The following changes since commit 3d48b6b687c558a042d91370633b91c6e29e0e05: Merge tag 'pull-request-2024-05-14' of https://gitlab.com/thuth/qemu into staging (2024-05-14 17:24:04 +0200) are available in the Git repository at: https://gitlab.com/rth7680/qemu.git tags/pull-tcg-20240515 for you to fetch changes up to c9290dfebfdba5c13baa5e1f10e13a1c876b0643: tcg/loongarch64: Fill out tcg_out_{ld,st} for vector regs (2024-05-15 08:57:39 +0200) ---------------------------------------------------------------- tcg/loongarch64: Fill out tcg_out_{ld,st} for vector regs accel/tcg: Improve disassembly for target and plugin ---------------------------------------------------------------- Philippe Mathieu-Daudé (1): accel/tcg: Remove cpu_ldsb_code / cpu_ldsw_code Richard Henderson (33): accel/tcg: Use vaddr in translator_ld* accel/tcg: Hide in_same_page outside of a target-specific context accel/tcg: Pass DisasContextBase to translator_fake_ldb accel/tcg: Reorg translator_ld* accel/tcg: Cap the translation block when we encounter mmio accel/tcg: Record mmio bytes during translation accel/tcg: Record when translator_fake_ldb is used accel/tcg: Record DisasContextBase in tcg_ctx for plugins plugins: Copy memory in qemu_plugin_insn_data accel/tcg: Implement translator_st plugins: Use translator_st for qemu_plugin_insn_data plugins: Read mem_only directly from TB cflags plugins: Use DisasContextBase for qemu_plugin_insn_haddr plugins: Use DisasContextBase for qemu_plugin_tb_vaddr plugins: Merge alloc_tcg_plugin_context into plugin_gen_tb_start accel/tcg: Provide default implementation of disas_log accel/tcg: Return bool from TranslatorOps.disas_log disas: Split disas.c disas: Use translator_st to get disassembly data accel/tcg: Introduce translator_fake_ld target/s390x: Fix translator_fake_ld length target/s390x: Disassemble EXECUTEd instructions target/hexagon: Use translator_ldl in pkt_crosses_page target/microblaze: Use translator_ldl target/i386: Use translator_ldub for everything target/avr: Use translator_lduw target/cris: Use translator_ld* in cris_fetch target/cris: Use cris_fetch in translate_v10.c.inc target/riscv: Use translator_ld* for everything target/rx: Use translator_ld* target/xtensa: Use translator_ldub in xtensa_insn_len target/s390x: Use translator_lduw in get_next_pc tcg/loongarch64: Fill out tcg_out_{ld,st} for vector regs disas/disas-internal.h | 4 + include/disas/disas.h | 9 +- include/exec/cpu_ldst.h | 10 -- include/exec/plugin-gen.h | 7 +- include/exec/translator.h | 74 ++++++--- include/qemu/plugin.h | 22 +-- include/qemu/qemu-plugin.h | 15 +- include/qemu/typedefs.h | 1 + include/tcg/tcg.h | 1 + accel/tcg/plugin-gen.c | 63 +++----- accel/tcg/translator.c | 331 ++++++++++++++++++++++++-------------- contrib/plugins/execlog.c | 5 +- contrib/plugins/howvec.c | 4 +- disas/disas-common.c | 104 ++++++++++++ disas/disas-host.c | 129 +++++++++++++++ disas/disas-mon.c | 15 ++ disas/disas-target.c | 99 ++++++++++++ disas/disas.c | 338 --------------------------------------- disas/objdump.c | 37 +++++ plugins/api.c | 57 +++++-- target/alpha/translate.c | 9 -- target/arm/tcg/translate-a64.c | 11 -- target/arm/tcg/translate.c | 12 -- target/avr/translate.c | 11 +- target/cris/translate.c | 37 +---- target/hexagon/translate.c | 11 +- target/hppa/translate.c | 21 ++- target/i386/tcg/translate.c | 19 +-- target/loongarch/tcg/translate.c | 8 - target/m68k/translate.c | 9 -- target/microblaze/translate.c | 11 +- target/mips/tcg/translate.c | 9 -- target/openrisc/translate.c | 11 -- target/ppc/translate.c | 9 -- target/riscv/translate.c | 24 +-- target/rx/translate.c | 35 ++-- target/s390x/tcg/translate.c | 26 ++- target/sh4/translate.c | 9 -- target/sparc/translate.c | 9 -- target/tricore/translate.c | 9 -- target/xtensa/translate.c | 12 +- tcg/tcg.c | 12 -- target/cris/translate_v10.c.inc | 30 ++-- tcg/loongarch64/tcg-target.c.inc | 103 +++++++++--- disas/meson.build | 8 +- 45 files changed, 899 insertions(+), 891 deletions(-) create mode 100644 disas/disas-common.c create mode 100644 disas/disas-host.c create mode 100644 disas/disas-target.c delete mode 100644 disas/disas.c create mode 100644 disas/objdump.c Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 21 +++++++++------------ accel/tcg/translator.c | 15 ++++++++------- target/hexagon/translate.c | 1 + target/microblaze/translate.c | 1 + 4 files changed, 19 insertions(+), 19 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 @@ */ #include "qemu/bswap.h" -#include "exec/cpu-common.h" -#include "exec/cpu-defs.h" -#include "exec/abi_ptr.h" -#include "cpu.h" +#include "exec/vaddr.h" /** * gen_intermediate_code @@ -XXX,XX +XXX,XX @@ bool translator_io_start(DisasContextBase *db); * the relevant information at translation time. */ -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); +uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc); +uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, vaddr pc); +uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, vaddr pc); +uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc); static inline uint16_t translator_lduw_swap(CPUArchState *env, DisasContextBase *db, - abi_ptr pc, bool do_swap) + vaddr pc, bool do_swap) { uint16_t ret = translator_lduw(env, db, pc); if (do_swap) { @@ -XXX,XX +XXX,XX @@ translator_lduw_swap(CPUArchState *env, DisasContextBase *db, static inline uint32_t translator_ldl_swap(CPUArchState *env, DisasContextBase *db, - abi_ptr pc, bool do_swap) + vaddr pc, bool do_swap) { uint32_t ret = translator_ldl(env, db, pc); if (do_swap) { @@ -XXX,XX +XXX,XX @@ translator_ldl_swap(CPUArchState *env, DisasContextBase *db, static inline uint64_t translator_ldq_swap(CPUArchState *env, DisasContextBase *db, - abi_ptr pc, bool do_swap) + vaddr pc, bool do_swap) { uint64_t ret = translator_ldq(env, db, pc); if (do_swap) { @@ -XXX,XX +XXX,XX @@ translator_ldq_swap(CPUArchState *env, DisasContextBase *db, * re-synthesised for s390x "ex"). It ensures we update other areas of * the translator with details of the executed instruction. */ -void translator_fake_ldb(uint8_t insn8, abi_ptr pc); +void translator_fake_ldb(uint8_t insn8, vaddr pc); /* * Return whether addr is on the same page as where disassembly started. 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 @@ #include "exec/translator.h" #include "exec/cpu_ldst.h" #include "exec/plugin-gen.h" +#include "exec/cpu_ldst.h" #include "tcg/tcg-op-common.h" #include "internal-target.h" @@ -XXX,XX +XXX,XX @@ static void *translator_access(CPUArchState *env, DisasContextBase *db, return host + (pc - base); } -static void plugin_insn_append(abi_ptr pc, const void *from, size_t size) +static void plugin_insn_append(vaddr pc, const void *from, size_t size) { #ifdef CONFIG_PLUGIN struct qemu_plugin_insn *insn = tcg_ctx->plugin_insn; - abi_ptr off; + size_t off; if (insn == NULL) { return; @@ -XXX,XX +XXX,XX @@ static void plugin_insn_append(abi_ptr pc, const void *from, size_t size) #endif } -uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc) { uint8_t ret; void *p = translator_access(env, db, pc, sizeof(ret)); @@ -XXX,XX +XXX,XX @@ uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, abi_ptr pc) return ret; } -uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, vaddr pc) { uint16_t ret, plug; void *p = translator_access(env, db, pc, sizeof(ret)); @@ -XXX,XX +XXX,XX @@ uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, abi_ptr pc) return ret; } -uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, vaddr pc) { uint32_t ret, plug; void *p = translator_access(env, db, pc, sizeof(ret)); @@ -XXX,XX +XXX,XX @@ uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, abi_ptr pc) return ret; } -uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, abi_ptr pc) +uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) { uint64_t ret, plug; void *p = translator_access(env, db, pc, sizeof(ret)); @@ -XXX,XX +XXX,XX @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, abi_ptr pc) return ret; } -void translator_fake_ldb(uint8_t insn8, abi_ptr pc) +void translator_fake_ldb(uint8_t insn8, vaddr pc) { plugin_insn_append(pc, &insn8, sizeof(insn8)); } 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 @@ #include "exec/translation-block.h" #include "exec/cpu_ldst.h" #include "exec/log.h" +#include "exec/cpu_ldst.h" #include "internal.h" #include "attribs.h" #include "insn.h" 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 @@ #include "tcg/tcg-op.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" +#include "exec/cpu_ldst.h" #include "exec/translator.h" #include "qemu/qemu-print.h" -- 2.34.1 While there are other methods that could be used to replace TARGET_PAGE_MASK, the function is not really required outside the context of target-specific translation. This makes the header usable by target independent code. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 2 ++ 1 file changed, 2 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 @@ translator_ldq_swap(CPUArchState *env, DisasContextBase *db, */ void translator_fake_ldb(uint8_t insn8, vaddr pc); +#ifdef COMPILING_PER_TARGET /* * Return whether addr is on the same page as where disassembly started. * Translators can use this to enforce the rule that only single-insn @@ -XXX,XX +XXX,XX @@ static inline bool is_same_page(const DisasContextBase *db, vaddr addr) { return ((addr ^ db->pc_first) & TARGET_PAGE_MASK) == 0; } +#endif #endif /* EXEC__TRANSLATOR_H */ -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 5 +++-- accel/tcg/translator.c | 2 +- target/s390x/tcg/translate.c | 2 +- 3 files changed, 5 insertions(+), 4 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 @@ translator_ldq_swap(CPUArchState *env, DisasContextBase *db, /** * translator_fake_ldb - fake instruction load - * @insn8: byte of instruction + * @db: Disassembly context * @pc: program counter of instruction + * @insn8: byte of instruction * * This is a special case helper used where the instruction we are * about to translate comes from somewhere else (e.g. being * re-synthesised for s390x "ex"). It ensures we update other areas of * the translator with details of the executed instruction. */ -void translator_fake_ldb(uint8_t insn8, vaddr pc); +void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8); #ifdef COMPILING_PER_TARGET /* 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 @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) return ret; } -void translator_fake_ldb(uint8_t insn8, vaddr pc) +void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) { plugin_insn_append(pc, &insn8, sizeof(insn8)); } 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 DisasInsn *extract_insn(CPUS390XState *env, DisasContext *s) /* Register insn bytes with translator so plugins work. */ for (int i = 0; i < ilen; i++) { uint8_t byte = extract64(insn, 56 - (i * 8), 8); - translator_fake_ldb(byte, pc + i); + translator_fake_ldb(&s->base, pc + i, byte); } op = insn >> 56; } else { -- 2.34.1 Reorg translator_access into translator_ld, with a more memcpy-ish interface. If both pages are in ram, do not go through the caller's slow path. Assert that the access is within the two pages that we are prepared to protect, per TranslationBlock. Allow access prior to pc_first, so long as it is within the first page. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/translator.c | 189 ++++++++++++++++++++++------------------- 1 file changed, 101 insertions(+), 88 deletions(-) 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, } } -static void *translator_access(CPUArchState *env, DisasContextBase *db, - vaddr pc, size_t len) +static bool translator_ld(CPUArchState *env, DisasContextBase *db, + void *dest, vaddr pc, size_t len) { + TranslationBlock *tb = db->tb; + vaddr last = pc + len - 1; void *host; - vaddr base, end; - TranslationBlock *tb; - - tb = db->tb; + vaddr base; /* Use slow path if first page is MMIO. */ if (unlikely(tb_page_addr0(tb) == -1)) { - return NULL; + return false; } - 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[0]; + base = db->pc_first; + + if (likely(((base ^ last) & TARGET_PAGE_MASK) == 0)) { + /* Entire read is from the first page. */ + memcpy(dest, host + (pc - base), len); + return true; + } + + if (unlikely(((base ^ pc) & TARGET_PAGE_MASK) == 0)) { + /* Read begins on the first page and extends to the second. */ + size_t len0 = -(pc | TARGET_PAGE_MASK); + memcpy(dest, host + (pc - base), len0); + pc += len0; + dest += len0; + len -= len0; + } + + /* + * The read must conclude on the second page and not extend to a third. + * + * TODO: We could allow the two pages to be virtually discontiguous, + * since we already allow the two pages to be physically discontiguous. + * The only reasonable use case would be executing an insn at the end + * of the address space wrapping around to the beginning. For that, + * we would need to know the current width of the address space. + * In the meantime, assert. + */ + base = (base & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE; + assert(((base ^ pc) & TARGET_PAGE_MASK) == 0); + assert(((base ^ last) & TARGET_PAGE_MASK) == 0); + host = db->host_addr[1]; + + if (host == NULL) { + tb_page_addr_t page0, old_page1, new_page1; + + new_page1 = get_page_addr_code_hostp(env, base, &db->host_addr[1]); + + /* + * If the second page is MMIO, treat as if the first page + * was MMIO as well, so that we do not cache the TB. + */ + if (unlikely(new_page1 == -1)) { + tb_unlock_pages(tb); + tb_set_page_addr0(tb, -1); + return false; + } + + /* + * If this is not the first time around, and page1 matches, + * then we already have the page locked. Alternately, we're + * not doing anything to prevent the PTE from changing, so + * we might wind up with a different page, requiring us to + * re-do the locking. + */ + old_page1 = tb_page_addr1(tb); + if (likely(new_page1 != old_page1)) { + page0 = tb_page_addr0(tb); + if (unlikely(old_page1 != -1)) { + tb_unlock_page1(page0, old_page1); + } + tb_set_page_addr1(tb, new_page1); + tb_lock_page1(page0, new_page1); + } host = db->host_addr[1]; - base = TARGET_PAGE_ALIGN(db->pc_first); - if (host == NULL) { - tb_page_addr_t page0, old_page1, new_page1; - - new_page1 = get_page_addr_code_hostp(env, base, &db->host_addr[1]); - - /* - * If the second page is MMIO, treat as if the first page - * was MMIO as well, so that we do not cache the TB. - */ - if (unlikely(new_page1 == -1)) { - tb_unlock_pages(tb); - tb_set_page_addr0(tb, -1); - return NULL; - } - - /* - * If this is not the first time around, and page1 matches, - * then we already have the page locked. Alternately, we're - * not doing anything to prevent the PTE from changing, so - * we might wind up with a different page, requiring us to - * re-do the locking. - */ - old_page1 = tb_page_addr1(tb); - if (likely(new_page1 != old_page1)) { - page0 = tb_page_addr0(tb); - if (unlikely(old_page1 != -1)) { - tb_unlock_page1(page0, old_page1); - } - tb_set_page_addr1(tb, new_page1); - tb_lock_page1(page0, new_page1); - } - 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); + memcpy(dest, host + (pc - base), len); + return true; } static void plugin_insn_append(vaddr pc, const void *from, size_t size) @@ -XXX,XX +XXX,XX @@ static void plugin_insn_append(vaddr pc, const void *from, size_t size) uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc) { - uint8_t ret; - void *p = translator_access(env, db, pc, sizeof(ret)); + uint8_t raw; - if (p) { - plugin_insn_append(pc, p, sizeof(ret)); - return ldub_p(p); + if (!translator_ld(env, db, &raw, pc, sizeof(raw))) { + raw = cpu_ldub_code(env, pc); } - ret = cpu_ldub_code(env, pc); - plugin_insn_append(pc, &ret, sizeof(ret)); - return ret; + plugin_insn_append(pc, &raw, sizeof(raw)); + return raw; } uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, vaddr pc) { - uint16_t ret, plug; - void *p = translator_access(env, db, pc, sizeof(ret)); + uint16_t raw, tgt; - if (p) { - plugin_insn_append(pc, p, sizeof(ret)); - return lduw_p(p); + if (translator_ld(env, db, &raw, pc, sizeof(raw))) { + tgt = tswap16(raw); + } else { + tgt = cpu_lduw_code(env, pc); + raw = tswap16(tgt); } - ret = cpu_lduw_code(env, pc); - plug = tswap16(ret); - plugin_insn_append(pc, &plug, sizeof(ret)); - return ret; + plugin_insn_append(pc, &raw, sizeof(raw)); + return tgt; } uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, vaddr pc) { - uint32_t ret, plug; - void *p = translator_access(env, db, pc, sizeof(ret)); + uint32_t raw, tgt; - if (p) { - plugin_insn_append(pc, p, sizeof(ret)); - return ldl_p(p); + if (translator_ld(env, db, &raw, pc, sizeof(raw))) { + tgt = tswap32(raw); + } else { + tgt = cpu_ldl_code(env, pc); + raw = tswap32(tgt); } - ret = cpu_ldl_code(env, pc); - plug = tswap32(ret); - plugin_insn_append(pc, &plug, sizeof(ret)); - return ret; + plugin_insn_append(pc, &raw, sizeof(raw)); + return tgt; } uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) { - uint64_t ret, plug; - void *p = translator_access(env, db, pc, sizeof(ret)); + uint64_t raw, tgt; - if (p) { - plugin_insn_append(pc, p, sizeof(ret)); - return ldq_p(p); + if (translator_ld(env, db, &raw, pc, sizeof(raw))) { + tgt = tswap64(raw); + } else { + tgt = cpu_ldq_code(env, pc); + raw = tswap64(tgt); } - ret = cpu_ldq_code(env, pc); - plug = tswap64(ret); - plugin_insn_append(pc, &plug, sizeof(ret)); - return ret; + plugin_insn_append(pc, &raw, sizeof(raw)); + return tgt; } void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) -- 2.34.1 Do not allow translation to proceed beyond one insn with mmio, as we will not be caching the TranslationBlock. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/translator.c | 4 ++++ 1 file changed, 4 insertions(+) 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 bool translator_ld(CPUArchState *env, DisasContextBase *db, /* Use slow path if first page is MMIO. */ if (unlikely(tb_page_addr0(tb) == -1)) { + /* We capped translation with first page MMIO in tb_gen_code. */ + tcg_debug_assert(db->max_insns == 1); return false; } @@ -XXX,XX +XXX,XX @@ static bool translator_ld(CPUArchState *env, DisasContextBase *db, if (unlikely(new_page1 == -1)) { tb_unlock_pages(tb); tb_set_page_addr0(tb, -1); + /* Require that this be the final insn. */ + db->max_insns = db->num_insns; return false; } -- 2.34.1 This will be able to replace plugin_insn_append, and will be usable for disassembly. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 12 ++++++++++++ accel/tcg/translator.c | 41 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 53 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 @@ typedef struct DisasContextBase { bool plugin_enabled; struct TCGOp *insn_start; void *host_addr[2]; + + /* + * Record insn data that we cannot read directly from host memory. + * There are only two reasons we cannot use host memory: + * (1) We are executing from I/O, + * (2) We are executing a synthetic instruction (s390x EX). + * In both cases we need record exactly one instruction, + * and thus the maximum amount of data we record is limited. + */ + int record_start; + int record_len; + uint8_t record[32]; } DisasContextBase; /** 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, db->insn_start = NULL; db->host_addr[0] = host_pc; db->host_addr[1] = NULL; + db->record_start = 0; + db->record_len = 0; ops->init_disas_context(db, cpu); tcg_debug_assert(db->is_jmp == DISAS_NEXT); /* no early exit */ @@ -XXX,XX +XXX,XX @@ static bool translator_ld(CPUArchState *env, DisasContextBase *db, return true; } +static void record_save(DisasContextBase *db, vaddr pc, + const void *from, int size) +{ + int offset; + + /* Do not record probes before the start of TB. */ + if (pc < db->pc_first) { + return; + } + + /* + * In translator_access, we verified that pc is within 2 pages + * of pc_first, thus this will never overflow. + */ + offset = pc - db->pc_first; + + /* + * Either the first or second page may be I/O. If it is the second, + * then the first byte we need to record will be at a non-zero offset. + * In either case, we should not need to record but a single insn. + */ + if (db->record_len == 0) { + db->record_start = offset; + db->record_len = size; + } else { + assert(offset == db->record_start + db->record_len); + assert(db->record_len + size <= sizeof(db->record)); + db->record_len += size; + } + + memcpy(db->record + (offset - db->record_start), from, size); +} + static void plugin_insn_append(vaddr pc, const void *from, size_t size) { #ifdef CONFIG_PLUGIN @@ -XXX,XX +XXX,XX @@ uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc) if (!translator_ld(env, db, &raw, pc, sizeof(raw))) { raw = cpu_ldub_code(env, pc); + record_save(db, pc, &raw, sizeof(raw)); } plugin_insn_append(pc, &raw, sizeof(raw)); return raw; @@ -XXX,XX +XXX,XX @@ uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, vaddr pc) } else { tgt = cpu_lduw_code(env, pc); raw = tswap16(tgt); + record_save(db, pc, &raw, sizeof(raw)); } plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; @@ -XXX,XX +XXX,XX @@ uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, vaddr pc) } else { tgt = cpu_ldl_code(env, pc); raw = tswap32(tgt); + record_save(db, pc, &raw, sizeof(raw)); } plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; @@ -XXX,XX +XXX,XX @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) } else { tgt = cpu_ldq_code(env, pc); raw = tswap64(tgt); + record_save(db, pc, &raw, sizeof(raw)); } plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; @@ -XXX,XX +XXX,XX @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) { + assert(pc >= db->pc_first); + record_save(db, pc, &insn8, sizeof(insn8)); plugin_insn_append(pc, &insn8, sizeof(insn8)); } -- 2.34.1 Remove left-over comment from commit dcd092a063 ("accel/tcg: Improve can_do_io management"). Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 3 ++- accel/tcg/translator.c | 2 ++ 2 files changed, 4 insertions(+), 1 deletion(-) 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 { * @num_insns: Number of translated instructions (including current). * @max_insns: Maximum number of instructions to be translated in this TB. * @singlestep_enabled: "Hardware" single stepping enabled. - * @saved_can_do_io: Known value of cpu->neg.can_do_io, or -1 for unknown. * @plugin_enabled: TCG plugin enabled in this TB. + * @fake_insn: True if translator_fake_ldb used. * @insn_start: The last op emitted by the insn_start hook, * which is expected to be INDEX_op_insn_start. * @@ -XXX,XX +XXX,XX @@ typedef struct DisasContextBase { int max_insns; bool singlestep_enabled; bool plugin_enabled; + bool fake_insn; struct TCGOp *insn_start; void *host_addr[2]; 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, db->max_insns = *max_insns; db->singlestep_enabled = cflags & CF_SINGLE_STEP; db->insn_start = NULL; + db->fake_insn = false; db->host_addr[0] = host_pc; db->host_addr[1] = NULL; db->record_start = 0; @@ -XXX,XX +XXX,XX @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) { assert(pc >= db->pc_first); + db->fake_insn = true; record_save(db, pc, &insn8, sizeof(insn8)); plugin_insn_append(pc, &insn8, sizeof(insn8)); } -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/tcg/tcg.h | 1 + accel/tcg/plugin-gen.c | 1 + 2 files changed, 2 insertions(+) diff --git a/include/tcg/tcg.h b/include/tcg/tcg.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg.h +++ b/include/tcg/tcg.h @@ -XXX,XX +XXX,XX @@ struct TCGContext { * space for instructions (for variable-instruction-length ISAs). */ struct qemu_plugin_tb *plugin_tb; + const struct DisasContextBase *plugin_db; /* descriptor of the instruction being translated */ struct qemu_plugin_insn *plugin_insn; diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db, tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); } + tcg_ctx->plugin_db = db; tcg_ctx->plugin_insn = NULL; return ret; -- 2.34.1 Instead of returning a host pointer, copy the data into storage provided by the caller. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/qemu/qemu-plugin.h | 15 +++++++-------- contrib/plugins/execlog.c | 5 +++-- contrib/plugins/howvec.c | 4 ++-- plugins/api.c | 7 +++++-- 4 files changed, 17 insertions(+), 14 deletions(-) diff --git a/include/qemu/qemu-plugin.h b/include/qemu/qemu-plugin.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/qemu-plugin.h +++ b/include/qemu/qemu-plugin.h @@ -XXX,XX +XXX,XX @@ typedef uint64_t qemu_plugin_id_t; extern QEMU_PLUGIN_EXPORT int qemu_plugin_version; -#define QEMU_PLUGIN_VERSION 2 +#define QEMU_PLUGIN_VERSION 3 /** * struct qemu_info_t - system information for plugins @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_insn * qemu_plugin_tb_get_insn(const struct qemu_plugin_tb *tb, size_t idx); /** - * qemu_plugin_insn_data() - return ptr to instruction data + * qemu_plugin_insn_data() - copy instruction data * @insn: opaque instruction handle from qemu_plugin_tb_get_insn() + * @dest: destination into which data is copied + * @len: length of dest * - * Note: data is only valid for duration of callback. See - * qemu_plugin_insn_size() to calculate size of stream. - * - * Returns: pointer to a stream of bytes containing the value of this - * instructions opcode. + * Returns the number of bytes copied, minimum of @len and insn size. */ QEMU_PLUGIN_API -const void *qemu_plugin_insn_data(const struct qemu_plugin_insn *insn); +size_t qemu_plugin_insn_data(const struct qemu_plugin_insn *insn, + void *dest, size_t len); /** * qemu_plugin_insn_size() - return size of instruction diff --git a/contrib/plugins/execlog.c b/contrib/plugins/execlog.c index XXXXXXX..XXXXXXX 100644 --- a/contrib/plugins/execlog.c +++ b/contrib/plugins/execlog.c @@ -XXX,XX +XXX,XX @@ static void vcpu_tb_trans(qemu_plugin_id_t id, struct qemu_plugin_tb *tb) NULL); } } else { - uint32_t insn_opcode; - insn_opcode = *((uint32_t *)qemu_plugin_insn_data(insn)); + uint32_t insn_opcode = 0; + qemu_plugin_insn_data(insn, &insn_opcode, sizeof(insn_opcode)); + char *output = g_strdup_printf("0x%"PRIx64", 0x%"PRIx32", \"%s\"", insn_vaddr, insn_opcode, insn_disas); diff --git a/contrib/plugins/howvec.c b/contrib/plugins/howvec.c index XXXXXXX..XXXXXXX 100644 --- a/contrib/plugins/howvec.c +++ b/contrib/plugins/howvec.c @@ -XXX,XX +XXX,XX @@ static struct qemu_plugin_scoreboard *find_counter( { int i; uint64_t *cnt = NULL; - uint32_t opcode; + uint32_t opcode = 0; InsnClassExecCount *class = NULL; /* @@ -XXX,XX +XXX,XX @@ static struct qemu_plugin_scoreboard *find_counter( * They would probably benefit from a more tailored plugin. * However we can fall back to individual instruction counting. */ - opcode = *((uint32_t *)qemu_plugin_insn_data(insn)); + qemu_plugin_insn_data(insn, &opcode, sizeof(opcode)); for (i = 0; !cnt && i < class_table_sz; i++) { class = &class_table[i]; diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ qemu_plugin_tb_get_insn(const struct qemu_plugin_tb *tb, size_t idx) * instruction being translated. */ -const void *qemu_plugin_insn_data(const struct qemu_plugin_insn *insn) +size_t qemu_plugin_insn_data(const struct qemu_plugin_insn *insn, + void *dest, size_t len) { - return insn->data->data; + len = MIN(len, insn->data->len); + memcpy(dest, insn->data->data, len); + return len; } size_t qemu_plugin_insn_size(const struct qemu_plugin_insn *insn) -- 2.34.1 Copy data out of a completed translation. This will be used for both plugins and disassembly. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 23 ++++++++++++++++ accel/tcg/translator.c | 55 +++++++++++++++++++++++++++++++++++++++ 2 files changed, 78 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 @@ translator_ldq_swap(CPUArchState *env, DisasContextBase *db, */ void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8); +/** + * translator_st + * @db: disassembly context + * @dest: address to copy into + * @addr: virtual address within TB + * @len: length + * + * Copy @len bytes from @addr into @dest. + * All bytes must have been read during translation. + * Return true on success or false on failure. + */ +bool translator_st(const DisasContextBase *db, void *dest, + vaddr addr, size_t len); + +/** + * translator_st_len + * @db: disassembly context + * + * Return the number of bytes available to copy from the + * current translation block with translator_st. + */ +size_t translator_st_len(const DisasContextBase *db); + #ifdef COMPILING_PER_TARGET /* * Return whether addr is on the same page as where disassembly started. 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 void record_save(DisasContextBase *db, vaddr pc, memcpy(db->record + (offset - db->record_start), from, size); } +size_t translator_st_len(const DisasContextBase *db) +{ + return db->fake_insn ? db->record_len : db->tb->size; +} + +bool translator_st(const DisasContextBase *db, void *dest, + vaddr addr, size_t len) +{ + size_t offset, offset_end; + + if (addr < db->pc_first) { + return false; + } + offset = addr - db->pc_first; + offset_end = offset + len; + if (offset_end > translator_st_len(db)) { + return false; + } + + if (!db->fake_insn) { + size_t offset_page1 = -(db->pc_first | TARGET_PAGE_MASK); + + /* Get all the bytes from the first page. */ + if (db->host_addr[0]) { + if (offset_end <= offset_page1) { + memcpy(dest, db->host_addr[0] + offset, len); + return true; + } + if (offset < offset_page1) { + size_t len0 = offset_page1 - offset; + memcpy(dest, db->host_addr[0] + offset, len0); + offset += len0; + dest += len0; + } + } + + /* Get any bytes from the second page. */ + if (db->host_addr[1] && offset >= offset_page1) { + memcpy(dest, db->host_addr[1] + (offset - offset_page1), + offset_end - offset); + return true; + } + } + + /* Else get recorded bytes. */ + if (db->record_len != 0 && + offset >= db->record_start && + offset_end <= db->record_start + db->record_len) { + memcpy(dest, db->record + (offset - db->record_start), + offset_end - offset); + return true; + } + return false; +} + static void plugin_insn_append(vaddr pc, const void *from, size_t size) { #ifdef CONFIG_PLUGIN -- 2.34.1 Use the bytes that we record for the entire TB, rather than a per-insn GByteArray. Record the length of the insn in plugin_gen_insn_end rather than infering from the length of the array. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/qemu/plugin.h | 14 +------------- accel/tcg/plugin-gen.c | 7 +++++-- accel/tcg/translator.c | 26 -------------------------- plugins/api.c | 12 +++++++----- tcg/tcg.c | 3 +-- 5 files changed, 14 insertions(+), 48 deletions(-) diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/plugin.h +++ b/include/qemu/plugin.h @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_dyn_cb { /* Internal context for instrumenting an instruction */ struct qemu_plugin_insn { - GByteArray *data; uint64_t vaddr; void *haddr; GArray *insn_cbs; GArray *mem_cbs; + uint8_t len; bool calls_helpers; /* if set, the instruction calls helpers that might access guest memory */ @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_scoreboard { QLIST_ENTRY(qemu_plugin_scoreboard) entry; }; -/* - * qemu_plugin_insn allocate and cleanup functions. We don't expect to - * cleanup many of these structures. They are reused for each fresh - * translation. - */ - -static inline void qemu_plugin_insn_cleanup_fn(gpointer data) -{ - struct qemu_plugin_insn *insn = (struct qemu_plugin_insn *) data; - g_byte_array_free(insn->data, true); -} - /* Internal context for this TranslationBlock */ struct qemu_plugin_tb { GPtrArray *insns; diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ void plugin_gen_insn_start(CPUState *cpu, const DisasContextBase *db) ptb->n = n; if (n <= ptb->insns->len) { insn = g_ptr_array_index(ptb->insns, n - 1); - g_byte_array_set_size(insn->data, 0); } else { assert(n - 1 == ptb->insns->len); insn = g_new0(struct qemu_plugin_insn, 1); - insn->data = g_byte_array_sized_new(4); g_ptr_array_add(ptb->insns, insn); } @@ -XXX,XX +XXX,XX @@ void plugin_gen_insn_start(CPUState *cpu, const DisasContextBase *db) void plugin_gen_insn_end(void) { + const DisasContextBase *db = tcg_ctx->plugin_db; + struct qemu_plugin_insn *pinsn = tcg_ctx->plugin_insn; + + pinsn->len = db->fake_insn ? db->record_len : db->pc_next - pinsn->vaddr; + tcg_gen_plugin_cb(PLUGIN_GEN_AFTER_INSN); } 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_st(const DisasContextBase *db, void *dest, return false; } -static void plugin_insn_append(vaddr pc, const void *from, size_t size) -{ -#ifdef CONFIG_PLUGIN - struct qemu_plugin_insn *insn = tcg_ctx->plugin_insn; - size_t off; - - if (insn == NULL) { - return; - } - off = pc - insn->vaddr; - if (off < insn->data->len) { - g_byte_array_set_size(insn->data, off); - } else if (off > insn->data->len) { - /* we have an unexpected gap */ - g_assert_not_reached(); - } - - insn->data = g_byte_array_append(insn->data, from, size); -#endif -} - uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc) { uint8_t raw; @@ -XXX,XX +XXX,XX @@ uint8_t translator_ldub(CPUArchState *env, DisasContextBase *db, vaddr pc) raw = cpu_ldub_code(env, pc); record_save(db, pc, &raw, sizeof(raw)); } - plugin_insn_append(pc, &raw, sizeof(raw)); return raw; } @@ -XXX,XX +XXX,XX @@ uint16_t translator_lduw(CPUArchState *env, DisasContextBase *db, vaddr pc) raw = tswap16(tgt); record_save(db, pc, &raw, sizeof(raw)); } - plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; } @@ -XXX,XX +XXX,XX @@ uint32_t translator_ldl(CPUArchState *env, DisasContextBase *db, vaddr pc) raw = tswap32(tgt); record_save(db, pc, &raw, sizeof(raw)); } - plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; } @@ -XXX,XX +XXX,XX @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) raw = tswap64(tgt); record_save(db, pc, &raw, sizeof(raw)); } - plugin_insn_append(pc, &raw, sizeof(raw)); return tgt; } @@ -XXX,XX +XXX,XX @@ void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) assert(pc >= db->pc_first); db->fake_insn = true; record_save(db, pc, &insn8, sizeof(insn8)); - plugin_insn_append(pc, &insn8, sizeof(insn8)); } diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ #include "tcg/tcg.h" #include "exec/exec-all.h" #include "exec/gdbstub.h" +#include "exec/translator.h" #include "disas/disas.h" #include "plugin.h" #ifndef CONFIG_USER_ONLY @@ -XXX,XX +XXX,XX @@ qemu_plugin_tb_get_insn(const struct qemu_plugin_tb *tb, size_t idx) size_t qemu_plugin_insn_data(const struct qemu_plugin_insn *insn, void *dest, size_t len) { - len = MIN(len, insn->data->len); - memcpy(dest, insn->data->data, len); - return len; + const DisasContextBase *db = tcg_ctx->plugin_db; + + len = MIN(len, insn->len); + return translator_st(db, dest, insn->vaddr, len) ? len : 0; } size_t qemu_plugin_insn_size(const struct qemu_plugin_insn *insn) { - return insn->data->len; + return insn->len; } uint64_t qemu_plugin_insn_vaddr(const struct qemu_plugin_insn *insn) @@ -XXX,XX +XXX,XX @@ void *qemu_plugin_insn_haddr(const struct qemu_plugin_insn *insn) char *qemu_plugin_insn_disas(const struct qemu_plugin_insn *insn) { CPUState *cpu = current_cpu; - return plugin_disas(cpu, insn->vaddr, insn->data->len); + return plugin_disas(cpu, insn->vaddr, insn->len); } const char *qemu_plugin_insn_symbol(const struct qemu_plugin_insn *insn) 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 alloc_tcg_plugin_context(TCGContext *s) { #ifdef CONFIG_PLUGIN s->plugin_tb = g_new0(struct qemu_plugin_tb, 1); - s->plugin_tb->insns = - g_ptr_array_new_with_free_func(qemu_plugin_insn_cleanup_fn); + s->plugin_tb->insns = g_ptr_array_new(); #endif } -- 2.34.1 Do not pass around a boolean between multiple structures, just read it from the TranslationBlock in the TCGContext. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/plugin-gen.h | 7 +++---- include/qemu/plugin.h | 3 --- accel/tcg/plugin-gen.c | 4 +--- accel/tcg/translator.c | 2 +- plugins/api.c | 14 +++++++++----- 5 files changed, 14 insertions(+), 16 deletions(-) diff --git a/include/exec/plugin-gen.h b/include/exec/plugin-gen.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/plugin-gen.h +++ b/include/exec/plugin-gen.h @@ -XXX,XX +XXX,XX @@ struct DisasContextBase; #ifdef CONFIG_PLUGIN -bool plugin_gen_tb_start(CPUState *cpu, const struct DisasContextBase *db, - bool supress); +bool plugin_gen_tb_start(CPUState *cpu, const struct DisasContextBase *db); void plugin_gen_tb_end(CPUState *cpu, size_t num_insns); void plugin_gen_insn_start(CPUState *cpu, const struct DisasContextBase *db); void plugin_gen_insn_end(void); @@ -XXX,XX +XXX,XX @@ void plugin_gen_disable_mem_helpers(void); #else /* !CONFIG_PLUGIN */ -static inline bool -plugin_gen_tb_start(CPUState *cpu, const struct DisasContextBase *db, bool sup) +static inline +bool plugin_gen_tb_start(CPUState *cpu, const struct DisasContextBase *db) { return false; } diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/plugin.h +++ b/include/qemu/plugin.h @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_insn { /* if set, the instruction calls helpers that might access guest memory */ bool mem_helper; - - bool mem_only; }; /* A scoreboard is an array of values, indexed by vcpu_index */ @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_tb { uint64_t vaddr2; void *haddr1; void *haddr2; - bool mem_only; /* if set, the TB calls helpers that might access guest memory */ bool mem_helper; diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ static void plugin_gen_inject(struct qemu_plugin_tb *plugin_tb) } } -bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db, - bool mem_only) +bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db) { bool ret = false; @@ -XXX,XX +XXX,XX @@ bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db, ptb->vaddr2 = -1; ptb->haddr1 = db->host_addr[0]; ptb->haddr2 = NULL; - ptb->mem_only = mem_only; ptb->mem_helper = false; tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, ops->tb_start(db, cpu); tcg_debug_assert(db->is_jmp == DISAS_NEXT); /* no early exit */ - plugin_enabled = plugin_gen_tb_start(cpu, db, cflags & CF_MEMI_ONLY); + plugin_enabled = plugin_gen_tb_start(cpu, db); db->plugin_enabled = plugin_enabled; while (true) { diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ void qemu_plugin_register_vcpu_exit_cb(qemu_plugin_id_t id, plugin_register_cb(id, QEMU_PLUGIN_EV_VCPU_EXIT, cb); } +static bool tb_is_mem_only(void) +{ + return tb_cflags(tcg_ctx->gen_tb) & CF_MEMI_ONLY; +} + void qemu_plugin_register_vcpu_tb_exec_cb(struct qemu_plugin_tb *tb, qemu_plugin_vcpu_udata_cb_t cb, enum qemu_plugin_cb_flags flags, void *udata) { - if (!tb->mem_only) { + if (!tb_is_mem_only()) { plugin_register_dyn_cb__udata(&tb->cbs, cb, flags, udata); } } @@ -XXX,XX +XXX,XX @@ void qemu_plugin_register_vcpu_tb_exec_inline_per_vcpu( qemu_plugin_u64 entry, uint64_t imm) { - if (!tb->mem_only) { + if (!tb_is_mem_only()) { plugin_register_inline_op_on_entry(&tb->cbs, 0, op, entry, imm); } } @@ -XXX,XX +XXX,XX @@ void qemu_plugin_register_vcpu_insn_exec_cb(struct qemu_plugin_insn *insn, enum qemu_plugin_cb_flags flags, void *udata) { - if (!insn->mem_only) { + if (!tb_is_mem_only()) { plugin_register_dyn_cb__udata(&insn->insn_cbs, cb, flags, udata); } } @@ -XXX,XX +XXX,XX @@ void qemu_plugin_register_vcpu_insn_exec_inline_per_vcpu( qemu_plugin_u64 entry, uint64_t imm) { - if (!insn->mem_only) { + if (!tb_is_mem_only()) { plugin_register_inline_op_on_entry(&insn->insn_cbs, 0, op, entry, imm); } } @@ -XXX,XX +XXX,XX @@ qemu_plugin_tb_get_insn(const struct qemu_plugin_tb *tb, size_t idx) return NULL; } insn = g_ptr_array_index(tb->insns, idx); - insn->mem_only = tb->mem_only; return insn; } -- 2.34.1 We can delay the computation of haddr until the plugin actually requests it. Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/qemu/plugin.h | 4 ---- accel/tcg/plugin-gen.c | 20 -------------------- plugins/api.c | 25 ++++++++++++++++++++++++- 3 files changed, 24 insertions(+), 25 deletions(-) diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/plugin.h +++ b/include/qemu/plugin.h @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_dyn_cb { /* Internal context for instrumenting an instruction */ struct qemu_plugin_insn { uint64_t vaddr; - void *haddr; GArray *insn_cbs; GArray *mem_cbs; uint8_t len; @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_tb { GPtrArray *insns; size_t n; uint64_t vaddr; - uint64_t vaddr2; - void *haddr1; - void *haddr2; /* if set, the TB calls helpers that might access guest memory */ bool mem_helper; diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db) ret = true; ptb->vaddr = db->pc_first; - ptb->vaddr2 = -1; - ptb->haddr1 = db->host_addr[0]; - ptb->haddr2 = NULL; ptb->mem_helper = false; tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); @@ -XXX,XX +XXX,XX @@ void plugin_gen_insn_start(CPUState *cpu, const DisasContextBase *db) pc = db->pc_next; insn->vaddr = pc; - /* - * Detect page crossing to get the new host address. - * Note that we skip this when haddr1 == NULL, e.g. when we're - * fetching instructions from a region not backed by RAM. - */ - if (ptb->haddr1 == NULL) { - insn->haddr = NULL; - } else if (is_same_page(db, db->pc_next)) { - insn->haddr = ptb->haddr1 + pc - ptb->vaddr; - } else { - if (ptb->vaddr2 == -1) { - ptb->vaddr2 = TARGET_PAGE_ALIGN(db->pc_first); - get_page_addr_code_hostp(cpu_env(cpu), ptb->vaddr2, &ptb->haddr2); - } - insn->haddr = ptb->haddr2 + pc - ptb->vaddr2; - } - tcg_gen_plugin_cb(PLUGIN_GEN_FROM_INSN); } diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ uint64_t qemu_plugin_insn_vaddr(const struct qemu_plugin_insn *insn) void *qemu_plugin_insn_haddr(const struct qemu_plugin_insn *insn) { - return insn->haddr; + const DisasContextBase *db = tcg_ctx->plugin_db; + vaddr page0_last = db->pc_first | ~TARGET_PAGE_MASK; + + if (db->fake_insn) { + return NULL; + } + + /* + * ??? The return value is not intended for use of host memory, + * but as a proxy for address space and physical address. + * Thus we are only interested in the first byte and do not + * care about spanning pages. + */ + if (insn->vaddr <= page0_last) { + if (db->host_addr[0] == NULL) { + return NULL; + } + return db->host_addr[0] + insn->vaddr - db->pc_first; + } else { + if (db->host_addr[1] == NULL) { + return NULL; + } + return db->host_addr[1] + insn->vaddr - (page0_last + 1); + } } char *qemu_plugin_insn_disas(const struct qemu_plugin_insn *insn) -- 2.34.1 We do not need to separately record the start of the TB. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/qemu/plugin.h | 1 - accel/tcg/plugin-gen.c | 3 +-- plugins/api.c | 3 ++- 3 files changed, 3 insertions(+), 4 deletions(-) diff --git a/include/qemu/plugin.h b/include/qemu/plugin.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/plugin.h +++ b/include/qemu/plugin.h @@ -XXX,XX +XXX,XX @@ struct qemu_plugin_scoreboard { struct qemu_plugin_tb { GPtrArray *insns; size_t n; - uint64_t vaddr; /* if set, the TB calls helpers that might access guest memory */ bool mem_helper; diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ static void plugin_gen_inject(struct qemu_plugin_tb *plugin_tb) int insn_idx = -1; if (unlikely(qemu_loglevel_mask(LOG_TB_OP_PLUGIN) - && qemu_log_in_addr_range(plugin_tb->vaddr))) { + && qemu_log_in_addr_range(tcg_ctx->plugin_db->pc_first))) { FILE *logfile = qemu_log_trylock(); if (logfile) { fprintf(logfile, "OP before plugin injection:\n"); @@ -XXX,XX +XXX,XX @@ bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db) ret = true; - ptb->vaddr = db->pc_first; ptb->mem_helper = false; tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ size_t qemu_plugin_tb_n_insns(const struct qemu_plugin_tb *tb) uint64_t qemu_plugin_tb_vaddr(const struct qemu_plugin_tb *tb) { - return tb->vaddr; + const DisasContextBase *db = tcg_ctx->plugin_db; + return db->pc_first; } struct qemu_plugin_insn * -- 2.34.1 We don't need to allocate plugin context at startup, we can wait until we actually use it. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/plugin-gen.c | 36 ++++++++++++++++++++---------------- tcg/tcg.c | 11 ----------- 2 files changed, 20 insertions(+), 27 deletions(-) diff --git a/accel/tcg/plugin-gen.c b/accel/tcg/plugin-gen.