The following changes since commit e93ded1bf6c94ab95015b33e188bc8b0b0c32670: Merge tag 'testing-pull-request-2022-08-30' of https://gitlab.com/thuth/qemu into staging (2022-08-31 18:19:03 -0400) are available in the Git repository at: https://gitlab.com/rth7680/qemu.git tags/pull-tcg-20220901 for you to fetch changes up to 20011be2e30b8aa8ef1fc258485f00c688703deb: target/riscv: Make translator stop before the end of a page (2022-09-01 07:43:08 +0100) ---------------------------------------------------------------- Respect PROT_EXEC in user-only mode. Fix s390x, i386 and riscv for translations crossing a page. ---------------------------------------------------------------- Ilya Leoshkevich (4): linux-user: Clear translations on mprotect() accel/tcg: Introduce is_same_page() target/s390x: Make translator stop before the end of a page target/i386: Make translator stop before the end of a page Richard Henderson (16): linux-user/arm: Mark the commpage executable linux-user/hppa: Allocate page zero as a commpage linux-user/x86_64: Allocate vsyscall page as a commpage linux-user: Honor PT_GNU_STACK tests/tcg/i386: Move smc_code2 to an executable section accel/tcg: Properly implement get_page_addr_code for user-only accel/tcg: Unlock mmap_lock after longjmp accel/tcg: Make tb_htable_lookup static accel/tcg: Move qemu_ram_addr_from_host_nofail to physmem.c accel/tcg: Use probe_access_internal for softmmu get_page_addr_code_hostp accel/tcg: Document the faulting lookup in tb_lookup_cmp accel/tcg: Remove translator_ldsw accel/tcg: Add pc and host_pc params to gen_intermediate_code accel/tcg: Add fast path for translator_ld* target/riscv: Add MAX_INSN_LEN and insn_len target/riscv: Make translator stop before the end of a page include/elf.h | 1 + include/exec/cpu-common.h | 1 + include/exec/exec-all.h | 89 ++++++++---------------- include/exec/translator.h | 96 ++++++++++++++++--------- linux-user/arm/target_cpu.h | 4 +- linux-user/qemu.h | 1 + accel/tcg/cpu-exec.c | 143 ++++++++++++++++++++------------------ accel/tcg/cputlb.c | 93 +++++++------------------ accel/tcg/translate-all.c | 29 ++++---- accel/tcg/translator.c | 135 ++++++++++++++++++++++++++--------- accel/tcg/user-exec.c | 17 ++++- linux-user/elfload.c | 82 ++++++++++++++++++++-- linux-user/mmap.c | 6 +- softmmu/physmem.c | 12 ++++ target/alpha/translate.c | 5 +- target/arm/translate.c | 5 +- target/avr/translate.c | 5 +- target/cris/translate.c | 5 +- target/hexagon/translate.c | 6 +- target/hppa/translate.c | 5 +- target/i386/tcg/translate.c | 71 +++++++++++-------- target/loongarch/translate.c | 6 +- target/m68k/translate.c | 5 +- target/microblaze/translate.c | 5 +- target/mips/tcg/translate.c | 5 +- target/nios2/translate.c | 5 +- target/openrisc/translate.c | 6 +- target/ppc/translate.c | 5 +- target/riscv/translate.c | 32 +++++++-- target/rx/translate.c | 5 +- target/s390x/tcg/translate.c | 20 ++++-- target/sh4/translate.c | 5 +- target/sparc/translate.c | 5 +- target/tricore/translate.c | 6 +- target/xtensa/translate.c | 6 +- tests/tcg/i386/test-i386.c | 2 +- tests/tcg/riscv64/noexec.c | 79 +++++++++++++++++++++ tests/tcg/s390x/noexec.c | 106 ++++++++++++++++++++++++++++ tests/tcg/x86_64/noexec.c | 75 ++++++++++++++++++++ tests/tcg/multiarch/noexec.c.inc | 139 ++++++++++++++++++++++++++++++++++++ tests/tcg/riscv64/Makefile.target | 1 + tests/tcg/s390x/Makefile.target | 1 + tests/tcg/x86_64/Makefile.target | 3 +- 43 files changed, 966 insertions(+), 367 deletions(-) create mode 100644 tests/tcg/riscv64/noexec.c create mode 100644 tests/tcg/s390x/noexec.c create mode 100644 tests/tcg/x86_64/noexec.c create mode 100644 tests/tcg/multiarch/noexec.c.inc We're about to start validating PAGE_EXEC, which means that we've got to mark the commpage executable. We had been placing the commpage outside of reserved_va, which was incorrect and lead to an abort. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- linux-user/arm/target_cpu.h | 4 ++-- linux-user/elfload.c | 6 +++++- 2 files changed, 7 insertions(+), 3 deletions(-) diff --git a/linux-user/arm/target_cpu.h b/linux-user/arm/target_cpu.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/arm/target_cpu.h +++ b/linux-user/arm/target_cpu.h @@ -XXX,XX +XXX,XX @@ static inline unsigned long arm_max_reserved_va(CPUState *cs) } else { /* * We need to be able to map the commpage. - * See validate_guest_space in linux-user/elfload.c. + * See init_guest_commpage in linux-user/elfload.c. */ - return 0xffff0000ul; + return 0xfffffffful; } } #define MAX_RESERVED_VA arm_max_reserved_va diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ enum { static bool init_guest_commpage(void) { - void *want = g2h_untagged(HI_COMMPAGE & -qemu_host_page_size); + abi_ptr commpage = HI_COMMPAGE & -qemu_host_page_size; + void *want = g2h_untagged(commpage); void *addr = mmap(want, qemu_host_page_size, PROT_READ | PROT_WRITE, MAP_ANONYMOUS | MAP_PRIVATE | MAP_FIXED, -1, 0); @@ -XXX,XX +XXX,XX @@ static bool init_guest_commpage(void) perror("Protecting guest commpage"); exit(EXIT_FAILURE); } + + page_set_flags(commpage, commpage + qemu_host_page_size, + PAGE_READ | PAGE_EXEC | PAGE_VALID); return true; } -- 2.34.1 While there are no target-specific nonfaulting probes, generic code may grow some uses at some point. Note that the attrs argument was incorrect -- it should have been MEMTXATTRS_UNSPECIFIED. Just use the simpler interface. Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/avr/helper.c | 46 ++++++++++++++++++++++++++++----------------- 1 file changed, 29 insertions(+), 17 deletions(-) diff --git a/target/avr/helper.c b/target/avr/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/helper.c +++ b/target/avr/helper.c @@ -XXX,XX +XXX,XX @@ bool avr_cpu_tlb_fill(CPUState *cs, vaddr address, int size, MMUAccessType access_type, int mmu_idx, bool probe, uintptr_t retaddr) { - int prot = 0; - MemTxAttrs attrs = {}; + int prot, page_size = TARGET_PAGE_SIZE; uint32_t paddr; address &= TARGET_PAGE_MASK; if (mmu_idx == MMU_CODE_IDX) { - /* access to code in flash */ + /* Access to code in flash. */ paddr = OFFSET_CODE + address; prot = PAGE_READ | PAGE_EXEC; - if (paddr + TARGET_PAGE_SIZE > OFFSET_DATA) { + if (paddr >= OFFSET_DATA) { + /* + * This should not be possible via any architectural operations. + * There is certainly not an exception that we can deliver. + * Accept probing that might come from generic code. + */ + if (probe) { + return false; + } error_report("execution left flash memory"); abort(); } - } else if (address < NUMBER_OF_CPU_REGISTERS + NUMBER_OF_IO_REGISTERS) { - /* - * access to CPU registers, exit and rebuilt this TB to use full access - * incase it touches specially handled registers like SREG or SP - */ - AVRCPU *cpu = AVR_CPU(cs); - CPUAVRState *env = &cpu->env; - env->fullacc = 1; - cpu_loop_exit_restore(cs, retaddr); } else { - /* access to memory. nothing special */ + /* Access to memory. */ paddr = OFFSET_DATA + address; prot = PAGE_READ | PAGE_WRITE; + if (address < NUMBER_OF_CPU_REGISTERS + NUMBER_OF_IO_REGISTERS) { + /* + * Access to CPU registers, exit and rebuilt this TB to use + * full access in case it touches specially handled registers + * like SREG or SP. For probing, set page_size = 1, in order + * to force tlb_fill to be called for the next access. + */ + if (probe) { + page_size = 1; + } else { + AVRCPU *cpu = AVR_CPU(cs); + CPUAVRState *env = &cpu->env; + env->fullacc = 1; + cpu_loop_exit_restore(cs, retaddr); + } + } } - tlb_set_page_with_attrs(cs, address, paddr, attrs, prot, - mmu_idx, TARGET_PAGE_SIZE); - + tlb_set_page(cs, address, paddr, prot, mmu_idx, page_size); return true; } -- 2.34.1 There is no need to go through cc->tcg_ops when we know what value that must have. Reviewed-by: Michael Rolnik <mrolnik@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/avr/helper.c | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/target/avr/helper.c b/target/avr/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/helper.c +++ b/target/avr/helper.c @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) { bool ret = false; - CPUClass *cc = CPU_GET_CLASS(cs); AVRCPU *cpu = AVR_CPU(cs); CPUAVRState *env = &cpu->env; if (interrupt_request & CPU_INTERRUPT_RESET) { if (cpu_interrupts_enabled(env)) { cs->exception_index = EXCP_RESET; - cc->tcg_ops->do_interrupt(cs); + avr_cpu_do_interrupt(cs); cs->interrupt_request &= ~CPU_INTERRUPT_RESET; @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) if (cpu_interrupts_enabled(env) && env->intsrc != 0) { int index = ctz32(env->intsrc); cs->exception_index = EXCP_INT(index); - cc->tcg_ops->do_interrupt(cs); + avr_cpu_do_interrupt(cs); env->intsrc &= env->intsrc - 1; /* clear the interrupt */ if (!env->intsrc) { -- 2.34.1 We're about to start validating PAGE_EXEC, which means that we've got to mark page zero executable. We had been special casing this entirely within translate. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- linux-user/elfload.c | 34 +++++++++++++++++++++++++++++++--- 1 file changed, 31 insertions(+), 3 deletions(-) diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ static inline void init_thread(struct target_pt_regs *regs, regs->gr[31] = infop->entry; } +#define LO_COMMPAGE 0 + +static bool init_guest_commpage(void) +{ + void *want = g2h_untagged(LO_COMMPAGE); + void *addr = mmap(want, qemu_host_page_size, PROT_NONE, + MAP_ANONYMOUS | MAP_PRIVATE | MAP_FIXED, -1, 0); + + if (addr == MAP_FAILED) { + perror("Allocating guest commpage"); + exit(EXIT_FAILURE); + } + if (addr != want) { + return false; + } + + /* + * On Linux, page zero is normally marked execute only + gateway. + * Normal read or write is supposed to fail (thus PROT_NONE above), + * but specific offsets have kernel code mapped to raise permissions + * and implement syscalls. Here, simply mark the page executable. + * Special case the entry points during translation (see do_page_zero). + */ + page_set_flags(LO_COMMPAGE, LO_COMMPAGE + TARGET_PAGE_SIZE, + PAGE_EXEC | PAGE_VALID); + return true; +} + #endif /* TARGET_HPPA */ #ifdef TARGET_XTENSA @@ -XXX,XX +XXX,XX @@ static abi_ulong create_elf_tables(abi_ulong p, int argc, int envc, } #if defined(HI_COMMPAGE) -#define LO_COMMPAGE 0 +#define LO_COMMPAGE -1 #elif defined(LO_COMMPAGE) #define HI_COMMPAGE 0 #else #define HI_COMMPAGE 0 -#define LO_COMMPAGE 0 +#define LO_COMMPAGE -1 #define init_guest_commpage() true #endif @@ -XXX,XX +XXX,XX @@ static void pgb_static(const char *image_name, abi_ulong orig_loaddr, } else { offset = -(HI_COMMPAGE & -align); } - } else if (LO_COMMPAGE != 0) { + } else if (LO_COMMPAGE != -1) { loaddr = MIN(loaddr, LO_COMMPAGE & -align); } -- 2.34.1 We're about to start validating PAGE_EXEC, which means that we've got to mark the vsyscall page executable. We had been special casing this entirely within translate. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- linux-user/elfload.c | 23 +++++++++++++++++++++++ 1 file changed, 23 insertions(+) diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ static void elf_core_copy_regs(target_elf_gregset_t *regs, const CPUX86State *en (*regs)[26] = tswapreg(env->segs[R_GS].selector & 0xffff); } +#if ULONG_MAX >= TARGET_VSYSCALL_PAGE +#define INIT_GUEST_COMMPAGE +static bool init_guest_commpage(void) +{ + /* + * The vsyscall page is at a high negative address aka kernel space, + * which means that we cannot actually allocate it with target_mmap. + * We still should be able to use page_set_flags, unless the user + * has specified -R reserved_va, which would trigger an assert(). + */ + if (reserved_va != 0 && + TARGET_VSYSCALL_PAGE + TARGET_PAGE_SIZE >= reserved_va) { + error_report("Cannot allocate vsyscall page"); + exit(EXIT_FAILURE); + } + page_set_flags(TARGET_VSYSCALL_PAGE, + TARGET_VSYSCALL_PAGE + TARGET_PAGE_SIZE, + PAGE_EXEC | PAGE_VALID); + return true; +} +#endif #else #define ELF_START_MMAP 0x80000000 @@ -XXX,XX +XXX,XX @@ static abi_ulong create_elf_tables(abi_ulong p, int argc, int envc, #else #define HI_COMMPAGE 0 #define LO_COMMPAGE -1 +#ifndef INIT_GUEST_COMMPAGE #define init_guest_commpage() true #endif +#endif static void pgb_fail_in_use(const char *image_name) { -- 2.34.1 We cannot deliver two interrupts simultaneously; the first interrupt handler must execute first. Reviewed-by: Michael Rolnik <mrolnik@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/avr/helper.c | 9 +++------ 1 file changed, 3 insertions(+), 6 deletions(-) diff --git a/target/avr/helper.c b/target/avr/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/helper.c +++ b/target/avr/helper.c @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) { - bool ret = false; AVRCPU *cpu = AVR_CPU(cs); CPUAVRState *env = &cpu->env; @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) avr_cpu_do_interrupt(cs); cs->interrupt_request &= ~CPU_INTERRUPT_RESET; - - ret = true; + return true; } } if (interrupt_request & CPU_INTERRUPT_HARD) { @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) if (!env->intsrc) { cs->interrupt_request &= ~CPU_INTERRUPT_HARD; } - - ret = true; + return true; } } - return ret; + return false; } void avr_cpu_do_interrupt(CPUState *cs) -- 2.34.1 This bit is not saved across interrupts, so we must delay delivering the interrupt until the skip has been processed. Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1118 Reviewed-by: Michael Rolnik <mrolnik@gmail.com> Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/avr/helper.c | 9 +++++++++ target/avr/translate.c | 26 ++++++++++++++++++++++---- 2 files changed, 31 insertions(+), 4 deletions(-) diff --git a/target/avr/helper.c b/target/avr/helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/helper.c +++ b/target/avr/helper.c @@ -XXX,XX +XXX,XX @@ bool avr_cpu_exec_interrupt(CPUState *cs, int interrupt_request) AVRCPU *cpu = AVR_CPU(cs); CPUAVRState *env = &cpu->env; + /* + * We cannot separate a skip from the next instruction, + * as the skip would not be preserved across the interrupt. + * Separating the two insn normally only happens at page boundaries. + */ + if (env->skip) { + return false; + } + if (interrupt_request & CPU_INTERRUPT_RESET) { if (cpu_interrupts_enabled(env)) { cs->exception_index = EXCP_RESET; diff --git a/target/avr/translate.c b/target/avr/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/translate.c +++ b/target/avr/translate.c @@ -XXX,XX +XXX,XX @@ static void avr_tr_translate_insn(DisasContextBase *dcbase, CPUState *cs) if (skip_label) { canonicalize_skip(ctx); gen_set_label(skip_label); - if (ctx->base.is_jmp == DISAS_NORETURN) { + + switch (ctx->base.is_jmp) { + case DISAS_NORETURN: ctx->base.is_jmp = DISAS_CHAIN; + break; + case DISAS_NEXT: + if (ctx->base.tb->flags & TB_FLAGS_SKIP) { + ctx->base.is_jmp = DISAS_TOO_MANY; + } + break; + default: + break; } } @@ -XXX,XX +XXX,XX @@ static void avr_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) { DisasContext *ctx = container_of(dcbase, DisasContext, base); bool nonconst_skip = canonicalize_skip(ctx); + /* + * Because we disable interrupts while env->skip is set, + * we must return to the main loop to re-evaluate afterward. + */ + bool force_exit = ctx->base.tb->flags & TB_FLAGS_SKIP; switch (ctx->base.is_jmp) { case DISAS_NORETURN: @@ -XXX,XX +XXX,XX @@ static void avr_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) case DISAS_NEXT: case DISAS_TOO_MANY: case DISAS_CHAIN: - if (!nonconst_skip) { + if (!nonconst_skip && !force_exit) { /* Note gen_goto_tb checks singlestep. */ gen_goto_tb(ctx, 1, ctx->npc); break; @@ -XXX,XX +XXX,XX @@ static void avr_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) tcg_gen_movi_tl(cpu_pc, ctx->npc); /* fall through */ case DISAS_LOOKUP: - tcg_gen_lookup_and_goto_ptr(); - break; + if (!force_exit) { + tcg_gen_lookup_and_goto_ptr(); + break; + } + /* fall through */ case DISAS_EXIT: tcg_gen_exit_tb(NULL, 0); break; -- 2.34.1 Map the stack executable if required by default or on demand. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/elf.h | 1 + linux-user/qemu.h | 1 + linux-user/elfload.c | 19 ++++++++++++++++++- 3 files changed, 20 insertions(+), 1 deletion(-) diff --git a/include/elf.h b/include/elf.h index XXXXXXX..XXXXXXX 100644 --- a/include/elf.h +++ b/include/elf.h @@ -XXX,XX +XXX,XX @@ typedef int64_t Elf64_Sxword; #define PT_LOPROC 0x70000000 #define PT_HIPROC 0x7fffffff +#define PT_GNU_STACK (PT_LOOS + 0x474e551) #define PT_GNU_PROPERTY (PT_LOOS + 0x474e553) #define PT_MIPS_REGINFO 0x70000000 diff --git a/linux-user/qemu.h b/linux-user/qemu.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/qemu.h +++ b/linux-user/qemu.h @@ -XXX,XX +XXX,XX @@ struct image_info { uint32_t elf_flags; int personality; abi_ulong alignment; + bool exec_stack; /* Generic semihosting knows about these pointers. */ abi_ulong arg_strings; /* strings for argv */ diff --git a/linux-user/elfload.c b/linux-user/elfload.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/elfload.c +++ b/linux-user/elfload.c @@ -XXX,XX +XXX,XX @@ static bool init_guest_commpage(void) #define ELF_ARCH EM_386 #define ELF_PLATFORM get_elf_platform() +#define EXSTACK_DEFAULT true static const char *get_elf_platform(void) { @@ -XXX,XX +XXX,XX @@ static void elf_core_copy_regs(target_elf_gregset_t *regs, const CPUX86State *en #define ELF_ARCH EM_ARM #define ELF_CLASS ELFCLASS32 +#define EXSTACK_DEFAULT true static inline void init_thread(struct target_pt_regs *regs, struct image_info *infop) @@ -XXX,XX +XXX,XX @@ static inline void init_thread(struct target_pt_regs *regs, #else #define ELF_CLASS ELFCLASS32 +#define EXSTACK_DEFAULT true #endif @@ -XXX,XX +XXX,XX @@ static void elf_core_copy_regs(target_elf_gregset_t *regs, const CPUPPCState *en #define ELF_CLASS ELFCLASS64 #define ELF_ARCH EM_LOONGARCH +#define EXSTACK_DEFAULT true #define elf_check_arch(x) ((x) == EM_LOONGARCH) @@ -XXX,XX +XXX,XX @@ static uint32_t get_elf_hwcap(void) #define ELF_CLASS ELFCLASS32 #endif #define ELF_ARCH EM_MIPS +#define EXSTACK_DEFAULT true #ifdef TARGET_ABI_MIPSN32 #define elf_check_abi(x) ((x) & EF_MIPS_ABI2) @@ -XXX,XX +XXX,XX @@ static inline void init_thread(struct target_pt_regs *regs, #define bswaptls(ptr) bswap32s(ptr) #endif +#ifndef EXSTACK_DEFAULT +#define EXSTACK_DEFAULT false +#endif + #include "elf.h" /* We must delay the following stanzas until after "elf.h". */ @@ -XXX,XX +XXX,XX @@ static abi_ulong setup_arg_pages(struct linux_binprm *bprm, struct image_info *info) { abi_ulong size, error, guard; + int prot; size = guest_stack_size; if (size < STACK_LOWER_LIMIT) { @@ -XXX,XX +XXX,XX @@ static abi_ulong setup_arg_pages(struct linux_binprm *bprm, guard = qemu_real_host_page_size(); } - error = target_mmap(0, size + guard, PROT_READ | PROT_WRITE, + prot = PROT_READ | PROT_WRITE; + if (info->exec_stack) { + prot |= PROT_EXEC; + } + error = target_mmap(0, size + guard, prot, MAP_PRIVATE | MAP_ANONYMOUS, -1, 0); if (error == -1) { perror("mmap stack"); @@ -XXX,XX +XXX,XX @@ static void load_elf_image(const char *image_name, int image_fd, */ loaddr = -1, hiaddr = 0; info->alignment = 0; + info->exec_stack = EXSTACK_DEFAULT; for (i = 0; i < ehdr->e_phnum; ++i) { struct elf_phdr *eppnt = phdr + i; if (eppnt->p_type == PT_LOAD) { @@ -XXX,XX +XXX,XX @@ static void load_elf_image(const char *image_name, int image_fd, if (!parse_elf_properties(image_fd, info, eppnt, bprm_buf, &err)) { goto exit_errmsg; } + } else if (eppnt->p_type == PT_GNU_STACK) { + info->exec_stack = eppnt->p_flags & PF_X; } } -- 2.34.1 From: Ilya Leoshkevich <iii@linux.ibm.com> Currently it's possible to execute pages that do not have PAGE_EXEC if there is an existing translation block. Fix by invalidating TBs that touch the affected pages. Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Message-Id: <20220817150506.592862-2-iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- linux-user/mmap.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/linux-user/mmap.c b/linux-user/mmap.c index XXXXXXX..XXXXXXX 100644 --- a/linux-user/mmap.c +++ b/linux-user/mmap.c @@ -XXX,XX +XXX,XX @@ int target_mprotect(abi_ulong start, abi_ulong len, int target_prot) goto error; } } + page_set_flags(start, start + len, page_flags); - mmap_unlock(); - return 0; + tb_invalidate_phys_range(start, start + len); + ret = 0; + error: mmap_unlock(); return ret; -- 2.34.1 We're about to start validating PAGE_EXEC, which means that we've got to put this code into a section that is both writable and executable. Note that this test did not run on hardware beforehand either. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tests/tcg/i386/test-i386.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/tcg/i386/test-i386.c b/tests/tcg/i386/test-i386.c index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/i386/test-i386.c +++ b/tests/tcg/i386/test-i386.c @@ -XXX,XX +XXX,XX @@ uint8_t code[] = { 0xc3, /* ret */ }; -asm(".section \".data\"\n" +asm(".section \".data_x\",\"awx\"\n" "smc_code2:\n" "movl 4(%esp), %eax\n" "movl %eax, smc_patch_addr2 + 1\n" -- 2.34.1 From: Ilya Leoshkevich <iii@linux.ibm.com> Introduce a function that checks whether a given address is on the same page as where disassembly started. Having it improves readability of the following patches. Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Message-Id: <20220811095534.241224-3-iii@linux.ibm.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> [rth: Make the DisasContextBase parameter const.] Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 10 ++++++++++ 1 file changed, 10 insertions(+) diff --git a/include/exec/translator.h b/include/exec/translator.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/translator.h +++ b/include/exec/translator.h @@ -XXX,XX +XXX,XX @@ FOR_EACH_TRANSLATOR_LD(GEN_TRANSLATOR_LD) #undef GEN_TRANSLATOR_LD +/* + * Return whether addr is on the same page as where disassembly started. + * Translators can use this to enforce the rule that only single-insn + * translation blocks are allowed to cross page boundaries. + */ +static inline bool is_same_page(const DisasContextBase *db, target_ulong addr) +{ + return ((addr ^ db->pc_first) & TARGET_PAGE_MASK) == 0; +} + #endif /* EXEC__TRANSLATOR_H */ -- 2.34.1 The current implementation is a no-op, simply returning addr. This is incorrect, because we ought to be checking the page permissions for execution. Make get_page_addr_code inline for both implementations. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Acked-by: Alistair Francis <alistair.francis@wdc.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/exec-all.h | 85 ++++++++++++++--------------------------- accel/tcg/cputlb.c | 5 --- accel/tcg/user-exec.c | 14 +++++++ 3 files changed, 42 insertions(+), 62 deletions(-) diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/exec-all.h +++ b/include/exec/exec-all.h @@ -XXX,XX +XXX,XX @@ struct MemoryRegionSection *iotlb_to_section(CPUState *cpu, hwaddr index, MemTxAttrs attrs); #endif -#if defined(CONFIG_USER_ONLY) -void mmap_lock(void); -void mmap_unlock(void); -bool have_mmap_lock(void); - /** - * get_page_addr_code() - user-mode version + * get_page_addr_code_hostp() * @env: CPUArchState * @addr: guest virtual address of guest code * - * Returns @addr. + * See get_page_addr_code() (full-system version) for documentation on the + * return value. + * + * Sets *@hostp (when @hostp is non-NULL) as follows. + * If the return value is -1, sets *@hostp to NULL. Otherwise, sets *@hostp + * to the host address where @addr's content is kept. + * + * Note: this function can trigger an exception. + */ +tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, + void **hostp); + +/** + * get_page_addr_code() + * @env: CPUArchState + * @addr: guest virtual address of guest code + * + * If we cannot translate and execute from the entire RAM page, or if + * the region is not backed by RAM, returns -1. Otherwise, returns the + * ram_addr_t corresponding to the guest code at @addr. + * + * Note: this function can trigger an exception. */ static inline tb_page_addr_t get_page_addr_code(CPUArchState *env, target_ulong addr) { - return addr; + return get_page_addr_code_hostp(env, addr, NULL); } -/** - * get_page_addr_code_hostp() - user-mode version - * @env: CPUArchState - * @addr: guest virtual address of guest code - * - * Returns @addr. - * - * If @hostp is non-NULL, sets *@hostp to the host address where @addr's content - * is kept. - */ -static inline tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, - target_ulong addr, - void **hostp) -{ - if (hostp) { - *hostp = g2h_untagged(addr); - } - return addr; -} +#if defined(CONFIG_USER_ONLY) +void mmap_lock(void); +void mmap_unlock(void); +bool have_mmap_lock(void); /** * adjust_signal_pc: @@ -XXX,XX +XXX,XX @@ G_NORETURN void cpu_loop_exit_sigbus(CPUState *cpu, target_ulong addr, static inline void mmap_lock(void) {} static inline void mmap_unlock(void) {} -/** - * get_page_addr_code() - full-system version - * @env: CPUArchState - * @addr: guest virtual address of guest code - * - * If we cannot translate and execute from the entire RAM page, or if - * the region is not backed by RAM, returns -1. Otherwise, returns the - * ram_addr_t corresponding to the guest code at @addr. - * - * Note: this function can trigger an exception. - */ -tb_page_addr_t get_page_addr_code(CPUArchState *env, target_ulong addr); - -/** - * get_page_addr_code_hostp() - full-system version - * @env: CPUArchState - * @addr: guest virtual address of guest code - * - * See get_page_addr_code() (full-system version) for documentation on the - * return value. - * - * Sets *@hostp (when @hostp is non-NULL) as follows. - * If the return value is -1, sets *@hostp to NULL. Otherwise, sets *@hostp - * to the host address where @addr's content is kept. - * - * Note: this function can trigger an exception. - */ -tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, - void **hostp); - void tlb_reset_dirty(CPUState *cpu, ram_addr_t start1, ram_addr_t length); void tlb_set_dirty(CPUState *cpu, target_ulong vaddr); diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cputlb.c +++ b/accel/tcg/cputlb.c @@ -XXX,XX +XXX,XX @@ tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, return qemu_ram_addr_from_host_nofail(p); } -tb_page_addr_t get_page_addr_code(CPUArchState *env, target_ulong addr) -{ - return get_page_addr_code_hostp(env, addr, NULL); -} - static void notdirty_write(CPUState *cpu, vaddr mem_vaddr, unsigned size, CPUIOTLBEntry *iotlbentry, uintptr_t retaddr) { diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/user-exec.c +++ b/accel/tcg/user-exec.c @@ -XXX,XX +XXX,XX @@ void *probe_access(CPUArchState *env, target_ulong addr, int size, return size ? g2h(env_cpu(env), addr) : NULL; } +tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, + void **hostp) +{ + int flags; + + flags = probe_access_internal(env, addr, 1, MMU_INST_FETCH, false, 0); + g_assert(flags == 0); + + if (hostp) { + *hostp = g2h_untagged(addr); + } + return addr; +} + /* The softmmu versions of these helpers are in cputlb.c. */ /* -- 2.34.1 The mmap_lock is held around tb_gen_code. While the comment is correct that the lock is dropped when tb_gen_code runs out of memory, the lock is *not* dropped when an exception is raised reading code for translation. Acked-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/cpu-exec.c | 12 ++++++------ accel/tcg/user-exec.c | 3 --- 2 files changed, 6 insertions(+), 9 deletions(-) diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cpu-exec.c +++ b/accel/tcg/cpu-exec.c @@ -XXX,XX +XXX,XX @@ void cpu_exec_step_atomic(CPUState *cpu) cpu_tb_exec(cpu, tb, &tb_exit); cpu_exec_exit(cpu); } else { - /* - * The mmap_lock is dropped by tb_gen_code if it runs out of - * memory. - */ #ifndef CONFIG_SOFTMMU clear_helper_retaddr(); - tcg_debug_assert(!have_mmap_lock()); + if (have_mmap_lock()) { + mmap_unlock(); + } #endif if (qemu_mutex_iothread_locked()) { qemu_mutex_unlock_iothread(); @@ -XXX,XX +XXX,XX @@ int cpu_exec(CPUState *cpu) #ifndef CONFIG_SOFTMMU clear_helper_retaddr(); - tcg_debug_assert(!have_mmap_lock()); + if (have_mmap_lock()) { + mmap_unlock(); + } #endif if (qemu_mutex_iothread_locked()) { qemu_mutex_unlock_iothread(); diff --git a/accel/tcg/user-exec.c b/accel/tcg/user-exec.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/user-exec.c +++ b/accel/tcg/user-exec.c @@ -XXX,XX +XXX,XX @@ MMUAccessType adjust_signal_pc(uintptr_t *pc, bool is_write) * (and if the translator doesn't handle page boundaries correctly * there's little we can do about that here). Therefore, do not * trigger the unwinder. - * - * Like tb_gen_code, release the memory lock before cpu_loop_exit. */ - mmap_unlock(); *pc = 0; return MMU_INST_FETCH; } -- 2.34.1 The function is not used outside of cpu-exec.c. Move it and its subroutines up in the file, before the first use. Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/exec-all.h | 3 - accel/tcg/cpu-exec.c | 122 ++++++++++++++++++++-------------------- 2 files changed, 61 insertions(+), 64 deletions(-) diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/exec-all.h +++ b/include/exec/exec-all.h @@ -XXX,XX +XXX,XX @@ void tb_invalidate_phys_addr(AddressSpace *as, hwaddr addr, MemTxAttrs attrs); #endif void tb_flush(CPUState *cpu); void tb_phys_invalidate(TranslationBlock *tb, tb_page_addr_t page_addr); -TranslationBlock *tb_htable_lookup(CPUState *cpu, target_ulong pc, - target_ulong cs_base, uint32_t flags, - uint32_t cflags); void tb_set_jmp_target(TranslationBlock *tb, int n, uintptr_t addr); /* GETPC is the true target of the return instruction that we'll execute. */ diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cpu-exec.c +++ b/accel/tcg/cpu-exec.c @@ -XXX,XX +XXX,XX @@ uint32_t curr_cflags(CPUState *cpu) return cflags; } +struct tb_desc { + target_ulong pc; + target_ulong cs_base; + CPUArchState *env; + tb_page_addr_t phys_page1; + uint32_t flags; + uint32_t cflags; + uint32_t trace_vcpu_dstate; +}; + +static bool tb_lookup_cmp(const void *p, const void *d) +{ + const TranslationBlock *tb = p; + const struct tb_desc *desc = d; + + if (tb->pc == desc->pc && + tb->page_addr[0] == desc->phys_page1 && + tb->cs_base == desc->cs_base && + tb->flags == desc->flags && + tb->trace_vcpu_dstate == desc->trace_vcpu_dstate && + tb_cflags(tb) == desc->cflags) { + /* check next page if needed */ + if (tb->page_addr[1] == -1) { + return true; + } else { + tb_page_addr_t phys_page2; + target_ulong virt_page2; + + virt_page2 = (desc->pc & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE; + phys_page2 = get_page_addr_code(desc->env, virt_page2); + if (tb->page_addr[1] == phys_page2) { + return true; + } + } + } + return false; +} + +static TranslationBlock *tb_htable_lookup(CPUState *cpu, target_ulong pc, + target_ulong cs_base, uint32_t flags, + uint32_t cflags) +{ + tb_page_addr_t phys_pc; + struct tb_desc desc; + uint32_t h; + + desc.env = cpu->env_ptr; + desc.cs_base = cs_base; + desc.flags = flags; + desc.cflags = cflags; + desc.trace_vcpu_dstate = *cpu->trace_dstate; + desc.pc = pc; + phys_pc = get_page_addr_code(desc.env, pc); + if (phys_pc == -1) { + return NULL; + } + desc.phys_page1 = phys_pc & TARGET_PAGE_MASK; + h = tb_hash_func(phys_pc, pc, flags, cflags, *cpu->trace_dstate); + return qht_lookup_custom(&tb_ctx.htable, &desc, h, tb_lookup_cmp); +} + /* Might cause an exception, so have a longjmp destination ready */ static inline TranslationBlock *tb_lookup(CPUState *cpu, target_ulong pc, target_ulong cs_base, @@ -XXX,XX +XXX,XX @@ void cpu_exec_step_atomic(CPUState *cpu) end_exclusive(); } -struct tb_desc { - target_ulong pc; - target_ulong cs_base; - CPUArchState *env; - tb_page_addr_t phys_page1; - uint32_t flags; - uint32_t cflags; - uint32_t trace_vcpu_dstate; -}; - -static bool tb_lookup_cmp(const void *p, const void *d) -{ - const TranslationBlock *tb = p; - const struct tb_desc *desc = d; - - if (tb->pc == desc->pc && - tb->page_addr[0] == desc->phys_page1 && - tb->cs_base == desc->cs_base && - tb->flags == desc->flags && - tb->trace_vcpu_dstate == desc->trace_vcpu_dstate && - tb_cflags(tb) == desc->cflags) { - /* check next page if needed */ - if (tb->page_addr[1] == -1) { - return true; - } else { - tb_page_addr_t phys_page2; - target_ulong virt_page2; - - virt_page2 = (desc->pc & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE; - phys_page2 = get_page_addr_code(desc->env, virt_page2); - if (tb->page_addr[1] == phys_page2) { - return true; - } - } - } - return false; -} - -TranslationBlock *tb_htable_lookup(CPUState *cpu, target_ulong pc, - target_ulong cs_base, uint32_t flags, - uint32_t cflags) -{ - tb_page_addr_t phys_pc; - struct tb_desc desc; - uint32_t h; - - desc.env = cpu->env_ptr; - desc.cs_base = cs_base; - desc.flags = flags; - desc.cflags = cflags; - desc.trace_vcpu_dstate = *cpu->trace_dstate; - desc.pc = pc; - phys_pc = get_page_addr_code(desc.env, pc); - if (phys_pc == -1) { - return NULL; - } - desc.phys_page1 = phys_pc & TARGET_PAGE_MASK; - h = tb_hash_func(phys_pc, pc, flags, cflags, *cpu->trace_dstate); - return qht_lookup_custom(&tb_ctx.htable, &desc, h, tb_lookup_cmp); -} - void tb_set_jmp_target(TranslationBlock *tb, int n, uintptr_t addr) { if (TCG_TARGET_HAS_direct_jump) { -- 2.34.1 The base qemu_ram_addr_from_host function is already in softmmu/physmem.c; move the nofail version to be adjacent. Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/cpu-common.h | 1 + accel/tcg/cputlb.c | 12 ------------ softmmu/physmem.c | 12 ++++++++++++ 3 files changed, 13 insertions(+), 12 deletions(-) diff --git a/include/exec/cpu-common.h b/include/exec/cpu-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/cpu-common.h +++ b/include/exec/cpu-common.h @@ -XXX,XX +XXX,XX @@ typedef uintptr_t ram_addr_t; void qemu_ram_remap(ram_addr_t addr, ram_addr_t length); /* This should not be used by devices. */ ram_addr_t qemu_ram_addr_from_host(void *ptr); +ram_addr_t qemu_ram_addr_from_host_nofail(void *ptr); RAMBlock *qemu_ram_block_by_name(const char *name); RAMBlock *qemu_ram_block_from_host(void *ptr, bool round_offset, ram_addr_t *offset); diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cputlb.c +++ b/accel/tcg/cputlb.c @@ -XXX,XX +XXX,XX @@ void tlb_set_page(CPUState *cpu, target_ulong vaddr, prot, mmu_idx, size); } -static inline ram_addr_t qemu_ram_addr_from_host_nofail(void *ptr) -{ - ram_addr_t ram_addr; - - ram_addr = qemu_ram_addr_from_host(ptr); - if (ram_addr == RAM_ADDR_INVALID) { - error_report("Bad ram pointer %p", ptr); - abort(); - } - return ram_addr; -} - /* * Note: tlb_fill() can trigger a resize of the TLB. This means that all of the * caller's prior references to the TLB table (e.g. CPUTLBEntry pointers) must diff --git a/softmmu/physmem.c b/softmmu/physmem.c index XXXXXXX..XXXXXXX 100644 --- a/softmmu/physmem.c +++ b/softmmu/physmem.c @@ -XXX,XX +XXX,XX @@ ram_addr_t qemu_ram_addr_from_host(void *ptr) return block->offset + offset; } +ram_addr_t qemu_ram_addr_from_host_nofail(void *ptr) +{ + ram_addr_t ram_addr; + + ram_addr = qemu_ram_addr_from_host(ptr); + if (ram_addr == RAM_ADDR_INVALID) { + error_report("Bad ram pointer %p", ptr); + abort(); + } + return ram_addr; +} + static MemTxResult flatview_read(FlatView *fv, hwaddr addr, MemTxAttrs attrs, void *buf, hwaddr len); static MemTxResult flatview_write(FlatView *fv, hwaddr addr, MemTxAttrs attrs, -- 2.34.1 Simplify the implementation of get_page_addr_code_hostp by reusing the existing probe_access infrastructure. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/cputlb.c | 76 ++++++++++++++++------------------------------ 1 file changed, 26 insertions(+), 50 deletions(-) diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cputlb.c +++ b/accel/tcg/cputlb.c @@ -XXX,XX +XXX,XX @@ static bool victim_tlb_hit(CPUArchState *env, size_t mmu_idx, size_t index, victim_tlb_hit(env, mmu_idx, index, offsetof(CPUTLBEntry, TY), \ (ADDR) & TARGET_PAGE_MASK) -/* - * Return a ram_addr_t for the virtual address for execution. - * - * Return -1 if we can't translate and execute from an entire page - * of RAM. This will force us to execute by loading and translating - * one insn at a time, without caching. - * - * NOTE: This function will trigger an exception if the page is - * not executable. - */ -tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, - void **hostp) -{ - uintptr_t mmu_idx = cpu_mmu_index(env, true); - uintptr_t index = tlb_index(env, mmu_idx, addr); - CPUTLBEntry *entry = tlb_entry(env, mmu_idx, addr); - void *p; - - if (unlikely(!tlb_hit(entry->addr_code, addr))) { - if (!VICTIM_TLB_HIT(addr_code, addr)) { - tlb_fill(env_cpu(env), addr, 0, MMU_INST_FETCH, mmu_idx, 0); - index = tlb_index(env, mmu_idx, addr); - entry = tlb_entry(env, mmu_idx, addr); - - if (unlikely(entry->addr_code & TLB_INVALID_MASK)) { - /* - * The MMU protection covers a smaller range than a target - * page, so we must redo the MMU check for every insn. - */ - return -1; - } - } - assert(tlb_hit(entry->addr_code, addr)); - } - - if (unlikely(entry->addr_code & TLB_MMIO)) { - /* The region is not backed by RAM. */ - if (hostp) { - *hostp = NULL; - } - return -1; - } - - p = (void *)((uintptr_t)addr + entry->addend); - if (hostp) { - *hostp = p; - } - return qemu_ram_addr_from_host_nofail(p); -} - static void notdirty_write(CPUState *cpu, vaddr mem_vaddr, unsigned size, CPUIOTLBEntry *iotlbentry, uintptr_t retaddr) { @@ -XXX,XX +XXX,XX @@ void *tlb_vaddr_to_host(CPUArchState *env, abi_ptr addr, return flags ? NULL : host; } +/* + * Return a ram_addr_t for the virtual address for execution. + * + * Return -1 if we can't translate and execute from an entire page + * of RAM. This will force us to execute by loading and translating + * one insn at a time, without caching. + * + * NOTE: This function will trigger an exception if the page is + * not executable. + */ +tb_page_addr_t get_page_addr_code_hostp(CPUArchState *env, target_ulong addr, + void **hostp) +{ + void *p; + + (void)probe_access_internal(env, addr, 1, MMU_INST_FETCH, + cpu_mmu_index(env, true), false, &p, 0); + if (p == NULL) { + return -1; + } + if (hostp) { + *hostp = p; + } + return qemu_ram_addr_from_host_nofail(p); +} + #ifdef CONFIG_PLUGIN /* * Perform a TLB lookup and populate the qemu_plugin_hwaddr structure. -- 2.34.1 It was non-obvious to me why we can raise an exception in the middle of a comparison function, but it works. While nearby, use TARGET_PAGE_ALIGN instead of open-coding. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/cpu-exec.c | 11 ++++++++++- 1 file changed, 10 insertions(+), 1 deletion(-) diff --git a/accel/tcg/cpu-exec.c b/accel/tcg/cpu-exec.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cpu-exec.c +++ b/accel/tcg/cpu-exec.c @@ -XXX,XX +XXX,XX @@ static bool tb_lookup_cmp(const void *p, const void *d) tb_page_addr_t phys_page2; target_ulong virt_page2; - virt_page2 = (desc->pc & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE; + /* + * We know that the first page matched, and an otherwise valid TB + * encountered an incomplete instruction at the end of that page, + * therefore we know that generating a new TB from the current PC + * must also require reading from the next page -- even if the + * second pages do not match, and therefore the resulting insn + * is different for the new TB. Therefore any exception raised + * here by the faulting lookup is not premature. + */ + virt_page2 = TARGET_PAGE_ALIGN(desc->pc); phys_page2 = get_page_addr_code(desc->env, virt_page2); if (tb->page_addr[1] == phys_page2) { return true; -- 2.34.1 The only user can easily use translator_lduw and adjust the type to signed during the return. Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 1 - target/i386/tcg/translate.c | 2 +- 2 files changed, 1 insertion(+), 2 deletions(-) diff --git a/include/exec/translator.h b/include/exec/translator.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/translator.h +++ b/include/exec/translator.h @@ -XXX,XX +XXX,XX @@ bool translator_use_goto_tb(DisasContextBase *db, target_ulong dest); #define FOR_EACH_TRANSLATOR_LD(F) \ F(translator_ldub, uint8_t, cpu_ldub_code, /* no swap */) \ - F(translator_ldsw, int16_t, cpu_ldsw_code, bswap16) \ F(translator_lduw, uint16_t, cpu_lduw_code, bswap16) \ F(translator_ldl, uint32_t, cpu_ldl_code, bswap32) \ F(translator_ldq, uint64_t, cpu_ldq_code, bswap64) diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/i386/tcg/translate.c +++ b/target/i386/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static inline uint8_t x86_ldub_code(CPUX86State *env, DisasContext *s) static inline int16_t x86_ldsw_code(CPUX86State *env, DisasContext *s) { - return translator_ldsw(env, &s->base, advance_pc(env, s, 2)); + return translator_lduw(env, &s->base, advance_pc(env, s, 2)); } static inline uint16_t x86_lduw_code(CPUX86State *env, DisasContext *s) -- 2.34.1 Pass these along to translator_loop -- pc may be used instead of tb->pc, and host_pc is currently unused. Adjust all targets at one time. Acked-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/exec-all.h | 1 - include/exec/translator.h | 24 ++++++++++++++++++++---- accel/tcg/translate-all.c | 6 ++++-- accel/tcg/translator.c | 9 +++++---- target/alpha/translate.c | 5 +++-- target/arm/translate.c | 5 +++-- target/avr/translate.c | 5 +++-- target/cris/translate.c | 5 +++-- target/hexagon/translate.c | 6 ++++-- target/hppa/translate.c | 5 +++-- target/i386/tcg/translate.c | 5 +++-- target/loongarch/translate.c | 6 ++++-- target/m68k/translate.c | 5 +++-- target/microblaze/translate.c | 5 +++-- target/mips/tcg/translate.c | 5 +++-- target/nios2/translate.c | 5 +++-- target/openrisc/translate.c | 6 ++++-- target/ppc/translate.c | 5 +++-- target/riscv/translate.c | 5 +++-- target/rx/translate.c | 5 +++-- target/s390x/tcg/translate.c | 5 +++-- target/sh4/translate.c | 5 +++-- target/sparc/translate.c | 5 +++-- target/tricore/translate.c | 6 ++++-- target/xtensa/translate.c | 6 ++++-- 25 files changed, 97 insertions(+), 53 deletions(-) diff --git a/include/exec/exec-all.h b/include/exec/exec-all.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/exec-all.h +++ b/include/exec/exec-all.h @@ -XXX,XX +XXX,XX @@ typedef ram_addr_t tb_page_addr_t; #define TB_PAGE_ADDR_FMT RAM_ADDR_FMT #endif -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns); void restore_state_to_opc(CPUArchState *env, TranslationBlock *tb, target_ulong *data); diff --git a/include/exec/translator.h b/include/exec/translator.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/translator.h +++ b/include/exec/translator.h @@ -XXX,XX +XXX,XX @@ #include "exec/translate-all.h" #include "tcg/tcg.h" +/** + * gen_intermediate_code + * @cpu: cpu context + * @tb: translation block + * @max_insns: max number of instructions to translate + * @pc: guest virtual program counter address + * @host_pc: host physical program counter address + * + * This function must be provided by the target, which should create + * the target-specific DisasContext, and then invoke translator_loop. + */ +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc); /** * DisasJumpType: @@ -XXX,XX +XXX,XX @@ typedef struct TranslatorOps { /** * translator_loop: - * @ops: Target-specific operations. - * @db: Disassembly context. * @cpu: Target vCPU. * @tb: Translation block. * @max_insns: Maximum number of insns to translate. + * @pc: guest virtual program counter address + * @host_pc: host physical program counter address + * @ops: Target-specific operations. + * @db: Disassembly context. * * Generic translator loop. * @@ -XXX,XX +XXX,XX @@ typedef struct TranslatorOps { * - When single-stepping is enabled (system-wide or on the current vCPU). * - When too many instructions have been translated. */ -void translator_loop(const TranslatorOps *ops, DisasContextBase *db, - CPUState *cpu, TranslationBlock *tb, int max_insns); +void translator_loop(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc, + const TranslatorOps *ops, DisasContextBase *db); void translator_loop_temp_check(DisasContextBase *db); diff --git a/accel/tcg/translate-all.c b/accel/tcg/translate-all.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/translate-all.c +++ b/accel/tcg/translate-all.c @@ -XXX,XX +XXX,XX @@ #include "exec/cputlb.h" #include "exec/translate-all.h" +#include "exec/translator.h" #include "qemu/bitmap.h" #include "qemu/qemu-print.h" #include "qemu/timer.h" @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, TCGProfile *prof = &tcg_ctx->prof; int64_t ti; #endif + void *host_pc; assert_memory_lock(); qemu_thread_jit_write(); - phys_pc = get_page_addr_code(env, pc); + phys_pc = get_page_addr_code_hostp(env, pc, &host_pc); if (phys_pc == -1) { /* Generate a one-shot TB with 1 insn in it */ @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, tcg_func_start(tcg_ctx); tcg_ctx->cpu = env_cpu(env); - gen_intermediate_code(cpu, tb, max_insns); + gen_intermediate_code(cpu, tb, max_insns, pc, host_pc); assert(tb->size != 0); tcg_ctx->cpu = NULL; max_insns = tb->icount; diff --git a/accel/tcg/translator.c b/accel/tcg/translator.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/translator.c +++ b/accel/tcg/translator.c @@ -XXX,XX +XXX,XX @@ static inline void translator_page_protect(DisasContextBase *dcbase, #endif } -void translator_loop(const TranslatorOps *ops, DisasContextBase *db, - CPUState *cpu, TranslationBlock *tb, int max_insns) +void translator_loop(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc, + const TranslatorOps *ops, DisasContextBase *db) { uint32_t cflags = tb_cflags(tb); bool plugin_enabled; /* Initialize DisasContext */ db->tb = tb; - db->pc_first = tb->pc; - db->pc_next = db->pc_first; + db->pc_first = pc; + db->pc_next = pc; db->is_jmp = DISAS_NEXT; db->num_insns = 0; db->max_insns = max_insns; diff --git a/target/alpha/translate.c b/target/alpha/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/alpha/translate.c +++ b/target/alpha/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps alpha_tr_ops = { .disas_log = alpha_tr_disas_log, }; -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&alpha_tr_ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, &alpha_tr_ops, &dc.base); } void restore_state_to_opc(CPUAlphaState *env, TranslationBlock *tb, 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 const TranslatorOps thumb_translator_ops = { }; /* generate intermediate code for basic block 'tb'. */ -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc = { }; const TranslatorOps *ops = &arm_translator_ops; @@ -XXX,XX +XXX,XX @@ void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) } #endif - translator_loop(ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, ops, &dc.base); } void restore_state_to_opc(CPUARMState *env, TranslationBlock *tb, diff --git a/target/avr/translate.c b/target/avr/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/translate.c +++ b/target/avr/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps avr_tr_ops = { .disas_log = avr_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc = { }; - translator_loop(&avr_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &avr_tr_ops, &dc.base); } void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb, diff --git a/target/cris/translate.c b/target/cris/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/cris/translate.c +++ b/target/cris/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps cris_tr_ops = { .disas_log = cris_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&cris_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &cris_tr_ops, &dc.base); } void cris_cpu_dump_state(CPUState *cs, FILE *f, int flags) diff --git a/target/hexagon/translate.c b/target/hexagon/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/hexagon/translate.c +++ b/target/hexagon/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps hexagon_tr_ops = { .disas_log = hexagon_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&hexagon_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, + &hexagon_tr_ops, &ctx.base); } #define NAME_LEN 64 diff --git a/target/hppa/translate.c b/target/hppa/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/hppa/translate.c +++ b/target/hppa/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps hppa_tr_ops = { .disas_log = hppa_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&hppa_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &hppa_tr_ops, &ctx.base); } void restore_state_to_opc(CPUHPPAState *env, TranslationBlock *tb, diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/i386/tcg/translate.c +++ b/target/i386/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps i386_tr_ops = { }; /* generate intermediate code for basic block 'tb'. */ -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&i386_tr_ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, &i386_tr_ops, &dc.base); } void restore_state_to_opc(CPUX86State *env, TranslationBlock *tb, diff --git a/target/loongarch/translate.c b/target/loongarch/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/loongarch/translate.c +++ b/target/loongarch/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps loongarch_tr_ops = { .disas_log = loongarch_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&loongarch_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, + &loongarch_tr_ops, &ctx.base); } void loongarch_translate_init(void) diff --git a/target/m68k/translate.c b/target/m68k/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/m68k/translate.c +++ b/target/m68k/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps m68k_tr_ops = { .disas_log = m68k_tr_disas_log, }; -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&m68k_tr_ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, &m68k_tr_ops, &dc.base); } static double floatx80_to_double(CPUM68KState *env, uint16_t high, uint64_t low) diff --git a/target/microblaze/translate.c b/target/microblaze/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/microblaze/translate.c +++ b/target/microblaze/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps mb_tr_ops = { .disas_log = mb_tr_disas_log, }; -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&mb_tr_ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, &mb_tr_ops, &dc.base); } void mb_cpu_dump_state(CPUState *cs, FILE *f, int flags) diff --git a/target/mips/tcg/translate.c b/target/mips/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/mips/tcg/translate.c +++ b/target/mips/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps mips_tr_ops = { .disas_log = mips_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&mips_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &mips_tr_ops, &ctx.base); } void mips_tcg_init(void) diff --git a/target/nios2/translate.c b/target/nios2/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/nios2/translate.c +++ b/target/nios2/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps nios2_tr_ops = { .disas_log = nios2_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&nios2_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &nios2_tr_ops, &dc.base); } void nios2_cpu_dump_state(CPUState *cs, FILE *f, int flags) diff --git a/target/openrisc/translate.c b/target/openrisc/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/openrisc/translate.c +++ b/target/openrisc/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps openrisc_tr_ops = { .disas_log = openrisc_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&openrisc_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, + &openrisc_tr_ops, &ctx.base); } void openrisc_cpu_dump_state(CPUState *cs, FILE *f, int flags) diff --git a/target/ppc/translate.c b/target/ppc/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/ppc/translate.c +++ b/target/ppc/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps ppc_tr_ops = { .disas_log = ppc_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&ppc_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &ppc_tr_ops, &ctx.base); } void restore_state_to_opc(CPUPPCState *env, TranslationBlock *tb, diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps riscv_tr_ops = { .disas_log = riscv_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&riscv_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &riscv_tr_ops, &ctx.base); } void riscv_translate_init(void) diff --git a/target/rx/translate.c b/target/rx/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/rx/translate.c +++ b/target/rx/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps rx_tr_ops = { .disas_log = rx_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&rx_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &rx_tr_ops, &dc.base); } void restore_state_to_opc(CPURXState *env, TranslationBlock *tb, diff --git a/target/s390x/tcg/translate.c b/target/s390x/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/s390x/tcg/translate.c +++ b/target/s390x/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps s390x_tr_ops = { .disas_log = s390x_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc; - translator_loop(&s390x_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &s390x_tr_ops, &dc.base); } void restore_state_to_opc(CPUS390XState *env, TranslationBlock *tb, diff --git a/target/sh4/translate.c b/target/sh4/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/sh4/translate.c +++ b/target/sh4/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps sh4_tr_ops = { .disas_log = sh4_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&sh4_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &sh4_tr_ops, &ctx.base); } void restore_state_to_opc(CPUSH4State *env, TranslationBlock *tb, diff --git a/target/sparc/translate.c b/target/sparc/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/translate.c +++ b/target/sparc/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps sparc_tr_ops = { .disas_log = sparc_tr_disas_log, }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc = {}; - translator_loop(&sparc_tr_ops, &dc.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, &sparc_tr_ops, &dc.base); } void sparc_tcg_init(void) diff --git a/target/tricore/translate.c b/target/tricore/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/tricore/translate.c +++ b/target/tricore/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps tricore_tr_ops = { }; -void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext ctx; - translator_loop(&tricore_tr_ops, &ctx.base, cs, tb, max_insns); + translator_loop(cs, tb, max_insns, pc, host_pc, + &tricore_tr_ops, &ctx.base); } void diff --git a/target/xtensa/translate.c b/target/xtensa/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/xtensa/translate.c +++ b/target/xtensa/translate.c @@ -XXX,XX +XXX,XX @@ static const TranslatorOps xtensa_translator_ops = { .disas_log = xtensa_tr_disas_log, }; -void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns) +void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int max_insns, + target_ulong pc, void *host_pc) { DisasContext dc = {}; - translator_loop(&xtensa_translator_ops, &dc.base, cpu, tb, max_insns); + translator_loop(cpu, tb, max_insns, pc, host_pc, + &xtensa_translator_ops, &dc.base); } void xtensa_cpu_dump_state(CPUState *cs, FILE *f, int flags) -- 2.34.1 Cache the translation from guest to host address, so we may use direct loads when we hit on the primary translation page. Look up the second translation page only once, during translation. This obviates another lookup of the second page within tb_gen_code after translation. Fixes a bug in that plugin_insn_append should be passed the bytes in the original memory order, not bswapped by pieces. Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 63 +++++++++++-------- accel/tcg/translate-all.c | 23 +++---- accel/tcg/translator.c | 126 +++++++++++++++++++++++++++++--------- 3 files changed, 141 insertions(+), 71 deletions(-) diff --git a/include/exec/translator.h b/include/exec/translator.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/translator.h +++ b/include/exec/translator.h @@ -XXX,XX +XXX,XX @@ typedef enum DisasJumpType { * Architecture-agnostic disassembly context. */ typedef struct DisasContextBase { - const TranslationBlock *tb; + TranslationBlock *tb; target_ulong pc_first; target_ulong pc_next; DisasJumpType is_jmp; int num_insns; int max_insns; bool singlestep_enabled; -#ifdef CONFIG_USER_ONLY - /* - * Guest address of the last byte of the last protected page. - * - * Pages containing the translated instructions are made non-writable in - * order to achieve consistency in case another thread is modifying the - * code while translate_insn() fetches the instruction bytes piecemeal. - * Such writer threads are blocked on mmap_lock() in page_unprotect(). - */ - target_ulong page_protect_end; -#endif + void *host_addr[2]; } DisasContextBase; /** @@ -XXX,XX +XXX,XX @@ bool translator_use_goto_tb(DisasContextBase *db, target_ulong dest); * the relevant information at translation time. */ -#define GEN_TRANSLATOR_LD(fullname, type, load_fn, swap_fn) \ - type fullname ## _swap(CPUArchState *env, DisasContextBase *dcbase, \ - abi_ptr pc, bool do_swap); \ - static inline type fullname(CPUArchState *env, \ - DisasContextBase *dcbase, abi_ptr pc) \ - { \ - return fullname ## _swap(env, dcbase, pc, false); \ +uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, abi_ptr pc); +uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, abi_ptr pc); +uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, abi_ptr pc); +uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, abi_ptr pc); + +static inline uint16_t +translator_lduw_swap(CPUArchState *env, DisasContextBase *db, + abi_ptr pc, bool do_swap) +{ + uint16_t ret = translator_lduw(env, db, pc); + if (do_swap) { + ret = bswap16(ret); } + return ret; +} -#define FOR_EACH_TRANSLATOR_LD(F) \ - F(translator_ldub, uint8_t, cpu_ldub_code, /* no swap */) \ - F(translator_lduw, uint16_t, cpu_lduw_code, bswap16) \ - F(translator_ldl, uint32_t, cpu_ldl_code, bswap32) \ - F(translator_ldq, uint64_t, cpu_ldq_code, bswap64) +static inline uint32_t +translator_ldl_swap(CPUArchState *env, DisasContextBase *db, + abi_ptr pc, bool do_swap) +{ + uint32_t ret = translator_ldl(env, db, pc); + if (do_swap) { + ret = bswap32(ret); + } + return ret; +} -FOR_EACH_TRANSLATOR_LD(GEN_TRANSLATOR_LD) - -#undef GEN_TRANSLATOR_LD +static inline uint64_t +translator_ldq_swap(CPUArchState *env, DisasContextBase *db, + abi_ptr pc, bool do_swap) +{ + uint64_t ret = translator_ldq_swap(env, db, pc, false); + if (do_swap) { + ret = bswap64(ret); + } + return ret; +} /* * Return whether addr is on the same page as where disassembly started. diff --git a/accel/tcg/translate-all.c b/accel/tcg/translate-all.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/translate-all.c +++ b/accel/tcg/translate-all.c @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, { CPUArchState *env = cpu->env_ptr; TranslationBlock *tb, *existing_tb; - tb_page_addr_t phys_pc, phys_page2; - target_ulong virt_page2; + tb_page_addr_t phys_pc; tcg_insn_unit *gen_code_buf; int gen_code_size, search_size, max_insns; #ifdef CONFIG_PROFILER @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, tb->flags = flags; tb->cflags = cflags; tb->trace_vcpu_dstate = *cpu->trace_dstate; + tb->page_addr[0] = phys_pc; + tb->page_addr[1] = -1; tcg_ctx->tb_cflags = cflags; tb_overflow: @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, } /* - * If the TB is not associated with a physical RAM page then - * it must be a temporary one-insn TB, and we have nothing to do - * except fill in the page_addr[] fields. Return early before - * attempting to link to other TBs or add to the lookup table. + * If the TB is not associated with a physical RAM page then it must be + * a temporary one-insn TB, and we have nothing left to do. Return early + * before attempting to link to other TBs or add to the lookup table. */ - if (phys_pc == -1) { - tb->page_addr[0] = tb->page_addr[1] = -1; + if (tb->page_addr[0] == -1) { return tb; } @@ -XXX,XX +XXX,XX @@ TranslationBlock *tb_gen_code(CPUState *cpu, */ tcg_tb_insert(tb); - /* check next page if needed */ - virt_page2 = (pc + tb->size - 1) & TARGET_PAGE_MASK; - phys_page2 = -1; - if ((pc & TARGET_PAGE_MASK) != virt_page2) { - phys_page2 = get_page_addr_code(env, virt_page2); - } /* * No explicit memory barrier is required -- tb_link_page() makes the * TB visible in a consistent state. */ - existing_tb = tb_link_page(tb, phys_pc, phys_page2); + existing_tb = tb_link_page(tb, tb->page_addr[0], tb->page_addr[1]); /* if the TB already exists, discard what we just translated */ if (unlikely(existing_tb != tb)) { uintptr_t orig_aligned = (uintptr_t)gen_code_buf; diff --git a/accel/tcg/translator.c b/accel/tcg/translator.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/translator.c +++ b/accel/tcg/translator.c @@ -XXX,XX +XXX,XX @@ bool translator_use_goto_tb(DisasContextBase *db, target_ulong dest) return ((db->pc_first ^ dest) & TARGET_PAGE_MASK) == 0; } -static inline void translator_page_protect(DisasContextBase *dcbase, - target_ulong pc) -{ -#ifdef CONFIG_USER_ONLY - dcbase->page_protect_end = pc | ~TARGET_PAGE_MASK; - page_protect(pc); -#endif -} - void translator_loop(CPUState *cpu, TranslationBlock *tb, int max_insns, target_ulong pc, void *host_pc, const TranslatorOps *ops, DisasContextBase *db) @@ -XXX,XX +XXX,XX @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int max_insns, db->num_insns = 0; db->max_insns = max_insns; db->singlestep_enabled = cflags & CF_SINGLE_STEP; - translator_page_protect(db, db->pc_next); + db->host_addr[0] = host_pc; + db->host_addr[1] = NULL; + +#ifdef CONFIG_USER_ONLY + page_protect(pc); +#endif ops->init_disas_context(db, cpu); tcg_debug_assert(db->is_jmp == DISAS_NEXT); /* no early exit */ @@ -XXX,XX +XXX,XX @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int max_insns, #endif } -static inline void translator_maybe_page_protect(DisasContextBase *dcbase, - target_ulong pc, size_t len) +static void *translator_access(CPUArchState *env, DisasContextBase *db, + target_ulong pc, size_t len) { -#ifdef CONFIG_USER_ONLY - target_ulong end = pc + len - 1; + void *host; + target_ulong base, end; + TranslationBlock *tb; - if (end > dcbase->page_protect_end) { - translator_page_protect(dcbase, end); + tb = db->tb; + + /* Use slow path if first page is MMIO. */ + if (unlikely(tb->page_addr[0] == -1)) { + return NULL; } + + end = pc + len - 1; + if (likely(is_same_page(db, end))) { + host = db->host_addr[0]; + base = db->pc_first; + } else { + host = db->host_addr[1]; + base = TARGET_PAGE_ALIGN(db->pc_first); + if (host == NULL) { + tb->page_addr[1] = + get_page_addr_code_hostp(env, base, &db->host_addr[1]); +#ifdef CONFIG_USER_ONLY + page_protect(end); #endif + /* We cannot handle MMIO as second page. */ + assert(tb->page_addr[1] != -1); + host = db->host_addr[1]; + } + + /* Use slow path when crossing pages. */ + if (is_same_page(db, pc)) { + return NULL; + } + } + + tcg_debug_assert(pc >= base); + return host + (pc - base); } -#define GEN_TRANSLATOR_LD(fullname, type, load_fn, swap_fn) \ - type fullname ## _swap(CPUArchState *env, DisasContextBase *dcbase, \ - abi_ptr pc, bool do_swap) \ - { \ - translator_maybe_page_protect(dcbase, pc, sizeof(type)); \ - type ret = load_fn(env, pc); \ - if (do_swap) { \ - ret = swap_fn(ret); \ - } \ - plugin_insn_append(pc, &ret, sizeof(ret)); \ - return ret; \ +uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +{ + uint8_t ret; + void *p = translator_access(env, db, pc, sizeof(ret)); + + if (p) { + plugin_insn_append(pc, p, sizeof(ret)); + return ldub_p(p); } + ret = cpu_ldub_code(env, pc); + plugin_insn_append(pc, &ret, sizeof(ret)); + return ret; +} -FOR_EACH_TRANSLATOR_LD(GEN_TRANSLATOR_LD) +uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +{ + uint16_t ret, plug; + void *p = translator_access(env, db, pc, sizeof(ret)); -#undef GEN_TRANSLATOR_LD + if (p) { + plugin_insn_append(pc, p, sizeof(ret)); + return lduw_p(p); + } + ret = cpu_lduw_code(env, pc); + plug = tswap16(ret); + plugin_insn_append(pc, &plug, sizeof(ret)); + return ret; +} + +uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +{ + uint32_t ret, plug; + void *p = translator_access(env, db, pc, sizeof(ret)); + + if (p) { + plugin_insn_append(pc, p, sizeof(ret)); + return ldl_p(p); + } + ret = cpu_ldl_code(env, pc); + plug = tswap32(ret); + plugin_insn_append(pc, &plug, sizeof(ret)); + return ret; +} + +uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +{ + uint64_t ret, plug; + void *p = translator_access(env, db, pc, sizeof(ret)); + + if (p) { + plugin_insn_append(pc, p, sizeof(ret)); + return ldq_p(p); + } + ret = cpu_ldq_code(env, pc); + plug = tswap64(ret); + plugin_insn_append(pc, &plug, sizeof(ret)); + return ret; +} -- 2.34.1 From: Ilya Leoshkevich <iii@linux.ibm.com> Right now translator stops right *after* the end of a page, which breaks reporting of fault locations when the last instruction of a multi-insn translation block crosses a page boundary. Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Message-Id: <20220817150506.592862-3-iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/s390x/tcg/translate.c | 15 +++- tests/tcg/s390x/noexec.c | 106 +++++++++++++++++++++++ tests/tcg/multiarch/noexec.c.inc | 139 +++++++++++++++++++++++++++++++ tests/tcg/s390x/Makefile.target | 1 + 4 files changed, 257 insertions(+), 4 deletions(-) create mode 100644 tests/tcg/s390x/noexec.c create mode 100644 tests/tcg/multiarch/noexec.c.inc diff --git a/target/s390x/tcg/translate.c b/target/s390x/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/s390x/tcg/translate.c +++ b/target/s390x/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static void s390x_tr_insn_start(DisasContextBase *dcbase, CPUState *cs) dc->insn_start = tcg_last_op(); } +static target_ulong get_next_pc(CPUS390XState *env, DisasContext *s, + uint64_t pc) +{ + uint64_t insn = ld_code2(env, s, pc); + + return pc + get_ilen((insn >> 8) & 0xff); +} + static void s390x_tr_translate_insn(DisasContextBase *dcbase, CPUState *cs) { CPUS390XState *env = cs->env_ptr; @@ -XXX,XX +XXX,XX @@ static void s390x_tr_translate_insn(DisasContextBase *dcbase, CPUState *cs) dc->base.is_jmp = translate_one(env, dc); if (dc->base.is_jmp == DISAS_NEXT) { - uint64_t page_start; - - page_start = dc->base.pc_first & TARGET_PAGE_MASK; - if (dc->base.pc_next - page_start >= TARGET_PAGE_SIZE || dc->ex_value) { + if (!is_same_page(dcbase, dc->base.pc_next) || + !is_same_page(dcbase, get_next_pc(env, dc, dc->base.pc_next)) || + dc->ex_value) { dc->base.is_jmp = DISAS_TOO_MANY; } } diff --git a/tests/tcg/s390x/noexec.c b/tests/tcg/s390x/noexec.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/s390x/noexec.c @@ -XXX,XX +XXX,XX @@ +#include "../multiarch/noexec.c.inc" + +static void *arch_mcontext_pc(const mcontext_t *ctx) +{ + return (void *)ctx->psw.addr; +} + +static int arch_mcontext_arg(const mcontext_t *ctx) +{ + return ctx->gregs[2]; +} + +static void arch_flush(void *p, int len) +{ +} + +extern char noexec_1[]; +extern char noexec_2[]; +extern char noexec_end[]; + +asm("noexec_1:\n" + " lgfi %r2,1\n" /* %r2 is 0 on entry, set 1. */ + "noexec_2:\n" + " lgfi %r2,2\n" /* %r2 is 0/1; set 2. */ + " br %r14\n" /* return */ + "noexec_end:"); + +extern char exrl_1[]; +extern char exrl_2[]; +extern char exrl_end[]; + +asm("exrl_1:\n" + " exrl %r0, exrl_2\n" + " br %r14\n" + "exrl_2:\n" + " lgfi %r2,2\n" + "exrl_end:"); + +int main(void) +{ + struct noexec_test noexec_tests[] = { + { + .name = "fallthrough", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = noexec_1 - noexec_2, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 1, + }, + { + .name = "jump", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = 0, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 0, + }, + { + .name = "exrl", + .test_code = exrl_1, + .test_len = exrl_end - exrl_1, + .page_ofs = exrl_1 - exrl_2, + .entry_ofs = exrl_1 - exrl_2, + .expected_si_ofs = 0, + .expected_pc_ofs = exrl_1 - exrl_2, + .expected_arg = 0, + }, + { + .name = "fallthrough [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = noexec_1 - noexec_2 - 2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 1, + }, + { + .name = "jump [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = -2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 0, + }, + { + .name = "exrl [cross]", + .test_code = exrl_1, + .test_len = exrl_end - exrl_1, + .page_ofs = exrl_1 - exrl_2 - 2, + .entry_ofs = exrl_1 - exrl_2 - 2, + .expected_si_ofs = 0, + .expected_pc_ofs = exrl_1 - exrl_2 - 2, + .expected_arg = 0, + }, + }; + + return test_noexec(noexec_tests, + sizeof(noexec_tests) / sizeof(noexec_tests[0])); +} diff --git a/tests/tcg/multiarch/noexec.c.inc b/tests/tcg/multiarch/noexec.c.inc new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/multiarch/noexec.c.inc @@ -XXX,XX +XXX,XX @@ +/* + * Common code for arch-specific MMU_INST_FETCH fault testing. + */ + +#define _GNU_SOURCE + +#include <assert.h> +#include <signal.