target-arm queue for rc1 -- these are all bug fixes. thanks -- PMM The following changes since commit b9404bf592e7ba74180e1a54ed7a266ec6ee67f2: Merge remote-tracking branch 'remotes/dgilbert/tags/pull-hmp-20190715' into staging (2019-07-15 12:22:07 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20190715 for you to fetch changes up to 51c9122e92b776a3f16af0b9282f1dc5012e2a19: target/arm: NS BusFault on vector table fetch escalates to NS HardFault (2019-07-15 14:17:04 +0100) ---------------------------------------------------------------- target-arm queue: * report ARMv8-A FP support for AArch32 -cpu max * hw/ssi/xilinx_spips: Avoid AXI writes to the LQSPI linear memory * hw/ssi/xilinx_spips: Avoid out-of-bound access to lqspi_buf[] * hw/ssi/mss-spi: Avoid crash when reading empty RX FIFO * hw/display/xlnx_dp: Avoid crash when reading empty RX FIFO * hw/arm/virt: Fix non-secure flash mode * pl031: Correctly migrate state when using -rtc clock=host * fix regression that meant arm926 and arm1026 lost VFP double-precision support * v8M: NS BusFault on vector table fetch escalates to NS HardFault ---------------------------------------------------------------- Alex Bennée (1): target/arm: report ARMv8-A FP support for AArch32 -cpu max David Engraf (1): hw/arm/virt: Fix non-secure flash mode Peter Maydell (3): pl031: Correctly migrate state when using -rtc clock=host target/arm: Set VFP-related MVFR0 fields for arm926 and arm1026 target/arm: NS BusFault on vector table fetch escalates to NS HardFault Philippe Mathieu-Daudé (5): hw/ssi/xilinx_spips: Convert lqspi_read() to read_with_attrs hw/ssi/xilinx_spips: Avoid AXI writes to the LQSPI linear memory hw/ssi/xilinx_spips: Avoid out-of-bound access to lqspi_buf[] hw/ssi/mss-spi: Avoid crash when reading empty RX FIFO hw/display/xlnx_dp: Avoid crash when reading empty RX FIFO include/hw/timer/pl031.h | 2 ++ hw/arm/virt.c | 2 +- hw/core/machine.c | 1 + hw/display/xlnx_dp.c | 15 +++++--- hw/ssi/mss-spi.c | 8 ++++- hw/ssi/xilinx_spips.c | 43 +++++++++++++++------- hw/timer/pl031.c | 92 +++++++++++++++++++++++++++++++++++++++++++++--- target/arm/cpu.c | 16 +++++++++ target/arm/m_helper.c | 21 ++++++++--- 9 files changed, 174 insertions(+), 26 deletions(-) From: Alex Bennée <alex.bennee@linaro.org> When we converted to using feature bits in 602f6e42cfbf we missed out the fact (dp && arm_dc_feature(s, ARM_FEATURE_V8)) was supported for -cpu max configurations. This caused a regression in the GCC test suite. Fix this by setting the appropriate bits in mvfr1.FPHP to report ARMv8-A with FP support (but not ARMv8.2-FP16). Fixes: https://bugs.launchpad.net/qemu/+bug/1836078 Signed-off-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20190711103737.10017-1-alex.bennee@linaro.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/cpu.c | 4 ++++ 1 file changed, 4 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_max_initfn(Object *obj) t = FIELD_DP32(t, ID_ISAR6, SPECRES, 1); cpu->isar.id_isar6 = t; + t = cpu->isar.mvfr1; + t = FIELD_DP32(t, MVFR1, FPHP, 2); /* v8.0 FP support */ + cpu->isar.mvfr1 = t; + t = cpu->isar.mvfr2; t = FIELD_DP32(t, MVFR2, SIMDMISC, 3); /* SIMD MaxNum */ t = FIELD_DP32(t, MVFR2, FPMISC, 4); /* FP MaxNum */ -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> In the next commit we will implement the write_with_attrs() handler. To avoid using different APIs, convert the read() handler first. Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Tested-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/xilinx_spips.c | 23 +++++++++++------------ 1 file changed, 11 insertions(+), 12 deletions(-) diff --git a/hw/ssi/xilinx_spips.c b/hw/ssi/xilinx_spips.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/xilinx_spips.c +++ b/hw/ssi/xilinx_spips.c @@ -XXX,XX +XXX,XX @@ static void lqspi_load_cache(void *opaque, hwaddr addr) } } -static uint64_t -lqspi_read(void *opaque, hwaddr addr, unsigned int size) +static MemTxResult lqspi_read(void *opaque, hwaddr addr, uint64_t *value, + unsigned size, MemTxAttrs attrs) { - XilinxQSPIPS *q = opaque; - uint32_t ret; + XilinxQSPIPS *q = XILINX_QSPIPS(opaque); if (addr >= q->lqspi_cached_addr && addr <= q->lqspi_cached_addr + LQSPI_CACHE_SIZE - 4) { uint8_t *retp = &q->lqspi_buf[addr - q->lqspi_cached_addr]; - ret = cpu_to_le32(*(uint32_t *)retp); - DB_PRINT_L(1, "addr: %08x, data: %08x\n", (unsigned)addr, - (unsigned)ret); - return ret; - } else { - lqspi_load_cache(opaque, addr); - return lqspi_read(opaque, addr, size); + *value = cpu_to_le32(*(uint32_t *)retp); + DB_PRINT_L(1, "addr: %08" HWADDR_PRIx ", data: %08" PRIx64 "\n", + addr, *value); + return MEMTX_OK; } + + lqspi_load_cache(opaque, addr); + return lqspi_read(opaque, addr, value, size, attrs); } static const MemoryRegionOps lqspi_ops = { - .read = lqspi_read, + .read_with_attrs = lqspi_read, .endianness = DEVICE_NATIVE_ENDIAN, .valid = { .min_access_size = 1, -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Lei Sun found while auditing the code that a CPU write would trigger a NULL pointer dereference. >From UG1085 datasheet [*] AXI writes in this region are ignored and generates an AXI Slave Error (SLVERR). Fix by implementing the write_with_attrs() handler. Return MEMTX_ERROR when the region is accessed (this error maps to an AXI slave error). [*] https://www.xilinx.com/support/documentation/user_guides/ug1085-zynq-ultrascale-trm.pdf Reported-by: Lei Sun <slei.casper@gmail.com> Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Tested-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/xilinx_spips.c | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/hw/ssi/xilinx_spips.c b/hw/ssi/xilinx_spips.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/xilinx_spips.c +++ b/hw/ssi/xilinx_spips.c @@ -XXX,XX +XXX,XX @@ static MemTxResult lqspi_read(void *opaque, hwaddr addr, uint64_t *value, return lqspi_read(opaque, addr, value, size, attrs); } +static MemTxResult lqspi_write(void *opaque, hwaddr offset, uint64_t value, + unsigned size, MemTxAttrs attrs) +{ + /* + * From UG1085, Chapter 24 (Quad-SPI controllers): + * - Writes are ignored + * - AXI writes generate an external AXI slave error (SLVERR) + */ + qemu_log_mask(LOG_GUEST_ERROR, "%s Unexpected %u-bit access to 0x%" PRIx64 + " (value: 0x%" PRIx64 "\n", + __func__, size << 3, offset, value); + + return MEMTX_ERROR; +} + static const MemoryRegionOps lqspi_ops = { .read_with_attrs = lqspi_read, + .write_with_attrs = lqspi_write, .endianness = DEVICE_NATIVE_ENDIAN, .valid = { .min_access_size = 1, -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Both lqspi_read() and lqspi_load_cache() expect a 32-bit aligned address. >From UG1085 datasheet [*] chapter on 'Quad-SPI Controller': Transfer Size Limitations Because of the 32-bit wide TX, RX, and generic FIFO, all APB/AXI transfers must be an integer multiple of 4-bytes. Shorter transfers are not possible. Set MemoryRegionOps.impl values to force 32-bit accesses, this way we are sure we do not access the lqspi_buf[] array out of bound. [*] https://www.xilinx.com/support/documentation/user_guides/ug1085-zynq-ultrascale-trm.pdf Reviewed-by: Francisco Iglesias <frasse.iglesias@gmail.com> Tested-by: Francisco Iglesias <frasse.iglesias@gmail.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/xilinx_spips.c | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/hw/ssi/xilinx_spips.c b/hw/ssi/xilinx_spips.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/xilinx_spips.c +++ b/hw/ssi/xilinx_spips.c @@ -XXX,XX +XXX,XX @@ static const MemoryRegionOps lqspi_ops = { .read_with_attrs = lqspi_read, .write_with_attrs = lqspi_write, .endianness = DEVICE_NATIVE_ENDIAN, + .impl = { + .min_access_size = 4, + .max_access_size = 4, + }, .valid = { .min_access_size = 1, .max_access_size = 4 -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Reading the RX_DATA register when the RX_FIFO is empty triggers an abort. This can be easily reproduced: $ qemu-system-arm -M emcraft-sf2 -monitor stdio -S QEMU 4.0.50 monitor - type 'help' for more information (qemu) x 0x40001010 Aborted (core dumped) (gdb) bt #1 0x00007f035874f895 in abort () at /lib64/libc.so.6 #2 0x00005628686591ff in fifo8_pop (fifo=0x56286a9a4c68) at util/fifo8.c:66 #3 0x00005628683e0b8e in fifo32_pop (fifo=0x56286a9a4c68) at include/qemu/fifo32.h:137 #4 0x00005628683e0efb in spi_read (opaque=0x56286a9a4850, addr=4, size=4) at hw/ssi/mss-spi.c:168 #5 0x0000562867f96801 in memory_region_read_accessor (mr=0x56286a9a4b60, addr=16, value=0x7ffeecb0c5c8, size=4, shift=0, mask=4294967295, attrs=...) at memory.c:439 #6 0x0000562867f96cdb in access_with_adjusted_size (addr=16, value=0x7ffeecb0c5c8, size=4, access_size_min=1, access_size_max=4, access_fn=0x562867f967c3 <memory_region_read_accessor>, mr=0x56286a9a4b60, attrs=...) at memory.c:569 #7 0x0000562867f99940 in memory_region_dispatch_read1 (mr=0x56286a9a4b60, addr=16, pval=0x7ffeecb0c5c8, size=4, attrs=...) at memory.c:1420 #8 0x0000562867f99a08 in memory_region_dispatch_read (mr=0x56286a9a4b60, addr=16, pval=0x7ffeecb0c5c8, size=4, attrs=...) at memory.c:1447 #9 0x0000562867f38721 in flatview_read_continue (fv=0x56286aec6360, addr=1073745936, attrs=..., buf=0x7ffeecb0c7c0 "\340ǰ\354\376\177", len=4, addr1=16, l=4, mr=0x56286a9a4b60) at exec.c:3385 #10 0x0000562867f38874 in flatview_read (fv=0x56286aec6360, addr=1073745936, attrs=..., buf=0x7ffeecb0c7c0 "\340ǰ\354\376\177", len=4) at exec.c:3423 #11 0x0000562867f388ea in address_space_read_full (as=0x56286aa3e890, addr=1073745936, attrs=..., buf=0x7ffeecb0c7c0 "\340ǰ\354\376\177", len=4) at exec.c:3436 #12 0x0000562867f389c5 in address_space_rw (as=0x56286aa3e890, addr=1073745936, attrs=..., buf=0x7ffeecb0c7c0 "\340ǰ\354\376\177", len=4, is_write=false) at exec.c:3466 #13 0x0000562867f3bdd7 in cpu_memory_rw_debug (cpu=0x56286aa19d00, addr=1073745936, buf=0x7ffeecb0c7c0 "\340ǰ\354\376\177", len=4, is_write=0) at exec.c:3976 #14 0x000056286811ed51 in memory_dump (mon=0x56286a8c32d0, count=1, format=120, wsize=4, addr=1073745936, is_physical=0) at monitor/misc.c:730 #15 0x000056286811eff1 in hmp_memory_dump (mon=0x56286a8c32d0, qdict=0x56286b15c400) at monitor/misc.c:785 #16 0x00005628684740ee in handle_hmp_command (mon=0x56286a8c32d0, cmdline=0x56286a8caeb2 "0x40001010") at monitor/hmp.c:1082 From the datasheet "Actel SmartFusion Microcontroller Subsystem User's Guide" Rev.1, Table 13-3 "SPI Register Summary", this register has a reset value of 0. Check the FIFO is not empty before accessing it, else log an error message. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20190709113715.7761-3-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/ssi/mss-spi.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/hw/ssi/mss-spi.c b/hw/ssi/mss-spi.c index XXXXXXX..XXXXXXX 100644 --- a/hw/ssi/mss-spi.c +++ b/hw/ssi/mss-spi.c @@ -XXX,XX +XXX,XX @@ spi_read(void *opaque, hwaddr addr, unsigned int size) case R_SPI_RX: s->regs[R_SPI_STATUS] &= ~S_RXFIFOFUL; s->regs[R_SPI_STATUS] &= ~S_RXCHOVRF; - ret = fifo32_pop(&s->rx_fifo); + if (fifo32_is_empty(&s->rx_fifo)) { + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Reading empty RX_FIFO\n", + __func__); + } else { + ret = fifo32_pop(&s->rx_fifo); + } if (fifo32_is_empty(&s->rx_fifo)) { s->regs[R_SPI_STATUS] |= S_RXFIFOEMP; } -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> In the previous commit we fixed a crash when the guest read a register that pop from an empty FIFO. By auditing the repository, we found another similar use with an easy way to reproduce: $ qemu-system-aarch64 -M xlnx-zcu102 -monitor stdio -S QEMU 4.0.50 monitor - type 'help' for more information (qemu) xp/b 0xfd4a0134 Aborted (core dumped) (gdb) bt #0 0x00007f6936dea57f in raise () at /lib64/libc.so.6 #1 0x00007f6936dd4895 in abort () at /lib64/libc.so.6 #2 0x0000561ad32975ec in xlnx_dp_aux_pop_rx_fifo (s=0x7f692babee70) at hw/display/xlnx_dp.c:431 #3 0x0000561ad3297dc0 in xlnx_dp_read (opaque=0x7f692babee70, offset=77, size=4) at hw/display/xlnx_dp.c:667 #4 0x0000561ad321b896 in memory_region_read_accessor (mr=0x7f692babf620, addr=308, value=0x7ffe05c1db88, size=4, shift=0, mask=4294967295, attrs=...) at memory.c:439 #5 0x0000561ad321bd70 in access_with_adjusted_size (addr=308, value=0x7ffe05c1db88, size=1, access_size_min=4, access_size_max=4, access_fn=0x561ad321b858 <memory_region_read_accessor>, mr=0x7f692babf620, attrs=...) at memory.c:569 #6 0x0000561ad321e9d5 in memory_region_dispatch_read1 (mr=0x7f692babf620, addr=308, pval=0x7ffe05c1db88, size=1, attrs=...) at memory.c:1420 #7 0x0000561ad321ea9d in memory_region_dispatch_read (mr=0x7f692babf620, addr=308, pval=0x7ffe05c1db88, size=1, attrs=...) at memory.c:1447 #8 0x0000561ad31bd742 in flatview_read_continue (fv=0x561ad69c04f0, addr=4249485620, attrs=..., buf=0x7ffe05c1dcf0 "\020\335\301\005\376\177", len=1, addr1=308, l=1, mr=0x7f692babf620) at exec.c:3385 #9 0x0000561ad31bd895 in flatview_read (fv=0x561ad69c04f0, addr=4249485620, attrs=..., buf=0x7ffe05c1dcf0 "\020\335\301\005\376\177", len=1) at exec.c:3423 #10 0x0000561ad31bd90b in address_space_read_full (as=0x561ad5bb3020, addr=4249485620, attrs=..., buf=0x7ffe05c1dcf0 "\020\335\301\005\376\177", len=1) at exec.c:3436 #11 0x0000561ad33b1c42 in address_space_read (len=1, buf=0x7ffe05c1dcf0 "\020\335\301\005\376\177", attrs=..., addr=4249485620, as=0x561ad5bb3020) at include/exec/memory.h:2131 #12 0x0000561ad33b1c42 in memory_dump (mon=0x561ad59c4530, count=1, format=120, wsize=1, addr=4249485620, is_physical=1) at monitor/misc.c:723 #13 0x0000561ad33b1fc1 in hmp_physical_memory_dump (mon=0x561ad59c4530, qdict=0x561ad6c6fd00) at monitor/misc.c:795 #14 0x0000561ad37b4a9f in handle_hmp_command (mon=0x561ad59c4530, cmdline=0x561ad59d0f22 "/b 0x00000000fd4a0134") at monitor/hmp.c:1082 Fix by checking the FIFO is not empty before popping from it. The datasheet is not clear about the reset value of this register, we choose to return '0'. Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Message-id: 20190709113715.7761-4-philmd@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/display/xlnx_dp.c | 15 +++++++++++---- 1 file changed, 11 insertions(+), 4 deletions(-) diff --git a/hw/display/xlnx_dp.c b/hw/display/xlnx_dp.c index XXXXXXX..XXXXXXX 100644 --- a/hw/display/xlnx_dp.c +++ b/hw/display/xlnx_dp.c @@ -XXX,XX +XXX,XX @@ static uint8_t xlnx_dp_aux_pop_rx_fifo(XlnxDPState *s) uint8_t ret; if (fifo8_is_empty(&s->rx_fifo)) { - DPRINTF("rx_fifo underflow..\n"); - abort(); + qemu_log_mask(LOG_GUEST_ERROR, + "%s: Reading empty RX_FIFO\n", + __func__); + /* + * The datasheet is not clear about the reset value, it seems + * to be unspecified. We choose to return '0'. + */ + ret = 0; + } else { + ret = fifo8_pop(&s->rx_fifo); + DPRINTF("pop 0x%" PRIX8 " from rx_fifo.\n", ret); } - ret = fifo8_pop(&s->rx_fifo); - DPRINTF("pop 0x%" PRIX8 " from rx_fifo.\n", ret); return ret; } -- 2.20.1 From: David Engraf <david.engraf@sysgo.com> Using the whole 128 MiB flash in non-secure mode is not working because virt_flash_fdt() expects the same address for secure_sysmem and sysmem. This is not correctly handled by caller because it forwards NULL for secure_sysmem in non-secure flash mode. Fixed by using sysmem when secure_sysmem is NULL. Signed-off-by: David Engraf <david.engraf@sysgo.com> Message-id: 20190712075002.14326-1-david.engraf@sysgo.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/virt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) 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 machvirt_init(MachineState *machine) &machine->device_memory->mr); } - virt_flash_fdt(vms, sysmem, secure_sysmem); + virt_flash_fdt(vms, sysmem, secure_sysmem ?: sysmem); create_gic(vms, pic); -- 2.20.1 The PL031 RTC tracks the difference between the guest RTC and the host RTC using a tick_offset field. For migration, however, we currently always migrate the offset between the guest and the vm_clock, even if the RTC clock is not the same as the vm_clock; this was an attempt to retain migration backwards compatibility. Unfortunately this results in the RTC behaving oddly across a VM state save and restore -- since the VM clock stands still across save-then-restore, regardless of how much real world time has elapsed, the guest RTC ends up out of sync with the host RTC in the restored VM. Fix this by migrating the raw tick_offset. To retain migration compatibility as far as possible, we have a new property migrate-tick-offset; by default this is 'true' and we will migrate the true tick offset in a new subsection; if the incoming data has no subsection we fall back to the old vm_clock-based offset information, so old->new migration compatibility is preserved. For complete new->old migration compatibility, the property is set to 'false' for 4.0 and earlier machine types (this will only affect 'virt-4.0' and below, as none of the other pl031-using machines are versioned). Reported-by: Russell King <rmk@armlinux.org.uk> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com> Message-id: 20190709143912.28905-1-peter.maydell@linaro.org --- include/hw/timer/pl031.h | 2 + hw/core/machine.c | 1 + hw/timer/pl031.c | 92 ++++++++++++++++++++++++++++++++++++++-- 3 files changed, 91 insertions(+), 4 deletions(-) diff --git a/include/hw/timer/pl031.h b/include/hw/timer/pl031.h index XXXXXXX..XXXXXXX 100644 --- a/include/hw/timer/pl031.h +++ b/include/hw/timer/pl031.h @@ -XXX,XX +XXX,XX @@ typedef struct PL031State { */ uint32_t tick_offset_vmstate; uint32_t tick_offset; + bool tick_offset_migrated; + bool migrate_tick_offset; uint32_t mr; uint32_t lr; diff --git a/hw/core/machine.c b/hw/core/machine.c index XXXXXXX..XXXXXXX 100644 --- a/hw/core/machine.c +++ b/hw/core/machine.c @@ -XXX,XX +XXX,XX @@ GlobalProperty hw_compat_4_0[] = { { "virtio-gpu-pci", "edid", "false" }, { "virtio-device", "use-started", "false" }, { "virtio-balloon-device", "qemu-4-0-config-size", "true" }, + { "pl031", "migrate-tick-offset", "false" }, }; const size_t hw_compat_4_0_len = G_N_ELEMENTS(hw_compat_4_0); diff --git a/hw/timer/pl031.c b/hw/timer/pl031.c index XXXXXXX..XXXXXXX 100644 --- a/hw/timer/pl031.c +++ b/hw/timer/pl031.c @@ -XXX,XX +XXX,XX @@ static int pl031_pre_save(void *opaque) { PL031State *s = opaque; - /* tick_offset is base_time - rtc_clock base time. Instead, we want to - * store the base time relative to the QEMU_CLOCK_VIRTUAL for backwards-compatibility. */ + /* + * The PL031 device model code uses the tick_offset field, which is + * the offset between what the guest RTC should read and what the + * QEMU rtc_clock reads: + * guest_rtc = rtc_clock + tick_offset + * and so + * tick_offset = guest_rtc - rtc_clock + * + * We want to migrate this offset, which sounds straightforward. + * Unfortunately older versions of QEMU migrated a conversion of this + * offset into an offset from the vm_clock. (This was in turn an + * attempt to be compatible with even older QEMU versions, but it + * has incorrect behaviour if the rtc_clock is not the same as the + * vm_clock.) So we put the actual tick_offset into a migration + * subsection, and the backwards-compatible time-relative-to-vm_clock + * in the main migration state. + * + * Calculate base time relative to QEMU_CLOCK_VIRTUAL: + */ int64_t delta = qemu_clock_get_ns(rtc_clock) - qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); s->tick_offset_vmstate = s->tick_offset + delta / NANOSECONDS_PER_SECOND; return 0; } +static int pl031_pre_load(void *opaque) +{ + PL031State *s = opaque; + + s->tick_offset_migrated = false; + return 0; +} + static int pl031_post_load(void *opaque, int version_id) { PL031State *s = opaque; - int64_t delta = qemu_clock_get_ns(rtc_clock) - qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); - s->tick_offset = s->tick_offset_vmstate - delta / NANOSECONDS_PER_SECOND; + /* + * If we got the tick_offset subsection, then we can just use + * the value in that. Otherwise the source is an older QEMU and + * has given us the offset from the vm_clock; convert it back to + * an offset from the rtc_clock. This will cause time to incorrectly + * go backwards compared to the host RTC, but this is unavoidable. + */ + + if (!s->tick_offset_migrated) { + int64_t delta = qemu_clock_get_ns(rtc_clock) - + qemu_clock_get_ns(QEMU_CLOCK_VIRTUAL); + s->tick_offset = s->tick_offset_vmstate - + delta / NANOSECONDS_PER_SECOND; + } pl031_set_alarm(s); return 0; } +static int pl031_tick_offset_post_load(void *opaque, int version_id) +{ + PL031State *s = opaque; + + s->tick_offset_migrated = true; + return 0; +} + +static bool pl031_tick_offset_needed(void *opaque) +{ + PL031State *s = opaque; + + return s->migrate_tick_offset; +} + +static const VMStateDescription vmstate_pl031_tick_offset = { + .name = "pl031/tick-offset", + .version_id = 1, + .minimum_version_id = 1, + .needed = pl031_tick_offset_needed, + .post_load = pl031_tick_offset_post_load, + .fields = (VMStateField[]) { + VMSTATE_UINT32(tick_offset, PL031State), + VMSTATE_END_OF_LIST() + } +}; + static