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/plugin-gen.c +++ b/accel/tcg/plugin-gen.c @@ -XXX,XX +XXX,XX @@ static void plugin_gen_inject(struct qemu_plugin_tb *plugin_tb) bool plugin_gen_tb_start(CPUState *cpu, const DisasContextBase *db) { - bool ret = false; + struct qemu_plugin_tb *ptb; - if (test_bit(QEMU_PLUGIN_EV_VCPU_TB_TRANS, cpu->plugin_state->event_mask)) { - struct qemu_plugin_tb *ptb = tcg_ctx->plugin_tb; - - /* reset callbacks */ - if (ptb->cbs) { - g_array_set_size(ptb->cbs, 0); - } - ptb->n = 0; - - ret = true; - - ptb->mem_helper = false; - - tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); + if (!test_bit(QEMU_PLUGIN_EV_VCPU_TB_TRANS, + cpu->plugin_state->event_mask)) { + return false; } tcg_ctx->plugin_db = db; tcg_ctx->plugin_insn = NULL; + ptb = tcg_ctx->plugin_tb; - return ret; + if (ptb) { + /* Reset callbacks */ + if (ptb->cbs) { + g_array_set_size(ptb->cbs, 0); + } + ptb->n = 0; + ptb->mem_helper = false; + } else { + ptb = g_new0(struct qemu_plugin_tb, 1); + tcg_ctx->plugin_tb = ptb; + ptb->insns = g_ptr_array_new(); + } + + tcg_gen_plugin_cb(PLUGIN_GEN_FROM_TB); + return true; } void plugin_gen_insn_start(CPUState *cpu, const DisasContextBase *db) 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 @@ QEMU_BUILD_BUG_ON((int)(offsetof(CPUNegativeOffsetState, tlb.f[0]) - < MIN_TLB_MASK_TABLE_OFS); #endif -static void alloc_tcg_plugin_context(TCGContext *s) -{ -#ifdef CONFIG_PLUGIN - s->plugin_tb = g_new0(struct qemu_plugin_tb, 1); - s->plugin_tb->insns = g_ptr_array_new(); -#endif -} - /* * All TCG threads except the parent (i.e. the one that called tcg_context_init * and registered the target's TCG globals) must register with this function @@ -XXX,XX +XXX,XX @@ void tcg_register_thread(void) qatomic_set(&tcg_ctxs[n], s); if (n > 0) { - alloc_tcg_plugin_context(s); tcg_region_initial_alloc(s); } @@ -XXX,XX +XXX,XX @@ static void tcg_context_init(unsigned max_cpus) indirect_reg_alloc_order[i] = tcg_target_reg_alloc_order[i]; } - alloc_tcg_plugin_context(s); - tcg_ctx = s; /* * In user-mode we simply share the init context among threads, since we -- 2.34.1 Almost all of the disas_log implementations are identical. Unify them within translator_loop. Drop extra Priv/Virt logging from target/riscv. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/translator.c | 9 ++++++++- target/alpha/translate.c | 9 --------- target/arm/tcg/translate-a64.c | 11 ----------- target/arm/tcg/translate.c | 12 ------------ target/avr/translate.c | 8 -------- target/cris/translate.c | 11 ----------- target/hexagon/translate.c | 9 --------- target/hppa/translate.c | 6 ++++-- target/i386/tcg/translate.c | 11 ----------- target/loongarch/tcg/translate.c | 8 -------- target/m68k/translate.c | 9 --------- target/microblaze/translate.c | 9 --------- target/mips/tcg/translate.c | 9 --------- target/openrisc/translate.c | 11 ----------- target/ppc/translate.c | 9 --------- target/riscv/translate.c | 18 ------------------ target/rx/translate.c | 8 -------- target/sh4/translate.c | 9 --------- target/sparc/translate.c | 9 --------- target/tricore/translate.c | 9 --------- target/xtensa/translate.c | 9 --------- 21 files changed, 12 insertions(+), 191 deletions(-) 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 @@ #include "exec/cpu_ldst.h" #include "tcg/tcg-op-common.h" #include "internal-target.h" +#include "disas/disas.h" static void set_can_do_io(DisasContextBase *db, bool val) { @@ -XXX,XX +XXX,XX @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, FILE *logfile = qemu_log_trylock(); if (logfile) { fprintf(logfile, "----------------\n"); - ops->disas_log(db, cpu, logfile); + + if (ops->disas_log) { + ops->disas_log(db, cpu, logfile); + } else { + fprintf(logfile, "IN: %s\n", lookup_symbol(db->pc_first)); + target_disas(logfile, cpu, db->pc_first, db->tb->size); + } fprintf(logfile, "\n"); qemu_log_unlock(logfile); } 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 @@ #include "qemu/osdep.h" #include "cpu.h" #include "sysemu/cpus.h" -#include "disas/disas.h" #include "qemu/host-utils.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" @@ -XXX,XX +XXX,XX @@ static void alpha_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void alpha_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps alpha_tr_ops = { .init_disas_context = alpha_tr_init_disas_context, .tb_start = alpha_tr_tb_start, .insn_start = alpha_tr_insn_start, .translate_insn = alpha_tr_translate_insn, .tb_stop = alpha_tr_tb_stop, - .disas_log = alpha_tr_disas_log, }; void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int *max_insns, diff --git a/target/arm/tcg/translate-a64.c b/target/arm/tcg/translate-a64.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate-a64.c +++ b/target/arm/tcg/translate-a64.c @@ -XXX,XX +XXX,XX @@ #include "translate.h" #include "translate-a64.h" #include "qemu/log.h" -#include "disas/disas.h" #include "arm_ldst.h" #include "semihosting/semihost.h" #include "cpregs.h" @@ -XXX,XX +XXX,XX @@ static void aarch64_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void aarch64_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - DisasContext *dc = container_of(dcbase, DisasContext, base); - - fprintf(logfile, "IN: %s\n", lookup_symbol(dc->base.pc_first)); - target_disas(logfile, cpu, dc->base.pc_first, dc->base.tb->size); -} - const TranslatorOps aarch64_translator_ops = { .init_disas_context = aarch64_tr_init_disas_context, .tb_start = aarch64_tr_tb_start, .insn_start = aarch64_tr_insn_start, .translate_insn = aarch64_tr_translate_insn, .tb_stop = aarch64_tr_tb_stop, - .disas_log = aarch64_tr_disas_log, }; diff --git a/target/arm/tcg/translate.c b/target/arm/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/tcg/translate.c +++ b/target/arm/tcg/translate.c @@ -XXX,XX +XXX,XX @@ #include "translate.h" #include "translate-a32.h" #include "qemu/log.h" -#include "disas/disas.h" #include "arm_ldst.h" #include "semihosting/semihost.h" #include "cpregs.h" @@ -XXX,XX +XXX,XX @@ static void arm_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void arm_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - DisasContext *dc = container_of(dcbase, DisasContext, base); - - fprintf(logfile, "IN: %s\n", lookup_symbol(dc->base.pc_first)); - target_disas(logfile, cpu, dc->base.pc_first, dc->base.tb->size); -} - static const TranslatorOps arm_translator_ops = { .init_disas_context = arm_tr_init_disas_context, .tb_start = arm_tr_tb_start, .insn_start = arm_tr_insn_start, .translate_insn = arm_tr_translate_insn, .tb_stop = arm_tr_tb_stop, - .disas_log = arm_tr_disas_log, }; static const TranslatorOps thumb_translator_ops = { @@ -XXX,XX +XXX,XX @@ static const TranslatorOps thumb_translator_ops = { .insn_start = arm_tr_insn_start, .translate_insn = thumb_tr_translate_insn, .tb_stop = arm_tr_tb_stop, - .disas_log = arm_tr_disas_log, }; /* generate intermediate code for basic block '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 void avr_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void avr_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cs, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps avr_tr_ops = { .init_disas_context = avr_tr_init_disas_context, .tb_start = avr_tr_tb_start, .insn_start = avr_tr_insn_start, .translate_insn = avr_tr_translate_insn, .tb_stop = avr_tr_tb_stop, - .disas_log = avr_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "exec/helper-proto.h" @@ -XXX,XX +XXX,XX @@ static void cris_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void cris_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - if (!DISAS_CRIS) { - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); - } -} - static const TranslatorOps cris_tr_ops = { .init_disas_context = cris_tr_init_disas_context, .tb_start = cris_tr_tb_start, .insn_start = cris_tr_insn_start, .translate_insn = cris_tr_translate_insn, .tb_stop = cris_tr_tb_stop, - .disas_log = cris_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 void hexagon_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void hexagon_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - - static const TranslatorOps hexagon_tr_ops = { .init_disas_context = hexagon_tr_init_disas_context, .tb_start = hexagon_tr_tb_start, .insn_start = hexagon_tr_insn_start, .translate_insn = hexagon_tr_translate_packet, .tb_stop = hexagon_tr_tb_stop, - .disas_log = hexagon_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 void hppa_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } +#ifdef CONFIG_USER_ONLY static void hppa_tr_disas_log(const DisasContextBase *dcbase, CPUState *cs, FILE *logfile) { target_ulong pc = dcbase->pc_first; -#ifdef CONFIG_USER_ONLY switch (pc) { case 0x00: fprintf(logfile, "IN:\n0x00000000: (null)\n"); @@ -XXX,XX +XXX,XX @@ static void hppa_tr_disas_log(const DisasContextBase *dcbase, fprintf(logfile, "IN:\n0x00000100: syscall\n"); return; } -#endif fprintf(logfile, "IN: %s\n", lookup_symbol(pc)); target_disas(logfile, cs, pc, dcbase->tb->size); } +#endif static const TranslatorOps hppa_tr_ops = { .init_disas_context = hppa_tr_init_disas_context, @@ -XXX,XX +XXX,XX @@ static const TranslatorOps hppa_tr_ops = { .insn_start = hppa_tr_insn_start, .translate_insn = hppa_tr_translate_insn, .tb_stop = hppa_tr_tb_stop, +#ifdef CONFIG_USER_ONLY .disas_log = hppa_tr_disas_log, +#endif }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/host-utils.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "tcg/tcg-op-gvec.h" @@ -XXX,XX +XXX,XX @@ static void i386_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void i386_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - DisasContext *dc = container_of(dcbase, DisasContext, base); - - fprintf(logfile, "IN: %s\n", lookup_symbol(dc->base.pc_first)); - target_disas(logfile, cpu, dc->base.pc_first, dc->base.tb->size); -} - static const TranslatorOps i386_tr_ops = { .init_disas_context = i386_tr_init_disas_context, .tb_start = i386_tr_tb_start, .insn_start = i386_tr_insn_start, .translate_insn = i386_tr_translate_insn, .tb_stop = i386_tr_tb_stop, - .disas_log = i386_tr_disas_log, }; /* generate intermediate code for basic block 'tb'. */ diff --git a/target/loongarch/tcg/translate.c b/target/loongarch/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/loongarch/tcg/translate.c +++ b/target/loongarch/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static void loongarch_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void loongarch_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - qemu_log("IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps loongarch_tr_ops = { .init_disas_context = loongarch_tr_init_disas_context, .tb_start = loongarch_tr_tb_start, .insn_start = loongarch_tr_insn_start, .translate_insn = loongarch_tr_translate_insn, .tb_stop = loongarch_tr_tb_stop, - .disas_log = loongarch_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "qemu/log.h" @@ -XXX,XX +XXX,XX @@ static void m68k_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void m68k_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps m68k_tr_ops = { .init_disas_context = m68k_tr_init_disas_context, .tb_start = m68k_tr_tb_start, .insn_start = m68k_tr_insn_start, .translate_insn = m68k_tr_translate_insn, .tb_stop = m68k_tr_tb_stop, - .disas_log = m68k_tr_disas_log, }; void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "exec/cpu_ldst.h" #include "tcg/tcg-op.h" @@ -XXX,XX +XXX,XX @@ static void mb_tr_tb_stop(DisasContextBase *dcb, CPUState *cs) } } -static void mb_tr_disas_log(const DisasContextBase *dcb, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcb->pc_first)); - target_disas(logfile, cs, dcb->pc_first, dcb->tb->size); -} - static const TranslatorOps mb_tr_ops = { .init_disas_context = mb_tr_init_disas_context, .tb_start = mb_tr_tb_start, .insn_start = mb_tr_insn_start, .translate_insn = mb_tr_translate_insn, .tb_stop = mb_tr_tb_stop, - .disas_log = mb_tr_disas_log, }; void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int *max_insns, 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 @@ #include "exec/translation-block.h" #include "semihosting/semihost.h" #include "trace.h" -#include "disas/disas.h" #include "fpu_helper.h" #define HELPER_H "helper.h" @@ -XXX,XX +XXX,XX @@ static void mips_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void mips_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cs, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps mips_tr_ops = { .init_disas_context = mips_tr_init_disas_context, .tb_start = mips_tr_tb_start, .insn_start = mips_tr_insn_start, .translate_insn = mips_tr_translate_insn, .tb_stop = mips_tr_tb_stop, - .disas_log = mips_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" #include "exec/exec-all.h" -#include "disas/disas.h" #include "tcg/tcg-op.h" #include "qemu/log.h" #include "qemu/bitops.h" @@ -XXX,XX +XXX,XX @@ static void openrisc_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void openrisc_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - DisasContext *s = container_of(dcbase, DisasContext, base); - - fprintf(logfile, "IN: %s\n", lookup_symbol(s->base.pc_first)); - target_disas(logfile, cs, s->base.pc_first, s->base.tb->size); -} - static const TranslatorOps openrisc_tr_ops = { .init_disas_context = openrisc_tr_init_disas_context, .tb_start = openrisc_tr_tb_start, .insn_start = openrisc_tr_insn_start, .translate_insn = openrisc_tr_translate_insn, .tb_stop = openrisc_tr_tb_stop, - .disas_log = openrisc_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" #include "internal.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "tcg/tcg-op-gvec.h" @@ -XXX,XX +XXX,XX @@ static void ppc_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void ppc_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cs, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps ppc_tr_ops = { .init_disas_context = ppc_tr_init_disas_context, .tb_start = ppc_tr_tb_start, .insn_start = ppc_tr_insn_start, .translate_insn = ppc_tr_translate_insn, .tb_stop = ppc_tr_tb_stop, - .disas_log = ppc_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/log.h" #include "cpu.h" #include "tcg/tcg-op.h" -#include "disas/disas.h" #include "exec/cpu_ldst.h" #include "exec/exec-all.h" #include "exec/helper-proto.h" @@ -XXX,XX +XXX,XX @@ static void riscv_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void riscv_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ -#ifndef CONFIG_USER_ONLY - RISCVCPU *rvcpu = RISCV_CPU(cpu); - CPURISCVState *env = &rvcpu->env; -#endif - - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); -#ifndef CONFIG_USER_ONLY - fprintf(logfile, "Priv: "TARGET_FMT_ld"; Virt: %d\n", - env->priv, env->virt_enabled); -#endif - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps riscv_tr_ops = { .init_disas_context = riscv_tr_init_disas_context, .tb_start = riscv_tr_tb_start, .insn_start = riscv_tr_insn_start, .translate_insn = riscv_tr_translate_insn, .tb_stop = riscv_tr_tb_stop, - .disas_log = riscv_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 void rx_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void rx_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cs, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps rx_tr_ops = { .init_disas_context = rx_tr_init_disas_context, .tb_start = rx_tr_tb_start, .insn_start = rx_tr_insn_start, .translate_insn = rx_tr_translate_insn, .tb_stop = rx_tr_tb_stop, - .disas_log = rx_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "exec/helper-proto.h" @@ -XXX,XX +XXX,XX @@ static void sh4_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void sh4_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cs, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cs, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps sh4_tr_ops = { .init_disas_context = sh4_tr_init_disas_context, .tb_start = sh4_tr_tb_start, .insn_start = sh4_tr_insn_start, .translate_insn = sh4_tr_translate_insn, .tb_stop = sh4_tr_tb_stop, - .disas_log = sh4_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/helper-proto.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" @@ -XXX,XX +XXX,XX @@ static void sparc_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void sparc_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps sparc_tr_ops = { .init_disas_context = sparc_tr_init_disas_context, .tb_start = sparc_tr_tb_start, .insn_start = sparc_tr_insn_start, .translate_insn = sparc_tr_translate_insn, .tb_stop = sparc_tr_tb_stop, - .disas_log = sparc_tr_disas_log, }; void gen_intermediate_code(CPUState *cs, TranslationBlock *tb, int *max_insns, 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "exec/cpu_ldst.h" @@ -XXX,XX +XXX,XX @@ static void tricore_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void tricore_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps tricore_tr_ops = { .init_disas_context = tricore_tr_init_disas_context, .tb_start = tricore_tr_tb_start, .insn_start = tricore_tr_insn_start, .translate_insn = tricore_tr_translate_insn, .tb_stop = tricore_tr_tb_stop, - .disas_log = tricore_tr_disas_log, }; 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 @@ #include "cpu.h" #include "exec/exec-all.h" -#include "disas/disas.h" #include "tcg/tcg-op.h" #include "qemu/log.h" #include "qemu/qemu-print.h" @@ -XXX,XX +XXX,XX @@ static void xtensa_tr_tb_stop(DisasContextBase *dcbase, CPUState *cpu) } } -static void xtensa_tr_disas_log(const DisasContextBase *dcbase, - CPUState *cpu, FILE *logfile) -{ - fprintf(logfile, "IN: %s\n", lookup_symbol(dcbase->pc_first)); - target_disas(logfile, cpu, dcbase->pc_first, dcbase->tb->size); -} - static const TranslatorOps xtensa_translator_ops = { .init_disas_context = xtensa_tr_init_disas_context, .tb_start = xtensa_tr_tb_start, .insn_start = xtensa_tr_insn_start, .translate_insn = xtensa_tr_translate_insn, .tb_stop = xtensa_tr_tb_stop, - .disas_log = xtensa_tr_disas_log, }; void gen_intermediate_code(CPUState *cpu, TranslationBlock *tb, int *max_insns, -- 2.34.1 We have eliminated most uses of this hook. Reduce further by allowing the hook to handle only the special cases, returning false for normal processing. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 2 +- accel/tcg/translator.c | 5 ++--- target/hppa/translate.c | 15 ++++++--------- target/s390x/tcg/translate.c | 8 +++----- 4 files changed, 12 insertions(+), 18 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 struct TranslatorOps { void (*insn_start)(DisasContextBase *db, CPUState *cpu); void (*translate_insn)(DisasContextBase *db, CPUState *cpu); void (*tb_stop)(DisasContextBase *db, CPUState *cpu); - void (*disas_log)(const DisasContextBase *db, CPUState *cpu, FILE *f); + bool (*disas_log)(const DisasContextBase *db, CPUState *cpu, FILE *f); } TranslatorOps; /** 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, if (logfile) { fprintf(logfile, "----------------\n"); - if (ops->disas_log) { - ops->disas_log(db, cpu, logfile); - } else { + if (!ops->disas_log || + !ops->disas_log(db, cpu, logfile)) { fprintf(logfile, "IN: %s\n", lookup_symbol(db->pc_first)); target_disas(logfile, cpu, db->pc_first, db->tb->size); } 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 @@ #include "qemu/osdep.h" #include "cpu.h" -#include "disas/disas.h" #include "qemu/host-utils.h" #include "exec/exec-all.h" #include "exec/page-protection.h" @@ -XXX,XX +XXX,XX @@ static void hppa_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } #ifdef CONFIG_USER_ONLY -static void hppa_tr_disas_log(const DisasContextBase *dcbase, +static bool hppa_tr_disas_log(const DisasContextBase *dcbase, CPUState *cs, FILE *logfile) { target_ulong pc = dcbase->pc_first; @@ -XXX,XX +XXX,XX @@ static void hppa_tr_disas_log(const DisasContextBase *dcbase, switch (pc) { case 0x00: fprintf(logfile, "IN:\n0x00000000: (null)\n"); - return; + return true; case 0xb0: fprintf(logfile, "IN:\n0x000000b0: light-weight-syscall\n"); - return; + return true; case 0xe0: fprintf(logfile, "IN:\n0x000000e0: set-thread-pointer-syscall\n"); - return; + return true; case 0x100: fprintf(logfile, "IN:\n0x00000100: syscall\n"); - return; + return true; } - - fprintf(logfile, "IN: %s\n", lookup_symbol(pc)); - target_disas(logfile, cs, pc, dcbase->tb->size); + return false; } #endif 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 @@ #include "qemu/osdep.h" #include "cpu.h" #include "s390x-internal.h" -#include "disas/disas.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "tcg/tcg-op-gvec.h" @@ -XXX,XX +XXX,XX @@ static void s390x_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs) } } -static void s390x_tr_disas_log(const DisasContextBase *dcbase, +static bool s390x_tr_disas_log(const DisasContextBase *dcbase, CPUState *cs, FILE *logfile) { DisasContext *dc = container_of(dcbase, DisasContext, base); @@ -XXX,XX +XXX,XX @@ static void s390x_tr_disas_log(const DisasContextBase *dcbase, if (unlikely(dc->ex_value)) { /* ??? Unfortunately target_disas can't use host memory. */ fprintf(logfile, "IN: EXECUTE %016" PRIx64, dc->ex_value); - } else { - fprintf(logfile, "IN: %s\n", lookup_symbol(dc->base.pc_first)); - target_disas(logfile, cs, dc->base.pc_first, dc->base.tb->size); + return true; } + return false; } static const TranslatorOps s390x_tr_ops = { -- 2.34.1 The routines in disas-common.c are also used from disas-mon.c. Otherwise the rest of disassembly is only used from tcg. While we're at it, put host and target code into separate files. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- disas/disas-internal.h | 4 + include/disas/disas.h | 4 + disas/disas-common.c | 118 ++++++++++++++ disas/disas-host.c | 129 ++++++++++++++++ disas/disas-target.c | 84 ++++++++++ disas/disas.c | 338 ----------------------------------------- disas/objdump.c | 37 +++++ disas/meson.build | 8 +- 8 files changed, 382 insertions(+), 340 deletions(-) create mode 100644 disas/disas-common.c create mode 100644 disas/disas-host.c create mode 100644 disas/disas-target.c delete mode 100644 disas/disas.c create mode 100644 disas/objdump.c diff --git a/disas/disas-internal.h b/disas/disas-internal.h index XXXXXXX..XXXXXXX 100644 --- a/disas/disas-internal.h +++ b/disas/disas-internal.h @@ -XXX,XX +XXX,XX @@ typedef struct CPUDebug { CPUState *cpu; } CPUDebug; +void disas_initialize_debug(CPUDebug *s); void disas_initialize_debug_target(CPUDebug *s, CPUState *cpu); int disas_gstring_printf(FILE *stream, const char *fmt, ...) G_GNUC_PRINTF(2, 3); +int print_insn_od_host(bfd_vma pc, disassemble_info *info); +int print_insn_od_target(bfd_vma pc, disassemble_info *info); + #endif diff --git a/include/disas/disas.h b/include/disas/disas.h index XXXXXXX..XXXXXXX 100644 --- a/include/disas/disas.h +++ b/include/disas/disas.h @@ -XXX,XX +XXX,XX @@ #define QEMU_DISAS_H /* Disassemble this for me please... (debugging). */ +#ifdef CONFIG_TCG void disas(FILE *out, const void *code, size_t size); void target_disas(FILE *out, CPUState *cpu, uint64_t code, size_t size); +#endif void monitor_disas(Monitor *mon, CPUState *cpu, uint64_t pc, int nb_insn, bool is_physical); +#ifdef CONFIG_PLUGIN char *plugin_disas(CPUState *cpu, uint64_t addr, size_t size); +#endif /* Look up symbol for debugging purpose. Returns "" if unknown. */ const char *lookup_symbol(uint64_t orig_addr); diff --git a/disas/disas-common.c b/disas/disas-common.