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <errno.h> +#include <unistd.h> +#include <sys/mman.h> +#include <sys/ucontext.h> + +/* Forward declarations. */ + +static void *arch_mcontext_pc(const mcontext_t *ctx); +static int arch_mcontext_arg(const mcontext_t *ctx); +static void arch_flush(void *p, int len); + +/* Testing infrastructure. */ + +struct noexec_test { + const char *name; + const char *test_code; + int test_len; + int page_ofs; + int entry_ofs; + int expected_si_ofs; + int expected_pc_ofs; + int expected_arg; +}; + +static void *page_base; +static int page_size; +static const struct noexec_test *current_noexec_test; + +static void handle_err(const char *syscall) +{ + printf("[ FAILED ] %s: %s\n", syscall, strerror(errno)); + exit(EXIT_FAILURE); +} + +static void handle_segv(int sig, siginfo_t *info, void *ucontext) +{ + const struct noexec_test *test = current_noexec_test; + const mcontext_t *mc = &((ucontext_t *)ucontext)->uc_mcontext; + void *expected_si; + void *expected_pc; + void *pc; + int arg; + + if (test == NULL) { + printf("[ FAILED ] unexpected SEGV\n"); + exit(EXIT_FAILURE); + } + current_noexec_test = NULL; + + expected_si = page_base + test->expected_si_ofs; + if (info->si_addr != expected_si) { + printf("[ FAILED ] wrong si_addr (%p != %p)\n", + info->si_addr, expected_si); + exit(EXIT_FAILURE); + } + + pc = arch_mcontext_pc(mc); + expected_pc = page_base + test->expected_pc_ofs; + if (pc != expected_pc) { + printf("[ FAILED ] wrong pc (%p != %p)\n", pc, expected_pc); + exit(EXIT_FAILURE); + } + + arg = arch_mcontext_arg(mc); + if (arg != test->expected_arg) { + printf("[ FAILED ] wrong arg (%d != %d)\n", arg, test->expected_arg); + exit(EXIT_FAILURE); + } + + if (mprotect(page_base, page_size, + PROT_READ | PROT_WRITE | PROT_EXEC) < 0) { + handle_err("mprotect"); + } +} + +static void test_noexec_1(const struct noexec_test *test) +{ + void *start = page_base + test->page_ofs; + void (*fn)(int arg) = page_base + test->entry_ofs; + + memcpy(start, test->test_code, test->test_len); + arch_flush(start, test->test_len); + + /* Trigger TB creation in order to test invalidation. */ + fn(0); + + if (mprotect(page_base, page_size, PROT_NONE) < 0) { + handle_err("mprotect"); + } + + /* Trigger SEGV and check that handle_segv() ran. */ + current_noexec_test = test; + fn(0); + assert(current_noexec_test == NULL); +} + +static int test_noexec(struct noexec_test *tests, size_t n_tests) +{ + struct sigaction act; + size_t i; + + memset(&act, 0, sizeof(act)); + act.sa_sigaction = handle_segv; + act.sa_flags = SA_SIGINFO; + if (sigaction(SIGSEGV, &act, NULL) < 0) { + handle_err("sigaction"); + } + + page_size = getpagesize(); + page_base = mmap(NULL, 2 * page_size, + PROT_READ | PROT_WRITE | PROT_EXEC, + MAP_ANONYMOUS | MAP_PRIVATE, -1, 0); + if (page_base == MAP_FAILED) { + handle_err("mmap"); + } + page_base += page_size; + + for (i = 0; i < n_tests; i++) { + struct noexec_test *test = &tests[i]; + + printf("[ RUN ] %s\n", test->name); + test_noexec_1(test); + printf("[ OK ]\n"); + } + + printf("[ PASSED ]\n"); + return EXIT_SUCCESS; +} diff --git a/tests/tcg/s390x/Makefile.target b/tests/tcg/s390x/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/s390x/Makefile.target +++ b/tests/tcg/s390x/Makefile.target @@ -XXX,XX +XXX,XX @@ TESTS+=shift TESTS+=trap TESTS+=signals-s390x TESTS+=branch-relative-long +TESTS+=noexec Z14_TESTS=vfminmax vfminmax: LDFLAGS+=-lm -- 2.34.1 From: Ilya Leoshkevich <iii@linux.ibm.com> Right now translator stops right *after* the end of a page, which breaks reporting of fault locations when the last instruction of a multi-insn translation block crosses a page boundary. An implementation, like the one arm and s390x have, would require an i386 length disassembler, which is burdensome to maintain. Another alternative would be to single-step at the end of a guest page, but this may come with a performance impact. Fix by snapshotting disassembly state and restoring it after we figure out we crossed a page boundary. This includes rolling back cc_op updates and emitted ops. Signed-off-by: Ilya Leoshkevich <iii@linux.ibm.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1143 Message-Id: <20220817150506.592862-4-iii@linux.ibm.com> [rth: Simplify end-of-insn cross-page checks.] Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/i386/tcg/translate.c | 64 ++++++++++++++++----------- tests/tcg/x86_64/noexec.c | 75 ++++++++++++++++++++++++++++++++ tests/tcg/x86_64/Makefile.target | 3 +- 3 files changed, 116 insertions(+), 26 deletions(-) create mode 100644 tests/tcg/x86_64/noexec.c diff --git a/target/i386/tcg/translate.c b/target/i386/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/i386/tcg/translate.c +++ b/target/i386/tcg/translate.c @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { TCGv_i64 tmp1_i64; sigjmp_buf jmpbuf; + TCGOp *prev_insn_end; } DisasContext; /* The environment in which user-only runs is constrained. */ @@ -XXX,XX +XXX,XX @@ static uint64_t advance_pc(CPUX86State *env, DisasContext *s, int num_bytes) { uint64_t pc = s->pc; + /* This is a subsequent insn that crosses a page boundary. */ + if (s->base.num_insns > 1 && + !is_same_page(&s->base, s->pc + num_bytes - 1)) { + siglongjmp(s->jmpbuf, 2); + } + s->pc += num_bytes; if (unlikely(s->pc - s->pc_start > X86_MAX_INSN_LENGTH)) { /* If the instruction's 16th byte is on a different page than the 1st, a @@ -XXX,XX +XXX,XX @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu) int modrm, reg, rm, mod, op, opreg, val; target_ulong next_eip, tval; target_ulong pc_start = s->base.pc_next; + bool orig_cc_op_dirty = s->cc_op_dirty; + CCOp orig_cc_op = s->cc_op; s->pc_start = s->pc = pc_start; s->override = -1; @@ -XXX,XX +XXX,XX @@ static target_ulong disas_insn(DisasContext *s, CPUState *cpu) s->rip_offset = 0; /* for relative ip address */ s->vex_l = 0; s->vex_v = 0; - if (sigsetjmp(s->jmpbuf, 0) != 0) { + switch (sigsetjmp(s->jmpbuf, 0)) { + case 0: + break; + case 1: gen_exception_gpf(s); return s->pc; + case 2: + /* Restore state that may affect the next instruction. */ + s->cc_op_dirty = orig_cc_op_dirty; + s->cc_op = orig_cc_op; + s->base.num_insns--; + tcg_remove_ops_after(s->prev_insn_end); + s->base.is_jmp = DISAS_TOO_MANY; + return pc_start; + default: + g_assert_not_reached(); } prefixes = 0; @@ -XXX,XX +XXX,XX @@ static void i386_tr_insn_start(DisasContextBase *dcbase, CPUState *cpu) { DisasContext *dc = container_of(dcbase, DisasContext, base); + dc->prev_insn_end = tcg_last_op(); tcg_gen_insn_start(dc->base.pc_next, dc->cc_op); } @@ -XXX,XX +XXX,XX @@ static void i386_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) #endif pc_next = disas_insn(dc, cpu); - - if (dc->flags & (HF_TF_MASK | HF_INHIBIT_IRQ_MASK)) { - /* if single step mode, we generate only one instruction and - generate an exception */ - /* if irq were inhibited with HF_INHIBIT_IRQ_MASK, we clear - the flag and abort the translation to give the irqs a - chance to happen */ - dc->base.is_jmp = DISAS_TOO_MANY; - } else if ((tb_cflags(dc->base.tb) & CF_USE_ICOUNT) - && ((pc_next & TARGET_PAGE_MASK) - != ((pc_next + TARGET_MAX_INSN_SIZE - 1) - & TARGET_PAGE_MASK) - || (pc_next & ~TARGET_PAGE_MASK) == 0)) { - /* Do not cross the boundary of the pages in icount mode, - it can cause an exception. Do it only when boundary is - crossed by the first instruction in the block. - If current instruction already crossed the bound - it's ok, - because an exception hasn't stopped this code. - */ - dc->base.is_jmp = DISAS_TOO_MANY; - } else if ((pc_next - dc->base.pc_first) >= (TARGET_PAGE_SIZE - 32)) { - dc->base.is_jmp = DISAS_TOO_MANY; - } - dc->base.pc_next = pc_next; + + if (dc->base.is_jmp == DISAS_NEXT) { + if (dc->flags & (HF_TF_MASK | HF_INHIBIT_IRQ_MASK)) { + /* + * If single step mode, we generate only one instruction and + * generate an exception. + * If irq were inhibited with HF_INHIBIT_IRQ_MASK, we clear + * the flag and abort the translation to give the irqs a + * chance to happen. + */ + dc->base.is_jmp = DISAS_TOO_MANY; + } else if (!is_same_page(&dc->base, pc_next)) { + dc->base.is_jmp = DISAS_TOO_MANY; + } + } } static void i386_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) diff --git a/tests/tcg/x86_64/noexec.c b/tests/tcg/x86_64/noexec.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/x86_64/noexec.c @@ -XXX,XX +XXX,XX @@ +#include "../multiarch/noexec.c.inc" + +static void *arch_mcontext_pc(const mcontext_t *ctx) +{ + return (void *)ctx->gregs[REG_RIP]; +} + +int arch_mcontext_arg(const mcontext_t *ctx) +{ + return ctx->gregs[REG_RDI]; +} + +static void arch_flush(void *p, int len) +{ +} + +extern char noexec_1[]; +extern char noexec_2[]; +extern char noexec_end[]; + +asm("noexec_1:\n" + " movq $1,%rdi\n" /* %rdi is 0 on entry, set 1. */ + "noexec_2:\n" + " movq $2,%rdi\n" /* %rdi is 0/1; set 2. */ + " ret\n" + "noexec_end:"); + +int main(void) +{ + struct noexec_test noexec_tests[] = { + { + .name = "fallthrough", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = noexec_1 - noexec_2, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 1, + }, + { + .name = "jump", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = 0, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 0, + }, + { + .name = "fallthrough [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = noexec_1 - noexec_2 - 2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 1, + }, + { + .name = "jump [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = -2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 0, + }, + }; + + return test_noexec(noexec_tests, + sizeof(noexec_tests) / sizeof(noexec_tests[0])); +} diff --git a/tests/tcg/x86_64/Makefile.target b/tests/tcg/x86_64/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/x86_64/Makefile.target +++ b/tests/tcg/x86_64/Makefile.target @@ -XXX,XX +XXX,XX @@ include $(SRC_PATH)/tests/tcg/i386/Makefile.target ifeq ($(filter %-linux-user, $(TARGET)),$(TARGET)) X86_64_TESTS += vsyscall +X86_64_TESTS += noexec TESTS=$(MULTIARCH_TESTS) $(X86_64_TESTS) test-x86_64 else TESTS=$(MULTIARCH_TESTS) @@ -XXX,XX +XXX,XX @@ test-x86_64: LDFLAGS+=-lm -lc test-x86_64: test-i386.c test-i386.h test-i386-shift.h test-i386-muldiv.h $(CC) $(CFLAGS) $< -o $@ $(LDFLAGS) -vsyscall: $(SRC_PATH)/tests/tcg/x86_64/vsyscall.c +%: $(SRC_PATH)/tests/tcg/x86_64/%.c $(CC) $(CFLAGS) $< -o $@ $(LDFLAGS) -- 2.34.1 These will be useful in properly ending the TB. Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/riscv/translate.c | 10 +++++++++- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) /* Include decoders for factored-out extensions */ #include "decode-XVentanaCondOps.c.inc" +/* The specification allows for longer insns, but not supported by qemu. */ +#define MAX_INSN_LEN 4 + +static inline int insn_len(uint16_t first_word) +{ + return (first_word & 3) == 3 ? 4 : 2; +} + static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) { /* @@ -XXX,XX +XXX,XX @@ static void decode_opc(CPURISCVState *env, DisasContext *ctx, uint16_t opcode) }; /* Check for compressed insn */ - if (extract16(opcode, 0, 2) != 3) { + if (insn_len(opcode) == 2) { if (!has_ext(ctx, RVC)) { gen_exception_illegal(ctx); } else { -- 2.34.1 Right now the translator stops right *after* the end of a page, which breaks reporting of fault locations when the last instruction of a multi-insn translation block crosses a page boundary. Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1155 Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Acked-by: Ilya Leoshkevich <iii@linux.ibm.com> Tested-by: Ilya Leoshkevich <iii@linux.ibm.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/riscv/translate.c | 17 +++++-- tests/tcg/riscv64/noexec.c | 79 +++++++++++++++++++++++++++++++ tests/tcg/riscv64/Makefile.target | 1 + 3 files changed, 93 insertions(+), 4 deletions(-) create mode 100644 tests/tcg/riscv64/noexec.c diff --git a/target/riscv/translate.c b/target/riscv/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/translate.c +++ b/target/riscv/translate.c @@ -XXX,XX +XXX,XX @@ static void riscv_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) } ctx->nftemp = 0; + /* Only the first insn within a TB is allowed to cross a page boundary. */ if (ctx->base.is_jmp == DISAS_NEXT) { - target_ulong page_start; - - page_start = ctx->base.pc_first & TARGET_PAGE_MASK; - if (ctx->base.pc_next - page_start >= TARGET_PAGE_SIZE) { + if (!is_same_page(&ctx->base, ctx->base.pc_next)) { ctx->base.is_jmp = DISAS_TOO_MANY; + } else { + unsigned page_ofs = ctx->base.pc_next & ~TARGET_PAGE_MASK; + + if (page_ofs > TARGET_PAGE_SIZE - MAX_INSN_LEN) { + uint16_t next_insn = cpu_lduw_code(env, ctx->base.pc_next); + int len = insn_len(next_insn); + + if (!is_same_page(&ctx->base, ctx->base.pc_next + len)) { + ctx->base.is_jmp = DISAS_TOO_MANY; + } + } } } } diff --git a/tests/tcg/riscv64/noexec.c b/tests/tcg/riscv64/noexec.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/tests/tcg/riscv64/noexec.c @@ -XXX,XX +XXX,XX @@ +#include "../multiarch/noexec.c.inc" + +static void *arch_mcontext_pc(const mcontext_t *ctx) +{ + return (void *)ctx->__gregs[REG_PC]; +} + +static int arch_mcontext_arg(const mcontext_t *ctx) +{ + return ctx->__gregs[REG_A0]; +} + +static void arch_flush(void *p, int len) +{ + __builtin___clear_cache(p, p + len); +} + +extern char noexec_1[]; +extern char noexec_2[]; +extern char noexec_end[]; + +asm(".option push\n" + ".option norvc\n" + "noexec_1:\n" + " li a0,1\n" /* a0 is 0 on entry, set 1. */ + "noexec_2:\n" + " li a0,2\n" /* a0 is 0/1; set 2. */ + " ret\n" + "noexec_end:\n" + ".option pop"); + +int main(void) +{ + struct noexec_test noexec_tests[] = { + { + .name = "fallthrough", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = noexec_1 - noexec_2, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 1, + }, + { + .name = "jump", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2, + .entry_ofs = 0, + .expected_si_ofs = 0, + .expected_pc_ofs = 0, + .expected_arg = 0, + }, + { + .name = "fallthrough [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = noexec_1 - noexec_2 - 2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 1, + }, + { + .name = "jump [cross]", + .test_code = noexec_1, + .test_len = noexec_end - noexec_1, + .page_ofs = noexec_1 - noexec_2 - 2, + .entry_ofs = -2, + .expected_si_ofs = 0, + .expected_pc_ofs = -2, + .expected_arg = 0, + }, + }; + + return test_noexec(noexec_tests, + sizeof(noexec_tests) / sizeof(noexec_tests[0])); +} diff --git a/tests/tcg/riscv64/Makefile.target b/tests/tcg/riscv64/Makefile.target index XXXXXXX..XXXXXXX 100644 --- a/tests/tcg/riscv64/Makefile.target +++ b/tests/tcg/riscv64/Makefile.target @@ -XXX,XX +XXX,XX @@ VPATH += $(SRC_PATH)/tests/tcg/riscv64 TESTS += test-div +TESTS += noexec -- 2.34.1 The following changes since commit a9fe9e191b4305b88c356a1ed9ac3baf89eb18aa: Merge tag 'pull-riscv-to-apply-20230505-1' of https://github.com/alistair23/qemu into staging (2023-05-05 09:25:13 +0100) are available in the Git repository at: https://gitlab.com/rth7680/qemu.git tags/pull-tcg-20230505 for you to fetch changes up to 35a0bd63b458f30389b6bc6b7471c1665fe7b9d8: tcg: Widen helper_*_st[bw]_mmu val arguments (2023-05-05 17:21:03 +0100) ---------------------------------------------------------------- softfloat: Fix the incorrect computation in float32_exp2 tcg: Remove compatability helpers for qemu ld/st target/alpha: Remove TARGET_ALIGNED_ONLY target/hppa: Remove TARGET_ALIGNED_ONLY target/sparc: Remove TARGET_ALIGNED_ONLY tcg: Cleanups preparing to unify calls to qemu_ld/st helpers ---------------------------------------------------------------- Richard Henderson (41): target/avr: Finish conversion to tcg_gen_qemu_{ld,st}_* target/cris: Finish conversion to tcg_gen_qemu_{ld,st}_* target/Hexagon: Finish conversion to tcg_gen_qemu_{ld, st}_* target/m68k: Finish conversion to tcg_gen_qemu_{ld,st}_* target/mips: Finish conversion to tcg_gen_qemu_{ld,st}_* target/s390x: Finish conversion to tcg_gen_qemu_{ld, st}_* target/sparc: Finish conversion to tcg_gen_qemu_{ld, st}_* target/xtensa: Finish conversion to tcg_gen_qemu_{ld, st}_* tcg: Remove compatability helpers for qemu ld/st target/alpha: Use MO_ALIGN for system UNALIGN() target/alpha: Use MO_ALIGN where required target/alpha: Remove TARGET_ALIGNED_ONLY target/hppa: Use MO_ALIGN for system UNALIGN() target/hppa: Remove TARGET_ALIGNED_ONLY target/sparc: Use MO_ALIGN where required target/sparc: Use cpu_ld*_code_mmu target/sparc: Remove TARGET_ALIGNED_ONLY tcg/i386: Rationalize args to tcg_out_qemu_{ld,st} tcg/i386: Generalize multi-part load overlap test tcg/i386: Introduce HostAddress tcg/i386: Drop r0+r1 local variables from tcg_out_tlb_load tcg/i386: Introduce tcg_out_testi tcg/aarch64: Rationalize args to tcg_out_qemu_{ld,st} tcg/aarch64: Introduce HostAddress tcg/arm: Rationalize args to tcg_out_qemu_{ld,st} tcg/arm: Introduce HostAddress tcg/loongarch64: Rationalize args to tcg_out_qemu_{ld,st} tcg/loongarch64: Introduce HostAddress tcg/mips: Rationalize args to tcg_out_qemu_{ld,st} tcg/ppc: Rationalize args to tcg_out_qemu_{ld,st} tcg/ppc: Introduce HostAddress tcg/riscv: Require TCG_TARGET_REG_BITS == 64 tcg/riscv: Rationalize args to tcg_out_qemu_{ld,st} tcg/s390x: Pass TCGType to tcg_out_qemu_{ld,st} tcg/s390x: Introduce HostAddress tcg/sparc64: Drop is_64 test from tcg_out_qemu_ld data return tcg/sparc64: Pass TCGType to tcg_out_qemu_{ld,st} tcg: Move TCGLabelQemuLdst to tcg.c tcg: Replace REG_P with arg_loc_reg_p tcg: Introduce arg_slot_stk_ofs tcg: Widen helper_*_st[bw]_mmu val arguments Shivaprasad G Bhat (1): softfloat: Fix the incorrect computation in float32_exp2 configs/targets/alpha-linux-user.mak | 1 - configs/targets/alpha-softmmu.mak | 1 - configs/targets/hppa-linux-user.mak | 1 - configs/targets/hppa-softmmu.mak | 1 - configs/targets/sparc-linux-user.mak | 1 - configs/targets/sparc-softmmu.mak | 1 - configs/targets/sparc32plus-linux-user.mak | 1 - configs/targets/sparc64-linux-user.mak | 1 - configs/targets/sparc64-softmmu.mak | 1 - include/tcg/tcg-ldst.h | 10 +- include/tcg/tcg-op.h | 55 ----- target/hexagon/macros.h | 14 +- tcg/riscv/tcg-target-con-set.h | 8 - tcg/riscv/tcg-target.h | 22 +- tcg/tcg-internal.h | 4 - accel/tcg/cputlb.c | 6 +- fpu/softfloat.c | 2 +- target/alpha/translate.c | 38 +-- target/avr/translate.c | 16 +- target/hexagon/genptr.c | 8 +- target/hexagon/idef-parser/parser-helpers.c | 28 +-- target/hexagon/translate.c | 32 +-- target/hppa/translate.c | 2 +- target/m68k/translate.c | 76 ++---- target/mips/tcg/translate.c | 8 +- target/s390x/tcg/translate.c | 152 ++++++------ target/sparc/ldst_helper.c | 10 +- target/sparc/translate.c | 85 ++++--- target/xtensa/translate.c | 4 +- tcg/tcg.c | 58 +++-- target/cris/translate_v10.c.inc | 18 +- target/mips/tcg/nanomips_translate.c.inc | 2 +- tcg/aarch64/tcg-target.c.inc | 108 ++++++--- tcg/arm/tcg-target.c.inc | 357 +++++++++++++--------------- tcg/i386/tcg-target.c.inc | 345 ++++++++++++++------------- tcg/loongarch64/tcg-target.c.inc | 135 +++++------ tcg/mips/tcg-target.c.inc | 186 ++++++++------- tcg/ppc/tcg-target.c.inc | 192 ++++++++------- tcg/riscv/tcg-target.c.inc | 268 ++++++--------------- tcg/s390x/tcg-target.c.inc | 131 +++++----- tcg/sparc64/tcg-target.c.inc | 8 +- tcg/tcg-ldst.c.inc | 14 -- 42 files changed, 1120 insertions(+), 1291 deletions(-) From: Shivaprasad G Bhat <sbhat@linux.ibm.com> The float32_exp2 function is computing wrong exponent of 2. For example, with the following set of values {0.1, 2.0, 2.0, -1.0}, the expected output would be {1.071773, 4.000000, 4.000000, 0.500000}. Instead, the function is computing {1.119102, 3.382044, 3.382044, -0.191022} Looking at the code, the float32_exp2() attempts to do this 2 3 4 5 n x x x x x x x e = 1 + --- + --- + --- + --- + --- + ... + --- + ... 1! 2! 3! 4! 5! n! But because of the typo it ends up doing x x x x x x x e = 1 + --- + --- + --- + --- + --- + ... + --- + ... 1! 2! 3! 4! 5! n! This is because instead of the xnp which holds the numerator, parts_muladd is using the xp which is just 'x'. Commit '572c4d862ff2' refactored this function, and mistakenly used xp instead of xnp. Cc: qemu-stable@nongnu.org Fixes: 572c4d862ff2 "softfloat: Convert float32_exp2 to FloatParts" Partially-Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1623 Reported-By: Luca Barbato (https://gitlab.com/lu-zero) Signed-off-by: Shivaprasad G Bhat <sbhat@linux.ibm.com> Signed-off-by: Vaibhav Jain <vaibhav@linux.ibm.com> Message-Id: <168304110865.537992.13059030916325018670.stgit@localhost.localdomain> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- fpu/softfloat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fpu/softfloat.c b/fpu/softfloat.c index XXXXXXX..XXXXXXX 100644 --- a/fpu/softfloat.c +++ b/fpu/softfloat.c @@ -XXX,XX +XXX,XX @@ float32 float32_exp2(float32 a, float_status *status) float64_unpack_canonical(&rp, float64_one, status); for (i = 0 ; i < 15 ; i++) { float64_unpack_canonical(&tp, float32_exp2_coefficients[i], status); - rp = *parts_muladd(&tp, &xp, &rp, 0, status); + rp = *parts_muladd(&tp, &xnp, &rp, 0, status); xnp = *parts_mul(&xnp, &xp, status); } -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-2-richard.henderson@linaro.org> --- target/avr/translate.c | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/target/avr/translate.c b/target/avr/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/avr/translate.c +++ b/target/avr/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_data_store(DisasContext *ctx, TCGv data, TCGv addr) if (ctx->base.tb->flags & TB_FLAGS_FULL_ACCESS) { gen_helper_fullwr(cpu_env, data, addr); } else { - tcg_gen_qemu_st8(data, addr, MMU_DATA_IDX); /* mem[addr] = data */ + tcg_gen_qemu_st_tl(data, addr, MMU_DATA_IDX, MO_UB); } } @@ -XXX,XX +XXX,XX @@ static void gen_data_load(DisasContext *ctx, TCGv data, TCGv addr) if (ctx->base.tb->flags & TB_FLAGS_FULL_ACCESS) { gen_helper_fullrd(data, cpu_env, addr); } else { - tcg_gen_qemu_ld8u(data, addr, MMU_DATA_IDX); /* data = mem[addr] */ + tcg_gen_qemu_ld_tl(data, addr, MMU_DATA_IDX, MO_UB); } } @@ -XXX,XX +XXX,XX @@ static bool trans_LPM1(DisasContext *ctx, arg_LPM1 *a) tcg_gen_shli_tl(addr, H, 8); /* addr = H:L */ tcg_gen_or_tl(addr, addr, L); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LPM2(DisasContext *ctx, arg_LPM2 *a) tcg_gen_shli_tl(addr, H, 8); /* addr = H:L */ tcg_gen_or_tl(addr, addr, L); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_LPMX(DisasContext *ctx, arg_LPMX *a) tcg_gen_shli_tl(addr, H, 8); /* addr = H:L */ tcg_gen_or_tl(addr, addr, L); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); tcg_gen_addi_tl(addr, addr, 1); /* addr = addr + 1 */ tcg_gen_andi_tl(L, addr, 0xff); tcg_gen_shri_tl(addr, addr, 8); @@ -XXX,XX +XXX,XX @@ static bool trans_ELPM1(DisasContext *ctx, arg_ELPM1 *a) TCGv Rd = cpu_r[0]; TCGv addr = gen_get_zaddr(); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ELPM2(DisasContext *ctx, arg_ELPM2 *a) TCGv Rd = cpu_r[a->rd]; TCGv addr = gen_get_zaddr(); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); return true; } @@ -XXX,XX +XXX,XX @@ static bool trans_ELPMX(DisasContext *ctx, arg_ELPMX *a) TCGv Rd = cpu_r[a->rd]; TCGv addr = gen_get_zaddr(); - tcg_gen_qemu_ld8u(Rd, addr, MMU_CODE_IDX); /* Rd = mem[addr] */ + tcg_gen_qemu_ld_tl(Rd, addr, MMU_CODE_IDX, MO_UB); tcg_gen_addi_tl(addr, addr, 1); /* addr = addr + 1 */ gen_set_zaddr(addr); return true; -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. In this case we can fold the calls using the size bits of MemOp. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-3-richard.henderson@linaro.org> --- target/cris/translate_v10.c.inc | 18 ++++-------------- 1 file changed, 4 insertions(+), 14 deletions(-) diff --git a/target/cris/translate_v10.c.inc b/target/cris/translate_v10.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/cris/translate_v10.c.inc +++ b/target/cris/translate_v10.c.inc @@ -XXX,XX +XXX,XX @@ static void gen_store_v10_conditional(DisasContext *dc, TCGv addr, TCGv val, /* Store only if F flag isn't set */ tcg_gen_andi_tl(t1, cpu_PR[PR_CCS], F_FLAG_V10); tcg_gen_brcondi_tl(TCG_COND_NE, t1, 0, l1); - if (size == 1) { - tcg_gen_qemu_st8(tval, taddr, mem_index); - } else if (size == 2) { - tcg_gen_qemu_st16(tval, taddr, mem_index); - } else { - tcg_gen_qemu_st32(tval, taddr, mem_index); - } + + tcg_gen_qemu_st_tl(tval, taddr, mem_index, ctz32(size) | MO_TE); + gen_set_label(l1); tcg_gen_shri_tl(t1, t1, 1); /* shift F to P position */ tcg_gen_or_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], t1); /*P=F*/ @@ -XXX,XX +XXX,XX @@ static void gen_store_v10(DisasContext *dc, TCGv addr, TCGv val, return; } - if (size == 1) { - tcg_gen_qemu_st8(val, addr, mem_index); - } else if (size == 2) { - tcg_gen_qemu_st16(val, addr, mem_index); - } else { - tcg_gen_qemu_st32(val, addr, mem_index); - } + tcg_gen_qemu_st_tl(val, addr, mem_index, ctz32(size) | MO_TE); } -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Importantly, this removes some incorrect casts generated by idef-parser's gen_load(). Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Tested-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Taylor Simpson <tsimpson@quicinc.com> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-4-richard.henderson@linaro.org> --- target/hexagon/macros.h | 14 ++++----- target/hexagon/genptr.c | 8 +++--- target/hexagon/idef-parser/parser-helpers.c | 28 +++++++++--------- target/hexagon/translate.c | 32 ++++++++++----------- 4 files changed, 40 insertions(+), 42 deletions(-) diff --git a/target/hexagon/macros.h b/target/hexagon/macros.h index XXXXXXX..XXXXXXX 100644 --- a/target/hexagon/macros.h +++ b/target/hexagon/macros.h @@ -XXX,XX +XXX,XX @@ #define MEM_LOAD1s(DST, VA) \ do { \ CHECK_NOSHUF(VA, 1); \ - tcg_gen_qemu_ld8s(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_SB); \ } while (0) #define MEM_LOAD1u(DST, VA) \ do { \ CHECK_NOSHUF(VA, 1); \ - tcg_gen_qemu_ld8u(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_UB); \ } while (0) #define MEM_LOAD2s(DST, VA) \ do { \ CHECK_NOSHUF(VA, 2); \ - tcg_gen_qemu_ld16s(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_TESW); \ } while (0) #define MEM_LOAD2u(DST, VA) \ do { \ CHECK_NOSHUF(VA, 2); \ - tcg_gen_qemu_ld16u(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_TEUW); \ } while (0) #define MEM_LOAD4s(DST, VA) \ do { \ CHECK_NOSHUF(VA, 4); \ - tcg_gen_qemu_ld32s(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_TESL); \ } while (0) #define MEM_LOAD4u(DST, VA) \ do { \ CHECK_NOSHUF(VA, 4); \ - tcg_gen_qemu_ld32s(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(DST, VA, ctx->mem_idx, MO_TEUL); \ } while (0) #define MEM_LOAD8u(DST, VA) \ do { \ CHECK_NOSHUF(VA, 8); \ - tcg_gen_qemu_ld64(DST, VA, ctx->mem_idx); \ + tcg_gen_qemu_ld_i64(DST, VA, ctx->mem_idx, MO_TEUQ); \ } while (0) #define MEM_STORE1_FUNC(X) \ diff --git a/target/hexagon/genptr.c b/target/hexagon/genptr.c index XXXXXXX..XXXXXXX 100644 --- a/target/hexagon/genptr.c +++ b/target/hexagon/genptr.c @@ -XXX,XX +XXX,XX @@ void gen_set_byte_i64(int N, TCGv_i64 result, TCGv src) static inline void gen_load_locked4u(TCGv dest, TCGv vaddr, int mem_index) { - tcg_gen_qemu_ld32u(dest, vaddr, mem_index); + tcg_gen_qemu_ld_tl(dest, vaddr, mem_index, MO_TEUL); tcg_gen_mov_tl(hex_llsc_addr, vaddr); tcg_gen_mov_tl(hex_llsc_val, dest); } static inline void gen_load_locked8u(TCGv_i64 dest, TCGv vaddr, int mem_index) { - tcg_gen_qemu_ld64(dest, vaddr, mem_index); + tcg_gen_qemu_ld_i64(dest, vaddr, mem_index, MO_TEUQ); tcg_gen_mov_tl(hex_llsc_addr, vaddr); tcg_gen_mov_i64(hex_llsc_val_i64, dest); } @@ -XXX,XX +XXX,XX @@ static void gen_load_frame(DisasContext *ctx, TCGv_i64 frame, TCGv EA) { Insn *insn = ctx->insn; /* Needed for CHECK_NOSHUF */ CHECK_NOSHUF(EA, 8); - tcg_gen_qemu_ld64(frame, EA, ctx->mem_idx); + tcg_gen_qemu_ld_i64(frame, EA, ctx->mem_idx, MO_TEUQ); } static void gen_return(DisasContext *ctx, TCGv_i64 dst, TCGv src) @@ -XXX,XX +XXX,XX @@ static void gen_vreg_load(DisasContext *ctx, intptr_t dstoff, TCGv src, tcg_gen_andi_tl(src, src, ~((int32_t)sizeof(MMVector) - 1)); } for (int i = 0; i < sizeof(MMVector) / 8; i++) { - tcg_gen_qemu_ld64(tmp, src, ctx->mem_idx); + tcg_gen_qemu_ld_i64(tmp, src, ctx->mem_idx, MO_TEUQ); tcg_gen_addi_tl(src, src, 8); tcg_gen_st_i64(tmp, cpu_env, dstoff + i * 8); } diff --git a/target/hexagon/idef-parser/parser-helpers.c b/target/hexagon/idef-parser/parser-helpers.c index XXXXXXX..XXXXXXX 100644 --- a/target/hexagon/idef-parser/parser-helpers.c +++ b/target/hexagon/idef-parser/parser-helpers.c @@ -XXX,XX +XXX,XX @@ void gen_load_cancel(Context *c, YYLTYPE *locp) void gen_load(Context *c, YYLTYPE *locp, HexValue *width, HexSignedness signedness, HexValue *ea, HexValue *dst) { - char size_suffix[4] = {0}; - const char *sign_suffix; + unsigned dst_bit_width; + unsigned src_bit_width; + /* Memop width is specified in the load macro */ assert_signedness(c, locp, signedness); - sign_suffix = (width->imm.value > 4) - ? "" - : ((signedness == UNSIGNED) ? "u" : "s"); + /* If dst is a variable, assert that is declared and load the type info */ if (dst->type == VARID) { find_variable(c, locp, dst, dst); } - snprintf(size_suffix, 4, "%" PRIu64, width->imm.value * 8); + src_bit_width = width->imm.value * 8; + dst_bit_width = MAX(dst->bit_width, 32); + /* Lookup the effective address EA */ find_variable(c, locp, ea, ea); OUT(c, locp, "if (insn->slot == 0 && pkt->pkt_has_store_s1) {\n"); OUT(c, locp, "probe_noshuf_load(", ea, ", ", width, ", ctx->mem_idx);\n"); OUT(c, locp, "process_store(ctx, 1);\n"); OUT(c, locp, "}\n"); - OUT(c, locp, "tcg_gen_qemu_ld", size_suffix, sign_suffix); + + OUT(c, locp, "tcg_gen_qemu_ld_i", &dst_bit_width); OUT(c, locp, "("); - if (dst->bit_width > width->imm.value * 8) { - /* - * Cast to the correct TCG type if necessary, to avoid implict cast - * warnings. This is needed when the width of the destination var is - * larger than the size of the requested load. - */ - OUT(c, locp, "(TCGv) "); + OUT(c, locp, dst, ", ", ea, ", ctx->mem_idx, MO_", &src_bit_width); + if (signedness == SIGNED) { + OUT(c, locp, " | MO_SIGN"); } - OUT(c, locp, dst, ", ", ea, ", ctx->mem_idx);\n"); + OUT(c, locp, " | MO_TE);\n"); } void gen_store(Context *c, YYLTYPE *locp, HexValue *width, HexValue *ea, diff --git a/target/hexagon/translate.c b/target/hexagon/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/hexagon/translate.c +++ b/target/hexagon/translate.c @@ -XXX,XX +XXX,XX @@ void process_store(DisasContext *ctx, int slot_num) switch (ctx->store_width[slot_num]) { case 1: gen_check_store_width(ctx, slot_num); - tcg_gen_qemu_st8(hex_store_val32[slot_num], - hex_store_addr[slot_num], - ctx->mem_idx); + tcg_gen_qemu_st_tl(hex_store_val32[slot_num], + hex_store_addr[slot_num], + ctx->mem_idx, MO_UB); break; case 2: gen_check_store_width(ctx, slot_num); - tcg_gen_qemu_st16(hex_store_val32[slot_num], - hex_store_addr[slot_num], - ctx->mem_idx); + tcg_gen_qemu_st_tl(hex_store_val32[slot_num], + hex_store_addr[slot_num], + ctx->mem_idx, MO_TEUW); break; case 4: gen_check_store_width(ctx, slot_num); - tcg_gen_qemu_st32(hex_store_val32[slot_num], - hex_store_addr[slot_num], - ctx->mem_idx); + tcg_gen_qemu_st_tl(hex_store_val32[slot_num], + hex_store_addr[slot_num], + ctx->mem_idx, MO_TEUL); break; case 8: gen_check_store_width(ctx, slot_num); - tcg_gen_qemu_st64(hex_store_val64[slot_num], - hex_store_addr[slot_num], - ctx->mem_idx); + tcg_gen_qemu_st_i64(hex_store_val64[slot_num], + hex_store_addr[slot_num], + ctx->mem_idx, MO_TEUQ); break; default: { @@ -XXX,XX +XXX,XX @@ static void process_dczeroa(DisasContext *ctx) TCGv_i64 zero = tcg_constant_i64(0); tcg_gen_andi_tl(addr, hex_dczero_addr, ~0x1f); - tcg_gen_qemu_st64(zero, addr, ctx->mem_idx); + tcg_gen_qemu_st_i64(zero, addr, ctx->mem_idx, MO_UQ); tcg_gen_addi_tl(addr, addr, 8); - tcg_gen_qemu_st64(zero, addr, ctx->mem_idx); + tcg_gen_qemu_st_i64(zero, addr, ctx->mem_idx, MO_UQ); tcg_gen_addi_tl(addr, addr, 8); - tcg_gen_qemu_st64(zero, addr, ctx->mem_idx); + tcg_gen_qemu_st_i64(zero, addr, ctx->mem_idx, MO_UQ); tcg_gen_addi_tl(addr, addr, 8); - tcg_gen_qemu_st64(zero, addr, ctx->mem_idx); + tcg_gen_qemu_st_i64(zero, addr, ctx->mem_idx, MO_UQ); } } -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-5-richard.henderson@linaro.org> --- target/m68k/translate.c | 76 ++++++++++++++--------------------------- 1 file changed, 25 insertions(+), 51 deletions(-) diff --git a/target/m68k/translate.c b/target/m68k/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/m68k/translate.c +++ b/target/m68k/translate.c @@ -XXX,XX +XXX,XX @@ static inline void gen_addr_fault(DisasContext *s) static inline TCGv gen_load(DisasContext *s, int opsize, TCGv addr, int sign, int index) { - TCGv tmp; - tmp = tcg_temp_new_i32(); - switch(opsize) { + TCGv tmp = tcg_temp_new_i32(); + + switch (opsize) { case OS_BYTE: - if (sign) - tcg_gen_qemu_ld8s(tmp, addr, index); - else - tcg_gen_qemu_ld8u(tmp, addr, index); - break; case OS_WORD: - if (sign) - tcg_gen_qemu_ld16s(tmp, addr, index); - else - tcg_gen_qemu_ld16u(tmp, addr, index); - break; case OS_LONG: - tcg_gen_qemu_ld32u(tmp, addr, index); + tcg_gen_qemu_ld_tl(tmp, addr, index, + opsize | (sign ? MO_SIGN : 0) | MO_TE); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static inline TCGv gen_load(DisasContext *s, int opsize, TCGv addr, static inline void gen_store(DisasContext *s, int opsize, TCGv addr, TCGv val, int index) { - switch(opsize) { + switch (opsize) { case OS_BYTE: - tcg_gen_qemu_st8(val, addr, index); - break; case OS_WORD: - tcg_gen_qemu_st16(val, addr, index); - break; case OS_LONG: - tcg_gen_qemu_st32(val, addr, index); + tcg_gen_qemu_st_tl(val, addr, index, opsize | MO_TE); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void gen_load_fp(DisasContext *s, int opsize, TCGv addr, TCGv_ptr fp, tmp = tcg_temp_new(); switch (opsize) { case OS_BYTE: - tcg_gen_qemu_ld8s(tmp, addr, index); - gen_helper_exts32(cpu_env, fp, tmp); - break; case OS_WORD: - tcg_gen_qemu_ld16s(tmp, addr, index); - gen_helper_exts32(cpu_env, fp, tmp); - break; - case OS_LONG: - tcg_gen_qemu_ld32u(tmp, addr, index); + tcg_gen_qemu_ld_tl(tmp, addr, index, opsize | MO_SIGN | MO_TE); gen_helper_exts32(cpu_env, fp, tmp); break; case OS_SINGLE: - tcg_gen_qemu_ld32u(tmp, addr, index); + tcg_gen_qemu_ld_tl(tmp, addr, index, MO_TEUL); gen_helper_extf32(cpu_env, fp, tmp); break; case OS_DOUBLE: - tcg_gen_qemu_ld64(t64, addr, index); + tcg_gen_qemu_ld_i64(t64, addr, index, MO_TEUQ); gen_helper_extf64(cpu_env, fp, t64); break; case OS_EXTENDED: @@ -XXX,XX +XXX,XX @@ static void gen_load_fp(DisasContext *s, int opsize, TCGv addr, TCGv_ptr fp, gen_exception(s, s->base.pc_next, EXCP_FP_UNIMP); break; } - tcg_gen_qemu_ld32u(tmp, addr, index); + tcg_gen_qemu_ld_i32(tmp, addr, index, MO_TEUL); tcg_gen_shri_i32(tmp, tmp, 16); tcg_gen_st16_i32(tmp, fp, offsetof(FPReg, l.upper)); tcg_gen_addi_i32(tmp, addr, 4); - tcg_gen_qemu_ld64(t64, tmp, index); + tcg_gen_qemu_ld_i64(t64, tmp, index, MO_TEUQ); tcg_gen_st_i64(t64, fp, offsetof(FPReg, l.lower)); break; case OS_PACKED: @@ -XXX,XX +XXX,XX @@ static void gen_store_fp(DisasContext *s, int opsize, TCGv addr, TCGv_ptr fp, tmp = tcg_temp_new(); switch (opsize) { case OS_BYTE: - gen_helper_reds32(tmp, cpu_env, fp); - tcg_gen_qemu_st8(tmp, addr, index); - break; case OS_WORD: - gen_helper_reds32(tmp, cpu_env, fp); - tcg_gen_qemu_st16(tmp, addr, index); - break; case OS_LONG: gen_helper_reds32(tmp, cpu_env, fp); - tcg_gen_qemu_st32(tmp, addr, index); + tcg_gen_qemu_st_tl(tmp, addr, index, opsize | MO_TE); break; case OS_SINGLE: gen_helper_redf32(tmp, cpu_env, fp); - tcg_gen_qemu_st32(tmp, addr, index); + tcg_gen_qemu_st_tl(tmp, addr, index, MO_TEUL); break; case OS_DOUBLE: gen_helper_redf64(t64, cpu_env, fp); - tcg_gen_qemu_st64(t64, addr, index); + tcg_gen_qemu_st_i64(t64, addr, index, MO_TEUQ); break; case OS_EXTENDED: if (m68k_feature(s->env, M68K_FEATURE_CF_FPU)) { @@ -XXX,XX +XXX,XX @@ static void gen_store_fp(DisasContext *s, int opsize, TCGv addr, TCGv_ptr fp, } tcg_gen_ld16u_i32(tmp, fp, offsetof(FPReg, l.upper)); tcg_gen_shli_i32(tmp, tmp, 16); - tcg_gen_qemu_st32(tmp, addr, index); + tcg_gen_qemu_st_i32(tmp, addr, index, MO_TEUL); tcg_gen_addi_i32(tmp, addr, 4); tcg_gen_ld_i64(t64, fp, offsetof(FPReg, l.lower)); - tcg_gen_qemu_st64(t64, tmp, index); + tcg_gen_qemu_st_i64(t64, tmp, index, MO_TEUQ); break; case OS_PACKED: /* @@ -XXX,XX +XXX,XX @@ DISAS_INSN(movep) if (insn & 0x80) { for ( ; i > 0 ; i--) { tcg_gen_shri_i32(dbuf, reg, (i - 1) * 8); - tcg_gen_qemu_st8(dbuf, abuf, IS_USER(s)); + tcg_gen_qemu_st_i32(dbuf, abuf, IS_USER(s), MO_UB); if (i > 1) { tcg_gen_addi_i32(abuf, abuf, 2); } } } else { for ( ; i > 0 ; i--) { - tcg_gen_qemu_ld8u(dbuf, abuf, IS_USER(s)); + tcg_gen_qemu_ld_tl(dbuf, abuf, IS_USER(s), MO_UB); tcg_gen_deposit_i32(reg, reg, dbuf, (i - 1) * 8, 8); if (i > 1) { tcg_gen_addi_i32(abuf, abuf, 2); @@ -XXX,XX +XXX,XX @@ static void m68k_copy_line(TCGv dst, TCGv src, int index) t1 = tcg_temp_new_i64(); tcg_gen_andi_i32(addr, src, ~15); - tcg_gen_qemu_ld64(t0, addr, index); + tcg_gen_qemu_ld_i64(t0, addr, index, MO_TEUQ); tcg_gen_addi_i32(addr, addr, 8); - tcg_gen_qemu_ld64(t1, addr, index); + tcg_gen_qemu_ld_i64(t1, addr, index, MO_TEUQ); tcg_gen_andi_i32(addr, dst, ~15); - tcg_gen_qemu_st64(t0, addr, index); + tcg_gen_qemu_st_i64(t0, addr, index, MO_TEUQ); tcg_gen_addi_i32(addr, addr, 8); - tcg_gen_qemu_st64(t1, addr, index); + tcg_gen_qemu_st_i64(t1, addr, index, MO_TEUQ); } DISAS_INSN(move16_reg) @@ -XXX,XX +XXX,XX @@ static void gen_qemu_store_fcr(DisasContext *s, TCGv addr, int reg) tmp = tcg_temp_new(); gen_load_fcr(s, tmp, reg); - tcg_gen_qemu_st32(tmp, addr, index); + tcg_gen_qemu_st_tl(tmp, addr, index, MO_TEUL); } static void gen_qemu_load_fcr(DisasContext *s, TCGv addr, int reg) @@ -XXX,XX +XXX,XX @@ static void gen_qemu_load_fcr(DisasContext *s, TCGv addr, int reg) TCGv tmp; tmp = tcg_temp_new(); - tcg_gen_qemu_ld32u(tmp, addr, index); + tcg_gen_qemu_ld_tl(tmp, addr, index, MO_TEUL); gen_store_fcr(s, tmp, reg); } -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-6-richard.henderson@linaro.org> --- target/mips/tcg/translate.c | 8 ++++---- target/mips/tcg/nanomips_translate.c.inc | 2 +- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git a/target/mips/tcg/translate.c b/target/mips/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/mips/tcg/translate.c +++ b/target/mips/tcg/translate.c @@ -XXX,XX +XXX,XX @@ FOP_CONDNS(s, FMT_S, 32, gen_store_fpr32(ctx, fp0, fd)) /* load/store instructions. */ #ifdef CONFIG_USER_ONLY -#define OP_LD_ATOMIC(insn, fname) \ +#define OP_LD_ATOMIC(insn, memop) \ static inline void op_ld_##insn(TCGv ret, TCGv arg1, int mem_idx, \ DisasContext *ctx) \ { \ TCGv t0 = tcg_temp_new(); \ tcg_gen_mov_tl(t0, arg1); \ - tcg_gen_qemu_##fname(ret, arg1, ctx->mem_idx); \ + tcg_gen_qemu_ld_tl(ret, arg1, ctx->mem_idx, memop); \ tcg_gen_st_tl(t0, cpu_env, offsetof(CPUMIPSState, lladdr)); \ tcg_gen_st_tl(ret, cpu_env, offsetof(CPUMIPSState, llval)); \ } @@ -XXX,XX +XXX,XX @@ static inline void op_ld_##insn(TCGv ret, TCGv arg1, int mem_idx, \ gen_helper_##insn(ret, cpu_env, arg1, tcg_constant_i32(mem_idx)); \ } #endif -OP_LD_ATOMIC(ll, ld32s); +OP_LD_ATOMIC(ll, MO_TESL); #if defined(TARGET_MIPS64) -OP_LD_ATOMIC(lld, ld64); +OP_LD_ATOMIC(lld, MO_TEUQ); #endif #undef OP_LD_ATOMIC diff --git a/target/mips/tcg/nanomips_translate.c.inc b/target/mips/tcg/nanomips_translate.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/mips/tcg/nanomips_translate.c.inc +++ b/target/mips/tcg/nanomips_translate.c.inc @@ -XXX,XX +XXX,XX @@ static void gen_llwp(DisasContext *ctx, uint32_t base, int16_t offset, TCGv tmp2 = tcg_temp_new(); gen_base_offset_addr(ctx, taddr, base, offset); - tcg_gen_qemu_ld64(tval, taddr, ctx->mem_idx); + tcg_gen_qemu_ld_i64(tval, taddr, ctx->mem_idx, MO_TEUQ); if (cpu_is_bigendian(ctx)) { tcg_gen_extr_i64_tl(tmp2, tmp1, tval); } else { -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: David Hildenbrand <david@redhat.com> Reviewed-by: Ilya Leoshkevich <iii@linux.ibm.com> Message-Id: <20230502135741.1158035-7-richard.henderson@linaro.org> --- target/s390x/tcg/translate.c | 152 ++++++++++++++++------------------- 1 file changed, 71 insertions(+), 81 deletions(-) diff --git a/target/s390x/tcg/translate.c b/target/s390x/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/s390x/tcg/translate.c +++ b/target/s390x/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_clc(DisasContext *s, DisasOps *o) { int l = get_field(s, l1); TCGv_i32 vl; + MemOp mop; switch (l + 1) { case 1: - tcg_gen_qemu_ld8u(cc_src, o->addr1, get_mem_index(s)); - tcg_gen_qemu_ld8u(cc_dst, o->in2, get_mem_index(s)); - break; case 2: - tcg_gen_qemu_ld16u(cc_src, o->addr1, get_mem_index(s)); - tcg_gen_qemu_ld16u(cc_dst, o->in2, get_mem_index(s)); - break; case 4: - tcg_gen_qemu_ld32u(cc_src, o->addr1, get_mem_index(s)); - tcg_gen_qemu_ld32u(cc_dst, o->in2, get_mem_index(s)); - break; case 8: - tcg_gen_qemu_ld64(cc_src, o->addr1, get_mem_index(s)); - tcg_gen_qemu_ld64(cc_dst, o->in2, get_mem_index(s)); - break; + mop = ctz32(l + 1) | MO_TE; + tcg_gen_qemu_ld_tl(cc_src, o->addr1, get_mem_index(s), mop); + tcg_gen_qemu_ld_tl(cc_dst, o->in2, get_mem_index(s), mop); + gen_op_update2_cc_i64(s, CC_OP_LTUGTU_64, cc_src, cc_dst); + return DISAS_NEXT; default: vl = tcg_constant_i32(l); gen_helper_clc(cc_op, cpu_env, vl, o->addr1, o->in2); set_cc_static(s); return DISAS_NEXT; } - gen_op_update2_cc_i64(s, CC_OP_LTUGTU_64, cc_src, cc_dst); - return DISAS_NEXT; } static DisasJumpType op_clcl(DisasContext *s, DisasOps *o) @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_cvd(DisasContext *s, DisasOps *o) TCGv_i32 t2 = tcg_temp_new_i32(); tcg_gen_extrl_i64_i32(t2, o->in1); gen_helper_cvd(t1, t2); - tcg_gen_qemu_st64(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(t1, o->in2, get_mem_index(s), MO_TEUQ); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_icm(DisasContext *s, DisasOps *o) switch (m3) { case 0xf: /* Effectively a 32-bit load. */ - tcg_gen_qemu_ld32u(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(tmp, o->in2, get_mem_index(s), MO_TEUL); len = 32; goto one_insert; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_icm(DisasContext *s, DisasOps *o) case 0x6: case 0x3: /* Effectively a 16-bit load. */ - tcg_gen_qemu_ld16u(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(tmp, o->in2, get_mem_index(s), MO_TEUW); len = 16; goto one_insert; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_icm(DisasContext *s, DisasOps *o) case 0x2: case 0x1: /* Effectively an 8-bit load. */ - tcg_gen_qemu_ld8u(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(tmp, o->in2, get_mem_index(s), MO_UB); len = 8; goto one_insert; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_icm(DisasContext *s, DisasOps *o) ccm = 0; while (m3) { if (m3 & 0x8) { - tcg_gen_qemu_ld8u(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(tmp, o->in2, get_mem_index(s), MO_UB); tcg_gen_addi_i64(o->in2, o->in2, 1); tcg_gen_deposit_i64(o->out, o->out, tmp, pos, 8); ccm |= 0xffull << pos; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_llgt(DisasContext *s, DisasOps *o) static DisasJumpType op_ld8s(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_ld8s(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_SB); return DISAS_NEXT; } static DisasJumpType op_ld8u(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_ld8u(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_UB); return DISAS_NEXT; } static DisasJumpType op_ld16s(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_ld16s(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_TESW); return DISAS_NEXT; } static DisasJumpType op_ld16u(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_ld16u(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_TEUW); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lat(DisasContext *s, DisasOps *o) static DisasJumpType op_lgat(DisasContext *s, DisasOps *o) { TCGLabel *lab = gen_new_label(); - tcg_gen_qemu_ld64(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_TEUQ); /* The value is stored even in case of trap. */ tcg_gen_brcondi_i64(TCG_COND_NE, o->out, 0, lab); gen_trap(s); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lfhat(DisasContext *s, DisasOps *o) static DisasJumpType op_llgfat(DisasContext *s, DisasOps *o) { TCGLabel *lab = gen_new_label(); - tcg_gen_qemu_ld32u(o->out, o->in2, get_mem_index(s)); + + tcg_gen_qemu_ld_i64(o->out, o->in2, get_mem_index(s), MO_TEUL); /* The value is stored even in case of trap. */ tcg_gen_brcondi_i64(TCG_COND_NE, o->out, 0, lab); gen_trap(s); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lpswe(DisasContext *s, DisasOps *o) tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUQ | MO_ALIGN_8); tcg_gen_addi_i64(o->in2, o->in2, 8); - tcg_gen_qemu_ld64(t2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t2, o->in2, get_mem_index(s), MO_TEUQ); gen_helper_load_psw(cpu_env, t1, t2); return DISAS_NORETURN; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm32(DisasContext *s, DisasOps *o) /* Only one register to read. */ t1 = tcg_temp_new_i64(); if (unlikely(r1 == r3)) { - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); store_reg32_i64(r1, t1); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm32(DisasContext *s, DisasOps *o) /* First load the values of the first and last registers to trigger possible page faults. */ t2 = tcg_temp_new_i64(); - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); tcg_gen_addi_i64(t2, o->in2, 4 * ((r3 - r1) & 15)); - tcg_gen_qemu_ld32u(t2, t2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t2, t2, get_mem_index(s), MO_TEUL); store_reg32_i64(r1, t1); store_reg32_i64(r3, t2); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm32(DisasContext *s, DisasOps *o) while (r1 != r3) { r1 = (r1 + 1) & 15; tcg_gen_add_i64(o->in2, o->in2, t2); - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); store_reg32_i64(r1, t1); } return DISAS_NEXT; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lmh(DisasContext *s, DisasOps *o) /* Only one register to read. */ t1 = tcg_temp_new_i64(); if (unlikely(r1 == r3)) { - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); store_reg32h_i64(r1, t1); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lmh(DisasContext *s, DisasOps *o) /* First load the values of the first and last registers to trigger possible page faults. */ t2 = tcg_temp_new_i64(); - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); tcg_gen_addi_i64(t2, o->in2, 4 * ((r3 - r1) & 15)); - tcg_gen_qemu_ld32u(t2, t2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t2, t2, get_mem_index(s), MO_TEUL); store_reg32h_i64(r1, t1); store_reg32h_i64(r3, t2); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lmh(DisasContext *s, DisasOps *o) while (r1 != r3) { r1 = (r1 + 1) & 15; tcg_gen_add_i64(o->in2, o->in2, t2); - tcg_gen_qemu_ld32u(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUL); store_reg32h_i64(r1, t1); } return DISAS_NEXT; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm64(DisasContext *s, DisasOps *o) /* Only one register to read. */ if (unlikely(r1 == r3)) { - tcg_gen_qemu_ld64(regs[r1], o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(regs[r1], o->in2, get_mem_index(s), MO_TEUQ); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm64(DisasContext *s, DisasOps *o) possible page faults. */ t1 = tcg_temp_new_i64(); t2 = tcg_temp_new_i64(); - tcg_gen_qemu_ld64(t1, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(t1, o->in2, get_mem_index(s), MO_TEUQ); tcg_gen_addi_i64(t2, o->in2, 8 * ((r3 - r1) & 15)); - tcg_gen_qemu_ld64(regs[r3], t2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(regs[r3], t2, get_mem_index(s), MO_TEUQ); tcg_gen_mov_i64(regs[r1], t1); /* Only two registers to read. */ @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_lm64(DisasContext *s, DisasOps *o) while (r1 != r3) { r1 = (r1 + 1) & 15; tcg_gen_add_i64(o->in2, o->in2, t1); - tcg_gen_qemu_ld64(regs[r1], o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(regs[r1], o->in2, get_mem_index(s), MO_TEUQ); } return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_soc(DisasContext *s, DisasOps *o) a = get_address(s, 0, get_field(s, b2), get_field(s, d2)); switch (s->insn->data) { case 1: /* STOCG */ - tcg_gen_qemu_st64(regs[r1], a, get_mem_index(s)); + tcg_gen_qemu_st_i64(regs[r1], a, get_mem_index(s), MO_TEUQ); break; case 0: /* STOC */ - tcg_gen_qemu_st32(regs[r1], a, get_mem_index(s)); + tcg_gen_qemu_st_i64(regs[r1], a, get_mem_index(s), MO_TEUL); break; case 2: /* STOCFH */ h = tcg_temp_new_i64(); tcg_gen_shri_i64(h, regs[r1], 32); - tcg_gen_qemu_st32(h, a, get_mem_index(s)); + tcg_gen_qemu_st_i64(h, a, get_mem_index(s), MO_TEUL); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_ectg(DisasContext *s, DisasOps *o) gen_addi_and_wrap_i64(s, o->addr1, regs[r3], 0); /* load the third operand into r3 before modifying anything */ - tcg_gen_qemu_ld64(regs[r3], o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(regs[r3], o->addr1, get_mem_index(s), MO_TEUQ); /* subtract CPU timer from first operand and store in GR0 */ gen_helper_stpt(tmp, cpu_env); @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stcke(DisasContext *s, DisasOps *o) tcg_gen_shri_i64(c1, c1, 8); tcg_gen_ori_i64(c2, c2, 0x10000); tcg_gen_or_i64(c2, c2, todpr); - tcg_gen_qemu_st64(c1, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(c1, o->in2, get_mem_index(s), MO_TEUQ); tcg_gen_addi_i64(o->in2, o->in2, 8); - tcg_gen_qemu_st64(c2, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(c2, o->in2, get_mem_index(s), MO_TEUQ); /* ??? We don't implement clock states. */ gen_op_movi_cc(s, 0); return DISAS_NEXT; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stnosm(DisasContext *s, DisasOps *o) restart, we'll have the wrong SYSTEM MASK in place. */ t = tcg_temp_new_i64(); tcg_gen_shri_i64(t, psw_mask, 56); - tcg_gen_qemu_st8(t, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(t, o->addr1, get_mem_index(s), MO_UB); if (s->fields.op == 0xac) { tcg_gen_andi_i64(psw_mask, psw_mask, @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stfle(DisasContext *s, DisasOps *o) static DisasJumpType op_st8(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st8(o->in1, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in1, o->in2, get_mem_index(s), MO_UB); return DISAS_NEXT; } static DisasJumpType op_st16(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st16(o->in1, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in1, o->in2, get_mem_index(s), MO_TEUW); return DISAS_NEXT; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stcm(DisasContext *s, DisasOps *o) case 0xf: /* Effectively a 32-bit store. */ tcg_gen_shri_i64(tmp, o->in1, pos); - tcg_gen_qemu_st32(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(tmp, o->in2, get_mem_index(s), MO_TEUL); break; case 0xc: @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stcm(DisasContext *s, DisasOps *o) case 0x3: /* Effectively a 16-bit store. */ tcg_gen_shri_i64(tmp, o->in1, pos); - tcg_gen_qemu_st16(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(tmp, o->in2, get_mem_index(s), MO_TEUW); break; case 0x8: @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stcm(DisasContext *s, DisasOps *o) case 0x1: /* Effectively an 8-bit store. */ tcg_gen_shri_i64(tmp, o->in1, pos); - tcg_gen_qemu_st8(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(tmp, o->in2, get_mem_index(s), MO_UB); break; default: @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stcm(DisasContext *s, DisasOps *o) while (m3) { if (m3 & 0x8) { tcg_gen_shri_i64(tmp, o->in1, pos); - tcg_gen_qemu_st8(tmp, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(tmp, o->in2, get_mem_index(s), MO_UB); tcg_gen_addi_i64(o->in2, o->in2, 1); } m3 = (m3 << 1) & 0xf; @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stm(DisasContext *s, DisasOps *o) TCGv_i64 tsize = tcg_constant_i64(size); while (1) { - if (size == 8) { - tcg_gen_qemu_st64(regs[r1], o->in2, get_mem_index(s)); - } else { - tcg_gen_qemu_st32(regs[r1], o->in2, get_mem_index(s)); - } + tcg_gen_qemu_st_i64(regs[r1], o->in2, get_mem_index(s), + size == 8 ? MO_TEUQ : MO_TEUL); if (r1 == r3) { break; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_stmh(DisasContext *s, DisasOps *o) while (1) { tcg_gen_shl_i64(t, regs[r1], t32); - tcg_gen_qemu_st32(t, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(t, o->in2, get_mem_index(s), MO_TEUL); if (r1 == r3) { break; } @@ -XXX,XX +XXX,XX @@ static DisasJumpType op_xc(DisasContext *s, DisasOps *o) l++; while (l >= 8) { - tcg_gen_qemu_st64(o->in2, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in2, o->addr1, get_mem_index(s), MO_UQ); l -= 8; if (l > 0) { tcg_gen_addi_i64(o->addr1, o->addr1, 8); } } if (l >= 4) { - tcg_gen_qemu_st32(o->in2, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in2, o->addr1, get_mem_index(s), MO_UL); l -= 4; if (l > 0) { tcg_gen_addi_i64(o->addr1, o->addr1, 4); } } if (l >= 2) { - tcg_gen_qemu_st16(o->in2, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in2, o->addr1, get_mem_index(s), MO_UW); l -= 2; if (l > 0) { tcg_gen_addi_i64(o->addr1, o->addr1, 2); } } if (l) { - tcg_gen_qemu_st8(o->in2, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->in2, o->addr1, get_mem_index(s), MO_UB); } gen_op_movi_cc(s, 0); return DISAS_NEXT; @@ -XXX,XX +XXX,XX @@ static void wout_cond_e1e2(DisasContext *s, DisasOps *o) static void wout_m1_8(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st8(o->out, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->out, o->addr1, get_mem_index(s), MO_UB); } #define SPEC_wout_m1_8 0 static void wout_m1_16(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st16(o->out, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->out, o->addr1, get_mem_index(s), MO_TEUW); } #define SPEC_wout_m1_16 0 @@ -XXX,XX +XXX,XX @@ static void wout_m1_16a(DisasContext *s, DisasOps *o) static void wout_m1_32(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st32(o->out, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->out, o->addr1, get_mem_index(s), MO_TEUL); } #define SPEC_wout_m1_32 0 @@ -XXX,XX +XXX,XX @@ static void wout_m1_32a(DisasContext *s, DisasOps *o) static void wout_m1_64(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st64(o->out, o->addr1, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->out, o->addr1, get_mem_index(s), MO_TEUQ); } #define SPEC_wout_m1_64 0 @@ -XXX,XX +XXX,XX @@ static void wout_m1_64a(DisasContext *s, DisasOps *o) static void wout_m2_32(DisasContext *s, DisasOps *o) { - tcg_gen_qemu_st32(o->out, o->in2, get_mem_index(s)); + tcg_gen_qemu_st_i64(o->out, o->in2, get_mem_index(s), MO_TEUL); } #define SPEC_wout_m2_32 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_8u(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld8u(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_UB); } #define SPEC_in1_m1_8u 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_16s(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld16s(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_TESW); } #define SPEC_in1_m1_16s 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_16u(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld16u(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_TEUW); } #define SPEC_in1_m1_16u 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_32s(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld32s(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_TESL); } #define SPEC_in1_m1_32s 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_32u(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld32u(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_TEUL); } #define SPEC_in1_m1_32u 0 @@ -XXX,XX +XXX,XX @@ static void in1_m1_64(DisasContext *s, DisasOps *o) { in1_la1(s, o); o->in1 = tcg_temp_new_i64(); - tcg_gen_qemu_ld64(o->in1, o->addr1, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in1, o->addr1, get_mem_index(s), MO_TEUQ); } #define SPEC_in1_m1_64 0 @@ -XXX,XX +XXX,XX @@ static void in2_sh(DisasContext *s, DisasOps *o) static void in2_m2_8u(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld8u(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_UB); } #define SPEC_in2_m2_8u 0 static void in2_m2_16s(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld16s(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TESW); } #define SPEC_in2_m2_16s 0 static void in2_m2_16u(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld16u(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TEUW); } #define SPEC_in2_m2_16u 0 static void in2_m2_32s(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld32s(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TESL); } #define SPEC_in2_m2_32s 0 static void in2_m2_32u(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld32u(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TEUL); } #define SPEC_in2_m2_32u 0 @@ -XXX,XX +XXX,XX @@ static void in2_m2_32ua(DisasContext *s, DisasOps *o) static void in2_m2_64(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld64(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TEUQ); } #define SPEC_in2_m2_64 0 static void in2_m2_64w(DisasContext *s, DisasOps *o) { in2_a2(s, o); - tcg_gen_qemu_ld64(o->in2, o->in2, get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, o->in2, get_mem_index(s), MO_TEUQ); gen_addi_and_wrap_i64(s, o->in2, o->in2, 0); } #define SPEC_in2_m2_64w 0 @@ -XXX,XX +XXX,XX @@ static void in2_m2_64a(DisasContext *s, DisasOps *o) static void in2_mri2_16s(DisasContext *s, DisasOps *o) { o->in2 = tcg_temp_new_i64(); - tcg_gen_qemu_ld16s(o->in2, gen_ri2(s), get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, gen_ri2(s), get_mem_index(s), MO_TESW); } #define SPEC_in2_mri2_16s 0 static void in2_mri2_16u(DisasContext *s, DisasOps *o) { o->in2 = tcg_temp_new_i64(); - tcg_gen_qemu_ld16u(o->in2, gen_ri2(s), get_mem_index(s)); + tcg_gen_qemu_ld_i64(o->in2, gen_ri2(s), get_mem_index(s), MO_TEUW); } #define SPEC_in2_mri2_16u 0 -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Anton Johansson <anjo@rev.ng> Message-Id: <20230502135741.1158035-8-richard.henderson@linaro.org> --- target/sparc/translate.c | 43 ++++++++++++++++++++++++++-------------- 1 file changed, 28 insertions(+), 15 deletions(-) diff --git a/target/sparc/translate.c b/target/sparc/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/translate.c +++ b/target/sparc/translate.c @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) switch (xop) { case 0x0: /* ld, V9 lduw, load unsigned word */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld32u(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUL); break; case 0x1: /* ldub, load unsigned byte */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld8u(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_UB); break; case 0x2: /* lduh, load unsigned halfword */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld16u(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUW); break; case 0x3: /* ldd, load double word */ if (rd & 1) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) gen_address_mask(dc, cpu_addr); t64 = tcg_temp_new_i64(); - tcg_gen_qemu_ld64(t64, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_i64(t64, cpu_addr, + dc->mem_idx, MO_TEUQ); tcg_gen_trunc_i64_tl(cpu_val, t64); tcg_gen_ext32u_tl(cpu_val, cpu_val); gen_store_gpr(dc, rd + 1, cpu_val); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) break; case 0x9: /* ldsb, load signed byte */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld8s(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, dc->mem_idx, MO_SB); break; case 0xa: /* ldsh, load signed halfword */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld16s(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TESW); break; case 0xd: /* ldstub */ gen_ldstub(dc, cpu_val, cpu_addr, dc->mem_idx); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) #ifdef TARGET_SPARC64 case 0x08: /* V9 ldsw */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld32s(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TESL); break; case 0x0b: /* V9 ldx */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_ld64(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUQ); break; case 0x18: /* V9 ldswa */ gen_ld_asi(dc, cpu_val, cpu_addr, insn, MO_TESL); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) switch (xop) { case 0x4: /* st, store word */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_st32(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUL); break; case 0x5: /* stb, store byte */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_st8(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_val, cpu_addr, dc->mem_idx, MO_UB); break; case 0x6: /* sth, store halfword */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_st16(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUW); break; case 0x7: /* std, store double word */ if (rd & 1) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) lo = gen_load_gpr(dc, rd + 1); t64 = tcg_temp_new_i64(); tcg_gen_concat_tl_i64(t64, lo, cpu_val); - tcg_gen_qemu_st64(t64, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_i64(t64, cpu_addr, + dc->mem_idx, MO_TEUQ); } break; #if !defined(CONFIG_USER_ONLY) || defined(TARGET_SPARC64) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) #ifdef TARGET_SPARC64 case 0x0e: /* V9 stx */ gen_address_mask(dc, cpu_addr); - tcg_gen_qemu_st64(cpu_val, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_val, cpu_addr, + dc->mem_idx, MO_TEUQ); break; case 0x1e: /* V9 stxa */ gen_st_asi(dc, cpu_val, cpu_addr, insn, MO_TEUQ); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) #ifdef TARGET_SPARC64 gen_address_mask(dc, cpu_addr); if (rd == 1) { - tcg_gen_qemu_st64(cpu_fsr, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_fsr, cpu_addr, + dc->mem_idx, MO_TEUQ); break; } #endif - tcg_gen_qemu_st32(cpu_fsr, cpu_addr, dc->mem_idx); + tcg_gen_qemu_st_tl(cpu_fsr, cpu_addr, + dc->mem_idx, MO_TEUL); } break; case 0x26: -- 2.34.1 Convert away from the old interface with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Max Filippov <jcmvbkbc@gmail.com> Message-Id: <20230502135741.1158035-9-richard.henderson@linaro.org> --- target/xtensa/translate.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/target/xtensa/translate.c b/target/xtensa/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/xtensa/translate.c +++ b/target/xtensa/translate.c @@ -XXX,XX +XXX,XX @@ static void translate_dcache(DisasContext *dc, const OpcodeArg arg[], TCGv_i32 res = tcg_temp_new_i32(); tcg_gen_addi_i32(addr, arg[0].in, arg[1].imm); - tcg_gen_qemu_ld8u(res, addr, dc->cring); + tcg_gen_qemu_ld_i32(res, addr, dc->cring, MO_UB); } static void translate_depbits(DisasContext *dc, const OpcodeArg arg[], @@ -XXX,XX +XXX,XX @@ static void translate_l32r(DisasContext *dc, const OpcodeArg arg[], } else { tmp = tcg_constant_i32(arg[1].imm); } - tcg_gen_qemu_ld32u(arg[0].out, tmp, dc->cring); + tcg_gen_qemu_ld_i32(arg[0].out, tmp, dc->cring, MO_TEUL); } static void translate_loop(DisasContext *dc, const OpcodeArg arg[], -- 2.34.1 Remove the old interfaces with the implicit MemOp argument. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Acked-by: David Hildenbrand <david@redhat.com> Message-Id: <20230502135741.1158035-10-richard.henderson@linaro.org> --- include/tcg/tcg-op.h | 55 -------------------------------------------- 1 file changed, 55 deletions(-) diff --git a/include/tcg/tcg-op.h b/include/tcg/tcg-op.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg-op.h +++ b/include/tcg/tcg-op.h @@ -XXX,XX +XXX,XX @@ void tcg_gen_qemu_st_i64(TCGv_i64, TCGv, TCGArg, MemOp); void tcg_gen_qemu_ld_i128(TCGv_i128, TCGv, TCGArg, MemOp); void tcg_gen_qemu_st_i128(TCGv_i128, TCGv, TCGArg, MemOp); -static inline void tcg_gen_qemu_ld8u(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_UB); -} - -static inline void tcg_gen_qemu_ld8s(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_SB); -} - -static inline void tcg_gen_qemu_ld16u(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_TEUW); -} - -static inline void tcg_gen_qemu_ld16s(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_TESW); -} - -static inline void tcg_gen_qemu_ld32u(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_TEUL); -} - -static inline void tcg_gen_qemu_ld32s(TCGv ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_tl(ret, addr, mem_index, MO_TESL); -} - -static inline void tcg_gen_qemu_ld64(TCGv_i64 ret, TCGv addr, int mem_index) -{ - tcg_gen_qemu_ld_i64(ret, addr, mem_index, MO_TEUQ); -} - -static inline void tcg_gen_qemu_st8(TCGv arg, TCGv addr, int mem_index) -{ - tcg_gen_qemu_st_tl(arg, addr, mem_index, MO_UB); -} - -static inline void tcg_gen_qemu_st16(TCGv arg, TCGv addr, int mem_index) -{ - tcg_gen_qemu_st_tl(arg, addr, mem_index, MO_TEUW); -} - -static inline void tcg_gen_qemu_st32(TCGv arg, TCGv addr, int mem_index) -{ - tcg_gen_qemu_st_tl(arg, addr, mem_index, MO_TEUL); -} - -static inline void tcg_gen_qemu_st64(TCGv_i64 arg, TCGv addr, int mem_index) -{ - tcg_gen_qemu_st_i64(arg, addr, mem_index, MO_TEUQ); -} - void tcg_gen_atomic_cmpxchg_i32(TCGv_i32, TCGv, TCGv_i32, TCGv_i32, TCGArg, MemOp); void tcg_gen_atomic_cmpxchg_i64(TCGv_i64, TCGv, TCGv_i64, TCGv_i64, -- 2.34.1 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/alpha/translate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/alpha/translate.c b/target/alpha/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/alpha/translate.c +++ b/target/alpha/translate.c @@ -XXX,XX +XXX,XX @@ struct DisasContext { #ifdef CONFIG_USER_ONLY #define UNALIGN(C) (C)->unalign #else -#define UNALIGN(C) 0 +#define UNALIGN(C) MO_ALIGN #endif /* Target-specific return values from translate_one, indicating the -- 2.34.1 Mark all memory operations that are not already marked with UNALIGN. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/alpha/translate.c | 36 ++++++++++++++++++++---------------- 1 file changed, 20 insertions(+), 16 deletions(-) diff --git a/target/alpha/translate.c b/target/alpha/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/alpha/translate.c +++ b/target/alpha/translate.c @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) switch ((insn >> 12) & 0xF) { case 0x0: /* Longword physical access (hw_ldl/p) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LESL); + tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LESL | MO_ALIGN); break; case 0x1: /* Quadword physical access (hw_ldq/p) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LEUQ); + tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LEUQ | MO_ALIGN); break; case 0x2: /* Longword physical access with lock (hw_ldl_l/p) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LESL); + tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LESL | MO_ALIGN); tcg_gen_mov_i64(cpu_lock_addr, addr); tcg_gen_mov_i64(cpu_lock_value, va); break; case 0x3: /* Quadword physical access with lock (hw_ldq_l/p) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LEUQ); + tcg_gen_qemu_ld_i64(va, addr, MMU_PHYS_IDX, MO_LEUQ | MO_ALIGN); tcg_gen_mov_i64(cpu_lock_addr, addr); tcg_gen_mov_i64(cpu_lock_value, va); break; @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) goto invalid_opc; case 0xA: /* Longword virtual access with protection check (hw_ldl/w) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_KERNEL_IDX, MO_LESL); + tcg_gen_qemu_ld_i64(va, addr, MMU_KERNEL_IDX, + MO_LESL | MO_ALIGN); break; case 0xB: /* Quadword virtual access with protection check (hw_ldq/w) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_KERNEL_IDX, MO_LEUQ); + tcg_gen_qemu_ld_i64(va, addr, MMU_KERNEL_IDX, + MO_LEUQ | MO_ALIGN); break; case 0xC: /* Longword virtual access with alt access mode (hw_ldl/a)*/ @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) case 0xE: /* Longword virtual access with alternate access mode and protection checks (hw_ldl/wa) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_USER_IDX, MO_LESL); + tcg_gen_qemu_ld_i64(va, addr, MMU_USER_IDX, + MO_LESL | MO_ALIGN); break; case 0xF: /* Quadword virtual access with alternate access mode and protection checks (hw_ldq/wa) */ - tcg_gen_qemu_ld_i64(va, addr, MMU_USER_IDX, MO_LEUQ); + tcg_gen_qemu_ld_i64(va, addr, MMU_USER_IDX, + MO_LEUQ | MO_ALIGN); break; } break; @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) vb = load_gpr(ctx, rb); tmp = tcg_temp_new(); tcg_gen_addi_i64(tmp, vb, disp12); - tcg_gen_qemu_st_i64(va, tmp, MMU_PHYS_IDX, MO_LESL); + tcg_gen_qemu_st_i64(va, tmp, MMU_PHYS_IDX, MO_LESL | MO_ALIGN); break; case 0x1: /* Quadword physical access */ @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) vb = load_gpr(ctx, rb); tmp = tcg_temp_new(); tcg_gen_addi_i64(tmp, vb, disp12); - tcg_gen_qemu_st_i64(va, tmp, MMU_PHYS_IDX, MO_LEUQ); + tcg_gen_qemu_st_i64(va, tmp, MMU_PHYS_IDX, MO_LEUQ | MO_ALIGN); break; case 0x2: /* Longword physical access with lock */ ret = gen_store_conditional(ctx, ra, rb, disp12, - MMU_PHYS_IDX, MO_LESL); + MMU_PHYS_IDX, MO_LESL | MO_ALIGN); break; case 0x3: /* Quadword physical access with lock */ ret = gen_store_conditional(ctx, ra, rb, disp12, - MMU_PHYS_IDX, MO_LEUQ); + MMU_PHYS_IDX, MO_LEUQ | MO_ALIGN); break; case 0x4: /* Longword virtual access */ @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) break; case 0x2A: /* LDL_L */ - gen_load_int(ctx, ra, rb, disp16, MO_LESL, 0, 1); + gen_load_int(ctx, ra, rb, disp16, MO_LESL | MO_ALIGN, 0, 1); break; case 0x2B: /* LDQ_L */ - gen_load_int(ctx, ra, rb, disp16, MO_LEUQ, 0, 1); + gen_load_int(ctx, ra, rb, disp16, MO_LEUQ | MO_ALIGN, 0, 1); break; case 0x2C: /* STL */ @@ -XXX,XX +XXX,XX @@ static DisasJumpType translate_one(DisasContext *ctx, uint32_t insn) case 0x2E: /* STL_C */ ret = gen_store_conditional(ctx, ra, rb, disp16, - ctx->mem_idx, MO_LESL); + ctx->mem_idx, MO_LESL | MO_ALIGN); break; case 0x2F: /* STQ_C */ ret = gen_store_conditional(ctx, ra, rb, disp16, - ctx->mem_idx, MO_LEUQ); + ctx->mem_idx, MO_LEUQ | MO_ALIGN); break; case 0x30: /* BR */ -- 2.34.1 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- configs/targets/alpha-linux-user.mak | 1 - configs/targets/alpha-softmmu.mak | 1 - 2 files changed, 2 deletions(-) diff --git a/configs/targets/alpha-linux-user.mak b/configs/targets/alpha-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/alpha-linux-user.mak +++ b/configs/targets/alpha-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=alpha TARGET_SYSTBL_ABI=common TARGET_SYSTBL=syscall.tbl -TARGET_ALIGNED_ONLY=y diff --git a/configs/targets/alpha-softmmu.mak b/configs/targets/alpha-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/alpha-softmmu.mak +++ b/configs/targets/alpha-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=alpha -TARGET_ALIGNED_ONLY=y TARGET_SUPPORTS_MTTCG=y -- 2.34.1 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/hppa/translate.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/target/hppa/translate.c b/target/hppa/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/hppa/translate.c +++ b/target/hppa/translate.c @@ -XXX,XX +XXX,XX @@ typedef struct DisasContext { #ifdef CONFIG_USER_ONLY #define UNALIGN(C) (C)->unalign #else -#define UNALIGN(C) 0 +#define UNALIGN(C) MO_ALIGN #endif /* Note that ssm/rsm instructions number PSW_W and PSW_E differently. */ -- 2.34.1 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- configs/targets/hppa-linux-user.mak | 1 - configs/targets/hppa-softmmu.mak | 1 - 2 files changed, 2 deletions(-) diff --git a/configs/targets/hppa-linux-user.mak b/configs/targets/hppa-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/hppa-linux-user.mak +++ b/configs/targets/hppa-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=hppa TARGET_SYSTBL_ABI=common,32 TARGET_SYSTBL=syscall.tbl -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y diff --git a/configs/targets/hppa-softmmu.mak b/configs/targets/hppa-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/hppa-softmmu.mak +++ b/configs/targets/hppa-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=hppa -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y TARGET_SUPPORTS_MTTCG=y -- 2.34.1 Acked-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/sparc/translate.c | 66 +++++++++++++++++++++------------------- 1 file changed, 34 insertions(+), 32 deletions(-) diff --git a/target/sparc/translate.c b/target/sparc/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/translate.c +++ b/target/sparc/translate.c @@ -XXX,XX +XXX,XX @@ static void gen_swap(DisasContext *dc, TCGv dst, TCGv src, TCGv addr, int mmu_idx, MemOp memop) { gen_address_mask(dc, addr); - tcg_gen_atomic_xchg_tl(dst, addr, src, mmu_idx, memop); + tcg_gen_atomic_xchg_tl(dst, addr, src, mmu_idx, memop | MO_ALIGN); } static void gen_ldstub(DisasContext *dc, TCGv dst, TCGv addr, int mmu_idx) @@ -XXX,XX +XXX,XX @@ static void gen_ld_asi(DisasContext *dc, TCGv dst, TCGv addr, break; case GET_ASI_DIRECT: gen_address_mask(dc, addr); - tcg_gen_qemu_ld_tl(dst, addr, da.mem_idx, da.memop); + tcg_gen_qemu_ld_tl(dst, addr, da.mem_idx, da.memop | MO_ALIGN); break; default: { TCGv_i32 r_asi = tcg_constant_i32(da.asi); - TCGv_i32 r_mop = tcg_constant_i32(memop); + TCGv_i32 r_mop = tcg_constant_i32(memop | MO_ALIGN); save_state(dc); #ifdef TARGET_SPARC64 @@ -XXX,XX +XXX,XX @@ static void gen_st_asi(DisasContext *dc, TCGv src, TCGv addr, /* fall through */ case GET_ASI_DIRECT: gen_address_mask(dc, addr); - tcg_gen_qemu_st_tl(src, addr, da.mem_idx, da.memop); + tcg_gen_qemu_st_tl(src, addr, da.mem_idx, da.memop | MO_ALIGN); break; #if !defined(TARGET_SPARC64) && !defined(CONFIG_USER_ONLY) case GET_ASI_BCOPY: @@ -XXX,XX +XXX,XX @@ static void gen_st_asi(DisasContext *dc, TCGv src, TCGv addr, default: { TCGv_i32 r_asi = tcg_constant_i32(da.asi); - TCGv_i32 r_mop = tcg_constant_i32(memop & MO_SIZE); + TCGv_i32 r_mop = tcg_constant_i32(memop | MO_ALIGN); save_state(dc); #ifdef TARGET_SPARC64 @@ -XXX,XX +XXX,XX @@ static void gen_cas_asi(DisasContext *dc, TCGv addr, TCGv cmpv, case GET_ASI_DIRECT: oldv = tcg_temp_new(); tcg_gen_atomic_cmpxchg_tl(oldv, addr, cmpv, gen_load_gpr(dc, rd), - da.mem_idx, da.memop); + da.mem_idx, da.memop | MO_ALIGN); gen_store_gpr(dc, rd, oldv); break; default: @@ -XXX,XX +XXX,XX @@ static void gen_ldf_asi(DisasContext *dc, TCGv addr, switch (size) { case 4: d32 = gen_dest_fpr_F(dc); - tcg_gen_qemu_ld_i32(d32, addr, da.mem_idx, da.memop); + tcg_gen_qemu_ld_i32(d32, addr, da.mem_idx, da.memop | MO_ALIGN); gen_store_fpr_F(dc, rd, d32); break; case 8: @@ -XXX,XX +XXX,XX @@ static void gen_ldf_asi(DisasContext *dc, TCGv addr, /* Valid for lddfa only. */ if (size == 8) { gen_address_mask(dc, addr); - tcg_gen_qemu_ld_i64(cpu_fpr[rd / 2], addr, da.mem_idx, da.memop); + tcg_gen_qemu_ld_i64(cpu_fpr[rd / 2], addr, da.mem_idx, + da.memop | MO_ALIGN); } else { gen_exception(dc, TT_ILL_INSN); } @@ -XXX,XX +XXX,XX @@ static void gen_ldf_asi(DisasContext *dc, TCGv addr, default: { TCGv_i32 r_asi = tcg_constant_i32(da.asi); - TCGv_i32 r_mop = tcg_constant_i32(da.memop); + TCGv_i32 r_mop = tcg_constant_i32(da.memop | MO_ALIGN); save_state(dc); /* According to the table in the UA2011 manual, the only @@ -XXX,XX +XXX,XX @@ static void gen_stf_asi(DisasContext *dc, TCGv addr, switch (size) { case 4: d32 = gen_load_fpr_F(dc, rd); - tcg_gen_qemu_st_i32(d32, addr, da.mem_idx, da.memop); + tcg_gen_qemu_st_i32(d32, addr, da.mem_idx, da.memop | MO_ALIGN); break; case 8: tcg_gen_qemu_st_i64(cpu_fpr[rd / 2], addr, da.mem_idx, @@ -XXX,XX +XXX,XX @@ static void gen_stf_asi(DisasContext *dc, TCGv addr, /* Valid for stdfa only. */ if (size == 8) { gen_address_mask(dc, addr); - tcg_gen_qemu_st_i64(cpu_fpr[rd / 2], addr, da.mem_idx, da.memop); + tcg_gen_qemu_st_i64(cpu_fpr[rd / 2], addr, da.mem_idx, + da.memop | MO_ALIGN); } else { gen_exception(dc, TT_ILL_INSN); } @@ -XXX,XX +XXX,XX @@ static void gen_ldda_asi(DisasContext *dc, TCGv addr, int insn, int rd) TCGv_i64 tmp = tcg_temp_new_i64(); gen_address_mask(dc, addr); - tcg_gen_qemu_ld_i64(tmp, addr, da.mem_idx, da.memop); + tcg_gen_qemu_ld_i64(tmp, addr, da.mem_idx, da.memop | MO_ALIGN); /* Note that LE ldda acts as if each 32-bit register result is byte swapped. Having just performed one @@ -XXX,XX +XXX,XX @@ static void gen_stda_asi(DisasContext *dc, TCGv hi, TCGv addr, tcg_gen_concat32_i64(t64, hi, lo); } gen_address_mask(dc, addr); - tcg_gen_qemu_st_i64(t64, addr, da.mem_idx, da.memop); + tcg_gen_qemu_st_i64(t64, addr, da.mem_idx, da.memop | MO_ALIGN); } break; @@ -XXX,XX +XXX,XX @@ static void gen_casx_asi(DisasContext *dc, TCGv addr, TCGv cmpv, case GET_ASI_DIRECT: oldv = tcg_temp_new(); tcg_gen_atomic_cmpxchg_tl(oldv, addr, cmpv, gen_load_gpr(dc, rd), - da.mem_idx, da.memop); + da.mem_idx, da.memop | MO_ALIGN); gen_store_gpr(dc, rd, oldv); break; default: @@ -XXX,XX +XXX,XX @@ static void gen_ldda_asi(DisasContext *dc, TCGv addr, int insn, int rd) return; case GET_ASI_DIRECT: gen_address_mask(dc, addr); - tcg_gen_qemu_ld_i64(t64, addr, da.mem_idx, da.memop); + tcg_gen_qemu_ld_i64(t64, addr, da.mem_idx, da.memop | MO_ALIGN); break; default: { @@ -XXX,XX +XXX,XX @@ static void gen_stda_asi(DisasContext *dc, TCGv hi, TCGv addr, break; case GET_ASI_DIRECT: gen_address_mask(dc, addr); - tcg_gen_qemu_st_i64(t64, addr, da.mem_idx, da.memop); + tcg_gen_qemu_st_i64(t64, addr, da.mem_idx, da.memop | MO_ALIGN); break; case GET_ASI_BFILL: /* Store 32 bytes of T64 to ADDR. */ @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x0: /* ld, V9 lduw, load unsigned word */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); break; case 0x1: /* ldub, load unsigned byte */ gen_address_mask(dc, cpu_addr); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x2: /* lduh, load unsigned halfword */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUW); + dc->mem_idx, MO_TEUW | MO_ALIGN); break; case 0x3: /* ldd, load double word */ if (rd & 1) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) gen_address_mask(dc, cpu_addr); t64 = tcg_temp_new_i64(); tcg_gen_qemu_ld_i64(t64, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); tcg_gen_trunc_i64_tl(cpu_val, t64); tcg_gen_ext32u_tl(cpu_val, cpu_val); gen_store_gpr(dc, rd + 1, cpu_val); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0xa: /* ldsh, load signed halfword */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TESW); + dc->mem_idx, MO_TESW | MO_ALIGN); break; case 0xd: /* ldstub */ gen_ldstub(dc, cpu_val, cpu_addr, dc->mem_idx); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x08: /* V9 ldsw */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TESL); + dc->mem_idx, MO_TESL | MO_ALIGN); break; case 0x0b: /* V9 ldx */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_ld_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); break; case 0x18: /* V9 ldswa */ gen_ld_asi(dc, cpu_val, cpu_addr, insn, MO_TESL); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) gen_address_mask(dc, cpu_addr); cpu_dst_32 = gen_dest_fpr_F(dc); tcg_gen_qemu_ld_i32(cpu_dst_32, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); gen_store_fpr_F(dc, rd, cpu_dst_32); break; case 0x21: /* ldfsr, V9 ldxfsr */ @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) if (rd == 1) { TCGv_i64 t64 = tcg_temp_new_i64(); tcg_gen_qemu_ld_i64(t64, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); gen_helper_ldxfsr(cpu_fsr, cpu_env, cpu_fsr, t64); break; } #endif cpu_dst_32 = tcg_temp_new_i32(); tcg_gen_qemu_ld_i32(cpu_dst_32, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); gen_helper_ldfsr(cpu_fsr, cpu_env, cpu_fsr, cpu_dst_32); break; case 0x22: /* ldqf, load quad fpreg */ @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x4: /* st, store word */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_st_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); break; case 0x5: /* stb, store byte */ gen_address_mask(dc, cpu_addr); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x6: /* sth, store halfword */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_st_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUW); + dc->mem_idx, MO_TEUW | MO_ALIGN); break; case 0x7: /* std, store double word */ if (rd & 1) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) t64 = tcg_temp_new_i64(); tcg_gen_concat_tl_i64(t64, lo, cpu_val); tcg_gen_qemu_st_i64(t64, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); } break; #if !defined(CONFIG_USER_ONLY) || defined(TARGET_SPARC64) @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) case 0x0e: /* V9 stx */ gen_address_mask(dc, cpu_addr); tcg_gen_qemu_st_tl(cpu_val, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); break; case 0x1e: /* V9 stxa */ gen_st_asi(dc, cpu_val, cpu_addr, insn, MO_TEUQ); @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) gen_address_mask(dc, cpu_addr); cpu_src1_32 = gen_load_fpr_F(dc, rd); tcg_gen_qemu_st_i32(cpu_src1_32, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); break; case 0x25: /* stfsr, V9 stxfsr */ { @@ -XXX,XX +XXX,XX @@ static void disas_sparc_insn(DisasContext * dc, unsigned int insn) gen_address_mask(dc, cpu_addr); if (rd == 1) { tcg_gen_qemu_st_tl(cpu_fsr, cpu_addr, - dc->mem_idx, MO_TEUQ); + dc->mem_idx, MO_TEUQ | MO_ALIGN); break; } #endif tcg_gen_qemu_st_tl(cpu_fsr, cpu_addr, - dc->mem_idx, MO_TEUL); + dc->mem_idx, MO_TEUL | MO_ALIGN); } break; case 0x26: -- 2.34.1 This passes on the memop as given as argument to helper_ld_asi to the ultimate load primitive. Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/sparc/ldst_helper.c | 10 ++++++---- 1 file changed, 6 insertions(+), 4 deletions(-) diff --git a/target/sparc/ldst_helper.c b/target/sparc/ldst_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/ldst_helper.c +++ b/target/sparc/ldst_helper.c @@ -XXX,XX +XXX,XX @@ uint64_t helper_ld_asi(CPUSPARCState *env, target_ulong addr, #if defined(DEBUG_MXCC) || defined(DEBUG_ASI) uint32_t last_addr = addr; #endif + MemOpIdx oi; do_check_align(env, addr, size - 1, GETPC()); switch (asi) { @@ -XXX,XX +XXX,XX @@ uint64_t helper_ld_asi(CPUSPARCState *env, target_ulong addr, case ASI_M_IODIAG: /* Turbosparc IOTLB Diagnostic */ break; case ASI_KERNELTXT: /* Supervisor code access */ + oi = make_memop_idx(memop, cpu_mmu_index(env, true)); switch (size) { case 1: - ret = cpu_ldub_code(env, addr); + ret = cpu_ldb_code_mmu(env, addr, oi, GETPC()); break; case 2: - ret = cpu_lduw_code(env, addr); + ret = cpu_ldw_code_mmu(env, addr, oi, GETPC()); break; default: case 4: - ret = cpu_ldl_code(env, addr); + ret = cpu_ldl_code_mmu(env, addr, oi, GETPC()); break; case 8: - ret = cpu_ldq_code(env, addr); + ret = cpu_ldq_code_mmu(env, addr, oi, GETPC()); break; } break; -- 2.34.1 Reviewed-by: Mark Cave-Ayland <mark.cave-ayland@ilande.co.uk> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- configs/targets/sparc-linux-user.mak | 1 - configs/targets/sparc-softmmu.mak | 1 - configs/targets/sparc32plus-linux-user.mak | 1 - configs/targets/sparc64-linux-user.mak | 1 - configs/targets/sparc64-softmmu.mak | 1 - 5 files changed, 5 deletions(-) diff --git a/configs/targets/sparc-linux-user.mak b/configs/targets/sparc-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/sparc-linux-user.mak +++ b/configs/targets/sparc-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=sparc TARGET_SYSTBL_ABI=common,32 TARGET_SYSTBL=syscall.tbl -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y diff --git a/configs/targets/sparc-softmmu.mak b/configs/targets/sparc-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/sparc-softmmu.mak +++ b/configs/targets/sparc-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=sparc -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y diff --git a/configs/targets/sparc32plus-linux-user.mak b/configs/targets/sparc32plus-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/sparc32plus-linux-user.mak +++ b/configs/targets/sparc32plus-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_BASE_ARCH=sparc TARGET_ABI_DIR=sparc TARGET_SYSTBL_ABI=common,32 TARGET_SYSTBL=syscall.tbl -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y diff --git a/configs/targets/sparc64-linux-user.mak b/configs/targets/sparc64-linux-user.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/sparc64-linux-user.mak +++ b/configs/targets/sparc64-linux-user.mak @@ -XXX,XX +XXX,XX @@ TARGET_BASE_ARCH=sparc TARGET_ABI_DIR=sparc TARGET_SYSTBL_ABI=common,64 TARGET_SYSTBL=syscall.tbl -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y diff --git a/configs/targets/sparc64-softmmu.mak b/configs/targets/sparc64-softmmu.mak index XXXXXXX..XXXXXXX 100644 --- a/configs/targets/sparc64-softmmu.mak +++ b/configs/targets/sparc64-softmmu.mak @@ -XXX,XX +XXX,XX @@ TARGET_ARCH=sparc64 TARGET_BASE_ARCH=sparc -TARGET_ALIGNED_ONLY=y TARGET_BIG_ENDIAN=y -- 2.34.1 Interpret the variable argument placement in the caller. Pass data_type instead of is64 -- there are several places where we already convert back from bool to type. Clean things up by using type throughout. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 111 +++++++++++++++++--------------------- 1 file changed, 50 insertions(+), 61 deletions(-) diff --git a/tcg/i386/tcg-target.c.inc b/tcg/i386/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/i386/tcg-target.c.inc +++ b/tcg/i386/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, * Record the context of a call to the out of line helper code for the slow path * for a load or store, so that we can later generate the correct helper code */ -static void add_qemu_ldst_label(TCGContext *s, bool is_ld, bool is_64, - MemOpIdx oi, +static void add_qemu_ldst_label(TCGContext *s, bool is_ld, + TCGType type, MemOpIdx oi, TCGReg datalo, TCGReg datahi, TCGReg addrlo, TCGReg addrhi, tcg_insn_unit *raddr, @@ -XXX,XX +XXX,XX @@ static void add_qemu_ldst_label(TCGContext *s, bool is_ld, bool is_64, label->is_ld = is_ld; label->oi = oi; - label->type = is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32; + label->type = type; label->datalo_reg = datalo; label->datahi_reg = datahi; label->addrlo_reg = addrlo; @@ -XXX,XX +XXX,XX @@ static inline int setup_guest_base_seg(void) static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, TCGReg base, int index, intptr_t ofs, - int seg, bool is64, MemOp memop) + int seg, TCGType type, MemOp memop) { - TCGType type = is64 ? TCG_TYPE_I64 : TCG_TYPE_I32; bool use_movbe = false; - int rexw = is64 * P_REXW; + int rexw = (type == TCG_TYPE_I32 ? 0 : P_REXW); int movop = OPC_MOVL_GvEv; /* Do big-endian loads with movbe. */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, } } -/* XXX: qemu_ld and qemu_st could be modified to clobber only EDX and - EAX. It will be useful once fixed registers globals are less - common. */ -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is64) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg datalo, datahi, addrlo; - TCGReg addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc; + MemOp opc = get_memop(oi); + #if defined(CONFIG_SOFTMMU) - int mem_index; tcg_insn_unit *label_ptr[2]; -#else - unsigned a_bits; -#endif - datalo = *args++; - datahi = (TCG_TARGET_REG_BITS == 32 && is64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TARGET_LONG_BITS > TCG_TARGET_REG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - -#if defined(CONFIG_SOFTMMU) - mem_index = get_mmuidx(oi); - - tcg_out_tlb_load(s, addrlo, addrhi, mem_index, opc, + tcg_out_tlb_load(s, addrlo, addrhi, get_mmuidx(oi), opc, label_ptr, offsetof(CPUTLBEntry, addr_read)); /* TLB Hit. */ - tcg_out_qemu_ld_direct(s, datalo, datahi, TCG_REG_L1, -1, 0, 0, is64, opc); + tcg_out_qemu_ld_direct(s, datalo, datahi, TCG_REG_L1, + -1, 0, 0, data_type, opc); /* Record the current context of a load into ldst label */ - add_qemu_ldst_label(s, true, is64, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, true, data_type, oi, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } tcg_out_qemu_ld_direct(s, datalo, datahi, addrlo, x86_guest_base_index, x86_guest_base_offset, x86_guest_base_seg, - is64, opc); + data_type, opc); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, } } -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is64) +static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg datalo, datahi, addrlo; - TCGReg addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc; + MemOp opc = get_memop(oi); + #if defined(CONFIG_SOFTMMU) - int mem_index; tcg_insn_unit *label_ptr[2]; -#else - unsigned a_bits; -#endif - datalo = *args++; - datahi = (TCG_TARGET_REG_BITS == 32 && is64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TARGET_LONG_BITS > TCG_TARGET_REG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - -#if defined(CONFIG_SOFTMMU) - mem_index = get_mmuidx(oi); - - tcg_out_tlb_load(s, addrlo, addrhi, mem_index, opc, + tcg_out_tlb_load(s, addrlo, addrhi, get_mmuidx(oi), opc, label_ptr, offsetof(CPUTLBEntry, addr_write)); /* TLB Hit. */ tcg_out_qemu_st_direct(s, datalo, datahi, TCG_REG_L1, -1, 0, 0, opc); /* Record the current context of a store into ldst label */ - add_qemu_ldst_label(s, false, is64, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, data_type, oi, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); } @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, 0); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); + } else { + tcg_out_qemu_ld(s, a0, -1, a1, a2, args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, 1); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_ld(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_ld(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); + } break; case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st8_i32: - tcg_out_qemu_st(s, args, 0); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); + } else { + tcg_out_qemu_st(s, a0, -1, a1, a2, args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, 1); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_st(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_st(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); + } break; OP_32_64(mulu2): -- 2.34.1 Test for both base and index; use datahi as a temporary, overwritten by the final load. Always perform the loads in ascending order, so that any (user-only) fault sees the correct address. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 31 +++++++++++++++---------------- 1 file changed, 15 insertions(+), 16 deletions(-) diff --git a/tcg/i386/tcg-target.c.inc b/tcg/i386/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/i386/tcg-target.c.inc +++ b/tcg/i386/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, if (TCG_TARGET_REG_BITS == 64) { tcg_out_modrm_sib_offset(s, movop + P_REXW + seg, datalo, base, index, 0, ofs); + break; + } + if (use_movbe) { + TCGReg t = datalo; + datalo = datahi; + datahi = t; + } + if (base == datalo || index == datalo) { + tcg_out_modrm_sib_offset(s, OPC_LEA, datahi, base, index, 0, ofs); + tcg_out_modrm_offset(s, movop + seg, datalo, datahi, 0); + tcg_out_modrm_offset(s, movop + seg, datahi, datahi, 4); } else { - if (use_movbe) { - TCGReg t = datalo; - datalo = datahi; - datahi = t; - } - if (base != datalo) { - tcg_out_modrm_sib_offset(s, movop + seg, datalo, - base, index, 0, ofs); - tcg_out_modrm_sib_offset(s, movop + seg, datahi, - base, index, 0, ofs + 4); - } else { - tcg_out_modrm_sib_offset(s, movop + seg, datahi, - base, index, 0, ofs + 4); - tcg_out_modrm_sib_offset(s, movop + seg, datalo, - base, index, 0, ofs); - } + tcg_out_modrm_sib_offset(s, movop + seg, datalo, + base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + seg, datahi, + base, index, 0, ofs + 4); } break; default: -- 2.34.1 Collect the 4 potential parts of the host address into a struct. Reorg tcg_out_qemu_{ld,st}_direct to use it. Reorg guest_base handling to use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 165 +++++++++++++++++++++----------------- 1 file changed, 90 insertions(+), 75 deletions(-) diff --git a/tcg/i386/tcg-target.c.inc b/tcg/i386/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/i386/tcg-target.c.inc +++ b/tcg/i386/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_nopn(TCGContext *s, int n) tcg_out8(s, 0x90); } +typedef struct { + TCGReg base; + int index; + int ofs; + int seg; +} HostAddress; + #if defined(CONFIG_SOFTMMU) /* helper signature: helper_ret_ld_mmu(CPUState *env, target_ulong addr, * int mmu_idx, uintptr_t ra) @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) return tcg_out_fail_alignment(s, l); } -#if TCG_TARGET_REG_BITS == 32 -# define x86_guest_base_seg 0 -# define x86_guest_base_index -1 -# define x86_guest_base_offset guest_base -#else -static int x86_guest_base_seg; -static int x86_guest_base_index = -1; -static int32_t x86_guest_base_offset; -# if defined(__x86_64__) && defined(__linux__) -# include <asm/prctl.h> -# include <sys/prctl.h> +static HostAddress x86_guest_base = { + .index = -1 +}; + +#if defined(__x86_64__) && defined(__linux__) +# include <asm/prctl.h> +# include <sys/prctl.h> int arch_prctl(int code, unsigned long addr); static inline int setup_guest_base_seg(void) { @@ -XXX,XX +XXX,XX @@ static inline int setup_guest_base_seg(void) } return 0; } -# elif defined (__FreeBSD__) || defined (__FreeBSD_kernel__) -# include <machine/sysarch.h> +#elif defined(__x86_64__) && \ + (defined (__FreeBSD__) || defined (__FreeBSD_kernel__)) +# include <machine/sysarch.h> static inline int setup_guest_base_seg(void) { if (sysarch(AMD64_SET_GSBASE, &guest_base) == 0) { @@ -XXX,XX +XXX,XX @@ static inline int setup_guest_base_seg(void) } return 0; } -# else +#else static inline int setup_guest_base_seg(void) { return 0; } -# endif -#endif +#endif /* setup_guest_base_seg */ #endif /* SOFTMMU */ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg base, int index, intptr_t ofs, - int seg, TCGType type, MemOp memop) + HostAddress h, TCGType type, MemOp memop) { bool use_movbe = false; int rexw = (type == TCG_TYPE_I32 ? 0 : P_REXW); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, switch (memop & MO_SSIZE) { case MO_UB: - tcg_out_modrm_sib_offset(s, OPC_MOVZBL + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVZBL + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; case MO_SB: - tcg_out_modrm_sib_offset(s, OPC_MOVSBL + rexw + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVSBL + rexw + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; case MO_UW: if (use_movbe) { /* There is no extending movbe; only low 16-bits are modified. */ - if (datalo != base && datalo != index) { + if (datalo != h.base && datalo != h.index) { /* XOR breaks dependency chains. */ tgen_arithr(s, ARITH_XOR, datalo, datalo); - tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + seg, - datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + h.seg, + datalo, h.base, h.index, 0, h.ofs); } else { - tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + seg, - datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + h.seg, + datalo, h.base, h.index, 0, h.ofs); tcg_out_ext16u(s, datalo, datalo); } } else { - tcg_out_modrm_sib_offset(s, OPC_MOVZWL + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVZWL + h.seg, datalo, + h.base, h.index, 0, h.ofs); } break; case MO_SW: if (use_movbe) { - tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + seg, - datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + P_DATA16 + h.seg, + datalo, h.base, h.index, 0, h.ofs); tcg_out_ext16s(s, type, datalo, datalo); } else { - tcg_out_modrm_sib_offset(s, OPC_MOVSWL + rexw + seg, - datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVSWL + rexw + h.seg, + datalo, h.base, h.index, 0, h.ofs); } break; case MO_UL: - tcg_out_modrm_sib_offset(s, movop + seg, datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; #if TCG_TARGET_REG_BITS == 64 case MO_SL: if (use_movbe) { - tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVBE_GyMy + h.seg, datalo, + h.base, h.index, 0, h.ofs); tcg_out_ext32s(s, datalo, datalo); } else { - tcg_out_modrm_sib_offset(s, OPC_MOVSLQ + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVSLQ + h.seg, datalo, + h.base, h.index, 0, h.ofs); } break; #endif case MO_UQ: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_modrm_sib_offset(s, movop + P_REXW + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + P_REXW + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; } if (use_movbe) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, datalo = datahi; datahi = t; } - if (base == datalo || index == datalo) { - tcg_out_modrm_sib_offset(s, OPC_LEA, datahi, base, index, 0, ofs); - tcg_out_modrm_offset(s, movop + seg, datalo, datahi, 0); - tcg_out_modrm_offset(s, movop + seg, datahi, datahi, 4); + if (h.base == datalo || h.index == datalo) { + tcg_out_modrm_sib_offset(s, OPC_LEA, datahi, + h.base, h.index, 0, h.ofs); + tcg_out_modrm_offset(s, movop + h.seg, datalo, datahi, 0); + tcg_out_modrm_offset(s, movop + h.seg, datahi, datahi, 4); } else { - tcg_out_modrm_sib_offset(s, movop + seg, datalo, - base, index, 0, ofs); - tcg_out_modrm_sib_offset(s, movop + seg, datahi, - base, index, 0, ofs + 4); + tcg_out_modrm_sib_offset(s, movop + h.seg, datalo, + h.base, h.index, 0, h.ofs); + tcg_out_modrm_sib_offset(s, movop + h.seg, datahi, + h.base, h.index, 0, h.ofs + 4); } break; default: @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; #if defined(CONFIG_SOFTMMU) tcg_insn_unit *label_ptr[2]; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, label_ptr, offsetof(CPUTLBEntry, addr_read)); /* TLB Hit. */ - tcg_out_qemu_ld_direct(s, datalo, datahi, TCG_REG_L1, - -1, 0, 0, data_type, opc); + h.base = TCG_REG_L1; + h.index = -1; + h.ofs = 0; + h.seg = 0; + tcg_out_qemu_ld_direct(s, datalo, datahi, h, data_type, opc); /* Record the current context of a load into ldst label */ add_qemu_ldst_label(s, true, data_type, oi, datalo, datahi, @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } - tcg_out_qemu_ld_direct(s, datalo, datahi, addrlo, x86_guest_base_index, - x86_guest_base_offset, x86_guest_base_seg, - data_type, opc); + h = x86_guest_base; + h.base = addrlo; + tcg_out_qemu_ld_direct(s, datalo, datahi, h, data_type, opc); #endif } static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg base, int index, intptr_t ofs, - int seg, MemOp memop) + HostAddress h, MemOp memop) { bool use_movbe = false; int movop = OPC_MOVL_EvGv; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, case MO_8: /* This is handled with constraints on INDEX_op_qemu_st8_i32. */ tcg_debug_assert(TCG_TARGET_REG_BITS == 64 || datalo < 4); - tcg_out_modrm_sib_offset(s, OPC_MOVB_EvGv + P_REXB_R + seg, - datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, OPC_MOVB_EvGv + P_REXB_R + h.seg, + datalo, h.base, h.index, 0, h.ofs); break; case MO_16: - tcg_out_modrm_sib_offset(s, movop + P_DATA16 + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + P_DATA16 + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; case MO_32: - tcg_out_modrm_sib_offset(s, movop + seg, datalo, base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + h.seg, datalo, + h.base, h.index, 0, h.ofs); break; case MO_64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_modrm_sib_offset(s, movop + P_REXW + seg, datalo, - base, index, 0, ofs); + tcg_out_modrm_sib_offset(s, movop + P_REXW + h.seg, datalo, + h.base, h.index, 0, h.ofs); } else { if (use_movbe) { TCGReg t = datalo; datalo = datahi; datahi = t; } - tcg_out_modrm_sib_offset(s, movop + seg, datalo, - base, index, 0, ofs); - tcg_out_modrm_sib_offset(s, movop + seg, datahi, - base, index, 0, ofs + 4); + tcg_out_modrm_sib_offset(s, movop + h.seg, datalo, + h.base, h.index, 0, h.ofs); + tcg_out_modrm_sib_offset(s, movop + h.seg, datahi, + h.base, h.index, 0, h.ofs + 4); } break; default: @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; #if defined(CONFIG_SOFTMMU) tcg_insn_unit *label_ptr[2]; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, label_ptr, offsetof(CPUTLBEntry, addr_write)); /* TLB Hit. */ - tcg_out_qemu_st_direct(s, datalo, datahi, TCG_REG_L1, -1, 0, 0, opc); + h.base = TCG_REG_L1; + h.index = -1; + h.ofs = 0; + h.seg = 0; + tcg_out_qemu_st_direct(s, datalo, datahi, h, opc); /* Record the current context of a store into ldst label */ add_qemu_ldst_label(s, false, data_type, oi, datalo, datahi, @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); } - tcg_out_qemu_st_direct(s, datalo, datahi, addrlo, x86_guest_base_index, - x86_guest_base_offset, x86_guest_base_seg, opc); + h = x86_guest_base; + h.base = addrlo; + + tcg_out_qemu_st_direct(s, datalo, datahi, h, opc); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_target_qemu_prologue(TCGContext *s) (ARRAY_SIZE(tcg_target_callee_save_regs) + 2) * 4 + stack_addend); #else -# if !defined(CONFIG_SOFTMMU) && TCG_TARGET_REG_BITS == 64 +# if !defined(CONFIG_SOFTMMU) if (guest_base) { int seg = setup_guest_base_seg(); if (seg != 0) { - x86_guest_base_seg = seg; + x86_guest_base.seg = seg; } else if (guest_base == (int32_t)guest_base) { - x86_guest_base_offset = guest_base; + x86_guest_base.ofs = guest_base; } else { /* Choose R12 because, as a base, it requires a SIB byte. */ - x86_guest_base_index = TCG_REG_R12; - tcg_out_movi(s, TCG_TYPE_PTR, x86_guest_base_index, guest_base); - tcg_regset_set_reg(s->reserved_regs, x86_guest_base_index); + x86_guest_base.index = TCG_REG_R12; + tcg_out_movi(s, TCG_TYPE_PTR, x86_guest_base.index, guest_base); + tcg_regset_set_reg(s->reserved_regs, x86_guest_base.index); } } # endif -- 2.34.1 Use TCG_REG_L[01] constants directly. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 32 ++++++++++++++++---------------- 1 file changed, 16 insertions(+), 16 deletions(-) diff --git a/tcg/i386/tcg-target.c.inc b/tcg/i386/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/i386/tcg-target.c.inc +++ b/tcg/i386/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, int mem_index, MemOp opc, tcg_insn_unit **label_ptr, int which) { - const TCGReg r0 = TCG_REG_L0; - const TCGReg r1 = TCG_REG_L1; TCGType ttype = TCG_TYPE_I32; TCGType tlbtype = TCG_TYPE_I32; int trexw = 0, hrexw = 0, tlbrexw = 0; @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, } } - tcg_out_mov(s, tlbtype, r0, addrlo); - tcg_out_shifti(s, SHIFT_SHR + tlbrexw, r0, + tcg_out_mov(s, tlbtype, TCG_REG_L0, addrlo); + tcg_out_shifti(s, SHIFT_SHR + tlbrexw, TCG_REG_L0, TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS); - tcg_out_modrm_offset(s, OPC_AND_GvEv + trexw, r0, TCG_AREG0, + tcg_out_modrm_offset(s, OPC_AND_GvEv + trexw, TCG_REG_L0, TCG_AREG0, TLB_MASK_TABLE_OFS(mem_index) + offsetof(CPUTLBDescFast, mask)); - tcg_out_modrm_offset(s, OPC_ADD_GvEv + hrexw, r0, TCG_AREG0, + tcg_out_modrm_offset(s, OPC_ADD_GvEv + hrexw, TCG_REG_L0, TCG_AREG0, TLB_MASK_TABLE_OFS(mem_index) + offsetof(CPUTLBDescFast, table)); @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, copy the address and mask. For lesser alignments, check that we don't cross pages for the complete access. */ if (a_bits >= s_bits) { - tcg_out_mov(s, ttype, r1, addrlo); + tcg_out_mov(s, ttype, TCG_REG_L1, addrlo); } else { - tcg_out_modrm_offset(s, OPC_LEA + trexw, r1, addrlo, s_mask - a_mask); + tcg_out_modrm_offset(s, OPC_LEA + trexw, TCG_REG_L1, + addrlo, s_mask - a_mask); } tlb_mask = (target_ulong)TARGET_PAGE_MASK | a_mask; - tgen_arithi(s, ARITH_AND + trexw, r1, tlb_mask, 0); + tgen_arithi(s, ARITH_AND + trexw, TCG_REG_L1, tlb_mask, 0); - /* cmp 0(r0), r1 */ - tcg_out_modrm_offset(s, OPC_CMP_GvEv + trexw, r1, r0, which); + /* cmp 0(TCG_REG_L0), TCG_REG_L1 */ + tcg_out_modrm_offset(s, OPC_CMP_GvEv + trexw, + TCG_REG_L1, TCG_REG_L0, which); /* Prepare for both the fast path add of the tlb addend, and the slow path function argument setup. */ - tcg_out_mov(s, ttype, r1, addrlo); + tcg_out_mov(s, ttype, TCG_REG_L1, addrlo); /* jne slow_path */ tcg_out_opc(s, OPC_JCC_long + JCC_JNE, 0, 0, 0); @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, s->code_ptr += 4; if (TARGET_LONG_BITS > TCG_TARGET_REG_BITS) { - /* cmp 4(r0), addrhi */ - tcg_out_modrm_offset(s, OPC_CMP_GvEv, addrhi, r0, which + 4); + /* cmp 4(TCG_REG_L0), addrhi */ + tcg_out_modrm_offset(s, OPC_CMP_GvEv, addrhi, TCG_REG_L0, which + 4); /* jne slow_path */ tcg_out_opc(s, OPC_JCC_long + JCC_JNE, 0, 0, 0); @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_tlb_load(TCGContext *s, TCGReg addrlo, TCGReg addrhi, /* TLB Hit. */ - /* add addend(r0), r1 */ - tcg_out_modrm_offset(s, OPC_ADD_GvEv + hrexw, r1, r0, + /* add addend(TCG_REG_L0), TCG_REG_L1 */ + tcg_out_modrm_offset(s, OPC_ADD_GvEv + hrexw, TCG_REG_L1, TCG_REG_L0, offsetof(CPUTLBEntry, addend)); } -- 2.34.1 Split out a helper for choosing testb vs testl. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 30 ++++++++++++++++++------------ 1 file changed, 18 insertions(+), 12 deletions(-) diff --git a/tcg/i386/tcg-target.c.inc b/tcg/i386/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/i386/tcg-target.c.inc +++ b/tcg/i386/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_nopn(TCGContext *s, int n) tcg_out8(s, 0x90); } +/* Test register R vs immediate bits I, setting Z flag for EQ/NE. */ +static void __attribute__((unused)) +tcg_out_testi(TCGContext *s, TCGReg r, uint32_t i) +{ + /* + * This is used for testing alignment, so we can usually use testb. + * For i686, we have to use testl for %esi/%edi. + */ + if (i <= 0xff && (TCG_TARGET_REG_BITS == 64 || r < 4)) { + tcg_out_modrm(s, OPC_GRP3_Eb | P_REXB_RM, EXT3_TESTi, r); + tcg_out8(s, i); + } else { + tcg_out_modrm(s, OPC_GRP3_Ev, EXT3_TESTi, r); + tcg_out32(s, i); + } +} + typedef struct { TCGReg base; int index; @@ -XXX,XX +XXX,XX @@ static void tcg_out_test_alignment(TCGContext *s, bool is_ld, TCGReg addrlo, unsigned a_mask = (1 << a_bits) - 1; TCGLabelQemuLdst *label; - /* - * We are expecting a_bits to max out at 7, so we can usually use testb. - * For i686, we have to use testl for %esi/%edi. - */ - if (a_mask <= 0xff && (TCG_TARGET_REG_BITS == 64 || addrlo < 4)) { - tcg_out_modrm(s, OPC_GRP3_Eb | P_REXB_RM, EXT3_TESTi, addrlo); - tcg_out8(s, a_mask); - } else { - tcg_out_modrm(s, OPC_GRP3_Ev, EXT3_TESTi, addrlo); - tcg_out32(s, a_mask); - } - + tcg_out_testi(s, addrlo, a_mask); /* jne slow_path */ tcg_out_opc(s, OPC_JCC_long + JCC_JNE, 0, 0, 0); -- 2.34.1 Rename the 'ext' parameter 'data_type' to make the use clearer; pass it to tcg_out_qemu_st as well to even out the interfaces. Rename the 'otype' local 'addr_type' to make the use clearer. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/aarch64/tcg-target.c.inc | 36 +++++++++++++++++------------------- 1 file changed, 17 insertions(+), 19 deletions(-) diff --git a/tcg/aarch64/tcg-target.c.inc b/tcg/aarch64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/aarch64/tcg-target.c.inc +++ b/tcg/aarch64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, MemOp memop, } static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, - MemOpIdx oi, TCGType ext) + MemOpIdx oi, TCGType data_type) { MemOp memop = get_memop(oi); - const TCGType otype = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; + TCGType addr_type = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((memop & MO_BSWAP) == 0); #ifdef CONFIG_SOFTMMU - unsigned mem_index = get_mmuidx(oi); tcg_insn_unit *label_ptr; - tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, mem_index, 1); - tcg_out_qemu_ld_direct(s, memop, ext, data_reg, - TCG_REG_X1, otype, addr_reg); - add_qemu_ldst_label(s, true, oi, ext, data_reg, addr_reg, + tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, get_mmuidx(oi), 1); + tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, + TCG_REG_X1, addr_type, addr_reg); + add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ unsigned a_bits = get_alignment_bits(memop); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, tcg_out_test_alignment(s, true, addr_reg, a_bits); } if (USE_GUEST_BASE) { - tcg_out_qemu_ld_direct(s, memop, ext, data_reg, - TCG_REG_GUEST_BASE, otype, addr_reg); + tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, + TCG_REG_GUEST_BASE, addr_type, addr_reg); } else { - tcg_out_qemu_ld_direct(s, memop, ext, data_reg, + tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, addr_reg, TCG_TYPE_I64, TCG_REG_XZR); } #endif /* CONFIG_SOFTMMU */ } static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, - MemOpIdx oi) + MemOpIdx oi, TCGType data_type) { MemOp memop = get_memop(oi); - const TCGType otype = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; + TCGType addr_type = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((memop & MO_BSWAP) == 0); #ifdef CONFIG_SOFTMMU - unsigned mem_index = get_mmuidx(oi); tcg_insn_unit *label_ptr; - tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, mem_index, 0); + tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, get_mmuidx(oi), 0); tcg_out_qemu_st_direct(s, memop, data_reg, - TCG_REG_X1, otype, addr_reg); - add_qemu_ldst_label(s, false, oi, (memop & MO_SIZE)== MO_64, - data_reg, addr_reg, s->code_ptr, label_ptr); + TCG_REG_X1, addr_type, addr_reg); + add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ unsigned a_bits = get_alignment_bits(memop); if (a_bits) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, } if (USE_GUEST_BASE) { tcg_out_qemu_st_direct(s, memop, data_reg, - TCG_REG_GUEST_BASE, otype, addr_reg); + TCG_REG_GUEST_BASE, addr_type, addr_reg); } else { tcg_out_qemu_st_direct(s, memop, data_reg, addr_reg, TCG_TYPE_I64, TCG_REG_XZR); @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, REG0(0), a1, a2); + tcg_out_qemu_st(s, REG0(0), a1, a2, ext); break; case INDEX_op_bswap64_i64: -- 2.34.1 Collect the 3 potential parts of the host address into a struct. Reorg tcg_out_qemu_{ld,st}_direct to use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/aarch64/tcg-target.c.inc | 86 +++++++++++++++++++++++++----------- 1 file changed, 59 insertions(+), 27 deletions(-) diff --git a/tcg/aarch64/tcg-target.c.inc b/tcg/aarch64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/aarch64/tcg-target.c.inc +++ b/tcg/aarch64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_adr(TCGContext *s, TCGReg rd, const void *target) tcg_out_insn(s, 3406, ADR, rd, offset); } +typedef struct { + TCGReg base; + TCGReg index; + TCGType index_ext; +} HostAddress; + #ifdef CONFIG_SOFTMMU /* helper signature: helper_ret_ld_mmu(CPUState *env, target_ulong addr, * MemOpIdx oi, uintptr_t ra) @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) #endif /* CONFIG_SOFTMMU */ static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp memop, TCGType ext, - TCGReg data_r, TCGReg addr_r, - TCGType otype, TCGReg off_r) + TCGReg data_r, HostAddress h) { switch (memop & MO_SSIZE) { case MO_UB: - tcg_out_ldst_r(s, I3312_LDRB, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_LDRB, data_r, h.base, h.index_ext, h.index); break; case MO_SB: tcg_out_ldst_r(s, ext ? I3312_LDRSBX : I3312_LDRSBW, - data_r, addr_r, otype, off_r); + data_r, h.base, h.index_ext, h.index); break; case MO_UW: - tcg_out_ldst_r(s, I3312_LDRH, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_LDRH, data_r, h.base, h.index_ext, h.index); break; case MO_SW: tcg_out_ldst_r(s, (ext ? I3312_LDRSHX : I3312_LDRSHW), - data_r, addr_r, otype, off_r); + data_r, h.base, h.index_ext, h.index); break; case MO_UL: - tcg_out_ldst_r(s, I3312_LDRW, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_LDRW, data_r, h.base, h.index_ext, h.index); break; case MO_SL: - tcg_out_ldst_r(s, I3312_LDRSWX, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_LDRSWX, data_r, h.base, h.index_ext, h.index); break; case MO_UQ: - tcg_out_ldst_r(s, I3312_LDRX, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_LDRX, data_r, h.base, h.index_ext, h.index); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp memop, TCGType ext, } static void tcg_out_qemu_st_direct(TCGContext *s, MemOp memop, - TCGReg data_r, TCGReg addr_r, - TCGType otype, TCGReg off_r) + TCGReg data_r, HostAddress h) { switch (memop & MO_SIZE) { case MO_8: - tcg_out_ldst_r(s, I3312_STRB, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_STRB, data_r, h.base, h.index_ext, h.index); break; case MO_16: - tcg_out_ldst_r(s, I3312_STRH, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_STRH, data_r, h.base, h.index_ext, h.index); break; case MO_32: - tcg_out_ldst_r(s, I3312_STRW, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_STRW, data_r, h.base, h.index_ext, h.index); break; case MO_64: - tcg_out_ldst_r(s, I3312_STRX, data_r, addr_r, otype, off_r); + tcg_out_ldst_r(s, I3312_STRX, data_r, h.base, h.index_ext, h.index); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, { MemOp memop = get_memop(oi); TCGType addr_type = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; + HostAddress h; /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((memop & MO_BSWAP) == 0); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, tcg_insn_unit *label_ptr; tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, get_mmuidx(oi), 1); - tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, - TCG_REG_X1, addr_type, addr_reg); + + h = (HostAddress){ + .base = TCG_REG_X1, + .index = addr_reg, + .index_ext = addr_type + }; + tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, h); + add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, tcg_out_test_alignment(s, true, addr_reg, a_bits); } if (USE_GUEST_BASE) { - tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, - TCG_REG_GUEST_BASE, addr_type, addr_reg); + h = (HostAddress){ + .base = TCG_REG_GUEST_BASE, + .index = addr_reg, + .index_ext = addr_type + }; } else { - tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, - addr_reg, TCG_TYPE_I64, TCG_REG_XZR); + h = (HostAddress){ + .base = addr_reg, + .index = TCG_REG_XZR, + .index_ext = TCG_TYPE_I64 + }; } + tcg_out_qemu_ld_direct(s, memop, data_type, data_reg, h); #endif /* CONFIG_SOFTMMU */ } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, { MemOp memop = get_memop(oi); TCGType addr_type = TARGET_LONG_BITS == 64 ? TCG_TYPE_I64 : TCG_TYPE_I32; + HostAddress h; /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((memop & MO_BSWAP) == 0); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, tcg_insn_unit *label_ptr; tcg_out_tlb_read(s, addr_reg, memop, &label_ptr, get_mmuidx(oi), 0); - tcg_out_qemu_st_direct(s, memop, data_reg, - TCG_REG_X1, addr_type, addr_reg); + + h = (HostAddress){ + .base = TCG_REG_X1, + .index = addr_reg, + .index_ext = addr_type + }; + tcg_out_qemu_st_direct(s, memop, data_reg, h); + add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, tcg_out_test_alignment(s, false, addr_reg, a_bits); } if (USE_GUEST_BASE) { - tcg_out_qemu_st_direct(s, memop, data_reg, - TCG_REG_GUEST_BASE, addr_type, addr_reg); + h = (HostAddress){ + .base = TCG_REG_GUEST_BASE, + .index = addr_reg, + .index_ext = addr_type + }; } else { - tcg_out_qemu_st_direct(s, memop, data_reg, - addr_reg, TCG_TYPE_I64, TCG_REG_XZR); + h = (HostAddress){ + .base = addr_reg, + .index = TCG_REG_XZR, + .index_ext = TCG_TYPE_I64 + }; } + tcg_out_qemu_st_direct(s, memop, data_reg, h); #endif /* CONFIG_SOFTMMU */ } -- 2.34.1 Interpret the variable argument placement in the caller. Pass data_type instead of is_64. We need to set this in TCGLabelQemuLdst, so plumb this all the way through from tcg_out_op. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/arm/tcg-target.c.inc | 113 +++++++++++++++++++-------------------- 1 file changed, 56 insertions(+), 57 deletions(-) diff --git a/tcg/arm/tcg-target.c.inc b/tcg/arm/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/arm/tcg-target.c.inc +++ b/tcg/arm/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_read(TCGContext *s, TCGReg addrlo, TCGReg addrhi, /* Record the context of a call to the out of line helper code for the slow path for a load or store, so that we can later generate the correct helper code. */ -static void add_qemu_ldst_label(TCGContext *s, bool is_ld, MemOpIdx oi, - TCGReg datalo, TCGReg datahi, TCGReg addrlo, - TCGReg addrhi, tcg_insn_unit *raddr, +static void add_qemu_ldst_label(TCGContext *s, bool is_ld, + MemOpIdx oi, TCGType type, + TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + tcg_insn_unit *raddr, tcg_insn_unit *label_ptr) { TCGLabelQemuLdst *label = new_ldst_label(s); label->is_ld = is_ld; label->oi = oi; + label->type = type; label->datalo_reg = datalo; label->datahi_reg = datahi; label->addrlo_reg = addrlo; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg datalo, } #endif -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is64) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg addrlo, datalo, datahi, addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc; -#ifdef CONFIG_SOFTMMU - int mem_index; - TCGReg addend; - tcg_insn_unit *label_ptr; -#else - unsigned a_bits; -#endif - - datalo = *args++; - datahi = (is64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TARGET_LONG_BITS == 64 ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); + MemOp opc = get_memop(oi); #ifdef CONFIG_SOFTMMU - mem_index = get_mmuidx(oi); - addend = tcg_out_tlb_read(s, addrlo, addrhi, opc, mem_index, 1); + TCGReg addend= tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 1); - /* This a conditional BL only to load a pointer within this opcode into LR - for the slow path. We will not be using the value for a tail call. */ - label_ptr = s->code_ptr; + /* + * This a conditional BL only to load a pointer within this opcode into + * LR for the slow path. We will not be using the value for a tail call. + */ + tcg_insn_unit *label_ptr = s->code_ptr; tcg_out_bl_imm(s, COND_NE, 0); tcg_out_qemu_ld_index(s, opc, datalo, datahi, addrlo, addend, true); - add_qemu_ldst_label(s, true, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, true, oi, data_type, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, MemOp opc, TCGReg datalo, } #endif -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is64) +static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg addrlo, datalo, datahi, addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc; -#ifdef CONFIG_SOFTMMU - int mem_index; - TCGReg addend; - tcg_insn_unit *label_ptr; -#else - unsigned a_bits; -#endif - - datalo = *args++; - datahi = (is64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TARGET_LONG_BITS == 64 ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); + MemOp opc = get_memop(oi); #ifdef CONFIG_SOFTMMU - mem_index = get_mmuidx(oi); - addend = tcg_out_tlb_read(s, addrlo, addrhi, opc, mem_index, 0); + TCGReg addend = tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 0); tcg_out_qemu_st_index(s, COND_EQ, opc, datalo, datahi, addrlo, addend, true); /* The conditional call must come last, as we're going to return here. */ - label_ptr = s->code_ptr; + tcg_insn_unit *label_ptr = s->code_ptr; tcg_out_bl_imm(s, COND_NE, 0); - add_qemu_ldst_label(s, false, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, oi, data_type, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else /* !CONFIG_SOFTMMU */ - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, 0); + if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_ld(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I32); + } else { + tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], + args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, 1); + if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_ld(s, args[0], args[1], args[2], -1, + args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_ld(s, args[0], args[1], args[2], args[3], + args[4], TCG_TYPE_I64); + } break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args, 0); + if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_st(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I32); + } else { + tcg_out_qemu_st(s, args[0], -1, args[1], args[2], + args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, 1); + if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_st(s, args[0], args[1], args[2], -1, + args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_st(s, args[0], args[1], args[2], args[3], + args[4], TCG_TYPE_I64); + } break; case INDEX_op_bswap16_i32: -- 2.34.1 Collect the parts of the host address, and condition, into a struct. Merge tcg_out_qemu_*_{index,direct} and use it. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/arm/tcg-target.c.inc | 248 ++++++++++++++++++--------------------- 1 file changed, 115 insertions(+), 133 deletions(-) diff --git a/tcg/arm/tcg-target.c.inc b/tcg/arm/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/arm/tcg-target.c.inc +++ b/tcg/arm/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_vldst(TCGContext *s, ARMInsn insn, tcg_out32(s, insn | (rn << 16) | encode_vd(rd) | 0xf); } +typedef struct { + ARMCond cond; + TCGReg base; + int index; + bool index_scratch; +} HostAddress; + #ifdef CONFIG_SOFTMMU /* helper signature: helper_ret_ld_mmu(CPUState *env, target_ulong addr, * int mmu_idx, uintptr_t ra) @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) } #endif /* SOFTMMU */ -static void tcg_out_qemu_ld_index(TCGContext *s, MemOp opc, - TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addend, - bool scratch_addend) +static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg datalo, + TCGReg datahi, HostAddress h) { + TCGReg base; + /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((opc & MO_BSWAP) == 0); switch (opc & MO_SSIZE) { case MO_UB: - tcg_out_ld8_r(s, COND_AL, datalo, addrlo, addend); + if (h.index < 0) { + tcg_out_ld8_12(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_ld8_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_SB: - tcg_out_ld8s_r(s, COND_AL, datalo, addrlo, addend); + if (h.index < 0) { + tcg_out_ld8s_8(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_ld8s_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_UW: - tcg_out_ld16u_r(s, COND_AL, datalo, addrlo, addend); + if (h.index < 0) { + tcg_out_ld16u_8(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_ld16u_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_SW: - tcg_out_ld16s_r(s, COND_AL, datalo, addrlo, addend); + if (h.index < 0) { + tcg_out_ld16s_8(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_ld16s_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_UL: - tcg_out_ld32_r(s, COND_AL, datalo, addrlo, addend); + if (h.index < 0) { + tcg_out_ld32_12(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_ld32_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_UQ: /* We used pair allocation for datalo, so already should be aligned. */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_index(TCGContext *s, MemOp opc, tcg_debug_assert(datahi == datalo + 1); /* LDRD requires alignment; double-check that. */ if (get_alignment_bits(opc) >= MO_64) { + if (h.index < 0) { + tcg_out_ldrd_8(s, h.cond, datalo, h.base, 0); + break; + } /* * Rm (the second address op) must not overlap Rt or Rt + 1. * Since datalo is aligned, we can simplify the test via alignment. * Flip the two address arguments if that works. */ - if ((addend & ~1) != datalo) { - tcg_out_ldrd_r(s, COND_AL, datalo, addrlo, addend); + if ((h.index & ~1) != datalo) { + tcg_out_ldrd_r(s, h.cond, datalo, h.base, h.index); break; } - if ((addrlo & ~1) != datalo) { - tcg_out_ldrd_r(s, COND_AL, datalo, addend, addrlo); + if ((h.base & ~1) != datalo) { + tcg_out_ldrd_r(s, h.cond, datalo, h.index, h.base); break; } } - if (scratch_addend) { - tcg_out_ld32_rwb(s, COND_AL, datalo, addend, addrlo); - tcg_out_ld32_12(s, COND_AL, datahi, addend, 4); + if (h.index < 0) { + base = h.base; + if (datalo == h.base) { + tcg_out_mov_reg(s, h.cond, TCG_REG_TMP, base); + base = TCG_REG_TMP; + } + } else if (h.index_scratch) { + tcg_out_ld32_rwb(s, h.cond, datalo, h.index, h.base); + tcg_out_ld32_12(s, h.cond, datahi, h.index, 4); + break; } else { - tcg_out_dat_reg(s, COND_AL, ARITH_ADD, TCG_REG_TMP, - addend, addrlo, SHIFT_IMM_LSL(0)); - tcg_out_ld32_12(s, COND_AL, datalo, TCG_REG_TMP, 0); - tcg_out_ld32_12(s, COND_AL, datahi, TCG_REG_TMP, 4); + tcg_out_dat_reg(s, h.cond, ARITH_ADD, TCG_REG_TMP, + h.base, h.index, SHIFT_IMM_LSL(0)); + base = TCG_REG_TMP; } + tcg_out_ld32_12(s, h.cond, datalo, base, 0); + tcg_out_ld32_12(s, h.cond, datahi, base, 4); break; default: g_assert_not_reached(); } } -#ifndef CONFIG_SOFTMMU -static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg datalo, - TCGReg datahi, TCGReg addrlo) -{ - /* Byte swapping is left to middle-end expansion. */ - tcg_debug_assert((opc & MO_BSWAP) == 0); - - switch (opc & MO_SSIZE) { - case MO_UB: - tcg_out_ld8_12(s, COND_AL, datalo, addrlo, 0); - break; - case MO_SB: - tcg_out_ld8s_8(s, COND_AL, datalo, addrlo, 0); - break; - case MO_UW: - tcg_out_ld16u_8(s, COND_AL, datalo, addrlo, 0); - break; - case MO_SW: - tcg_out_ld16s_8(s, COND_AL, datalo, addrlo, 0); - break; - case MO_UL: - tcg_out_ld32_12(s, COND_AL, datalo, addrlo, 0); - break; - case MO_UQ: - /* We used pair allocation for datalo, so already should be aligned. */ - tcg_debug_assert((datalo & 1) == 0); - tcg_debug_assert(datahi == datalo + 1); - /* LDRD requires alignment; double-check that. */ - if (get_alignment_bits(opc) >= MO_64) { - tcg_out_ldrd_8(s, COND_AL, datalo, addrlo, 0); - } else if (datalo == addrlo) { - tcg_out_ld32_12(s, COND_AL, datahi, addrlo, 4); - tcg_out_ld32_12(s, COND_AL, datalo, addrlo, 0); - } else { - tcg_out_ld32_12(s, COND_AL, datalo, addrlo, 0); - tcg_out_ld32_12(s, COND_AL, datahi, addrlo, 4); - } - break; - default: - g_assert_not_reached(); - } -} -#endif - static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, TCGReg addrlo, TCGReg addrhi, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; #ifdef CONFIG_SOFTMMU - TCGReg addend= tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 1); + h.cond = COND_AL; + h.base = addrlo; + h.index_scratch = true; + h.index = tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 1); /* * This a conditional BL only to load a pointer within this opcode into @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, tcg_insn_unit *label_ptr = s->code_ptr; tcg_out_bl_imm(s, COND_NE, 0); - tcg_out_qemu_ld_index(s, opc, datalo, datahi, addrlo, addend, true); + tcg_out_qemu_ld_direct(s, opc, datalo, datahi, h); add_qemu_ldst_label(s, true, oi, data_type, datalo, datahi, addrlo, addrhi, s->code_ptr, label_ptr); -#else /* !CONFIG_SOFTMMU */ +#else unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } - if (guest_base) { - tcg_out_qemu_ld_index(s, opc, datalo, datahi, - addrlo, TCG_REG_GUEST_BASE, false); - } else { - tcg_out_qemu_ld_direct(s, opc, datalo, datahi, addrlo); - } + + h.cond = COND_AL; + h.base = addrlo; + h.index = guest_base ? TCG_REG_GUEST_BASE : -1; + h.index_scratch = false; + tcg_out_qemu_ld_direct(s, opc, datalo, datahi, h); #endif } -static void tcg_out_qemu_st_index(TCGContext *s, ARMCond cond, MemOp opc, - TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addend, - bool scratch_addend) -{ - /* Byte swapping is left to middle-end expansion. */ - tcg_debug_assert((opc & MO_BSWAP) == 0); - - switch (opc & MO_SIZE) { - case MO_8: - tcg_out_st8_r(s, cond, datalo, addrlo, addend); - break; - case MO_16: - tcg_out_st16_r(s, cond, datalo, addrlo, addend); - break; - case MO_32: - tcg_out_st32_r(s, cond, datalo, addrlo, addend); - break; - case MO_64: - /* We used pair allocation for datalo, so already should be aligned. */ - tcg_debug_assert((datalo & 1) == 0); - tcg_debug_assert(datahi == datalo + 1); - /* STRD requires alignment; double-check that. */ - if (get_alignment_bits(opc) >= MO_64) { - tcg_out_strd_r(s, cond, datalo, addrlo, addend); - } else if (scratch_addend) { - tcg_out_st32_rwb(s, cond, datalo, addend, addrlo); - tcg_out_st32_12(s, cond, datahi, addend, 4); - } else { - tcg_out_dat_reg(s, cond, ARITH_ADD, TCG_REG_TMP, - addend, addrlo, SHIFT_IMM_LSL(0)); - tcg_out_st32_12(s, cond, datalo, TCG_REG_TMP, 0); - tcg_out_st32_12(s, cond, datahi, TCG_REG_TMP, 4); - } - break; - default: - g_assert_not_reached(); - } -} - -#ifndef CONFIG_SOFTMMU static void tcg_out_qemu_st_direct(TCGContext *s, MemOp opc, TCGReg datalo, - TCGReg datahi, TCGReg addrlo) + TCGReg datahi, HostAddress h) { /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((opc & MO_BSWAP) == 0); switch (opc & MO_SIZE) { case MO_8: - tcg_out_st8_12(s, COND_AL, datalo, addrlo, 0); + if (h.index < 0) { + tcg_out_st8_12(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_st8_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_16: - tcg_out_st16_8(s, COND_AL, datalo, addrlo, 0); + if (h.index < 0) { + tcg_out_st16_8(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_st16_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_32: - tcg_out_st32_12(s, COND_AL, datalo, addrlo, 0); + if (h.index < 0) { + tcg_out_st32_12(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_st32_r(s, h.cond, datalo, h.base, h.index); + } break; case MO_64: /* We used pair allocation for datalo, so already should be aligned. */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, MemOp opc, TCGReg datalo, tcg_debug_assert(datahi == datalo + 1); /* STRD requires alignment; double-check that. */ if (get_alignment_bits(opc) >= MO_64) { - tcg_out_strd_8(s, COND_AL, datalo, addrlo, 0); + if (h.index < 0) { + tcg_out_strd_8(s, h.cond, datalo, h.base, 0); + } else { + tcg_out_strd_r(s, h.cond, datalo, h.base, h.index); + } + } else if (h.index_scratch) { + tcg_out_st32_rwb(s, h.cond, datalo, h.index, h.base); + tcg_out_st32_12(s, h.cond, datahi, h.index, 4); } else { - tcg_out_st32_12(s, COND_AL, datalo, addrlo, 0); - tcg_out_st32_12(s, COND_AL, datahi, addrlo, 4); + tcg_out_dat_reg(s, h.cond, ARITH_ADD, TCG_REG_TMP, + h.base, h.index, SHIFT_IMM_LSL(0)); + tcg_out_st32_12(s, h.cond, datalo, TCG_REG_TMP, 0); + tcg_out_st32_12(s, h.cond, datahi, TCG_REG_TMP, 4); } break; default: g_assert_not_reached(); } } -#endif static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, TCGReg addrlo, TCGReg addrhi, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; #ifdef CONFIG_SOFTMMU - TCGReg addend = tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 0); - - tcg_out_qemu_st_index(s, COND_EQ, opc, datalo, datahi, - addrlo, addend, true); + h.cond = COND_EQ; + h.base = addrlo; + h.index_scratch = true; + h.index = tcg_out_tlb_read(s, addrlo, addrhi, opc, get_mmuidx(oi), 0); + tcg_out_qemu_st_direct(s, opc, datalo, datahi, h); /* The conditional call must come last, as we're going to return here. */ tcg_insn_unit *label_ptr = s->code_ptr; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, add_qemu_ldst_label(s, false, oi, data_type, datalo, datahi, addrlo, addrhi, s->code_ptr, label_ptr); -#else /* !CONFIG_SOFTMMU */ +#else unsigned a_bits = get_alignment_bits(opc); + + h.cond = COND_AL; if (a_bits) { tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); + h.cond = COND_EQ; } - if (guest_base) { - tcg_out_qemu_st_index(s, COND_AL, opc, datalo, datahi, - addrlo, TCG_REG_GUEST_BASE, false); - } else { - tcg_out_qemu_st_direct(s, opc, datalo, datahi, addrlo); - } + + h.base = addrlo; + h.index = guest_base ? TCG_REG_GUEST_BASE : -1; + h.index_scratch = false; + tcg_out_qemu_st_direct(s, opc, datalo, datahi, h); #endif } -- 2.34.1 Interpret the variable argument placement in the caller. Shift some code around slightly to share more between softmmu and user-only. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/loongarch64/tcg-target.c.inc | 100 +++++++++++++------------------ 1 file changed, 42 insertions(+), 58 deletions(-) diff --git a/tcg/loongarch64/tcg-target.c.inc b/tcg/loongarch64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target.c.inc +++ b/tcg/loongarch64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_indexed(TCGContext *s, TCGReg rd, TCGReg rj, } } -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, TCGType type) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, + MemOpIdx oi, TCGType data_type) { - TCGReg addr_regl; - TCGReg data_regl; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) + MemOp opc = get_memop(oi); + TCGReg base, index; + +#ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr[1]; -#else - unsigned a_bits; -#endif - TCGReg base; - data_regl = *args++; - addr_regl = *args++; - oi = *args++; - opc = get_memop(oi); - -#if defined(CONFIG_SOFTMMU) - tcg_out_tlb_load(s, addr_regl, oi, label_ptr, 1); - base = tcg_out_zext_addr_if_32_bit(s, addr_regl, TCG_REG_TMP0); - tcg_out_qemu_ld_indexed(s, data_regl, base, TCG_REG_TMP2, opc, type); - add_qemu_ldst_label(s, 1, oi, type, - data_regl, addr_regl, - s->code_ptr, label_ptr); + tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 1); + index = TCG_REG_TMP2; #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { - tcg_out_test_alignment(s, true, addr_regl, a_bits); + tcg_out_test_alignment(s, true, addr_reg, a_bits); } - base = tcg_out_zext_addr_if_32_bit(s, addr_regl, TCG_REG_TMP0); - TCGReg guest_base_reg = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; - tcg_out_qemu_ld_indexed(s, data_regl, base, guest_base_reg, opc, type); + index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; +#endif + + base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); + tcg_out_qemu_ld_indexed(s, data_reg, base, index, opc, data_type); + +#ifdef CONFIG_SOFTMMU + add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_indexed(TCGContext *s, TCGReg data, } } -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, TCGType type) +static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, + MemOpIdx oi, TCGType data_type) { - TCGReg addr_regl; - TCGReg data_regl; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) + MemOp opc = get_memop(oi); + TCGReg base, index; + +#ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr[1]; -#else - unsigned a_bits; -#endif - TCGReg base; - data_regl = *args++; - addr_regl = *args++; - oi = *args++; - opc = get_memop(oi); - -#if defined(CONFIG_SOFTMMU) - tcg_out_tlb_load(s, addr_regl, oi, label_ptr, 0); - base = tcg_out_zext_addr_if_32_bit(s, addr_regl, TCG_REG_TMP0); - tcg_out_qemu_st_indexed(s, data_regl, base, TCG_REG_TMP2, opc); - add_qemu_ldst_label(s, 0, oi, type, - data_regl, addr_regl, - s->code_ptr, label_ptr); + tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 0); + index = TCG_REG_TMP2; #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { - tcg_out_test_alignment(s, false, addr_regl, a_bits); + tcg_out_test_alignment(s, false, addr_reg, a_bits); } - base = tcg_out_zext_addr_if_32_bit(s, addr_regl, TCG_REG_TMP0); - TCGReg guest_base_reg = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; - tcg_out_qemu_st_indexed(s, data_regl, base, guest_base_reg, opc); + index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; +#endif + + base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); + tcg_out_qemu_st_indexed(s, data_reg, base, index, opc); + +#ifdef CONFIG_SOFTMMU + add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, TCG_TYPE_I32); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, TCG_TYPE_I64); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args, TCG_TYPE_I32); + tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, TCG_TYPE_I64); + tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I64); break; case INDEX_op_mov_i32: /* Always emitted via tcg_out_mov. */ -- 2.34.1 Collect the 2 parts of the host address into a struct. Reorg tcg_out_qemu_{ld,st}_direct to use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/loongarch64/tcg-target.c.inc | 55 +++++++++++++++++--------------- 1 file changed, 30 insertions(+), 25 deletions(-) diff --git a/tcg/loongarch64/tcg-target.c.inc b/tcg/loongarch64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target.c.inc +++ b/tcg/loongarch64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_zext_addr_if_32_bit(TCGContext *s, return addr; } -static void tcg_out_qemu_ld_indexed(TCGContext *s, TCGReg rd, TCGReg rj, - TCGReg rk, MemOp opc, TCGType type) +typedef struct { + TCGReg base; + TCGReg index; +} HostAddress; + +static void tcg_out_qemu_ld_indexed(TCGContext *s, MemOp opc, TCGType type, + TCGReg rd, HostAddress h) { /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((opc & MO_BSWAP) == 0); switch (opc & MO_SSIZE) { case MO_UB: - tcg_out_opc_ldx_bu(s, rd, rj, rk); + tcg_out_opc_ldx_bu(s, rd, h.base, h.index); break; case MO_SB: - tcg_out_opc_ldx_b(s, rd, rj, rk); + tcg_out_opc_ldx_b(s, rd, h.base, h.index); break; case MO_UW: - tcg_out_opc_ldx_hu(s, rd, rj, rk); + tcg_out_opc_ldx_hu(s, rd, h.base, h.index); break; case MO_SW: - tcg_out_opc_ldx_h(s, rd, rj, rk); + tcg_out_opc_ldx_h(s, rd, h.base, h.index); break; case MO_UL: if (type == TCG_TYPE_I64) { - tcg_out_opc_ldx_wu(s, rd, rj, rk); + tcg_out_opc_ldx_wu(s, rd, h.base, h.index); break; } /* fallthrough */ case MO_SL: - tcg_out_opc_ldx_w(s, rd, rj, rk); + tcg_out_opc_ldx_w(s, rd, h.base, h.index); break; case MO_UQ: - tcg_out_opc_ldx_d(s, rd, rj, rk); + tcg_out_opc_ldx_d(s, rd, h.base, h.index); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); - TCGReg base, index; + HostAddress h; #ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr[1]; tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 1); - index = TCG_REG_TMP2; + h.index = TCG_REG_TMP2; #else unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addr_reg, a_bits); } - index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; + h.index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; #endif - base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); - tcg_out_qemu_ld_indexed(s, data_reg, base, index, opc, data_type); + h.base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); + tcg_out_qemu_ld_indexed(s, opc, data_type, data_reg, h); #ifdef CONFIG_SOFTMMU add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, #endif } -static void tcg_out_qemu_st_indexed(TCGContext *s, TCGReg data, - TCGReg rj, TCGReg rk, MemOp opc) +static void tcg_out_qemu_st_indexed(TCGContext *s, MemOp opc, + TCGReg rd, HostAddress h) { /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((opc & MO_BSWAP) == 0); switch (opc & MO_SIZE) { case MO_8: - tcg_out_opc_stx_b(s, data, rj, rk); + tcg_out_opc_stx_b(s, rd, h.base, h.index); break; case MO_16: - tcg_out_opc_stx_h(s, data, rj, rk); + tcg_out_opc_stx_h(s, rd, h.base, h.index); break; case MO_32: - tcg_out_opc_stx_w(s, data, rj, rk); + tcg_out_opc_stx_w(s, rd, h.base, h.index); break; case MO_64: - tcg_out_opc_stx_d(s, data, rj, rk); + tcg_out_opc_stx_d(s, rd, h.base, h.index); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); - TCGReg base, index; + HostAddress h; #ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr[1]; tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 0); - index = TCG_REG_TMP2; + h.index = TCG_REG_TMP2; #else unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addr_reg, a_bits); } - index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; + h.index = USE_GUEST_BASE ? TCG_GUEST_BASE_REG : TCG_REG_ZERO; #endif - base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); - tcg_out_qemu_st_indexed(s, data_reg, base, index, opc); + h.base = tcg_out_zext_addr_if_32_bit(s, addr_reg, TCG_REG_TMP0); + tcg_out_qemu_st_indexed(s, opc, data_reg, h); #ifdef CONFIG_SOFTMMU add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, -- 2.34.1 Interpret the variable argument placement in the caller. There are several places where we already convert back from bool to type. Clean things up by using type throughout. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/mips/tcg-target.c.inc | 186 +++++++++++++++++++------------------- 1 file changed, 95 insertions(+), 91 deletions(-) diff --git a/tcg/mips/tcg-target.c.inc b/tcg/mips/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/mips/tcg-target.c.inc +++ b/tcg/mips/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) #endif /* SOFTMMU */ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, - TCGReg base, MemOp opc, bool is_64) + TCGReg base, MemOp opc, TCGType type) { switch (opc & (MO_SSIZE | MO_BSWAP)) { case MO_UB: @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, tcg_out_opc_imm(s, OPC_LH, lo, base, 0); break; case MO_UL | MO_BSWAP: - if (TCG_TARGET_REG_BITS == 64 && is_64) { + if (TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I64) { if (use_mips32r2_instructions) { tcg_out_opc_imm(s, OPC_LWU, lo, base, 0); tcg_out_bswap32(s, lo, lo, TCG_BSWAP_IZ | TCG_BSWAP_OZ); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, } break; case MO_UL: - if (TCG_TARGET_REG_BITS == 64 && is_64) { + if (TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I64) { tcg_out_opc_imm(s, OPC_LWU, lo, base, 0); break; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, } static void tcg_out_qemu_ld_unalign(TCGContext *s, TCGReg lo, TCGReg hi, - TCGReg base, MemOp opc, bool is_64) + TCGReg base, MemOp opc, TCGType type) { const MIPSInsn lw1 = MIPS_BE ? OPC_LWL : OPC_LWR; const MIPSInsn lw2 = MIPS_BE ? OPC_LWR : OPC_LWL; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_unalign(TCGContext *s, TCGReg lo, TCGReg hi, case MO_UL: tcg_out_opc_imm(s, lw1, lo, base, 0); tcg_out_opc_imm(s, lw2, lo, base, 3); - if (TCG_TARGET_REG_BITS == 64 && is_64 && !sgn) { + if (TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I64 && !sgn) { tcg_out_ext32u(s, lo, lo); } break; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_unalign(TCGContext *s, TCGReg lo, TCGReg hi, tcg_out_opc_imm(s, lw1, lo, base, 0); tcg_out_opc_imm(s, lw2, lo, base, 3); tcg_out_bswap32(s, lo, lo, - TCG_TARGET_REG_BITS == 64 && is_64 + TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I64 ? (sgn ? TCG_BSWAP_OS : TCG_BSWAP_OZ) : 0); } else { const tcg_insn_unit *subr = - (TCG_TARGET_REG_BITS == 64 && is_64 && !sgn + (TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I64 && !sgn ? bswap32u_addr : bswap32_addr); tcg_out_opc_imm(s, lw1, TCG_TMP0, base, 0); tcg_out_bswap_subr(s, subr); /* delay slot */ tcg_out_opc_imm(s, lw2, TCG_TMP0, base, 3); - tcg_out_mov(s, is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32, lo, TCG_TMP3); + tcg_out_mov(s, type, lo, TCG_TMP3); } break; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_unalign(TCGContext *s, TCGReg lo, TCGReg hi, } } -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg addr_regl, addr_regh __attribute__((unused)); - TCGReg data_regl, data_regh; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) - tcg_insn_unit *label_ptr[2]; -#else -#endif - unsigned a_bits, s_bits; - TCGReg base = TCG_REG_A0; - - data_regl = *args++; - data_regh = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addr_regl = *args++; - addr_regh = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - a_bits = get_alignment_bits(opc); - s_bits = opc & MO_SIZE; + MemOp opc = get_memop(oi); + unsigned a_bits = get_alignment_bits(opc); + unsigned s_bits = opc & MO_SIZE; + TCGReg base; /* * R6 removes the left/right instructions but requires the * system to support misaligned memory accesses. */ #if defined(CONFIG_SOFTMMU) - tcg_out_tlb_load(s, base, addr_regl, addr_regh, oi, label_ptr, 1); + tcg_insn_unit *label_ptr[2]; + + base = TCG_REG_A0; + tcg_out_tlb_load(s, base, addrlo, addrhi, oi, label_ptr, 1); if (use_mips32r6_instructions || a_bits >= s_bits) { - tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_direct(s, datalo, datahi, base, opc, data_type); } else { - tcg_out_qemu_ld_unalign(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_unalign(s, datalo, datahi, base, opc, data_type); } - add_qemu_ldst_label(s, 1, oi, - (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_regl, data_regh, addr_regl, addr_regh, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, true, oi, data_type, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else + base = addrlo; if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { - tcg_out_ext32u(s, base, addr_regl); - addr_regl = base; + tcg_out_ext32u(s, TCG_REG_A0, base); + base = TCG_REG_A0; } - if (guest_base == 0 && data_regl != addr_regl) { - base = addr_regl; - } else if (guest_base == (int16_t)guest_base) { - tcg_out_opc_imm(s, ALIAS_PADDI, base, addr_regl, guest_base); - } else { - tcg_out_opc_reg(s, ALIAS_PADD, base, TCG_GUEST_BASE_REG, addr_regl); + if (guest_base) { + if (guest_base == (int16_t)guest_base) { + tcg_out_opc_imm(s, ALIAS_PADDI, TCG_REG_A0, base, guest_base); + } else { + tcg_out_opc_reg(s, ALIAS_PADD, TCG_REG_A0, base, + TCG_GUEST_BASE_REG); + } + base = TCG_REG_A0; } if (use_mips32r6_instructions) { if (a_bits) { - tcg_out_test_alignment(s, true, addr_regl, addr_regh, a_bits); + tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } - tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_direct(s, datalo, datahi, base, opc, data_type); } else { if (a_bits && a_bits != s_bits) { - tcg_out_test_alignment(s, true, addr_regl, addr_regh, a_bits); + tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } if (a_bits >= s_bits) { - tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_direct(s, datalo, datahi, base, opc, data_type); } else { - tcg_out_qemu_ld_unalign(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_unalign(s, datalo, datahi, base, opc, data_type); } } #endif @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_unalign(TCGContext *s, TCGReg lo, TCGReg hi, g_assert_not_reached(); } } -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) -{ - TCGReg addr_regl, addr_regh __attribute__((unused)); - TCGReg data_regl, data_regh; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) - tcg_insn_unit *label_ptr[2]; -#endif - unsigned a_bits, s_bits; - TCGReg base = TCG_REG_A0; - data_regl = *args++; - data_regh = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addr_regl = *args++; - addr_regh = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - a_bits = get_alignment_bits(opc); - s_bits = opc & MO_SIZE; +static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) +{ + MemOp opc = get_memop(oi); + unsigned a_bits = get_alignment_bits(opc); + unsigned s_bits = opc & MO_SIZE; + TCGReg base; /* * R6 removes the left/right instructions but requires the * system to support misaligned memory accesses. */ #if defined(CONFIG_SOFTMMU) - tcg_out_tlb_load(s, base, addr_regl, addr_regh, oi, label_ptr, 0); + tcg_insn_unit *label_ptr[2]; + + base = TCG_REG_A0; + tcg_out_tlb_load(s, base, addrlo, addrhi, oi, label_ptr, 0); if (use_mips32r6_instructions || a_bits >= s_bits) { - tcg_out_qemu_st_direct(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_direct(s, datalo, datahi, base, opc); } else { - tcg_out_qemu_st_unalign(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_unalign(s, datalo, datahi, base, opc); } - add_qemu_ldst_label(s, 0, oi, - (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_regl, data_regh, addr_regl, addr_regh, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, oi, data_type, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #else + base = addrlo; if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { - tcg_out_ext32u(s, base, addr_regl); - addr_regl = base; + tcg_out_ext32u(s, TCG_REG_A0, base); + base = TCG_REG_A0; } - if (guest_base == 0) { - base = addr_regl; - } else if (guest_base == (int16_t)guest_base) { - tcg_out_opc_imm(s, ALIAS_PADDI, base, addr_regl, guest_base); - } else { - tcg_out_opc_reg(s, ALIAS_PADD, base, TCG_GUEST_BASE_REG, addr_regl); + if (guest_base) { + if (guest_base == (int16_t)guest_base) { + tcg_out_opc_imm(s, ALIAS_PADDI, TCG_REG_A0, base, guest_base); + } else { + tcg_out_opc_reg(s, ALIAS_PADD, TCG_REG_A0, base, + TCG_GUEST_BASE_REG); + } + base = TCG_REG_A0; } if (use_mips32r6_instructions) { if (a_bits) { - tcg_out_test_alignment(s, true, addr_regl, addr_regh, a_bits); + tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } - tcg_out_qemu_st_direct(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_direct(s, datalo, datahi, base, opc); } else { if (a_bits && a_bits != s_bits) { - tcg_out_test_alignment(s, true, addr_regl, addr_regh, a_bits); + tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } if (a_bits >= s_bits) { - tcg_out_qemu_st_direct(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_direct(s, datalo, datahi, base, opc); } else { - tcg_out_qemu_st_unalign(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_unalign(s, datalo, datahi, base, opc); } } #endif @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, false); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); + } else { + tcg_out_qemu_ld(s, a0, 0, a1, a2, args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, true); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_ld(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_ld(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); + } break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args, false); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); + } else { + tcg_out_qemu_st(s, a0, 0, a1, a2, args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, true); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_st(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_st(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); + } break; case INDEX_op_add2_i32: -- 2.34.1 Interpret the variable argument placement in the caller. Pass data_type instead of is64 -- there are several places where we already convert back from bool to type. Clean things up by using type throughout. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/ppc/tcg-target.c.inc | 110 +++++++++++++++++++++------------------ 1 file changed, 59 insertions(+), 51 deletions(-) diff --git a/tcg/ppc/tcg-target.c.inc b/tcg/ppc/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/ppc/tcg-target.c.inc +++ b/tcg/ppc/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_read(TCGContext *s, MemOp opc, /* Record the context of a call to the out of line helper code for the slow path for a load or store, so that we can later generate the correct helper code. */ -static void add_qemu_ldst_label(TCGContext *s, bool is_ld, MemOpIdx oi, +static void add_qemu_ldst_label(TCGContext *s, bool is_ld, + TCGType type, MemOpIdx oi, TCGReg datalo_reg, TCGReg datahi_reg, TCGReg addrlo_reg, TCGReg addrhi_reg, tcg_insn_unit *raddr, tcg_insn_unit *lptr) @@ -XXX,XX +XXX,XX @@ static void add_qemu_ldst_label(TCGContext *s, bool is_ld, MemOpIdx oi, TCGLabelQemuLdst *label = new_ldst_label(s); label->is_ld = is_ld; + label->type = type; label->oi = oi; label->datalo_reg = datalo_reg; label->datahi_reg = datahi_reg; @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) #endif /* SOFTMMU */ -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg datalo, datahi, addrlo, rbase; - TCGReg addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc, s_bits; + MemOp opc = get_memop(oi); + MemOp s_bits = opc & MO_SIZE; + TCGReg rbase; + #ifdef CONFIG_SOFTMMU - int mem_index; tcg_insn_unit *label_ptr; -#else - unsigned a_bits; -#endif - datalo = *args++; - datahi = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - s_bits = opc & MO_SIZE; - -#ifdef CONFIG_SOFTMMU - mem_index = get_mmuidx(oi); - addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, mem_index, true); + addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), true); /* Load a pointer into the current opcode w/conditional branch-link. */ label_ptr = s->code_ptr; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) rbase = TCG_REG_R3; #else /* !CONFIG_SOFTMMU */ - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) } #ifdef CONFIG_SOFTMMU - add_qemu_ldst_label(s, true, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, true, data_type, oi, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #endif } -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) +static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, + TCGReg addrlo, TCGReg addrhi, + MemOpIdx oi, TCGType data_type) { - TCGReg datalo, datahi, addrlo, rbase; - TCGReg addrhi __attribute__((unused)); - MemOpIdx oi; - MemOp opc, s_bits; + MemOp opc = get_memop(oi); + MemOp s_bits = opc & MO_SIZE; + TCGReg rbase; + #ifdef CONFIG_SOFTMMU - int mem_index; tcg_insn_unit *label_ptr; -#else - unsigned a_bits; -#endif - datalo = *args++; - datahi = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addrlo = *args++; - addrhi = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); - oi = *args++; - opc = get_memop(oi); - s_bits = opc & MO_SIZE; - -#ifdef CONFIG_SOFTMMU - mem_index = get_mmuidx(oi); - addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, mem_index, false); + addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), false); /* Load a pointer into the current opcode w/conditional branch-link. */ label_ptr = s->code_ptr; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) rbase = TCG_REG_R3; #else /* !CONFIG_SOFTMMU */ - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) } #ifdef CONFIG_SOFTMMU - add_qemu_ldst_label(s, false, oi, datalo, datahi, addrlo, addrhi, - s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, data_type, oi, datalo, datahi, + addrlo, addrhi, s->code_ptr, label_ptr); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, false); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_ld(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I32); + } else { + tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], + args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, true); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_ld(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_ld(s, args[0], args[1], args[2], -1, + args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_ld(s, args[0], args[1], args[2], args[3], + args[4], TCG_TYPE_I64); + } break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args, false); + if (TCG_TARGET_REG_BITS >= TARGET_LONG_BITS) { + tcg_out_qemu_st(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I32); + } else { + tcg_out_qemu_st(s, args[0], -1, args[1], args[2], + args[3], TCG_TYPE_I32); + } break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, true); + if (TCG_TARGET_REG_BITS == 64) { + tcg_out_qemu_st(s, args[0], -1, args[1], -1, + args[2], TCG_TYPE_I64); + } else if (TARGET_LONG_BITS == 32) { + tcg_out_qemu_st(s, args[0], args[1], args[2], -1, + args[3], TCG_TYPE_I64); + } else { + tcg_out_qemu_st(s, args[0], args[1], args[2], args[3], + args[4], TCG_TYPE_I64); + } break; case INDEX_op_setcond_i32: -- 2.34.1 Collect the parts of the host address into a struct. Reorg tcg_out_qemu_{ld,st} to use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/ppc/tcg-target.c.inc | 90 +++++++++++++++++++++------------------- 1 file changed, 47 insertions(+), 43 deletions(-) diff --git a/tcg/ppc/tcg-target.c.inc b/tcg/ppc/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/ppc/tcg-target.c.inc +++ b/tcg/ppc/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) { return tcg_out_fail_alignment(s, l); } - #endif /* SOFTMMU */ +typedef struct { + TCGReg base; + TCGReg index; +} HostAddress; + static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, TCGReg addrlo, TCGReg addrhi, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); MemOp s_bits = opc & MO_SIZE; - TCGReg rbase; + HostAddress h; #ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr; - addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), true); + h.index = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), true); + h.base = TCG_REG_R3; /* Load a pointer into the current opcode w/conditional branch-link. */ label_ptr = s->code_ptr; tcg_out32(s, BC | BI(7, CR_EQ) | BO_COND_FALSE | LK); - - rbase = TCG_REG_R3; #else /* !CONFIG_SOFTMMU */ unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addrlo, addrhi, a_bits); } - rbase = guest_base ? TCG_GUEST_BASE_REG : 0; + h.base = guest_base ? TCG_GUEST_BASE_REG : 0; + h.index = addrlo; if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { tcg_out_ext32u(s, TCG_REG_TMP1, addrlo); - addrlo = TCG_REG_TMP1; + h.index = TCG_REG_TMP1; } #endif if (TCG_TARGET_REG_BITS == 32 && s_bits == MO_64) { if (opc & MO_BSWAP) { - tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4)); - tcg_out32(s, LWBRX | TAB(datalo, rbase, addrlo)); - tcg_out32(s, LWBRX | TAB(datahi, rbase, TCG_REG_R0)); - } else if (rbase != 0) { - tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4)); - tcg_out32(s, LWZX | TAB(datahi, rbase, addrlo)); - tcg_out32(s, LWZX | TAB(datalo, rbase, TCG_REG_R0)); - } else if (addrlo == datahi) { - tcg_out32(s, LWZ | TAI(datalo, addrlo, 4)); - tcg_out32(s, LWZ | TAI(datahi, addrlo, 0)); + tcg_out32(s, ADDI | TAI(TCG_REG_R0, h.index, 4)); + tcg_out32(s, LWBRX | TAB(datalo, h.base, h.index)); + tcg_out32(s, LWBRX | TAB(datahi, h.base, TCG_REG_R0)); + } else if (h.base != 0) { + tcg_out32(s, ADDI | TAI(TCG_REG_R0, h.index, 4)); + tcg_out32(s, LWZX | TAB(datahi, h.base, h.index)); + tcg_out32(s, LWZX | TAB(datalo, h.base, TCG_REG_R0)); + } else if (h.index == datahi) { + tcg_out32(s, LWZ | TAI(datalo, h.index, 4)); + tcg_out32(s, LWZ | TAI(datahi, h.index, 0)); } else { - tcg_out32(s, LWZ | TAI(datahi, addrlo, 0)); - tcg_out32(s, LWZ | TAI(datalo, addrlo, 4)); + tcg_out32(s, LWZ | TAI(datahi, h.index, 0)); + tcg_out32(s, LWZ | TAI(datalo, h.index, 4)); } } else { uint32_t insn = qemu_ldx_opc[opc & (MO_BSWAP | MO_SSIZE)]; if (!have_isa_2_06 && insn == LDBRX) { - tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4)); - tcg_out32(s, LWBRX | TAB(datalo, rbase, addrlo)); - tcg_out32(s, LWBRX | TAB(TCG_REG_R0, rbase, TCG_REG_R0)); + tcg_out32(s, ADDI | TAI(TCG_REG_R0, h.index, 4)); + tcg_out32(s, LWBRX | TAB(datalo, h.base, h.index)); + tcg_out32(s, LWBRX | TAB(TCG_REG_R0, h.base, TCG_REG_R0)); tcg_out_rld(s, RLDIMI, datalo, TCG_REG_R0, 32, 0); } else if (insn) { - tcg_out32(s, insn | TAB(datalo, rbase, addrlo)); + tcg_out32(s, insn | TAB(datalo, h.base, h.index)); } else { insn = qemu_ldx_opc[opc & (MO_SIZE | MO_BSWAP)]; - tcg_out32(s, insn | TAB(datalo, rbase, addrlo)); + tcg_out32(s, insn | TAB(datalo, h.base, h.index)); tcg_out_movext(s, TCG_TYPE_REG, datalo, TCG_TYPE_REG, opc & MO_SSIZE, datalo); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, { MemOp opc = get_memop(oi); MemOp s_bits = opc & MO_SIZE; - TCGReg rbase; + HostAddress h; #ifdef CONFIG_SOFTMMU tcg_insn_unit *label_ptr; - addrlo = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), false); + h.index = tcg_out_tlb_read(s, opc, addrlo, addrhi, get_mmuidx(oi), false); + h.base = TCG_REG_R3; /* Load a pointer into the current opcode w/conditional branch-link. */ label_ptr = s->code_ptr; tcg_out32(s, BC | BI(7, CR_EQ) | BO_COND_FALSE | LK); - - rbase = TCG_REG_R3; #else /* !CONFIG_SOFTMMU */ unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addrlo, addrhi, a_bits); } - rbase = guest_base ? TCG_GUEST_BASE_REG : 0; + h.base = guest_base ? TCG_GUEST_BASE_REG : 0; + h.index = addrlo; if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { tcg_out_ext32u(s, TCG_REG_TMP1, addrlo); - addrlo = TCG_REG_TMP1; + h.index = TCG_REG_TMP1; } #endif if (TCG_TARGET_REG_BITS == 32 && s_bits == MO_64) { if (opc & MO_BSWAP) { - tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4)); - tcg_out32(s, STWBRX | SAB(datalo, rbase, addrlo)); - tcg_out32(s, STWBRX | SAB(datahi, rbase, TCG_REG_R0)); - } else if (rbase != 0) { - tcg_out32(s, ADDI | TAI(TCG_REG_R0, addrlo, 4)); - tcg_out32(s, STWX | SAB(datahi, rbase, addrlo)); - tcg_out32(s, STWX | SAB(datalo, rbase, TCG_REG_R0)); + tcg_out32(s, ADDI | TAI(TCG_REG_R0, h.index, 4)); + tcg_out32(s, STWBRX | SAB(datalo, h.base, h.index)); + tcg_out32(s, STWBRX | SAB(datahi, h.base, TCG_REG_R0)); + } else if (h.base != 0) { + tcg_out32(s, ADDI | TAI(TCG_REG_R0, h.index, 4)); + tcg_out32(s, STWX | SAB(datahi, h.base, h.index)); + tcg_out32(s, STWX | SAB(datalo, h.base, TCG_REG_R0)); } else { - tcg_out32(s, STW | TAI(datahi, addrlo, 0)); - tcg_out32(s, STW | TAI(datalo, addrlo, 4)); + tcg_out32(s, STW | TAI(datahi, h.index, 0)); + tcg_out32(s, STW | TAI(datalo, h.index, 4)); } } else { uint32_t insn = qemu_stx_opc[opc & (MO_BSWAP | MO_SIZE)]; if (!have_isa_2_06 && insn == STDBRX) { - tcg_out32(s, STWBRX | SAB(datalo, rbase, addrlo)); - tcg_out32(s, ADDI | TAI(TCG_REG_TMP1, addrlo, 4)); + tcg_out32(s, STWBRX | SAB(datalo, h.base, h.index)); + tcg_out32(s, ADDI | TAI(TCG_REG_TMP1, h.index, 4)); tcg_out_shri64(s, TCG_REG_R0, datalo, 32); - tcg_out32(s, STWBRX | SAB(TCG_REG_R0, rbase, TCG_REG_TMP1)); + tcg_out32(s, STWBRX | SAB(TCG_REG_R0, h.base, TCG_REG_TMP1)); } else { - tcg_out32(s, insn | SAB(datalo, rbase, addrlo)); + tcg_out32(s, insn | SAB(datalo, h.base, h.index)); } } -- 2.34.1 The port currently does not support "oversize" guests, which means riscv32 can only target 32-bit guests. We will soon be building TCG once for all guests. This implies that we can only support riscv64. Since all Linux distributions target riscv64 not riscv32, this is not much of a restriction and simplifies the code. The brcond2 and setcond2 opcodes are exclusive to 32-bit hosts, so we can and should remove the stubs. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/riscv/tcg-target-con-set.h | 8 -- tcg/riscv/tcg-target.h | 22 ++-- tcg/riscv/tcg-target.c.inc | 232 +++++++++------------------------ 3 files changed, 72 insertions(+), 190 deletions(-) diff --git a/tcg/riscv/tcg-target-con-set.h b/tcg/riscv/tcg-target-con-set.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target-con-set.h +++ b/tcg/riscv/tcg-target-con-set.h @@ -XXX,XX +XXX,XX @@ C_O0_I1(r) C_O0_I2(LZ, L) C_O0_I2(rZ, r) C_O0_I2(rZ, rZ) -C_O0_I3(LZ, L, L) -C_O0_I3(LZ, LZ, L) -C_O0_I4(LZ, LZ, L, L) -C_O0_I4(rZ, rZ, rZ, rZ) C_O1_I1(r, L) C_O1_I1(r, r) -C_O1_I2(r, L, L) C_O1_I2(r, r, ri) C_O1_I2(r, r, rI) C_O1_I2(r, rZ, rN) C_O1_I2(r, rZ, rZ) -C_O1_I4(r, rZ, rZ, rZ, rZ) -C_O2_I1(r, r, L) -C_O2_I2(r, r, L, L) C_O2_I4(r, r, rZ, rZ, rM, rM) diff --git a/tcg/riscv/tcg-target.h b/tcg/riscv/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target.h +++ b/tcg/riscv/tcg-target.h @@ -XXX,XX +XXX,XX @@ #ifndef RISCV_TCG_TARGET_H #define RISCV_TCG_TARGET_H -#if __riscv_xlen == 32 -# define TCG_TARGET_REG_BITS 32 -#elif __riscv_xlen == 64 -# define TCG_TARGET_REG_BITS 64 +/* + * We don't support oversize guests. + * Since we will only build tcg once, this in turn requires a 64-bit host. + */ +#if __riscv_xlen != 64 +#error "unsupported code generation mode" #endif +#define TCG_TARGET_REG_BITS 64 #define TCG_TARGET_INSN_UNIT_SIZE 4 #define TCG_TARGET_TLB_DISPLACEMENT_BITS 20 @@ -XXX,XX +XXX,XX @@ typedef enum { #define TCG_TARGET_STACK_ALIGN 16 #define TCG_TARGET_CALL_STACK_OFFSET 0 #define TCG_TARGET_CALL_ARG_I32 TCG_CALL_ARG_NORMAL -#if TCG_TARGET_REG_BITS == 32 -#define TCG_TARGET_CALL_ARG_I64 TCG_CALL_ARG_EVEN -#define TCG_TARGET_CALL_ARG_I128 TCG_CALL_ARG_EVEN -#else #define TCG_TARGET_CALL_ARG_I64 TCG_CALL_ARG_NORMAL #define TCG_TARGET_CALL_ARG_I128 TCG_CALL_ARG_NORMAL -#endif #define TCG_TARGET_CALL_RET_I128 TCG_CALL_RET_NORMAL /* optional instructions */ @@ -XXX,XX +XXX,XX @@ typedef enum { #define TCG_TARGET_HAS_sub2_i32 1 #define TCG_TARGET_HAS_mulu2_i32 0 #define TCG_TARGET_HAS_muls2_i32 0 -#define TCG_TARGET_HAS_muluh_i32 (TCG_TARGET_REG_BITS == 32) -#define TCG_TARGET_HAS_mulsh_i32 (TCG_TARGET_REG_BITS == 32) +#define TCG_TARGET_HAS_muluh_i32 0 +#define TCG_TARGET_HAS_mulsh_i32 0 #define TCG_TARGET_HAS_ext8s_i32 1 #define TCG_TARGET_HAS_ext16s_i32 1 #define TCG_TARGET_HAS_ext8u_i32 1 @@ -XXX,XX +XXX,XX @@ typedef enum { #define TCG_TARGET_HAS_setcond2 1 #define TCG_TARGET_HAS_qemu_st8_i32 0 -#if TCG_TARGET_REG_BITS == 64 #define TCG_TARGET_HAS_movcond_i64 0 #define TCG_TARGET_HAS_div_i64 1 #define TCG_TARGET_HAS_rem_i64 1 @@ -XXX,XX +XXX,XX @@ typedef enum { #define TCG_TARGET_HAS_muls2_i64 0 #define TCG_TARGET_HAS_muluh_i64 1 #define TCG_TARGET_HAS_mulsh_i64 1 -#endif #define TCG_TARGET_DEFAULT_MO (0) diff --git a/tcg/riscv/tcg-target.c.inc b/tcg/riscv/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target.c.inc +++ b/tcg/riscv/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_target_call_oarg_reg(TCGCallReturnKind kind, int slot) #define SOFTMMU_RESERVE_REGS 0 #endif - -static inline tcg_target_long sextreg(tcg_target_long val, int pos, int len) -{ - if (TCG_TARGET_REG_BITS == 32) { - return sextract32(val, pos, len); - } else { - return sextract64(val, pos, len); - } -} +#define sextreg sextract64 /* test if a constant matches the constraint */ static bool tcg_target_const_match(int64_t val, TCGType type, int ct) @@ -XXX,XX +XXX,XX @@ typedef enum { OPC_XOR = 0x4033, OPC_XORI = 0x4013, -#if TCG_TARGET_REG_BITS == 64 OPC_ADDIW = 0x1b, OPC_ADDW = 0x3b, OPC_DIVUW = 0x200503b, @@ -XXX,XX +XXX,XX @@ typedef enum { OPC_SRLIW = 0x501b, OPC_SRLW = 0x503b, OPC_SUBW = 0x4000003b, -#else - /* Simplify code throughout by defining aliases for RV32. */ - OPC_ADDIW = OPC_ADDI, - OPC_ADDW = OPC_ADD, - OPC_DIVUW = OPC_DIVU, - OPC_DIVW = OPC_DIV, - OPC_MULW = OPC_MUL, - OPC_REMUW = OPC_REMU, - OPC_REMW = OPC_REM, - OPC_SLLIW = OPC_SLLI, - OPC_SLLW = OPC_SLL, - OPC_SRAIW = OPC_SRAI, - OPC_SRAW = OPC_SRA, - OPC_SRLIW = OPC_SRLI, - OPC_SRLW = OPC_SRL, - OPC_SUBW = OPC_SUB, -#endif OPC_FENCE = 0x0000000f, OPC_NOP = OPC_ADDI, /* nop = addi r0,r0,0 */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_movi(TCGContext *s, TCGType type, TCGReg rd, tcg_target_long lo, hi, tmp; int shift, ret; - if (TCG_TARGET_REG_BITS == 64 && type == TCG_TYPE_I32) { + if (type == TCG_TYPE_I32) { val = (int32_t)val; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_movi(TCGContext *s, TCGType type, TCGReg rd, } hi = val - lo; - if (TCG_TARGET_REG_BITS == 32 || val == (int32_t)val) { + if (val == (int32_t)val) { tcg_out_opc_upper(s, OPC_LUI, rd, hi); if (lo != 0) { tcg_out_opc_imm(s, OPC_ADDIW, rd, rd, lo); @@ -XXX,XX +XXX,XX @@ static void tcg_out_movi(TCGContext *s, TCGType type, TCGReg rd, return; } - /* We can only be here if TCG_TARGET_REG_BITS != 32 */ tmp = tcg_pcrel_diff(s, (void *)val); if (tmp == (int32_t)tmp) { tcg_out_opc_upper(s, OPC_AUIPC, rd, 0); @@ -XXX,XX +XXX,XX @@ static void tcg_out_ldst(TCGContext *s, RISCVInsn opc, TCGReg data, static void tcg_out_ld(TCGContext *s, TCGType type, TCGReg arg, TCGReg arg1, intptr_t arg2) { - bool is32bit = (TCG_TARGET_REG_BITS == 32 || type == TCG_TYPE_I32); - tcg_out_ldst(s, is32bit ? OPC_LW : OPC_LD, arg, arg1, arg2); + RISCVInsn insn = type == TCG_TYPE_I32 ? OPC_LW : OPC_LD; + tcg_out_ldst(s, insn, arg, arg1, arg2); } static void tcg_out_st(TCGContext *s, TCGType type, TCGReg arg, TCGReg arg1, intptr_t arg2) { - bool is32bit = (TCG_TARGET_REG_BITS == 32 || type == TCG_TYPE_I32); - tcg_out_ldst(s, is32bit ? OPC_SW : OPC_SD, arg, arg1, arg2); + RISCVInsn insn = type == TCG_TYPE_I32 ? OPC_SW : OPC_SD; + tcg_out_ldst(s, insn, arg, arg1, arg2); } static bool tcg_out_sti(TCGContext *s, TCGType type, TCGArg val, @@ -XXX,XX +XXX,XX @@ static void tcg_out_setcond(TCGContext *s, TCGCond cond, TCGReg ret, } } -static void tcg_out_brcond2(TCGContext *s, TCGCond cond, TCGReg al, TCGReg ah, - TCGReg bl, TCGReg bh, TCGLabel *l) -{ - /* todo */ - g_assert_not_reached(); -} - -static void tcg_out_setcond2(TCGContext *s, TCGCond cond, TCGReg ret, - TCGReg al, TCGReg ah, TCGReg bl, TCGReg bh) -{ - /* todo */ - g_assert_not_reached(); -} - static void tcg_out_call_int(TCGContext *s, const tcg_insn_unit *arg, bool tail) { TCGReg link = tail ? TCG_REG_ZERO : TCG_REG_RA; @@ -XXX,XX +XXX,XX @@ static void tcg_out_call_int(TCGContext *s, const tcg_insn_unit *arg, bool tail) if (offset == sextreg(offset, 0, 20)) { /* short jump: -2097150 to 2097152 */ tcg_out_opc_jump(s, OPC_JAL, link, offset); - } else if (TCG_TARGET_REG_BITS == 32 || offset == (int32_t)offset) { + } else if (offset == (int32_t)offset) { /* long jump: -2147483646 to 2147483648 */ tcg_out_opc_upper(s, OPC_AUIPC, TCG_REG_TMP0, 0); tcg_out_opc_imm(s, OPC_JALR, link, TCG_REG_TMP0, 0); ret = reloc_call(s->code_ptr - 2, arg); tcg_debug_assert(ret == true); - } else if (TCG_TARGET_REG_BITS == 64) { + } else { /* far jump: 64-bit */ tcg_target_long imm = sextreg((tcg_target_long)arg, 0, 12); tcg_target_long base = (tcg_target_long)arg - imm; tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP0, base); tcg_out_opc_imm(s, OPC_JALR, link, TCG_REG_TMP0, imm); - } else { - g_assert_not_reached(); } } @@ -XXX,XX +XXX,XX @@ static void * const qemu_st_helpers[MO_SIZE + 1] = { #endif }; -/* We don't support oversize guests */ -QEMU_BUILD_BUG_ON(TCG_TARGET_REG_BITS < TARGET_LONG_BITS); - /* We expect to use a 12-bit negative offset from ENV. */ QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) > 0); QEMU_BUILD_BUG_ON(TLB_MASK_TABLE_OFS(0) < -(1 << 11)); @@ -XXX,XX +XXX,XX @@ static void tcg_out_goto(TCGContext *s, const tcg_insn_unit *target) tcg_debug_assert(ok); } -static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addrl, - TCGReg addrh, MemOpIdx oi, +static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addr, MemOpIdx oi, tcg_insn_unit **label_ptr, bool is_load) { MemOp opc = get_memop(oi); @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addrl, tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP0, mask_base, mask_ofs); tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, table_base, table_ofs); - tcg_out_opc_imm(s, OPC_SRLI, TCG_REG_TMP2, addrl, + tcg_out_opc_imm(s, OPC_SRLI, TCG_REG_TMP2, addr, TARGET_PAGE_BITS - CPU_TLB_ENTRY_BITS); tcg_out_opc_reg(s, OPC_AND, TCG_REG_TMP2, TCG_REG_TMP2, TCG_REG_TMP0); tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP2, TCG_REG_TMP2, TCG_REG_TMP1); @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addrl, /* Clear the non-page, non-alignment bits from the address. */ compare_mask = (tcg_target_long)TARGET_PAGE_MASK | ((1 << a_bits) - 1); if (compare_mask == sextreg(compare_mask, 0, 12)) { - tcg_out_opc_imm(s, OPC_ANDI, TCG_REG_TMP1, addrl, compare_mask); + tcg_out_opc_imm(s, OPC_ANDI, TCG_REG_TMP1, addr, compare_mask); } else { tcg_out_movi(s, TCG_TYPE_TL, TCG_REG_TMP1, compare_mask); - tcg_out_opc_reg(s, OPC_AND, TCG_REG_TMP1, TCG_REG_TMP1, addrl); + tcg_out_opc_reg(s, OPC_AND, TCG_REG_TMP1, TCG_REG_TMP1, addr); } /* Compare masked address with the TLB entry. */ @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_load(TCGContext *s, TCGReg addrl, tcg_out_opc_branch(s, OPC_BNE, TCG_REG_TMP0, TCG_REG_TMP1, 0); /* TLB Hit - translate address using addend. */ - if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { - tcg_out_ext32u(s, TCG_REG_TMP0, addrl); - addrl = TCG_REG_TMP0; + if (TARGET_LONG_BITS == 32) { + tcg_out_ext32u(s, TCG_REG_TMP0, addr); + addr = TCG_REG_TMP0; } - tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP0, TCG_REG_TMP2, addrl); + tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP0, TCG_REG_TMP2, addr); return TCG_REG_TMP0; } static void add_qemu_ldst_label(TCGContext *s, int is_ld, MemOpIdx oi, - TCGType ext, - TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - void *raddr, tcg_insn_unit **label_ptr) + TCGType data_type, TCGReg data_reg, + TCGReg addr_reg, void *raddr, + tcg_insn_unit **label_ptr) { TCGLabelQemuLdst *label = new_ldst_label(s); label->is_ld = is_ld; label->oi = oi; - label->type = ext; - label->datalo_reg = datalo; - label->datahi_reg = datahi; - label->addrlo_reg = addrlo; - label->addrhi_reg = addrhi; + label->type = data_type; + label->datalo_reg = data_reg; + label->addrlo_reg = addr_reg; label->raddr = tcg_splitwx_to_rx(raddr); label->label_ptr[0] = label_ptr[0]; } @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_ld_slow_path(TCGContext *s, TCGLabelQemuLdst *l) TCGReg a2 = tcg_target_call_iarg_regs[2]; TCGReg a3 = tcg_target_call_iarg_regs[3]; - /* We don't support oversize