const VMStateDescription vmstate_pl031 = { .name = "pl031", .version_id = 1, .minimum_version_id = 1, .pre_save = pl031_pre_save, + .pre_load = pl031_pre_load, .post_load = pl031_post_load, .fields = (VMStateField[]) { VMSTATE_UINT32(tick_offset_vmstate, PL031State), @@ -XXX,XX +XXX,XX @@ static const VMStateDescription vmstate_pl031 = { VMSTATE_UINT32(im, PL031State), VMSTATE_UINT32(is, PL031State), VMSTATE_END_OF_LIST() + }, + .subsections = (const VMStateDescription*[]) { + &vmstate_pl031_tick_offset, + NULL } }; +static Property pl031_properties[] = { + /* + * True to correctly migrate the tick offset of the RTC. False to + * obtain backward migration compatibility with older QEMU versions, + * at the expense of the guest RTC going backwards compared with the + * host RTC when the VM is saved/restored if using -rtc host. + * (Even if set to 'true' older QEMU can migrate forward to newer QEMU; + * 'false' also permits newer QEMU to migrate to older QEMU.) + */ + DEFINE_PROP_BOOL("migrate-tick-offset", + PL031State, migrate_tick_offset, true), + DEFINE_PROP_END_OF_LIST() +}; + static void pl031_class_init(ObjectClass *klass, void *data) { DeviceClass *dc = DEVICE_CLASS(klass); dc->vmsd = &vmstate_pl031; + dc->props = pl031_properties; } static const TypeInfo pl031_info = { -- 2.20.1 The ARMv5 architecture didn't specify detailed per-feature ID registers. Now that we're using the MVFR0 register fields to gate the existence of VFP instructions, we need to set up the correct values in the cpu->isar structure so that we still provide an FPU to the guest. This fixes a regression in the arm926 and arm1026 CPUs, which are the only ones that both have VFP and are ARMv5 or earlier. This regression was introduced by the VFP refactoring, and more specifically by commits 1120827fa182f0e76 and 266bd25c485597c, which accidentally disabled VFP short-vector support and double-precision support on these CPUs. Fixes: 1120827fa182f0e Fixes: 266bd25c485597c Fixes: https://bugs.launchpad.net/qemu/+bug/1836192 Reported-by: Christophe Lyon <christophe.lyon@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Tested-by: Christophe Lyon <christophe.lyon@linaro.org> Message-id: 20190711131241.22231-1-peter.maydell@linaro.org --- target/arm/cpu.c | 12 ++++++++++++ 1 file changed, 12 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 arm926_initfn(Object *obj) * set the field to indicate Jazelle support within QEMU. */ cpu->isar.id_isar1 = FIELD_DP32(cpu->isar.id_isar1, ID_ISAR1, JAZELLE, 1); + /* + * Similarly, we need to set MVFR0 fields to enable double precision + * and short vector support even though ARMv5 doesn't have this register. + */ + cpu->isar.mvfr0 = FIELD_DP32(cpu->isar.mvfr0, MVFR0, FPSHVEC, 1); + cpu->isar.mvfr0 = FIELD_DP32(cpu->isar.mvfr0, MVFR0, FPDP, 1); } static void arm946_initfn(Object *obj) @@ -XXX,XX +XXX,XX @@ static void arm1026_initfn(Object *obj) * set the field to indicate Jazelle support within QEMU. */ cpu->isar.id_isar1 = FIELD_DP32(cpu->isar.id_isar1, ID_ISAR1, JAZELLE, 1); + /* + * Similarly, we need to set MVFR0 fields to enable double precision + * and short vector support even though ARMv5 doesn't have this register. + */ + cpu->isar.mvfr0 = FIELD_DP32(cpu->isar.mvfr0, MVFR0, FPSHVEC, 1); + cpu->isar.mvfr0 = FIELD_DP32(cpu->isar.mvfr0, MVFR0, FPDP, 1); { /* The 1026 had an IFAR at c6,c0,0,1 rather than the ARMv6 c6,c0,0,2 */ -- 2.20.1 In the M-profile architecture, when we do a vector table fetch and it fails, we need to report a HardFault. Whether this is a Secure HF or a NonSecure HF depends on several things. If AIRCR.BFHFNMINS is 0 then HF is always Secure, because there is no NonSecure HardFault. Otherwise, the answer depends on whether the 'underlying exception' (MemManage, BusFault, SecureFault) targets Secure or NonSecure. (In the pseudocode, this is handled in the Vector() function: the final exc.isSecure is calculated by looking at the exc.isSecure from the exception returned from the memory access, not the isSecure input argument.) We weren't doing this correctly, because we were looking at the target security domain of the exception we were trying to load the vector table entry for. This produces errors of two kinds: * a load from the NS vector table which hits the "NS access to S memory" SecureFault should end up as a Secure HardFault, but we were raising an NS HardFault * a load from the S vector table which causes a BusFault should raise an NS HardFault if BFHFNMINS == 1 (because in that case all BusFaults are NonSecure), but we were raising a Secure HardFault Correct the logic. We also fix a comment error where we claimed that we might be escalating MemManage to HardFault, and forgot about SecureFault. (Vector loads can never hit MPU access faults, because they're always aligned and always use the default address map.) Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Message-id: 20190705094823.28905-1-peter.maydell@linaro.org --- target/arm/m_helper.c | 21 +++++++++++++++++---- 1 file changed, 17 insertions(+), 4 deletions(-) diff --git a/target/arm/m_helper.c b/target/arm/m_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/m_helper.c +++ b/target/arm/m_helper.c @@ -XXX,XX +XXX,XX @@ static bool arm_v7m_load_vector(ARMCPU *cpu, int exc, bool targets_secure, if (sattrs.ns) { attrs.secure = false; } else if (!targets_secure) { - /* NS access to S memory */ + /* + * NS access to S memory: the underlying exception which we escalate + * to HardFault is SecureFault, which always targets Secure. + */ + exc_secure = true; goto load_fail; } } @@ -XXX,XX +XXX,XX @@ static bool arm_v7m_load_vector(ARMCPU *cpu, int exc, bool targets_secure, vector_entry = address_space_ldl(arm_addressspace(cs, attrs), addr, attrs, &result); if (result != MEMTX_OK) { + /* + * Underlying exception is BusFault: its target security state + * depends on BFHFNMINS. + */ + exc_secure = !(cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK); goto load_fail; } *pvec = vector_entry; @@ -XXX,XX +XXX,XX @@ load_fail: /* * All vector table fetch fails are reported as HardFault, with * HFSR.VECTTBL and .FORCED set. (FORCED is set because - * technically the underlying exception is a MemManage or BusFault + * technically the underlying exception is a SecureFault or BusFault * that is escalated to HardFault.) This is a terminal exception, * so we will either take the HardFault immediately or else enter * lockup (the latter case is handled in armv7m_nvic_set_pending_derived()). + * The HardFault is Secure if BFHFNMINS is 0 (meaning that all HFs are + * secure); otherwise it targets the same security state as the + * underlying exception. */ - exc_secure = targets_secure || - !(cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK); + if (!(cpu->env.v7m.aircr & R_V7M_AIRCR_BFHFNMINS_MASK)) { + exc_secure = true; + } env->v7m.hfsr |= R_V7M_HFSR_VECTTBL_MASK | R_V7M_HFSR_FORCED_MASK; armv7m_nvic_set_pending_derived(env->nvic, ARMV7M_EXCP_HARD, exc_secure); return false; -- 2.20.1 The following changes since commit 61fee7f45955cd0bf9b79be9fa9c7ebabb5e6a85: Merge remote-tracking branch 'remotes/philmd-gitlab/tags/acceptance-testing-20200622' into staging (2020-06-22 20:50:10 +0100) are available in the Git repository at: https://git.linaro.org/people/pmaydell/qemu-arm.git tags/pull-target-arm-20200623 for you to fetch changes up to 539533b85fbd269f777bed931de8ccae1dd837e9: arm/virt: Add memory hot remove support (2020-06-23 11:39:48 +0100) ---------------------------------------------------------------- target-arm queue: * util/oslib-posix : qemu_init_exec_dir implementation for Mac * target/arm: Last parts of neon decodetree conversion * hw/arm/virt: Add 5.0 HW compat props * hw/watchdog/cmsdk-apb-watchdog: Add trace event for lock status * mps2: Add CMSDK APB watchdog, FPGAIO block, S2I devices and I2C devices * mps2: Add some unimplemented-device stubs for audio and GPIO * mps2-tz: Use the ARM SBCon two-wire serial bus interface * target/arm: Check supported KVM features globally (not per vCPU) * tests/qtest/arm-cpu-features: Add feature setting tests * arm/virt: Add memory hot remove support ---------------------------------------------------------------- Andrew Jones (2): hw/arm/virt: Add 5.0 HW compat props tests/qtest/arm-cpu-features: Add feature setting tests David CARLIER (1): util/oslib-posix : qemu_init_exec_dir implementation for Mac Peter Maydell (23): target/arm: Convert Neon 2-reg-misc VREV64 to decodetree target/arm: Convert Neon 2-reg-misc pairwise ops to decodetree target/arm: Convert VZIP, VUZP to decodetree target/arm: Convert Neon narrowing moves to decodetree target/arm: Convert Neon 2-reg-misc VSHLL to decodetree target/arm: Convert Neon VCVT f16/f32 insns to decodetree target/arm: Convert vectorised 2-reg-misc Neon ops to decodetree target/arm: Convert Neon 2-reg-misc crypto operations to decodetree target/arm: Rename NeonGenOneOpFn to NeonGenOne64OpFn target/arm: Fix capitalization in NeonGenTwo{Single, Double}OPFn typedefs target/arm: Make gen_swap_half() take separate src and dest target/arm: Convert Neon 2-reg-misc VREV32 and VREV16 to decodetree target/arm: Convert remaining simple 2-reg-misc Neon ops target/arm: Convert Neon VQABS, VQNEG to decodetree target/arm: Convert simple fp Neon 2-reg-misc insns target/arm: Convert Neon 2-reg-misc fp-compare-with-zero insns to decodetree target/arm: Convert Neon 2-reg-misc VRINT insns to decodetree target/arm: Convert Neon 2-reg-misc VCVT insns to decodetree target/arm: Convert Neon VSWP to decodetree target/arm: Convert Neon VTRN to decodetree target/arm: Move some functions used only in translate-neon.inc.c to that file target/arm: Remove unnecessary gen_io_end() calls target/arm: Remove dead code relating to SABA and UABA Philippe Mathieu-Daudé (15): hw/watchdog/cmsdk-apb-watchdog: Add trace event for lock status hw/i2c/versatile_i2c: Add definitions for register addresses hw/i2c/versatile_i2c: Add SCL/SDA definitions hw/i2c: Add header for ARM SBCon two-wire serial bus interface hw/arm: Use TYPE_VERSATILE_I2C instead of hardcoded string hw/arm/mps2: Document CMSDK/FPGA APB subsystem sections hw/arm/mps2: Rename CMSDK AHB peripheral region hw/arm/mps2: Add CMSDK APB watchdog device hw/arm/mps2: Add CMSDK AHB GPIO peripherals as unimplemented devices hw/arm/mps2: Map the FPGA I/O block hw/arm/mps2: Add SPI devices hw/arm/mps2: Add I2C devices hw/arm/mps2: Add audio I2S interface as unimplemented device hw/arm/mps2-tz: Use the ARM SBCon two-wire serial bus interface target/arm: Check supported KVM features globally (not per vCPU) Shameer Kolothum (1): arm/virt: Add memory hot remove support include/hw/i2c/arm_sbcon_i2c.h | 35 ++ target/arm/cpu.h | 2 +- target/arm/kvm_arm.h | 21 +- target/arm/translate.h | 8 +- target/arm/neon-dp.decode | 106 ++++ hw/acpi/generic_event_device.c | 29 + hw/arm/mps2-tz.c | 23 +- hw/arm/mps2.c | 65 ++- hw/arm/realview.c | 3 +- hw/arm/versatilepb.c | 3 +- hw/arm/vexpress.c | 3 +- hw/arm/virt.c | 63 +- hw/i2c/versatile_i2c.c | 38 +- hw/watchdog/cmsdk-apb-watchdog.c | 1 + target/arm/cpu.c | 2 +- target/arm/cpu64.c | 10 +- target/arm/kvm.c | 4 +- target/arm/kvm64.c | 14 +- target/arm/translate-a64.c | 20 +- target/arm/translate-neon.inc.c | 1191 +++++++++++++++++++++++++++++++++++++- target/arm/translate-vfp.inc.c | 7 +- target/arm/translate.c | 1064 +--------------------------------- tests/qtest/arm-cpu-features.c | 38 +- util/oslib-posix.c | 15 + MAINTAINERS | 1 + hw/arm/Kconfig | 8 +- hw/watchdog/trace-events | 1 + 27 files changed, 1624 insertions(+), 1151 deletions(-) create mode 100644 include/hw/i2c/arm_sbcon_i2c.h From: Andrew Jones <drjones@redhat.com> Cc: Cornelia Huck <cohuck@redhat.com> Signed-off-by: Andrew Jones <drjones@redhat.com> Reviewed-by: Cornelia Huck <cohuck@redhat.com> Message-id: 20200616140803.25515-1-drjones@redhat.com Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/virt.c | 1 + 1 file changed, 1 insertion(+) diff --git a/hw/arm/virt.c b/hw/arm/virt.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/virt.c +++ b/hw/arm/virt.c @@ -XXX,XX +XXX,XX @@ DEFINE_VIRT_MACHINE_AS_LATEST(5, 1) static void virt_machine_5_0_options(MachineClass *mc) { virt_machine_5_1_options(mc); + compat_props_add(mc->compat_props, hw_compat_5_0, hw_compat_5_0_len); } DEFINE_VIRT_MACHINE(5, 0) -- 2.20.1 From: David CARLIER <devnexen@gmail.com> From 3025a0ce3fdf7d3559fc35a52c659f635f5c750c Mon Sep 17 00:00:00 2001 From: David Carlier <devnexen@gmail.com> Date: Tue, 26 May 2020 21:35:27 +0100 Subject: [PATCH] util/oslib-posix : qemu_init_exec_dir implementation for Mac Using dyld API to get the full path of the current process. Signed-off-by: David Carlier <devnexen@gmail.com> Message-id: CA+XhMqxwC10XHVs4Z-JfE0-WLAU3ztDuU9QKVi31mjr59HWCxg@mail.gmail.com Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- util/oslib-posix.c | 15 +++++++++++++++ 1 file changed, 15 insertions(+) diff --git a/util/oslib-posix.c b/util/oslib-posix.c index XXXXXXX..XXXXXXX 100644 --- a/util/oslib-posix.c +++ b/util/oslib-posix.c @@ -XXX,XX +XXX,XX @@ #include <lwp.h> #endif +#ifdef __APPLE__ +#include <mach-o/dyld.h> +#endif + #include "qemu/mmap-alloc.h" #ifdef CONFIG_DEBUG_STACK_USAGE @@ -XXX,XX +XXX,XX @@ void qemu_init_exec_dir(const char *argv0) p = buf; } } +#elif defined(__APPLE__) + { + char fpath[PATH_MAX]; + uint32_t len = sizeof(fpath); + if (_NSGetExecutablePath(fpath, &len) == 0) { + p = realpath(fpath, buf); + if (!p) { + return; + } + } + } #endif /* If we don't have any way of figuring out the actual executable location then try argv[0]. */ -- 2.20.1 Convert the Neon VREV64 insn from the 2-reg-misc grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-2-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 12 ++++++++ target/arm/translate-neon.inc.c | 50 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 24 ++-------------- 3 files changed, 64 insertions(+), 22 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm vm=%vm_dp vd=%vd_dp size=1 VDUP_scalar 1111 001 1 1 . 11 index:1 100 .... 11 000 q:1 . 0 .... \ vm=%vm_dp vd=%vd_dp size=2 + + ################################################################## + # 2-reg-misc grouping: + # 1111 001 11 D 11 size:2 opc1:2 Vd:4 0 opc2:4 q:1 M 0 Vm:4 + ################################################################## + + &2misc vd vm q size + + @2misc .... ... .. . .. size:2 .. .... . .... q:1 . . .... \ + &2misc vm=%vm_dp vd=%vd_dp + + VREV64 1111 001 11 . 11 .. 00 .... 0 0000 . . 0 .... @2misc ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VDUP_scalar(DisasContext *s, arg_VDUP_scalar *a) a->q ? 16 : 8, a->q ? 16 : 8); return true; } + +static bool trans_VREV64(DisasContext *s, arg_VREV64 *a) +{ + int pass, half; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (a->size == 3) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + for (pass = 0; pass < (a->q ? 2 : 1); pass++) { + TCGv_i32 tmp[2]; + + for (half = 0; half < 2; half++) { + tmp[half] = neon_load_reg(a->vm, pass * 2 + half); + switch (a->size) { + case 0: + tcg_gen_bswap32_i32(tmp[half], tmp[half]); + break; + case 1: + gen_swap_half(tmp[half]); + break; + case 2: + break; + default: + g_assert_not_reached(); + } + } + neon_store_reg(a->vd, pass * 2, tmp[1]); + neon_store_reg(a->vd, pass * 2 + 1, tmp[0]); + } + return true; +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) } switch (op) { case NEON_2RM_VREV64: - for (pass = 0; pass < (q ? 2 : 1); pass++) { - tmp = neon_load_reg(rm, pass * 2); - tmp2 = neon_load_reg(rm, pass * 2 + 1); - switch (size) { - case 0: tcg_gen_bswap32_i32(tmp, tmp); break; - case 1: gen_swap_half(tmp); break; - case 2: /* no-op */ break; - default: abort(); - } - neon_store_reg(rd, pass * 2 + 1, tmp); - if (size == 2) { - neon_store_reg(rd, pass * 2, tmp2); - } else { - switch (size) { - case 0: tcg_gen_bswap32_i32(tmp2, tmp2); break; - case 1: gen_swap_half(tmp2); break; - default: abort(); - } - neon_store_reg(rd, pass * 2, tmp2); - } - } - break; + /* handled by decodetree */ + return 1; case NEON_2RM_VPADDL: case NEON_2RM_VPADDL_U: case NEON_2RM_VPADAL: case NEON_2RM_VPADAL_U: for (pass = 0; pass < q + 1; pass++) { -- 2.20.1 Convert the pairwise ops VPADDL and VPADAL in the 2-reg-misc grouping to decodetree. At this point we can get rid of the weird CPU_V001 #define that was used to avoid having to explicitly list all the arguments being passed to some TCG gen/helper functions. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-3-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 6 ++ target/arm/translate-neon.inc.c | 149 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 35 +------- 3 files changed, 157 insertions(+), 33 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm &2misc vm=%vm_dp vd=%vd_dp VREV64 1111 001 11 . 11 .. 00 .... 0 0000 . . 0 .... @2misc + + VPADDL_S 1111 001 11 . 11 .. 00 .... 0 0100 . . 0 .... @2misc + VPADDL_U 1111 001 11 . 11 .. 00 .... 0 0101 . . 0 .... @2misc + + VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc + VPADAL_U 1111 001 11 . 11 .. 00 .... 0 1101 . . 0 .... @2misc ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VREV64(DisasContext *s, arg_VREV64 *a) } return true; } + +static bool do_2misc_pairwise(DisasContext *s, arg_2misc *a, + NeonGenWidenFn *widenfn, + NeonGenTwo64OpFn *opfn, + NeonGenTwo64OpFn *accfn) +{ + /* + * Pairwise long operations: widen both halves of the pair, + * combine the pairs with the opfn, and then possibly accumulate + * into the destination with the accfn. + */ + int pass; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!widenfn) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + for (pass = 0; pass < a->q + 1; pass++) { + TCGv_i32 tmp; + TCGv_i64 rm0_64, rm1_64, rd_64; + + rm0_64 = tcg_temp_new_i64(); + rm1_64 = tcg_temp_new_i64(); + rd_64 = tcg_temp_new_i64(); + tmp = neon_load_reg(a->vm, pass * 2); + widenfn(rm0_64, tmp); + tcg_temp_free_i32(tmp); + tmp = neon_load_reg(a->vm, pass * 2 + 1); + widenfn(rm1_64, tmp); + tcg_temp_free_i32(tmp); + opfn(rd_64, rm0_64, rm1_64); + tcg_temp_free_i64(rm0_64); + tcg_temp_free_i64(rm1_64); + + if (accfn) { + TCGv_i64 tmp64 = tcg_temp_new_i64(); + neon_load_reg64(tmp64, a->vd + pass); + accfn(rd_64, tmp64, rd_64); + tcg_temp_free_i64(tmp64); + } + neon_store_reg64(rd_64, a->vd + pass); + tcg_temp_free_i64(rd_64); + } + return true; +} + +static bool trans_VPADDL_S(DisasContext *s, arg_2misc *a) +{ + static NeonGenWidenFn * const widenfn[] = { + gen_helper_neon_widen_s8, + gen_helper_neon_widen_s16, + tcg_gen_ext_i32_i64, + NULL, + }; + static NeonGenTwo64OpFn * const opfn[] = { + gen_helper_neon_paddl_u16, + gen_helper_neon_paddl_u32, + tcg_gen_add_i64, + NULL, + }; + + return do_2misc_pairwise(s, a, widenfn[a->size], opfn[a->size], NULL); +} + +static bool trans_VPADDL_U(DisasContext *s, arg_2misc *a) +{ + static NeonGenWidenFn * const widenfn[] = { + gen_helper_neon_widen_u8, + gen_helper_neon_widen_u16, + tcg_gen_extu_i32_i64, + NULL, + }; + static NeonGenTwo64OpFn * const opfn[] = { + gen_helper_neon_paddl_u16, + gen_helper_neon_paddl_u32, + tcg_gen_add_i64, + NULL, + }; + + return do_2misc_pairwise(s, a, widenfn[a->size], opfn[a->size], NULL); +} + +static bool trans_VPADAL_S(DisasContext *s, arg_2misc *a) +{ + static NeonGenWidenFn * const widenfn[] = { + gen_helper_neon_widen_s8, + gen_helper_neon_widen_s16, + tcg_gen_ext_i32_i64, + NULL, + }; + static NeonGenTwo64OpFn * const opfn[] = { + gen_helper_neon_paddl_u16, + gen_helper_neon_paddl_u32, + tcg_gen_add_i64, + NULL, + }; + static NeonGenTwo64OpFn * const accfn[] = { + gen_helper_neon_addl_u16, + gen_helper_neon_addl_u32, + tcg_gen_add_i64, + NULL, + }; + + return do_2misc_pairwise(s, a, widenfn[a->size], opfn[a->size], + accfn[a->size]); +} + +static bool trans_VPADAL_U(DisasContext *s, arg_2misc *a) +{ + static NeonGenWidenFn * const widenfn[] = { + gen_helper_neon_widen_u8, + gen_helper_neon_widen_u16, + tcg_gen_extu_i32_i64, + NULL, + }; + static NeonGenTwo64OpFn * const opfn[] = { + gen_helper_neon_paddl_u16, + gen_helper_neon_paddl_u32, + tcg_gen_add_i64, + NULL, + }; + static NeonGenTwo64OpFn * const accfn[] = { + gen_helper_neon_addl_u16, + gen_helper_neon_addl_u32, + tcg_gen_add_i64, + NULL, + }; + + return do_2misc_pairwise(s, a, widenfn[a->size], opfn[a->size], + accfn[a->size]); +} 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 gen_exception_return(DisasContext *s, TCGv_i32 pc) gen_rfe(s, pc, load_cpu_field(spsr)); } -#define CPU_V001 cpu_V0, cpu_V0, cpu_V1 - static int gen_neon_unzip(int rd, int rm, int size, int q) { TCGv_ptr pd, pm; @@ -XXX,XX +XXX,XX @@ static inline void gen_neon_widen(TCGv_i64 dest, TCGv_i32 src, int size, int u) tcg_temp_free_i32(src); } -static inline void gen_neon_addl(int size) -{ - switch (size) { - case 0: gen_helper_neon_addl_u16(CPU_V001); break; - case 1: gen_helper_neon_addl_u32(CPU_V001); break; - case 2: tcg_gen_add_i64(CPU_V001); break; - default: abort(); - } -} - static void gen_neon_narrow_op(int op, int u, int size, TCGv_i32 dest, TCGv_i64 src) { @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } switch (op) { case NEON_2RM_VREV64: - /* handled by decodetree */ - return 1; case NEON_2RM_VPADDL: case NEON_2RM_VPADDL_U: case NEON_2RM_VPADAL: case NEON_2RM_VPADAL_U: - for (pass = 0; pass < q + 1; pass++) { - tmp = neon_load_reg(rm, pass * 2); - gen_neon_widen(cpu_V0, tmp, size, op & 1); - tmp = neon_load_reg(rm, pass * 2 + 1); - gen_neon_widen(cpu_V1, tmp, size, op & 1); - switch (size) { - case 0: gen_helper_neon_paddl_u16(CPU_V001); break; - case 1: gen_helper_neon_paddl_u32(CPU_V001); break; - case 2: tcg_gen_add_i64(CPU_V001); break; - default: abort(); - } - if (op >= NEON_2RM_VPADAL) { - /* Accumulate. */ - neon_load_reg64(cpu_V1, rd + pass); - gen_neon_addl(size); - } - neon_store_reg64(cpu_V0, rd + pass); - } - break; + /* handled by decodetree */ + return 1; case NEON_2RM_VTRN: if (size == 2) { int n; -- 2.20.1 Convert the Neon VZIP and VUZP insns in the 2-reg-misc group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-4-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 3 ++ target/arm/translate-neon.inc.c | 74 ++++++++++++++++++++++++++ target/arm/translate.c | 92 +-------------------------------- 3 files changed, 79 insertions(+), 90 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc VPADAL_U 1111 001 11 . 