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/disas/disas-common.c @@ -XXX,XX +XXX,XX @@ +/* + * Common routines for disassembly. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "disas/disas.h" +#include "disas/capstone.h" +#include "hw/core/cpu.h" +#include "exec/tswap.h" +#include "exec/memory.h" +#include "disas-internal.h" + + +/* Filled in by elfload.c. Simplistic, but will do for now. */ +struct syminfo *syminfos = NULL; + +/* + * Get LENGTH bytes from info's buffer, at target address memaddr. + * Transfer them to myaddr. + */ +static int target_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, + struct disassemble_info *info) +{ + CPUDebug *s = container_of(info, CPUDebug, info); + int r = cpu_memory_rw_debug(s->cpu, memaddr, myaddr, length, 0); + return r ? EIO : 0; +} + +/* + * Print an error message. We can assume that this is in response to + * an error return from {host,target}_read_memory. + */ +static void perror_memory(int status, bfd_vma memaddr, + struct disassemble_info *info) +{ + if (status != EIO) { + /* Can't happen. */ + info->fprintf_func(info->stream, "Unknown error %d\n", status); + } else { + /* Address between memaddr and memaddr + len was out of bounds. */ + info->fprintf_func(info->stream, + "Address 0x%" PRIx64 " is out of bounds.\n", + memaddr); + } +} + +/* Print address in hex. */ +static void print_address(bfd_vma addr, struct disassemble_info *info) +{ + info->fprintf_func(info->stream, "0x%" PRIx64, addr); +} + +/* Stub prevents some fruitless earching in optabs disassemblers. */ +static int symbol_at_address(bfd_vma addr, struct disassemble_info *info) +{ + return 1; +} + +void disas_initialize_debug(CPUDebug *s) +{ + memset(s, 0, sizeof(*s)); + s->info.arch = bfd_arch_unknown; + s->info.cap_arch = -1; + s->info.cap_insn_unit = 4; + s->info.cap_insn_split = 4; + s->info.memory_error_func = perror_memory; + s->info.symbol_at_address_func = symbol_at_address; +} + +void disas_initialize_debug_target(CPUDebug *s, CPUState *cpu) +{ + disas_initialize_debug(s); + + s->cpu = cpu; + s->info.read_memory_func = target_read_memory; + s->info.print_address_func = print_address; + if (target_words_bigendian()) { + s->info.endian = BFD_ENDIAN_BIG; + } else { + s->info.endian = BFD_ENDIAN_LITTLE; + } + + CPUClass *cc = CPU_GET_CLASS(cpu); + if (cc->disas_set_info) { + cc->disas_set_info(cpu, &s->info); + } +} + +int disas_gstring_printf(FILE *stream, const char *fmt, ...) +{ + /* We abuse the FILE parameter to pass a GString. */ + GString *s = (GString *)stream; + int initial_len = s->len; + va_list va; + + va_start(va, fmt); + g_string_append_vprintf(s, fmt, va); + va_end(va); + + return s->len - initial_len; +} + +/* Look up symbol for debugging purpose. Returns "" if unknown. */ +const char *lookup_symbol(uint64_t orig_addr) +{ + const char *symbol = ""; + struct syminfo *s; + + for (s = syminfos; s; s = s->next) { + symbol = s->lookup_symbol(s, orig_addr); + if (symbol[0] != '\0') { + break; + } + } + + return symbol; +} diff --git a/disas/disas-host.c b/disas/disas-host.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/disas/disas-host.c @@ -XXX,XX +XXX,XX @@ +/* + * Routines for host instruction disassembly. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "disas/disas.h" +#include "disas/capstone.h" +#include "disas-internal.h" + + +/* + * Get LENGTH bytes from info's buffer, at host address memaddr. + * Transfer them to myaddr. + */ +static int host_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, + struct disassemble_info *info) +{ + if (memaddr < info->buffer_vma + || memaddr + length > info->buffer_vma + info->buffer_length) { + /* Out of bounds. Use EIO because GDB uses it. */ + return EIO; + } + memcpy (myaddr, info->buffer + (memaddr - info->buffer_vma), length); + return 0; +} + +/* Print address in hex, truncated to the width of a host virtual address. */ +static void host_print_address(bfd_vma addr, struct disassemble_info *info) +{ + info->fprintf_func(info->stream, "0x%" PRIxPTR, (uintptr_t)addr); +} + +static void initialize_debug_host(CPUDebug *s) +{ + disas_initialize_debug(s); + + s->info.read_memory_func = host_read_memory; + s->info.print_address_func = host_print_address; +#if HOST_BIG_ENDIAN + s->info.endian = BFD_ENDIAN_BIG; +#else + s->info.endian = BFD_ENDIAN_LITTLE; +#endif +#if defined(CONFIG_TCG_INTERPRETER) + s->info.print_insn = print_insn_tci; +#elif defined(__i386__) + s->info.mach = bfd_mach_i386_i386; + s->info.cap_arch = CS_ARCH_X86; + s->info.cap_mode = CS_MODE_32; + s->info.cap_insn_unit = 1; + s->info.cap_insn_split = 8; +#elif defined(__x86_64__) + s->info.mach = bfd_mach_x86_64; + s->info.cap_arch = CS_ARCH_X86; + s->info.cap_mode = CS_MODE_64; + s->info.cap_insn_unit = 1; + s->info.cap_insn_split = 8; +#elif defined(_ARCH_PPC) + s->info.cap_arch = CS_ARCH_PPC; +# ifdef _ARCH_PPC64 + s->info.cap_mode = CS_MODE_64; +# endif +#elif defined(__riscv) +#if defined(_ILP32) || (__riscv_xlen == 32) + s->info.print_insn = print_insn_riscv32; +#elif defined(_LP64) + s->info.print_insn = print_insn_riscv64; +#else +#error unsupported RISC-V ABI +#endif +#elif defined(__aarch64__) + s->info.cap_arch = CS_ARCH_ARM64; +#elif defined(__alpha__) + s->info.print_insn = print_insn_alpha; +#elif defined(__sparc__) + s->info.print_insn = print_insn_sparc; + s->info.mach = bfd_mach_sparc_v9b; +#elif defined(__arm__) + /* TCG only generates code for arm mode. */ + s->info.cap_arch = CS_ARCH_ARM; +#elif defined(__MIPSEB__) + s->info.print_insn = print_insn_big_mips; +#elif defined(__MIPSEL__) + s->info.print_insn = print_insn_little_mips; +#elif defined(__m68k__) + s->info.print_insn = print_insn_m68k; +#elif defined(__s390__) + s->info.cap_arch = CS_ARCH_SYSZ; + s->info.cap_insn_unit = 2; + s->info.cap_insn_split = 6; +#elif defined(__hppa__) + s->info.print_insn = print_insn_hppa; +#elif defined(__loongarch__) + s->info.print_insn = print_insn_loongarch; +#endif +} + +/* Disassemble this for me please... (debugging). */ +void disas(FILE *out, const void *code, size_t size) +{ + uintptr_t pc; + int count; + CPUDebug s; + + initialize_debug_host(&s); + s.info.fprintf_func = fprintf; + s.info.stream = out; + s.info.buffer = code; + s.info.buffer_vma = (uintptr_t)code; + s.info.buffer_length = size; + s.info.show_opcodes = true; + + if (s.info.cap_arch >= 0 && cap_disas_host(&s.info, code, size)) { + return; + } + + if (s.info.print_insn == NULL) { + s.info.print_insn = print_insn_od_host; + } + for (pc = (uintptr_t)code; size > 0; pc += count, size -= count) { + fprintf(out, "0x%08" PRIxPTR ": ", pc); + count = s.info.print_insn(pc, &s.info); + fprintf(out, "\n"); + if (count < 0) { + break; + } + } +} diff --git a/disas/disas-target.c b/disas/disas-target.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/disas/disas-target.c @@ -XXX,XX +XXX,XX @@ +/* + * Routines for target instruction disassembly. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "disas/disas.h" +#include "disas/capstone.h" +#include "disas-internal.h" + + +void target_disas(FILE *out, CPUState *cpu, uint64_t code, size_t size) +{ + uint64_t pc; + int count; + CPUDebug s; + + disas_initialize_debug_target(&s, cpu); + s.info.fprintf_func = fprintf; + s.info.stream = out; + s.info.buffer_vma = code; + s.info.buffer_length = size; + s.info.show_opcodes = true; + + if (s.info.cap_arch >= 0 && cap_disas_target(&s.info, code, size)) { + return; + } + + if (s.info.print_insn == NULL) { + s.info.print_insn = print_insn_od_target; + } + + for (pc = code; size > 0; pc += count, size -= count) { + fprintf(out, "0x%08" PRIx64 ": ", pc); + count = s.info.print_insn(pc, &s.info); + fprintf(out, "\n"); + if (count < 0) { + break; + } + if (size < count) { + fprintf(out, + "Disassembler disagrees with translator over instruction " + "decoding\n" + "Please report this to qemu-devel@nongnu.org\n"); + break; + } + } +} + +#ifdef CONFIG_PLUGIN +static void plugin_print_address(bfd_vma addr, struct disassemble_info *info) +{ + /* does nothing */ +} + +/* + * We should only be dissembling one instruction at a time here. If + * there is left over it usually indicates the front end has read more + * bytes than it needed. + */ +char *plugin_disas(CPUState *cpu, uint64_t addr, size_t size) +{ + CPUDebug s; + GString *ds = g_string_new(NULL); + + disas_initialize_debug_target(&s, cpu); + s.info.fprintf_func = disas_gstring_printf; + s.info.stream = (FILE *)ds; /* abuse this slot */ + s.info.buffer_vma = addr; + s.info.buffer_length = size; + s.info.print_address_func = plugin_print_address; + + if (s.info.cap_arch >= 0 && cap_disas_plugin(&s.info, addr, size)) { + ; /* done */ + } else if (s.info.print_insn) { + s.info.print_insn(addr, &s.info); + } else { + ; /* cannot disassemble -- return empty string */ + } + + /* Return the buffer, freeing the GString container. */ + return g_string_free(ds, false); +} +#endif /* CONFIG_PLUGIN */ diff --git a/disas/disas.c b/disas/disas.c deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/disas/disas.c +++ /dev/null @@ -XXX,XX +XXX,XX @@ -/* General "disassemble this chunk" code. Used for debugging. */ -#include "qemu/osdep.h" -#include "disas/disas-internal.h" -#include "elf.h" -#include "qemu/qemu-print.h" -#include "disas/disas.h" -#include "disas/capstone.h" -#include "hw/core/cpu.h" -#include "exec/tswap.h" -#include "exec/memory.h" - -/* Filled in by elfload.c. Simplistic, but will do for now. */ -struct syminfo *syminfos = NULL; - -/* - * Get LENGTH bytes from info's buffer, at host address memaddr. - * Transfer them to myaddr. - */ -static int host_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, - struct disassemble_info *info) -{ - if (memaddr < info->buffer_vma - || memaddr + length > info->buffer_vma + info->buffer_length) { - /* Out of bounds. Use EIO because GDB uses it. */ - return EIO; - } - memcpy (myaddr, info->buffer + (memaddr - info->buffer_vma), length); - return 0; -} - -/* - * Get LENGTH bytes from info's buffer, at target address memaddr. - * Transfer them to myaddr. - */ -static int target_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, - struct disassemble_info *info) -{ - CPUDebug *s = container_of(info, CPUDebug, info); - int r = cpu_memory_rw_debug(s->cpu, memaddr, myaddr, length, 0); - return r ? EIO : 0; -} - -/* - * Print an error message. We can assume that this is in response to - * an error return from {host,target}_read_memory. - */ -static void perror_memory(int status, bfd_vma memaddr, - struct disassemble_info *info) -{ - if (status != EIO) { - /* Can't happen. */ - info->fprintf_func(info->stream, "Unknown error %d\n", status); - } else { - /* Address between memaddr and memaddr + len was out of bounds. */ - info->fprintf_func(info->stream, - "Address 0x%" PRIx64 " is out of bounds.\n", - memaddr); - } -} - -/* Print address in hex. */ -static void print_address(bfd_vma addr, struct disassemble_info *info) -{ - info->fprintf_func(info->stream, "0x%" PRIx64, addr); -} - -/* Print address in hex, truncated to the width of a host virtual address. */ -static void host_print_address(bfd_vma addr, struct disassemble_info *info) -{ - print_address((uintptr_t)addr, info); -} - -/* Stub prevents some fruitless earching in optabs disassemblers. */ -static int symbol_at_address(bfd_vma addr, struct disassemble_info *info) -{ - return 1; -} - -static int print_insn_objdump(bfd_vma pc, disassemble_info *info, - const char *prefix) -{ - int i, n = info->buffer_length; - g_autofree uint8_t *buf = g_malloc(n); - - if (info->read_memory_func(pc, buf, n, info) == 0) { - for (i = 0; i < n; ++i) { - if (i % 32 == 0) { - info->fprintf_func(info->stream, "\n%s: ", prefix); - } - info->fprintf_func(info->stream, "%02x", buf[i]); - } - } else { - info->fprintf_func(info->stream, "unable to read memory"); - } - return n; -} - -static int print_insn_od_host(bfd_vma pc, disassemble_info *info) -{ - return print_insn_objdump(pc, info, "OBJD-H"); -} - -static int print_insn_od_target(bfd_vma pc, disassemble_info *info) -{ - return print_insn_objdump(pc, info, "OBJD-T"); -} - -static void initialize_debug(CPUDebug *s) -{ - memset(s, 0, sizeof(*s)); - s->info.arch = bfd_arch_unknown; - s->info.cap_arch = -1; - s->info.cap_insn_unit = 4; - s->info.cap_insn_split = 4; - s->info.memory_error_func = perror_memory; - s->info.symbol_at_address_func = symbol_at_address; -} - -void disas_initialize_debug_target(CPUDebug *s, CPUState *cpu) -{ - initialize_debug(s); - - s->cpu = cpu; - s->info.read_memory_func = target_read_memory; - s->info.print_address_func = print_address; - if (target_words_bigendian()) { - s->info.endian = BFD_ENDIAN_BIG; - } else { - s->info.endian = BFD_ENDIAN_LITTLE; - } - - CPUClass *cc = CPU_GET_CLASS(cpu); - if (cc->disas_set_info) { - cc->disas_set_info(cpu, &s->info); - } -} - -static void initialize_debug_host(CPUDebug *s) -{ - initialize_debug(s); - - s->info.read_memory_func = host_read_memory; - s->info.print_address_func = host_print_address; -#if HOST_BIG_ENDIAN - s->info.endian = BFD_ENDIAN_BIG; -#else - s->info.endian = BFD_ENDIAN_LITTLE; -#endif -#if defined(CONFIG_TCG_INTERPRETER) - s->info.print_insn = print_insn_tci; -#elif defined(__i386__) - s->info.mach = bfd_mach_i386_i386; - s->info.cap_arch = CS_ARCH_X86; - s->info.cap_mode = CS_MODE_32; - s->info.cap_insn_unit = 1; - s->info.cap_insn_split = 8; -#elif defined(__x86_64__) - s->info.mach = bfd_mach_x86_64; - s->info.cap_arch = CS_ARCH_X86; - s->info.cap_mode = CS_MODE_64; - s->info.cap_insn_unit = 1; - s->info.cap_insn_split = 8; -#elif defined(_ARCH_PPC) - s->info.cap_arch = CS_ARCH_PPC; -# ifdef _ARCH_PPC64 - s->info.cap_mode = CS_MODE_64; -# endif -#elif defined(__riscv) -#if defined(_ILP32) || (__riscv_xlen == 32) - s->info.print_insn = print_insn_riscv32; -#elif defined(_LP64) - s->info.print_insn = print_insn_riscv64; -#else -#error unsupported RISC-V ABI -#endif -#elif defined(__aarch64__) - s->info.cap_arch = CS_ARCH_ARM64; -#elif defined(__alpha__) - s->info.print_insn = print_insn_alpha; -#elif defined(__sparc__) - s->info.print_insn = print_insn_sparc; - s->info.mach = bfd_mach_sparc_v9b; -#elif defined(__arm__) - /* TCG only generates code for arm mode. */ - s->info.cap_arch = CS_ARCH_ARM; -#elif defined(__MIPSEB__) - s->info.print_insn = print_insn_big_mips; -#elif defined(__MIPSEL__) - s->info.print_insn = print_insn_little_mips; -#elif defined(__m68k__) - s->info.print_insn = print_insn_m68k; -#elif defined(__s390__) - s->info.cap_arch = CS_ARCH_SYSZ; - s->info.cap_insn_unit = 2; - s->info.cap_insn_split = 6; -#elif defined(__hppa__) - s->info.print_insn = print_insn_hppa; -#elif defined(__loongarch__) - s->info.print_insn = print_insn_loongarch; -#endif -} - -/* Disassemble this for me please... (debugging). */ -void target_disas(FILE *out, CPUState *cpu, uint64_t code, size_t size) -{ - uint64_t pc; - int count; - CPUDebug s; - - disas_initialize_debug_target(&s, cpu); - s.info.fprintf_func = fprintf; - s.info.stream = out; - s.info.buffer_vma = code; - s.info.buffer_length = size; - s.info.show_opcodes = true; - - if (s.info.cap_arch >= 0 && cap_disas_target(&s.info, code, size)) { - return; - } - - if (s.info.print_insn == NULL) { - s.info.print_insn = print_insn_od_target; - } - - for (pc = code; size > 0; pc += count, size -= count) { - fprintf(out, "0x%08" PRIx64 ": ", pc); - count = s.info.print_insn(pc, &s.info); - fprintf(out, "\n"); - if (count < 0) { - break; - } - if (size < count) { - fprintf(out, - "Disassembler disagrees with translator over instruction " - "decoding\n" - "Please report this to qemu-devel@nongnu.org\n"); - break; - } - } -} - -int disas_gstring_printf(FILE *stream, const char *fmt, ...) -{ - /* We abuse the FILE parameter to pass a GString. */ - GString *s = (GString *)stream; - int initial_len = s->len; - va_list va; - - va_start(va, fmt); - g_string_append_vprintf(s, fmt, va); - va_end(va); - - return s->len - initial_len; -} - -static void plugin_print_address(bfd_vma addr, struct disassemble_info *info) -{ - /* does nothing */ -} - - -/* - * We should only be dissembling one instruction at a time here. If - * there is left over it usually indicates the front end has read more - * bytes than it needed. - */ -char *plugin_disas(CPUState *cpu, uint64_t addr, size_t size) -{ - CPUDebug s; - GString *ds = g_string_new(NULL); - - disas_initialize_debug_target(&s, cpu); - s.info.fprintf_func = disas_gstring_printf; - s.info.stream = (FILE *)ds; /* abuse this slot */ - s.info.buffer_vma = addr; - s.info.buffer_length = size; - s.info.print_address_func = plugin_print_address; - - if (s.info.cap_arch >= 0 && cap_disas_plugin(&s.info, addr, size)) { - ; /* done */ - } else if (s.info.print_insn) { - s.info.print_insn(addr, &s.info); - } else { - ; /* cannot disassemble -- return empty string */ - } - - /* Return the buffer, freeing the GString container. */ - return g_string_free(ds, false); -} - -/* Disassemble this for me please... (debugging). */ -void disas(FILE *out, const void *code, size_t size) -{ - uintptr_t pc; - int count; - CPUDebug s; - - initialize_debug_host(&s); - s.info.fprintf_func = fprintf; - s.info.stream = out; - s.info.buffer = code; - s.info.buffer_vma = (uintptr_t)code; - s.info.buffer_length = size; - s.info.show_opcodes = true; - - if (s.info.cap_arch >= 0 && cap_disas_host(&s.info, code, size)) { - return; - } - - if (s.info.print_insn == NULL) { - s.info.print_insn = print_insn_od_host; - } - for (pc = (uintptr_t)code; size > 0; pc += count, size -= count) { - fprintf(out, "0x%08" PRIxPTR ": ", pc); - count = s.info.print_insn(pc, &s.info); - fprintf(out, "\n"); - if (count < 0) { - break; - } - } - -} - -/* Look up symbol for debugging purpose. Returns "" if unknown. */ -const char *lookup_symbol(uint64_t orig_addr) -{ - const char *symbol = ""; - struct syminfo *s; - - for (s = syminfos; s; s = s->next) { - symbol = s->lookup_symbol(s, orig_addr); - if (symbol[0] != '\0') { - break; - } - } - - return symbol; -} diff --git a/disas/objdump.c b/disas/objdump.c new file mode 100644 index XXXXXXX..XXXXXXX --- /dev/null +++ b/disas/objdump.c @@ -XXX,XX +XXX,XX @@ +/* + * Dump disassembly as text, for processing by scripts/disas-objdump.pl. + * SPDX-License-Identifier: GPL-2.0-or-later + */ + +#include "qemu/osdep.h" +#include "disas-internal.h" + + +static int print_insn_objdump(bfd_vma pc, disassemble_info *info, + const char *prefix) +{ + int i, n = info->buffer_length; + g_autofree uint8_t *buf = g_malloc(n); + + if (info->read_memory_func(pc, buf, n, info) == 0) { + for (i = 0; i < n; ++i) { + if (i % 32 == 0) { + info->fprintf_func(info->stream, "\n%s: ", prefix); + } + info->fprintf_func(info->stream, "%02x", buf[i]); + } + } else { + info->fprintf_func(info->stream, "unable to read memory"); + } + return n; +} + +int print_insn_od_host(bfd_vma pc, disassemble_info *info) +{ + return print_insn_objdump(pc, info, "OBJD-H"); +} + +int print_insn_od_target(bfd_vma pc, disassemble_info *info) +{ + return print_insn_objdump(pc, info, "OBJD-T"); +} diff --git a/disas/meson.build b/disas/meson.build index XXXXXXX..XXXXXXX 100644 --- a/disas/meson.build +++ b/disas/meson.build @@ -XXX,XX +XXX,XX @@ common_ss.add(when: 'CONFIG_SH4_DIS', if_true: files('sh4.c')) common_ss.add(when: 'CONFIG_SPARC_DIS', if_true: files('sparc.c')) common_ss.add(when: 'CONFIG_XTENSA_DIS', if_true: files('xtensa.c')) common_ss.add(when: capstone, if_true: [files('capstone.c'), capstone]) -common_ss.add(files('disas.c')) - +common_ss.add(when: 'CONFIG_TCG', if_true: files( + 'disas-host.c', + 'disas-target.c', + 'objdump.c' +)) +common_ss.add(files('disas-common.c')) system_ss.add(files('disas-mon.c')) specific_ss.add(capstone) -- 2.34.1 Read from already translated pages, or saved mmio data. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/disas/disas.h | 5 +++-- include/exec/translator.h | 4 ++-- include/qemu/typedefs.h | 1 + accel/tcg/translator.c | 2 +- disas/disas-common.c | 14 -------------- disas/disas-mon.c | 15 +++++++++++++++ disas/disas-target.c | 19 +++++++++++++++++-- plugins/api.c | 4 ++-- 8 files changed, 41 insertions(+), 23 deletions(-) diff --git a/include/disas/disas.h b/include/disas/disas.h index XXXXXXX..XXXXXXX 100644 --- a/include/disas/disas.h +++ b/include/disas/disas.h @@ -XXX,XX +XXX,XX @@ /* Disassemble this for me please... (debugging). */ #ifdef CONFIG_TCG void disas(FILE *out, const void *code, size_t size); -void target_disas(FILE *out, CPUState *cpu, uint64_t code, size_t size); +void target_disas(FILE *out, CPUState *cpu, const DisasContextBase *db); #endif void monitor_disas(Monitor *mon, CPUState *cpu, uint64_t pc, int nb_insn, bool is_physical); #ifdef CONFIG_PLUGIN -char *plugin_disas(CPUState *cpu, uint64_t addr, size_t size); +char *plugin_disas(CPUState *cpu, const DisasContextBase *db, + uint64_t addr, size_t size); #endif /* Look up symbol for debugging purpose. Returns "" if unknown. */ 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 { +struct DisasContextBase { TranslationBlock *tb; vaddr pc_first; vaddr pc_next; @@ -XXX,XX +XXX,XX @@ typedef struct DisasContextBase { int record_start; int record_len; uint8_t record[32]; -} DisasContextBase; +}; /** * TranslatorOps: diff --git a/include/qemu/typedefs.h b/include/qemu/typedefs.