guests */ - if (TCG_TARGET_REG_BITS < TARGET_LONG_BITS) { - g_assert_not_reached(); - } - /* resolve label address */ if (!reloc_sbimm12(l->label_ptr[0], tcg_splitwx_to_rx(s->code_ptr))) { return false; @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) TCGReg a3 = tcg_target_call_iarg_regs[3]; TCGReg a4 = tcg_target_call_iarg_regs[4]; - /* We don't support oversize guests */ - if (TCG_TARGET_REG_BITS < TARGET_LONG_BITS) { - g_assert_not_reached(); - } - /* resolve label address */ if (!reloc_sbimm12(l->label_ptr[0], tcg_splitwx_to_rx(s->code_ptr))) { return false; @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) #endif /* CONFIG_SOFTMMU */ -static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, +static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg val, TCGReg base, MemOp opc, bool is_64) { /* Byte swapping is left to middle-end expansion. */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, switch (opc & (MO_SSIZE)) { case MO_UB: - tcg_out_opc_imm(s, OPC_LBU, lo, base, 0); + tcg_out_opc_imm(s, OPC_LBU, val, base, 0); break; case MO_SB: - tcg_out_opc_imm(s, OPC_LB, lo, base, 0); + tcg_out_opc_imm(s, OPC_LB, val, base, 0); break; case MO_UW: - tcg_out_opc_imm(s, OPC_LHU, lo, base, 0); + tcg_out_opc_imm(s, OPC_LHU, val, base, 0); break; case MO_SW: - tcg_out_opc_imm(s, OPC_LH, lo, base, 0); + tcg_out_opc_imm(s, OPC_LH, val, base, 0); break; case MO_UL: - if (TCG_TARGET_REG_BITS == 64 && is_64) { - tcg_out_opc_imm(s, OPC_LWU, lo, base, 0); + if (is_64) { + tcg_out_opc_imm(s, OPC_LWU, val, base, 0); break; } /* FALLTHRU */ case MO_SL: - tcg_out_opc_imm(s, OPC_LW, lo, base, 0); + tcg_out_opc_imm(s, OPC_LW, val, base, 0); break; case MO_UQ: - /* Prefer to load from offset 0 first, but allow for overlap. */ - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_opc_imm(s, OPC_LD, lo, base, 0); - } else if (lo != base) { - tcg_out_opc_imm(s, OPC_LW, lo, base, 0); - tcg_out_opc_imm(s, OPC_LW, hi, base, 4); - } else { - tcg_out_opc_imm(s, OPC_LW, hi, base, 4); - tcg_out_opc_imm(s, OPC_LW, lo, base, 0); - } + tcg_out_opc_imm(s, OPC_LD, val, base, 0); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg lo, TCGReg hi, static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) { - TCGReg addr_regl, addr_regh __attribute__((unused)); - TCGReg data_regl, data_regh; + TCGReg addr_reg, data_reg; MemOpIdx oi; MemOp opc; #if defined(CONFIG_SOFTMMU) @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) #endif TCGReg base; - data_regl = *args++; - data_regh = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addr_regl = *args++; - addr_regh = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); + data_reg = *args++; + addr_reg = *args++; oi = *args++; opc = get_memop(oi); #if defined(CONFIG_SOFTMMU) - base = tcg_out_tlb_load(s, addr_regl, addr_regh, oi, label_ptr, 1); - tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64); - add_qemu_ldst_label(s, 1, oi, - (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_regl, data_regh, addr_regl, addr_regh, - s->code_ptr, label_ptr); + base = tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 1); + tcg_out_qemu_ld_direct(s, data_reg, base, opc, is_64); + add_qemu_ldst_label(s, 1, oi, (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), + data_reg, addr_reg, s->code_ptr, label_ptr); #else a_bits = get_alignment_bits(opc); if (a_bits) { - tcg_out_test_alignment(s, true, addr_regl, a_bits); + tcg_out_test_alignment(s, true, addr_reg, a_bits); } - base = addr_regl; - if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { + base = addr_reg; + if (TARGET_LONG_BITS == 32) { tcg_out_ext32u(s, TCG_REG_TMP0, base); base = TCG_REG_TMP0; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP0, TCG_GUEST_BASE_REG, base); base = TCG_REG_TMP0; } - tcg_out_qemu_ld_direct(s, data_regl, data_regh, base, opc, is_64); + tcg_out_qemu_ld_direct(s, data_reg, base, opc, is_64); #endif } -static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg lo, TCGReg hi, +static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg val, TCGReg base, MemOp opc) { /* Byte swapping is left to middle-end expansion. */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg lo, TCGReg hi, switch (opc & (MO_SSIZE)) { case MO_8: - tcg_out_opc_store(s, OPC_SB, base, lo, 0); + tcg_out_opc_store(s, OPC_SB, base, val, 0); break; case MO_16: - tcg_out_opc_store(s, OPC_SH, base, lo, 0); + tcg_out_opc_store(s, OPC_SH, base, val, 0); break; case MO_32: - tcg_out_opc_store(s, OPC_SW, base, lo, 0); + tcg_out_opc_store(s, OPC_SW, base, val, 0); break; case MO_64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_opc_store(s, OPC_SD, base, lo, 0); - } else { - tcg_out_opc_store(s, OPC_SW, base, lo, 0); - tcg_out_opc_store(s, OPC_SW, base, hi, 4); - } + tcg_out_opc_store(s, OPC_SD, base, val, 0); break; default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg lo, TCGReg hi, static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) { - TCGReg addr_regl, addr_regh __attribute__((unused)); - TCGReg data_regl, data_regh; + TCGReg addr_reg, data_reg; MemOpIdx oi; MemOp opc; #if defined(CONFIG_SOFTMMU) @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) #endif TCGReg base; - data_regl = *args++; - data_regh = (TCG_TARGET_REG_BITS == 32 && is_64 ? *args++ : 0); - addr_regl = *args++; - addr_regh = (TCG_TARGET_REG_BITS < TARGET_LONG_BITS ? *args++ : 0); + data_reg = *args++; + addr_reg = *args++; oi = *args++; opc = get_memop(oi); #if defined(CONFIG_SOFTMMU) - base = tcg_out_tlb_load(s, addr_regl, addr_regh, oi, label_ptr, 0); - tcg_out_qemu_st_direct(s, data_regl, data_regh, base, opc); - add_qemu_ldst_label(s, 0, oi, - (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_regl, data_regh, addr_regl, addr_regh, - s->code_ptr, label_ptr); + base = tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 0); + tcg_out_qemu_st_direct(s, data_reg, base, opc); + add_qemu_ldst_label(s, 0, oi, (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), + data_reg, addr_reg, s->code_ptr, label_ptr); #else a_bits = get_alignment_bits(opc); if (a_bits) { - tcg_out_test_alignment(s, false, addr_regl, a_bits); + tcg_out_test_alignment(s, false, addr_reg, a_bits); } - base = addr_regl; - if (TCG_TARGET_REG_BITS > TARGET_LONG_BITS) { + base = addr_reg; + if (TARGET_LONG_BITS == 32) { tcg_out_ext32u(s, TCG_REG_TMP0, base); base = TCG_REG_TMP0; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP0, TCG_GUEST_BASE_REG, base); base = TCG_REG_TMP0; } - tcg_out_qemu_st_direct(s, data_regl, data_regh, base, opc); + tcg_out_qemu_st_direct(s, data_reg, base, opc); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, case INDEX_op_brcond_i64: tcg_out_brcond(s, a2, a0, a1, arg_label(args[3])); break; - case INDEX_op_brcond2_i32: - tcg_out_brcond2(s, args[4], a0, a1, a2, args[3], arg_label(args[5])); - break; case INDEX_op_setcond_i32: case INDEX_op_setcond_i64: tcg_out_setcond(s, args[3], a0, a1, a2); break; - case INDEX_op_setcond2_i32: - tcg_out_setcond2(s, args[5], a0, a1, a2, args[3], args[4]); - break; case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, args, false); @@ -XXX,XX +XXX,XX @@ static TCGConstraintSetIndex tcg_target_op_def(TCGOpcode op) case INDEX_op_sub2_i64: return C_O2_I4(r, r, rZ, rZ, rM, rM); - case INDEX_op_brcond2_i32: - return C_O0_I4(rZ, rZ, rZ, rZ); - - case INDEX_op_setcond2_i32: - return C_O1_I4(r, rZ, rZ, rZ, rZ); - case INDEX_op_qemu_ld_i32: - return (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS - ? C_O1_I1(r, L) : C_O1_I2(r, L, L)); - case INDEX_op_qemu_st_i32: - return (TARGET_LONG_BITS <= TCG_TARGET_REG_BITS - ? C_O0_I2(LZ, L) : C_O0_I3(LZ, L, L)); case INDEX_op_qemu_ld_i64: - return (TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, L) - : TARGET_LONG_BITS <= TCG_TARGET_REG_BITS ? C_O2_I1(r, r, L) - : C_O2_I2(r, r, L, L)); + return C_O1_I1(r, L); + case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - return (TCG_TARGET_REG_BITS == 64 ? C_O0_I2(LZ, L) - : TARGET_LONG_BITS <= TCG_TARGET_REG_BITS ? C_O0_I3(LZ, LZ, L) - : C_O0_I4(LZ, LZ, L, L)); + return C_O0_I2(LZ, L); default: g_assert_not_reached(); @@ -XXX,XX +XXX,XX @@ static void tcg_target_qemu_prologue(TCGContext *s) static void tcg_target_init(TCGContext *s) { tcg_target_available_regs[TCG_TYPE_I32] = 0xffffffff; - if (TCG_TARGET_REG_BITS == 64) { - tcg_target_available_regs[TCG_TYPE_I64] = 0xffffffff; - } + tcg_target_available_regs[TCG_TYPE_I64] = 0xffffffff; tcg_target_call_clobber_regs = -1u; tcg_regset_reset_reg(tcg_target_call_clobber_regs, TCG_REG_S0); -- 2.34.1 Interpret the variable argument placement in the caller. Pass data_type instead of is64 -- there are several places where we already convert back from bool to type. Clean things up by using type throughout. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/riscv/tcg-target.c.inc | 66 ++++++++++++++------------------------ 1 file changed, 24 insertions(+), 42 deletions(-) diff --git a/tcg/riscv/tcg-target.c.inc b/tcg/riscv/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target.c.inc +++ b/tcg/riscv/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) #endif /* CONFIG_SOFTMMU */ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg val, - TCGReg base, MemOp opc, bool is_64) + TCGReg base, MemOp opc, TCGType type) { /* Byte swapping is left to middle-end expansion. */ tcg_debug_assert((opc & MO_BSWAP) == 0); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg val, tcg_out_opc_imm(s, OPC_LH, val, base, 0); break; case MO_UL: - if (is_64) { + if (type == TCG_TYPE_I64) { tcg_out_opc_imm(s, OPC_LWU, val, base, 0); break; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg val, } } -static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) +static void tcg_out_qemu_ld(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, + MemOpIdx oi, TCGType data_type) { - TCGReg addr_reg, data_reg; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) - tcg_insn_unit *label_ptr[1]; -#else - unsigned a_bits; -#endif + MemOp opc = get_memop(oi); TCGReg base; - data_reg = *args++; - addr_reg = *args++; - oi = *args++; - opc = get_memop(oi); - #if defined(CONFIG_SOFTMMU) + tcg_insn_unit *label_ptr[1]; + base = tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 1); - tcg_out_qemu_ld_direct(s, data_reg, base, opc, is_64); - add_qemu_ldst_label(s, 1, oi, (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_reg, addr_reg, s->code_ptr, label_ptr); + tcg_out_qemu_ld_direct(s, data_reg, base, opc, data_type); + add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addr_reg, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, bool is_64) tcg_out_opc_reg(s, OPC_ADD, TCG_REG_TMP0, TCG_GUEST_BASE_REG, base); base = TCG_REG_TMP0; } - tcg_out_qemu_ld_direct(s, data_reg, base, opc, is_64); + tcg_out_qemu_ld_direct(s, data_reg, base, opc, data_type); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, TCGReg val, } } -static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, bool is_64) +static void tcg_out_qemu_st(TCGContext *s, TCGReg data_reg, TCGReg addr_reg, + MemOpIdx oi, TCGType data_type) { - TCGReg addr_reg, data_reg; - MemOpIdx oi; - MemOp opc; -#if defined(CONFIG_SOFTMMU) - tcg_insn_unit *label_ptr[1]; -#else - unsigned a_bits; -#endif + MemOp opc = get_memop(oi); TCGReg base; - data_reg = *args++; - addr_reg = *args++; - oi = *args++; - opc = get_memop(oi); - #if defined(CONFIG_SOFTMMU) + tcg_insn_unit *label_ptr[1]; + base = tcg_out_tlb_load(s, addr_reg, oi, label_ptr, 0); tcg_out_qemu_st_direct(s, data_reg, base, opc); - add_qemu_ldst_label(s, 0, oi, (is_64 ? TCG_TYPE_I64 : TCG_TYPE_I32), - data_reg, addr_reg, s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #else - a_bits = get_alignment_bits(opc); + unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addr_reg, a_bits); } @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args, false); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args, true); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args, false); + tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args, true); + tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I64); break; case INDEX_op_extrh_i64_i32: -- 2.34.1 We need to set this in TCGLabelQemuLdst, so plumb this all the way through from tcg_out_op. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/s390x/tcg-target.c.inc | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/tcg/s390x/tcg-target.c.inc b/tcg/s390x/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/s390x/tcg-target.c.inc +++ b/tcg/s390x/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static TCGReg tcg_out_tlb_read(TCGContext *s, TCGReg addr_reg, MemOp opc, } static void add_qemu_ldst_label(TCGContext *s, bool is_ld, MemOpIdx oi, - TCGReg data, TCGReg addr, + TCGType type, TCGReg data, TCGReg addr, tcg_insn_unit *raddr, tcg_insn_unit *label_ptr) { TCGLabelQemuLdst *label = new_ldst_label(s); label->is_ld = is_ld; label->oi = oi; + label->type = type; label->datalo_reg = data; label->addrlo_reg = addr; label->raddr = tcg_splitwx_to_rx(raddr); @@ -XXX,XX +XXX,XX @@ static void tcg_prepare_user_ldst(TCGContext *s, TCGReg *addr_reg, #endif /* CONFIG_SOFTMMU */ static void tcg_out_qemu_ld(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, - MemOpIdx oi) + MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); #ifdef CONFIG_SOFTMMU @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, tcg_out_qemu_ld_direct(s, opc, data_reg, base_reg, TCG_REG_R2, 0); - add_qemu_ldst_label(s, 1, oi, data_reg, addr_reg, s->code_ptr, label_ptr); + add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #else TCGReg index_reg; tcg_target_long disp; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, } static void tcg_out_qemu_st(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, - MemOpIdx oi) + MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); #ifdef CONFIG_SOFTMMU @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, tcg_out_qemu_st_direct(s, opc, data_reg, base_reg, TCG_REG_R2, 0); - add_qemu_ldst_label(s, 0, oi, data_reg, addr_reg, s->code_ptr, label_ptr); + add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, + s->code_ptr, label_ptr); #else TCGReg index_reg; tcg_target_long disp; @@ -XXX,XX +XXX,XX @@ static inline void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - /* ??? Technically we can use a non-extending instruction. */ + tcg_out_qemu_ld(s, args[0], args[1], args[2], TCG_TYPE_I32); + break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, args[0], args[1], args[2]); + tcg_out_qemu_ld(s, args[0], args[1], args[2], TCG_TYPE_I64); break; case INDEX_op_qemu_st_i32: + tcg_out_qemu_st(s, args[0], args[1], args[2], TCG_TYPE_I32); + break; case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, args[0], args[1], args[2]); + tcg_out_qemu_st(s, args[0], args[1], args[2], TCG_TYPE_I64); break; case INDEX_op_ld16s_i64: -- 2.34.1 Collect the 3 potential parts of the host address into a struct. Reorg tcg_out_qemu_{ld,st}_direct to use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/s390x/tcg-target.c.inc | 109 ++++++++++++++++++++----------------- 1 file changed, 60 insertions(+), 49 deletions(-) diff --git a/tcg/s390x/tcg-target.c.inc b/tcg/s390x/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/s390x/tcg-target.c.inc +++ b/tcg/s390x/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_call(TCGContext *s, const tcg_insn_unit *dest, tcg_out_call_int(s, dest); } +typedef struct { + TCGReg base; + TCGReg index; + int disp; +} HostAddress; + static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg data, - TCGReg base, TCGReg index, int disp) + HostAddress h) { switch (opc & (MO_SSIZE | MO_BSWAP)) { case MO_UB: - tcg_out_insn(s, RXY, LLGC, data, base, index, disp); + tcg_out_insn(s, RXY, LLGC, data, h.base, h.index, h.disp); break; case MO_SB: - tcg_out_insn(s, RXY, LGB, data, base, index, disp); + tcg_out_insn(s, RXY, LGB, data, h.base, h.index, h.disp); break; case MO_UW | MO_BSWAP: /* swapped unsigned halfword load with upper bits zeroed */ - tcg_out_insn(s, RXY, LRVH, data, base, index, disp); + tcg_out_insn(s, RXY, LRVH, data, h.base, h.index, h.disp); tcg_out_ext16u(s, data, data); break; case MO_UW: - tcg_out_insn(s, RXY, LLGH, data, base, index, disp); + tcg_out_insn(s, RXY, LLGH, data, h.base, h.index, h.disp); break; case MO_SW | MO_BSWAP: /* swapped sign-extended halfword load */ - tcg_out_insn(s, RXY, LRVH, data, base, index, disp); + tcg_out_insn(s, RXY, LRVH, data, h.base, h.index, h.disp); tcg_out_ext16s(s, TCG_TYPE_REG, data, data); break; case MO_SW: - tcg_out_insn(s, RXY, LGH, data, base, index, disp); + tcg_out_insn(s, RXY, LGH, data, h.base, h.index, h.disp); break; case MO_UL | MO_BSWAP: /* swapped unsigned int load with upper bits zeroed */ - tcg_out_insn(s, RXY, LRV, data, base, index, disp); + tcg_out_insn(s, RXY, LRV, data, h.base, h.index, h.disp); tcg_out_ext32u(s, data, data); break; case MO_UL: - tcg_out_insn(s, RXY, LLGF, data, base, index, disp); + tcg_out_insn(s, RXY, LLGF, data, h.base, h.index, h.disp); break; case MO_SL | MO_BSWAP: /* swapped sign-extended int load */ - tcg_out_insn(s, RXY, LRV, data, base, index, disp); + tcg_out_insn(s, RXY, LRV, data, h.base, h.index, h.disp); tcg_out_ext32s(s, data, data); break; case MO_SL: - tcg_out_insn(s, RXY, LGF, data, base, index, disp); + tcg_out_insn(s, RXY, LGF, data, h.base, h.index, h.disp); break; case MO_UQ | MO_BSWAP: - tcg_out_insn(s, RXY, LRVG, data, base, index, disp); + tcg_out_insn(s, RXY, LRVG, data, h.base, h.index, h.disp); break; case MO_UQ: - tcg_out_insn(s, RXY, LG, data, base, index, disp); + tcg_out_insn(s, RXY, LG, data, h.base, h.index, h.disp); break; default: @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg data, } static void tcg_out_qemu_st_direct(TCGContext *s, MemOp opc, TCGReg data, - TCGReg base, TCGReg index, int disp) + HostAddress h) { switch (opc & (MO_SIZE | MO_BSWAP)) { case MO_UB: - if (disp >= 0 && disp < 0x1000) { - tcg_out_insn(s, RX, STC, data, base, index, disp); + if (h.disp >= 0 && h.disp < 0x1000) { + tcg_out_insn(s, RX, STC, data, h.base, h.index, h.disp); } else { - tcg_out_insn(s, RXY, STCY, data, base, index, disp); + tcg_out_insn(s, RXY, STCY, data, h.base, h.index, h.disp); } break; case MO_UW | MO_BSWAP: - tcg_out_insn(s, RXY, STRVH, data, base, index, disp); + tcg_out_insn(s, RXY, STRVH, data, h.base, h.index, h.disp); break; case MO_UW: - if (disp >= 0 && disp < 0x1000) { - tcg_out_insn(s, RX, STH, data, base, index, disp); + if (h.disp >= 0 && h.disp < 0x1000) { + tcg_out_insn(s, RX, STH, data, h.base, h.index, h.disp); } else { - tcg_out_insn(s, RXY, STHY, data, base, index, disp); + tcg_out_insn(s, RXY, STHY, data, h.base, h.index, h.disp); } break; case MO_UL | MO_BSWAP: - tcg_out_insn(s, RXY, STRV, data, base, index, disp); + tcg_out_insn(s, RXY, STRV, data, h.base, h.index, h.disp); break; case MO_UL: - if (disp >= 0 && disp < 0x1000) { - tcg_out_insn(s, RX, ST, data, base, index, disp); + if (h.disp >= 0 && h.disp < 0x1000) { + tcg_out_insn(s, RX, ST, data, h.base, h.index, h.disp); } else { - tcg_out_insn(s, RXY, STY, data, base, index, disp); + tcg_out_insn(s, RXY, STY, data, h.base, h.index, h.disp); } break; case MO_UQ | MO_BSWAP: - tcg_out_insn(s, RXY, STRVG, data, base, index, disp); + tcg_out_insn(s, RXY, STRVG, data, h.base, h.index, h.disp); break; case MO_UQ: - tcg_out_insn(s, RXY, STG, data, base, index, disp); + tcg_out_insn(s, RXY, STG, data, h.base, h.index, h.disp); break; default: @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *l) return tcg_out_fail_alignment(s, l); } -static void tcg_prepare_user_ldst(TCGContext *s, TCGReg *addr_reg, - TCGReg *index_reg, tcg_target_long *disp) +static HostAddress tcg_prepare_user_ldst(TCGContext *s, TCGReg addr_reg) { + TCGReg index; + int disp; + if (TARGET_LONG_BITS == 32) { - tcg_out_ext32u(s, TCG_TMP0, *addr_reg); - *addr_reg = TCG_TMP0; + tcg_out_ext32u(s, TCG_TMP0, addr_reg); + addr_reg = TCG_TMP0; } if (guest_base < 0x80000) { - *index_reg = TCG_REG_NONE; - *disp = guest_base; + index = TCG_REG_NONE; + disp = guest_base; } else { - *index_reg = TCG_GUEST_BASE_REG; - *disp = 0; + index = TCG_GUEST_BASE_REG; + disp = 0; } + return (HostAddress){ .base = addr_reg, .index = index, .disp = disp }; } #endif /* CONFIG_SOFTMMU */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; + #ifdef CONFIG_SOFTMMU unsigned mem_index = get_mmuidx(oi); tcg_insn_unit *label_ptr; - TCGReg base_reg; - base_reg = tcg_out_tlb_read(s, addr_reg, opc, mem_index, 1); + h.base = tcg_out_tlb_read(s, addr_reg, opc, mem_index, 1); + h.index = TCG_REG_R2; + h.disp = 0; tcg_out16(s, RI_BRC | (S390_CC_NE << 4)); label_ptr = s->code_ptr; s->code_ptr += 1; - tcg_out_qemu_ld_direct(s, opc, data_reg, base_reg, TCG_REG_R2, 0); + tcg_out_qemu_ld_direct(s, opc, data_reg, h); add_qemu_ldst_label(s, true, oi, data_type, data_reg, addr_reg, s->code_ptr, label_ptr); #else - TCGReg index_reg; - tcg_target_long disp; unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, true, addr_reg, a_bits); } - tcg_prepare_user_ldst(s, &addr_reg, &index_reg, &disp); - tcg_out_qemu_ld_direct(s, opc, data_reg, addr_reg, index_reg, disp); + h = tcg_prepare_user_ldst(s, addr_reg); + tcg_out_qemu_ld_direct(s, opc, data_reg, h); #endif } @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st(TCGContext* s, TCGReg data_reg, TCGReg addr_reg, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); + HostAddress h; + #ifdef CONFIG_SOFTMMU unsigned mem_index = get_mmuidx(oi); tcg_insn_unit *label_ptr; - TCGReg base_reg; - base_reg = tcg_out_tlb_read(s, addr_reg, opc, mem_index, 0); + h.base = tcg_out_tlb_read(s, addr_reg, opc, mem_index, 0); + h.index = TCG_REG_R2; + h.disp = 0; tcg_out16(s, RI_BRC | (S390_CC_NE << 4)); label_ptr = s->code_ptr; s->code_ptr += 1; - tcg_out_qemu_st_direct(s, opc, data_reg, base_reg, TCG_REG_R2, 0); + tcg_out_qemu_st_direct(s, opc, data_reg, h); add_qemu_ldst_label(s, false, oi, data_type, data_reg, addr_reg, s->code_ptr, label_ptr); #else - TCGReg index_reg; - tcg_target_long disp; unsigned a_bits = get_alignment_bits(opc); if (a_bits) { tcg_out_test_alignment(s, false, addr_reg, a_bits); } - tcg_prepare_user_ldst(s, &addr_reg, &index_reg, &disp); - tcg_out_qemu_st_direct(s, opc, data_reg, addr_reg, index_reg, disp); + h = tcg_prepare_user_ldst(s, addr_reg); + tcg_out_qemu_st_direct(s, opc, data_reg, h); #endif } -- 2.34.1 In tcg_canonicalize_memop, we remove MO_SIGN from MO_32 operations with TCG_TYPE_I32. Thus this is never set. We already have an identical test just above which does not include is_64 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/sparc64/tcg-target.c.inc | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tcg/sparc64/tcg-target.c.inc b/tcg/sparc64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/sparc64/tcg-target.c.inc +++ b/tcg/sparc64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr, tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_O2, oi); /* We let the helper sign-extend SB and SW, but leave SL for here. */ - if (is_64 && (memop & MO_SSIZE) == MO_SL) { + if ((memop & MO_SSIZE) == MO_SL) { tcg_out_ext32s(s, data, TCG_REG_O0); } else { tcg_out_mov(s, TCG_TYPE_REG, data, TCG_REG_O0); -- 2.34.1 We need to set this in TCGLabelQemuLdst, so plumb this all the way through from tcg_out_op. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/sparc64/tcg-target.c.inc | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tcg/sparc64/tcg-target.c.inc b/tcg/sparc64/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/sparc64/tcg-target.c.inc +++ b/tcg/sparc64/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ static const int qemu_st_opc[(MO_SIZE | MO_BSWAP) + 1] = { }; static void tcg_out_qemu_ld(TCGContext *s, TCGReg data, TCGReg addr, - MemOpIdx oi, bool is_64) + MemOpIdx oi, TCGType data_type) { MemOp memop = get_memop(oi); tcg_insn_unit *label_ptr; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, a0, a1, a2, false); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: - tcg_out_qemu_ld(s, a0, a1, a2, true); + tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/tcg.c | 13 +++++++++++++ tcg/tcg-ldst.c.inc | 14 -------------- 2 files changed, 13 insertions(+), 14 deletions(-) diff --git a/tcg/tcg.c b/tcg/tcg.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg.c +++ b/tcg/tcg.c @@ -XXX,XX +XXX,XX @@ typedef struct QEMU_PACKED { DebugFrameFDEHeader fde; } DebugFrameHeader; +typedef struct TCGLabelQemuLdst { + bool is_ld; /* qemu_ld: true, qemu_st: false */ + MemOpIdx oi; + TCGType type; /* result type of a load */ + TCGReg addrlo_reg; /* reg index for low word of guest virtual addr */ + TCGReg addrhi_reg; /* reg index for high word of guest virtual addr */ + TCGReg datalo_reg; /* reg index for low word to be loaded or stored */ + TCGReg datahi_reg; /* reg index for high word to be loaded or stored */ + const tcg_insn_unit *raddr; /* addr of the next IR of qemu_ld/st IR */ + tcg_insn_unit *label_ptr[2]; /* label pointers to be updated */ + QSIMPLEQ_ENTRY(TCGLabelQemuLdst) next; +} TCGLabelQemuLdst; + static void tcg_register_jit_int(const void *buf, size_t size, const void *debug_frame, size_t debug_frame_size) diff --git a/tcg/tcg-ldst.c.inc b/tcg/tcg-ldst.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg-ldst.c.inc +++ b/tcg/tcg-ldst.c.inc @@ -XXX,XX +XXX,XX @@ * THE SOFTWARE. */ -typedef struct TCGLabelQemuLdst { - bool is_ld; /* qemu_ld: true, qemu_st: false */ - MemOpIdx oi; - TCGType type; /* result type of a load */ - TCGReg addrlo_reg; /* reg index for low word of guest virtual addr */ - TCGReg addrhi_reg; /* reg index for high word of guest virtual addr */ - TCGReg datalo_reg; /* reg index for low word to be loaded or stored */ - TCGReg datahi_reg; /* reg index for high word to be loaded or stored */ - const tcg_insn_unit *raddr; /* addr of the next IR of qemu_ld/st IR */ - tcg_insn_unit *label_ptr[2]; /* label pointers to be updated */ - QSIMPLEQ_ENTRY(TCGLabelQemuLdst) next; -} TCGLabelQemuLdst; - - /* * Generate TB finalization at the end of block */ -- 2.34.1 An inline function is safer than a macro, and REG_P was rather too generic. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/tcg-internal.h | 4 ---- tcg/tcg.c | 16 +++++++++++++--- 2 files changed, 13 insertions(+), 7 deletions(-) diff --git a/tcg/tcg-internal.h b/tcg/tcg-internal.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg-internal.h +++ b/tcg/tcg-internal.h @@ -XXX,XX +XXX,XX @@ typedef struct TCGCallArgumentLoc { unsigned tmp_subindex : 2; } TCGCallArgumentLoc; -/* Avoid "unsigned < 0 is always false" Werror, when iarg_regs is empty. */ -#define REG_P(L) \ - ((int)(L)->arg_slot < (int)ARRAY_SIZE(tcg_target_call_iarg_regs)) - typedef struct TCGHelperInfo { void *func; const char *name; diff --git a/tcg/tcg.c b/tcg/tcg.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg.c +++ b/tcg/tcg.c @@ -XXX,XX +XXX,XX @@ static void init_ffi_layouts(void) } #endif /* CONFIG_TCG_INTERPRETER */ +static inline bool arg_slot_reg_p(unsigned arg_slot) +{ + /* + * Split the sizeof away from the comparison to avoid Werror from + * "unsigned < 0 is always false", when iarg_regs is empty. + */ + unsigned nreg = ARRAY_SIZE(tcg_target_call_iarg_regs); + return arg_slot < nreg; +} + typedef struct TCGCumulativeArgs { int arg_idx; /* tcg_gen_callN args[] */ int info_in_idx; /* TCGHelperInfo in[] */ @@ -XXX,XX +XXX,XX @@ liveness_pass_1(TCGContext *s) case TCG_CALL_ARG_NORMAL: case TCG_CALL_ARG_EXTEND_U: case TCG_CALL_ARG_EXTEND_S: - if (REG_P(loc)) { + if (arg_slot_reg_p(loc->arg_slot)) { *la_temp_pref(ts) = 0; break; } @@ -XXX,XX +XXX,XX @@ liveness_pass_1(TCGContext *s) case TCG_CALL_ARG_NORMAL: case TCG_CALL_ARG_EXTEND_U: case TCG_CALL_ARG_EXTEND_S: - if (REG_P(loc)) { + if (arg_slot_reg_p(loc->arg_slot)) { tcg_regset_set_reg(*la_temp_pref(ts), tcg_target_call_iarg_regs[loc->arg_slot]); } @@ -XXX,XX +XXX,XX @@ static void load_arg_stk(TCGContext *s, int stk_slot, TCGTemp *ts, static void load_arg_normal(TCGContext *s, const TCGCallArgumentLoc *l, TCGTemp *ts, TCGRegSet *allocated_regs) { - if (REG_P(l)) { + if (arg_slot_reg_p(l->arg_slot)) { TCGReg reg = tcg_target_call_iarg_regs[l->arg_slot]; load_arg_reg(s, reg, ts, *allocated_regs); tcg_regset_set_reg(*allocated_regs, reg); -- 2.34.1 Unify all computation of argument stack offset in one function. This requires that we adjust ref_slot to be in the same units, by adding max_reg_slots during init_call_layout. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/tcg.c | 29 +++++++++++++++++------------ 1 file changed, 17 insertions(+), 12 deletions(-) diff --git a/tcg/tcg.c b/tcg/tcg.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg.c +++ b/tcg/tcg.c @@ -XXX,XX +XXX,XX @@ static inline bool arg_slot_reg_p(unsigned arg_slot) return arg_slot < nreg; } +static inline int arg_slot_stk_ofs(unsigned arg_slot) +{ + unsigned max = TCG_STATIC_CALL_ARGS_SIZE / sizeof(tcg_target_long); + unsigned stk_slot = arg_slot - ARRAY_SIZE(tcg_target_call_iarg_regs); + + tcg_debug_assert(stk_slot < max); + return TCG_TARGET_CALL_STACK_OFFSET + stk_slot * sizeof(tcg_target_long); +} + typedef struct TCGCumulativeArgs { int arg_idx; /* tcg_gen_callN args[] */ int info_in_idx; /* TCGHelperInfo in[] */ @@ -XXX,XX +XXX,XX @@ static void init_call_layout(TCGHelperInfo *info) } } assert(ref_base + cum.ref_slot <= max_stk_slots); + ref_base += max_reg_slots; if (ref_base != 0) { for (int i = cum.info_in_idx - 1; i >= 0; --i) { @@ -XXX,XX +XXX,XX @@ static void load_arg_reg(TCGContext *s, TCGReg reg, TCGTemp *ts, } } -static void load_arg_stk(TCGContext *s, int stk_slot, TCGTemp *ts, +static void load_arg_stk(TCGContext *s, unsigned arg_slot, TCGTemp *ts, TCGRegSet allocated_regs) { /* @@ -XXX,XX +XXX,XX @@ static void load_arg_stk(TCGContext *s, int stk_slot, TCGTemp *ts, */ temp_load(s, ts, tcg_target_available_regs[ts->type], allocated_regs, 0); tcg_out_st(s, ts->type, ts->reg, TCG_REG_CALL_STACK, - TCG_TARGET_CALL_STACK_OFFSET + - stk_slot * sizeof(tcg_target_long)); + arg_slot_stk_ofs(arg_slot)); } static void load_arg_normal(TCGContext *s, const TCGCallArgumentLoc *l, @@ -XXX,XX +XXX,XX @@ static void load_arg_normal(TCGContext *s, const TCGCallArgumentLoc *l, load_arg_reg(s, reg, ts, *allocated_regs); tcg_regset_set_reg(*allocated_regs, reg); } else { - load_arg_stk(s, l->arg_slot - ARRAY_SIZE(tcg_target_call_iarg_regs), - ts, *allocated_regs); + load_arg_stk(s, l->arg_slot, ts, *allocated_regs); } } -static void load_arg_ref(TCGContext *s, int arg_slot, TCGReg ref_base, +static void load_arg_ref(TCGContext *s, unsigned arg_slot, TCGReg ref_base, intptr_t ref_off, TCGRegSet *allocated_regs) { TCGReg reg; - int stk_slot = arg_slot - ARRAY_SIZE(tcg_target_call_iarg_regs); - if (stk_slot < 0) { + if (arg_slot_reg_p(arg_slot)) { reg = tcg_target_call_iarg_regs[arg_slot]; tcg_reg_free(s, reg, *allocated_regs); tcg_out_addi_ptr(s, reg, ref_base, ref_off); @@ -XXX,XX +XXX,XX @@ static void load_arg_ref(TCGContext *s, int arg_slot, TCGReg ref_base, *allocated_regs, 0, false); tcg_out_addi_ptr(s, reg, ref_base, ref_off); tcg_out_st(s, TCG_TYPE_PTR, reg, TCG_REG_CALL_STACK, - TCG_TARGET_CALL_STACK_OFFSET - + stk_slot * sizeof(tcg_target_long)); + arg_slot_stk_ofs(arg_slot)); } } @@ -XXX,XX +XXX,XX @@ static void tcg_reg_alloc_call(TCGContext *s, TCGOp *op) case TCG_CALL_ARG_BY_REF: load_arg_stk(s, loc->ref_slot, ts, allocated_regs); load_arg_ref(s, loc->arg_slot, TCG_REG_CALL_STACK, - TCG_TARGET_CALL_STACK_OFFSET - + loc->ref_slot * sizeof(tcg_target_long), + arg_slot_stk_ofs(loc->ref_slot), &allocated_regs); break; case TCG_CALL_ARG_BY_REF_N: -- 2.34.1 While the old type was correct in the ideal sense, some ABIs require the argument to be zero-extended. Using uint32_t for all such values is a decent compromise. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/tcg/tcg-ldst.h | 10 +++++++--- accel/tcg/cputlb.c | 6 +++--- 2 files changed, 10 insertions(+), 6 deletions(-) diff --git a/include/tcg/tcg-ldst.h b/include/tcg/tcg-ldst.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg-ldst.h +++ b/include/tcg/tcg-ldst.h @@ -XXX,XX +XXX,XX @@ tcg_target_ulong helper_be_ldsw_mmu(CPUArchState *env, target_ulong addr, tcg_target_ulong helper_be_ldsl_mmu(CPUArchState *env, target_ulong addr, MemOpIdx oi, uintptr_t retaddr); -void helper_ret_stb_mmu(CPUArchState *env, target_ulong addr, uint8_t val, +/* + * Value extended to at least uint32_t, so that some ABIs do not require + * zero-extension from uint8_t or uint16_t. + */ +void helper_ret_stb_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr); -void helper_le_stw_mmu(CPUArchState *env, target_ulong addr, uint16_t val, +void helper_le_stw_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr); void helper_le_stl_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr); void helper_le_stq_mmu(CPUArchState *env, target_ulong addr, uint64_t val, MemOpIdx oi, uintptr_t retaddr); -void helper_be_stw_mmu(CPUArchState *env, target_ulong addr, uint16_t val, +void helper_be_stw_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr); void helper_be_stl_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr); diff --git a/accel/tcg/cputlb.c b/accel/tcg/cputlb.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/cputlb.c +++ b/accel/tcg/cputlb.c @@ -XXX,XX +XXX,XX @@ full_stb_mmu(CPUArchState *env, target_ulong addr, uint64_t val, store_helper(env, addr, val, oi, retaddr, MO_UB); } -void helper_ret_stb_mmu(CPUArchState *env, target_ulong addr, uint8_t val, +void helper_ret_stb_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr) { full_stb_mmu(env, addr, val, oi, retaddr); @@ -XXX,XX +XXX,XX @@ static void full_le_stw_mmu(CPUArchState *env, target_ulong addr, uint64_t val, store_helper(env, addr, val, oi, retaddr, MO_LEUW); } -void helper_le_stw_mmu(CPUArchState *env, target_ulong addr, uint16_t val, +void helper_le_stw_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr) { full_le_stw_mmu(env, addr, val, oi, retaddr); @@ -XXX,XX +XXX,XX @@ static void full_be_stw_mmu(CPUArchState *env, target_ulong addr, uint64_t val, store_helper(env, addr, val, oi, retaddr, MO_BEUW); } -void helper_be_stw_mmu(CPUArchState *env, target_ulong addr, uint16_t val, +void helper_be_stw_mmu(CPUArchState *env, target_ulong addr, uint32_t val, MemOpIdx oi, uintptr_t retaddr) { full_be_stw_mmu(env, addr, val, oi, retaddr); -- 2.34.1