11 .. 00 .... 0 1101 . . 0 .... @2misc + + VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc + VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VPADAL_U(DisasContext *s, arg_2misc *a) return do_2misc_pairwise(s, a, widenfn[a->size], opfn[a->size], accfn[a->size]); } + +typedef void ZipFn(TCGv_ptr, TCGv_ptr); + +static bool do_zip_uzp(DisasContext *s, arg_2misc *a, + ZipFn *fn) +{ + TCGv_ptr pd, pm; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!fn) { + /* Bad size or size/q combination */ + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + pd = vfp_reg_ptr(true, a->vd); + pm = vfp_reg_ptr(true, a->vm); + fn(pd, pm); + tcg_temp_free_ptr(pd); + tcg_temp_free_ptr(pm); + return true; +} + +static bool trans_VUZP(DisasContext *s, arg_2misc *a) +{ + static ZipFn * const fn[2][4] = { + { + gen_helper_neon_unzip8, + gen_helper_neon_unzip16, + NULL, + NULL, + }, { + gen_helper_neon_qunzip8, + gen_helper_neon_qunzip16, + gen_helper_neon_qunzip32, + NULL, + } + }; + return do_zip_uzp(s, a, fn[a->q][a->size]); +} + +static bool trans_VZIP(DisasContext *s, arg_2misc *a) +{ + static ZipFn * const fn[2][4] = { + { + gen_helper_neon_zip8, + gen_helper_neon_zip16, + NULL, + NULL, + }, { + gen_helper_neon_qzip8, + gen_helper_neon_qzip16, + gen_helper_neon_qzip32, + NULL, + } + }; + return do_zip_uzp(s, a, fn[a->q][a->size]); +} 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 gen_exception_return(DisasContext *s, TCGv_i32 pc) gen_rfe(s, pc, load_cpu_field(spsr)); } -static int gen_neon_unzip(int rd, int rm, int size, int q) -{ - TCGv_ptr pd, pm; - - if (!q && size == 2) { - return 1; - } - pd = vfp_reg_ptr(true, rd); - pm = vfp_reg_ptr(true, rm); - if (q) { - switch (size) { - case 0: - gen_helper_neon_qunzip8(pd, pm); - break; - case 1: - gen_helper_neon_qunzip16(pd, pm); - break; - case 2: - gen_helper_neon_qunzip32(pd, pm); - break; - default: - abort(); - } - } else { - switch (size) { - case 0: - gen_helper_neon_unzip8(pd, pm); - break; - case 1: - gen_helper_neon_unzip16(pd, pm); - break; - default: - abort(); - } - } - tcg_temp_free_ptr(pd); - tcg_temp_free_ptr(pm); - return 0; -} - -static int gen_neon_zip(int rd, int rm, int size, int q) -{ - TCGv_ptr pd, pm; - - if (!q && size == 2) { - return 1; - } - pd = vfp_reg_ptr(true, rd); - pm = vfp_reg_ptr(true, rm); - if (q) { - switch (size) { - case 0: - gen_helper_neon_qzip8(pd, pm); - break; - case 1: - gen_helper_neon_qzip16(pd, pm); - break; - case 2: - gen_helper_neon_qzip32(pd, pm); - break; - default: - abort(); - } - } else { - switch (size) { - case 0: - gen_helper_neon_zip8(pd, pm); - break; - case 1: - gen_helper_neon_zip16(pd, pm); - break; - default: - abort(); - } - } - tcg_temp_free_ptr(pd); - tcg_temp_free_ptr(pm); - return 0; -} - static void gen_neon_trn_u8(TCGv_i32 t0, TCGv_i32 t1) { TCGv_i32 rd, tmp; @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VREV64: case NEON_2RM_VPADDL: case NEON_2RM_VPADDL_U: case NEON_2RM_VPADAL: case NEON_2RM_VPADAL_U: + case NEON_2RM_VUZP: + case NEON_2RM_VZIP: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) goto elementwise; } break; - case NEON_2RM_VUZP: - if (gen_neon_unzip(rd, rm, size, q)) { - return 1; - } - break; - case NEON_2RM_VZIP: - if (gen_neon_zip(rd, rm, size, q)) { - return 1; - } - break; case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: /* also VQMOVUN; op field and mnemonics don't line up */ if (rm & 1) { -- 2.20.1 Convert the Neon narrowing moves VMQNV, VQMOVN, VQMOVUN in the 2-reg-misc group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-5-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 9 ++++ target/arm/translate-neon.inc.c | 59 ++++++++++++++++++++++++ target/arm/translate.c | 81 +-------------------------------- 3 files changed, 70 insertions(+), 79 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm @2misc .... ... .. . .. size:2 .. .... . .... q:1 . . .... \ &2misc vm=%vm_dp vd=%vd_dp + @2misc_q0 .... ... .. . .. size:2 .. .... . .... . . . .... \ + &2misc vm=%vm_dp vd=%vd_dp q=0 VREV64 1111 001 11 . 11 .. 00 .... 0 0000 . . 0 .... @2misc @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc + + VMOVN 1111 001 11 . 11 .. 10 .... 0 0100 0 . 0 .... @2misc_q0 + # VQMOVUN: unsigned result (source is always signed) + VQMOVUN 1111 001 11 . 11 .. 10 .... 0 0100 1 . 0 .... @2misc_q0 + # VQMOVN: signed result, source may be signed (_S) or unsigned (_U) + VQMOVN_S 1111 001 11 . 11 .. 10 .... 0 0101 0 . 0 .... @2misc_q0 + VQMOVN_U 1111 001 11 . 11 .. 10 .... 0 0101 1 . 0 .... @2misc_q0 ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VZIP(DisasContext *s, arg_2misc *a) }; return do_zip_uzp(s, a, fn[a->q][a->size]); } + +static bool do_vmovn(DisasContext *s, arg_2misc *a, + NeonGenNarrowEnvFn *narrowfn) +{ + TCGv_i64 rm; + TCGv_i32 rd0, rd1; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->vm & 1) { + return false; + } + + if (!narrowfn) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + rm = tcg_temp_new_i64(); + rd0 = tcg_temp_new_i32(); + rd1 = tcg_temp_new_i32(); + + neon_load_reg64(rm, a->vm); + narrowfn(rd0, cpu_env, rm); + neon_load_reg64(rm, a->vm + 1); + narrowfn(rd1, cpu_env, rm); + neon_store_reg(a->vd, 0, rd0); + neon_store_reg(a->vd, 1, rd1); + tcg_temp_free_i64(rm); + return true; +} + +#define DO_VMOVN(INSN, FUNC) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + static NeonGenNarrowEnvFn * const narrowfn[] = { \ + FUNC##8, \ + FUNC##16, \ + FUNC##32, \ + NULL, \ + }; \ + return do_vmovn(s, a, narrowfn[a->size]); \ + } + +DO_VMOVN(VMOVN, gen_neon_narrow_u) +DO_VMOVN(VQMOVUN, gen_helper_neon_unarrow_sat) +DO_VMOVN(VQMOVN_S, gen_helper_neon_narrow_sat_s) +DO_VMOVN(VQMOVN_U, gen_helper_neon_narrow_sat_u) 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 gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) tcg_temp_free_i32(rd); } -static inline void gen_neon_narrow(int size, TCGv_i32 dest, TCGv_i64 src) -{ - switch (size) { - case 0: gen_helper_neon_narrow_u8(dest, src); break; - case 1: gen_helper_neon_narrow_u16(dest, src); break; - case 2: tcg_gen_extrl_i64_i32(dest, src); break; - default: abort(); - } -} - -static inline void gen_neon_narrow_sats(int size, TCGv_i32 dest, TCGv_i64 src) -{ - switch (size) { - case 0: gen_helper_neon_narrow_sat_s8(dest, cpu_env, src); break; - case 1: gen_helper_neon_narrow_sat_s16(dest, cpu_env, src); break; - case 2: gen_helper_neon_narrow_sat_s32(dest, cpu_env, src); break; - default: abort(); - } -} - -static inline void gen_neon_narrow_satu(int size, TCGv_i32 dest, TCGv_i64 src) -{ - switch (size) { - case 0: gen_helper_neon_narrow_sat_u8(dest, cpu_env, src); break; - case 1: gen_helper_neon_narrow_sat_u16(dest, cpu_env, src); break; - case 2: gen_helper_neon_narrow_sat_u32(dest, cpu_env, src); break; - default: abort(); - } -} - -static inline void gen_neon_unarrow_sats(int size, TCGv_i32 dest, TCGv_i64 src) -{ - switch (size) { - case 0: gen_helper_neon_unarrow_sat8(dest, cpu_env, src); break; - case 1: gen_helper_neon_unarrow_sat16(dest, cpu_env, src); break; - case 2: gen_helper_neon_unarrow_sat32(dest, cpu_env, src); break; - default: abort(); - } -} - static inline void gen_neon_widen(TCGv_i64 dest, TCGv_i32 src, int size, int u) { if (u) { @@ -XXX,XX +XXX,XX @@ static inline void gen_neon_widen(TCGv_i64 dest, TCGv_i32 src, int size, int u) tcg_temp_free_i32(src); } -static void gen_neon_narrow_op(int op, int u, int size, - TCGv_i32 dest, TCGv_i64 src) -{ - if (op) { - if (u) { - gen_neon_unarrow_sats(size, dest, src); - } else { - gen_neon_narrow(size, dest, src); - } - } else { - if (u) { - gen_neon_narrow_satu(size, dest, src); - } else { - gen_neon_narrow_sats(size, dest, src); - } - } -} - /* Symbolic constants for op fields for Neon 2-register miscellaneous. * The values correspond to bits [17:16,10:7]; see the ARM ARM DDI0406B * table A7-13. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) !arm_dc_feature(s, ARM_FEATURE_V8)) { return 1; } - if ((op != NEON_2RM_VMOVN && op != NEON_2RM_VQMOVN) && - q && ((rm | rd) & 1)) { + if (q && ((rm | rd) & 1)) { return 1; } switch (op) { @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VPADAL: case NEON_2RM_VPADAL_U: case NEON_2RM_VUZP: case NEON_2RM_VZIP: + case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) goto elementwise; } break; - case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: - /* also VQMOVUN; op field and mnemonics don't line up */ - if (rm & 1) { - return 1; - } - tmp2 = NULL; - for (pass = 0; pass < 2; pass++) { - neon_load_reg64(cpu_V0, rm + pass); - tmp = tcg_temp_new_i32(); - gen_neon_narrow_op(op == NEON_2RM_VMOVN, q, size, - tmp, cpu_V0); - if (pass == 0) { - tmp2 = tmp; - } else { - neon_store_reg(rd, 0, tmp2); - neon_store_reg(rd, 1, tmp); - } - } - break; case NEON_2RM_VSHLL: if (q || (rd & 1)) { return 1; -- 2.20.1 Convert the VSHLL insn in the 2-reg-misc Neon group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-6-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 2 ++ target/arm/translate-neon.inc.c | 52 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 35 +--------------------- 3 files changed, 55 insertions(+), 34 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm # VQMOVN: signed result, source may be signed (_S) or unsigned (_U) VQMOVN_S 1111 001 11 . 11 .. 10 .... 0 0101 0 . 0 .... @2misc_q0 VQMOVN_U 1111 001 11 . 11 .. 10 .... 0 0101 1 . 0 .... @2misc_q0 + + VSHLL 1111 001 11 . 11 .. 10 .... 0 0110 0 . 0 .... @2misc_q0 ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_VMOVN(VMOVN, gen_neon_narrow_u) DO_VMOVN(VQMOVUN, gen_helper_neon_unarrow_sat) DO_VMOVN(VQMOVN_S, gen_helper_neon_narrow_sat_s) DO_VMOVN(VQMOVN_U, gen_helper_neon_narrow_sat_u) + +static bool trans_VSHLL(DisasContext *s, arg_2misc *a) +{ + TCGv_i32 rm0, rm1; + TCGv_i64 rd; + static NeonGenWidenFn * const widenfns[] = { + gen_helper_neon_widen_u8, + gen_helper_neon_widen_u16, + tcg_gen_extu_i32_i64, + NULL, + }; + NeonGenWidenFn *widenfn = widenfns[a->size]; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->vd & 1) { + return false; + } + + if (!widenfn) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + rd = tcg_temp_new_i64(); + + rm0 = neon_load_reg(a->vm, 0); + rm1 = neon_load_reg(a->vm, 1); + + widenfn(rd, rm0); + tcg_gen_shli_i64(rd, rd, 8 << a->size); + neon_store_reg64(rd, a->vd); + widenfn(rd, rm1); + tcg_gen_shli_i64(rd, rd, 8 << a->size); + neon_store_reg64(rd, a->vd + 1); + + tcg_temp_free_i64(rd); + tcg_temp_free_i32(rm0); + tcg_temp_free_i32(rm1); + return true; +} 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 gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) tcg_temp_free_i32(rd); } -static inline void gen_neon_widen(TCGv_i64 dest, TCGv_i32 src, int size, int u) -{ - if (u) { - switch (size) { - case 0: gen_helper_neon_widen_u8(dest, src); break; - case 1: gen_helper_neon_widen_u16(dest, src); break; - case 2: tcg_gen_extu_i32_i64(dest, src); break; - default: abort(); - } - } else { - switch (size) { - case 0: gen_helper_neon_widen_s8(dest, src); break; - case 1: gen_helper_neon_widen_s16(dest, src); break; - case 2: tcg_gen_ext_i32_i64(dest, src); break; - default: abort(); - } - } - tcg_temp_free_i32(src); -} - /* Symbolic constants for op fields for Neon 2-register miscellaneous. * The values correspond to bits [17:16,10:7]; see the ARM ARM DDI0406B * table A7-13. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VUZP: case NEON_2RM_VZIP: case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: + case NEON_2RM_VSHLL: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) goto elementwise; } break; - case NEON_2RM_VSHLL: - if (q || (rd & 1)) { - return 1; - } - tmp = neon_load_reg(rm, 0); - tmp2 = neon_load_reg(rm, 1); - for (pass = 0; pass < 2; pass++) { - if (pass == 1) - tmp = tmp2; - gen_neon_widen(cpu_V0, tmp, size, 1); - tcg_gen_shli_i64(cpu_V0, cpu_V0, 8 << size); - neon_store_reg64(cpu_V0, rd + pass); - } - break; case NEON_2RM_VCVT_F16_F32: { TCGv_ptr fpst; -- 2.20.1 Convert the Neon insns in the 2-reg-misc group which are VCVT between f32 and f16 to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-7-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 3 ++ target/arm/translate-neon.inc.c | 96 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 65 ++-------------------- 3 files changed, 102 insertions(+), 62 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VQMOVN_U 1111 001 11 . 11 .. 10 .... 0 0101 1 . 0 .... @2misc_q0 VSHLL 1111 001 11 . 11 .. 10 .... 0 0110 0 . 0 .... @2misc_q0 + + VCVT_F16_F32 1111 001 11 . 11 .. 10 .... 0 1100 0 . 0 .... @2misc_q0 + VCVT_F32_F16 1111 001 11 . 11 .. 10 .... 0 1110 0 . 0 .... @2misc_q0 ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VSHLL(DisasContext *s, arg_2misc *a) tcg_temp_free_i32(rm1); return true; } + +static bool trans_VCVT_F16_F32(DisasContext *s, arg_2misc *a) +{ + TCGv_ptr fpst; + TCGv_i32 ahp, tmp, tmp2, tmp3; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON) || + !dc_isar_feature(aa32_fp16_spconv, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vm & 1) || (a->size != 1)) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fpst = get_fpstatus_ptr(true); + ahp = get_ahp_flag(); + tmp = neon_load_reg(a->vm, 0); + gen_helper_vfp_fcvt_f32_to_f16(tmp, tmp, fpst, ahp); + tmp2 = neon_load_reg(a->vm, 1); + gen_helper_vfp_fcvt_f32_to_f16(tmp2, tmp2, fpst, ahp); + tcg_gen_shli_i32(tmp2, tmp2, 16); + tcg_gen_or_i32(tmp2, tmp2, tmp); + tcg_temp_free_i32(tmp); + tmp = neon_load_reg(a->vm, 2); + gen_helper_vfp_fcvt_f32_to_f16(tmp, tmp, fpst, ahp); + tmp3 = neon_load_reg(a->vm, 3); + neon_store_reg(a->vd, 0, tmp2); + gen_helper_vfp_fcvt_f32_to_f16(tmp3, tmp3, fpst, ahp); + tcg_gen_shli_i32(tmp3, tmp3, 16); + tcg_gen_or_i32(tmp3, tmp3, tmp); + neon_store_reg(a->vd, 1, tmp3); + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(ahp); + tcg_temp_free_ptr(fpst); + + return true; +} + +static bool trans_VCVT_F32_F16(DisasContext *s, arg_2misc *a) +{ + TCGv_ptr fpst; + TCGv_i32 ahp, tmp, tmp2, tmp3; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON) || + !dc_isar_feature(aa32_fp16_spconv, s)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vd & 1) || (a->size != 1)) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fpst = get_fpstatus_ptr(true); + ahp = get_ahp_flag(); + tmp3 = tcg_temp_new_i32(); + tmp = neon_load_reg(a->vm, 0); + tmp2 = neon_load_reg(a->vm, 1); + tcg_gen_ext16u_i32(tmp3, tmp); + gen_helper_vfp_fcvt_f16_to_f32(tmp3, tmp3, fpst, ahp); + neon_store_reg(a->vd, 0, tmp3); + tcg_gen_shri_i32(tmp, tmp, 16); + gen_helper_vfp_fcvt_f16_to_f32(tmp, tmp, fpst, ahp); + neon_store_reg(a->vd, 1, tmp); + tmp3 = tcg_temp_new_i32(); + tcg_gen_ext16u_i32(tmp3, tmp2); + gen_helper_vfp_fcvt_f16_to_f32(tmp3, tmp3, fpst, ahp); + neon_store_reg(a->vd, 2, tmp3); + tcg_gen_shri_i32(tmp2, tmp2, 16); + gen_helper_vfp_fcvt_f16_to_f32(tmp2, tmp2, fpst, ahp); + neon_store_reg(a->vd, 3, tmp2); + tcg_temp_free_i32(ahp); + tcg_temp_free_ptr(fpst); + + return true; +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) int pass; int u; int vec_size; - TCGv_i32 tmp, tmp2, tmp3; + TCGv_i32 tmp, tmp2; if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { return 1; @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VZIP: case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: case NEON_2RM_VSHLL: + case NEON_2RM_VCVT_F16_F32: + case NEON_2RM_VCVT_F32_F16: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) goto elementwise; } break; - case NEON_2RM_VCVT_F16_F32: - { - TCGv_ptr fpst; - TCGv_i32 ahp; - - if (!dc_isar_feature(aa32_fp16_spconv, s) || - q || (rm & 1)) { - return 1; - } - fpst = get_fpstatus_ptr(true); - ahp = get_ahp_flag(); - tmp = neon_load_reg(rm, 0); - gen_helper_vfp_fcvt_f32_to_f16(tmp, tmp, fpst, ahp); - tmp2 = neon_load_reg(rm, 1); - gen_helper_vfp_fcvt_f32_to_f16(tmp2, tmp2, fpst, ahp); - tcg_gen_shli_i32(tmp2, tmp2, 16); - tcg_gen_or_i32(tmp2, tmp2, tmp); - tcg_temp_free_i32(tmp); - tmp = neon_load_reg(rm, 2); - gen_helper_vfp_fcvt_f32_to_f16(tmp, tmp, fpst, ahp); - tmp3 = neon_load_reg(rm, 3); - neon_store_reg(rd, 0, tmp2); - gen_helper_vfp_fcvt_f32_to_f16(tmp3, tmp3, fpst, ahp); - tcg_gen_shli_i32(tmp3, tmp3, 16); - tcg_gen_or_i32(tmp3, tmp3, tmp); - neon_store_reg(rd, 1, tmp3); - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(ahp); - tcg_temp_free_ptr(fpst); - break; - } - case NEON_2RM_VCVT_F32_F16: - { - TCGv_ptr fpst; - TCGv_i32 ahp; - if (!dc_isar_feature(aa32_fp16_spconv, s) || - q || (rd & 1)) { - return 1; - } - fpst = get_fpstatus_ptr(true); - ahp = get_ahp_flag(); - tmp3 = tcg_temp_new_i32(); - tmp = neon_load_reg(rm, 0); - tmp2 = neon_load_reg(rm, 1); - tcg_gen_ext16u_i32(tmp3, tmp); - gen_helper_vfp_fcvt_f16_to_f32(tmp3, tmp3, fpst, ahp); - neon_store_reg(rd, 0, tmp3); - tcg_gen_shri_i32(tmp, tmp, 16); - gen_helper_vfp_fcvt_f16_to_f32(tmp, tmp, fpst, ahp); - neon_store_reg(rd, 1, tmp); - tmp3 = tcg_temp_new_i32(); - tcg_gen_ext16u_i32(tmp3, tmp2); - gen_helper_vfp_fcvt_f16_to_f32(tmp3, tmp3, fpst, ahp); - neon_store_reg(rd, 2, tmp3); - tcg_gen_shri_i32(tmp2, tmp2, 16); - gen_helper_vfp_fcvt_f16_to_f32(tmp2, tmp2, fpst, ahp); - neon_store_reg(rd, 3, tmp2); - tcg_temp_free_i32(ahp); - tcg_temp_free_ptr(fpst); - break; - } case NEON_2RM_AESE: case NEON_2RM_AESMC: if (!dc_isar_feature(aa32_aes, s) || ((rm | rd) & 1)) { return 1; -- 2.20.1 Convert to decodetree the insns in the Neon 2-reg-misc grouping which we implement using gvec. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-8-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 11 +++++++ target/arm/translate-neon.inc.c | 55 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 35 +++++---------------- 3 files changed, 74 insertions(+), 27 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VPADDL_S 1111 001 11 . 11 .. 00 .... 0 0100 . . 0 .... @2misc VPADDL_U 1111 001 11 . 11 .. 00 .... 0 0101 . . 0 .... @2misc + VMVN 1111 001 11 . 11 .. 00 .... 0 1011 . . 0 .... @2misc + VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc VPADAL_U 1111 001 11 . 11 .. 00 .... 0 1101 . . 0 .... @2misc + VCGT0 1111 001 11 . 11 .. 01 .... 0 0000 . . 0 .... @2misc + VCGE0 1111 001 11 . 11 .. 01 .... 0 0001 . . 0 .... @2misc + VCEQ0 1111 001 11 . 11 .. 01 .... 0 0010 . . 0 .... @2misc + VCLE0 1111 001 11 . 11 .. 01 .... 0 0011 . . 0 .... @2misc + VCLT0 1111 001 11 . 11 .. 01 .... 0 0100 . . 0 .... @2misc + + VABS 1111 001 11 . 11 .. 01 .... 0 0110 . . 0 .... @2misc + VNEG 1111 001 11 . 11 .. 01 .... 0 0111 . . 0 .... @2misc + VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VCVT_F32_F16(DisasContext *s, arg_2misc *a) return true; } + +static bool do_2misc_vec(DisasContext *s, arg_2misc *a, GVecGen2Fn *fn) +{ + int vec_size = a->q ? 16 : 8; + int rd_ofs = neon_reg_offset(a->vd, 0); + int rm_ofs = neon_reg_offset(a->vm, 0); + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->size == 3) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fn(a->size, rd_ofs, rm_ofs, vec_size, vec_size); + + return true; +} + +#define DO_2MISC_VEC(INSN, FN) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + return do_2misc_vec(s, a, FN); \ + } + +DO_2MISC_VEC(VNEG, tcg_gen_gvec_neg) +DO_2MISC_VEC(VABS, tcg_gen_gvec_abs) +DO_2MISC_VEC(VCEQ0, gen_gvec_ceq0) +DO_2MISC_VEC(VCGT0, gen_gvec_cgt0) +DO_2MISC_VEC(VCLE0, gen_gvec_cle0) +DO_2MISC_VEC(VCGE0, gen_gvec_cge0) +DO_2MISC_VEC(VCLT0, gen_gvec_clt0) + +static bool trans_VMVN(DisasContext *s, arg_2misc *a) +{ + if (a->size != 0) { + return false; + } + return do_2misc_vec(s, a, tcg_gen_gvec_not); +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) int size; int pass; int u; - int vec_size; TCGv_i32 tmp, tmp2; if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) VFP_DREG_D(rd, insn); VFP_DREG_M(rm, insn); size = (insn >> 20) & 3; - vec_size = q ? 