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/typedefs.h +++ b/include/qemu/typedefs.h @@ -XXX,XX +XXX,XX @@ typedef struct CPUPluginState CPUPluginState; typedef struct CPUState CPUState; typedef struct DeviceState DeviceState; typedef struct DirtyBitmapSnapshot DirtyBitmapSnapshot; +typedef struct DisasContextBase DisasContextBase; typedef struct DisplayChangeListener DisplayChangeListener; typedef struct DriveInfo DriveInfo; typedef struct DumpState DumpState; 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 @@ void translator_loop(CPUState *cpu, TranslationBlock *tb, int *max_insns, if (!ops->disas_log || !ops->disas_log(db, cpu, logfile)) { fprintf(logfile, "IN: %s\n", lookup_symbol(db->pc_first)); - target_disas(logfile, cpu, db->pc_first, db->tb->size); + target_disas(logfile, cpu, db); } fprintf(logfile, "\n"); qemu_log_unlock(logfile); diff --git a/disas/disas-common.c b/disas/disas-common.c index XXXXXXX..XXXXXXX 100644 --- a/disas/disas-common.c +++ b/disas/disas-common.c @@ -XXX,XX +XXX,XX @@ #include "disas/capstone.h" #include "hw/core/cpu.h" #include "exec/tswap.h" -#include "exec/memory.h" #include "disas-internal.h" /* Filled in by elfload.c. Simplistic, but will do for now. */ struct syminfo *syminfos = NULL; -/* - * Get LENGTH bytes from info's buffer, at target address memaddr. - * Transfer them to myaddr. - */ -static int target_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, - struct disassemble_info *info) -{ - CPUDebug *s = container_of(info, CPUDebug, info); - int r = cpu_memory_rw_debug(s->cpu, memaddr, myaddr, length, 0); - return r ? EIO : 0; -} - /* * Print an error message. We can assume that this is in response to * an error return from {host,target}_read_memory. @@ -XXX,XX +XXX,XX @@ void disas_initialize_debug_target(CPUDebug *s, CPUState *cpu) disas_initialize_debug(s); s->cpu = cpu; - s->info.read_memory_func = target_read_memory; s->info.print_address_func = print_address; if (target_words_bigendian()) { s->info.endian = BFD_ENDIAN_BIG; diff --git a/disas/disas-mon.c b/disas/disas-mon.c index XXXXXXX..XXXXXXX 100644 --- a/disas/disas-mon.c +++ b/disas/disas-mon.c @@ -XXX,XX +XXX,XX @@ #include "hw/core/cpu.h" #include "monitor/monitor.h" +/* + * Get LENGTH bytes from info's buffer, at target address memaddr. + * Transfer them to myaddr. + */ +static int +virtual_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, + struct disassemble_info *info) +{ + CPUDebug *s = container_of(info, CPUDebug, info); + int r = cpu_memory_rw_debug(s->cpu, memaddr, myaddr, length, 0); + return r ? EIO : 0; +} + static int physical_read_memory(bfd_vma memaddr, bfd_byte *myaddr, int length, struct disassemble_info *info) @@ -XXX,XX +XXX,XX @@ void monitor_disas(Monitor *mon, CPUState *cpu, uint64_t pc, if (is_physical) { s.info.read_memory_func = physical_read_memory; + } else { + s.info.read_memory_func = virtual_read_memory; } s.info.buffer_vma = pc; diff --git a/disas/disas-target.c b/disas/disas-target.c index XXXXXXX..XXXXXXX 100644 --- a/disas/disas-target.c +++ b/disas/disas-target.c @@ -XXX,XX +XXX,XX @@ #include "qemu/osdep.h" #include "disas/disas.h" #include "disas/capstone.h" +#include "exec/translator.h" #include "disas-internal.h" -void target_disas(FILE *out, CPUState *cpu, uint64_t code, size_t size) +static int translator_read_memory(bfd_vma memaddr, bfd_byte *myaddr, + int length, struct disassemble_info *info) { + const DisasContextBase *db = info->application_data; + return translator_st(db, myaddr, memaddr, length) ? 0 : EIO; +} + +void target_disas(FILE *out, CPUState *cpu, const struct DisasContextBase *db) +{ + uint64_t code = db->pc_first; + size_t size = translator_st_len(db); uint64_t pc; int count; CPUDebug s; disas_initialize_debug_target(&s, cpu); + s.info.read_memory_func = translator_read_memory; + s.info.application_data = (void *)db; s.info.fprintf_func = fprintf; s.info.stream = out; s.info.buffer_vma = code; @@ -XXX,XX +XXX,XX @@ static void plugin_print_address(bfd_vma addr, struct disassemble_info *info) * there is left over it usually indicates the front end has read more * bytes than it needed. */ -char *plugin_disas(CPUState *cpu, uint64_t addr, size_t size) +char *plugin_disas(CPUState *cpu, const DisasContextBase *db, + uint64_t addr, size_t size) { CPUDebug s; GString *ds = g_string_new(NULL); disas_initialize_debug_target(&s, cpu); + s.info.read_memory_func = translator_read_memory; + s.info.application_data = (void *)db; s.info.fprintf_func = disas_gstring_printf; s.info.stream = (FILE *)ds; /* abuse this slot */ s.info.buffer_vma = addr; diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ void *qemu_plugin_insn_haddr(const struct qemu_plugin_insn *insn) char *qemu_plugin_insn_disas(const struct qemu_plugin_insn *insn) { - CPUState *cpu = current_cpu; - return plugin_disas(cpu, insn->vaddr, insn->len); + return plugin_disas(tcg_ctx->cpu, tcg_ctx->plugin_db, + insn->vaddr, insn->len); } const char *qemu_plugin_insn_symbol(const struct qemu_plugin_insn *insn) -- 2.34.1 Replace translator_fake_ldb, which required multiple calls, with translator_fake_ld, which can take all data at once. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/translator.h | 8 ++++---- accel/tcg/translator.c | 5 ++--- target/s390x/tcg/translate.c | 8 ++++---- 3 files changed, 10 insertions(+), 11 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 @@ translator_ldq_swap(CPUArchState *env, DisasContextBase *db, } /** - * translator_fake_ldb - fake instruction load + * translator_fake_ld - fake instruction load * @db: Disassembly context - * @pc: program counter of instruction - * @insn8: byte of instruction + * @data: bytes of instruction + * @len: number of bytes * * This is a special case helper used where the instruction we are * about to translate comes from somewhere else (e.g. being * re-synthesised for s390x "ex"). It ensures we update other areas of * the translator with details of the executed instruction. */ -void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8); +void translator_fake_ld(DisasContextBase *db, const void *data, size_t len); /** * translator_st 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 @@ uint64_t translator_ldq(CPUArchState *env, DisasContextBase *db, vaddr pc) return tgt; } -void translator_fake_ldb(DisasContextBase *db, vaddr pc, uint8_t insn8) +void translator_fake_ld(DisasContextBase *db, const void *data, size_t len) { - assert(pc >= db->pc_first); db->fake_insn = true; - record_save(db, pc, &insn8, sizeof(insn8)); + record_save(db, db->pc_first, data, len); } 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 DisasInsn *extract_insn(CPUS390XState *env, DisasContext *s) const DisasInsn *info; if (unlikely(s->ex_value)) { + uint64_t be_insn; + /* Drop the EX data now, so that it's clear on exception paths. */ tcg_gen_st_i64(tcg_constant_i64(0), tcg_env, offsetof(CPUS390XState, ex_value)); @@ -XXX,XX +XXX,XX @@ static const DisasInsn *extract_insn(CPUS390XState *env, DisasContext *s) ilen = s->ex_value & 0xf; /* Register insn bytes with translator so plugins work. */ - for (int i = 0; i < ilen; i++) { - uint8_t byte = extract64(insn, 56 - (i * 8), 8); - translator_fake_ldb(&s->base, pc + i, byte); - } + be_insn = cpu_to_be64(insn); + translator_fake_ld(&s->base, &be_insn, ilen); op = insn >> 56; } else { insn = ld_code2(env, s, pc); -- 2.34.1 The ilen value extracted from ex_value is the length of the EXECUTE instruction itself, and so is the increment to the pc. However, the length of the synthetic insn is located in the opcode like all other instructions. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/s390x/tcg/translate.c | 4 ++-- 1 file changed, 2 insertions(+), 2 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 const DisasInsn *extract_insn(CPUS390XState *env, DisasContext *s) /* Extract the values saved by EXECUTE. */ insn = s->ex_value & 0xffffffffffff0000ull; ilen = s->ex_value & 0xf; + op = insn >> 56; /* Register insn bytes with translator so plugins work. */ be_insn = cpu_to_be64(insn); - translator_fake_ld(&s->base, &be_insn, ilen); - op = insn >> 56; + translator_fake_ld(&s->base, &be_insn, get_ilen(op)); } else { insn = ld_code2(env, s, pc); op = (insn >> 8) & 0xff; -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/s390x/tcg/translate.c | 5 +++-- 1 file changed, 3 insertions(+), 2 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 bool s390x_tr_disas_log(const DisasContextBase *dcbase, DisasContext *dc = container_of(dcbase, DisasContext, base); if (unlikely(dc->ex_value)) { - /* ??? Unfortunately target_disas can't use host memory. */ - fprintf(logfile, "IN: EXECUTE %016" PRIx64, dc->ex_value); + /* The ex_value has been recorded with translator_fake_ld. */ + fprintf(logfile, "IN: EXECUTE\n"); + target_disas(logfile, cs, &dc->base); return true; } return false; -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/hexagon/translate.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) 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 @@ #include "exec/translation-block.h" #include "exec/cpu_ldst.h" #include "exec/log.h" -#include "exec/cpu_ldst.h" #include "internal.h" #include "attribs.h" #include "insn.h" @@ -XXX,XX +XXX,XX @@ static bool pkt_crosses_page(CPUHexagonState *env, DisasContext *ctx) int nwords; for (nwords = 0; !found_end && nwords < PACKET_WORDS_MAX; nwords++) { - uint32_t word = cpu_ldl_code(env, + uint32_t word = translator_ldl(env, &ctx->base, ctx->base.pc_next + nwords * sizeof(uint32_t)); found_end = is_packet_end(word); } -- 2.34.1 Reviewed-by: Edgar E. Iglesias <edgar.iglesias@amd.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/microblaze/translate.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) 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 @@ #include "tcg/tcg-op.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" -#include "exec/cpu_ldst.h" #include "exec/translator.h" #include "qemu/qemu-print.h" @@ -XXX,XX +XXX,XX @@ static void mb_tr_translate_insn(DisasContextBase *dcb, CPUState *cs) dc->tb_flags_to_set = 0; - ir = cpu_ldl_code(cpu_env(cs), dc->base.pc_next); + ir = translator_ldl(cpu_env(cs), &dc->base, dc->base.pc_next); if (!decode(dc, ir)) { trap_illegal(dc, true); } -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/i386/tcg/translate.c | 8 +++----- 1 file changed, 3 insertions(+), 5 deletions(-) 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 @@ #include "exec/exec-all.h" #include "tcg/tcg-op.h" #include "tcg/tcg-op-gvec.h" -#include "exec/cpu_ldst.h" #include "exec/translator.h" #include "fpu/softfloat.h" @@ -XXX,XX +XXX,XX @@ static uint64_t advance_pc(CPUX86State *env, DisasContext *s, int num_bytes) * This can happen even if the operand is only one byte long! */ if (((s->pc - 1) ^ (pc - 1)) & TARGET_PAGE_MASK) { - volatile uint8_t unused = - cpu_ldub_code(env, (s->pc - 1) & TARGET_PAGE_MASK); - (void) unused; + (void)translator_ldub(env, &s->base, + (s->pc - 1) & TARGET_PAGE_MASK); } siglongjmp(s->jmpbuf, 1); } @@ -XXX,XX +XXX,XX @@ static void gen_unknown_opcode(CPUX86State *env, DisasContext *s) fprintf(logfile, "ILLOPC: " TARGET_FMT_lx ":", pc); for (; pc < end; ++pc) { - fprintf(logfile, " %02x", cpu_ldub_code(env, pc)); + fprintf(logfile, " %02x", translator_ldub(env, &s->base, pc)); } fprintf(logfile, "\n"); qemu_log_unlock(logfile); -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/avr/translate.c | 3 +-- 1 file changed, 1 insertion(+), 2 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 @@ #include "cpu.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" -#include "exec/cpu_ldst.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" #include "exec/log.h" @@ -XXX,XX +XXX,XX @@ static int to_regs_00_30_by_two(DisasContext *ctx, int indx) static uint16_t next_word(DisasContext *ctx) { - return cpu_lduw_code(ctx->env, ctx->npc++ * 2); + return translator_lduw(ctx->env, &ctx->base, ctx->npc++ * 2); } static int append_16(DisasContext *ctx, int x) -- 2.34.1 Reviewed-by: Edgar E. Iglesias <edgar.iglesias@amd.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/cris/translate.c | 25 ++++++++----------------- 1 file changed, 8 insertions(+), 17 deletions(-) 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 int sign_extend(unsigned int val, unsigned int width) } static int cris_fetch(CPUCRISState *env, DisasContext *dc, uint32_t addr, - unsigned int size, unsigned int sign) + unsigned int size, bool sign) { int r; switch (size) { case 4: - { - r = cpu_ldl_code(env, addr); + r = translator_ldl(env, &dc->base, addr); break; - } case 2: - { + r = translator_lduw(env, &dc->base, addr); if (sign) { - r = cpu_ldsw_code(env, addr); - } else { - r = cpu_lduw_code(env, addr); + r = (int16_t)r; } break; - } case 1: - { + r = translator_ldub(env, &dc->base, addr); if (sign) { - r = cpu_ldsb_code(env, addr); - } else { - r = cpu_ldub_code(env, addr); + r = (int8_t)r; } break; - } default: - cpu_abort(CPU(dc->cpu), "Invalid fetch size %d\n", size); - break; + g_assert_not_reached(); } return r; } @@ -XXX,XX +XXX,XX @@ static unsigned int crisv32_decoder(CPUCRISState *env, DisasContext *dc) int i; /* Load a halfword onto the instruction register. */ - dc->ir = cris_fetch(env, dc, dc->pc, 2, 0); + dc->ir = cris_fetch(env, dc, dc->pc, 2, 0); /* Now decode it. */ dc->opcode = EXTRACT_FIELD(dc->ir, 4, 11); -- 2.34.1 Reviewed-by: Edgar E. Iglesias <edgar.iglesias@amd.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/cris/translate.c | 1 - target/cris/translate_v10.c.inc | 30 +++++++++--------------------- 2 files changed, 9 insertions(+), 22 deletions(-) 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 @@ #include "tcg/tcg-op.h" #include "exec/helper-proto.h" #include "mmu.h" -#include "exec/cpu_ldst.h" #include "exec/translator.h" #include "crisv32-decode.h" #include "qemu/qemu-print.h" 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 int dec10_prep_move_m(CPUCRISState *env, DisasContext *dc, /* Load [$rs] onto T1. */ if (is_imm) { - if (memsize != 4) { - if (s_ext) { - if (memsize == 1) - imm = cpu_ldsb_code(env, dc->pc + 2); - else - imm = cpu_ldsw_code(env, dc->pc + 2); - } else { - if (memsize == 1) - imm = cpu_ldub_code(env, dc->pc + 2); - else - imm = cpu_lduw_code(env, dc->pc + 2); - } - } else - imm = cpu_ldl_code(env, dc->pc + 2); + imm = cris_fetch(env, dc, dc->pc + 2, memsize, s_ext); tcg_gen_movi_tl(dst, imm); @@ -XXX,XX +XXX,XX @@ static int dec10_dip(CPUCRISState *env, DisasContext *dc) LOG_DIS("dip pc=%x opcode=%d r%d r%d\n", dc->pc, dc->opcode, dc->src, dc->dst); if (dc->src == 15) { - imm = cpu_ldl_code(env, dc->pc + 2); + imm = cris_fetch(env, dc, dc->pc + 2, 4, 0); tcg_gen_movi_tl(cpu_PR[PR_PREFIX], imm); - if (dc->postinc) + if (dc->postinc) { insn_len += 4; + } tcg_gen_addi_tl(cpu_R[15], cpu_R[15], insn_len - 2); } else { gen_load(dc, cpu_PR[PR_PREFIX], cpu_R[dc->src], 4, 0); @@ -XXX,XX +XXX,XX @@ static unsigned int dec10_ind(CPUCRISState *env, DisasContext *dc) if (dc->src == 15) { LOG_DIS("jump.%d %d r%d r%d direct\n", size, dc->opcode, dc->src, dc->dst); - imm = cpu_ldl_code(env, dc->pc + 2); - if (dc->mode == CRISV10_MODE_AUTOINC) + imm = cris_fetch(env, dc, dc->pc + 2, size, 0); + if (dc->mode == CRISV10_MODE_AUTOINC) { insn_len += size; - + } c = tcg_constant_tl(dc->pc + insn_len); t_gen_mov_preg_TN(dc, dc->dst, c); dc->jmp_pc = imm; @@ -XXX,XX +XXX,XX @@ static unsigned int dec10_ind(CPUCRISState *env, DisasContext *dc) case CRISV10_IND_BCC_M: cris_cc_mask(dc, 0); - simm = cpu_ldsw_code(env, dc->pc + 2); + simm = cris_fetch(env, dc, dc->pc + 2, 2, 1); simm += 4; LOG_DIS("bcc_m: b%s %x\n", cc_name(dc->cond), dc->pc + simm); @@ -XXX,XX +XXX,XX @@ static unsigned int crisv10_decoder(CPUCRISState *env, DisasContext *dc) unsigned int insn_len = 2; /* Load a halfword onto the instruction register. */ - dc->ir = cpu_lduw_code(env, dc->pc); + dc->ir = cris_fetch(env, dc, dc->pc, 2, 0); /* Now decode it. */ dc->opcode = EXTRACT_FIELD(dc->ir, 6, 9); -- 2.34.1 Reviewed-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/riscv/translate.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) 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 @@ #include "qemu/log.h" #include "cpu.h" #include "tcg/tcg-op.h" -#include "exec/cpu_ldst.h" #include "exec/exec-all.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" @@ -XXX,XX +XXX,XX @@ static uint32_t opcode_at(DisasContextBase *dcbase, target_ulong pc) CPUState *cpu = ctx->cs; CPURISCVState *env = cpu_env(cpu); - return cpu_ldl_code(env, pc); + return translator_ldl(env, &ctx->base, pc); } /* Include insn module translation function */ @@ -XXX,XX +XXX,XX @@ static void riscv_tr_translate_insn(DisasContextBase *dcbase, CPUState *cpu) 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); + uint16_t next_insn = + translator_lduw(env, &ctx->base, ctx->base.pc_next); int len = insn_len(next_insn); if (!is_same_page(&ctx->base, ctx->base.pc_next + len - 1)) { -- 2.34.1 Reviewed-by: Yoshinori Sato <ysato@users.sourceforge.jp> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/rx/translate.c | 27 ++++++++++++++------------- 1 file changed, 14 insertions(+), 13 deletions(-) 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 @@ #include "cpu.h" #include "exec/exec-all.h" #include "tcg/tcg-op.h" -#include "exec/cpu_ldst.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" #include "exec/translator.h" @@ -XXX,XX +XXX,XX @@ static TCGv_i64 cpu_acc; /* decoder helper */ static uint32_t decode_load_bytes(DisasContext *ctx, uint32_t insn, - int i, int n) + int i, int n) { while (++i <= n) { - uint8_t b = cpu_ldub_code(ctx->env, ctx->base.pc_next++); + uint8_t b = translator_ldub(ctx->env, &ctx->base, ctx->base.pc_next++); insn |= b << (32 - i * 8); } return insn; @@ -XXX,XX +XXX,XX @@ static uint32_t li(DisasContext *ctx, int sz) CPURXState *env = ctx->env; addr = ctx->base.pc_next; - tcg_debug_assert(sz < 4); switch (sz) { case 1: ctx->base.pc_next += 1; - return cpu_ldsb_code(env, addr); + return (int8_t)translator_ldub(env, &ctx->base, addr); case 2: ctx->base.pc_next += 2; - return cpu_ldsw_code(env, addr); + return (int16_t)translator_lduw(env, &ctx->base, addr); case 3: ctx->base.pc_next += 3; - tmp = cpu_ldsb_code(env, addr + 2) << 16; - tmp |= cpu_lduw_code(env, addr) & 0xffff; + tmp = (int8_t)translator_ldub(env, &ctx->base, addr + 2); + tmp <<= 16; + tmp |= translator_lduw(env, &ctx->base, addr); return tmp; case 0: ctx->base.pc_next += 4; - return cpu_ldl_code(env, addr); + return translator_ldl(env, &ctx->base, addr); + default: + g_assert_not_reached(); } return 0; } @@ -XXX,XX +XXX,XX @@ static inline TCGv rx_index_addr(DisasContext *ctx, TCGv mem, { uint32_t dsp; - tcg_debug_assert(ld < 3); switch (ld) { case 0: return cpu_regs[reg]; case 1: - dsp = cpu_ldub_code(ctx->env, ctx->base.pc_next) << size; + dsp = translator_ldub(ctx->env, &ctx->base, ctx->base.pc_next) << size; tcg_gen_addi_i32(mem, cpu_regs[reg], dsp); ctx->base.pc_next += 1; return mem; case 2: - dsp = cpu_lduw_code(ctx->env, ctx->base.pc_next) << size; + dsp = translator_lduw(ctx->env, &ctx->base, ctx->base.pc_next) << size; tcg_gen_addi_i32(mem, cpu_regs[reg], dsp); ctx->base.pc_next += 2; return mem; + default: + g_assert_not_reached(); } - return NULL; } static inline MemOp mi_to_mop(unsigned mi) -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/xtensa/translate.c | 3 +-- 1 file changed, 1 insertion(+), 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 @@ #include "tcg/tcg-op.h" #include "qemu/log.h" #include "qemu/qemu-print.h" -#include "exec/cpu_ldst.h" #include "semihosting/semihost.h" #include "exec/translator.h" @@ -XXX,XX +XXX,XX @@ static void disas_xtensa_insn(CPUXtensaState *env, DisasContext *dc) static inline unsigned xtensa_insn_len(CPUXtensaState *env, DisasContext *dc) { - uint8_t b0 = cpu_ldub_code(env, dc->pc); + uint8_t b0 = translator_ldub(env, &dc->base, dc->pc); return xtensa_op0_insn_len(dc, b0); } -- 2.34.1 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/s390x/tcg/translate.c | 3 +-- 1 file changed, 1 insertion(+), 2 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 @@ #include "tcg/tcg-op-gvec.h" #include "qemu/log.h" #include "qemu/host-utils.h" -#include "exec/cpu_ldst.h" #include "exec/helper-proto.h" #include "exec/helper-gen.h" @@ -XXX,XX +XXX,XX @@ static void s390x_tr_insn_start(DisasContextBase *dcbase, CPUState *cs) static target_ulong get_next_pc(CPUS390XState *env, DisasContext *s, uint64_t pc) { - uint64_t insn = cpu_lduw_code(env, pc); + uint64_t insn = translator_lduw(env, &s->base, pc); return pc + get_ilen((insn >> 8) & 0xff); } -- 2.34.1 From: Philippe Mathieu-Daudé <philmd@linaro.org> Previous commits replaced them by translator_ld* calls. Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org> Message-Id: <20240405131532.40913-1-philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/cpu_ldst.h | 10 ---------- 1 file changed, 10 deletions(-) diff --git a/include/exec/cpu_ldst.h b/include/exec/cpu_ldst.