE (1 << 7) +#define CTRL_PERIODIC (1 << 6) +#define CTRL_INTEN (1 << 5) +#define CTRL_PRESCALE_1 (0 << 2) +#define CTRL_PRESCALE_16 (1 << 2) +#define CTRL_PRESCALE_256 (2 << 2) +#define CTRL_32BIT (1 << 1) +#define CTRL_ONESHOT (1 << 0) + +static void test_dualtimer(void) +{ + g_assert_true(readl(TIMER_BASE + TIMER1RIS) == 0); + + /* Start timer: will fire after 40000 ns */ + writel(TIMER_BASE + TIMER1LOAD, 1000); + /* enable in free-running, wrapping, interrupt mode */ + writel(TIMER_BASE + TIMER1CONTROL, CTRL_ENABLE | CTRL_INTEN); + + /* Step to just past the 500th tick and check VALUE */ + clock_step(500 * 40 + 1); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1RIS), ==, 0); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1VALUE), ==, 500); + + /* Just past the 1000th tick: timer should have fired */ + clock_step(500 * 40); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1RIS), ==, 1); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1VALUE), ==, 0); + + /* + * We are in free-running wrapping 16-bit mode, so on the following + * tick VALUE should have wrapped round to 0xffff. + */ + clock_step(40); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1VALUE), ==, 0xffff); + + /* Check that any write to INTCLR clears interrupt */ + writel(TIMER_BASE + TIMER1INTCLR, 1); + g_assert_cmpuint(readl(TIMER_BASE + TIMER1RIS), ==, 0); + + /* Turn off the timer */ + writel(TIMER_BASE + TIMER1CONTROL, 0); +} + +static void test_prescale(void) +{ + g_assert_true(readl(TIMER_BASE + TIMER2RIS) == 0); + + /* Start timer: will fire after 40 * 256 * 1000 == 1024000 ns */ + writel(TIMER_BASE + TIMER2LOAD, 1000); + /* enable in periodic, wrapping, interrupt mode, prescale 256 */ + writel(TIMER_BASE + TIMER2CONTROL, + CTRL_ENABLE | CTRL_INTEN | CTRL_PERIODIC | CTRL_PRESCALE_256); + + /* Step to just past the 500th tick and check VALUE */ + clock_step(40 * 256 * 501); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2RIS), ==, 0); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2VALUE), ==, 500); + + /* Just past the 1000th tick: timer should have fired */ + clock_step(40 * 256 * 500); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2RIS), ==, 1); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2VALUE), ==, 0); + + /* In periodic mode the tick VALUE now reloads */ + clock_step(40 * 256); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2VALUE), ==, 1000); + + /* Check that any write to INTCLR clears interrupt */ + writel(TIMER_BASE + TIMER2INTCLR, 1); + g_assert_cmpuint(readl(TIMER_BASE + TIMER2RIS), ==, 0); + + /* Turn off the timer */ + writel(TIMER_BASE + TIMER2CONTROL, 0); +} + +int main(int argc, char **argv) +{ + int r; + + g_test_init(&argc, &argv, NULL); + + qtest_start("-machine mps2-an385"); + + qtest_add_func("/cmsdk-apb-dualtimer/dualtimer", test_dualtimer); + qtest_add_func("/cmsdk-apb-dualtimer/prescale", test_prescale); + + r = g_test_run(); + + qtest_end(); + + return r; +} diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ F: include/hw/timer/cmsdk-apb-timer.h F: tests/qtest/cmsdk-apb-timer-test.c F: hw/timer/cmsdk-apb-dualtimer.c F: include/hw/timer/cmsdk-apb-dualtimer.h +F: tests/qtest/cmsdk-apb-dualtimer-test.c F: hw/char/cmsdk-apb-uart.c F: include/hw/char/cmsdk-apb-uart.h F: hw/watchdog/cmsdk-apb-watchdog.c diff --git a/tests/qtest/meson.build b/tests/qtest/meson.build index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/meson.build +++ b/tests/qtest/meson.build @@ -XXX,XX +XXX,XX @@ qtests_npcm7xx = \ 'npcm7xx_timer-test', 'npcm7xx_watchdog_timer-test'] qtests_arm = \ + (config_all_devices.has_key('CONFIG_CMSDK_APB_DUALTIMER') ? ['cmsdk-apb-dualtimer-test'] : []) + \ (config_all_devices.has_key('CONFIG_CMSDK_APB_TIMER') ? ['cmsdk-apb-timer-test'] : []) + \ (config_all_devices.has_key('CONFIG_CMSDK_APB_WATCHDOG') ? ['cmsdk-apb-watchdog-test'] : []) + \ (config_all_devices.has_key('CONFIG_PFLASH_CFI02') ? ['pflash-cfi02-test'] : []) + \ -- 2.20.1 The state struct for the CMSDK APB timer device doesn't follow our usual naming convention of camelcase -- "CMSDK" and "APB" are both acronyms, but "TIMER" is not so should not be all-uppercase. Globally rename the struct to "CMSDKAPBTimer" (bringing it into line with CMSDKAPBWatchdog and CMSDKAPBDualTimer; CMSDKAPBUART remains as-is because "UART" is an acronym). Commit created with: perl -p -i -e 's/CMSDKAPBTIMER/CMSDKAPBTimer/g' hw/timer/cmsdk-apb-timer.c include/hw/arm/armsse.h include/hw/timer/cmsdk-apb-timer.h Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-7-peter.maydell@linaro.org Message-id: 20210121190622.22000-7-peter.maydell@linaro.org --- include/hw/arm/armsse.h | 6 +++--- include/hw/timer/cmsdk-apb-timer.h | 4 ++-- hw/timer/cmsdk-apb-timer.c | 28 ++++++++++++++-------------- 3 files changed, 19 insertions(+), 19 deletions(-) diff --git a/include/hw/arm/armsse.h b/include/hw/arm/armsse.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/armsse.h +++ b/include/hw/arm/armsse.h @@ -XXX,XX +XXX,XX @@ struct ARMSSE { TZPPC apb_ppc0; TZPPC apb_ppc1; TZMPC mpc[IOTS_NUM_MPC]; - CMSDKAPBTIMER timer0; - CMSDKAPBTIMER timer1; - CMSDKAPBTIMER s32ktimer; + CMSDKAPBTimer timer0; + CMSDKAPBTimer timer1; + CMSDKAPBTimer s32ktimer; qemu_or_irq ppc_irq_orgate; SplitIRQ sec_resp_splitter; SplitIRQ ppc_irq_splitter[NUM_PPCS]; diff --git a/include/hw/timer/cmsdk-apb-timer.h b/include/hw/timer/cmsdk-apb-timer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-timer.h +++ b/include/hw/timer/cmsdk-apb-timer.h @@ -XXX,XX +XXX,XX @@ #include "qom/object.h" #define TYPE_CMSDK_APB_TIMER "cmsdk-apb-timer" -OBJECT_DECLARE_SIMPLE_TYPE(CMSDKAPBTIMER, CMSDK_APB_TIMER) +OBJECT_DECLARE_SIMPLE_TYPE(CMSDKAPBTimer, CMSDK_APB_TIMER) -struct CMSDKAPBTIMER { +struct CMSDKAPBTimer { /*< private >*/ SysBusDevice parent_obj; diff --git a/hw/timer/cmsdk-apb-timer.c b/hw/timer/cmsdk-apb-timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-timer.c +++ b/hw/timer/cmsdk-apb-timer.c @@ -XXX,XX +XXX,XX @@ static const int timer_id[] = { 0x0d, 0xf0, 0x05, 0xb1, /* CID0..CID3 */ }; -static void cmsdk_apb_timer_update(CMSDKAPBTIMER *s) +static void cmsdk_apb_timer_update(CMSDKAPBTimer *s) { qemu_set_irq(s->timerint, !!(s->intstatus & R_INTSTATUS_IRQ_MASK)); } static uint64_t cmsdk_apb_timer_read(void *opaque, hwaddr offset, unsigned size) { - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(opaque); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(opaque); uint64_t r; switch (offset) { @@ -XXX,XX +XXX,XX @@ static uint64_t cmsdk_apb_timer_read(void *opaque, hwaddr offset, unsigned size) static void cmsdk_apb_timer_write(void *opaque, hwaddr offset, uint64_t value, unsigned size) { - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(opaque); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(opaque); trace_cmsdk_apb_timer_write(offset, value, size); @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps cmsdk_apb_timer_ops = { static void cmsdk_apb_timer_tick(void *opaque) { - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(opaque); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(opaque); if (s->ctrl & R_CTRL_IRQEN_MASK) { s->intstatus |= R_INTSTATUS_IRQ_MASK; @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_tick(void *opaque) static void cmsdk_apb_timer_reset(DeviceState *dev) { - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(dev); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(dev); trace_cmsdk_apb_timer_reset(); s->ctrl = 0; @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_reset(DeviceState *dev) static void cmsdk_apb_timer_init(Object *obj) { SysBusDevice *sbd = SYS_BUS_DEVICE(obj); - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(obj); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(obj); memory_region_init_io(&s->iomem, obj, &cmsdk_apb_timer_ops, s, "cmsdk-apb-timer", 0x1000); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_init(Object *obj) static void cmsdk_apb_timer_realize(DeviceState *dev, Error **errp) { - CMSDKAPBTIMER *s = CMSDK_APB_TIMER(dev); + CMSDKAPBTimer *s = CMSDK_APB_TIMER(dev); if (s->pclk_frq == 0) { error_setg(errp, "CMSDK APB timer: pclk-frq property must be set"); @@ -XXX,XX +XXX,XX @@ static const VMStateDescription cmsdk_apb_timer_vmstate = { .version_id = 1, .minimum_version_id = 1, .fields = (VMStateField[]) { - VMSTATE_PTIMER(timer, CMSDKAPBTIMER), - VMSTATE_UINT32(ctrl, CMSDKAPBTIMER), - VMSTATE_UINT32(value, CMSDKAPBTIMER), - VMSTATE_UINT32(reload, CMSDKAPBTIMER), - VMSTATE_UINT32(intstatus, CMSDKAPBTIMER), + VMSTATE_PTIMER(timer, CMSDKAPBTimer), + VMSTATE_UINT32(ctrl, CMSDKAPBTimer), + VMSTATE_UINT32(value, CMSDKAPBTimer), + VMSTATE_UINT32(reload, CMSDKAPBTimer), + VMSTATE_UINT32(intstatus, CMSDKAPBTimer), VMSTATE_END_OF_LIST() } }; static Property cmsdk_apb_timer_properties[] = { - DEFINE_PROP_UINT32("pclk-frq", CMSDKAPBTIMER, pclk_frq, 0), + DEFINE_PROP_UINT32("pclk-frq", CMSDKAPBTimer, pclk_frq, 0), DEFINE_PROP_END_OF_LIST(), }; @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_class_init(ObjectClass *klass, void *data) static const TypeInfo cmsdk_apb_timer_info = { .name = TYPE_CMSDK_APB_TIMER, .parent = TYPE_SYS_BUS_DEVICE, - .instance_size = sizeof(CMSDKAPBTIMER), + .instance_size = sizeof(CMSDKAPBTimer), .instance_init = cmsdk_apb_timer_init, .class_init = cmsdk_apb_timer_class_init, }; -- 2.20.1 As the first step in converting the CMSDK_APB_TIMER device to the Clock framework, add a Clock input. For the moment we do nothing with this clock; we will change the behaviour from using the pclk-frq property to using the Clock once all the users of this device have been converted to wire up the Clock. Since the device doesn't already have a doc comment for its "QEMU interface", we add one including the new Clock. This is a migration compatibility break for machines mps2-an505, mps2-an521, musca-a, musca-b1. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-8-peter.maydell@linaro.org Message-id: 20210121190622.22000-8-peter.maydell@linaro.org --- include/hw/timer/cmsdk-apb-timer.h | 9 +++++++++ hw/timer/cmsdk-apb-timer.c | 7 +++++-- 2 files changed, 14 insertions(+), 2 deletions(-) diff --git a/include/hw/timer/cmsdk-apb-timer.h b/include/hw/timer/cmsdk-apb-timer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-timer.h +++ b/include/hw/timer/cmsdk-apb-timer.h @@ -XXX,XX +XXX,XX @@ #include "hw/qdev-properties.h" #include "hw/sysbus.h" #include "hw/ptimer.h" +#include "hw/clock.h" #include "qom/object.h" #define TYPE_CMSDK_APB_TIMER "cmsdk-apb-timer" OBJECT_DECLARE_SIMPLE_TYPE(CMSDKAPBTimer, CMSDK_APB_TIMER) +/* + * QEMU interface: + * + QOM property "pclk-frq": frequency at which the timer is clocked + * + Clock input "pclk": clock for the timer + * + sysbus MMIO region 0: the register bank + * + sysbus IRQ 0: timer interrupt TIMERINT + */ struct CMSDKAPBTimer { /*< private >*/ SysBusDevice parent_obj; @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBTimer { qemu_irq timerint; uint32_t pclk_frq; struct ptimer_state *timer; + Clock *pclk; uint32_t ctrl; uint32_t value; diff --git a/hw/timer/cmsdk-apb-timer.c b/hw/timer/cmsdk-apb-timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-timer.c +++ b/hw/timer/cmsdk-apb-timer.c @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/irq.h" #include "hw/registerfields.h" +#include "hw/qdev-clock.h" #include "hw/timer/cmsdk-apb-timer.h" #include "migration/vmstate.h" @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_init(Object *obj) s, "cmsdk-apb-timer", 0x1000); sysbus_init_mmio(sbd, &s->iomem); sysbus_init_irq(sbd, &s->timerint); + s->pclk = qdev_init_clock_in(DEVICE(s), "pclk", NULL, NULL); } static void cmsdk_apb_timer_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_realize(DeviceState *dev, Error **errp) static const VMStateDescription cmsdk_apb_timer_vmstate = { .name = "cmsdk-apb-timer", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .fields = (VMStateField[]) { VMSTATE_PTIMER(timer, CMSDKAPBTimer), + VMSTATE_CLOCK(pclk, CMSDKAPBTimer), VMSTATE_UINT32(ctrl, CMSDKAPBTimer), VMSTATE_UINT32(value, CMSDKAPBTimer), VMSTATE_UINT32(reload, CMSDKAPBTimer), -- 2.20.1 As the first step in converting the CMSDK_APB_DUALTIMER device to the Clock framework, add a Clock input. For the moment we do nothing with this clock; we will change the behaviour from using the pclk-frq property to using the Clock once all the users of this device have been converted to wire up the Clock. We take the opportunity to correct the name of the clock input to match the hardware -- the dual timer names the clock which drives the timers TIMCLK. (It does also have a 'pclk' input, which is used only for the register and APB bus logic; on the SSE-200 these clocks are both connected together.) This is a migration compatibility break for machines mps2-an385, mps2-an386, mps2-an500, mps2-an511, mps2-an505, mps2-an521, musca-a, musca-b1. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-9-peter.maydell@linaro.org Message-id: 20210121190622.22000-9-peter.maydell@linaro.org --- include/hw/timer/cmsdk-apb-dualtimer.h | 3 +++ hw/timer/cmsdk-apb-dualtimer.c | 7 +++++-- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/include/hw/timer/cmsdk-apb-dualtimer.h b/include/hw/timer/cmsdk-apb-dualtimer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-dualtimer.h +++ b/include/hw/timer/cmsdk-apb-dualtimer.h @@ -XXX,XX +XXX,XX @@ * * QEMU interface: * + QOM property "pclk-frq": frequency at which the timer is clocked + * + Clock input "TIMCLK": clock (for both timers) * + sysbus MMIO region 0: the register bank * + sysbus IRQ 0: combined timer interrupt TIMINTC * + sysbus IRO 1: timer block 1 interrupt TIMINT1 @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/ptimer.h" +#include "hw/clock.h" #include "qom/object.h" #define TYPE_CMSDK_APB_DUALTIMER "cmsdk-apb-dualtimer" @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBDualTimer { MemoryRegion iomem; qemu_irq timerintc; uint32_t pclk_frq; + Clock *timclk; CMSDKAPBDualTimerModule timermod[CMSDK_APB_DUALTIMER_NUM_MODULES]; uint32_t timeritcr; diff --git a/hw/timer/cmsdk-apb-dualtimer.c b/hw/timer/cmsdk-apb-dualtimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-dualtimer.c +++ b/hw/timer/cmsdk-apb-dualtimer.c @@ -XXX,XX +XXX,XX @@ #include "hw/irq.h" #include "hw/qdev-properties.h" #include "hw/registerfields.h" +#include "hw/qdev-clock.h" #include "hw/timer/cmsdk-apb-dualtimer.h" #include "migration/vmstate.h" @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_init(Object *obj) for (i = 0; i < ARRAY_SIZE(s->timermod); i++) { sysbus_init_irq(sbd, &s->timermod[i].timerint); } + s->timclk = qdev_init_clock_in(DEVICE(s), "TIMCLK", NULL, NULL); } static void cmsdk_apb_dualtimer_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static const VMStateDescription cmsdk_dualtimermod_vmstate = { static const VMStateDescription cmsdk_apb_dualtimer_vmstate = { .name = "cmsdk-apb-dualtimer", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .fields = (VMStateField[]) { + VMSTATE_CLOCK(timclk, CMSDKAPBDualTimer), VMSTATE_STRUCT_ARRAY(timermod, CMSDKAPBDualTimer, CMSDK_APB_DUALTIMER_NUM_MODULES, 1, cmsdk_dualtimermod_vmstate, -- 2.20.1 As the first step in converting the CMSDK_APB_TIMER device to the Clock framework, add a Clock input. For the moment we do nothing with this clock; we will change the behaviour from using the wdogclk-frq property to using the Clock once all the users of this device have been converted to wire up the Clock. This is a migration compatibility break for machines mps2-an385, mps2-an386, mps2-an500, mps2-an511, mps2-an505, mps2-an521, musca-a, musca-b1, lm3s811evb, lm3s6965evb. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-10-peter.maydell@linaro.org Message-id: 20210121190622.22000-10-peter.maydell@linaro.org --- include/hw/watchdog/cmsdk-apb-watchdog.h | 3 +++ hw/watchdog/cmsdk-apb-watchdog.c | 7 +++++-- 2 files changed, 8 insertions(+), 2 deletions(-) diff --git a/include/hw/watchdog/cmsdk-apb-watchdog.h b/include/hw/watchdog/cmsdk-apb-watchdog.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/watchdog/cmsdk-apb-watchdog.h +++ b/include/hw/watchdog/cmsdk-apb-watchdog.h @@ -XXX,XX +XXX,XX @@ * * QEMU interface: * + QOM property "wdogclk-frq": frequency at which the watchdog is clocked + * + Clock input "WDOGCLK": clock for the watchdog's timer * + sysbus MMIO region 0: the register bank * + sysbus IRQ 0: watchdog interrupt * @@ -XXX,XX +XXX,XX @@ #include "hw/sysbus.h" #include "hw/ptimer.h" +#include "hw/clock.h" #include "qom/object.h" #define TYPE_CMSDK_APB_WATCHDOG "cmsdk-apb-watchdog" @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBWatchdog { uint32_t wdogclk_frq; bool is_luminary; struct ptimer_state *timer; + Clock *wdogclk; uint32_t control; uint32_t intstatus; diff --git a/hw/watchdog/cmsdk-apb-watchdog.c b/hw/watchdog/cmsdk-apb-watchdog.c index XXXXXXX..XXXXXXX 100644 --- a/hw/watchdog/cmsdk-apb-watchdog.c +++ b/hw/watchdog/cmsdk-apb-watchdog.c @@ -XXX,XX +XXX,XX @@ #include "hw/irq.h" #include "hw/qdev-properties.h" #include "hw/registerfields.h" +#include "hw/qdev-clock.h" #include "hw/watchdog/cmsdk-apb-watchdog.h" #include "migration/vmstate.h" @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_init(Object *obj) s, "cmsdk-apb-watchdog", 0x1000); sysbus_init_mmio(sbd, &s->iomem); sysbus_init_irq(sbd, &s->wdogint); + s->wdogclk = qdev_init_clock_in(DEVICE(s), "WDOGCLK", NULL, NULL); s->is_luminary = false; s->id = cmsdk_apb_watchdog_id; @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_realize(DeviceState *dev, Error **errp) static const VMStateDescription cmsdk_apb_watchdog_vmstate = { .name = "cmsdk-apb-watchdog", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .fields = (VMStateField[]) { + VMSTATE_CLOCK(wdogclk, CMSDKAPBWatchdog), VMSTATE_PTIMER(timer, CMSDKAPBWatchdog), VMSTATE_UINT32(control, CMSDKAPBWatchdog), VMSTATE_UINT32(intstatus, CMSDKAPBWatchdog), -- 2.20.1 While we transition the ARMSSE code from integer properties specifying clock frequencies to Clock objects, we want to have the device provide both at once. We want the final name of the main input Clock to be "MAINCLK", following the hardware name. Unfortunately creating an input Clock with a name X creates an under-the-hood QOM property X; for "MAINCLK" this clashes with the existing UINT32 property of that name. Rename the UINT32 property to MAINCLK_FRQ so it can coexist with the MAINCLK Clock; once the transition is complete MAINCLK_FRQ will be deleted. Commit created with: perl -p -i -e 's/MAINCLK/MAINCLK_FRQ/g' hw/arm/{armsse,mps2-tz,musca}.c include/hw/arm/armsse.h Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-11-peter.maydell@linaro.org Message-id: 20210121190622.22000-11-peter.maydell@linaro.org --- include/hw/arm/armsse.h | 2 +- hw/arm/armsse.c | 6 +++--- hw/arm/mps2-tz.c | 2 +- hw/arm/musca.c | 2 +- 4 files changed, 6 insertions(+), 6 deletions(-) diff --git a/include/hw/arm/armsse.h b/include/hw/arm/armsse.