16 : 8; rd_ofs = neon_reg_offset(rd, 0); rm_ofs = neon_reg_offset(rm, 0); @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VSHLL: case NEON_2RM_VCVT_F16_F32: case NEON_2RM_VCVT_F32_F16: + case NEON_2RM_VMVN: + case NEON_2RM_VNEG: + case NEON_2RM_VABS: + case NEON_2RM_VCEQ0: + case NEON_2RM_VCGT0: + case NEON_2RM_VCLE0: + case NEON_2RM_VCGE0: + case NEON_2RM_VCLT0: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) q ? gen_helper_crypto_sha256su0 : gen_helper_crypto_sha1su1); break; - case NEON_2RM_VMVN: - tcg_gen_gvec_not(0, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VNEG: - tcg_gen_gvec_neg(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VABS: - tcg_gen_gvec_abs(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - - case NEON_2RM_VCEQ0: - gen_gvec_ceq0(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VCGT0: - gen_gvec_cgt0(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VCLE0: - gen_gvec_cle0(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VCGE0: - gen_gvec_cge0(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; - case NEON_2RM_VCLT0: - gen_gvec_clt0(size, rd_ofs, rm_ofs, vec_size, vec_size); - break; default: elementwise: -- 2.20.1 Convert the Neon-2-reg misc crypto ops (AESE, AESMC, SHA1H, SHA1SU1) to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-9-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 12 ++++++++ target/arm/translate-neon.inc.c | 42 ++++++++++++++++++++++++++ target/arm/translate.c | 52 +++------------------------------ 3 files changed, 58 insertions(+), 48 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm &2misc vm=%vm_dp vd=%vd_dp @2misc_q0 .... ... .. . .. size:2 .. .... . .... . . . .... \ &2misc vm=%vm_dp vd=%vd_dp q=0 + @2misc_q1 .... ... .. . .. size:2 .. .... . .... . . . .... \ + &2misc vm=%vm_dp vd=%vd_dp q=1 VREV64 1111 001 11 . 11 .. 00 .... 0 0000 . . 0 .... @2misc VPADDL_S 1111 001 11 . 11 .. 00 .... 0 0100 . . 0 .... @2misc VPADDL_U 1111 001 11 . 11 .. 00 .... 0 0101 . . 0 .... @2misc + AESE 1111 001 11 . 11 .. 00 .... 0 0110 0 . 0 .... @2misc_q1 + AESD 1111 001 11 . 11 .. 00 .... 0 0110 1 . 0 .... @2misc_q1 + AESMC 1111 001 11 . 11 .. 00 .... 0 0111 0 . 0 .... @2misc_q1 + AESIMC 1111 001 11 . 11 .. 00 .... 0 0111 1 . 0 .... @2misc_q1 + VMVN 1111 001 11 . 11 .. 00 .... 0 1011 . . 0 .... @2misc VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VCLE0 1111 001 11 . 11 .. 01 .... 0 0011 . . 0 .... @2misc VCLT0 1111 001 11 . 11 .. 01 .... 0 0100 . . 0 .... @2misc + SHA1H 1111 001 11 . 11 .. 01 .... 0 0101 1 . 0 .... @2misc_q1 + VABS 1111 001 11 . 11 .. 01 .... 0 0110 . . 0 .... @2misc VNEG 1111 001 11 . 11 .. 01 .... 0 0111 . . 0 .... @2misc @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VSHLL 1111 001 11 . 11 .. 10 .... 0 0110 0 . 0 .... @2misc_q0 + SHA1SU1 1111 001 11 . 11 .. 10 .... 0 0111 0 . 0 .... @2misc_q1 + SHA256SU0 1111 001 11 . 11 .. 10 .... 0 0111 1 . 0 .... @2misc_q1 + VCVT_F16_F32 1111 001 11 . 11 .. 10 .... 0 1100 0 . 0 .... @2misc_q0 VCVT_F32_F16 1111 001 11 . 11 .. 10 .... 0 1110 0 . 0 .... @2misc_q0 ] diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VMVN(DisasContext *s, arg_2misc *a) } return do_2misc_vec(s, a, tcg_gen_gvec_not); } + +#define WRAP_2M_3_OOL_FN(WRAPNAME, FUNC, DATA) \ + static void WRAPNAME(unsigned vece, uint32_t rd_ofs, \ + uint32_t rm_ofs, uint32_t oprsz, \ + uint32_t maxsz) \ + { \ + tcg_gen_gvec_3_ool(rd_ofs, rd_ofs, rm_ofs, oprsz, maxsz, \ + DATA, FUNC); \ + } + +#define WRAP_2M_2_OOL_FN(WRAPNAME, FUNC, DATA) \ + static void WRAPNAME(unsigned vece, uint32_t rd_ofs, \ + uint32_t rm_ofs, uint32_t oprsz, \ + uint32_t maxsz) \ + { \ + tcg_gen_gvec_2_ool(rd_ofs, rm_ofs, oprsz, maxsz, DATA, FUNC); \ + } + +WRAP_2M_3_OOL_FN(gen_AESE, gen_helper_crypto_aese, 0) +WRAP_2M_3_OOL_FN(gen_AESD, gen_helper_crypto_aese, 1) +WRAP_2M_2_OOL_FN(gen_AESMC, gen_helper_crypto_aesmc, 0) +WRAP_2M_2_OOL_FN(gen_AESIMC, gen_helper_crypto_aesmc, 1) +WRAP_2M_2_OOL_FN(gen_SHA1H, gen_helper_crypto_sha1h, 0) +WRAP_2M_2_OOL_FN(gen_SHA1SU1, gen_helper_crypto_sha1su1, 0) +WRAP_2M_2_OOL_FN(gen_SHA256SU0, gen_helper_crypto_sha256su0, 0) + +#define DO_2M_CRYPTO(INSN, FEATURE, SIZE) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + if (!dc_isar_feature(FEATURE, s) || a->size != SIZE) { \ + return false; \ + } \ + return do_2misc_vec(s, a, gen_##INSN); \ + } + +DO_2M_CRYPTO(AESE, aa32_aes, 0) +DO_2M_CRYPTO(AESD, aa32_aes, 0) +DO_2M_CRYPTO(AESMC, aa32_aes, 0) +DO_2M_CRYPTO(AESIMC, aa32_aes, 0) +DO_2M_CRYPTO(SHA1H, aa32_sha1, 2) +DO_2M_CRYPTO(SHA1SU1, aa32_sha1, 2) +DO_2M_CRYPTO(SHA256SU0, aa32_sha2, 2) 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) { int op; int q; - int rd, rm, rd_ofs, rm_ofs; + int rd, rm; int size; int pass; int u; @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) VFP_DREG_D(rd, insn); VFP_DREG_M(rm, insn); size = (insn >> 20) & 3; - rd_ofs = neon_reg_offset(rd, 0); - rm_ofs = neon_reg_offset(rm, 0); if ((insn & (1 << 23)) == 0) { /* Three register same length: handled by decodetree */ @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VCLE0: case NEON_2RM_VCGE0: case NEON_2RM_VCLT0: + case NEON_2RM_AESE: case NEON_2RM_AESMC: + case NEON_2RM_SHA1H: + case NEON_2RM_SHA1SU1: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) goto elementwise; } break; - case NEON_2RM_AESE: case NEON_2RM_AESMC: - if (!dc_isar_feature(aa32_aes, s) || ((rm | rd) & 1)) { - return 1; - } - /* - * Bit 6 is the lowest opcode bit; it distinguishes - * between encryption (AESE/AESMC) and decryption - * (AESD/AESIMC). - */ - if (op == NEON_2RM_AESE) { - tcg_gen_gvec_3_ool(vfp_reg_offset(true, rd), - vfp_reg_offset(true, rd), - vfp_reg_offset(true, rm), - 16, 16, extract32(insn, 6, 1), - gen_helper_crypto_aese); - } else { - tcg_gen_gvec_2_ool(vfp_reg_offset(true, rd), - vfp_reg_offset(true, rm), - 16, 16, extract32(insn, 6, 1), - gen_helper_crypto_aesmc); - } - break; - case NEON_2RM_SHA1H: - if (!dc_isar_feature(aa32_sha1, s) || ((rm | rd) & 1)) { - return 1; - } - tcg_gen_gvec_2_ool(rd_ofs, rm_ofs, 16, 16, 0, - gen_helper_crypto_sha1h); - break; - case NEON_2RM_SHA1SU1: - if ((rm | rd) & 1) { - return 1; - } - /* bit 6 (q): set -> SHA256SU0, cleared -> SHA1SU1 */ - if (q) { - if (!dc_isar_feature(aa32_sha2, s)) { - return 1; - } - } else if (!dc_isar_feature(aa32_sha1, s)) { - return 1; - } - tcg_gen_gvec_2_ool(rd_ofs, rm_ofs, 16, 16, 0, - q ? gen_helper_crypto_sha256su0 - : gen_helper_crypto_sha1su1); - break; default: elementwise: -- 2.20.1 The NeonGenOneOpFn typedef breaks with the pattern of the other NeonGen*Fn typedefs, because it is a TCGv_i64 -> TCGv_i64 operation but it does not have '64' in its name. Rename it to NeonGenOne64OpFn, so that the old name is available for a TCGv_i32 -> TCGv_i32 operation (which we will need in a subsequent commit). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-10-peter.maydell@linaro.org --- target/arm/translate.h | 2 +- target/arm/translate-a64.c | 4 ++-- 2 files changed, 3 insertions(+), 3 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 @@ typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); typedef void NeonGenTwoOpWidenFn(TCGv_i64, TCGv_i32, TCGv_i32); typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); -typedef void NeonGenOneOpFn(TCGv_i64, TCGv_i64); +typedef void NeonGenOne64OpFn(TCGv_i64, TCGv_i64); typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); typedef void CryptoThreeOpFn(TCGv_ptr, TCGv_ptr, TCGv_ptr); 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 handle_2misc_pairwise(DisasContext *s, int opcode, bool u, } else { for (pass = 0; pass < maxpass; pass++) { TCGv_i64 tcg_op = tcg_temp_new_i64(); - NeonGenOneOpFn *genfn; - static NeonGenOneOpFn * const fns[2][2] = { + NeonGenOne64OpFn *genfn; + static NeonGenOne64OpFn * const fns[2][2] = { { gen_helper_neon_addlp_s8, gen_helper_neon_addlp_u8 }, { gen_helper_neon_addlp_s16, gen_helper_neon_addlp_u16 }, }; -- 2.20.1 All the other typedefs like these spell "Op" with a lowercase 'p'; remane the NeonGenTwoSingleOPFn and NeonGenTwoDoubleOPFn typedefs to match. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-11-peter.maydell@linaro.org --- target/arm/translate.h | 4 ++-- target/arm/translate-a64.c | 4 ++-- target/arm/translate-neon.inc.c | 2 +- 3 files changed, 5 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 @@ typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); typedef void NeonGenTwoOpWidenFn(TCGv_i64, TCGv_i32, TCGv_i32); -typedef void NeonGenTwoSingleOPFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); -typedef void NeonGenTwoDoubleOPFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); +typedef void NeonGenTwoSingleOpFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); +typedef void NeonGenTwoDoubleOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); typedef void NeonGenOne64OpFn(TCGv_i64, TCGv_i64); typedef void CryptoTwoOpFn(TCGv_ptr, TCGv_ptr); typedef void CryptoThreeOpIntFn(TCGv_ptr, TCGv_ptr, TCGv_i32); 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 handle_2misc_fcmp_zero(DisasContext *s, int opcode, TCGv_i64 tcg_op = tcg_temp_new_i64(); TCGv_i64 tcg_zero = tcg_const_i64(0); TCGv_i64 tcg_res = tcg_temp_new_i64(); - NeonGenTwoDoubleOPFn *genfn; + NeonGenTwoDoubleOpFn *genfn; bool swap = false; int pass; @@ -XXX,XX +XXX,XX @@ static void handle_2misc_fcmp_zero(DisasContext *s, int opcode, TCGv_i32 tcg_op = tcg_temp_new_i32(); TCGv_i32 tcg_zero = tcg_const_i32(0); TCGv_i32 tcg_res = tcg_temp_new_i32(); - NeonGenTwoSingleOPFn *genfn; + NeonGenTwoSingleOpFn *genfn; bool swap = false; int pass, maxpasses; diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VSHLL_U_2sh(DisasContext *s, arg_2reg_shift *a) } static bool do_fp_2sh(DisasContext *s, arg_2reg_shift *a, - NeonGenTwoSingleOPFn *fn) + NeonGenTwoSingleOpFn *fn) { /* FP operations in 2-reg-and-shift group */ TCGv_i32 tmp, shiftv; -- 2.20.1 Make gen_swap_half() take a source and destination TCGv_i32 rather than modifying the input TCGv_i32; we're going to want to be able to use it with the more flexible function signature, and this also brings it into line with other functions like gen_rev16() and gen_revsh(). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-12-peter.maydell@linaro.org --- target/arm/translate-neon.inc.c | 2 +- target/arm/translate.c | 10 +++++----- 2 files changed, 6 insertions(+), 6 deletions(-) diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VREV64(DisasContext *s, arg_VREV64 *a) tcg_gen_bswap32_i32(tmp[half], tmp[half]); break; case 1: - gen_swap_half(tmp[half]); + gen_swap_half(tmp[half], tmp[half]); break; case 2: break; 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 gen_revsh(TCGv_i32 dest, TCGv_i32 var) } /* Swap low and high halfwords. */ -static void gen_swap_half(TCGv_i32 var) +static void gen_swap_half(TCGv_i32 dest, TCGv_i32 var) { - tcg_gen_rotri_i32(var, var, 16); + tcg_gen_rotri_i32(dest, var, 16); } /* Dual 16-bit add. Result placed in t0 and t1 is marked as dead. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VREV32: switch (size) { case 0: tcg_gen_bswap32_i32(tmp, tmp); break; - case 1: gen_swap_half(tmp); break; + case 1: gen_swap_half(tmp, tmp); break; default: abort(); } break; @@ -XXX,XX +XXX,XX @@ static bool op_smlad(DisasContext *s, arg_rrrr *a, bool m_swap, bool sub) t1 = load_reg(s, a->rn); t2 = load_reg(s, a->rm); if (m_swap) { - gen_swap_half(t2); + gen_swap_half(t2, t2); } gen_smul_dual(t1, t2); @@ -XXX,XX +XXX,XX @@ static bool op_smlald(DisasContext *s, arg_rrrr *a, bool m_swap, bool sub) t1 = load_reg(s, a->rn); t2 = load_reg(s, a->rm); if (m_swap) { - gen_swap_half(t2); + gen_swap_half(t2, t2); } gen_smul_dual(t1, t2); -- 2.20.1 Convert the VREV32 and VREV16 insns in the Neon 2-reg-misc group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-13-peter.maydell@linaro.org --- target/arm/translate.h | 1 + target/arm/neon-dp.decode | 2 ++ target/arm/translate-neon.inc.c | 55 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 12 ++----- 4 files changed, 60 insertions(+), 10 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 @@ typedef void GVecGen4Fn(unsigned, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t, uint32_t); /* Function prototype for gen_ functions for calling Neon helpers */ +typedef void NeonGenOneOpFn(TCGv_i32, TCGv_i32); typedef void NeonGenOneOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32); typedef void NeonGenTwoOpFn(TCGv_i32, TCGv_i32, TCGv_i32); typedef void NeonGenTwoOpEnvFn(TCGv_i32, TCGv_ptr, TCGv_i32, TCGv_i32); diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm &2misc vm=%vm_dp vd=%vd_dp q=1 VREV64 1111 001 11 . 11 .. 00 .... 0 0000 . . 0 .... @2misc + VREV32 1111 001 11 . 11 .. 00 .... 0 0001 . . 0 .... @2misc + VREV16 1111 001 11 . 11 .. 00 .... 0 0010 . . 0 .... @2misc VPADDL_S 1111 001 11 . 11 .. 00 .... 0 0100 . . 0 .... @2misc VPADDL_U 1111 001 11 . 11 .. 00 .... 0 0101 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_2M_CRYPTO(AESIMC, aa32_aes, 0) DO_2M_CRYPTO(SHA1H, aa32_sha1, 2) DO_2M_CRYPTO(SHA1SU1, aa32_sha1, 2) DO_2M_CRYPTO(SHA256SU0, aa32_sha2, 2) + +static bool do_2misc(DisasContext *s, arg_2misc *a, NeonGenOneOpFn *fn) +{ + int pass; + + /* Handle a 2-reg-misc operation by iterating 32 bits at a time */ + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (!fn) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + for (pass = 0; pass < (a->q ? 4 : 2); pass++) { + TCGv_i32 tmp = neon_load_reg(a->vm, pass); + fn(tmp, tmp); + neon_store_reg(a->vd, pass, tmp); + } + + return true; +} + +static bool trans_VREV32(DisasContext *s, arg_2misc *a) +{ + static NeonGenOneOpFn * const fn[] = { + tcg_gen_bswap32_i32, + gen_swap_half, + NULL, + NULL, + }; + return do_2misc(s, a, fn[a->size]); +} + +static bool trans_VREV16(DisasContext *s, arg_2misc *a) +{ + if (a->size != 0) { + return false; + } + return do_2misc(s, a, gen_rev16); +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_AESE: case NEON_2RM_AESMC: case NEON_2RM_SHA1H: case NEON_2RM_SHA1SU1: + case NEON_2RM_VREV32: + case NEON_2RM_VREV16: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) for (pass = 0; pass < (q ? 4 : 2); pass++) { tmp = neon_load_reg(rm, pass); switch (op) { - case NEON_2RM_VREV32: - switch (size) { - case 0: tcg_gen_bswap32_i32(tmp, tmp); break; - case 1: gen_swap_half(tmp, tmp); break; - default: abort(); - } - break; - case NEON_2RM_VREV16: - gen_rev16(tmp, tmp); - break; case NEON_2RM_VCLS: switch (size) { case 0: gen_helper_neon_cls_s8(tmp, tmp); break; -- 2.20.1 Convert the remaining ops in the Neon 2-reg-misc group which can be implemented simply with our do_2misc() helper. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-14-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 10 +++++ target/arm/translate-neon.inc.c | 69 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 38 ++++-------------- 3 files changed, 86 insertions(+), 31 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm AESMC 1111 001 11 . 11 .. 00 .... 0 0111 0 . 0 .... @2misc_q1 AESIMC 1111 001 11 . 11 .. 00 .... 0 0111 1 . 0 .... @2misc_q1 + VCLS 1111 001 11 . 11 .. 00 .... 0 1000 . . 0 .... @2misc + VCLZ 1111 001 11 . 11 .. 00 .... 0 1001 . . 0 .... @2misc + VCNT 1111 001 11 . 11 .. 00 .... 0 1010 . . 0 .... @2misc + VMVN 1111 001 11 . 11 .. 00 .... 0 1011 . . 0 .... @2misc VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VABS 1111 001 11 . 11 .. 01 .... 0 0110 . . 0 .... @2misc VNEG 1111 001 11 . 11 .. 01 .... 0 0111 . . 0 .... @2misc + VABS_F 1111 001 11 . 11 .. 01 .... 0 1110 . . 0 .... @2misc + VNEG_F 1111 001 11 . 11 .. 01 .... 0 1111 . . 0 .... @2misc + VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VCVT_F16_F32 1111 001 11 . 11 .. 10 .... 0 1100 0 . 0 .... @2misc_q0 VCVT_F32_F16 1111 001 11 . 11 .. 10 .... 0 1110 0 . 0 .... @2misc_q0 + + VRECPE 1111 001 11 . 11 .. 11 .... 0 1000 . . 0 .... @2misc + VRSQRTE 1111 001 11 . 11 .. 11 .... 0 1001 . . 0 .... @2misc ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VREV16(DisasContext *s, arg_2misc *a) } return do_2misc(s, a, gen_rev16); } + +static bool trans_VCLS(DisasContext *s, arg_2misc *a) +{ + static NeonGenOneOpFn * const fn[] = { + gen_helper_neon_cls_s8, + gen_helper_neon_cls_s16, + gen_helper_neon_cls_s32, + NULL, + }; + return do_2misc(s, a, fn[a->size]); +} + +static void do_VCLZ_32(TCGv_i32 rd, TCGv_i32 rm) +{ + tcg_gen_clzi_i32(rd, rm, 32); +} + +static bool trans_VCLZ(DisasContext *s, arg_2misc *a) +{ + static NeonGenOneOpFn * const fn[] = { + gen_helper_neon_clz_u8, + gen_helper_neon_clz_u16, + do_VCLZ_32, + NULL, + }; + return do_2misc(s, a, fn[a->size]); +} + +static bool trans_VCNT(DisasContext *s, arg_2misc *a) +{ + if (a->size != 0) { + return false; + } + return do_2misc(s, a, gen_helper_neon_cnt_u8); +} + +static bool trans_VABS_F(DisasContext *s, arg_2misc *a) +{ + if (a->size != 2) { + return false; + } + /* TODO: FP16 : size == 1 */ + return do_2misc(s, a, gen_helper_vfp_abss); +} + +static bool trans_VNEG_F(DisasContext *s, arg_2misc *a) +{ + if (a->size != 2) { + return false; + } + /* TODO: FP16 : size == 1 */ + return do_2misc(s, a, gen_helper_vfp_negs); +} + +static bool trans_VRECPE(DisasContext *s, arg_2misc *a) +{ + if (a->size != 2) { + return false; + } + return do_2misc(s, a, gen_helper_recpe_u32); +} + +static bool trans_VRSQRTE(DisasContext *s, arg_2misc *a) +{ + if (a->size != 2) { + return false; + } + return do_2misc(s, a, gen_helper_rsqrte_u32); +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_SHA1SU1: case NEON_2RM_VREV32: case NEON_2RM_VREV16: + case NEON_2RM_VCLS: + case NEON_2RM_VCLZ: + case NEON_2RM_VCNT: + case NEON_2RM_VABS_F: + case NEON_2RM_VNEG_F: + case NEON_2RM_VRECPE: + case NEON_2RM_VRSQRTE: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) for (pass = 0; pass < (q ? 4 : 2); pass++) { tmp = neon_load_reg(rm, pass); switch (op) { - case NEON_2RM_VCLS: - switch (size) { - case 0: gen_helper_neon_cls_s8(tmp, tmp); break; - case 1: gen_helper_neon_cls_s16(tmp, tmp); break; - case 2: gen_helper_neon_cls_s32(tmp, tmp); break; - default: abort(); - } - break; - case NEON_2RM_VCLZ: - switch (size) { - case 0: gen_helper_neon_clz_u8(tmp, tmp); break; - case 1: gen_helper_neon_clz_u16(tmp, tmp); break; - case 2: tcg_gen_clzi_i32(tmp, tmp, 32); break; - default: abort(); - } - break; - case NEON_2RM_VCNT: - gen_helper_neon_cnt_u8(tmp, tmp); - break; case NEON_2RM_VQABS: switch (size) { case 0: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_temp_free_ptr(fpstatus); break; } - case NEON_2RM_VABS_F: - gen_helper_vfp_abss(tmp, tmp); - break; - case NEON_2RM_VNEG_F: - gen_helper_vfp_negs(tmp, tmp); - break; case NEON_2RM_VSWP: tmp2 = neon_load_reg(rd, pass); neon_store_reg(rm, pass, tmp2); @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_temp_free_ptr(fpst); break; } - case NEON_2RM_VRECPE: - gen_helper_recpe_u32(tmp, tmp); - break; - case NEON_2RM_VRSQRTE: - gen_helper_rsqrte_u32(tmp, tmp); - break; case NEON_2RM_VRECPE_F: { TCGv_ptr fpstatus = get_fpstatus_ptr(1); -- 2.20.1 Convert the Neon VQABS and VQNEG insns to decodetree. Since these are the only ones which need cpu_env passing to the helper, we wrap the helper rather than creating a whole new do_2misc_env() function. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-15-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 3 +++ target/arm/translate-neon.inc.c | 35 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 30 ++-------------------------- 3 files changed, 40 insertions(+), 28 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VPADAL_S 1111 001 11 . 11 .. 00 .... 0 1100 . . 0 .... @2misc VPADAL_U 1111 001 11 . 11 .. 00 .... 0 1101 . . 0 .... @2misc + VQABS 1111 001 11 . 11 .. 00 .... 0 1110 . . 0 .... @2misc + VQNEG 1111 001 11 . 11 .. 00 .... 0 1111 . . 0 .... @2misc + VCGT0 1111 001 11 . 11 .. 01 .... 0 0000 . . 0 .... @2misc VCGE0 1111 001 11 . 11 .. 01 .... 0 0001 . . 0 .... @2misc VCEQ0 1111 001 11 . 11 .. 01 .... 0 0010 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VRSQRTE(DisasContext *s, arg_2misc *a) } return do_2misc(s, a, gen_helper_rsqrte_u32); } + +#define WRAP_1OP_ENV_FN(WRAPNAME, FUNC) \ + static void WRAPNAME(TCGv_i32 d, TCGv_i32 m) \ + { \ + FUNC(d, cpu_env, m); \ + } + +WRAP_1OP_ENV_FN(gen_VQABS_s8, gen_helper_neon_qabs_s8) +WRAP_1OP_ENV_FN(gen_VQABS_s16, gen_helper_neon_qabs_s16) +WRAP_1OP_ENV_FN(gen_VQABS_s32, gen_helper_neon_qabs_s32) +WRAP_1OP_ENV_FN(gen_VQNEG_s8, gen_helper_neon_qneg_s8) +WRAP_1OP_ENV_FN(gen_VQNEG_s16, gen_helper_neon_qneg_s16) +WRAP_1OP_ENV_FN(gen_VQNEG_s32, gen_helper_neon_qneg_s32) + +static bool trans_VQABS(DisasContext *s, arg_2misc *a) +{ + static NeonGenOneOpFn * const fn[] = { + gen_VQABS_s8, + gen_VQABS_s16, + gen_VQABS_s32, + NULL, + }; + return do_2misc(s, a, fn[a->size]); +} + +static bool trans_VQNEG(DisasContext *s, arg_2misc *a) +{ + static NeonGenOneOpFn * const fn[] = { + gen_VQNEG_s8, + gen_VQNEG_s16, + gen_VQNEG_s32, + NULL, + }; + return do_2misc(s, a, fn[a->size]); +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VNEG_F: case NEON_2RM_VRECPE: case NEON_2RM_VRSQRTE: + case NEON_2RM_VQABS: + case NEON_2RM_VQNEG: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) for (pass = 0; pass < (q ? 