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/cpu_ldst.h +++ b/include/exec/cpu_ldst.h @@ -XXX,XX +XXX,XX @@ uint32_t cpu_lduw_code(CPUArchState *env, abi_ptr addr); uint32_t cpu_ldl_code(CPUArchState *env, abi_ptr addr); uint64_t cpu_ldq_code(CPUArchState *env, abi_ptr addr); -static inline int cpu_ldsb_code(CPUArchState *env, abi_ptr addr) -{ - return (int8_t)cpu_ldub_code(env, addr); -} - -static inline int cpu_ldsw_code(CPUArchState *env, abi_ptr addr) -{ - return (int16_t)cpu_lduw_code(env, addr); -} - /** * tlb_vaddr_to_host: * @env: CPUArchState -- 2.34.1 TCG register spill/fill uses tcg_out_ld/st with all types, not necessarily going through INDEX_op_{ld,st}_vec. Cc: qemu-stable@nongnu.org Fixes: 16288ded944 ("tcg/loongarch64: Lower basic tcg vec ops to LSX") Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2336 Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Song Gao <gaosong@loongson.cn> Tested-by: Song Gao <gaosong@loongson.cn> --- tcg/loongarch64/tcg-target.c.inc | 103 ++++++++++++++++++++++++------- 1 file changed, 80 insertions(+), 23 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_ldst(TCGContext *s, LoongArchInsn opc, TCGReg data, } } -static void tcg_out_ld(TCGContext *s, TCGType type, TCGReg arg, - TCGReg arg1, intptr_t arg2) +static void tcg_out_ld(TCGContext *s, TCGType type, TCGReg dest, + TCGReg base, intptr_t offset) { - bool is_32bit = type == TCG_TYPE_I32; - tcg_out_ldst(s, is_32bit ? OPC_LD_W : OPC_LD_D, arg, arg1, arg2); + switch (type) { + case TCG_TYPE_I32: + if (dest < TCG_REG_V0) { + tcg_out_ldst(s, OPC_LD_W, dest, base, offset); + } else { + tcg_out_dupm_vec(s, TCG_TYPE_I128, MO_32, dest, base, offset); + } + break; + case TCG_TYPE_I64: + if (dest < TCG_REG_V0) { + tcg_out_ldst(s, OPC_LD_D, dest, base, offset); + } else { + tcg_out_dupm_vec(s, TCG_TYPE_I128, MO_64, dest, base, offset); + } + break; + case TCG_TYPE_V128: + if (-0x800 <= offset && offset <= 0x7ff) { + tcg_out_opc_vld(s, dest, base, offset); + } else { + tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP0, offset); + tcg_out_opc_vldx(s, dest, base, TCG_REG_TMP0); + } + break; + default: + g_assert_not_reached(); + } } -static void tcg_out_st(TCGContext *s, TCGType type, TCGReg arg, - TCGReg arg1, intptr_t arg2) +static void tcg_out_st(TCGContext *s, TCGType type, TCGReg src, + TCGReg base, intptr_t offset) { - bool is_32bit = type == TCG_TYPE_I32; - tcg_out_ldst(s, is_32bit ? OPC_ST_W : OPC_ST_D, arg, arg1, arg2); + switch (type) { + case TCG_TYPE_I32: + if (src < TCG_REG_V0) { + tcg_out_ldst(s, OPC_ST_W, src, base, offset); + } else { + /* TODO: Could use fst_s, fstx_s */ + if (offset < -0x100 || offset > 0xff || (offset & 3)) { + if (-0x800 <= offset && offset <= 0x7ff) { + tcg_out_opc_addi_d(s, TCG_REG_TMP0, base, offset); + } else { + tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP0, offset); + tcg_out_opc_add_d(s, TCG_REG_TMP0, TCG_REG_TMP0, base); + } + base = TCG_REG_TMP0; + offset = 0; + } + tcg_out_opc_vstelm_w(s, src, base, offset, 0); + } + break; + case TCG_TYPE_I64: + if (src < TCG_REG_V0) { + tcg_out_ldst(s, OPC_ST_D, src, base, offset); + } else { + /* TODO: Could use fst_d, fstx_d */ + if (offset < -0x100 || offset > 0xff || (offset & 7)) { + if (-0x800 <= offset && offset <= 0x7ff) { + tcg_out_opc_addi_d(s, TCG_REG_TMP0, base, offset); + } else { + tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP0, offset); + tcg_out_opc_add_d(s, TCG_REG_TMP0, TCG_REG_TMP0, base); + } + base = TCG_REG_TMP0; + offset = 0; + } + tcg_out_opc_vstelm_d(s, src, base, offset, 0); + } + break; + case TCG_TYPE_V128: + if (-0x800 <= offset && offset <= 0x7ff) { + tcg_out_opc_vst(s, src, base, offset); + } else { + tcg_out_movi(s, TCG_TYPE_PTR, TCG_REG_TMP0, offset); + tcg_out_opc_vstx(s, src, base, TCG_REG_TMP0); + } + break; + default: + g_assert_not_reached(); + } } static bool tcg_out_sti(TCGContext *s, TCGType type, TCGArg val, @@ -XXX,XX +XXX,XX @@ static void tcg_out_vec_op(TCGContext *s, TCGOpcode opc, { TCGType type = vecl + TCG_TYPE_V64; TCGArg a0, a1, a2, a3; - TCGReg temp = TCG_REG_TMP0; TCGReg temp_vec = TCG_VEC_TMP0; static const LoongArchInsn cmp_vec_insn[16][4] = { @@ -XXX,XX +XXX,XX @@ static void tcg_out_vec_op(TCGContext *s, TCGOpcode opc, switch (opc) { case INDEX_op_st_vec: - /* Try to fit vst imm */ - if (-0x800 <= a2 && a2 <= 0x7ff) { - tcg_out_opc_vst(s, a0, a1, a2); - } else { - tcg_out_movi(s, TCG_TYPE_I64, temp, a2); - tcg_out_opc_vstx(s, a0, a1, temp); - } + tcg_out_st(s, type, a0, a1, a2); break; case INDEX_op_ld_vec: - /* Try to fit vld imm */ - if (-0x800 <= a2 && a2 <= 0x7ff) { - tcg_out_opc_vld(s, a0, a1, a2); - } else { - tcg_out_movi(s, TCG_TYPE_I64, temp, a2); - tcg_out_opc_vldx(s, a0, a1, temp); - } + tcg_out_ld(s, type, a0, a1, a2); break; case INDEX_op_and_vec: tcg_out_opc_vand_v(s, a0, a1, a2); -- 2.34.1 v2: Fix target/loongarch printf formats for vaddr Include two more reviewed patches. r~ These should have been removed with the rest. There are a couple of hosts which can emit guest_base into the constant pool: aarch64, mips64, ppc64, riscv64. Fixes: a417ef835058 ("tcg: Remove TCG_TARGET_NEED_LDST_LABELS and TCG_TARGET_NEED_POOL_LABELS") Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> --- tcg/tci/tcg-target.h | 1 - tcg/tcg.c | 4 ---- 2 files changed, 5 deletions(-) diff --git a/tcg/tci/tcg-target.h b/tcg/tci/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/tci/tcg-target.h +++ b/tcg/tci/tcg-target.h @@ -XXX,XX +XXX,XX @@ typedef enum { } TCGReg; #define HAVE_TCG_QEMU_TB_EXEC -#define TCG_TARGET_NEED_POOL_LABELS #endif /* TCG_TARGET_H */ 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 @@ void tcg_prologue_init(void) tcg_qemu_tb_exec = (tcg_prologue_fn *)tcg_splitwx_to_rx(s->code_ptr); #endif -#ifdef TCG_TARGET_NEED_POOL_LABELS s->pool_labels = NULL; -#endif qemu_thread_jit_write(); /* Generate the prologue. */ tcg_target_qemu_prologue(s); -#ifdef TCG_TARGET_NEED_POOL_LABELS /* Allow the prologue to put e.g. guest_base into a pool entry. */ { int result = tcg_out_pool_finalize(s); tcg_debug_assert(result == 0); } -#endif prologue_size = tcg_current_code_size(s); perf_report_prologue(s->code_gen_ptr, prologue_size); -- 2.43.0 This is now prohibited in configuration. Reviewed-by: Alex Bennée <alex.bennee@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/qemu/atomic.h | 18 +++-------------- include/tcg/oversized-guest.h | 23 ---------------------- accel/tcg/cputlb.c | 7 ------- accel/tcg/tcg-all.c | 9 ++++----- target/arm/ptw.c | 34 --------------------------------- target/riscv/cpu_helper.c | 13 +------------ docs/devel/multi-thread-tcg.rst | 1 - 7 files changed, 8 insertions(+), 97 deletions(-) delete mode 100644 include/tcg/oversized-guest.h diff --git a/include/qemu/atomic.h b/include/qemu/atomic.h index XXXXXXX..XXXXXXX 100644 --- a/include/qemu/atomic.h +++ b/include/qemu/atomic.h @@ -XXX,XX +XXX,XX @@ */ #define signal_barrier() __atomic_signal_fence(__ATOMIC_SEQ_CST) -/* Sanity check that the size of an atomic operation isn't "overly large". +/* + * Sanity check that the size of an atomic operation isn't "overly large". * Despite the fact that e.g. i686 has 64-bit atomic operations, we do not * want to use them because we ought not need them, and this lets us do a * bit of sanity checking that other 32-bit hosts might build. - * - * That said, we have a problem on 64-bit ILP32 hosts in that in order to - * sync with TCG_OVERSIZED_GUEST, this must match TCG_TARGET_REG_BITS. - * We'd prefer not want to pull in everything else TCG related, so handle - * those few cases by hand. - * - * Note that x32 is fully detected with __x86_64__ + _ILP32, and that for - * Sparc we always force the use of sparcv9 in configure. MIPS n32 (ILP32) & - * n64 (LP64) ABIs are both detected using __mips64. */ -#if defined(__x86_64__) || defined(__sparc__) || defined(__mips64) -# define ATOMIC_REG_SIZE 8 -#else -# define ATOMIC_REG_SIZE sizeof(void *) -#endif +#define ATOMIC_REG_SIZE sizeof(void *) /* Weak atomic operations prevent the compiler moving other * loads/stores past the atomic operation load/store. However there is diff --git a/include/tcg/oversized-guest.h b/include/tcg/oversized-guest.h deleted file mode 100644 index XXXXXXX..XXXXXXX --- a/include/tcg/oversized-guest.h +++ /dev/null @@ -XXX,XX +XXX,XX @@ -/* SPDX-License-Identifier: MIT */ -/* - * Define TCG_OVERSIZED_GUEST - * Copyright (c) 2008 Fabrice Bellard - */ - -#ifndef EXEC_TCG_OVERSIZED_GUEST_H -#define EXEC_TCG_OVERSIZED_GUEST_H - -#include "tcg-target-reg-bits.h" -#include "cpu-param.h" - -/* - * Oversized TCG guests make things like MTTCG hard - * as we can't use atomics for cputlb updates. - */ -#if TARGET_LONG_BITS > TCG_TARGET_REG_BITS -#define TCG_OVERSIZED_GUEST 1 -#else -#define TCG_OVERSIZED_GUEST 0 -#endif - -#endif 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 @@ #include "qemu/plugin-memory.h" #endif #include "tcg/tcg-ldst.h" -#include "tcg/oversized-guest.h" /* DEBUG defines, enable DEBUG_TLB_LOG to log to the CPU_LOG_MMU target */ /* #define DEBUG_TLB */ @@ -XXX,XX +XXX,XX @@ static inline uint64_t tlb_read_idx(const CPUTLBEntry *entry, return qatomic_read(ptr); #else const uint64_t *ptr = &entry->addr_idx[access_type]; -# if TCG_OVERSIZED_GUEST - return *ptr; -# else /* ofs might correspond to .addr_write, so use qatomic_read */ return qatomic_read(ptr); -# endif #endif } @@ -XXX,XX +XXX,XX @@ static void tlb_reset_dirty_range_locked(CPUTLBEntry *tlb_entry, uint32_t *ptr_write = (uint32_t *)&tlb_entry->addr_write; ptr_write += HOST_BIG_ENDIAN; qatomic_set(ptr_write, *ptr_write | TLB_NOTDIRTY); -#elif TCG_OVERSIZED_GUEST - tlb_entry->addr_write |= TLB_NOTDIRTY; #else qatomic_set(&tlb_entry->addr_write, tlb_entry->addr_write | TLB_NOTDIRTY); diff --git a/accel/tcg/tcg-all.c b/accel/tcg/tcg-all.c index XXXXXXX..XXXXXXX 100644 --- a/accel/tcg/tcg-all.c +++ b/accel/tcg/tcg-all.c @@ -XXX,XX +XXX,XX @@ #include "exec/replay-core.h" #include "system/cpu-timers.h" #include "tcg/startup.h" -#include "tcg/oversized-guest.h" #include "qapi/error.h" #include "qemu/error-report.h" #include "qemu/accel.h" @@ -XXX,XX +XXX,XX @@ #include "hw/boards.h" #endif #include "internal-common.h" +#include "cpu-param.h" + struct TCGState { AccelState parent_obj; @@ -XXX,XX +XXX,XX @@ DECLARE_INSTANCE_CHECKER(TCGState, TCG_STATE, static bool default_mttcg_enabled(void) { - if (icount_enabled() || TCG_OVERSIZED_GUEST) { + if (icount_enabled()) { return false; } #ifdef TARGET_SUPPORTS_MTTCG @@ -XXX,XX +XXX,XX @@ static void tcg_set_thread(Object *obj, const char *value, Error **errp) TCGState *s = TCG_STATE(obj); if (strcmp(value, "multi") == 0) { - if (TCG_OVERSIZED_GUEST) { - error_setg(errp, "No MTTCG when guest word size > hosts"); - } else if (icount_enabled()) { + if (icount_enabled()) { error_setg(errp, "No MTTCG when icount is enabled"); } else { #ifndef TARGET_SUPPORTS_MTTCG diff --git a/target/arm/ptw.c b/target/arm/ptw.c index XXXXXXX..XXXXXXX 100644 --- a/target/arm/ptw.c +++ b/target/arm/ptw.c @@ -XXX,XX +XXX,XX @@ #include "internals.h" #include "cpu-features.h" #include "idau.h" -#ifdef CONFIG_TCG -# include "tcg/oversized-guest.h" -#endif typedef struct S1Translate { /* @@ -XXX,XX +XXX,XX @@ static uint64_t arm_casq_ptw(CPUARMState *env, uint64_t old_val, ptw->out_rw = true; } -#ifdef CONFIG_ATOMIC64 if (ptw->out_be) { old_val = cpu_to_be64(old_val); new_val = cpu_to_be64(new_val); @@ -XXX,XX +XXX,XX @@ static uint64_t arm_casq_ptw(CPUARMState *env, uint64_t old_val, cur_val = qatomic_cmpxchg__nocheck((uint64_t *)host, old_val, new_val); cur_val = le64_to_cpu(cur_val); } -#else - /* - * We can't support the full 64-bit atomic cmpxchg on the host. - * Because this is only used for FEAT_HAFDBS, which is only for AA64, - * we know that TCG_OVERSIZED_GUEST is set, which means that we are - * running in round-robin mode and could only race with dma i/o. - */ -#if !TCG_OVERSIZED_GUEST -# error "Unexpected configuration" -#endif - bool locked = bql_locked(); - if (!locked) { - bql_lock(); - } - if (ptw->out_be) { - cur_val = ldq_be_p(host); - if (cur_val == old_val) { - stq_be_p(host, new_val); - } - } else { - cur_val = ldq_le_p(host); - if (cur_val == old_val) { - stq_le_p(host, new_val); - } - } - if (!locked) { - bql_unlock(); - } -#endif - return cur_val; #else /* AArch32 does not have FEAT_HADFS; non-TCG guests only use debug-mode. */ diff --git a/target/riscv/cpu_helper.c b/target/riscv/cpu_helper.c index XXXXXXX..XXXXXXX 100644 --- a/target/riscv/cpu_helper.c +++ b/target/riscv/cpu_helper.c @@ -XXX,XX +XXX,XX @@ #include "system/cpu-timers.h" #include "cpu_bits.h" #include "debug.h" -#include "tcg/oversized-guest.h" #include "pmp.h" int riscv_env_mmu_index(CPURISCVState *env, bool ifetch) @@ -XXX,XX +XXX,XX @@ static int get_physical_address(CPURISCVState *env, hwaddr *physical, hwaddr pte_addr; int i; -#if !TCG_OVERSIZED_GUEST -restart: -#endif + restart: for (i = 0; i < levels; i++, ptshift -= ptidxbits) { target_ulong idx; if (i == 0) { @@ -XXX,XX +XXX,XX @@ restart: false, MEMTXATTRS_UNSPECIFIED); if (memory_region_is_ram(mr)) { target_ulong *pte_pa = qemu_map_ram_ptr(mr->ram_block, addr1); -#if TCG_OVERSIZED_GUEST - /* - * MTTCG is not enabled on oversized TCG guests so - * page table updates do not need to be atomic - */ - *pte_pa = pte = updated_pte; -#else target_ulong old_pte; if (riscv_cpu_sxl(env) == MXL_RV32) { old_pte = qatomic_cmpxchg((uint32_t *)pte_pa, pte, updated_pte); @@ -XXX,XX +XXX,XX @@ restart: goto restart; } pte = updated_pte; -#endif } else { /* * Misconfigured PTE in ROM (AD bits are not preset) or diff --git a/docs/devel/multi-thread-tcg.rst b/docs/devel/multi-thread-tcg.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/devel/multi-thread-tcg.rst +++ b/docs/devel/multi-thread-tcg.rst @@ -XXX,XX +XXX,XX @@ if: * forced by --accel tcg,thread=single * enabling --icount mode -* 64 bit guests on 32 bit hosts (TCG_OVERSIZED_GUEST) In the general case of running translated code there should be no inter-vCPU dependencies and all vCPUs should be able to run at full -- 2.43.0 Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/tcg-op-ldst.c | 21 +++------------------ tcg/tcg.c | 4 +--- 2 files changed, 4 insertions(+), 21 deletions(-) diff --git a/tcg/tcg-op-ldst.c b/tcg/tcg-op-ldst.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg-op-ldst.c +++ b/tcg/tcg-op-ldst.c @@ -XXX,XX +XXX,XX @@ static MemOp tcg_canonicalize_memop(MemOp op, bool is64, bool st) static void gen_ldst(TCGOpcode opc, TCGType type, TCGTemp *vl, TCGTemp *vh, TCGTemp *addr, MemOpIdx oi) { - if (TCG_TARGET_REG_BITS == 64 || tcg_ctx->addr_type == TCG_TYPE_I32) { - if (vh) { - tcg_gen_op4(opc, type, temp_arg(vl), temp_arg(vh), - temp_arg(addr), oi); - } else { - tcg_gen_op3(opc, type, temp_arg(vl), temp_arg(addr), oi); - } + if (vh) { + tcg_gen_op4(opc, type, temp_arg(vl), temp_arg(vh), temp_arg(addr), oi); } else { - /* See TCGV_LOW/HIGH. */ - TCGTemp *al = addr + HOST_BIG_ENDIAN; - TCGTemp *ah = addr + !HOST_BIG_ENDIAN; - - if (vh) { - tcg_gen_op5(opc, type, temp_arg(vl), temp_arg(vh), - temp_arg(al), temp_arg(ah), oi); - } else { - tcg_gen_op4(opc, type, temp_arg(vl), - temp_arg(al), temp_arg(ah), oi); - } + tcg_gen_op3(opc, type, temp_arg(vl), temp_arg(addr), oi); } } 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 @@ void tcg_func_start(TCGContext *s) s->emit_before_op = NULL; QSIMPLEQ_INIT(&s->labels); - tcg_debug_assert(s->addr_type == TCG_TYPE_I32 || - s->addr_type == TCG_TYPE_I64); - + tcg_debug_assert(s->addr_type <= TCG_TYPE_REG); tcg_debug_assert(s->insn_start_words > 0); } -- 2.43.0 Since 64-on-32 is now unsupported, guest addresses always fit in one host register. Drop the replication of opcodes. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/tcg/tcg-opc.h | 28 ++------ tcg/optimize.c | 21 ++---- tcg/tcg-op-ldst.c | 82 +++++---------------- tcg/tcg.c | 42 ++++------- tcg/tci.c | 119 ++++++------------------------- tcg/aarch64/tcg-target.c.inc | 36 ++++------ tcg/arm/tcg-target.c.inc | 40 +++-------- tcg/i386/tcg-target.c.inc | 69 ++++-------------- tcg/loongarch64/tcg-target.c.inc | 36 ++++------ tcg/mips/tcg-target.c.inc | 51 +++---------- tcg/ppc/tcg-target.c.inc | 68 ++++-------------- tcg/riscv/tcg-target.c.inc | 24 +++---- tcg/s390x/tcg-target.c.inc | 36 ++++------ tcg/sparc64/tcg-target.c.inc | 24 +++---- tcg/tci/tcg-target.c.inc | 60 ++++------------ 15 files changed, 177 insertions(+), 559 deletions(-) diff --git a/include/tcg/tcg-opc.h b/include/tcg/tcg-opc.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg-opc.h +++ b/include/tcg/tcg-opc.h @@ -XXX,XX +XXX,XX @@ DEF(goto_ptr, 0, 1, 0, TCG_OPF_BB_EXIT | TCG_OPF_BB_END) DEF(plugin_cb, 0, 0, 1, TCG_OPF_NOT_PRESENT) DEF(plugin_mem_cb, 0, 1, 1, TCG_OPF_NOT_PRESENT) -/* Replicate ld/st ops for 32 and 64-bit guest addresses. */ -DEF(qemu_ld_a32_i32, 1, 1, 1, +DEF(qemu_ld_i32, 1, 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a32_i32, 0, 1 + 1, 1, +DEF(qemu_st_i32, 0, 1 + 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_ld_a32_i64, DATA64_ARGS, 1, 1, +DEF(qemu_ld_i64, DATA64_ARGS, 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a32_i64, 0, DATA64_ARGS + 1, 1, - TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) - -DEF(qemu_ld_a64_i32, 1, DATA64_ARGS, 1, - TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a64_i32, 0, 1 + DATA64_ARGS, 1, - TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_ld_a64_i64, DATA64_ARGS, DATA64_ARGS, 1, - TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a64_i64, 0, DATA64_ARGS + DATA64_ARGS, 1, +DEF(qemu_st_i64, 0, DATA64_ARGS + 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) /* Only used by i386 to cope with stupid register constraints. */ -DEF(qemu_st8_a32_i32, 0, 1 + 1, 1, - TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st8_a64_i32, 0, 1 + DATA64_ARGS, 1, +DEF(qemu_st8_i32, 0, 1 + 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) /* Only for 64-bit hosts at the moment. */ -DEF(qemu_ld_a32_i128, 2, 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_ld_a64_i128, 2, 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a32_i128, 0, 3, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) -DEF(qemu_st_a64_i128, 0, 3, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) +DEF(qemu_ld_i128, 2, 1, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) +DEF(qemu_st_i128, 0, 3, 1, TCG_OPF_CALL_CLOBBER | TCG_OPF_SIDE_EFFECTS) /* Host vector support. */ diff --git a/tcg/optimize.c b/tcg/optimize.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/optimize.c +++ b/tcg/optimize.c @@ -XXX,XX +XXX,XX @@ void tcg_optimize(TCGContext *s) CASE_OP_32_64_VEC(orc): done = fold_orc(&ctx, op); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: + case INDEX_op_qemu_ld_i32: done = fold_qemu_ld_1reg(&ctx, op); break; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { done = fold_qemu_ld_1reg(&ctx, op); break; } QEMU_FALLTHROUGH; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: done = fold_qemu_ld_2reg(&ctx, op); break; - case INDEX_op_qemu_st8_a32_i32: - case INDEX_op_qemu_st8_a64_i32: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st8_i32: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: + case INDEX_op_qemu_st_i128: done = fold_qemu_st(&ctx, op); break; CASE_OP_32_64(rem): diff --git a/tcg/tcg-op-ldst.c b/tcg/tcg-op-ldst.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tcg-op-ldst.c +++ b/tcg/tcg-op-ldst.c @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i32_int(TCGv_i32 val, TCGTemp *addr, MemOp orig_memop; MemOpIdx orig_oi, oi; TCGv_i64 copy_addr; - TCGOpcode opc; tcg_gen_req_mo(TCG_MO_LD_LD | TCG_MO_ST_LD); orig_memop = memop = tcg_canonicalize_memop(memop, 0, 0); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i32_int(TCGv_i32 val, TCGTemp *addr, } copy_addr = plugin_maybe_preserve_addr(addr); - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_ld_a32_i32; - } else { - opc = INDEX_op_qemu_ld_a64_i32; - } - gen_ldst(opc, TCG_TYPE_I32, tcgv_i32_temp(val), NULL, addr, oi); + gen_ldst(INDEX_op_qemu_ld_i32, TCG_TYPE_I32, + tcgv_i32_temp(val), NULL, addr, oi); plugin_gen_mem_callbacks_i32(val, copy_addr, addr, orig_oi, QEMU_PLUGIN_MEM_R); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i32_int(TCGv_i32 val, TCGTemp *addr, } if (TCG_TARGET_HAS_qemu_st8_i32 && (memop & MO_SIZE) == MO_8) { - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_st8_a32_i32; - } else { - opc = INDEX_op_qemu_st8_a64_i32; - } + opc = INDEX_op_qemu_st8_i32; } else { - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_st_a32_i32; - } else { - opc = INDEX_op_qemu_st_a64_i32; - } + opc = INDEX_op_qemu_st_i32; } gen_ldst(opc, TCG_TYPE_I32, tcgv_i32_temp(val), NULL, addr, oi); plugin_gen_mem_callbacks_i32(val, NULL, addr, orig_oi, QEMU_PLUGIN_MEM_W); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i64_int(TCGv_i64 val, TCGTemp *addr, MemOp orig_memop; MemOpIdx orig_oi, oi; TCGv_i64 copy_addr; - TCGOpcode opc; if (TCG_TARGET_REG_BITS == 32 && (memop & MO_SIZE) < MO_64) { tcg_gen_qemu_ld_i32_int(TCGV_LOW(val), addr, idx, memop); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i64_int(TCGv_i64 val, TCGTemp *addr, } copy_addr = plugin_maybe_preserve_addr(addr); - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_ld_a32_i64; - } else { - opc = INDEX_op_qemu_ld_a64_i64; - } - gen_ldst_i64(opc, val, addr, oi); + gen_ldst_i64(INDEX_op_qemu_ld_i64, val, addr, oi); plugin_gen_mem_callbacks_i64(val, copy_addr, addr, orig_oi, QEMU_PLUGIN_MEM_R); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i64_int(TCGv_i64 val, TCGTemp *addr, { TCGv_i64 swap = NULL; MemOpIdx orig_oi, oi; - TCGOpcode opc; if (TCG_TARGET_REG_BITS == 32 && (memop & MO_SIZE) < MO_64) { tcg_gen_qemu_st_i32_int(TCGV_LOW(val), addr, idx, memop); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i64_int(TCGv_i64 val, TCGTemp *addr, oi = make_memop_idx(memop, idx); } - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_st_a32_i64; - } else { - opc = INDEX_op_qemu_st_a64_i64; - } - gen_ldst_i64(opc, val, addr, oi); + gen_ldst_i64(INDEX_op_qemu_st_i64, val, addr, oi); plugin_gen_mem_callbacks_i64(val, NULL, addr, orig_oi, QEMU_PLUGIN_MEM_W); if (swap) { @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i128_int(TCGv_i128 val, TCGTemp *addr, { MemOpIdx orig_oi; TCGv_i64 ext_addr = NULL; - TCGOpcode opc; check_max_alignment(memop_alignment_bits(memop)); tcg_gen_req_mo(TCG_MO_LD_LD | TCG_MO_ST_LD); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i128_int(TCGv_i128 val, TCGTemp *addr, hi = TCGV128_HIGH(val); } - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_ld_a32_i128; - } else { - opc = INDEX_op_qemu_ld_a64_i128; - } - gen_ldst(opc, TCG_TYPE_I128, tcgv_i64_temp(lo), + gen_ldst(INDEX_op_qemu_ld_i128, TCG_TYPE_I128, tcgv_i64_temp(lo), tcgv_i64_temp(hi), addr, oi); if (need_bswap) { @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i128_int(TCGv_i128 val, TCGTemp *addr, canonicalize_memop_i128_as_i64(mop, memop); need_bswap = (mop[0] ^ memop) & MO_BSWAP; - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_ld_a32_i64; - } else { - opc = INDEX_op_qemu_ld_a64_i64; - } - /* * Since there are no global TCGv_i128, there is no visible state * changed if the second load faults. Load directly into the two @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i128_int(TCGv_i128 val, TCGTemp *addr, y = TCGV128_LOW(val); } - gen_ldst_i64(opc, x, addr, make_memop_idx(mop[0], idx)); + gen_ldst_i64(INDEX_op_qemu_ld_i64, x, addr, + make_memop_idx(mop[0], idx)); if (need_bswap) { tcg_gen_bswap64_i64(x, x); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_ld_i128_int(TCGv_i128 val, TCGTemp *addr, addr_p8 = tcgv_i64_temp(t); } - gen_ldst_i64(opc, y, addr_p8, make_memop_idx(mop[1], idx)); + gen_ldst_i64(INDEX_op_qemu_ld_i64, y, addr_p8, + make_memop_idx(mop[1], idx)); tcg_temp_free_internal(addr_p8); if (need_bswap) { @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i128_int(TCGv_i128 val, TCGTemp *addr, { MemOpIdx orig_oi; TCGv_i64 ext_addr = NULL; - TCGOpcode opc; check_max_alignment(memop_alignment_bits(memop)); tcg_gen_req_mo(TCG_MO_ST_LD | TCG_MO_ST_ST); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i128_int(TCGv_i128 