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/armsse.h +++ b/include/hw/arm/armsse.h @@ -XXX,XX +XXX,XX @@ * QEMU interface: * + QOM property "memory" is a MemoryRegion containing the devices provided * by the board model. - * + QOM property "MAINCLK" is the frequency of the main system clock + * + QOM property "MAINCLK_FRQ" is the frequency of the main system clock * + QOM property "EXP_NUMIRQ" sets the number of expansion interrupts. * (In hardware, the SSE-200 permits the number of expansion interrupts * for the two CPUs to be configured separately, but we restrict it to diff --git a/hw/arm/armsse.c b/hw/arm/armsse.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/armsse.c +++ b/hw/arm/armsse.c @@ -XXX,XX +XXX,XX @@ static Property iotkit_properties[] = { DEFINE_PROP_LINK("memory", ARMSSE, board_memory, TYPE_MEMORY_REGION, MemoryRegion *), DEFINE_PROP_UINT32("EXP_NUMIRQ", ARMSSE, exp_numirq, 64), - DEFINE_PROP_UINT32("MAINCLK", ARMSSE, mainclk_frq, 0), + DEFINE_PROP_UINT32("MAINCLK_FRQ", ARMSSE, mainclk_frq, 0), DEFINE_PROP_UINT32("SRAM_ADDR_WIDTH", ARMSSE, sram_addr_width, 15), DEFINE_PROP_UINT32("init-svtor", ARMSSE, init_svtor, 0x10000000), DEFINE_PROP_BOOL("CPU0_FPU", ARMSSE, cpu_fpu[0], true), @@ -XXX,XX +XXX,XX @@ static Property armsse_properties[] = { DEFINE_PROP_LINK("memory", ARMSSE, board_memory, TYPE_MEMORY_REGION, MemoryRegion *), DEFINE_PROP_UINT32("EXP_NUMIRQ", ARMSSE, exp_numirq, 64), - DEFINE_PROP_UINT32("MAINCLK", ARMSSE, mainclk_frq, 0), + DEFINE_PROP_UINT32("MAINCLK_FRQ", ARMSSE, mainclk_frq, 0), DEFINE_PROP_UINT32("SRAM_ADDR_WIDTH", ARMSSE, sram_addr_width, 15), DEFINE_PROP_UINT32("init-svtor", ARMSSE, init_svtor, 0x10000000), DEFINE_PROP_BOOL("CPU0_FPU", ARMSSE, cpu_fpu[0], false), @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) } if (!s->mainclk_frq) { - error_setg(errp, "MAINCLK property was not set"); + error_setg(errp, "MAINCLK_FRQ property was not set"); return; } diff --git a/hw/arm/mps2-tz.c b/hw/arm/mps2-tz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2-tz.c +++ b/hw/arm/mps2-tz.c @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) object_property_set_link(OBJECT(&mms->iotkit), "memory", OBJECT(system_memory), &error_abort); qdev_prop_set_uint32(iotkitdev, "EXP_NUMIRQ", MPS2TZ_NUMIRQ); - qdev_prop_set_uint32(iotkitdev, "MAINCLK", SYSCLK_FRQ); + qdev_prop_set_uint32(iotkitdev, "MAINCLK_FRQ", SYSCLK_FRQ); sysbus_realize(SYS_BUS_DEVICE(&mms->iotkit), &error_fatal); /* diff --git a/hw/arm/musca.c b/hw/arm/musca.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/musca.c +++ b/hw/arm/musca.c @@ -XXX,XX +XXX,XX @@ static void musca_init(MachineState *machine) qdev_prop_set_uint32(ssedev, "EXP_NUMIRQ", mmc->num_irqs); qdev_prop_set_uint32(ssedev, "init-svtor", mmc->init_svtor); qdev_prop_set_uint32(ssedev, "SRAM_ADDR_WIDTH", mmc->sram_addr_width); - qdev_prop_set_uint32(ssedev, "MAINCLK", SYSCLK_FRQ); + qdev_prop_set_uint32(ssedev, "MAINCLK_FRQ", SYSCLK_FRQ); /* * Musca-A takes the default SSE-200 FPU/DSP settings (ie no for * CPU0 and yes for CPU1); Musca-B1 explicitly enables them for CPU0. -- 2.20.1 Create two input clocks on the ARMSSE devices, one for the normal MAINCLK, and one for the 32KHz S32KCLK, and wire these up to the appropriate devices. The old property-based clock frequency setting will remain in place until conversion is complete. This is a migration compatibility break for machines mps2-an505, mps2-an521, musca-a, musca-b1. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-12-peter.maydell@linaro.org Message-id: 20210121190622.22000-12-peter.maydell@linaro.org --- include/hw/arm/armsse.h | 6 ++++++ hw/arm/armsse.c | 17 +++++++++++++++-- 2 files changed, 21 insertions(+), 2 deletions(-) diff --git a/include/hw/arm/armsse.h b/include/hw/arm/armsse.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/armsse.h +++ b/include/hw/arm/armsse.h @@ -XXX,XX +XXX,XX @@ * per-CPU identity and control register blocks * * QEMU interface: + * + Clock input "MAINCLK": clock for CPUs and most peripherals + * + Clock input "S32KCLK": slow 32KHz clock used for a few peripherals * + QOM property "memory" is a MemoryRegion containing the devices provided * by the board model. * + QOM property "MAINCLK_FRQ" is the frequency of the main system clock @@ -XXX,XX +XXX,XX @@ #include "hw/misc/armsse-mhu.h" #include "hw/misc/unimp.h" #include "hw/or-irq.h" +#include "hw/clock.h" #include "hw/core/split-irq.h" #include "hw/cpu/cluster.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct ARMSSE { uint32_t nsccfg; + Clock *mainclk; + Clock *s32kclk; + /* Properties */ MemoryRegion *board_memory; uint32_t exp_numirq; diff --git a/hw/arm/armsse.c b/hw/arm/armsse.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/armsse.c +++ b/hw/arm/armsse.c @@ -XXX,XX +XXX,XX @@ #include "hw/arm/armsse.h" #include "hw/arm/boot.h" #include "hw/irq.h" +#include "hw/qdev-clock.h" /* Format of the System Information block SYS_CONFIG register */ typedef enum SysConfigFormat { @@ -XXX,XX +XXX,XX @@ static void armsse_init(Object *obj) assert(info->sram_banks <= MAX_SRAM_BANKS); assert(info->num_cpus <= SSE_MAX_CPUS); + s->mainclk = qdev_init_clock_in(DEVICE(s), "MAINCLK", NULL, NULL); + s->s32kclk = qdev_init_clock_in(DEVICE(s), "S32KCLK", NULL, NULL); + memory_region_init(&s->container, obj, "armsse-container", UINT64_MAX); for (i = 0; i < info->num_cpus; i++) { @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) * map its upstream ends to the right place in the container. */ qdev_prop_set_uint32(DEVICE(&s->timer0), "pclk-frq", s->mainclk_frq); + qdev_connect_clock_in(DEVICE(&s->timer0), "pclk", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->timer0), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) &error_abort); qdev_prop_set_uint32(DEVICE(&s->timer1), "pclk-frq", s->mainclk_frq); + qdev_connect_clock_in(DEVICE(&s->timer1), "pclk", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->timer1), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) &error_abort); qdev_prop_set_uint32(DEVICE(&s->dualtimer), "pclk-frq", s->mainclk_frq); + qdev_connect_clock_in(DEVICE(&s->dualtimer), "TIMCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->dualtimer), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) * 0x4002f000: S32K timer */ qdev_prop_set_uint32(DEVICE(&s->s32ktimer), "pclk-frq", S32KCLK); + qdev_connect_clock_in(DEVICE(&s->s32ktimer), "pclk", s->s32kclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->s32ktimer), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) qdev_get_gpio_in_named(DEVICE(&s->armv7m), "NMI", 0)); qdev_prop_set_uint32(DEVICE(&s->s32kwatchdog), "wdogclk-frq", S32KCLK); + qdev_connect_clock_in(DEVICE(&s->s32kwatchdog), "WDOGCLK", s->s32kclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->s32kwatchdog), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) /* 0x40080000 .. 0x4008ffff : ARMSSE second Base peripheral region */ qdev_prop_set_uint32(DEVICE(&s->nswatchdog), "wdogclk-frq", s->mainclk_frq); + qdev_connect_clock_in(DEVICE(&s->nswatchdog), "WDOGCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->nswatchdog), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) sysbus_mmio_map(SYS_BUS_DEVICE(&s->nswatchdog), 0, 0x40081000); qdev_prop_set_uint32(DEVICE(&s->swatchdog), "wdogclk-frq", s->mainclk_frq); + qdev_connect_clock_in(DEVICE(&s->swatchdog), "WDOGCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->swatchdog), errp)) { return; } @@ -XXX,XX +XXX,XX @@ static void armsse_idau_check(IDAUInterface *ii, uint32_t address, static const VMStateDescription armsse_vmstate = { .name = "iotkit", - .version_id = 1, - .minimum_version_id = 1, + .version_id = 2, + .minimum_version_id = 2, .fields = (VMStateField[]) { + VMSTATE_CLOCK(mainclk, ARMSSE), + VMSTATE_CLOCK(s32kclk, ARMSSE), VMSTATE_UINT32(nsccfg, ARMSSE), VMSTATE_END_OF_LIST() } -- 2.20.1 The old-style convenience function cmsdk_apb_timer_create() for creating CMSDK_APB_TIMER objects is used in only two places in mps2.c. Most of the rest of the code in that file uses the new "initialize in place" coding style. We want to connect up a Clock object which should be done between the object creation and realization; rather than adding a Clock* argument to the convenience function, convert the timer creation code in mps2.c to the same style as is used already for the watchdog, dualtimer and other devices, and delete the now-unused convenience function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-13-peter.maydell@linaro.org Message-id: 20210121190622.22000-13-peter.maydell@linaro.org --- include/hw/timer/cmsdk-apb-timer.h | 21 --------------------- hw/arm/mps2.c | 18 ++++++++++++++++-- 2 files changed, 16 insertions(+), 23 deletions(-) diff --git a/include/hw/timer/cmsdk-apb-timer.h b/include/hw/timer/cmsdk-apb-timer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-timer.h +++ b/include/hw/timer/cmsdk-apb-timer.h @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBTimer { uint32_t intstatus; }; -/** - * cmsdk_apb_timer_create - convenience function to create TYPE_CMSDK_APB_TIMER - * @addr: location in system memory to map registers - * @pclk_frq: frequency in Hz of the PCLK clock (used for calculating baud rate) - */ -static inline DeviceState *cmsdk_apb_timer_create(hwaddr addr, - qemu_irq timerint, - uint32_t pclk_frq) -{ - DeviceState *dev; - SysBusDevice *s; - - dev = qdev_new(TYPE_CMSDK_APB_TIMER); - s = SYS_BUS_DEVICE(dev); - qdev_prop_set_uint32(dev, "pclk-frq", pclk_frq); - sysbus_realize_and_unref(s, &error_fatal); - sysbus_mmio_map(s, 0, addr); - sysbus_connect_irq(s, 0, timerint); - return dev; -} - #endif diff --git a/hw/arm/mps2.c b/hw/arm/mps2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2.c +++ b/hw/arm/mps2.c @@ -XXX,XX +XXX,XX @@ struct MPS2MachineState { /* CMSDK APB subsystem */ CMSDKAPBDualTimer dualtimer; CMSDKAPBWatchdog watchdog; + CMSDKAPBTimer timer[2]; }; #define TYPE_MPS2_MACHINE "mps2" @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) } /* CMSDK APB subsystem */ - cmsdk_apb_timer_create(0x40000000, qdev_get_gpio_in(armv7m, 8), SYSCLK_FRQ); - cmsdk_apb_timer_create(0x40001000, qdev_get_gpio_in(armv7m, 9), SYSCLK_FRQ); + for (i = 0; i < ARRAY_SIZE(mms->timer); i++) { + g_autofree char *name = g_strdup_printf("timer%d", i); + hwaddr base = 0x40000000 + i * 0x1000; + int irqno = 8 + i; + SysBusDevice *sbd; + + object_initialize_child(OBJECT(mms), name, &mms->timer[i], + TYPE_CMSDK_APB_TIMER); + sbd = SYS_BUS_DEVICE(&mms->timer[i]); + qdev_prop_set_uint32(DEVICE(&mms->timer[i]), "pclk-frq", SYSCLK_FRQ); + sysbus_realize_and_unref(sbd, &error_fatal); + sysbus_mmio_map(sbd, 0, base); + sysbus_connect_irq(sbd, 0, qdev_get_gpio_in(armv7m, irqno)); + } + object_initialize_child(OBJECT(mms), "dualtimer", &mms->dualtimer, TYPE_CMSDK_APB_DUALTIMER); qdev_prop_set_uint32(DEVICE(&mms->dualtimer), "pclk-frq", SYSCLK_FRQ); -- 2.20.1 Create a fixed-frequency Clock object to be the SYSCLK, and wire it up to the devices that require it. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-14-peter.maydell@linaro.org Message-id: 20210121190622.22000-14-peter.maydell@linaro.org --- hw/arm/mps2.c | 9 +++++++++ 1 file changed, 9 insertions(+) diff --git a/hw/arm/mps2.c b/hw/arm/mps2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2.c +++ b/hw/arm/mps2.c @@ -XXX,XX +XXX,XX @@ #include "hw/net/lan9118.h" #include "net/net.h" #include "hw/watchdog/cmsdk-apb-watchdog.h" +#include "hw/qdev-clock.h" #include "qom/object.h" typedef enum MPS2FPGAType { @@ -XXX,XX +XXX,XX @@ struct MPS2MachineState { CMSDKAPBDualTimer dualtimer; CMSDKAPBWatchdog watchdog; CMSDKAPBTimer timer[2]; + Clock *sysclk; }; #define TYPE_MPS2_MACHINE "mps2" @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) exit(EXIT_FAILURE); } + /* This clock doesn't need migration because it is fixed-frequency */ + mms->sysclk = clock_new(OBJECT(machine), "SYSCLK"); + clock_set_hz(mms->sysclk, SYSCLK_FRQ); + /* The FPGA images have an odd combination of different RAMs, * because in hardware they are different implementations and * connected to different buses, giving varying performance/size @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) TYPE_CMSDK_APB_TIMER); sbd = SYS_BUS_DEVICE(&mms->timer[i]); qdev_prop_set_uint32(DEVICE(&mms->timer[i]), "pclk-frq", SYSCLK_FRQ); + qdev_connect_clock_in(DEVICE(&mms->timer[i]), "pclk", mms->sysclk); sysbus_realize_and_unref(sbd, &error_fatal); sysbus_mmio_map(sbd, 0, base); sysbus_connect_irq(sbd, 0, qdev_get_gpio_in(armv7m, irqno)); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) object_initialize_child(OBJECT(mms), "dualtimer", &mms->dualtimer, TYPE_CMSDK_APB_DUALTIMER); qdev_prop_set_uint32(DEVICE(&mms->dualtimer), "pclk-frq", SYSCLK_FRQ); + qdev_connect_clock_in(DEVICE(&mms->dualtimer), "TIMCLK", mms->sysclk); sysbus_realize(SYS_BUS_DEVICE(&mms->dualtimer), &error_fatal); sysbus_connect_irq(SYS_BUS_DEVICE(&mms->dualtimer), 0, qdev_get_gpio_in(armv7m, 10)); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) object_initialize_child(OBJECT(mms), "watchdog", &mms->watchdog, TYPE_CMSDK_APB_WATCHDOG); qdev_prop_set_uint32(DEVICE(&mms->watchdog), "wdogclk-frq", SYSCLK_FRQ); + qdev_connect_clock_in(DEVICE(&mms->watchdog), "WDOGCLK", mms->sysclk); sysbus_realize(SYS_BUS_DEVICE(&mms->watchdog), &error_fatal); sysbus_connect_irq(SYS_BUS_DEVICE(&mms->watchdog), 0, qdev_get_gpio_in_named(armv7m, "NMI", 0)); -- 2.20.1 Create and connect the two clocks needed by the ARMSSE. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-15-peter.maydell@linaro.org Message-id: 20210121190622.22000-15-peter.maydell@linaro.org --- hw/arm/mps2-tz.c | 13 +++++++++++++ 1 file changed, 13 insertions(+) diff --git a/hw/arm/mps2-tz.c b/hw/arm/mps2-tz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2-tz.c +++ b/hw/arm/mps2-tz.c @@ -XXX,XX +XXX,XX @@ #include "hw/net/lan9118.h" #include "net/net.h" #include "hw/core/split-irq.h" +#include "hw/qdev-clock.h" #include "qom/object.h" #define MPS2TZ_NUMIRQ 92 @@ -XXX,XX +XXX,XX @@ struct MPS2TZMachineState { qemu_or_irq uart_irq_orgate; DeviceState *lan9118; SplitIRQ cpu_irq_splitter[MPS2TZ_NUMIRQ]; + Clock *sysclk; + Clock *s32kclk; }; #define TYPE_MPS2TZ_MACHINE "mps2tz" @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_TYPE(MPS2TZMachineState, MPS2TZMachineClass, MPS2TZ_MACHINE) /* Main SYSCLK frequency in Hz */ #define SYSCLK_FRQ 20000000 +/* Slow 32Khz S32KCLK frequency in Hz */ +#define S32KCLK_FRQ (32 * 1000) /* Create an alias of an entire original MemoryRegion @orig * located at @base in the memory map. @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) exit(EXIT_FAILURE); } + /* These clocks don't need migration because they are fixed-frequency */ + mms->sysclk = clock_new(OBJECT(machine), "SYSCLK"); + clock_set_hz(mms->sysclk, SYSCLK_FRQ); + mms->s32kclk = clock_new(OBJECT(machine), "S32KCLK"); + clock_set_hz(mms->s32kclk, S32KCLK_FRQ); + object_initialize_child(OBJECT(machine), TYPE_IOTKIT, &mms->iotkit, mmc->armsse_type); iotkitdev = DEVICE(&mms->iotkit); @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) OBJECT(system_memory), &error_abort); qdev_prop_set_uint32(iotkitdev, "EXP_NUMIRQ", MPS2TZ_NUMIRQ); qdev_prop_set_uint32(iotkitdev, "MAINCLK_FRQ", SYSCLK_FRQ); + qdev_connect_clock_in(iotkitdev, "MAINCLK", mms->sysclk); + qdev_connect_clock_in(iotkitdev, "S32KCLK", mms->s32kclk); sysbus_realize(SYS_BUS_DEVICE(&mms->iotkit), &error_fatal); /* -- 2.20.1 Create and connect the two clocks needed by the ARMSSE. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-16-peter.maydell@linaro.org Message-id: 20210121190622.22000-16-peter.maydell@linaro.org --- hw/arm/musca.c | 12 ++++++++++++ 1 file changed, 12 insertions(+) diff --git a/hw/arm/musca.c b/hw/arm/musca.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/musca.c +++ b/hw/arm/musca.c @@ -XXX,XX +XXX,XX @@ #include "hw/misc/tz-ppc.h" #include "hw/misc/unimp.h" #include "hw/rtc/pl031.h" +#include "hw/qdev-clock.h" #include "qom/object.h" #define MUSCA_NUMIRQ_MAX 96 @@ -XXX,XX +XXX,XX @@ struct MuscaMachineState { UnimplementedDeviceState sdio; UnimplementedDeviceState gpio; UnimplementedDeviceState cryptoisland; + Clock *sysclk; + Clock *s32kclk; }; #define TYPE_MUSCA_MACHINE "musca" @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_TYPE(MuscaMachineState, MuscaMachineClass, MUSCA_MACHINE) * don't model that in our SSE-200 model yet. */ #define SYSCLK_FRQ 40000000 +/* Slow 32Khz S32KCLK frequency in Hz */ +#define S32KCLK_FRQ (32 * 1000) static qemu_irq get_sse_irq_in(MuscaMachineState *mms, int irqno) { @@ -XXX,XX +XXX,XX @@ static void musca_init(MachineState *machine) exit(1); } + mms->sysclk = clock_new(OBJECT(machine), "SYSCLK"); + clock_set_hz(mms->sysclk, SYSCLK_FRQ); + mms->s32kclk = clock_new(OBJECT(machine), "S32KCLK"); + clock_set_hz(mms->s32kclk, S32KCLK_FRQ); + object_initialize_child(OBJECT(machine), "sse-200", &mms->sse, TYPE_SSE200); ssedev = DEVICE(&mms->sse); @@ -XXX,XX +XXX,XX @@ static void musca_init(MachineState *machine) qdev_prop_set_uint32(ssedev, "init-svtor", mmc->init_svtor); qdev_prop_set_uint32(ssedev, "SRAM_ADDR_WIDTH", mmc->sram_addr_width); qdev_prop_set_uint32(ssedev, "MAINCLK_FRQ", SYSCLK_FRQ); + qdev_connect_clock_in(ssedev, "MAINCLK", mms->sysclk); + qdev_connect_clock_in(ssedev, "S32KCLK", mms->s32kclk); /* * Musca-A takes the default SSE-200 FPU/DSP settings (ie no for * CPU0 and yes for CPU1); Musca-B1 explicitly enables them for CPU0. -- 2.20.1 Convert the SSYS code in the Stellaris boards (which encapsulates the system registers) to a proper QOM device. This will provide us with somewhere to put the output Clock whose frequency depends on the setting of the PLL configuration registers. This is a migration compatibility break for lm3s811evb, lm3s6965evb. We use 3-phase reset here because the Clock will need to propagate its value in the hold phase. For the moment we reset the device during the board creation so that the system_clock_scale global gets set; this will be removed in a subsequent commit. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-17-peter.maydell@linaro.org Message-id: 20210121190622.22000-17-peter.maydell@linaro.org Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> --- hw/arm/stellaris.c | 132 ++++++++++++++++++++++++++++++++++++--------- 1 file changed, 107 insertions(+), 25 deletions(-) diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ static void stellaris_gptm_realize(DeviceState *dev, Error **errp) /* System controller. */ -typedef struct { +#define TYPE_STELLARIS_SYS "stellaris-sys" +OBJECT_DECLARE_SIMPLE_TYPE(ssys_state, STELLARIS_SYS) + +struct ssys_state { + SysBusDevice parent_obj; + MemoryRegion iomem; uint32_t pborctl; uint32_t ldopctl; @@ -XXX,XX +XXX,XX @@ typedef struct { uint32_t dcgc[3]; uint32_t clkvclr; uint32_t ldoarst; + qemu_irq irq; + /* Properties (all read-only registers) */ uint32_t user0; uint32_t user1; - qemu_irq irq; - stellaris_board_info *board; -} ssys_state; + uint32_t did0; + uint32_t did1; + uint32_t dc0; + uint32_t dc1; + uint32_t dc2; + uint32_t dc3; + uint32_t dc4; +}; static void ssys_update(ssys_state *s) { @@ -XXX,XX +XXX,XX @@ static uint32_t pllcfg_fury[16] = { static int ssys_board_class(const ssys_state *s) { - uint32_t did0 = s->board->did0; + uint32_t did0 = s->did0; switch (did0 & DID0_VER_MASK) { case DID0_VER_0: return DID0_CLASS_SANDSTORM; @@ -XXX,XX +XXX,XX @@ static uint64_t ssys_read(void *opaque, hwaddr offset, switch (offset) { case 0x000: /* DID0 */ - return s->board->did0; + return s->did0; case 0x004: /* DID1 */ - return s->board->did1; + return s->did1; case 0x008: /* DC0 */ - return s->board->dc0; + return s->dc0; case 0x010: /* DC1 */ - return s->board->dc1; + return s->dc1; case 0x014: /* DC2 */ - return s->board->dc2; + return s->dc2; case 0x018: /* DC3 */ - return s->board->dc3; + return s->dc3; case 0x01c: /* DC4 */ - return s->board->dc4; + return s->dc4; case 0x030: /* PBORCTL */ return s->pborctl; case 0x034: /* LDOPCTL */ @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps ssys_ops = { .endianness = DEVICE_NATIVE_ENDIAN, }; -static void ssys_reset(void *opaque) +static void stellaris_sys_reset_enter(Object *obj, ResetType type) { - ssys_state *s = (ssys_state *)opaque; + ssys_state *s = STELLARIS_SYS(obj); s->pborctl = 0x7ffd; s->rcc = 0x078e3ac0; @@ -XXX,XX +XXX,XX @@ static void ssys_reset(void *opaque) s->rcgc[0] = 1; s->scgc[0] = 1; s->dcgc[0] = 1; +} + +static void stellaris_sys_reset_hold(Object *obj) +{ + ssys_state *s = STELLARIS_SYS(obj); + ssys_calculate_system_clock(s); } +static void stellaris_sys_reset_exit(Object *obj) +{ +} + static int stellaris_sys_post_load(void *opaque, int version_id) { ssys_state *s = opaque; @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_stellaris_sys = { } }; +static Property stellaris_sys_properties[] = { + DEFINE_PROP_UINT32("user0", ssys_state, user0, 0), + DEFINE_PROP_UINT32("user1", ssys_state, user1, 0), + DEFINE_PROP_UINT32("did0", ssys_state, did0, 0), + DEFINE_PROP_UINT32("did1", ssys_state, did1, 0), + DEFINE_PROP_UINT32("dc0", ssys_state, dc0, 0), + DEFINE_PROP_UINT32("dc1", ssys_state, dc1, 0), + DEFINE_PROP_UINT32("dc2", ssys_state, dc2, 0), + DEFINE_PROP_UINT32("dc3", ssys_state, dc3, 0), + DEFINE_PROP_UINT32("dc4", ssys_state, dc4, 0), + DEFINE_PROP_END_OF_LIST() +}; + +static void stellaris_sys_instance_init(Object *obj) +{ + ssys_state *s = STELLARIS_SYS(obj); + SysBusDevice *sbd = SYS_BUS_DEVICE(s); + + memory_region_init_io(&s->iomem, obj, &ssys_ops, s, "ssys", 0x00001000); + sysbus_init_mmio(sbd, &s->iomem); + sysbus_init_irq(sbd, &s->irq); +} + static int stellaris_sys_init(uint32_t base, qemu_irq irq, stellaris_board_info * board, uint8_t *macaddr) { - ssys_state *s; + DeviceState *dev = qdev_new(TYPE_STELLARIS_SYS); + SysBusDevice *sbd = SYS_BUS_DEVICE(dev); - s = g_new0(ssys_state, 1); - s->irq = irq; - s->board = board; /* Most devices come preprogrammed with a MAC address in the user data. */ - s->user0 = macaddr[0] | (macaddr[1] << 8) | (macaddr[2] << 16); - s->user1 = macaddr[3] | (macaddr[4] << 8) | (macaddr[5] << 16); + qdev_prop_set_uint32(dev, "user0", + macaddr[0] | (macaddr[1] << 8) | (macaddr[2] << 16)); + qdev_prop_set_uint32(dev, "user1", + macaddr[3] | (macaddr[4] << 8) | (macaddr[5] << 16)); + qdev_prop_set_uint32(dev, "did0", board->did0); + qdev_prop_set_uint32(dev, "did1", board->did1); + qdev_prop_set_uint32(dev, "dc0", board->dc0); + qdev_prop_set_uint32(dev, "dc1", board->dc1); + qdev_prop_set_uint32(dev, "dc2", board->dc2); + qdev_prop_set_uint32(dev, "dc3", board->dc3); + qdev_prop_set_uint32(dev, "dc4", board->dc4); + + sysbus_realize_and_unref(sbd, &error_fatal); + sysbus_mmio_map(sbd, 0, base); + sysbus_connect_irq(sbd, 0, irq); + + /* + * Normally we should not be resetting devices like this during + * board creation. For the moment we need to do so, because + * system_clock_scale will only get set when the STELLARIS_SYS + * device is reset, and we need its initial value to pass to + * the watchdog device. This hack can be removed once the + * watchdog has been converted to use a Clock input instead. + */ + device_cold_reset(dev); - memory_region_init_io(&s->iomem, NULL, &ssys_ops, s, "ssys", 0x00001000); - memory_region_add_subregion(get_system_memory(), base, &s->iomem); - ssys_reset(s); - vmstate_register(NULL, VMSTATE_INSTANCE_ID_ANY, &vmstate_stellaris_sys, s); return 0; } - /* I2C controller. */ #define TYPE_STELLARIS_I2C "stellaris-i2c" @@ -XXX,XX +XXX,XX @@ static const TypeInfo stellaris_adc_info = { .class_init = stellaris_adc_class_init, }; +static void stellaris_sys_class_init(ObjectClass *klass, void *data) +{ + DeviceClass *dc = DEVICE_CLASS(klass); + ResettableClass *rc = RESETTABLE_CLASS(klass); + + dc->vmsd = &vmstate_stellaris_sys; + rc->phases.enter = stellaris_sys_reset_enter; + rc->phases.hold = stellaris_sys_reset_hold; + rc->phases.exit = stellaris_sys_reset_exit; + device_class_set_props(dc, stellaris_sys_properties); +} + +static const TypeInfo stellaris_sys_info = { + .name = TYPE_STELLARIS_SYS, + .parent = TYPE_SYS_BUS_DEVICE, + .instance_size = sizeof(ssys_state), + .instance_init = stellaris_sys_instance_init, + .class_init = stellaris_sys_class_init, +}; + static void stellaris_register_types(void) { type_register_static(&stellaris_i2c_info); type_register_static(&stellaris_gptm_info); type_register_static(&stellaris_adc_info); + type_register_static(&stellaris_sys_info); } type_init(stellaris_register_types) -- 2.20.1 Create and connect the Clock input for the watchdog device on the Stellaris boards. Because the Stellaris boards model the ability to change the clock rate by programming PLL registers, we have to create an output Clock on the ssys_state device and wire it up to the watchdog. Note that the old comment on ssys_calculate_system_clock() got the units wrong -- system_clock_scale is in nanoseconds, not milliseconds. Improve the commentary to clarify how we are calculating the period. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-18-peter.maydell@linaro.org Message-id: 20210121190622.22000-18-peter.maydell@linaro.org Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> --- hw/arm/stellaris.c | 43 +++++++++++++++++++++++++++++++------------ 1 file changed, 31 insertions(+), 12 deletions(-) diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ #include "hw/watchdog/cmsdk-apb-watchdog.h" #include "migration/vmstate.h" #include "hw/misc/unimp.h" +#include "hw/qdev-clock.h" #include "cpu.h" #include "qom/object.h" @@ -XXX,XX +XXX,XX @@ struct ssys_state { uint32_t clkvclr; uint32_t ldoarst; qemu_irq irq; + Clock *sysclk; /* Properties (all read-only registers) */ uint32_t user0; uint32_t user1; @@ -XXX,XX +XXX,XX @@ static bool ssys_use_rcc2(ssys_state *s) } /* - * Caculate the sys. clock period in ms. + * Calculate the system clock period. We only want to propagate + * this change to the rest of the system if we're not being called + * from migration post-load. */ -static void ssys_calculate_system_clock(ssys_state *s) +static void ssys_calculate_system_clock(ssys_state *s, bool propagate_clock) { + /* + * SYSDIV field specifies divisor: 0 == /1, 1 == /2, etc. Input + * clock is 200MHz, which is a period of 5 ns. Dividing the clock + * frequency by X is the same as multiplying the period by X. + */ if (ssys_use_rcc2(s)) { system_clock_scale = 5 * (((s->rcc2 >> 23) & 0x3f) + 1); } else { system_clock_scale = 5 * (((s->rcc >> 23) & 0xf) + 1); } + clock_set_ns(s->sysclk, system_clock_scale); + if (propagate_clock) { + clock_propagate(s->sysclk); + } } static void ssys_write(void *opaque, hwaddr offset, @@ -XXX,XX +XXX,XX @@ static void ssys_write(void *opaque, hwaddr offset, s->int_status |= (1 << 6); } s->rcc = value; - ssys_calculate_system_clock(s); + ssys_calculate_system_clock(s, true); break; case 0x070: /* RCC2 */ if (ssys_board_class(s) == DID0_CLASS_SANDSTORM) { @@ -XXX,XX +XXX,XX @@ static void ssys_write(void *opaque, hwaddr offset, s->int_status |= (1 << 6); } s->rcc2 = value; - ssys_calculate_system_clock(s); + ssys_calculate_system_clock(s, true); break; case 0x100: /* RCGC0 */ s->rcgc[0] = value; @@ -XXX,XX +XXX,XX @@ static void stellaris_sys_reset_hold(Object *obj) { ssys_state *s = STELLARIS_SYS(obj); - ssys_calculate_system_clock(s); + /* OK to propagate clocks from the hold phase */ + ssys_calculate_system_clock(s, true); } static void stellaris_sys_reset_exit(Object *obj) @@ -XXX,XX +XXX,XX @@ static int stellaris_sys_post_load(void *opaque, int version_id) { ssys_state *s = opaque; - ssys_calculate_system_clock(s); + ssys_calculate_system_clock(s, false); return 0; } @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_stellaris_sys = { VMSTATE_UINT32_ARRAY(dcgc, ssys_state, 3), VMSTATE_UINT32(clkvclr, ssys_state), VMSTATE_UINT32(ldoarst, ssys_state), + /* No field for sysclk -- handled in post-load instead */ VMSTATE_END_OF_LIST() } }; @@ -XXX,XX +XXX,XX @@ static void stellaris_sys_instance_init(Object *obj) memory_region_init_io(&s->iomem, obj, &ssys_ops, s, "ssys", 0x00001000); sysbus_init_mmio(sbd, &s->iomem); sysbus_init_irq(sbd, &s->irq); + s->sysclk = qdev_init_clock_out(DEVICE(s), "SYSCLK"); } -static int stellaris_sys_init(uint32_t base, qemu_irq irq, - stellaris_board_info * board, - uint8_t *macaddr) +static DeviceState *stellaris_sys_init(uint32_t base, qemu_irq irq, + stellaris_board_info *board, + uint8_t *macaddr) { DeviceState *dev = qdev_new(TYPE_STELLARIS_SYS); SysBusDevice *sbd = SYS_BUS_DEVICE(dev); @@ -XXX,XX +XXX,XX @@ static int stellaris_sys_init(uint32_t base, qemu_irq irq, */ device_cold_reset(dev); - return 0; + return dev; } /* I2C controller. */ @@ -XXX,XX +XXX,XX @@ static void stellaris_init(MachineState *ms, stellaris_board_info *board) int flash_size; I2CBus *i2c; DeviceState *dev; + DeviceState *ssys_dev; int i; int j; @@ -XXX,XX +XXX,XX @@ static void stellaris_init(MachineState *ms, stellaris_board_info *board) } } - stellaris_sys_init(0x400fe000, qdev_get_gpio_in(nvic, 28), - board, nd_table[0].macaddr.a); + ssys_dev = stellaris_sys_init(0x400fe000, qdev_get_gpio_in(nvic, 28), + board, nd_table[0].macaddr.a); if (board->dc1 & (1 << 3)) { /* watchdog present */ @@ -XXX,XX +XXX,XX @@ static void stellaris_init(MachineState *ms, stellaris_board_info *board) /* system_clock_scale is valid now */ uint32_t mainclk = NANOSECONDS_PER_SECOND / system_clock_scale; qdev_prop_set_uint32(dev, "wdogclk-frq", mainclk); + qdev_connect_clock_in(dev, "WDOGCLK", + qdev_get_clock_out(ssys_dev, "SYSCLK")); sysbus_realize_and_unref(SYS_BUS_DEVICE(dev), &error_fatal); sysbus_mmio_map(SYS_BUS_DEVICE(dev), -- 2.20.1 Switch the CMSDK APB timer device over to using its Clock input; the pclk-frq property is now ignored. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-19-peter.maydell@linaro.org Message-id: 20210121190622.22000-19-peter.maydell@linaro.org --- hw/timer/cmsdk-apb-timer.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/hw/timer/cmsdk-apb-timer.c b/hw/timer/cmsdk-apb-timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-timer.c +++ b/hw/timer/cmsdk-apb-timer.c @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_reset(DeviceState *dev) ptimer_transaction_commit(s->timer); } +static void cmsdk_apb_timer_clk_update(void *opaque) +{ + CMSDKAPBTimer *s = CMSDK_APB_TIMER(opaque); + + ptimer_transaction_begin(s->timer); + ptimer_set_period_from_clock(s->timer, s->pclk, 1); + ptimer_transaction_commit(s->timer); +} + static void cmsdk_apb_timer_init(Object *obj) { SysBusDevice *sbd = SYS_BUS_DEVICE(obj); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_init(Object *obj) s, "cmsdk-apb-timer", 0x1000); sysbus_init_mmio(sbd, &s->iomem); sysbus_init_irq(sbd, &s->timerint); - s->pclk = qdev_init_clock_in(DEVICE(s), "pclk", NULL, NULL); + s->pclk = qdev_init_clock_in(DEVICE(s), "pclk", + cmsdk_apb_timer_clk_update, s); } static void cmsdk_apb_timer_realize(DeviceState *dev, Error **errp) { CMSDKAPBTimer *s = CMSDK_APB_TIMER(dev); - if (s->pclk_frq == 0) { - error_setg(errp, "CMSDK APB timer: pclk-frq property must be set"); + if (!clock_has_source(s->pclk)) { + error_setg(errp, "CMSDK APB timer: pclk clock must be connected"); return; } @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_realize(DeviceState *dev, Error **errp) PTIMER_POLICY_NO_COUNTER_ROUND_DOWN); ptimer_transaction_begin(s->timer); - ptimer_set_freq(s->timer, s->pclk_frq); + ptimer_set_period_from_clock(s->timer, s->pclk, 1); ptimer_transaction_commit(s->timer); } -- 2.20.1 Switch the CMSDK APB dualtimer device over to using its Clock input; the pclk-frq property is now ignored. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-20-peter.maydell@linaro.org Message-id: 20210121190622.22000-20-peter.maydell@linaro.org Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> --- hw/timer/cmsdk-apb-dualtimer.c | 42 ++++++++++++++++++++++++++++++---- 1 file changed, 37 insertions(+), 5 deletions(-) diff --git a/hw/timer/cmsdk-apb-dualtimer.c b/hw/timer/cmsdk-apb-dualtimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-dualtimer.c +++ b/hw/timer/cmsdk-apb-dualtimer.c @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_update(CMSDKAPBDualTimer *s) qemu_set_irq(s->timerintc, timintc); } +static int cmsdk_dualtimermod_divisor(CMSDKAPBDualTimerModule *m) +{ + /* Return the divisor set by the current CONTROL.PRESCALE value */ + switch (FIELD_EX32(m->control, CONTROL, PRESCALE)) { + case 0: + return 1; + case 1: + return 16; + case 2: + case 3: /* UNDEFINED, we treat like 2 (and complained when it was set) */ + return 256; + default: + g_assert_not_reached(); + } +} + static void cmsdk_dualtimermod_write_control(CMSDKAPBDualTimerModule *m, uint32_t newctrl) { @@ -XXX,XX +XXX,XX @@ static void cmsdk_dualtimermod_write_control(CMSDKAPBDualTimerModule *m, default: g_assert_not_reached(); } - ptimer_set_freq(m->timer, m->parent->pclk_frq / divisor); + ptimer_set_period_from_clock(m->timer, m->parent->timclk, divisor); } if (changed & R_CONTROL_MODE_MASK) { @@ -XXX,XX +XXX,XX @@ static void cmsdk_dualtimermod_reset(CMSDKAPBDualTimerModule *m) * limit must both be set to 0xffff, so we wrap at 16 bits. */ ptimer_set_limit(m->timer, 0xffff, 1); - ptimer_set_freq(m->timer, m->parent->pclk_frq); + ptimer_set_period_from_clock(m->timer, m->parent->timclk, + cmsdk_dualtimermod_divisor(m)); ptimer_transaction_commit(m->timer); } @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_reset(DeviceState *dev) s->timeritop = 0; } +static void cmsdk_apb_dualtimer_clk_update(void *opaque) +{ + CMSDKAPBDualTimer *s = CMSDK_APB_DUALTIMER(opaque); + int i; + + for (i = 0; i < ARRAY_SIZE(s->timermod); i++) { + CMSDKAPBDualTimerModule *m = &s->timermod[i]; + ptimer_transaction_begin(m->timer); + ptimer_set_period_from_clock(m->timer, m->parent->timclk, + cmsdk_dualtimermod_divisor(m)); + ptimer_transaction_commit(m->timer); + } +} + static void cmsdk_apb_dualtimer_init(Object *obj) { SysBusDevice *sbd = SYS_BUS_DEVICE(obj); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_init(Object *obj) for (i = 0; i < ARRAY_SIZE(s->timermod); i++) { sysbus_init_irq(sbd, &s->timermod[i].timerint); } - s->timclk = qdev_init_clock_in(DEVICE(s), "TIMCLK", NULL, NULL); + s->timclk = qdev_init_clock_in(DEVICE(s), "TIMCLK", + cmsdk_apb_dualtimer_clk_update, s); } static void cmsdk_apb_dualtimer_realize(DeviceState *dev, Error **errp) @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_realize(DeviceState *dev, Error **errp) CMSDKAPBDualTimer *s = CMSDK_APB_DUALTIMER(dev); int i; - if (s->pclk_frq == 0) { - error_setg(errp, "CMSDK APB timer: pclk-frq property must be set"); + if (!clock_has_source(s->timclk)) { + error_setg(errp, "CMSDK APB dualtimer: TIMCLK clock must be connected"); return; } -- 2.20.1 Switch the CMSDK APB watchdog device over to using its Clock input; the wdogclk_frq property is now ignored. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-21-peter.maydell@linaro.org Message-id: 20210121190622.22000-21-peter.maydell@linaro.org --- hw/watchdog/cmsdk-apb-watchdog.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/hw/watchdog/cmsdk-apb-watchdog.c b/hw/watchdog/cmsdk-apb-watchdog.c index XXXXXXX..XXXXXXX 100644 --- a/hw/watchdog/cmsdk-apb-watchdog.c +++ b/hw/watchdog/cmsdk-apb-watchdog.c @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_reset(DeviceState *dev) ptimer_transaction_commit(s->timer); } +static void cmsdk_apb_watchdog_clk_update(void *opaque) +{ + CMSDKAPBWatchdog *s = CMSDK_APB_WATCHDOG(opaque); + + ptimer_transaction_begin(s->timer); + ptimer_set_period_from_clock(s->timer, s->wdogclk, 1); + ptimer_transaction_commit(s->timer); +} + static void cmsdk_apb_watchdog_init(Object *obj) { SysBusDevice *sbd = SYS_BUS_DEVICE(obj); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_init(Object *obj) s, "cmsdk-apb-watchdog", 0x1000); sysbus_init_mmio(sbd, &s->iomem); sysbus_init_irq(sbd, &s->wdogint); - s->wdogclk = qdev_init_clock_in(DEVICE(s), "WDOGCLK", NULL, NULL); + s->wdogclk = qdev_init_clock_in(DEVICE(s), "WDOGCLK", + cmsdk_apb_watchdog_clk_update, s); s->is_luminary = false; s->id = cmsdk_apb_watchdog_id; @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_realize(DeviceState *dev, Error **errp) { CMSDKAPBWatchdog *s = CMSDK_APB_WATCHDOG(dev); - if (s->wdogclk_frq == 0) { + if (!clock_has_source(s->wdogclk)) { error_setg(errp, - "CMSDK APB watchdog: wdogclk-frq property must be set"); + "CMSDK APB watchdog: WDOGCLK clock must be connected"); return; } @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_realize(DeviceState *dev, Error **errp) PTIMER_POLICY_NO_COUNTER_ROUND_DOWN); ptimer_transaction_begin(s->timer); - ptimer_set_freq(s->timer, s->wdogclk_frq); + ptimer_set_period_from_clock(s->timer, s->wdogclk, 1); ptimer_transaction_commit(s->timer); } -- 2.20.1 Now that the CMSDK APB watchdog uses its Clock input, it will correctly respond when the system clock frequency is changed using the RCC register on in the Stellaris board system registers. Test that when the RCC register is written it causes the watchdog timer to change speed. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-22-peter.maydell@linaro.org Message-id: 20210121190622.22000-22-peter.maydell@linaro.org --- tests/qtest/cmsdk-apb-watchdog-test.c | 52 +++++++++++++++++++++++++++ 1 file changed, 52 insertions(+) diff --git a/tests/qtest/cmsdk-apb-watchdog-test.c b/tests/qtest/cmsdk-apb-watchdog-test.