4 : 2); pass++) { tmp = neon_load_reg(rm, pass); switch (op) { - case NEON_2RM_VQABS: - switch (size) { - case 0: - gen_helper_neon_qabs_s8(tmp, cpu_env, tmp); - break; - case 1: - gen_helper_neon_qabs_s16(tmp, cpu_env, tmp); - break; - case 2: - gen_helper_neon_qabs_s32(tmp, cpu_env, tmp); - break; - default: abort(); - } - break; - case NEON_2RM_VQNEG: - switch (size) { - case 0: - gen_helper_neon_qneg_s8(tmp, cpu_env, tmp); - break; - case 1: - gen_helper_neon_qneg_s16(tmp, cpu_env, tmp); - break; - case 2: - gen_helper_neon_qneg_s32(tmp, cpu_env, tmp); - break; - default: abort(); - } - break; case NEON_2RM_VCGT0_F: { TCGv_ptr fpstatus = get_fpstatus_ptr(1); -- 2.20.1 Convert the Neon 2-reg-misc insns which are implemented with simple calls to functions that take the input, output and fpstatus pointer. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-16-peter.maydell@linaro.org --- target/arm/translate.h | 1 + target/arm/neon-dp.decode | 8 +++++ target/arm/translate-neon.inc.c | 62 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 56 ++++------------------------- 4 files changed, 78 insertions(+), 49 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 @@ typedef void NeonGenNarrowFn(TCGv_i32, TCGv_i64); typedef void NeonGenNarrowEnvFn(TCGv_i32, TCGv_ptr, TCGv_i64); typedef void NeonGenWidenFn(TCGv_i64, TCGv_i32); typedef void NeonGenTwoOpWidenFn(TCGv_i64, TCGv_i32, TCGv_i32); +typedef void NeonGenOneSingleOpFn(TCGv_i32, TCGv_i32, TCGv_ptr); typedef void NeonGenTwoSingleOpFn(TCGv_i32, TCGv_i32, TCGv_i32, TCGv_ptr); typedef void NeonGenTwoDoubleOpFn(TCGv_i64, TCGv_i64, TCGv_i64, TCGv_ptr); typedef void NeonGenOne64OpFn(TCGv_i64, TCGv_i64); diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm SHA1SU1 1111 001 11 . 11 .. 10 .... 0 0111 0 . 0 .... @2misc_q1 SHA256SU0 1111 001 11 . 11 .. 10 .... 0 0111 1 . 0 .... @2misc_q1 + VRINTX 1111 001 11 . 11 .. 10 .... 0 1001 . . 0 .... @2misc + VCVT_F16_F32 1111 001 11 . 11 .. 10 .... 0 1100 0 . 0 .... @2misc_q0 VCVT_F32_F16 1111 001 11 . 11 .. 10 .... 0 1110 0 . 0 .... @2misc_q0 VRECPE 1111 001 11 . 11 .. 11 .... 0 1000 . . 0 .... @2misc VRSQRTE 1111 001 11 . 11 .. 11 .... 0 1001 . . 0 .... @2misc + VRECPE_F 1111 001 11 . 11 .. 11 .... 0 1010 . . 0 .... @2misc + VRSQRTE_F 1111 001 11 . 11 .. 11 .... 0 1011 . . 0 .... @2misc + VCVT_FS 1111 001 11 . 11 .. 11 .... 0 1100 . . 0 .... @2misc + VCVT_FU 1111 001 11 . 11 .. 11 .... 0 1101 . . 0 .... @2misc + VCVT_SF 1111 001 11 . 11 .. 11 .... 0 1110 . . 0 .... @2misc + VCVT_UF 1111 001 11 . 11 .. 11 .... 0 1111 . . 0 .... @2misc ] # Subgroup for size != 0b11 diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VQNEG(DisasContext *s, arg_2misc *a) }; return do_2misc(s, a, fn[a->size]); } + +static bool do_2misc_fp(DisasContext *s, arg_2misc *a, + NeonGenOneSingleOpFn *fn) +{ + int pass; + TCGv_ptr fpst; + + /* Handle a 2-reg-misc operation by iterating 32 bits at a time */ + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->size != 2) { + /* TODO: FP16 will be the size == 1 case */ + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fpst = get_fpstatus_ptr(1); + for (pass = 0; pass < (a->q ? 4 : 2); pass++) { + TCGv_i32 tmp = neon_load_reg(a->vm, pass); + fn(tmp, tmp, fpst); + neon_store_reg(a->vd, pass, tmp); + } + tcg_temp_free_ptr(fpst); + + return true; +} + +#define DO_2MISC_FP(INSN, FUNC) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + return do_2misc_fp(s, a, FUNC); \ + } + +DO_2MISC_FP(VRECPE_F, gen_helper_recpe_f32) +DO_2MISC_FP(VRSQRTE_F, gen_helper_rsqrte_f32) +DO_2MISC_FP(VCVT_FS, gen_helper_vfp_sitos) +DO_2MISC_FP(VCVT_FU, gen_helper_vfp_uitos) +DO_2MISC_FP(VCVT_SF, gen_helper_vfp_tosizs) +DO_2MISC_FP(VCVT_UF, gen_helper_vfp_touizs) + +static bool trans_VRINTX(DisasContext *s, arg_2misc *a) +{ + if (!arm_dc_feature(s, ARM_FEATURE_V8)) { + return false; + } + return do_2misc_fp(s, a, gen_helper_rints_exact); +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VRSQRTE: case NEON_2RM_VQABS: case NEON_2RM_VQNEG: + case NEON_2RM_VRECPE_F: + case NEON_2RM_VRSQRTE_F: + case NEON_2RM_VCVT_FS: + case NEON_2RM_VCVT_FU: + case NEON_2RM_VCVT_SF: + case NEON_2RM_VCVT_UF: + case NEON_2RM_VRINTX: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_temp_free_i32(tcg_rmode); break; } - case NEON_2RM_VRINTX: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_rints_exact(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } case NEON_2RM_VCVTAU: case NEON_2RM_VCVTAS: case NEON_2RM_VCVTNU: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) tcg_temp_free_ptr(fpst); break; } - case NEON_2RM_VRECPE_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_recpe_f32(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VRSQRTE_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_rsqrte_f32(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCVT_FS: /* VCVT.F32.S32 */ - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_vfp_sitos(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCVT_FU: /* VCVT.F32.U32 */ - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_vfp_uitos(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCVT_SF: /* VCVT.S32.F32 */ - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_vfp_tosizs(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCVT_UF: /* VCVT.U32.F32 */ - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - gen_helper_vfp_touizs(tmp, tmp, fpstatus); - tcg_temp_free_ptr(fpstatus); - break; - } default: /* Reserved op values were caught by the * neon_2rm_sizes[] check earlier. -- 2.20.1 Convert the fp-compare-with-zero insns in the Neon 2-reg-misc group to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-17-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 6 ++++ target/arm/translate-neon.inc.c | 28 ++++++++++++++++++ target/arm/translate.c | 50 ++++----------------------------- 3 files changed, 39 insertions(+), 45 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VABS 1111 001 11 . 11 .. 01 .... 0 0110 . . 0 .... @2misc VNEG 1111 001 11 . 11 .. 01 .... 0 0111 . . 0 .... @2misc + VCGT0_F 1111 001 11 . 11 .. 01 .... 0 1000 . . 0 .... @2misc + VCGE0_F 1111 001 11 . 11 .. 01 .... 0 1001 . . 0 .... @2misc + VCEQ0_F 1111 001 11 . 11 .. 01 .... 0 1010 . . 0 .... @2misc + VCLE0_F 1111 001 11 . 11 .. 01 .... 0 1011 . . 0 .... @2misc + VCLT0_F 1111 001 11 . 11 .. 01 .... 0 1100 . . 0 .... @2misc + VABS_F 1111 001 11 . 11 .. 01 .... 0 1110 . . 0 .... @2misc VNEG_F 1111 001 11 . 11 .. 01 .... 0 1111 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VRINTX(DisasContext *s, arg_2misc *a) } return do_2misc_fp(s, a, gen_helper_rints_exact); } + +#define WRAP_FP_CMP0_FWD(WRAPNAME, FUNC) \ + static void WRAPNAME(TCGv_i32 d, TCGv_i32 m, TCGv_ptr fpst) \ + { \ + TCGv_i32 zero = tcg_const_i32(0); \ + FUNC(d, m, zero, fpst); \ + tcg_temp_free_i32(zero); \ + } +#define WRAP_FP_CMP0_REV(WRAPNAME, FUNC) \ + static void WRAPNAME(TCGv_i32 d, TCGv_i32 m, TCGv_ptr fpst) \ + { \ + TCGv_i32 zero = tcg_const_i32(0); \ + FUNC(d, zero, m, fpst); \ + tcg_temp_free_i32(zero); \ + } + +#define DO_FP_CMP0(INSN, FUNC, REV) \ + WRAP_FP_CMP0_##REV(gen_##INSN, FUNC) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + return do_2misc_fp(s, a, gen_##INSN); \ + } + +DO_FP_CMP0(VCGT0_F, gen_helper_neon_cgt_f32, FWD) +DO_FP_CMP0(VCGE0_F, gen_helper_neon_cge_f32, FWD) +DO_FP_CMP0(VCEQ0_F, gen_helper_neon_ceq_f32, FWD) +DO_FP_CMP0(VCLE0_F, gen_helper_neon_cge_f32, REV) +DO_FP_CMP0(VCLT0_F, gen_helper_neon_cgt_f32, REV) 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VCVT_SF: case NEON_2RM_VCVT_UF: case NEON_2RM_VRINTX: + case NEON_2RM_VCGT0_F: + case NEON_2RM_VCGE0_F: + case NEON_2RM_VCEQ0_F: + case NEON_2RM_VCLE0_F: + case NEON_2RM_VCLT0_F: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) for (pass = 0; pass < (q ? 4 : 2); pass++) { tmp = neon_load_reg(rm, pass); switch (op) { - case NEON_2RM_VCGT0_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - tmp2 = tcg_const_i32(0); - gen_helper_neon_cgt_f32(tmp, tmp, tmp2, fpstatus); - tcg_temp_free_i32(tmp2); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCGE0_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - tmp2 = tcg_const_i32(0); - gen_helper_neon_cge_f32(tmp, tmp, tmp2, fpstatus); - tcg_temp_free_i32(tmp2); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCEQ0_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - tmp2 = tcg_const_i32(0); - gen_helper_neon_ceq_f32(tmp, tmp, tmp2, fpstatus); - tcg_temp_free_i32(tmp2); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCLE0_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - tmp2 = tcg_const_i32(0); - gen_helper_neon_cge_f32(tmp, tmp2, tmp, fpstatus); - tcg_temp_free_i32(tmp2); - tcg_temp_free_ptr(fpstatus); - break; - } - case NEON_2RM_VCLT0_F: - { - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - tmp2 = tcg_const_i32(0); - gen_helper_neon_cgt_f32(tmp, tmp2, tmp, fpstatus); - tcg_temp_free_i32(tmp2); - tcg_temp_free_ptr(fpstatus); - break; - } case NEON_2RM_VSWP: tmp2 = neon_load_reg(rd, pass); neon_store_reg(rm, pass, tmp2); -- 2.20.1 Convert the Neon 2-reg-misc VRINT insns to decodetree. Giving these insns their own do_vrint() function allows us to change the rounding mode just once at the start and end rather than doing it for every element in the vector. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-18-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 8 +++++ target/arm/translate-neon.inc.c | 61 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 31 +++-------------- 3 files changed, 74 insertions(+), 26 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm SHA1SU1 1111 001 11 . 11 .. 10 .... 0 0111 0 . 0 .... @2misc_q1 SHA256SU0 1111 001 11 . 11 .. 10 .... 0 0111 1 . 0 .... @2misc_q1 + VRINTN 1111 001 11 . 11 .. 10 .... 0 1000 . . 0 .... @2misc VRINTX 1111 001 11 . 11 .. 10 .... 0 1001 . . 0 .... @2misc + VRINTA 1111 001 11 . 11 .. 10 .... 0 1010 . . 0 .... @2misc + VRINTZ 1111 001 11 . 11 .. 10 .... 0 1011 . . 0 .... @2misc VCVT_F16_F32 1111 001 11 . 11 .. 10 .... 0 1100 0 . 0 .... @2misc_q0 + + VRINTM 1111 001 11 . 11 .. 10 .... 0 1101 . . 0 .... @2misc + VCVT_F32_F16 1111 001 11 . 11 .. 10 .... 0 1110 0 . 0 .... @2misc_q0 + VRINTP 1111 001 11 . 11 .. 10 .... 0 1111 . . 0 .... @2misc + VRECPE 1111 001 11 . 11 .. 11 .... 0 1000 . . 0 .... @2misc VRSQRTE 1111 001 11 . 11 .. 11 .... 0 1001 . . 0 .... @2misc VRECPE_F 1111 001 11 . 11 .. 11 .... 0 1010 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_FP_CMP0(VCGE0_F, gen_helper_neon_cge_f32, FWD) DO_FP_CMP0(VCEQ0_F, gen_helper_neon_ceq_f32, FWD) DO_FP_CMP0(VCLE0_F, gen_helper_neon_cge_f32, REV) DO_FP_CMP0(VCLT0_F, gen_helper_neon_cgt_f32, REV) + +static bool do_vrint(DisasContext *s, arg_2misc *a, int rmode) +{ + /* + * Handle a VRINT* operation by iterating 32 bits at a time, + * with a specified rounding mode in operation. + */ + int pass; + TCGv_ptr fpst; + TCGv_i32 tcg_rmode; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON) || + !arm_dc_feature(s, ARM_FEATURE_V8)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->size != 2) { + /* TODO: FP16 will be the size == 1 case */ + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fpst = get_fpstatus_ptr(1); + tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode)); + gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env); + for (pass = 0; pass < (a->q ? 4 : 2); pass++) { + TCGv_i32 tmp = neon_load_reg(a->vm, pass); + gen_helper_rints(tmp, tmp, fpst); + neon_store_reg(a->vd, pass, tmp); + } + gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env); + tcg_temp_free_i32(tcg_rmode); + tcg_temp_free_ptr(fpst); + + return true; +} + +#define DO_VRINT(INSN, RMODE) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + return do_vrint(s, a, RMODE); \ + } + +DO_VRINT(VRINTN, FPROUNDING_TIEEVEN) +DO_VRINT(VRINTA, FPROUNDING_TIEAWAY) +DO_VRINT(VRINTZ, FPROUNDING_ZERO) +DO_VRINT(VRINTM, FPROUNDING_NEGINF) +DO_VRINT(VRINTP, FPROUNDING_POSINF) 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VCEQ0_F: case NEON_2RM_VCLE0_F: case NEON_2RM_VCLT0_F: + case NEON_2RM_VRINTN: + case NEON_2RM_VRINTA: + case NEON_2RM_VRINTM: + case NEON_2RM_VRINTP: + case NEON_2RM_VRINTZ: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } neon_store_reg(rm, pass, tmp2); break; - case NEON_2RM_VRINTN: - case NEON_2RM_VRINTA: - case NEON_2RM_VRINTM: - case NEON_2RM_VRINTP: - case NEON_2RM_VRINTZ: - { - TCGv_i32 tcg_rmode; - TCGv_ptr fpstatus = get_fpstatus_ptr(1); - int rmode; - - if (op == NEON_2RM_VRINTZ) { - rmode = FPROUNDING_ZERO; - } else { - rmode = fp_decode_rm[((op & 0x6) >> 1) ^ 1]; - } - - tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode)); - gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, - cpu_env); - gen_helper_rints(tmp, tmp, fpstatus); - gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, - cpu_env); - tcg_temp_free_ptr(fpstatus); - tcg_temp_free_i32(tcg_rmode); - break; - } case NEON_2RM_VCVTAU: case NEON_2RM_VCVTAS: case NEON_2RM_VCVTNU: -- 2.20.1 Convert the VCVT instructions in the 2-reg-misc grouping to decodetree. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-19-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 9 +++++ target/arm/translate-neon.inc.c | 70 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 70 ++++----------------------------- 3 files changed, 87 insertions(+), 62 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VRINTP 1111 001 11 . 11 .. 10 .... 0 1111 . . 0 .... @2misc + VCVTAS 1111 001 11 . 11 .. 11 .... 0 0000 . . 0 .... @2misc + VCVTAU 1111 001 11 . 11 .. 11 .... 0 0001 . . 0 .... @2misc + VCVTNS 1111 001 11 . 11 .. 11 .... 0 0010 . . 0 .... @2misc + VCVTNU 1111 001 11 . 11 .. 11 .... 0 0011 . . 0 .... @2misc + VCVTPS 1111 001 11 . 11 .. 11 .... 0 0100 . . 0 .... @2misc + VCVTPU 1111 001 11 . 11 .. 11 .... 0 0101 . . 0 .... @2misc + VCVTMS 1111 001 11 . 11 .. 11 .... 0 0110 . . 0 .... @2misc + VCVTMU 1111 001 11 . 11 .. 11 .... 0 0111 . . 0 .... @2misc + VRECPE 1111 001 11 . 11 .. 11 .... 0 1000 . . 0 .... @2misc VRSQRTE 1111 001 11 . 11 .. 11 .... 0 1001 . . 0 .... @2misc VRECPE_F 1111 001 11 . 11 .. 11 .... 0 1010 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_VRINT(VRINTA, FPROUNDING_TIEAWAY) DO_VRINT(VRINTZ, FPROUNDING_ZERO) DO_VRINT(VRINTM, FPROUNDING_NEGINF) DO_VRINT(VRINTP, FPROUNDING_POSINF) + +static bool do_vcvt(DisasContext *s, arg_2misc *a, int rmode, bool is_signed) +{ + /* + * Handle a VCVT* operation by iterating 32 bits at a time, + * with a specified rounding mode in operation. + */ + int pass; + TCGv_ptr fpst; + TCGv_i32 tcg_rmode, tcg_shift; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON) || + !arm_dc_feature(s, ARM_FEATURE_V8)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->size != 2) { + /* TODO: FP16 will be the size == 1 case */ + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + fpst = get_fpstatus_ptr(1); + tcg_shift = tcg_const_i32(0); + tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode)); + gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env); + for (pass = 0; pass < (a->q ? 4 : 2); pass++) { + TCGv_i32 tmp = neon_load_reg(a->vm, pass); + if (is_signed) { + gen_helper_vfp_tosls(tmp, tmp, tcg_shift, fpst); + } else { + gen_helper_vfp_touls(tmp, tmp, tcg_shift, fpst); + } + neon_store_reg(a->vd, pass, tmp); + } + gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, cpu_env); + tcg_temp_free_i32(tcg_rmode); + tcg_temp_free_i32(tcg_shift); + tcg_temp_free_ptr(fpst); + + return true; +} + +#define DO_VCVT(INSN, RMODE, SIGNED) \ + static bool trans_##INSN(DisasContext *s, arg_2misc *a) \ + { \ + return do_vcvt(s, a, RMODE, SIGNED); \ + } + +DO_VCVT(VCVTAU, FPROUNDING_TIEAWAY, false) +DO_VCVT(VCVTAS, FPROUNDING_TIEAWAY, true) +DO_VCVT(VCVTNU, FPROUNDING_TIEEVEN, false) +DO_VCVT(VCVTNS, FPROUNDING_TIEEVEN, true) +DO_VCVT(VCVTPU, FPROUNDING_POSINF, false) +DO_VCVT(VCVTPS, FPROUNDING_POSINF, true) +DO_VCVT(VCVTMU, FPROUNDING_NEGINF, false) +DO_VCVT(VCVTMS, FPROUNDING_NEGINF, true) 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 gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) #define NEON_2RM_VCVT_SF 62 #define NEON_2RM_VCVT_UF 63 -static bool neon_2rm_is_v8_op(int op) -{ - /* Return true if this neon 2reg-misc op is ARMv8 and up */ - switch (op) { - case NEON_2RM_VRINTN: - case NEON_2RM_VRINTA: - case NEON_2RM_VRINTM: - case NEON_2RM_VRINTP: - case NEON_2RM_VRINTZ: - case NEON_2RM_VRINTX: - case NEON_2RM_VCVTAU: - case NEON_2RM_VCVTAS: - case NEON_2RM_VCVTNU: - case NEON_2RM_VCVTNS: - case NEON_2RM_VCVTPU: - case NEON_2RM_VCVTPS: - case NEON_2RM_VCVTMU: - case NEON_2RM_VCVTMS: - return true; - default: - return false; - } -} - /* Each entry in this array has bit n set if the insn allows * size value n (otherwise it will UNDEF). Since unallocated * op values will have no bits set they always UNDEF. @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) if ((neon_2rm_sizes[op] & (1 << size)) == 0) { return 1; } - if (neon_2rm_is_v8_op(op) && - !arm_dc_feature(s, ARM_FEATURE_V8)) { - return 1; - } if (q && ((rm | rd) & 1)) { return 1; } @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VRINTM: case NEON_2RM_VRINTP: case NEON_2RM_VRINTZ: + case NEON_2RM_VCVTAU: + case NEON_2RM_VCVTAS: + case NEON_2RM_VCVTNU: + case NEON_2RM_VCVTNS: + case NEON_2RM_VCVTPU: + case NEON_2RM_VCVTPS: + case NEON_2RM_VCVTMU: + case NEON_2RM_VCVTMS: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) } neon_store_reg(rm, pass, tmp2); break; - case NEON_2RM_VCVTAU: - case NEON_2RM_VCVTAS: - case NEON_2RM_VCVTNU: - case NEON_2RM_VCVTNS: - case NEON_2RM_VCVTPU: - case NEON_2RM_VCVTPS: - case NEON_2RM_VCVTMU: - case NEON_2RM_VCVTMS: - { - bool is_signed = !extract32(insn, 7, 1); - TCGv_ptr fpst = get_fpstatus_ptr(1); - TCGv_i32 tcg_rmode, tcg_shift; - int rmode = fp_decode_rm[extract32(insn, 8, 2)]; - - tcg_shift = tcg_const_i32(0); - tcg_rmode = tcg_const_i32(arm_rmode_to_sf(rmode)); - gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, - cpu_env); - - if (is_signed) { - gen_helper_vfp_tosls(tmp, tmp, - tcg_shift, fpst); - } else { - gen_helper_vfp_touls(tmp, tmp, - tcg_shift, fpst); - } - - gen_helper_set_neon_rmode(tcg_rmode, tcg_rmode, - cpu_env); - tcg_temp_free_i32(tcg_rmode); - tcg_temp_free_i32(tcg_shift); - tcg_temp_free_ptr(fpst); - break; - } default: /* Reserved op values were caught by the * neon_2rm_sizes[] check earlier. -- 2.20.1 Convert the Neon VSWP insn to decodetree. Since the new implementation doesn't have to share a pass-loop with the other 2-reg-misc operations we can implement the swap with 64-bit accesses rather than 32-bits (which brings us into line with the pseudocode and is more efficient). Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-20-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 2 ++ target/arm/translate-neon.inc.c | 41 +++++++++++++++++++++++++++++++++ target/arm/translate.c | 5 +--- 3 files changed, 44 insertions(+), 4 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VABS_F 1111 001 11 . 11 .. 01 .... 0 1110 . . 0 .... @2misc VNEG_F 1111 001 11 . 11 .. 01 .... 0 1111 . . 0 .... @2misc + VSWP 1111 001 11 . 11 .. 10 .... 0 0000 . . 0 .... @2misc + VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ DO_VCVT(VCVTPU, FPROUNDING_POSINF, false) DO_VCVT(VCVTPS, FPROUNDING_POSINF, true) DO_VCVT(VCVTMU, FPROUNDING_NEGINF, false) DO_VCVT(VCVTMS, FPROUNDING_NEGINF, true) + +static bool trans_VSWP(DisasContext *s, arg_2misc *a) +{ + TCGv_i64 rm, rd; + int pass; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if (a->size != 0) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + rm = tcg_temp_new_i64(); + rd = tcg_temp_new_i64(); + for (pass = 0; pass < (a->q ? 2 : 1); pass++) { + neon_load_reg64(rm, a->vm + pass); + neon_load_reg64(rd, a->vd + pass); + neon_store_reg64(rm, a->vd + pass); + neon_store_reg64(rd, a->vm + pass); + } + tcg_temp_free_i64(rm); + tcg_temp_free_i64(rd); + + return true; +} 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 int disas_neon_data_insn(DisasContext *s, uint32_t insn) case NEON_2RM_VCVTPS: case NEON_2RM_VCVTMU: case NEON_2RM_VCVTMS: + case NEON_2RM_VSWP: /* handled by decodetree */ return 1; case NEON_2RM_VTRN: @@ -XXX,XX +XXX,XX @@ static int disas_neon_data_insn(DisasContext *s, uint32_t insn) for (pass = 0; pass < (q ? 4 : 2); pass++) { tmp = neon_load_reg(rm, pass); switch (op) { - case NEON_2RM_VSWP: - tmp2 = neon_load_reg(rd, pass); - neon_store_reg(rm, pass, tmp2); - break; case NEON_2RM_VTRN: tmp2 = neon_load_reg(rd, pass); switch (size) { -- 2.20.1 Convert the Neon VTRN insn to decodetree. This is the last insn in the Neon data-processing group, so we can remove all the now-unused old decoder framework. It's possible that there's a more efficient implementation of VTRN, but for this conversion we just copy the existing approach. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-21-peter.maydell@linaro.org --- target/arm/neon-dp.decode | 2 +- target/arm/translate-neon.inc.c | 90 ++++++++ target/arm/translate.c | 363 +------------------------------- 3 files changed, 93 insertions(+), 362 deletions(-) diff --git a/target/arm/neon-dp.decode b/target/arm/neon-dp.decode index XXXXXXX..XXXXXXX 100644 --- a/target/arm/neon-dp.decode +++ b/target/arm/neon-dp.decode @@ -XXX,XX +XXX,XX @@ Vimm_1r 1111 001 . 1 . 000 ... .... cmode:4 0 . op:1 1 .... @1reg_imm VNEG_F 1111 001 11 . 11 .. 01 .... 0 1111 . . 0 .... @2misc VSWP 1111 001 11 . 11 .. 10 .... 0 0000 . . 0 .... @2misc - + VTRN 1111 001 11 . 11 .. 10 .... 0 0001 . . 0 .... @2misc VUZP 1111 001 11 . 11 .. 10 .... 0 0010 . . 0 .... @2misc VZIP 1111 001 11 . 11 .. 10 .... 0 0011 . . 0 .... @2misc diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static bool trans_VSWP(DisasContext *s, arg_2misc *a) return true; } +static void gen_neon_trn_u8(TCGv_i32 t0, TCGv_i32 t1) +{ + TCGv_i32 rd, tmp; + + rd = tcg_temp_new_i32(); + tmp = tcg_temp_new_i32(); + + tcg_gen_shli_i32(rd, t0, 8); + tcg_gen_andi_i32(rd, rd, 0xff00ff00); + tcg_gen_andi_i32(tmp, t1, 0x00ff00ff); + tcg_gen_or_i32(rd, rd, tmp); + + tcg_gen_shri_i32(t1, t1, 8); + tcg_gen_andi_i32(t1, t1, 0x00ff00ff); + tcg_gen_andi_i32(tmp, t0, 0xff00ff00); + tcg_gen_or_i32(t1, t1, tmp); + tcg_gen_mov_i32(t0, rd); + + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(rd); +} + +static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) +{ + TCGv_i32 rd, tmp; + + rd = tcg_temp_new_i32(); + tmp = tcg_temp_new_i32(); + + tcg_gen_shli_i32(rd, t0, 16); + tcg_gen_andi_i32(tmp, t1, 0xffff); + tcg_gen_or_i32(rd, rd, tmp); + tcg_gen_shri_i32(t1, t1, 16); + tcg_gen_andi_i32(tmp, t0, 0xffff0000); + tcg_gen_or_i32(t1, t1, tmp); + tcg_gen_mov_i32(t0, rd); + + tcg_temp_free_i32(tmp); + tcg_temp_free_i32(rd); +} + +static bool trans_VTRN(DisasContext *s, arg_2misc *a) +{ + TCGv_i32 tmp, tmp2; + int pass; + + if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { + return false; + } + + /* UNDEF accesses to D16-D31 if they don't exist. */ + if (!dc_isar_feature(aa32_simd_r32, s) && + ((a->vd | a->vm) & 0x10)) { + return false; + } + + if ((a->vd | a->vm) & a->q) { + return false; + } + + if (a->size == 3) { + return false; + } + + if (!vfp_access_check(s)) { + return true; + } + + if (a->size == 2) { + for (pass = 0; pass < (a->q ? 4 : 2); pass += 2) { + tmp = neon_load_reg(a->vm, pass); + tmp2 = neon_load_reg(a->vd, pass + 1); + neon_store_reg(a->vm, pass, tmp2); + neon_store_reg(a->vd, pass + 1, tmp); + } + } else { + for (pass = 0; pass < (a->q ? 4 : 2); pass++) { + tmp = neon_load_reg(a->vm, pass); + tmp2 = neon_load_reg(a->vd, pass); + if (a->size == 0) { + gen_neon_trn_u8(tmp, tmp2); + } else { + gen_neon_trn_u16(tmp, tmp2); + } + neon_store_reg(a->vm, pass, tmp2); + neon_store_reg(a->vd, pass, tmp); + } + } + return true; +} 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 gen_exception_return(DisasContext *s, TCGv_i32 pc) gen_rfe(s, pc, load_cpu_field(spsr)); } -static void gen_neon_trn_u8(TCGv_i32 t0, TCGv_i32 t1) -{ - TCGv_i32 rd, tmp; - - rd = tcg_temp_new_i32(); - tmp = tcg_temp_new_i32(); - - tcg_gen_shli_i32(rd, t0, 8); - tcg_gen_andi_i32(rd, rd, 0xff00ff00); - tcg_gen_andi_i32(tmp, t1, 0x00ff00ff); - tcg_gen_or_i32(rd, rd, tmp); - - tcg_gen_shri_i32(t1, t1, 8); - tcg_gen_andi_i32(t1, t1, 0x00ff00ff); - tcg_gen_andi_i32(tmp, t0, 0xff00ff00); - tcg_gen_or_i32(t1, t1, tmp); - tcg_gen_mov_i32(t0, rd); - - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(rd); -} - -static void gen_neon_trn_u16(TCGv_i32 t0, TCGv_i32 t1) -{ - TCGv_i32 rd, tmp; - - rd = tcg_temp_new_i32(); - tmp = tcg_temp_new_i32(); - - tcg_gen_shli_i32(rd, t0, 16); - tcg_gen_andi_i32(tmp, t1, 0xffff); - tcg_gen_or_i32(rd, rd, tmp); - tcg_gen_shri_i32(t1, t1, 16); - tcg_gen_andi_i32(tmp, t0, 0xffff0000); - tcg_gen_or_i32(t1, t1, tmp); - tcg_gen_mov_i32(t0, rd); - - tcg_temp_free_i32(tmp); - tcg_temp_free_i32(rd); -} - -/* Symbolic constants for op fields for Neon 2-register miscellaneous. - * The values correspond to bits [17:16,10:7]; see the ARM ARM DDI0406B - * table A7-13. - */ -#define NEON_2RM_VREV64 0 -#define NEON_2RM_VREV32 1 -#define NEON_2RM_VREV16 2 -#define NEON_2RM_VPADDL 4 -#define NEON_2RM_VPADDL_U 5 -#define NEON_2RM_AESE 6 /* Includes AESD */ -#define NEON_2RM_AESMC 7 /* Includes AESIMC */ -#define NEON_2RM_VCLS 8 -#define NEON_2RM_VCLZ 9 -#define NEON_2RM_VCNT 10 -#define NEON_2RM_VMVN 11 -#define NEON_2RM_VPADAL 12 -#define NEON_2RM_VPADAL_U 13 -#define NEON_2RM_VQABS 14 -#define NEON_2RM_VQNEG 15 -#define NEON_2RM_VCGT0 16 -#define NEON_2RM_VCGE0 17 -#define NEON_2RM_VCEQ0 18 -#define NEON_2RM_VCLE0 19 -#define NEON_2RM_VCLT0 20 -#define NEON_2RM_SHA1H 21 -#define NEON_2RM_VABS 22 -#define NEON_2RM_VNEG 23 -#define NEON_2RM_VCGT0_F 24 -#define NEON_2RM_VCGE0_F 25 -#define NEON_2RM_VCEQ0_F 26 -#define NEON_2RM_VCLE0_F 27 -#define NEON_2RM_VCLT0_F 28 -#define NEON_2RM_VABS_F 30 -#define NEON_2RM_VNEG_F 31 -#define NEON_2RM_VSWP 32 -#define NEON_2RM_VTRN 33 -#define NEON_2RM_VUZP 34 -#define NEON_2RM_VZIP 35 -#define NEON_2RM_VMOVN 36 /* Includes VQMOVN, VQMOVUN */ -#define NEON_2RM_VQMOVN 37 /* Includes VQMOVUN */ -#define NEON_2RM_VSHLL 38 -#define NEON_2RM_SHA1SU1 39 /* Includes SHA256SU0 */ -#define NEON_2RM_VRINTN 40 -#define NEON_2RM_VRINTX 41 -#define NEON_2RM_VRINTA 42 -#define NEON_2RM_VRINTZ 43 -#define NEON_2RM_VCVT_F16_F32 44 -#define NEON_2RM_VRINTM 45 -#define NEON_2RM_VCVT_F32_F16 46 -#define NEON_2RM_VRINTP 47 -#define NEON_2RM_VCVTAU 48 -#define NEON_2RM_VCVTAS 49 -#define NEON_2RM_VCVTNU 50 -#define NEON_2RM_VCVTNS 51 -#define NEON_2RM_VCVTPU 52 -#define NEON_2RM_VCVTPS 53 -#define NEON_2RM_VCVTMU 54 -#define NEON_2RM_VCVTMS 55 -#define NEON_2RM_VRECPE 56 -#define NEON_2RM_VRSQRTE 57 -#define NEON_2RM_VRECPE_F 58 -#define NEON_2RM_VRSQRTE_F 59 -#define NEON_2RM_VCVT_FS 60 -#define NEON_2RM_VCVT_FU 61 -#define NEON_2RM_VCVT_SF 62 -#define NEON_2RM_VCVT_UF 63 - -/* Each entry in this array has bit n set if the insn allows - * size value n (otherwise it will UNDEF). Since unallocated - * op values will have no bits set they always UNDEF. - */ -static const uint8_t neon_2rm_sizes[] = { - [NEON_2RM_VREV64] = 0x7, - [NEON_2RM_VREV32] = 0x3, - [NEON_2RM_VREV16] = 0x1, - [NEON_2RM_VPADDL] = 0x7, - [NEON_2RM_VPADDL_U] = 0x7, - [NEON_2RM_AESE] = 0x1, - [NEON_2RM_AESMC] = 0x1, - [NEON_2RM_VCLS] = 0x7, - [NEON_2RM_VCLZ] = 0x7, - [NEON_2RM_VCNT] = 0x1, - [NEON_2RM_VMVN] = 0x1, - [NEON_2RM_VPADAL] = 0x7, - [NEON_2RM_VPADAL_U] = 0x7, - [NEON_2RM_VQABS] = 0x7, - [NEON_2RM_VQNEG] = 0x7, - [NEON_2RM_VCGT0] = 0x7, - [NEON_2RM_VCGE0] = 0x7, - [NEON_2RM_VCEQ0] = 0x7, - [NEON_2RM_VCLE0] = 0x7, - [NEON_2RM_VCLT0] = 0x7, - [NEON_2RM_SHA1H] = 0x4, - [NEON_2RM_VABS] = 0x7, - [NEON_2RM_VNEG] = 0x7, - [NEON_2RM_VCGT0_F] = 0x4, - [NEON_2RM_VCGE0_F] = 0x4, - [NEON_2RM_VCEQ0_F] = 0x4, - [NEON_2RM_VCLE0_F] = 0x4, - [NEON_2RM_VCLT0_F] = 0x4, - [NEON_2RM_VABS_F] = 0x4, - [NEON_2RM_VNEG_F] = 0x4, - [NEON_2RM_VSWP] = 0x1, - [NEON_2RM_VTRN] = 0x7, - [NEON_2RM_VUZP] = 0x7, - [NEON_2RM_VZIP] = 0x7, - [NEON_2RM_VMOVN] = 0x7, - [NEON_2RM_VQMOVN] = 0x7, - [NEON_2RM_VSHLL] = 0x7, - [NEON_2RM_SHA1SU1] = 0x4, - [NEON_2RM_VRINTN] = 0x4, - [NEON_2RM_VRINTX] = 0x4, - [NEON_2RM_VRINTA] = 0x4, - [NEON_2RM_VRINTZ] = 0x4, - [NEON_2RM_VCVT_F16_F32] = 0x2, - [NEON_2RM_VRINTM] = 0x4, - [NEON_2RM_VCVT_F32_F16] = 0x2, - [NEON_2RM_VRINTP] = 0x4, - [NEON_2RM_VCVTAU] = 0x4, - [NEON_2RM_VCVTAS] = 0x4, - [NEON_2RM_VCVTNU] = 0x4, - [NEON_2RM_VCVTNS] = 0x4, - [NEON_2RM_VCVTPU] = 0x4, - [NEON_2RM_VCVTPS] = 0x4, - [NEON_2RM_VCVTMU] = 0x4, - [NEON_2RM_VCVTMS] = 0x4, - [NEON_2RM_VRECPE] = 0x4, - [NEON_2RM_VRSQRTE] = 0x4, - [NEON_2RM_VRECPE_F] = 0x4, - [NEON_2RM_VRSQRTE_F] = 0x4, - [NEON_2RM_VCVT_FS] = 0x4, - [NEON_2RM_VCVT_FU] = 0x4, - [NEON_2RM_VCVT_SF] = 0x4, - [NEON_2RM_VCVT_UF] = 0x4, -}; - static void gen_gvec_fn3_qc(uint32_t rd_ofs, uint32_t rn_ofs, uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz, gen_helper_gvec_3_ptr *fn) @@ -XXX,XX +XXX,XX @@ void gen_gvec_uaba(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs, tcg_gen_gvec_3(rd_ofs, rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]); } -/* Translate a NEON data processing instruction. Return nonzero if the - instruction is invalid. - We process data in a mixture of 32-bit and 64-bit chunks. - Mostly we use 32-bit chunks so we can use normal scalar instructions. */ - -static int disas_neon_data_insn(DisasContext *s, uint32_t insn) -{ - int op; - int q; - int rd, rm; - int size; - int pass; - int u; - TCGv_i32 tmp, tmp2; - - if (!arm_dc_feature(s, ARM_FEATURE_NEON)) { - return 1; - } - - /* FIXME: this access check should not take precedence over UNDEF - * for invalid encodings; we will generate incorrect syndrome information - * for attempts to execute invalid vfp/neon encodings with FP disabled. - */ - if (s->fp_excp_el) { - gen_exception_insn(s, s->pc_curr, EXCP_UDEF, - syn_simd_access_trap(1, 0xe, false), s->fp_excp_el); - return 0; - } - - if (!s->vfp_enabled) - return 1; - q = (insn & (1 << 6)) != 0; - u = (insn >> 24) & 1; - VFP_DREG_D(rd, insn); - VFP_DREG_M(rm, insn); - size = (insn >> 20) & 3; - - if ((insn & (1 << 23)) == 0) { - /* Three register same length: handled by decodetree */ - return 1; - } else if (insn & (1 << 4)) { - /* Two registers and shift or reg and imm: handled by decodetree */ - return 1; - } else { /* (insn & 0x00800010 == 0x00800000) */ - if (size != 3) { - /* - * Three registers of different lengths, or two registers and - * a scalar: handled by decodetree - */ - return 1; - } else { /* size == 3 */ - if (!u) { - /* Extract: handled by decodetree */ - return 1; - } else if ((insn & (1 << 11)) == 0) { - /* Two register misc. */ - op = ((insn >> 12) & 0x30) | ((insn >> 7) & 0xf); - size = (insn >> 18) & 3; - /* UNDEF for unknown op values and bad op-size combinations */ - if ((neon_2rm_sizes[op] & (1 << size)) == 0) { - return 1; - } - if (q && ((rm | rd) & 1)) { - return 1; - } - switch (op) { - case NEON_2RM_VREV64: - case NEON_2RM_VPADDL: case NEON_2RM_VPADDL_U: - case NEON_2RM_VPADAL: case NEON_2RM_VPADAL_U: - case NEON_2RM_VUZP: - case NEON_2RM_VZIP: - case NEON_2RM_VMOVN: case NEON_2RM_VQMOVN: - case NEON_2RM_VSHLL: - case NEON_2RM_VCVT_F16_F32: - case NEON_2RM_VCVT_F32_F16: - case NEON_2RM_VMVN: - case NEON_2RM_VNEG: - case NEON_2RM_VABS: - case NEON_2RM_VCEQ0: - case NEON_2RM_VCGT0: - case NEON_2RM_VCLE0: - case NEON_2RM_VCGE0: - case NEON_2RM_VCLT0: - case NEON_2RM_AESE: case NEON_2RM_AESMC: - case NEON_2RM_SHA1H: - case NEON_2RM_SHA1SU1: - case NEON_2RM_VREV32: - case NEON_2RM_VREV16: - case NEON_2RM_VCLS: - case NEON_2RM_VCLZ: - case NEON_2RM_VCNT: - case NEON_2RM_VABS_F: - case NEON_2RM_VNEG_F: - case NEON_2RM_VRECPE: - case NEON_2RM_VRSQRTE: - case NEON_2RM_VQABS: - case NEON_2RM_VQNEG: - case NEON_2RM_VRECPE_F: - case NEON_2RM_VRSQRTE_F: - case NEON_2RM_VCVT_FS: - case NEON_2RM_VCVT_FU: - case NEON_2RM_VCVT_SF: - case NEON_2RM_VCVT_UF: - case NEON_2RM_VRINTX: - case NEON_2RM_VCGT0_F: - case NEON_2RM_VCGE0_F: - case NEON_2RM_VCEQ0_F: - case NEON_2RM_VCLE0_F: - case NEON_2RM_VCLT0_F: - case NEON_2RM_VRINTN: - case NEON_2RM_VRINTA: - case NEON_2RM_VRINTM: - case NEON_2RM_VRINTP: - case NEON_2RM_VRINTZ: - case NEON_2RM_VCVTAU: - case NEON_2RM_VCVTAS: - case NEON_2RM_VCVTNU: - case NEON_2RM_VCVTNS: - case NEON_2RM_VCVTPU: - case NEON_2RM_VCVTPS: - case NEON_2RM_VCVTMU: - case NEON_2RM_VCVTMS: - case NEON_2RM_VSWP: - /* handled by decodetree */ - return 1; - case NEON_2RM_VTRN: - if (size == 2) { - int n; - for (n = 0; n < (q ? 4 : 2); n += 2) { - tmp = neon_load_reg(rm, n); - tmp2 = neon_load_reg(rd, n + 1); - neon_store_reg(rm, n, tmp2); - neon_store_reg(rd, n + 1, tmp); - } - } else { - goto elementwise; - } - break; - - default: - elementwise: - for (pass = 0; pass < (q ? 4 : 2); pass++) { - tmp = neon_load_reg(rm, pass); - switch (op) { - case NEON_2RM_VTRN: - tmp2 = neon_load_reg(rd, pass); - switch (size) { - case 0: gen_neon_trn_u8(tmp, tmp2); break; - case 1: gen_neon_trn_u16(tmp, tmp2); break; - default: abort(); - } - neon_store_reg(rm, pass, tmp2); - break; - default: - /* Reserved op values were caught by the - * neon_2rm_sizes[] check earlier. - */ - abort(); - } - neon_store_reg(rd, pass, tmp); - } - break; - } - } else { - /* VTBL, VTBX, VDUP: handled by decodetree */ - return 1; - } - } - } - return 0; -} - static int disas_coproc_insn(DisasContext *s, uint32_t insn) { int cpnum, is64, crn, crm, opc1, opc2, isread, rt, rt2; @@ -XXX,XX +XXX,XX @@ static void disas_arm_insn(DisasContext *s, unsigned int insn) } /* fall back to legacy decoder */ - if (((insn >> 25) & 7) == 1) { - /* NEON Data processing. */ - if (disas_neon_data_insn(s, insn)) { - goto illegal_op; - } - return; - } if ((insn & 0x0e000f00) == 0x0c000100) { if (arm_dc_feature(s, ARM_FEATURE_IWMMXT)) { /* iWMMXt register transfer. */ @@ -XXX,XX +XXX,XX @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn) break; } if (((insn >> 24) & 3) == 3) { - /* Translate into the equivalent ARM encoding. */ - insn = (insn & 0xe2ffffff) | ((insn & (1 << 28)) >> 4) | (1 << 28); - if (disas_neon_data_insn(s, insn)) { - goto illegal_op; - } + /* Neon DP, but failed disas_neon_dp() */ + goto illegal_op; } else if (((insn >> 8) & 0xe) == 10) { /* VFP, but failed disas_vfp. */ goto illegal_op; -- 2.20.1 The functions neon_element_offset(), neon_load_element(), neon_load_element64(), neon_store_element() and neon_store_element64() are used only in the translate-neon.inc.c file, so move their definitions there. Since the .inc.c file is #included in translate.c this doesn't make much difference currently, but it's a more logical place to put the functions and it might be helpful if we ever decide to try to make the .inc.c files genuinely separate compilation units. Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200616170844.13318-22-peter.maydell@linaro.org --- target/arm/translate-neon.inc.c | 101 ++++++++++++++++++++++++++++++++ target/arm/translate.c | 101 -------------------------------- 2 files changed, 101 insertions(+), 101 deletions(-) diff --git a/target/arm/translate-neon.inc.c b/target/arm/translate-neon.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-neon.inc.c +++ b/target/arm/translate-neon.inc.c @@ -XXX,XX +XXX,XX @@ static inline int rsub_8(DisasContext *s, int x) #include "decode-neon-ls.inc.c" #include "decode-neon-shared.inc.c" +/* Return the offset of a 2**SIZE piece of a NEON register, at index ELE, + * where 0 is the least significant end of the register. + */ +static inline long +neon_element_offset(int reg, int element, MemOp size) +{ + int element_size = 1 << size; + int ofs = element * element_size; +#ifdef HOST_WORDS_BIGENDIAN + /* Calculate the offset assuming fully little-endian, + * then XOR to account for the order of the 8-byte units. + */ + if (element_size < 8) { + ofs ^= 8 - element_size; + } +#endif + return neon_reg_offset(reg, 0) + ofs; +} + +static void neon_load_element(TCGv_i32 var, int reg, int ele, MemOp mop) +{ + long offset = neon_element_offset(reg, ele, mop & MO_SIZE); + + switch (mop) { + case MO_UB: + tcg_gen_ld8u_i32(var, cpu_env, offset); + break; + case MO_UW: + tcg_gen_ld16u_i32(var, cpu_env, offset); + break; + case MO_UL: + tcg_gen_ld_i32(var, cpu_env, offset); + break; + default: + g_assert_not_reached(); + } +} + +static void neon_load_element64(TCGv_i64 var, int reg, int ele, MemOp mop) +{ + long offset = neon_element_offset(reg, ele, mop & MO_SIZE); + + switch (mop) { + case MO_UB: + tcg_gen_ld8u_i64(var, cpu_env, offset); + break; + case MO_UW: + tcg_gen_ld16u_i64(var, cpu_env, offset); + break; + case MO_UL: + tcg_gen_ld32u_i64(var, cpu_env, offset); + break; + case MO_Q: + tcg_gen_ld_i64(var, cpu_env, offset); + break; + default: + g_assert_not_reached(); + } +} + +static void neon_store_element(int reg, int ele, MemOp size, TCGv_i32 var) +{ + long offset = neon_element_offset(reg, ele, size); + + switch (size) { + case MO_8: + tcg_gen_st8_i32(var, cpu_env, offset); + break; + case MO_16: + tcg_gen_st16_i32(var, cpu_env, offset); + break; + case MO_32: + tcg_gen_st_i32(var, cpu_env, offset); + break; + default: + g_assert_not_reached(); + } +} + +static void neon_store_element64(int reg, int ele, MemOp size, TCGv_i64 var) +{ + long offset = neon_element_offset(reg, ele, size); + + switch (size) { + case MO_8: + tcg_gen_st8_i64(var, cpu_env, offset); + break; + case MO_16: + tcg_gen_st16_i64(var, cpu_env, offset); + break; + case MO_32: + tcg_gen_st32_i64(var, cpu_env, offset); + break; + case MO_64: + tcg_gen_st_i64(var, cpu_env, offset); + break; + default: + g_assert_not_reached(); + } +} + static bool trans_VCMLA(DisasContext *s, arg_VCMLA *a) { int opr_sz; 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 @@ neon_reg_offset (int reg, int n) return vfp_reg_offset(0, sreg); } -/* Return the offset of a 2**SIZE piece of a NEON register, at index ELE, - * where 0 is the least significant end of the register. - */ -static inline long -neon_element_offset(int reg, int element, MemOp size) -{ - int element_size = 1 << size; - int ofs = element * element_size; -#ifdef HOST_WORDS_BIGENDIAN - /* Calculate the offset assuming fully little-endian, - * then XOR to account for the order of the 8-byte units. - */ - if (element_size < 8) { - ofs ^= 8 - element_size; - } -#endif - return neon_reg_offset(reg, 0) + ofs; -} - static TCGv_i32 neon_load_reg(int reg, int pass) { TCGv_i32 tmp = tcg_temp_new_i32(); @@ -XXX,XX +XXX,XX @@ static TCGv_i32 neon_load_reg(int reg, int pass) return tmp; } -static void neon_load_element(TCGv_i32 var, int reg, int ele, MemOp mop) -{ - long offset = neon_element_offset(reg, ele, mop & MO_SIZE); - - switch (mop) { - case MO_UB: - tcg_gen_ld8u_i32(var, cpu_env, offset); - break; - case MO_UW: - tcg_gen_ld16u_i32(var, cpu_env, offset); - break; - case MO_UL: - tcg_gen_ld_i32(var, cpu_env, offset); - break; - default: - g_assert_not_reached(); - } -} - -static void neon_load_element64(TCGv_i64 var, int reg, int ele, MemOp mop) -{ - long offset = neon_element_offset(reg, ele, mop & MO_SIZE); - - switch (mop) { - case MO_UB: - tcg_gen_ld8u_i64(var, cpu_env, offset); - break; - case MO_UW: - tcg_gen_ld16u_i64(var, cpu_env, offset); - break; - case MO_UL: - tcg_gen_ld32u_i64(var, cpu_env, offset); - break; - case MO_Q: - tcg_gen_ld_i64(var, cpu_env, offset); - break; - default: - g_assert_not_reached(); - } -} - static void neon_store_reg(int reg, int pass, TCGv_i32 var) { tcg_gen_st_i32(var, cpu_env, neon_reg_offset(reg, pass)); tcg_temp_free_i32(var); } -static void neon_store_element(int reg, int ele, MemOp size, TCGv_i32 var) -{ - long offset = neon_element_offset(reg, ele, size); - - switch (size) { - case MO_8: - tcg_gen_st8_i32(var, cpu_env, offset); - break; - case MO_16: - tcg_gen_st16_i32(var, cpu_env, offset); - break; - case MO_32: - tcg_gen_st_i32(var, cpu_env, offset); - break; - default: - g_assert_not_reached(); - } -} - -static void neon_store_element64(int reg, int ele, MemOp size, TCGv_i64 var) -{ - long offset = neon_element_offset(reg, ele, size); - - switch (size) { - case MO_8: - tcg_gen_st8_i64(var, cpu_env, offset); - break; - case MO_16: - tcg_gen_st16_i64(var, cpu_env, offset); - break; - case MO_32: - tcg_gen_st32_i64(var, cpu_env, offset); - break; - case MO_64: - tcg_gen_st_i64(var, cpu_env, offset); - break; - default: - g_assert_not_reached(); - } -} - static inline void neon_load_reg64(TCGv_i64 var, int reg) { tcg_gen_ld_i64(var, cpu_env, vfp_reg_offset(1, reg)); -- 2.20.1 Since commit ba3e7926691ed3 it has been unnecessary for target code to call gen_io_end() after an IO instruction in icount mode; it is sufficient to call gen_io_start() before it and to force the end of the TB. Many now-unnecessary calls to gen_io_end() were removed in commit 9e9b10c6491153b, but some were missed or accidentally added later. Remove unneeded calls from the arm target: * the call in the handling of exception-return-via-LDM is unnecessary, and the code is already forcing end-of-TB * the call in the VFP access check code is more complicated: we weren't ending the TB, so we need to add the code to force that by setting DISAS_UPDATE * the doc comment for ARM_CP_IO doesn't need to mention gen_io_end() any more Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru> Message-id: 20200619170324.12093-1-peter.maydell@linaro.org --- target/arm/cpu.h | 2 +- target/arm/translate-vfp.inc.c | 7 +++---- target/arm/translate.c | 3 --- 3 files changed, 4 insertions(+), 8 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 @@ static inline uint64_t cpreg_to_kvm_id(uint32_t cpregid) * migration or KVM state synchronization. (Typically this is for "registers" * which are actually used as instructions for cache maintenance and so on.) * IO indicates that this register does I/O and therefore its accesses - * need to be surrounded by gen_io_start()/gen_io_end(). In particular, + * need to be marked with gen_io_start() and also end the TB. In particular, * registers which implement clocks or timers require this. * RAISES_EXC is for when the read or write hook might raise an exception; * the generated code will synchronize the CPU state before calling the hook diff --git a/target/arm/translate-vfp.inc.c b/target/arm/translate-vfp.inc.