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/cmsdk-apb-watchdog-test.c +++ b/tests/qtest/cmsdk-apb-watchdog-test.c @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" +#include "qemu/bitops.h" #include "libqtest-single.h" /* @@ -XXX,XX +XXX,XX @@ #define WDOGMIS 0x14 #define WDOGLOCK 0xc00 +#define SSYS_BASE 0x400fe000 +#define RCC 0x60 +#define SYSDIV_SHIFT 23 +#define SYSDIV_LENGTH 4 + static void test_watchdog(void) { g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); @@ -XXX,XX +XXX,XX @@ static void test_watchdog(void) g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); } +static void test_clock_change(void) +{ + uint32_t rcc; + + /* + * Test that writing to the stellaris board's RCC register to + * change the system clock frequency causes the watchdog + * to change the speed it counts at. + */ + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); + + writel(WDOG_BASE + WDOGCONTROL, 1); + writel(WDOG_BASE + WDOGLOAD, 1000); + + /* Step to just past the 500th tick */ + clock_step(80 * 500 + 1); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 500); + + /* Rewrite RCC.SYSDIV from 16 to 8, so the clock is now 40ns per tick */ + rcc = readl(SSYS_BASE + RCC); + g_assert_cmpuint(extract32(rcc, SYSDIV_SHIFT, SYSDIV_LENGTH), ==, 0xf); + rcc = deposit32(rcc, SYSDIV_SHIFT, SYSDIV_LENGTH, 7); + writel(SSYS_BASE + RCC, rcc); + + /* Just past the 1000th tick: timer should have fired */ + clock_step(40 * 500); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 1); + + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 0); + + /* VALUE reloads at following tick */ + clock_step(41); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 1000); + + /* Writing any value to WDOGINTCLR clears the interrupt and reloads */ + clock_step(40 * 500); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 500); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 1); + writel(WDOG_BASE + WDOGINTCLR, 0); + g_assert_cmpuint(readl(WDOG_BASE + WDOGVALUE), ==, 1000); + g_assert_cmpuint(readl(WDOG_BASE + WDOGRIS), ==, 0); +} + int main(int argc, char **argv) { int r; @@ -XXX,XX +XXX,XX @@ int main(int argc, char **argv) qtest_start("-machine lm3s811evb"); qtest_add_func("/cmsdk-apb-watchdog/watchdog", test_watchdog); + qtest_add_func("/cmsdk-apb-watchdog/watchdog_clock_change", + test_clock_change); r = g_test_run(); -- 2.20.1 Use the MAINCLK Clock input to set the system_clock_scale variable rather than using the mainclk_frq property. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Message-id: 20210128114145.20536-23-peter.maydell@linaro.org Message-id: 20210121190622.22000-23-peter.maydell@linaro.org --- hw/arm/armsse.c | 24 +++++++++++++++++++----- 1 file changed, 19 insertions(+), 5 deletions(-) diff --git a/hw/arm/armsse.c b/hw/arm/armsse.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/armsse.c +++ b/hw/arm/armsse.c @@ -XXX,XX +XXX,XX @@ static void armsse_forward_sec_resp_cfg(ARMSSE *s) qdev_connect_gpio_out(dev_splitter, 2, s->sec_resp_cfg_in); } +static void armsse_mainclk_update(void *opaque) +{ + ARMSSE *s = ARM_SSE(opaque); + /* + * Set system_clock_scale from our Clock input; this is what + * controls the tick rate of the CPU SysTick timer. + */ + system_clock_scale = clock_ticks_to_ns(s->mainclk, 1); +} + static void armsse_init(Object *obj) { ARMSSE *s = ARM_SSE(obj); @@ -XXX,XX +XXX,XX @@ static void armsse_init(Object *obj) assert(info->sram_banks <= MAX_SRAM_BANKS); assert(info->num_cpus <= SSE_MAX_CPUS); - s->mainclk = qdev_init_clock_in(DEVICE(s), "MAINCLK", NULL, NULL); + s->mainclk = qdev_init_clock_in(DEVICE(s), "MAINCLK", + armsse_mainclk_update, s); s->s32kclk = qdev_init_clock_in(DEVICE(s), "S32KCLK", NULL, NULL); memory_region_init(&s->container, obj, "armsse-container", UINT64_MAX); @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) return; } - if (!s->mainclk_frq) { - error_setg(errp, "MAINCLK_FRQ property was not set"); - return; + if (!clock_has_source(s->mainclk)) { + error_setg(errp, "MAINCLK clock was not connected"); + } + if (!clock_has_source(s->s32kclk)) { + error_setg(errp, "S32KCLK clock was not connected"); } assert(info->num_cpus <= SSE_MAX_CPUS); @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) */ sysbus_init_mmio(SYS_BUS_DEVICE(s), &s->container); - system_clock_scale = NANOSECONDS_PER_SECOND / s->mainclk_frq; + /* Set initial system_clock_scale from MAINCLK */ + armsse_mainclk_update(s); } static void armsse_idau_check(IDAUInterface *ii, uint32_t address, -- 2.20.1 Remove all the code that sets frequency properties on the CMSDK timer, dualtimer and watchdog devices and on the ARMSSE SoC device: these properties are unused now that the devices rely on their Clock inputs instead. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-24-peter.maydell@linaro.org Message-id: 20210121190622.22000-24-peter.maydell@linaro.org --- hw/arm/armsse.c | 7 ------- hw/arm/mps2-tz.c | 1 - hw/arm/mps2.c | 3 --- hw/arm/musca.c | 1 - hw/arm/stellaris.c | 3 --- 5 files changed, 15 deletions(-) diff --git a/hw/arm/armsse.c b/hw/arm/armsse.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/armsse.c +++ b/hw/arm/armsse.c @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) * it to the appropriate PPC port; then we can realize the PPC and * map its upstream ends to the right place in the container. */ - qdev_prop_set_uint32(DEVICE(&s->timer0), "pclk-frq", s->mainclk_frq); qdev_connect_clock_in(DEVICE(&s->timer0), "pclk", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->timer0), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) object_property_set_link(OBJECT(&s->apb_ppc0), "port[0]", OBJECT(mr), &error_abort); - qdev_prop_set_uint32(DEVICE(&s->timer1), "pclk-frq", s->mainclk_frq); qdev_connect_clock_in(DEVICE(&s->timer1), "pclk", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->timer1), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) object_property_set_link(OBJECT(&s->apb_ppc0), "port[1]", OBJECT(mr), &error_abort); - qdev_prop_set_uint32(DEVICE(&s->dualtimer), "pclk-frq", s->mainclk_frq); qdev_connect_clock_in(DEVICE(&s->dualtimer), "TIMCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->dualtimer), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) /* Devices behind APB PPC1: * 0x4002f000: S32K timer */ - qdev_prop_set_uint32(DEVICE(&s->s32ktimer), "pclk-frq", S32KCLK); qdev_connect_clock_in(DEVICE(&s->s32ktimer), "pclk", s->s32kclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->s32ktimer), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) qdev_connect_gpio_out(DEVICE(&s->nmi_orgate), 0, qdev_get_gpio_in_named(DEVICE(&s->armv7m), "NMI", 0)); - qdev_prop_set_uint32(DEVICE(&s->s32kwatchdog), "wdogclk-frq", S32KCLK); qdev_connect_clock_in(DEVICE(&s->s32kwatchdog), "WDOGCLK", s->s32kclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->s32kwatchdog), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) /* 0x40080000 .. 0x4008ffff : ARMSSE second Base peripheral region */ - qdev_prop_set_uint32(DEVICE(&s->nswatchdog), "wdogclk-frq", s->mainclk_frq); qdev_connect_clock_in(DEVICE(&s->nswatchdog), "WDOGCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->nswatchdog), errp)) { return; @@ -XXX,XX +XXX,XX @@ static void armsse_realize(DeviceState *dev, Error **errp) armsse_get_common_irq_in(s, 1)); sysbus_mmio_map(SYS_BUS_DEVICE(&s->nswatchdog), 0, 0x40081000); - qdev_prop_set_uint32(DEVICE(&s->swatchdog), "wdogclk-frq", s->mainclk_frq); qdev_connect_clock_in(DEVICE(&s->swatchdog), "WDOGCLK", s->mainclk); if (!sysbus_realize(SYS_BUS_DEVICE(&s->swatchdog), errp)) { return; diff --git a/hw/arm/mps2-tz.c b/hw/arm/mps2-tz.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2-tz.c +++ b/hw/arm/mps2-tz.c @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) object_property_set_link(OBJECT(&mms->iotkit), "memory", OBJECT(system_memory), &error_abort); qdev_prop_set_uint32(iotkitdev, "EXP_NUMIRQ", MPS2TZ_NUMIRQ); - qdev_prop_set_uint32(iotkitdev, "MAINCLK_FRQ", SYSCLK_FRQ); qdev_connect_clock_in(iotkitdev, "MAINCLK", mms->sysclk); qdev_connect_clock_in(iotkitdev, "S32KCLK", mms->s32kclk); sysbus_realize(SYS_BUS_DEVICE(&mms->iotkit), &error_fatal); diff --git a/hw/arm/mps2.c b/hw/arm/mps2.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/mps2.c +++ b/hw/arm/mps2.c @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) object_initialize_child(OBJECT(mms), name, &mms->timer[i], TYPE_CMSDK_APB_TIMER); sbd = SYS_BUS_DEVICE(&mms->timer[i]); - qdev_prop_set_uint32(DEVICE(&mms->timer[i]), "pclk-frq", SYSCLK_FRQ); qdev_connect_clock_in(DEVICE(&mms->timer[i]), "pclk", mms->sysclk); sysbus_realize_and_unref(sbd, &error_fatal); sysbus_mmio_map(sbd, 0, base); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) object_initialize_child(OBJECT(mms), "dualtimer", &mms->dualtimer, TYPE_CMSDK_APB_DUALTIMER); - qdev_prop_set_uint32(DEVICE(&mms->dualtimer), "pclk-frq", SYSCLK_FRQ); qdev_connect_clock_in(DEVICE(&mms->dualtimer), "TIMCLK", mms->sysclk); sysbus_realize(SYS_BUS_DEVICE(&mms->dualtimer), &error_fatal); sysbus_connect_irq(SYS_BUS_DEVICE(&mms->dualtimer), 0, @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) sysbus_mmio_map(SYS_BUS_DEVICE(&mms->dualtimer), 0, 0x40002000); object_initialize_child(OBJECT(mms), "watchdog", &mms->watchdog, TYPE_CMSDK_APB_WATCHDOG); - qdev_prop_set_uint32(DEVICE(&mms->watchdog), "wdogclk-frq", SYSCLK_FRQ); qdev_connect_clock_in(DEVICE(&mms->watchdog), "WDOGCLK", mms->sysclk); sysbus_realize(SYS_BUS_DEVICE(&mms->watchdog), &error_fatal); sysbus_connect_irq(SYS_BUS_DEVICE(&mms->watchdog), 0, diff --git a/hw/arm/musca.c b/hw/arm/musca.