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/translate-vfp.inc.c +++ b/target/arm/translate-vfp.inc.c @@ -XXX,XX +XXX,XX @@ static bool full_vfp_access_check(DisasContext *s, bool ignore_vfp_enabled) if (s->v7m_lspact) { /* * Lazy state saving affects external memory and also the NVIC, - * so we must mark it as an IO operation for icount. + * so we must mark it as an IO operation for icount (and cause + * this to be the last insn in the TB). */ if (tb_cflags(s->base.tb) & CF_USE_ICOUNT) { + s->base.is_jmp = DISAS_UPDATE; gen_io_start(); } gen_helper_v7m_preserve_fp_state(cpu_env); - if (tb_cflags(s->base.tb) & CF_USE_ICOUNT) { - gen_io_end(); - } /* * If the preserve_fp_state helper doesn't throw an exception * then it will clear LSPACT; we don't need to repeat this for 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 bool do_ldm(DisasContext *s, arg_ldst_block *a, int min_n) gen_io_start(); } gen_helper_cpsr_write_eret(cpu_env, tmp); - if (tb_cflags(s->base.tb) & CF_USE_ICOUNT) { - gen_io_end(); - } tcg_temp_free_i32(tmp); /* Must exit loop to check un-masked IRQs */ s->base.is_jmp = DISAS_EXIT; -- 2.20.1 In commit cfdb2c0c95ae9205b0 ("target/arm: Vectorize SABA/UABA") we replaced the old handling of SABA/UABA with a vectorized implementation which returns early rather than falling into the loop-ever-elements code. We forgot to delete the part of the old looping code that did the accumulate step, and Coverity correctly warns (CID 1428955) that this code is now dead. Delete it. Fixes: cfdb2c0c95ae9205b0 Signed-off-by: Peter Maydell <peter.maydell@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-id: 20200619171547.29780-1-peter.maydell@linaro.org --- target/arm/translate-a64.c | 12 ------------ 1 file changed, 12 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 void disas_simd_3same_int(DisasContext *s, uint32_t insn) genfn(tcg_res, tcg_op1, tcg_op2); } - if (opcode == 0xf) { - /* SABA, UABA: accumulating ops */ - static NeonGenTwoOpFn * const fns[3] = { - gen_helper_neon_add_u8, - gen_helper_neon_add_u16, - tcg_gen_add_i32, - }; - - read_vec_element_i32(s, tcg_op1, rd, pass, MO_32); - fns[size](tcg_res, tcg_op1, tcg_res); - } - write_vec_element_i32(s, tcg_res, rd, pass, MO_32); tcg_temp_free_i32(tcg_res); -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Add a trace event to see when a guest disable/enable the watchdog. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-2-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/watchdog/cmsdk-apb-watchdog.c | 1 + hw/watchdog/trace-events | 1 + 2 files changed, 2 insertions(+) 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_write(void *opaque, hwaddr offset, break; case A_WDOGLOCK: s->lock = (value != WDOG_UNLOCK_VALUE); + trace_cmsdk_apb_watchdog_lock(s->lock); break; case A_WDOGITCR: if (s->is_luminary) { diff --git a/hw/watchdog/trace-events b/hw/watchdog/trace-events index XXXXXXX..XXXXXXX 100644 --- a/hw/watchdog/trace-events +++ b/hw/watchdog/trace-events @@ -XXX,XX +XXX,XX @@ cmsdk_apb_watchdog_read(uint64_t offset, uint64_t data, unsigned size) "CMSDK APB watchdog read: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %u" cmsdk_apb_watchdog_write(uint64_t offset, uint64_t data, unsigned size) "CMSDK APB watchdog write: offset 0x%" PRIx64 " data 0x%" PRIx64 " size %u" cmsdk_apb_watchdog_reset(void) "CMSDK APB watchdog: reset" +cmsdk_apb_watchdog_lock(uint32_t lock) "CMSDK APB watchdog: lock %" PRIu32 -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Use self-explicit definitions instead of magic values. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-3-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/i2c/versatile_i2c.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-) diff --git a/hw/i2c/versatile_i2c.c b/hw/i2c/versatile_i2c.c index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/versatile_i2c.c +++ b/hw/i2c/versatile_i2c.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "hw/sysbus.h" #include "hw/i2c/bitbang_i2c.h" +#include "hw/registerfields.h" #include "qemu/log.h" #include "qemu/module.h" @@ -XXX,XX +XXX,XX @@ typedef struct VersatileI2CState { int in; } VersatileI2CState; +REG32(CONTROL_GET, 0) +REG32(CONTROL_SET, 0) +REG32(CONTROL_CLR, 4) + static uint64_t versatile_i2c_read(void *opaque, hwaddr offset, unsigned size) { VersatileI2CState *s = (VersatileI2CState *)opaque; - if (offset == 0) { + switch (offset) { + case A_CONTROL_SET: return (s->out & 1) | (s->in << 1); - } else { + default: qemu_log_mask(LOG_GUEST_ERROR, "%s: Bad offset 0x%x\n", __func__, (int)offset); return -1; @@ -XXX,XX +XXX,XX @@ static void versatile_i2c_write(void *opaque, hwaddr offset, VersatileI2CState *s = (VersatileI2CState *)opaque; switch (offset) { - case 0: + case A_CONTROL_SET: s->out |= value & 3; break; - case 4: + case A_CONTROL_CLR: s->out &= ~value; break; default: -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Use self-explicit definitions instead of magic values. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-4-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/i2c/versatile_i2c.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) diff --git a/hw/i2c/versatile_i2c.c b/hw/i2c/versatile_i2c.c index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/versatile_i2c.c +++ b/hw/i2c/versatile_i2c.c @@ -XXX,XX +XXX,XX @@ REG32(CONTROL_GET, 0) REG32(CONTROL_SET, 0) REG32(CONTROL_CLR, 4) +#define SCL BIT(0) +#define SDA BIT(1) + static uint64_t versatile_i2c_read(void *opaque, hwaddr offset, unsigned size) { @@ -XXX,XX +XXX,XX @@ static void versatile_i2c_write(void *opaque, hwaddr offset, qemu_log_mask(LOG_GUEST_ERROR, "%s: Bad offset 0x%x\n", __func__, (int)offset); } - bitbang_i2c_set(&s->bitbang, BITBANG_I2C_SCL, (s->out & 1) != 0); - s->in = bitbang_i2c_set(&s->bitbang, BITBANG_I2C_SDA, (s->out & 2) != 0); + bitbang_i2c_set(&s->bitbang, BITBANG_I2C_SCL, (s->out & SCL) != 0); + s->in = bitbang_i2c_set(&s->bitbang, BITBANG_I2C_SDA, (s->out & SDA) != 0); } static const MemoryRegionOps versatile_i2c_ops = { -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> 'ARM SBCon two-wire serial bus interface' is the official name describing the pair of registers used to bitbanging I2C in the Versatile boards. Make the private VersatileI2CState structure as public ArmSbconI2CState. Add the TYPE_ARM_SBCON_I2C, alias to our current TYPE_VERSATILE_I2C model. Rename the memory region description as 'arm_sbcon_i2c'. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-5-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- include/hw/i2c/arm_sbcon_i2c.h | 35 ++++++++++++++++++++++++++++++++++ hw/i2c/versatile_i2c.c | 17 +++++------------ MAINTAINERS | 1 + 3 files changed, 41 insertions(+), 12 deletions(-) create mode 100644 include/hw/i2c/arm_sbcon_i2c.h diff --git a/include/hw/i2c/arm_sbcon_i2c.h b/include/hw/i2c/arm_sbcon_i2c.h new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/include/hw/i2c/arm_sbcon_i2c.h @@ -XXX,XX +XXX,XX @@ +/* + * ARM SBCon two-wire serial bus interface (I2C bitbang) + * a.k.a. + * ARM Versatile I2C controller + * + * Copyright (c) 2006-2007 CodeSourcery. + * Copyright (c) 2012 Oskar Andero <oskar.andero@gmail.com> + * Copyright (C) 2020 Philippe Mathieu-Daudé <f4bug@amsat.org> + * + * SPDX-License-Identifier: GPL-2.0-or-later + */ +#ifndef HW_I2C_ARM_SBCON_H +#define HW_I2C_ARM_SBCON_H + +#include "hw/sysbus.h" +#include "hw/i2c/bitbang_i2c.h" + +#define TYPE_VERSATILE_I2C "versatile_i2c" +#define TYPE_ARM_SBCON_I2C TYPE_VERSATILE_I2C + +#define ARM_SBCON_I2C(obj) \ + OBJECT_CHECK(ArmSbconI2CState, (obj), TYPE_ARM_SBCON_I2C) + +typedef struct ArmSbconI2CState { + /*< private >*/ + SysBusDevice parent_obj; + /*< public >*/ + + MemoryRegion iomem; + bitbang_i2c_interface bitbang; + int out; + int in; +} ArmSbconI2CState; + +#endif /* HW_I2C_ARM_SBCON_H */ diff --git a/hw/i2c/versatile_i2c.c b/hw/i2c/versatile_i2c.c index XXXXXXX..XXXXXXX 100644 --- a/hw/i2c/versatile_i2c.c +++ b/hw/i2c/versatile_i2c.c @@ -XXX,XX +XXX,XX @@ /* - * ARM Versatile I2C controller + * ARM SBCon two-wire serial bus interface (I2C bitbang) + * a.k.a. ARM Versatile I2C controller * * Copyright (c) 2006-2007 CodeSourcery. * Copyright (c) 2012 Oskar Andero <oskar.andero@gmail.com> @@ -XXX,XX +XXX,XX @@ */ #include "qemu/osdep.h" -#include "hw/sysbus.h" -#include "hw/i2c/bitbang_i2c.h" +#include "hw/i2c/arm_sbcon_i2c.h" #include "hw/registerfields.h" #include "qemu/log.h" #include "qemu/module.h" -#define TYPE_VERSATILE_I2C "versatile_i2c" #define VERSATILE_I2C(obj) \ OBJECT_CHECK(VersatileI2CState, (obj), TYPE_VERSATILE_I2C) -typedef struct VersatileI2CState { - SysBusDevice parent_obj; +typedef ArmSbconI2CState VersatileI2CState; - MemoryRegion iomem; - bitbang_i2c_interface bitbang; - int out; - int in; -} VersatileI2CState; REG32(CONTROL_GET, 0) REG32(CONTROL_SET, 0) @@ -XXX,XX +XXX,XX @@ static void versatile_i2c_init(Object *obj) bus = i2c_init_bus(dev, "i2c"); bitbang_i2c_init(&s->bitbang, bus); memory_region_init_io(&s->iomem, obj, &versatile_i2c_ops, s, - "versatile_i2c", 0x1000); + "arm_sbcon_i2c", 0x1000); sysbus_init_mmio(sbd, &s->iomem); } diff --git a/MAINTAINERS b/MAINTAINERS index XXXXXXX..XXXXXXX 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -XXX,XX +XXX,XX @@ M: Peter Maydell <peter.maydell@linaro.org> L: qemu-arm@nongnu.org S: Maintained F: hw/*/versatile* +F: include/hw/i2c/arm_sbcon_i2c.h F: hw/misc/arm_sysctl.c F: docs/system/arm/versatile.rst -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> By using the TYPE_* definitions for devices, we can: - quickly find where devices are used with 'git-grep' - easily rename a device (one-line change). Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-6-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/realview.c | 3 ++- hw/arm/versatilepb.c | 3 ++- hw/arm/vexpress.c | 3 ++- 3 files changed, 6 insertions(+), 3 deletions(-) diff --git a/hw/arm/realview.c b/hw/arm/realview.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/realview.c +++ b/hw/arm/realview.c @@ -XXX,XX +XXX,XX @@ #include "hw/cpu/a9mpcore.h" #include "hw/intc/realview_gic.h" #include "hw/irq.h" +#include "hw/i2c/arm_sbcon_i2c.h" #define SMP_BOOT_ADDR 0xe0000000 #define SMP_BOOTREG_ADDR 0x10000030 @@ -XXX,XX +XXX,XX @@ static void realview_init(MachineState *machine, } } - dev = sysbus_create_simple("versatile_i2c", 0x10002000, NULL); + dev = sysbus_create_simple(TYPE_VERSATILE_I2C, 0x10002000, NULL); i2c = (I2CBus *)qdev_get_child_bus(dev, "i2c"); i2c_create_slave(i2c, "ds1338", 0x68); diff --git a/hw/arm/versatilepb.c b/hw/arm/versatilepb.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/versatilepb.c +++ b/hw/arm/versatilepb.c @@ -XXX,XX +XXX,XX @@ #include "sysemu/sysemu.h" #include "hw/pci/pci.h" #include "hw/i2c/i2c.h" +#include "hw/i2c/arm_sbcon_i2c.h" #include "hw/irq.h" #include "hw/boards.h" #include "exec/address-spaces.h" @@ -XXX,XX +XXX,XX @@ static void versatile_init(MachineState *machine, int board_id) /* Add PL031 Real Time Clock. */ sysbus_create_simple("pl031", 0x101e8000, pic[10]); - dev = sysbus_create_simple("versatile_i2c", 0x10002000, NULL); + dev = sysbus_create_simple(TYPE_VERSATILE_I2C, 0x10002000, NULL); i2c = (I2CBus *)qdev_get_child_bus(dev, "i2c"); i2c_create_slave(i2c, "ds1338", 0x68); diff --git a/hw/arm/vexpress.c b/hw/arm/vexpress.c index XXXXXXX..XXXXXXX 100644 --- a/hw/arm/vexpress.c +++ b/hw/arm/vexpress.c @@ -XXX,XX +XXX,XX @@ #include "hw/char/pl011.h" #include "hw/cpu/a9mpcore.h" #include "hw/cpu/a15mpcore.h" +#include "hw/i2c/arm_sbcon_i2c.h" #define VEXPRESS_BOARD_ID 0x8e0 #define VEXPRESS_FLASH_SIZE (64 * 1024 * 1024) @@ -XXX,XX +XXX,XX @@ static void vexpress_common_init(MachineState *machine) sysbus_create_simple("sp804", map[VE_TIMER01], pic[2]); sysbus_create_simple("sp804", map[VE_TIMER23], pic[3]); - dev = sysbus_create_simple("versatile_i2c", map[VE_SERIALDVI], NULL); + dev = sysbus_create_simple(TYPE_VERSATILE_I2C, map[VE_SERIALDVI], NULL); i2c = (I2CBus *)qdev_get_child_bus(dev, "i2c"); i2c_create_slave(i2c, "sii9022", 0x39); -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-7-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) 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 @@ typedef struct { MemoryRegion blockram_m2; MemoryRegion blockram_m3; MemoryRegion sram; + /* FPGA APB subsystem */ MPS2SCC scc; + /* CMSDK APB subsystem */ CMSDKAPBDualTimer dualtimer; } MPS2MachineState; @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) g_assert_not_reached(); } + /* 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); - object_initialize_child(OBJECT(mms), "dualtimer", &mms->dualtimer, TYPE_CMSDK_APB_DUALTIMER); qdev_prop_set_uint32(DEVICE(&mms->dualtimer), "pclk-frq", SYSCLK_FRQ); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) qdev_get_gpio_in(armv7m, 10)); sysbus_mmio_map(SYS_BUS_DEVICE(&mms->dualtimer), 0, 0x40002000); + /* FPGA APB subsystem */ object_initialize_child(OBJECT(mms), "scc", &mms->scc, TYPE_MPS2_SCC); sccdev = DEVICE(&mms->scc); qdev_prop_set_uint32(sccdev, "scc-cfg4", 0x2); -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> To differenciate with the CMSDK APB peripheral region, rename this region 'CMSDK AHB peripheral region'. Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-8-f4bug@amsat.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) 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) */ create_unimplemented_device("CMSDK APB peripheral region @0x40000000", 0x40000000, 0x00010000); - create_unimplemented_device("CMSDK peripheral region @0x40010000", + create_unimplemented_device("CMSDK AHB peripheral region @0x40010000", 0x40010000, 0x00010000); create_unimplemented_device("Extra peripheral region @0x40020000", 0x40020000, 0x00010000); + create_unimplemented_device("RESERVED 4", 0x40030000, 0x001D0000); create_unimplemented_device("VGA", 0x41000000, 0x0200000); -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> We already model the CMSDK APB watchdog device, let's use it! Suggested-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-9-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 7 +++++++ hw/arm/Kconfig | 1 + 2 files changed, 8 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 @@ static void mps2_common_init(MachineState *machine) sysbus_connect_irq(SYS_BUS_DEVICE(&mms->dualtimer), 0, qdev_get_gpio_in(armv7m, 10)); 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); + 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)); + sysbus_mmio_map(SYS_BUS_DEVICE(&mms->watchdog), 0, 0x40008000); /* FPGA APB subsystem */ object_initialize_child(OBJECT(mms), "scc", &mms->scc, TYPE_MPS2_SCC); 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 MPS2 select PL080 # DMA controller select SPLIT_IRQ select UNIMP + select CMSDK_APB_WATCHDOG config FSL_IMX7 bool -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Register the GPIO peripherals as unimplemented to better follow their accesses, for example booting Zephyr: ---------------- IN: arm_mps2_pinmux_init 0x00001160: f64f 0231 movw r2, #0xf831 0x00001164: 4b06 ldr r3, [pc, #0x18] 0x00001166: 2000 movs r0, #0 0x00001168: 619a str r2, [r3, #0x18] 0x0000116a: f24c 426f movw r2, #0xc46f 0x0000116e: f503 5380 add.w r3, r3, #0x1000 0x00001172: 619a str r2, [r3, #0x18] 0x00001174: f44f 529e mov.w r2, #0x13c0 0x00001178: f503 5380 add.w r3, r3, #0x1000 0x0000117c: 619a str r2, [r3, #0x18] 0x0000117e: 4770 bx lr cmsdk-ahb-gpio: unimplemented device write (size 4, value 0xf831, offset 0x18) cmsdk-ahb-gpio: unimplemented device write (size 4, value 0xc46f, offset 0x18) cmsdk-ahb-gpio: unimplemented device write (size 4, value 0x13c0, offset 0x18) Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-10-f4bug@amsat.org Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-) 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) MemoryRegion *system_memory = get_system_memory(); MachineClass *mc = MACHINE_GET_CLASS(machine); DeviceState *armv7m, *sccdev; + int i; if (strcmp(machine->cpu_type, mc->default_cpu_type) != 0) { error_report("This board can only be used with CPU %s", @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) */ Object *orgate; DeviceState *orgate_dev; - int i; orgate = object_new(TYPE_OR_IRQ); object_property_set_int(orgate, 6, "num-lines", &error_fatal); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) */ Object *orgate; DeviceState *orgate_dev; - int i; orgate = object_new(TYPE_OR_IRQ); object_property_set_int(orgate, 10, "num-lines", &error_fatal); @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) default: g_assert_not_reached(); } + for (i = 0; i < 4; i++) { + static const hwaddr gpiobase[] = {0x40010000, 0x40011000, + 0x40012000, 0x40013000}; + create_unimplemented_device("cmsdk-ahb-gpio", gpiobase[i], 0x1000); + } /* CMSDK APB subsystem */ cmsdk_apb_timer_create(0x40000000, qdev_get_gpio_in(armv7m, 8), SYSCLK_FRQ); -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-11-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <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/timer/cmsdk-apb-timer.h" #include "hw/timer/cmsdk-apb-dualtimer.h" #include "hw/misc/mps2-scc.h" +#include "hw/misc/mps2-fpgaio.h" #include "hw/net/lan9118.h" #include "net/net.h" +#include "hw/watchdog/cmsdk-apb-watchdog.h" typedef enum MPS2FPGAType { FPGA_AN385, @@ -XXX,XX +XXX,XX @@ typedef struct { MemoryRegion sram; /* FPGA APB subsystem */ MPS2SCC scc; + MPS2FPGAIO fpgaio; /* CMSDK APB subsystem */ CMSDKAPBDualTimer dualtimer; + CMSDKAPBWatchdog watchdog; } MPS2MachineState; #define TYPE_MPS2_MACHINE "mps2" @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) qdev_prop_set_uint32(sccdev, "scc-id", mmc->scc_id); sysbus_realize(SYS_BUS_DEVICE(&mms->scc), &error_fatal); sysbus_mmio_map(SYS_BUS_DEVICE(sccdev), 0, 0x4002f000); + object_initialize_child(OBJECT(mms), "fpgaio", + &mms->fpgaio, TYPE_MPS2_FPGAIO); + qdev_prop_set_uint32(DEVICE(&mms->fpgaio), "prescale-clk", 25000000); + sysbus_realize(SYS_BUS_DEVICE(&mms->fpgaio), &error_fatal); + sysbus_mmio_map(SYS_BUS_DEVICE(&mms->fpgaio), 0, 0x40028000); /* In hardware this is a LAN9220; the LAN9118 is software compatible * except that it doesn't support the checksum-offload feature. -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> From 'Application Note AN385', chapter 3.9, SPI: The SMM implements five PL022 SPI modules. Two pairs of modules share the same OR-gated IRQ. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-12-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 24 ++++++++++++++++++++++++ hw/arm/Kconfig | 6 +++--- 2 files changed, 27 insertions(+), 3 deletions(-) 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/timer/cmsdk-apb-dualtimer.h" #include "hw/misc/mps2-scc.h" #include "hw/misc/mps2-fpgaio.h" +#include "hw/ssi/pl022.h" #include "hw/net/lan9118.h" #include "net/net.h" #include "hw/watchdog/cmsdk-apb-watchdog.h" @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) qdev_prop_set_uint32(DEVICE(&mms->fpgaio), "prescale-clk", 25000000); sysbus_realize(SYS_BUS_DEVICE(&mms->fpgaio), &error_fatal); sysbus_mmio_map(SYS_BUS_DEVICE(&mms->fpgaio), 0, 0x40028000); + sysbus_create_simple(TYPE_PL022, 0x40025000, /* External ADC */ + qdev_get_gpio_in(armv7m, 22)); + for (i = 0; i < 2; i++) { + static const int spi_irqno[] = {11, 24}; + static const hwaddr spibase[] = {0x40020000, /* APB */ + 0x40021000, /* LCD */ + 0x40026000, /* Shield0 */ + 0x40027000}; /* Shield1 */ + DeviceState *orgate_dev; + Object *orgate; + int j; + + orgate = object_new(TYPE_OR_IRQ); + object_property_set_int(orgate, 2, "num-lines", &error_fatal); + orgate_dev = DEVICE(orgate); + qdev_realize(orgate_dev, NULL, &error_fatal); + qdev_connect_gpio_out(orgate_dev, 0, + qdev_get_gpio_in(armv7m, spi_irqno[i])); + for (j = 0; j < 2; j++) { + sysbus_create_simple(TYPE_PL022, spibase[2 * i + j], + qdev_get_gpio_in(orgate_dev, j)); + } + } /* In hardware this is a LAN9220; the LAN9118 is software compatible * except that it doesn't support the checksum-offload feature. 