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/musca.c +++ b/hw/arm/musca.c @@ -XXX,XX +XXX,XX @@ static void musca_init(MachineState *machine) qdev_prop_set_uint32(ssedev, "EXP_NUMIRQ", mmc->num_irqs); qdev_prop_set_uint32(ssedev, "init-svtor", mmc->init_svtor); qdev_prop_set_uint32(ssedev, "SRAM_ADDR_WIDTH", mmc->sram_addr_width); - qdev_prop_set_uint32(ssedev, "MAINCLK_FRQ", SYSCLK_FRQ); qdev_connect_clock_in(ssedev, "MAINCLK", mms->sysclk); qdev_connect_clock_in(ssedev, "S32KCLK", mms->s32kclk); /* diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ static void stellaris_init(MachineState *ms, stellaris_board_info *board) if (board->dc1 & (1 << 3)) { /* watchdog present */ dev = qdev_new(TYPE_LUMINARY_WATCHDOG); - /* system_clock_scale is valid now */ - uint32_t mainclk = NANOSECONDS_PER_SECOND / system_clock_scale; - qdev_prop_set_uint32(dev, "wdogclk-frq", mainclk); qdev_connect_clock_in(dev, "WDOGCLK", qdev_get_clock_out(ssys_dev, "SYSCLK")); -- 2.20.1 Now no users are setting the frq properties on the CMSDK timer, dualtimer, watchdog or ARMSSE SoC devices, we can remove the properties and the struct fields that back them. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-25-peter.maydell@linaro.org Message-id: 20210121190622.22000-25-peter.maydell@linaro.org --- include/hw/arm/armsse.h | 2 -- include/hw/timer/cmsdk-apb-dualtimer.h | 2 -- include/hw/timer/cmsdk-apb-timer.h | 2 -- include/hw/watchdog/cmsdk-apb-watchdog.h | 2 -- hw/arm/armsse.c | 2 -- hw/timer/cmsdk-apb-dualtimer.c | 6 ------ hw/timer/cmsdk-apb-timer.c | 6 ------ hw/watchdog/cmsdk-apb-watchdog.c | 6 ------ 8 files changed, 28 deletions(-) diff --git a/include/hw/arm/armsse.h b/include/hw/arm/armsse.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/arm/armsse.h +++ b/include/hw/arm/armsse.h @@ -XXX,XX +XXX,XX @@ * + Clock input "S32KCLK": slow 32KHz clock used for a few peripherals * + QOM property "memory" is a MemoryRegion containing the devices provided * by the board model. - * + QOM property "MAINCLK_FRQ" is the frequency of the main system clock * + QOM property "EXP_NUMIRQ" sets the number of expansion interrupts. * (In hardware, the SSE-200 permits the number of expansion interrupts * for the two CPUs to be configured separately, but we restrict it to @@ -XXX,XX +XXX,XX @@ struct ARMSSE { /* Properties */ MemoryRegion *board_memory; uint32_t exp_numirq; - uint32_t mainclk_frq; uint32_t sram_addr_width; uint32_t init_svtor; bool cpu_fpu[SSE_MAX_CPUS]; diff --git a/include/hw/timer/cmsdk-apb-dualtimer.h b/include/hw/timer/cmsdk-apb-dualtimer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-dualtimer.h +++ b/include/hw/timer/cmsdk-apb-dualtimer.h @@ -XXX,XX +XXX,XX @@ * https://developer.arm.com/products/system-design/system-design-kits/cortex-m-system-design-kit * * QEMU interface: - * + QOM property "pclk-frq": frequency at which the timer is clocked * + Clock input "TIMCLK": clock (for both timers) * + sysbus MMIO region 0: the register bank * + sysbus IRQ 0: combined timer interrupt TIMINTC @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBDualTimer { /*< public >*/ MemoryRegion iomem; qemu_irq timerintc; - uint32_t pclk_frq; Clock *timclk; CMSDKAPBDualTimerModule timermod[CMSDK_APB_DUALTIMER_NUM_MODULES]; diff --git a/include/hw/timer/cmsdk-apb-timer.h b/include/hw/timer/cmsdk-apb-timer.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/cmsdk-apb-timer.h +++ b/include/hw/timer/cmsdk-apb-timer.h @@ -XXX,XX +XXX,XX @@ OBJECT_DECLARE_SIMPLE_TYPE(CMSDKAPBTimer, CMSDK_APB_TIMER) /* * QEMU interface: - * + QOM property "pclk-frq": frequency at which the timer is clocked * + Clock input "pclk": clock for the timer * + sysbus MMIO region 0: the register bank * + sysbus IRQ 0: timer interrupt TIMERINT @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBTimer { /*< public >*/ MemoryRegion iomem; qemu_irq timerint; - uint32_t pclk_frq; struct ptimer_state *timer; Clock *pclk; diff --git a/include/hw/watchdog/cmsdk-apb-watchdog.h b/include/hw/watchdog/cmsdk-apb-watchdog.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/watchdog/cmsdk-apb-watchdog.h +++ b/include/hw/watchdog/cmsdk-apb-watchdog.h @@ -XXX,XX +XXX,XX @@ * https://developer.arm.com/products/system-design/system-design-kits/cortex-m-system-design-kit * * QEMU interface: - * + QOM property "wdogclk-frq": frequency at which the watchdog is clocked * + Clock input "WDOGCLK": clock for the watchdog's timer * + sysbus MMIO region 0: the register bank * + sysbus IRQ 0: watchdog interrupt @@ -XXX,XX +XXX,XX @@ struct CMSDKAPBWatchdog { /*< public >*/ MemoryRegion iomem; qemu_irq wdogint; - uint32_t wdogclk_frq; bool is_luminary; struct ptimer_state *timer; Clock *wdogclk; diff --git a/hw/arm/armsse.c b/hw/arm/armsse.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/armsse.c +++ b/hw/arm/armsse.c @@ -XXX,XX +XXX,XX @@ static Property iotkit_properties[] = { DEFINE_PROP_LINK("memory", ARMSSE, board_memory, TYPE_MEMORY_REGION, MemoryRegion *), DEFINE_PROP_UINT32("EXP_NUMIRQ", ARMSSE, exp_numirq, 64), - DEFINE_PROP_UINT32("MAINCLK_FRQ", ARMSSE, mainclk_frq, 0), DEFINE_PROP_UINT32("SRAM_ADDR_WIDTH", ARMSSE, sram_addr_width, 15), DEFINE_PROP_UINT32("init-svtor", ARMSSE, init_svtor, 0x10000000), DEFINE_PROP_BOOL("CPU0_FPU", ARMSSE, cpu_fpu[0], true), @@ -XXX,XX +XXX,XX @@ static Property armsse_properties[] = { DEFINE_PROP_LINK("memory", ARMSSE, board_memory, TYPE_MEMORY_REGION, MemoryRegion *), DEFINE_PROP_UINT32("EXP_NUMIRQ", ARMSSE, exp_numirq, 64), - DEFINE_PROP_UINT32("MAINCLK_FRQ", ARMSSE, mainclk_frq, 0), DEFINE_PROP_UINT32("SRAM_ADDR_WIDTH", ARMSSE, sram_addr_width, 15), DEFINE_PROP_UINT32("init-svtor", ARMSSE, init_svtor, 0x10000000), DEFINE_PROP_BOOL("CPU0_FPU", ARMSSE, cpu_fpu[0], false), diff --git a/hw/timer/cmsdk-apb-dualtimer.c b/hw/timer/cmsdk-apb-dualtimer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-dualtimer.c +++ b/hw/timer/cmsdk-apb-dualtimer.c @@ -XXX,XX +XXX,XX @@ static const VMStateDescription cmsdk_apb_dualtimer_vmstate = { } }; -static Property cmsdk_apb_dualtimer_properties[] = { - DEFINE_PROP_UINT32("pclk-frq", CMSDKAPBDualTimer, pclk_frq, 0), - DEFINE_PROP_END_OF_LIST(), -}; - static void cmsdk_apb_dualtimer_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_dualtimer_class_init(ObjectClass *klass, void *data) dc->realize = cmsdk_apb_dualtimer_realize; dc->vmsd = &cmsdk_apb_dualtimer_vmstate; dc->reset = cmsdk_apb_dualtimer_reset; - device_class_set_props(dc, cmsdk_apb_dualtimer_properties); } static const TypeInfo cmsdk_apb_dualtimer_info = { diff --git a/hw/timer/cmsdk-apb-timer.c b/hw/timer/cmsdk-apb-timer.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/cmsdk-apb-timer.c +++ b/hw/timer/cmsdk-apb-timer.c @@ -XXX,XX +XXX,XX @@ static const VMStateDescription cmsdk_apb_timer_vmstate = { } }; -static Property cmsdk_apb_timer_properties[] = { - DEFINE_PROP_UINT32("pclk-frq", CMSDKAPBTimer, pclk_frq, 0), - DEFINE_PROP_END_OF_LIST(), -}; - static void cmsdk_apb_timer_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_timer_class_init(ObjectClass *klass, void *data) dc->realize = cmsdk_apb_timer_realize; dc->vmsd = &cmsdk_apb_timer_vmstate; dc->reset = cmsdk_apb_timer_reset; - device_class_set_props(dc, cmsdk_apb_timer_properties); } static const TypeInfo cmsdk_apb_timer_info = { diff --git a/hw/watchdog/cmsdk-apb-watchdog.c b/hw/watchdog/cmsdk-apb-watchdog.c index XXXXXXX..XXXXXXX 100644 --- a/hw/watchdog/cmsdk-apb-watchdog.c +++ b/hw/watchdog/cmsdk-apb-watchdog.c @@ -XXX,XX +XXX,XX @@ static const VMStateDescription cmsdk_apb_watchdog_vmstate = { } }; -static Property cmsdk_apb_watchdog_properties[] = { - DEFINE_PROP_UINT32("wdogclk-frq", CMSDKAPBWatchdog, wdogclk_frq, 0), - DEFINE_PROP_END_OF_LIST(), -}; - static void cmsdk_apb_watchdog_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); @@ -XXX,XX +XXX,XX @@ static void cmsdk_apb_watchdog_class_init(ObjectClass *klass, void *data) dc->realize = cmsdk_apb_watchdog_realize; dc->vmsd = &cmsdk_apb_watchdog_vmstate; dc->reset = cmsdk_apb_watchdog_reset; - device_class_set_props(dc, cmsdk_apb_watchdog_properties); } static const TypeInfo cmsdk_apb_watchdog_info = { -- 2.20.1 Now that the watchdog device uses its Clock input rather than being passed the value of system_clock_scale at creation time, we can remove the hack where we reset the STELLARIS_SYS at board creation time to force it to set system_clock_scale. Instead it will be reset at the usual point in startup and will inform the watchdog of the clock frequency at that point. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Luc Michel <luc@lmichel.fr> Tested-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20210128114145.20536-26-peter.maydell@linaro.org Message-id: 20210121190622.22000-26-peter.maydell@linaro.org Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> --- hw/arm/stellaris.c | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/hw/arm/stellaris.c b/hw/arm/stellaris.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/stellaris.c +++ b/hw/arm/stellaris.c @@ -XXX,XX +XXX,XX @@ static DeviceState *stellaris_sys_init(uint32_t base, qemu_irq irq, sysbus_mmio_map(sbd, 0, base); sysbus_connect_irq(sbd, 0, irq); - /* - * Normally we should not be resetting devices like this during - * board creation. For the moment we need to do so, because - * system_clock_scale will only get set when the STELLARIS_SYS - * device is reset, and we need its initial value to pass to - * the watchdog device. This hack can be removed once the - * watchdog has been converted to use a Clock input instead. - */ - device_cold_reset(dev); - return dev; } -- 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