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 HIGHBANK select ARM_TIMER # sp804 select ARM_V7M select PL011 # UART - select PL022 # Serial port + select PL022 # SPI select PL031 # RTC select PL061 # GPIO select PL310 # cache controller @@ -XXX,XX +XXX,XX @@ config STELLARIS select CMSDK_APB_WATCHDOG select I2C select PL011 # UART - select PL022 # Serial port + select PL022 # SPI select PL061 # GPIO select SSD0303 # OLED display select SSD0323 # OLED display @@ -XXX,XX +XXX,XX @@ config MPS2 select MPS2_FPGAIO select MPS2_SCC select OR_IRQ - select PL022 # Serial port + select PL022 # SPI select PL080 # DMA controller select SPLIT_IRQ select UNIMP -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> From 'Application Note AN385', chapter 3.14: The SMM implements a simple SBCon interface based on I2C. There are 4 SBCon interfaces on the FPGA APB subsystem. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-13-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 8 ++++++++ hw/arm/Kconfig | 1 + 2 files 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/misc/mps2-scc.h" #include "hw/misc/mps2-fpgaio.h" #include "hw/ssi/pl022.h" +#include "hw/i2c/arm_sbcon_i2c.h" #include "hw/net/lan9118.h" #include "net/net.h" #include "hw/watchdog/cmsdk-apb-watchdog.h" @@ -XXX,XX +XXX,XX @@ static void mps2_common_init(MachineState *machine) qdev_get_gpio_in(orgate_dev, j)); } } + for (i = 0; i < 4; i++) { + static const hwaddr i2cbase[] = {0x40022000, /* Touch */ + 0x40023000, /* Audio */ + 0x40029000, /* Shield0 */ + 0x4002a000}; /* Shield1 */ + sysbus_create_simple(TYPE_ARM_SBCON_I2C, i2cbase[i], NULL); + } /* In hardware this is a LAN9220; the LAN9118 is software compatible * except that it doesn't support the checksum-offload feature. 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 MPS2 select SPLIT_IRQ select UNIMP select CMSDK_APB_WATCHDOG + select VERSATILE_I2C config FSL_IMX7 bool -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-14-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2.c | 1 + 1 file changed, 1 insertion(+) 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) 0x4002a000}; /* Shield1 */ sysbus_create_simple(TYPE_ARM_SBCON_I2C, i2cbase[i], NULL); } + create_unimplemented_device("i2s", 0x40024000, 0x400); /* In hardware this is a LAN9220; the LAN9118 is software compatible * except that it doesn't support the checksum-offload feature. -- 2.20.1 From: Philippe Mathieu-Daudé <f4bug@amsat.org> From 'Application Note AN521', chapter 4.7: The SMM implements four SBCon serial modules: One SBCon module for use by the Color LCD touch interface. One SBCon module to configure the audio controller. Two general purpose SBCon modules, that connect to the Expansion headers J7 and J8, are intended for use with the V2C-Shield1 which provide an I2C interface on the headers. Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Message-id: 20200617072539.32686-15-f4bug@amsat.org Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/arm/mps2-tz.c | 23 ++++++++++++++++++----- 1 file changed, 18 insertions(+), 5 deletions(-) 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/arm/armsse.h" #include "hw/dma/pl080.h" #include "hw/ssi/pl022.h" +#include "hw/i2c/arm_sbcon_i2c.h" #include "hw/net/lan9118.h" #include "net/net.h" #include "hw/core/split-irq.h" @@ -XXX,XX +XXX,XX @@ typedef struct { TZPPC ppc[5]; TZMPC ssram_mpc[3]; PL022State spi[5]; - UnimplementedDeviceState i2c[4]; + ArmSbconI2CState i2c[4]; UnimplementedDeviceState i2s_audio; UnimplementedDeviceState gpio[4]; UnimplementedDeviceState gfx; @@ -XXX,XX +XXX,XX @@ static MemoryRegion *make_spi(MPS2TZMachineState *mms, void *opaque, return sysbus_mmio_get_region(s, 0); } +static MemoryRegion *make_i2c(MPS2TZMachineState *mms, void *opaque, + const char *name, hwaddr size) +{ + ArmSbconI2CState *i2c = opaque; + SysBusDevice *s; + + object_initialize_child(OBJECT(mms), name, i2c, TYPE_ARM_SBCON_I2C); + s = SYS_BUS_DEVICE(i2c); + sysbus_realize(s, &error_fatal); + return sysbus_mmio_get_region(s, 0); +} + static void mps2tz_common_init(MachineState *machine) { MPS2TZMachineState *mms = MPS2TZ_MACHINE(machine); @@ -XXX,XX +XXX,XX @@ static void mps2tz_common_init(MachineState *machine) { "uart2", make_uart, &mms->uart[2], 0x40202000, 0x1000 }, { "uart3", make_uart, &mms->uart[3], 0x40203000, 0x1000 }, { "uart4", make_uart, &mms->uart[4], 0x40204000, 0x1000 }, - { "i2c0", make_unimp_dev, &mms->i2c[0], 0x40207000, 0x1000 }, - { "i2c1", make_unimp_dev, &mms->i2c[1], 0x40208000, 0x1000 }, - { "i2c2", make_unimp_dev, &mms->i2c[2], 0x4020c000, 0x1000 }, - { "i2c3", make_unimp_dev, &mms->i2c[3], 0x4020d000, 0x1000 }, + { "i2c0", make_i2c, &mms->i2c[0], 0x40207000, 0x1000 }, + { "i2c1", make_i2c, &mms->i2c[1], 0x40208000, 0x1000 }, + { "i2c2", make_i2c, &mms->i2c[2], 0x4020c000, 0x1000 }, + { "i2c3", make_i2c, &mms->i2c[3], 0x4020d000, 0x1000 }, }, }, { .name = "apb_ppcexp2", -- 2.20.1 From: Philippe Mathieu-Daudé <philmd@redhat.com> Since commit d70c996df23f, when enabling the PMU we get: $ qemu-system-aarch64 -cpu host,pmu=on -M virt,accel=kvm,gic-version=3 Segmentation fault (core dumped) Thread 1 "qemu-system-aar" received signal SIGSEGV, Segmentation fault. 0x0000aaaaaae356d0 in kvm_ioctl (s=0x0, type=44547) at accel/kvm/kvm-all.c:2588 2588 ret = ioctl(s->fd, type, arg); (gdb) bt #0 0x0000aaaaaae356d0 in kvm_ioctl (s=0x0, type=44547) at accel/kvm/kvm-all.c:2588 #1 0x0000aaaaaae31568 in kvm_check_extension (s=0x0, extension=126) at accel/kvm/kvm-all.c:916 #2 0x0000aaaaaafce254 in kvm_arm_pmu_supported (cpu=0xaaaaac214ab0) at target/arm/kvm.c:213 #3 0x0000aaaaaafc0f94 in arm_set_pmu (obj=0xaaaaac214ab0, value=true, errp=0xffffffffe438) at target/arm/cpu.c:1111 #4 0x0000aaaaab5533ac in property_set_bool (obj=0xaaaaac214ab0, v=0xaaaaac223a80, name=0xaaaaac11a970 "pmu", opaque=0xaaaaac222730, errp=0xffffffffe438) at qom/object.c:2170 #5 0x0000aaaaab5512f0 in object_property_set (obj=0xaaaaac214ab0, v=0xaaaaac223a80, name=0xaaaaac11a970 "pmu", errp=0xffffffffe438) at qom/object.c:1328 #6 0x0000aaaaab551e10 in object_property_parse (obj=0xaaaaac214ab0, string=0xaaaaac11b4c0 "on", name=0xaaaaac11a970 "pmu", errp=0xffffffffe438) at qom/object.c:1561 #7 0x0000aaaaab54ee8c in object_apply_global_props (obj=0xaaaaac214ab0, props=0xaaaaac018e20, errp=0xaaaaabd6fd88 <error_fatal>) at qom/object.c:407 #8 0x0000aaaaab1dd5a4 in qdev_prop_set_globals (dev=0xaaaaac214ab0) at hw/core/qdev-properties.c:1218 #9 0x0000aaaaab1d9fac in device_post_init (obj=0xaaaaac214ab0) at hw/core/qdev.c:1050 ... #15 0x0000aaaaab54f310 in object_initialize_with_type (obj=0xaaaaac214ab0, size=52208, type=0xaaaaabe237f0) at qom/object.c:512 #16 0x0000aaaaab54fa24 in object_new_with_type (type=0xaaaaabe237f0) at qom/object.c:687 #17 0x0000aaaaab54fa80 in object_new (typename=0xaaaaabe23970 "host-arm-cpu") at qom/object.c:702 #18 0x0000aaaaaaf04a74 in machvirt_init (machine=0xaaaaac0a8550) at hw/arm/virt.c:1770 #19 0x0000aaaaab1e8720 in machine_run_board_init (machine=0xaaaaac0a8550) at hw/core/machine.c:1138 #20 0x0000aaaaaaf95394 in qemu_init (argc=5, argv=0xffffffffea58, envp=0xffffffffea88) at softmmu/vl.c:4348 #21 0x0000aaaaaada3f74 in main (argc=<optimized out>, argv=<optimized out>, envp=<optimized out>) at softmmu/main.c:48 This is because in frame #2, cpu->kvm_state is still NULL (the vCPU is not yet realized). KVM has a hard requirement of all cores supporting the same feature set. We only need to check if the accelerator supports a feature, not each vCPU individually. Fix by removing the 'CPUState *cpu' argument from the kvm_arm_<FEATURE>_supported() functions. Fixes: d70c996df23f ('Use CPUState::kvm_state in kvm_arm_pmu_supported') Reported-by: Haibo Xu <haibo.xu@linaro.org> Reviewed-by: Andrew Jones <drjones@redhat.com> Acked-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Suggested-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- target/arm/kvm_arm.h | 21 +++++++++------------ target/arm/cpu.c | 2 +- target/arm/cpu64.c | 10 +++++----- target/arm/kvm.c | 4 ++-- target/arm/kvm64.c | 14 +++++--------- 5 files changed, 22 insertions(+), 29 deletions(-) diff --git a/target/arm/kvm_arm.h b/target/arm/kvm_arm.h index XXXXXXX..XXXXXXX 100644 --- a/target/arm/kvm_arm.h +++ b/target/arm/kvm_arm.h @@ -XXX,XX +XXX,XX @@ void kvm_arm_add_vcpu_properties(Object *obj); /** * kvm_arm_aarch32_supported: - * @cs: CPUState * - * Returns: true if the KVM VCPU can enable AArch32 mode + * Returns: true if KVM can enable AArch32 mode * and false otherwise. */ -bool kvm_arm_aarch32_supported(CPUState *cs); +bool kvm_arm_aarch32_supported(void); /** * kvm_arm_pmu_supported: - * @cs: CPUState * - * Returns: true if the KVM VCPU can enable its PMU + * Returns: true if KVM can enable the PMU * and false otherwise. */ -bool kvm_arm_pmu_supported(CPUState *cs); +bool kvm_arm_pmu_supported(void); /** * kvm_arm_sve_supported: - * @cs: CPUState * - * Returns true if the KVM VCPU can enable SVE and false otherwise. + * Returns true if KVM can enable SVE and false otherwise. */ -bool kvm_arm_sve_supported(CPUState *cs); +bool kvm_arm_sve_supported(void); /** * kvm_arm_get_max_vm_ipa_size: @@ -XXX,XX +XXX,XX @@ static inline void kvm_arm_set_cpu_features_from_host(ARMCPU *cpu) static inline void kvm_arm_add_vcpu_properties(Object *obj) {} -static inline bool kvm_arm_aarch32_supported(CPUState *cs) +static inline bool kvm_arm_aarch32_supported(void) { return false; } -static inline bool kvm_arm_pmu_supported(CPUState *cs) +static inline bool kvm_arm_pmu_supported(void) { return false; } -static inline bool kvm_arm_sve_supported(CPUState *cs) +static inline bool kvm_arm_sve_supported(void) { return false; } 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_set_pmu(Object *obj, bool value, Error **errp) ARMCPU *cpu = ARM_CPU(obj); if (value) { - if (kvm_enabled() && !kvm_arm_pmu_supported(CPU(cpu))) { + if (kvm_enabled() && !kvm_arm_pmu_supported()) { error_setg(errp, "'pmu' feature not supported by KVM on this host"); return; } 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 @@ void arm_cpu_sve_finalize(ARMCPU *cpu, Error **errp) /* Collect the set of vector lengths supported by KVM. */ bitmap_zero(kvm_supported, ARM_MAX_VQ); - if (kvm_enabled() && kvm_arm_sve_supported(CPU(cpu))) { + if (kvm_enabled() && kvm_arm_sve_supported()) { kvm_arm_sve_get_vls(CPU(cpu), kvm_supported); } else if (kvm_enabled()) { assert(!cpu_isar_feature(aa64_sve, cpu)); @@ -XXX,XX +XXX,XX @@ static void cpu_max_set_sve_max_vq(Object *obj, Visitor *v, const char *name, return; } - if (kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) { + if (kvm_enabled() && !kvm_arm_sve_supported()) { error_setg(errp, "cannot set sve-max-vq"); error_append_hint(errp, "SVE not supported by KVM on this host\n"); return; @@ -XXX,XX +XXX,XX @@ static void cpu_arm_set_sve_vq(Object *obj, Visitor *v, const char *name, return; } - if (value && kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) { + if (value && kvm_enabled() && !kvm_arm_sve_supported()) { error_setg(errp, "cannot enable %s", name); error_append_hint(errp, "SVE not supported by KVM on this host\n"); return; @@ -XXX,XX +XXX,XX @@ static void cpu_arm_set_sve(Object *obj, Visitor *v, const char *name, return; } - if (value && kvm_enabled() && !kvm_arm_sve_supported(CPU(cpu))) { + if (value && kvm_enabled() && !kvm_arm_sve_supported()) { error_setg(errp, "'sve' feature not supported by KVM on this host"); return; } @@ -XXX,XX +XXX,XX @@ static void aarch64_cpu_set_aarch64(Object *obj, bool value, Error **errp) * uniform execution state like do_interrupt. */ if (value == false) { - if (!kvm_enabled() || !kvm_arm_aarch32_supported(CPU(cpu))) { + if (!kvm_enabled() || !kvm_arm_aarch32_supported()) { error_setg(errp, "'aarch64' feature cannot be disabled " "unless KVM is enabled and 32-bit EL1 " "is supported"); diff --git a/target/arm/kvm.c b/target/arm/kvm.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/kvm.c +++ b/target/arm/kvm.c @@ -XXX,XX +XXX,XX @@ void kvm_arm_add_vcpu_properties(Object *obj) } } -bool kvm_arm_pmu_supported(CPUState *cpu) +bool kvm_arm_pmu_supported(void) { - return kvm_check_extension(cpu->kvm_state, KVM_CAP_ARM_PMU_V3); + return kvm_check_extension(kvm_state, KVM_CAP_ARM_PMU_V3); } int kvm_arm_get_max_vm_ipa_size(MachineState *ms) 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) return true; } -bool kvm_arm_aarch32_supported(CPUState *cpu) +bool kvm_arm_aarch32_supported(void) { - KVMState *s = KVM_STATE(current_accel()); - - return kvm_check_extension(s, KVM_CAP_ARM_EL1_32BIT); + return kvm_check_extension(kvm_state, KVM_CAP_ARM_EL1_32BIT); } -bool kvm_arm_sve_supported(CPUState *cpu) +bool kvm_arm_sve_supported(void) { - KVMState *s = KVM_STATE(current_accel()); - - return kvm_check_extension(s, KVM_CAP_ARM_SVE); + return kvm_check_extension(kvm_state, KVM_CAP_ARM_SVE); } QEMU_BUILD_BUG_ON(KVM_ARM64_SVE_VQ_MIN != 1); @@ -XXX,XX +XXX,XX @@ int kvm_arch_init_vcpu(CPUState *cs) env->features &= ~(1ULL << ARM_FEATURE_PMU); } if (cpu_isar_feature(aa64_sve, cpu)) { - assert(kvm_arm_sve_supported(cs)); + assert(kvm_arm_sve_supported()); cpu->kvm_init_features[0] |= 1 << KVM_ARM_VCPU_SVE; } -- 2.20.1 From: Andrew Jones <drjones@redhat.com> Some cpu features may be enabled and disabled for all configurations that support the feature. Let's test that. A recent regression[*] inspired adding these tests. [*] '-cpu host,pmu=on' caused a segfault Signed-off-by: Andrew Jones <drjones@redhat.com> Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com> Message-id: 20200623090622.30365-2-philmd@redhat.com Message-Id: <20200623082310.17577-1-drjones@redhat.com> Reviewed-by: Peter Maydell <peter.maydell@linaro.org> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- tests/qtest/arm-cpu-features.c | 38 ++++++++++++++++++++++++++++++---- 1 file changed, 34 insertions(+), 4 deletions(-) diff --git a/tests/qtest/arm-cpu-features.c b/tests/qtest/arm-cpu-features.c index XXXXXXX..XXXXXXX 100644 --- a/tests/qtest/arm-cpu-features.c +++ b/tests/qtest/arm-cpu-features.c @@ -XXX,XX +XXX,XX @@ static bool resp_get_feature(QDict *resp, const char *feature) qobject_unref(_resp); \ }) -#define assert_feature(qts, cpu_type, feature, expected_value) \ +#define resp_assert_feature(resp, feature, expected_value) \ ({ \ - QDict *_resp, *_props; \ + QDict *_props; \ \ - _resp = do_query_no_props(qts, cpu_type); \ g_assert(_resp); \ g_assert(resp_has_props(_resp)); \ _props = resp_get_props(_resp); \ g_assert(qdict_get(_props, feature)); \ g_assert(qdict_get_bool(_props, feature) == (expected_value)); \ +}) + +#define assert_feature(qts, cpu_type, feature, expected_value) \ +({ \ + QDict *_resp; \ + \ + _resp = do_query_no_props(qts, cpu_type); \ + g_assert(_resp); \ + resp_assert_feature(_resp, feature, expected_value); \ + qobject_unref(_resp); \ +}) + +#define assert_set_feature(qts, cpu_type, feature, value) \ +({ \ + const char *_fmt = (value) ? "{ %s: true }" : "{ %s: false }"; \ + QDict *_resp; \ + \ + _resp = do_query(qts, cpu_type, _fmt, feature); \ + g_assert(_resp); \ + resp_assert_feature(_resp, feature, value); \ qobject_unref(_resp); \ }) @@ -XXX,XX +XXX,XX @@ static void test_query_cpu_model_expansion(const void *data) assert_error(qts, "host", "The CPU type 'host' requires KVM", NULL); /* Test expected feature presence/absence for some cpu types */ - assert_has_feature_enabled(qts, "max", "pmu"); assert_has_feature_enabled(qts, "cortex-a15", "pmu"); assert_has_not_feature(qts, "cortex-a15", "aarch64"); + /* Enabling and disabling pmu should always work. */ + assert_has_feature_enabled(qts, "max", "pmu"); + assert_set_feature(qts, "max", "pmu", false); + assert_set_feature(qts, "max", "pmu", true); + assert_has_not_feature(qts, "max", "kvm-no-adjvtime"); if (g_str_equal(qtest_get_arch(), "aarch64")) { @@ -XXX,XX +XXX,XX @@ static void test_query_cpu_model_expansion_kvm(const void *data) return; } + /* Enabling and disabling kvm-no-adjvtime should always work. */ assert_has_feature_disabled(qts, "host", "kvm-no-adjvtime"); + assert_set_feature(qts, "host", "kvm-no-adjvtime", true); + assert_set_feature(qts, "host", "kvm-no-adjvtime", false); if (g_str_equal(qtest_get_arch(), "aarch64")) { bool kvm_supports_sve; @@ -XXX,XX +XXX,XX @@ static void test_query_cpu_model_expansion_kvm(const void *data) char *error; assert_has_feature_enabled(qts, "host", "aarch64"); + + /* Enabling and disabling pmu should always work. */ assert_has_feature_enabled(qts, "host", "pmu"); + assert_set_feature(qts, "host", "pmu", false); + assert_set_feature(qts, "host", "pmu", true); assert_error(qts, "cortex-a15", "We cannot guarantee the CPU type 'cortex-a15' works " -- 2.20.1 From: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com> This adds support for memory(pc-dimm) hot remove on arm/virt that uses acpi ged device. NVDIMM hot removal is not yet supported. Signed-off-by: Shameer Kolothum <shameerali.kolothum.thodi@huawei.com> Message-id: 20200622124157.20360-1-shameerali.kolothum.thodi@huawei.com Reviewed-by: Eric Auger <eric.auger@redhat.com> Tested-by: Eric Auger <eric.auger@redhat.com> Signed-off-by: Peter Maydell <peter.maydell@linaro.org> --- hw/acpi/generic_event_device.c | 29 ++++++++++++++++ hw/arm/virt.c | 62 ++++++++++++++++++++++++++++++++-- 2 files changed, 89 insertions(+), 2 deletions(-) diff --git a/hw/acpi/generic_event_device.c b/hw/acpi/generic_event_device.c index XXXXXXX..XXXXXXX 100644 --- a/hw/acpi/generic_event_device.c +++ b/hw/acpi/generic_event_device.c @@ -XXX,XX +XXX,XX @@ static void acpi_ged_device_plug_cb(HotplugHandler *hotplug_dev, } } +static void acpi_ged_unplug_request_cb(HotplugHandler *hotplug_dev, + DeviceState *dev, Error **errp) +{ + AcpiGedState *s = ACPI_GED(hotplug_dev); + + if ((object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM) && + !(object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM)))) { + acpi_memory_unplug_request_cb(hotplug_dev, &s->memhp_state, dev, errp); + } else { + error_setg(errp, "acpi: device unplug request for unsupported device" + " type: %s", object_get_typename(OBJECT(dev))); + } +} + +static void acpi_ged_unplug_cb(HotplugHandler *hotplug_dev, + DeviceState *dev, Error **errp) +{ + AcpiGedState *s = ACPI_GED(hotplug_dev); + + if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) { + acpi_memory_unplug_cb(&s->memhp_state, dev, errp); + } else { + error_setg(errp, "acpi: device unplug for unsupported device" + " type: %s", object_get_typename(OBJECT(dev))); + } +} + static void acpi_ged_send_event(AcpiDeviceIf *adev, AcpiEventStatusBits ev) { AcpiGedState *s = ACPI_GED(adev); @@ -XXX,XX +XXX,XX @@ static void acpi_ged_class_init(ObjectClass *class, void *data) dc->vmsd = &vmstate_acpi_ged; hc->plug = acpi_ged_device_plug_cb; + hc->unplug_request = acpi_ged_unplug_request_cb; + hc->unplug = acpi_ged_unplug_cb; adevc->send_event = acpi_ged_send_event; } 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_machine_device_plug_cb(HotplugHandler *hotplug_dev, } } +static void virt_dimm_unplug_request(HotplugHandler *hotplug_dev, + DeviceState *dev, Error **errp) +{ + VirtMachineState *vms = VIRT_MACHINE(hotplug_dev); + Error *local_err = NULL; + + if (!vms->acpi_dev) { + error_setg(&local_err, + "memory hotplug is not enabled: missing acpi-ged device"); + goto out; + } + + if (object_dynamic_cast(OBJECT(dev), TYPE_NVDIMM)) { + error_setg(&local_err, + "nvdimm device hot unplug is not supported yet."); + goto out; + } + + hotplug_handler_unplug_request(HOTPLUG_HANDLER(vms->acpi_dev), dev, + &local_err); +out: + error_propagate(errp, local_err); +} + +static void virt_dimm_unplug(HotplugHandler *hotplug_dev, + DeviceState *dev, Error **errp) +{ + VirtMachineState *vms = VIRT_MACHINE(hotplug_dev); + Error *local_err = NULL; + + hotplug_handler_unplug(HOTPLUG_HANDLER(vms->acpi_dev), dev, &local_err); + if (local_err) { + goto out; + } + + pc_dimm_unplug(PC_DIMM(dev), MACHINE(vms)); + qdev_unrealize(dev); + +out: + error_propagate(errp, local_err); +} + static void virt_machine_device_unplug_request_cb(HotplugHandler *hotplug_dev, DeviceState *dev, Error **errp) { - error_setg(errp, "device unplug request for unsupported device" - " type: %s", object_get_typename(OBJECT(dev))); + if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) { + virt_dimm_unplug_request(hotplug_dev, dev, errp); + } else { + error_setg(errp, "device unplug request for unsupported device" + " type: %s", object_get_typename(OBJECT(dev))); + } +} + +static void virt_machine_device_unplug_cb(HotplugHandler *hotplug_dev, + DeviceState *dev, Error **errp) +{ + if (object_dynamic_cast(OBJECT(dev), TYPE_PC_DIMM)) { + virt_dimm_unplug(hotplug_dev, dev, errp); + } else { + error_setg(errp, "virt: device unplug for unsupported device" + " type: %s", object_get_typename(OBJECT(dev))); + } } static HotplugHandler *virt_machine_get_hotplug_handler(MachineState *machine, @@ -XXX,XX +XXX,XX @@ static void virt_machine_class_init(ObjectClass *oc, void *data) hc->pre_plug = virt_machine_device_pre_plug_cb; hc->plug = virt_machine_device_plug_cb; hc->unplug_request = virt_machine_device_unplug_request_cb; + hc->unplug = virt_machine_device_unplug_cb; mc->numa_mem_supported = true; mc->nvdimm_supported = true; mc->auto_enable_numa_with_memhp = true; -- 2.20.1