val, TCGTemp *addr, hi = TCGV128_HIGH(val); } - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_st_a32_i128; - } else { - opc = INDEX_op_qemu_st_a64_i128; - } - gen_ldst(opc, TCG_TYPE_I128, tcgv_i64_temp(lo), - tcgv_i64_temp(hi), addr, oi); + gen_ldst(INDEX_op_qemu_st_i128, TCG_TYPE_I128, + tcgv_i64_temp(lo), tcgv_i64_temp(hi), addr, oi); if (need_bswap) { tcg_temp_free_i64(lo); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i128_int(TCGv_i128 val, TCGTemp *addr, canonicalize_memop_i128_as_i64(mop, memop); - if (tcg_ctx->addr_type == TCG_TYPE_I32) { - opc = INDEX_op_qemu_st_a32_i64; - } else { - opc = INDEX_op_qemu_st_a64_i64; - } - if ((memop & MO_BSWAP) == MO_LE) { x = TCGV128_LOW(val); y = TCGV128_HIGH(val); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i128_int(TCGv_i128 val, TCGTemp *addr, x = b; } - gen_ldst_i64(opc, x, addr, make_memop_idx(mop[0], idx)); + gen_ldst_i64(INDEX_op_qemu_st_i64, x, addr, + make_memop_idx(mop[0], idx)); if (tcg_ctx->addr_type == TCG_TYPE_I32) { TCGv_i32 t = tcg_temp_ebb_new_i32(); @@ -XXX,XX +XXX,XX @@ static void tcg_gen_qemu_st_i128_int(TCGv_i128 val, TCGTemp *addr, if (b) { tcg_gen_bswap64_i64(b, y); - gen_ldst_i64(opc, b, addr_p8, make_memop_idx(mop[1], idx)); + gen_ldst_i64(INDEX_op_qemu_st_i64, b, addr_p8, + make_memop_idx(mop[1], idx)); tcg_temp_free_i64(b); } else { - gen_ldst_i64(opc, y, addr_p8, make_memop_idx(mop[1], idx)); + gen_ldst_i64(INDEX_op_qemu_st_i64, y, addr_p8, + make_memop_idx(mop[1], idx)); } tcg_temp_free_internal(addr_p8); } else { 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 @@ bool tcg_op_supported(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_exit_tb: case INDEX_op_goto_tb: case INDEX_op_goto_ptr: - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_ld_i64: + case INDEX_op_qemu_st_i64: return true; - case INDEX_op_qemu_st8_a32_i32: - case INDEX_op_qemu_st8_a64_i32: + case INDEX_op_qemu_st8_i32: return TCG_TARGET_HAS_qemu_st8_i32; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_ld_i128: + case INDEX_op_qemu_st_i128: return TCG_TARGET_HAS_qemu_ldst_i128; case INDEX_op_mov_i32: @@ -XXX,XX +XXX,XX @@ void tcg_dump_ops(TCGContext *s, FILE *f, bool have_prefs) } i = 1; break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st8_a32_i32: - case INDEX_op_qemu_st8_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st8_i32: + case INDEX_op_qemu_ld_i64: + case INDEX_op_qemu_st_i64: + case INDEX_op_qemu_ld_i128: + case INDEX_op_qemu_st_i128: { const char *s_al, *s_op, *s_at; MemOpIdx oi = op->args[k++]; diff --git a/tcg/tci.c b/tcg/tci.c index XXXXXXX..XXXXXXX 100644 --- a/tcg/tci.c +++ b/tcg/tci.c @@ -XXX,XX +XXX,XX @@ static void tci_args_rrrbb(uint32_t insn, TCGReg *r0, TCGReg *r1, *i4 = extract32(insn, 26, 6); } -static void tci_args_rrrrr(uint32_t insn, TCGReg *r0, TCGReg *r1, - TCGReg *r2, TCGReg *r3, TCGReg *r4) -{ - *r0 = extract32(insn, 8, 4); - *r1 = extract32(insn, 12, 4); - *r2 = extract32(insn, 16, 4); - *r3 = extract32(insn, 20, 4); - *r4 = extract32(insn, 24, 4); -} - static void tci_args_rrrr(uint32_t insn, TCGReg *r0, TCGReg *r1, TCGReg *r2, TCGReg *r3) { @@ -XXX,XX +XXX,XX @@ uintptr_t QEMU_DISABLE_CFI tcg_qemu_tb_exec(CPUArchState *env, tb_ptr = ptr; break; - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: tci_args_rrm(insn, &r0, &r1, &oi); - taddr = (uint32_t)regs[r1]; - goto do_ld_i32; - case INDEX_op_qemu_ld_a64_i32: - if (TCG_TARGET_REG_BITS == 64) { - tci_args_rrm(insn, &r0, &r1, &oi); - taddr = regs[r1]; - } else { - tci_args_rrrr(insn, &r0, &r1, &r2, &r3); - taddr = tci_uint64(regs[r2], regs[r1]); - oi = regs[r3]; - } - do_ld_i32: + taddr = regs[r1]; regs[r0] = tci_qemu_ld(env, taddr, oi, tb_ptr); break; - case INDEX_op_qemu_ld_a32_i64: - if (TCG_TARGET_REG_BITS == 64) { - tci_args_rrm(insn, &r0, &r1, &oi); - taddr = (uint32_t)regs[r1]; - } else { - tci_args_rrrr(insn, &r0, &r1, &r2, &r3); - taddr = (uint32_t)regs[r2]; - oi = regs[r3]; - } - goto do_ld_i64; - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { tci_args_rrm(insn, &r0, &r1, &oi); taddr = regs[r1]; } else { - tci_args_rrrrr(insn, &r0, &r1, &r2, &r3, &r4); - taddr = tci_uint64(regs[r3], regs[r2]); - oi = regs[r4]; + tci_args_rrrr(insn, &r0, &r1, &r2, &r3); + taddr = regs[r2]; + oi = regs[r3]; } - do_ld_i64: tmp64 = tci_qemu_ld(env, taddr, oi, tb_ptr); if (TCG_TARGET_REG_BITS == 32) { tci_write_reg64(regs, r1, r0, tmp64); @@ -XXX,XX +XXX,XX @@ uintptr_t QEMU_DISABLE_CFI tcg_qemu_tb_exec(CPUArchState *env, } break; - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: tci_args_rrm(insn, &r0, &r1, &oi); - taddr = (uint32_t)regs[r1]; - goto do_st_i32; - case INDEX_op_qemu_st_a64_i32: - if (TCG_TARGET_REG_BITS == 64) { - tci_args_rrm(insn, &r0, &r1, &oi); - taddr = regs[r1]; - } else { - tci_args_rrrr(insn, &r0, &r1, &r2, &r3); - taddr = tci_uint64(regs[r2], regs[r1]); - oi = regs[r3]; - } - do_st_i32: + taddr = regs[r1]; tci_qemu_st(env, taddr, regs[r0], oi, tb_ptr); break; - case INDEX_op_qemu_st_a32_i64: - if (TCG_TARGET_REG_BITS == 64) { - tci_args_rrm(insn, &r0, &r1, &oi); - tmp64 = regs[r0]; - taddr = (uint32_t)regs[r1]; - } else { - tci_args_rrrr(insn, &r0, &r1, &r2, &r3); - tmp64 = tci_uint64(regs[r1], regs[r0]); - taddr = (uint32_t)regs[r2]; - oi = regs[r3]; - } - goto do_st_i64; - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { tci_args_rrm(insn, &r0, &r1, &oi); tmp64 = regs[r0]; taddr = regs[r1]; } else { - tci_args_rrrrr(insn, &r0, &r1, &r2, &r3, &r4); + tci_args_rrrr(insn, &r0, &r1, &r2, &r3); tmp64 = tci_uint64(regs[r1], regs[r0]); - taddr = tci_uint64(regs[r3], regs[r2]); - oi = regs[r4]; + taddr = regs[r2]; + oi = regs[r3]; } - do_st_i64: tci_qemu_st(env, taddr, tmp64, oi, tb_ptr); break; @@ -XXX,XX +XXX,XX @@ int print_insn_tci(bfd_vma addr, disassemble_info *info) str_r(r3), str_r(r4), str_r(r5)); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_st_a32_i32: - len = 1 + 1; - goto do_qemu_ldst; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_st_a64_i32: - len = 1 + DIV_ROUND_UP(64, TCG_TARGET_REG_BITS); - goto do_qemu_ldst; - case INDEX_op_qemu_ld_a64_i64: - case INDEX_op_qemu_st_a64_i64: - len = 2 * DIV_ROUND_UP(64, TCG_TARGET_REG_BITS); - goto do_qemu_ldst; - do_qemu_ldst: - switch (len) { - case 2: - tci_args_rrm(insn, &r0, &r1, &oi); - info->fprintf_func(info->stream, "%-12s %s, %s, %x", - op_name, str_r(r0), str_r(r1), oi); - break; - case 3: + case INDEX_op_qemu_ld_i64: + case INDEX_op_qemu_st_i64: + if (TCG_TARGET_REG_BITS == 32) { tci_args_rrrr(insn, &r0, &r1, &r2, &r3); info->fprintf_func(info->stream, "%-12s %s, %s, %s, %s", op_name, str_r(r0), str_r(r1), str_r(r2), str_r(r3)); break; - case 4: - tci_args_rrrrr(insn, &r0, &r1, &r2, &r3, &r4); - info->fprintf_func(info->stream, "%-12s %s, %s, %s, %s, %s", - op_name, str_r(r0), str_r(r1), - str_r(r2), str_r(r3), str_r(r4)); - break; - default: - g_assert_not_reached(); } + /* fall through */ + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_st_i32: + tci_args_rrm(insn, &r0, &r1, &oi); + info->fprintf_func(info->stream, "%-12s %s, %s, %x", + op_name, str_r(r0), str_r(r1), oi); break; case 0: 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_op(TCGContext *s, TCGOpcode opc, TCGType ext, tcg_out_insn(s, 3506, CSEL, ext, a0, REG0(3), REG0(4), args[5]); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, a0, a1, a2, ext); break; - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, REG0(0), a1, a2, ext); break; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_out_qemu_ldst_i128(s, a0, a1, a2, args[3], true); break; - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_out_qemu_ldst_i128(s, REG0(0), REG0(1), a2, args[3], false); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_movcond_i64: return C_O1_I4(r, r, rC, rZ, rZ); - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: return C_O2_I1(r, r, r); - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: return C_O0_I2(rZ, r); - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: return C_O0_I3(rZ, rZ, r); case INDEX_op_deposit_i32: 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_op(TCGContext *s, TCGOpcode opc, TCGType type, ARITH_MOV, args[0], 0, 0); break; - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a64_i32: - tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], - args[3], TCG_TYPE_I32); - break; - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, args[0], args[1], args[2], -1, args[3], TCG_TYPE_I64); break; - case INDEX_op_qemu_ld_a64_i64: - tcg_out_qemu_ld(s, args[0], args[1], args[2], args[3], - args[4], TCG_TYPE_I64); - break; - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a64_i32: - tcg_out_qemu_st(s, args[0], -1, args[1], args[2], - args[3], TCG_TYPE_I32); - break; - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, args[0], args[1], args[2], -1, args[3], TCG_TYPE_I64); break; - case INDEX_op_qemu_st_a64_i64: - tcg_out_qemu_st(s, args[0], args[1], args[2], args[3], - args[4], TCG_TYPE_I64); - break; case INDEX_op_bswap16_i32: tcg_out_bswap16(s, COND_AL, args[0], args[1], args[2]); @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_setcond2_i32: return C_O1_I4(r, r, r, rI, rI); - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: return C_O1_I1(r, q); - case INDEX_op_qemu_ld_a64_i32: - return C_O1_I2(r, q, q); - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: return C_O2_I1(e, p, q); - case INDEX_op_qemu_ld_a64_i64: - return C_O2_I2(e, p, q, q); - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: return C_O0_I2(q, q); - case INDEX_op_qemu_st_a64_i32: - return C_O0_I3(q, q, q); - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: return C_O0_I3(Q, p, q); - case INDEX_op_qemu_st_a64_i64: - return C_O0_I4(Q, p, q, q); case INDEX_op_st_vec: return C_O0_I2(w, r); 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_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out_modrm(s, OPC_GRP3_Ev + rexw, EXT3_NOT, a0); break; - case INDEX_op_qemu_ld_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_ld(s, a0, -1, a1, a2, args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); } else { tcg_out_qemu_ld(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_ld_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); - } else { - tcg_out_qemu_ld(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); - } - break; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); tcg_out_qemu_ld(s, a0, a1, a2, -1, args[3], TCG_TYPE_I128); break; - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st8_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_st(s, a0, -1, a1, a2, args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st8_a32_i32: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st8_i32: tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); } else { tcg_out_qemu_st(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_st_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); - } else { - tcg_out_qemu_st(s, a0, a1, a2, args[3], args[4], TCG_TYPE_I64); - } - break; - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); tcg_out_qemu_st(s, a0, a1, a2, -1, args[3], TCG_TYPE_I128); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_clz_i64: return have_lzcnt ? C_N1_I2(r, r, rW) : C_N1_I2(r, r, r); - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: return C_O1_I1(r, L); - case INDEX_op_qemu_ld_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, L) : C_O1_I2(r, L, L); - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: return C_O0_I2(L, L); - case INDEX_op_qemu_st_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(L, L) : C_O0_I3(L, L, L); - case INDEX_op_qemu_st8_a32_i32: + case INDEX_op_qemu_st8_i32: return C_O0_I2(s, L); - case INDEX_op_qemu_st8_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(s, L) : C_O0_I3(s, L, L); - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, L) : C_O2_I1(r, r, L); - case INDEX_op_qemu_ld_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, L) : C_O2_I2(r, r, L, L); - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(L, L) : C_O0_I3(L, L, L); - case INDEX_op_qemu_st_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(L, L) : C_O0_I4(L, L, L, L); - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); return C_O2_I1(r, r, L); - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); return C_O0_I3(L, L, L); 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_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out_ldst(s, OPC_ST_D, a0, a1, a2); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_out_qemu_ldst_i128(s, a0, a1, a2, a3, true); break; - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I64); break; - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_out_qemu_ldst_i128(s, a0, a1, a2, a3, false); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st32_i64: case INDEX_op_st_i32: case INDEX_op_st_i64: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: return C_O0_I2(rZ, r); - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: return C_N2_I1(r, r, r); - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: return C_O0_I3(r, r, r); case INDEX_op_brcond_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_ld32u_i64: case INDEX_op_ld_i32: case INDEX_op_ld_i64: - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); case INDEX_op_andc_i32: 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 void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out_setcond2(s, args[5], a0, a1, a2, args[3], args[4]); break; - case INDEX_op_qemu_ld_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_ld(s, a0, 0, a1, a2, args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); } else { tcg_out_qemu_ld(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_ld_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, 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_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_st(s, a0, 0, a1, a2, args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); } else { tcg_out_qemu_st(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_st_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, a0, 0, a1, 0, a2, 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: tcg_out_addsub2(s, a0, a1, a2, args[3], args[4], args[5], @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_brcond2_i32: return C_O0_I4(rZ, rZ, rZ, rZ); - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: return C_O1_I1(r, r); - case INDEX_op_qemu_ld_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O1_I2(r, r, r); - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: return C_O0_I2(rZ, r); - case INDEX_op_qemu_st_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(rZ, r) : C_O0_I3(rZ, r, r); - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I1(r, r, r); - case INDEX_op_qemu_ld_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I2(r, r, r, r); - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(rZ, r) : C_O0_I3(rZ, rZ, r); - case INDEX_op_qemu_st_a64_i64: - return (TCG_TARGET_REG_BITS == 64 ? C_O0_I2(rZ, r) - : C_O0_I4(rZ, rZ, r, r)); default: return C_NotImplemented; 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 void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out32(s, MODUD | TAB(args[0], args[1], args[2])); break; - case INDEX_op_qemu_ld_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], - args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_ld(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I64); @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_ld_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, args[0], -1, args[1], -1, - args[2], 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_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); tcg_out_qemu_ldst_i128(s, args[0], args[1], args[2], args[3], true); break; - case INDEX_op_qemu_st_a64_i32: - if (TCG_TARGET_REG_BITS == 32) { - tcg_out_qemu_st(s, args[0], -1, args[1], args[2], - args[3], TCG_TYPE_I32); - break; - } - /* fall through */ - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: + case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { tcg_out_qemu_st(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I64); @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, args[3], TCG_TYPE_I64); } break; - case INDEX_op_qemu_st_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, args[0], -1, args[1], -1, - args[2], 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_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); tcg_out_qemu_ldst_i128(s, args[0], args[1], args[2], args[3], false); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_sub2_i32: return C_O2_I4(r, r, rI, rZM, r, r); - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: return C_O1_I1(r, r); - case INDEX_op_qemu_ld_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O1_I2(r, r, r); - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I1(r, r, r); - case INDEX_op_qemu_ld_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I2(r, r, r, r); - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: return C_O0_I2(r, r); - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i64: return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I3(r, r, r); - case INDEX_op_qemu_st_a32_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I3(r, r, r); - case INDEX_op_qemu_st_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I4(r, r, r, r); - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: return C_N1O1_I1(o, m, r); - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: return C_O0_I3(o, m, r); case INDEX_op_add_vec: 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 void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, args[3], const_args[3], args[4], const_args[4]); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I64); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_sub2_i64: return C_O2_I4(r, r, rZ, rZ, rM, rM); - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: return C_O0_I2(rZ, r); case INDEX_op_st_vec: 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_op(TCGContext *s, TCGOpcode opc, TCGType type, args[2], const_args[2], args[3], const_args[3], args[4]); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, args[0], args[1], args[2], TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, args[0], args[1], args[2], TCG_TYPE_I64); break; - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: + 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_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, args[0], args[1], args[2], TCG_TYPE_I64); break; - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: tcg_out_qemu_ldst_i128(s, args[0], args[1], args[2], args[3], true); break; - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: tcg_out_qemu_ldst_i128(s, args[0], args[1], args[2], args[3], false); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_ctpop_i64: return C_O1_I1(r, r); - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i64: + case INDEX_op_qemu_st_i32: return C_O0_I2(r, r); - case INDEX_op_qemu_ld_a32_i128: - case INDEX_op_qemu_ld_a64_i128: + case INDEX_op_qemu_ld_i128: return C_O2_I1(o, m, r); - case INDEX_op_qemu_st_a32_i128: - case INDEX_op_qemu_st_a64_i128: + case INDEX_op_qemu_st_i128: return C_O0_I3(o, m, r); case INDEX_op_deposit_i32: 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_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out_arithi(s, a1, a0, 32, SHIFT_SRLX); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: + case INDEX_op_qemu_ld_i32: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i64: tcg_out_qemu_ld(s, a0, a1, a2, TCG_TYPE_I64); break; - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i32: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I32); break; - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i64: tcg_out_qemu_st(s, a0, a1, a2, TCG_TYPE_I64); break; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_extu_i32_i64: case INDEX_op_extract_i64: case INDEX_op_sextract_i64: - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_ld_a64_i64: + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); case INDEX_op_st8_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st_i32: case INDEX_op_st32_i64: case INDEX_op_st_i64: - case INDEX_op_qemu_st_a32_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_st_a32_i64: - case INDEX_op_qemu_st_a64_i64: + case INDEX_op_qemu_st_i32: + case INDEX_op_qemu_st_i64: return C_O0_I2(rZ, r); case INDEX_op_add_i32: diff --git a/tcg/tci/tcg-target.c.inc b/tcg/tci/tcg-target.c.inc index XXXXXXX..XXXXXXX 100644 --- a/tcg/tci/tcg-target.c.inc +++ b/tcg/tci/tcg-target.c.inc @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_setcond2_i32: return C_O1_I4(r, r, r, r, r); - case INDEX_op_qemu_ld_a32_i32: + case INDEX_op_qemu_ld_i32: return C_O1_I1(r, r); - case INDEX_op_qemu_ld_a64_i32: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O1_I2(r, r, r); - case INDEX_op_qemu_ld_a32_i64: + case INDEX_op_qemu_ld_i64: return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I1(r, r, r); - case INDEX_op_qemu_ld_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I2(r, r, r, r); - case INDEX_op_qemu_st_a32_i32: + case INDEX_op_qemu_st_i32: return C_O0_I2(r, r); - case INDEX_op_qemu_st_a64_i32: + case INDEX_op_qemu_st_i64: return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I3(r, r, r); - case INDEX_op_qemu_st_a32_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I3(r, r, r); - case INDEX_op_qemu_st_a64_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(r, r) : C_O0_I4(r, r, r, r); default: return C_NotImplemented; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op_rrrbb(TCGContext *s, TCGOpcode op, TCGReg r0, tcg_out32(s, insn); } -static void tcg_out_op_rrrrr(TCGContext *s, TCGOpcode op, TCGReg r0, - TCGReg r1, TCGReg r2, TCGReg r3, TCGReg r4) -{ - tcg_insn_unit insn = 0; - - insn = deposit32(insn, 0, 8, op); - insn = deposit32(insn, 8, 4, r0); - insn = deposit32(insn, 12, 4, r1); - insn = deposit32(insn, 16, 4, r2); - insn = deposit32(insn, 20, 4, r3); - insn = deposit32(insn, 24, 4, r4); - tcg_out32(s, insn); -} - static void tcg_out_op_rrrr(TCGContext *s, TCGOpcode op, TCGReg r0, TCGReg r1, TCGReg r2, TCGReg r3) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, tcg_out_op_rrrr(s, opc, args[0], args[1], args[2], args[3]); break; - case INDEX_op_qemu_ld_a32_i32: - case INDEX_op_qemu_st_a32_i32: - tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); - break; - case INDEX_op_qemu_ld_a64_i32: - case INDEX_op_qemu_st_a64_i32: - case INDEX_op_qemu_ld_a32_i64: - case INDEX_op_qemu_st_a32_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); - } else { + case INDEX_op_qemu_ld_i64: + case INDEX_op_qemu_st_i64: + if (TCG_TARGET_REG_BITS == 32) { tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_TMP, args[3]); tcg_out_op_rrrr(s, opc, args[0], args[1], args[2], TCG_REG_TMP); + break; } - break; - case INDEX_op_qemu_ld_a64_i64: - case INDEX_op_qemu_st_a64_i64: - if (TCG_TARGET_REG_BITS == 64) { - tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); + /* fall through */ + case INDEX_op_qemu_ld_i32: + case INDEX_op_qemu_st_i32: + if (TCG_TARGET_REG_BITS == 64 && s->addr_type == TCG_TYPE_I32) { + tcg_out_ext32u(s, TCG_REG_TMP, args[1]); + tcg_out_op_rrm(s, opc, args[0], TCG_REG_TMP, args[2]); } else { - tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_TMP, args[4]); - tcg_out_op_rrrrr(s, opc, args[0], args[1], - args[2], args[3], TCG_REG_TMP); + tcg_out_op_rrm(s, opc, args[0], args[1], args[2]); } break; -- 2.43.0 The guest address will now always be TCG_TYPE_I32. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/arm/tcg-target.c.inc | 73 +++++++++++++--------------------------- 1 file changed, 23 insertions(+), 50 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_ldrd_r(TCGContext *s, ARMCond cond, TCGReg rt, tcg_out_memop_r(s, cond, INSN_LDRD_REG, rt, rn, rm, 1, 1, 0); } -static void __attribute__((unused)) -tcg_out_ldrd_rwb(TCGContext *s, ARMCond cond, TCGReg rt, TCGReg rn, TCGReg rm) -{ - tcg_out_memop_r(s, cond, INSN_LDRD_REG, rt, rn, rm, 1, 1, 1); -} - -static void __attribute__((unused)) -tcg_out_strd_8(TCGContext *s, ARMCond cond, TCGReg rt, TCGReg rn, int imm8) +static void tcg_out_strd_8(TCGContext *s, ARMCond cond, TCGReg rt, + TCGReg rn, int imm8) { tcg_out_memop_8(s, cond, INSN_STRD_IMM, rt, rn, imm8, 1, 0); } @@ -XXX,XX +XXX,XX @@ static bool tcg_out_qemu_st_slow_path(TCGContext *s, TCGLabelQemuLdst *lb) #define MIN_TLB_MASK_TABLE_OFS -256 static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, bool is_ld) + TCGReg addr, MemOpIdx oi, bool is_ld) { TCGLabelQemuLdst *ldst = NULL; MemOp opc = get_memop(oi); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, if (tcg_use_softmmu) { *h = (HostAddress){ .cond = COND_AL, - .base = addrlo, + .base = addr, .index = TCG_REG_R1, .index_scratch = true, }; } else { *h = (HostAddress){ .cond = COND_AL, - .base = addrlo, + .base = addr, .index = guest_base ? TCG_REG_GUEST_BASE : -1, .index_scratch = false, }; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* Load cpu->neg.tlb.f[mmu_idx].{mask,table} into {r0,r1}. */ QEMU_BUILD_BUG_ON(offsetof(CPUTLBDescFast, mask) != 0); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, tcg_out_ldrd_8(s, COND_AL, TCG_REG_R0, TCG_AREG0, fast_off); /* Extract the tlb index from the address into R0. */ - tcg_out_dat_reg(s, COND_AL, ARITH_AND, TCG_REG_R0, TCG_REG_R0, addrlo, + tcg_out_dat_reg(s, COND_AL, ARITH_AND, TCG_REG_R0, TCG_REG_R0, addr, SHIFT_IMM_LSR(s->page_bits - CPU_TLB_ENTRY_BITS)); /* * Add the tlb_table pointer, creating the CPUTLBEntry address in R1. - * Load the tlb comparator into R2/R3 and the fast path addend into R1. + * Load the tlb comparator into R2 and the fast path addend into R1. */ QEMU_BUILD_BUG_ON(HOST_BIG_ENDIAN); if (cmp_off == 0) { - if (s->addr_type == TCG_TYPE_I32) { - tcg_out_ld32_rwb(s, COND_AL, TCG_REG_R2, - TCG_REG_R1, TCG_REG_R0); - } else { - tcg_out_ldrd_rwb(s, COND_AL, TCG_REG_R2, - TCG_REG_R1, TCG_REG_R0); - } + tcg_out_ld32_rwb(s, COND_AL, TCG_REG_R2, TCG_REG_R1, TCG_REG_R0); } else { tcg_out_dat_reg(s, COND_AL, ARITH_ADD, TCG_REG_R1, TCG_REG_R1, TCG_REG_R0, 0); - if (s->addr_type == TCG_TYPE_I32) { - tcg_out_ld32_12(s, COND_AL, TCG_REG_R2, TCG_REG_R1, cmp_off); - } else { - tcg_out_ldrd_8(s, COND_AL, TCG_REG_R2, TCG_REG_R1, cmp_off); - } + tcg_out_ld32_12(s, COND_AL, TCG_REG_R2, TCG_REG_R1, cmp_off); } /* Load the tlb addend. */ @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, * This leaves the least significant alignment bits unchanged, and of * course must be zero. */ - t_addr = addrlo; + t_addr = addr; if (a_mask < s_mask) { t_addr = TCG_REG_R0; tcg_out_dat_imm(s, COND_AL, ARITH_ADD, t_addr, - addrlo, s_mask - a_mask); + addr, s_mask - a_mask); } if (use_armv7_instructions && s->page_bits <= 16) { tcg_out_movi32(s, COND_AL, TCG_REG_TMP, ~(s->page_mask | a_mask)); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, } else { if (a_mask) { tcg_debug_assert(a_mask <= 0xff); - tcg_out_dat_imm(s, COND_AL, ARITH_TST, 0, addrlo, a_mask); + tcg_out_dat_imm(s, COND_AL, ARITH_TST, 0, addr, a_mask); } tcg_out_dat_reg(s, COND_AL, ARITH_MOV, TCG_REG_TMP, 0, t_addr, SHIFT_IMM_LSR(s->page_bits)); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, 0, TCG_REG_R2, TCG_REG_TMP, SHIFT_IMM_LSL(s->page_bits)); } - - if (s->addr_type != TCG_TYPE_I32) { - tcg_out_dat_reg(s, COND_EQ, ARITH_CMP, 0, TCG_REG_R3, addrhi, 0); - } } else if (a_mask) { ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* We are expecting alignment to max out at 7 */ tcg_debug_assert(a_mask <= 0xff); /* tst addr, #mask */ - tcg_out_dat_imm(s, COND_AL, ARITH_TST, 0, addrlo, a_mask); + tcg_out_dat_imm(s, COND_AL, ARITH_TST, 0, addr, a_mask); } return ldst; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, MemOp opc, TCGReg datalo, } static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, true); + ldst = prepare_host_addr(s, &h, addr, oi, true); if (ldst) { ldst->type = data_type; ldst->datalo_reg = datalo; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_direct(TCGContext *s, MemOp opc, TCGReg datalo, } static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, false); + ldst = prepare_host_addr(s, &h, addr, oi, false); if (ldst) { ldst->type = data_type; ldst->datalo_reg = datalo; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); + tcg_out_qemu_ld(s, args[0], -1, 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], -1, - args[3], TCG_TYPE_I64); + tcg_out_qemu_ld(s, args[0], args[1], args[2], args[3], TCG_TYPE_I64); break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); + tcg_out_qemu_st(s, args[0], -1, 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], -1, - args[3], TCG_TYPE_I64); + tcg_out_qemu_st(s, args[0], args[1], args[2], args[3], TCG_TYPE_I64); break; case INDEX_op_bswap16_i32: -- 2.43.0 The guest address will now always fit in one register. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 56 ++++++++++++++------------------------- 1 file changed, 20 insertions(+), 36 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 int setup_guest_base_seg(void) * is required and fill in @h with the host address for the fast path. */ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, bool is_ld) + TCGReg addr, MemOpIdx oi, bool is_ld) { TCGLabelQemuLdst *ldst = NULL; MemOp opc = get_memop(oi); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, } else { *h = x86_guest_base; } - h->base = addrlo; + h->base = addr; h->aa = atom_and_align_for_opc(s, opc, MO_ATOM_IFALIGN, s_bits == MO_128); a_mask = (1 << h->aa.align) - 1; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; if (TCG_TARGET_REG_BITS == 64) { ttype = s->addr_type; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, } } - tcg_out_mov(s, tlbtype, TCG_REG_L0, addrlo); + tcg_out_mov(s, tlbtype, TCG_REG_L0, addr); tcg_out_shifti(s, SHIFT_SHR + tlbrexw, TCG_REG_L0, s->page_bits - CPU_TLB_ENTRY_BITS); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, * check that we don't cross pages for the complete access. */ if (a_mask >= s_mask) { - tcg_out_mov(s, ttype, TCG_REG_L1, addrlo); + tcg_out_mov(s, ttype, TCG_REG_L1, addr); } else { tcg_out_modrm_offset(s, OPC_LEA + trexw, TCG_REG_L1, - addrlo, s_mask - a_mask); + addr, s_mask - a_mask); } tlb_mask = s->page_mask | a_mask; tgen_arithi(s, ARITH_AND + trexw, TCG_REG_L1, tlb_mask, 0); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst->label_ptr[0] = s->code_ptr; s->code_ptr += 4; - if (TCG_TARGET_REG_BITS == 32 && s->addr_type == TCG_TYPE_I64) { - /* cmp 4(TCG_REG_L0), addrhi */ - tcg_out_modrm_offset(s, OPC_CMP_GvEv, addrhi, - TCG_REG_L0, cmp_ofs + 4); - - /* jne slow_path */ - tcg_out_opc(s, OPC_JCC_long + JCC_JNE, 0, 0, 0); - ldst->label_ptr[1] = s->code_ptr; - s->code_ptr += 4; - } - /* TLB Hit. */ tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_L0, TCG_REG_L0, offsetof(CPUTLBEntry, addend)); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* jne slow_path */ - jcc = tcg_out_cmp(s, TCG_COND_TSTNE, addrlo, a_mask, true, false); + jcc = tcg_out_cmp(s, TCG_COND_TSTNE, addr, a_mask, true, false); tcg_out_opc(s, OPC_JCC_long + jcc, 0, 0, 0); ldst->label_ptr[0] = s->code_ptr; s->code_ptr += 4; @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld_direct(TCGContext *s, TCGReg datalo, TCGReg datahi, } static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, true); + ldst = prepare_host_addr(s, &h, addr, oi, true); tcg_out_qemu_ld_direct(s, datalo, datahi, h, data_type, get_memop(oi)); if (ldst) { @@ -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, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, false); + ldst = prepare_host_addr(s, &h, addr, oi, false); tcg_out_qemu_st_direct(s, datalo, datahi, h, get_memop(oi)); if (ldst) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); + tcg_out_qemu_ld(s, a0, -1, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); + tcg_out_qemu_ld(s, a0, -1, a1, a2, TCG_TYPE_I64); } else { - tcg_out_qemu_ld(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); + tcg_out_qemu_ld(s, a0, a1, a2, args[3], TCG_TYPE_I64); } break; case INDEX_op_qemu_ld_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); - tcg_out_qemu_ld(s, a0, a1, a2, -1, args[3], TCG_TYPE_I128); + tcg_out_qemu_ld(s, a0, a1, a2, args[3], TCG_TYPE_I128); break; case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st8_i32: - tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I32); + tcg_out_qemu_st(s, a0, -1, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, a0, -1, a1, -1, a2, TCG_TYPE_I64); + tcg_out_qemu_st(s, a0, -1, a1, a2, TCG_TYPE_I64); } else { - tcg_out_qemu_st(s, a0, a1, a2, -1, args[3], TCG_TYPE_I64); + tcg_out_qemu_st(s, a0, a1, a2, args[3], TCG_TYPE_I64); } break; case INDEX_op_qemu_st_i128: tcg_debug_assert(TCG_TARGET_REG_BITS == 64); - tcg_out_qemu_st(s, a0, a1, a2, -1, args[3], TCG_TYPE_I128); + tcg_out_qemu_st(s, a0, a1, a2, args[3], TCG_TYPE_I128); break; OP_32_64(mulu2): -- 2.43.0 The guest address will now always fit in one register. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/mips/tcg-target.c.inc | 62 ++++++++++++++------------------------- 1 file changed, 22 insertions(+), 40 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 @@ bool tcg_target_has_memory_bswap(MemOp memop) * is required and fill in @h with the host address for the fast path. */ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, bool is_ld) + TCGReg addr, MemOpIdx oi, bool is_ld) { TCGType addr_type = s->addr_type; TCGLabelQemuLdst *ldst = NULL; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */ tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP0, TCG_AREG0, mask_off); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, /* Extract the TLB index from the address into TMP3. */ if (TCG_TARGET_REG_BITS == 32 || addr_type == TCG_TYPE_I32) { - tcg_out_opc_sa(s, OPC_SRL, TCG_TMP3, addrlo, + tcg_out_opc_sa(s, OPC_SRL, TCG_TMP3, addr, s->page_bits - CPU_TLB_ENTRY_BITS); } else { - tcg_out_dsrl(s, TCG_TMP3, addrlo, - s->page_bits - CPU_TLB_ENTRY_BITS); + tcg_out_dsrl(s, TCG_TMP3, addr, s->page_bits - CPU_TLB_ENTRY_BITS); } tcg_out_opc_reg(s, OPC_AND, TCG_TMP3, TCG_TMP3, TCG_TMP0); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, tcg_out_opc_imm(s, (TCG_TARGET_REG_BITS == 32 || addr_type == TCG_TYPE_I32 ? OPC_ADDIU : OPC_DADDIU), - TCG_TMP2, addrlo, s_mask - a_mask); + TCG_TMP2, addr, s_mask - a_mask); tcg_out_opc_reg(s, OPC_AND, TCG_TMP1, TCG_TMP1, TCG_TMP2); } else { - tcg_out_opc_reg(s, OPC_AND, TCG_TMP1, TCG_TMP1, addrlo); + tcg_out_opc_reg(s, OPC_AND, TCG_TMP1, TCG_TMP1, addr); } /* Zero extend a 32-bit guest address for a 64-bit host. */ if (TCG_TARGET_REG_BITS == 64 && addr_type == TCG_TYPE_I32) { - tcg_out_ext32u(s, TCG_TMP2, addrlo); - addrlo = TCG_TMP2; + tcg_out_ext32u(s, TCG_TMP2, addr); + addr = TCG_TMP2; } ldst->label_ptr[0] = s->code_ptr; tcg_out_opc_br(s, OPC_BNE, TCG_TMP1, TCG_TMP0); - /* Load and test the high half tlb comparator. */ - if (TCG_TARGET_REG_BITS == 32 && addr_type != TCG_TYPE_I32) { - /* delay slot */ - tcg_out_ldst(s, OPC_LW, TCG_TMP0, TCG_TMP3, cmp_off + HI_OFF); - - /* Load the tlb addend for the fast path. */ - tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP3, TCG_TMP3, add_off); - - ldst->label_ptr[1] = s->code_ptr; - tcg_out_opc_br(s, OPC_BNE, addrhi, TCG_TMP0); - } - /* delay slot */ base = TCG_TMP3; - tcg_out_opc_reg(s, ALIAS_PADD, base, TCG_TMP3, addrlo); + tcg_out_opc_reg(s, ALIAS_PADD, base, TCG_TMP3, addr); } else { if (a_mask && (use_mips32r6_instructions || a_bits != s_bits)) { ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* We are expecting a_bits to max out at 7, much lower than ANDI. */ tcg_debug_assert(a_bits < 16); - tcg_out_opc_imm(s, OPC_ANDI, TCG_TMP0, addrlo, a_mask); + tcg_out_opc_imm(s, OPC_ANDI, TCG_TMP0, addr, a_mask); ldst->label_ptr[0] = s->code_ptr; if (use_mips32r6_instructions) { @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, } } - base = addrlo; + base = addr; if (TCG_TARGET_REG_BITS == 64 && addr_type == TCG_TYPE_I32) { tcg_out_ext32u(s, TCG_REG_A0, base); base = TCG_REG_A0; @@ -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, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, true); + ldst = prepare_host_addr(s, &h, addr, oi, true); if (use_mips32r6_instructions || h.aa.align >= (opc & MO_SIZE)) { tcg_out_qemu_ld_direct(s, datalo, datahi, h.base, opc, data_type); @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_st_unalign(TCGContext *s, TCGReg lo, TCGReg hi, } static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, false); + ldst = prepare_host_addr(s, &h, addr, oi, false); if (use_mips32r6_instructions || h.aa.align >= (opc & MO_SIZE)) { tcg_out_qemu_st_direct(s, datalo, datahi, h.base, opc); @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); + tcg_out_qemu_ld(s, a0, 0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); + tcg_out_qemu_ld(s, a0, 0, a1, a2, TCG_TYPE_I64); } else { - tcg_out_qemu_ld(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); + tcg_out_qemu_ld(s, a0, a1, a2, args[3], TCG_TYPE_I64); } break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I32); + tcg_out_qemu_st(s, a0, 0, a1, a2, TCG_TYPE_I32); break; case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, a0, 0, a1, 0, a2, TCG_TYPE_I64); + tcg_out_qemu_st(s, a0, 0, a1, a2, TCG_TYPE_I64); } else { - tcg_out_qemu_st(s, a0, a1, a2, 0, args[3], TCG_TYPE_I64); + tcg_out_qemu_st(s, a0, a1, a2, args[3], TCG_TYPE_I64); } break; -- 2.43.0 The guest address will now always fit in one register. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/ppc/tcg-target.c.inc | 75 ++++++++++++---------------------------- 1 file changed, 23 insertions(+), 52 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 @@ bool tcg_target_has_memory_bswap(MemOp memop) * is required and fill in @h with the host address for the fast path. */ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, bool is_ld) + TCGReg addr, MemOpIdx oi, bool is_ld) { TCGType addr_type = s->addr_type; TCGLabelQemuLdst *ldst = NULL; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */ tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, TCG_AREG0, mask_off); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, /* Extract the page index, shifted into place for tlb index. */ if (TCG_TARGET_REG_BITS == 32) { - tcg_out_shri32(s, TCG_REG_R0, addrlo, + tcg_out_shri32(s, TCG_REG_R0, addr, s->page_bits - CPU_TLB_ENTRY_BITS); } else { - tcg_out_shri64(s, TCG_REG_R0, addrlo, + tcg_out_shri64(s, TCG_REG_R0, addr, s->page_bits - CPU_TLB_ENTRY_BITS); } tcg_out32(s, AND | SAB(TCG_REG_TMP1, TCG_REG_TMP1, TCG_REG_R0)); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, if (a_bits < s_bits) { a_bits = s_bits; } - tcg_out_rlw(s, RLWINM, TCG_REG_R0, addrlo, 0, + tcg_out_rlw(s, RLWINM, TCG_REG_R0, addr, 0, (32 - a_bits) & 31, 31 - s->page_bits); } else { - TCGReg t = addrlo; + TCGReg t = addr; /* * If the access is unaligned, we need to make sure we fail if we @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, } } - if (TCG_TARGET_REG_BITS == 32 && addr_type != TCG_TYPE_I32) { - /* Low part comparison into cr7. */ - tcg_out_cmp(s, TCG_COND_EQ, TCG_REG_R0, TCG_REG_TMP2, - 0, 7, TCG_TYPE_I32); - - /* Load the high part TLB comparator into TMP2. */ - tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_TMP2, TCG_REG_TMP1, - cmp_off + 4 * !HOST_BIG_ENDIAN); - - /* Load addend, deferred for this case. */ - tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, TCG_REG_TMP1, - offsetof(CPUTLBEntry, addend)); - - /* High part comparison into cr6. */ - tcg_out_cmp(s, TCG_COND_EQ, addrhi, TCG_REG_TMP2, - 0, 6, TCG_TYPE_I32); - - /* Combine comparisons into cr0. */ - tcg_out32(s, CRAND | BT(0, CR_EQ) | BA(6, CR_EQ) | BB(7, CR_EQ)); - } else { - /* Full comparison into cr0. */ - tcg_out_cmp(s, TCG_COND_EQ, TCG_REG_R0, TCG_REG_TMP2, - 0, 0, addr_type); - } + /* Full comparison into cr0. */ + tcg_out_cmp(s, TCG_COND_EQ, TCG_REG_R0, TCG_REG_TMP2, 0, 0, addr_type); /* Load a pointer into the current opcode w/conditional branch-link. */ ldst->label_ptr[0] = s->code_ptr; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addrlo; - ldst->addrhi_reg = addrhi; + ldst->addrlo_reg = addr; /* We are expecting a_bits to max out at 7, much lower than ANDI. */ tcg_debug_assert(a_bits < 16); - tcg_out32(s, ANDI | SAI(addrlo, TCG_REG_R0, (1 << a_bits) - 1)); + tcg_out32(s, ANDI | SAI(addr, TCG_REG_R0, (1 << a_bits) - 1)); ldst->label_ptr[0] = s->code_ptr; tcg_out32(s, BC | BI(0, CR_EQ) | BO_COND_FALSE | LK); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, if (TCG_TARGET_REG_BITS == 64 && addr_type == TCG_TYPE_I32) { /* Zero-extend the guest address for use in the host address. */ - tcg_out_ext32u(s, TCG_REG_TMP2, addrlo); + tcg_out_ext32u(s, TCG_REG_TMP2, addr); h->index = TCG_REG_TMP2; } else { - h->index = addrlo; + h->index = addr; } return ldst; } static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, true); + ldst = prepare_host_addr(s, &h, addr, oi, true); if (TCG_TARGET_REG_BITS == 32 && (opc & MO_SIZE) == MO_64) { if (opc & MO_BSWAP) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ld(TCGContext *s, TCGReg datalo, TCGReg datahi, } static void tcg_out_qemu_st(TCGContext *s, TCGReg datalo, TCGReg datahi, - TCGReg addrlo, TCGReg addrhi, - MemOpIdx oi, TCGType data_type) + TCGReg addr, MemOpIdx oi, TCGType data_type) { MemOp opc = get_memop(oi); TCGLabelQemuLdst *ldst; HostAddress h; - ldst = prepare_host_addr(s, &h, addrlo, addrhi, oi, false); + ldst = prepare_host_addr(s, &h, addr, oi, false); if (TCG_TARGET_REG_BITS == 32 && (opc & MO_SIZE) == MO_64) { if (opc & MO_BSWAP) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_qemu_ldst_i128(TCGContext *s, TCGReg datalo, TCGReg datahi, uint32_t insn; TCGReg index; - ldst = prepare_host_addr(s, &h, addr_reg, -1, oi, is_ld); + ldst = prepare_host_addr(s, &h, addr_reg, oi, is_ld); /* Compose the final address, as LQ/STQ have no indexing. */ index = h.index; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, break; case INDEX_op_qemu_ld_i32: - tcg_out_qemu_ld(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); + tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], TCG_TYPE_I32); break; case INDEX_op_qemu_ld_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_ld(s, args[0], -1, args[1], -1, - args[2], TCG_TYPE_I64); + tcg_out_qemu_ld(s, args[0], -1, args[1], args[2], TCG_TYPE_I64); } else { - tcg_out_qemu_ld(s, args[0], args[1], args[2], -1, + tcg_out_qemu_ld(s, args[0], args[1], args[2], args[3], TCG_TYPE_I64); } break; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, break; case INDEX_op_qemu_st_i32: - tcg_out_qemu_st(s, args[0], -1, args[1], -1, args[2], TCG_TYPE_I32); + tcg_out_qemu_st(s, args[0], -1, args[1], args[2], TCG_TYPE_I32); break; case INDEX_op_qemu_st_i64: if (TCG_TARGET_REG_BITS == 64) { - tcg_out_qemu_st(s, args[0], -1, args[1], -1, - args[2], TCG_TYPE_I64); + tcg_out_qemu_st(s, args[0], -1, args[1], args[2], TCG_TYPE_I64); } else { - tcg_out_qemu_st(s, args[0], args[1], args[2], -1, + tcg_out_qemu_st(s, args[0], args[1], args[2], args[3], TCG_TYPE_I64); } break; -- 2.43.0 There is now always only one guest address register. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/tcg.c | 18 +++++++++--------- tcg/aarch64/tcg-target.c.inc | 4 ++-- tcg/arm/tcg-target.c.inc | 4 ++-- tcg/i386/tcg-target.c.inc | 4 ++-- tcg/loongarch64/tcg-target.c.inc | 4 ++-- tcg/mips/tcg-target.c.inc | 4 ++-- tcg/ppc/tcg-target.c.inc | 4 ++-- tcg/riscv/tcg-target.c.inc | 4 ++-- tcg/s390x/tcg-target.c.inc | 4 ++-- tcg/sparc64/tcg-target.c.inc | 4 ++-- 10 files changed, 27 insertions(+), 27 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 @@ 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 addr_reg; /* reg index for 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 */ @@ -XXX,XX +XXX,XX @@ static void tcg_out_ld_helper_args(TCGContext *s, const TCGLabelQemuLdst *ldst, */ tcg_out_helper_add_mov(mov, loc + HOST_BIG_ENDIAN, TCG_TYPE_I32, TCG_TYPE_I32, - ldst->addrlo_reg, -1); + ldst->addr_reg, -1); tcg_out_helper_load_slots(s, 1, mov, parm); tcg_out_helper_load_imm(s, loc[!HOST_BIG_ENDIAN].arg_slot, @@ -XXX,XX +XXX,XX @@ static void tcg_out_ld_helper_args(TCGContext *s, const TCGLabelQemuLdst *ldst, next_arg += 2; } else { nmov = tcg_out_helper_add_mov(mov, loc, TCG_TYPE_I64, s->addr_type, - ldst->addrlo_reg, ldst->addrhi_reg); + ldst->addr_reg, -1); tcg_out_helper_load_slots(s, nmov, mov, parm); next_arg += nmov; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_st_helper_args(TCGContext *s, const TCGLabelQemuLdst *ldst, /* Handle addr argument. */ loc = &info->in[next_arg]; - if (TCG_TARGET_REG_BITS == 32 && s->addr_type == TCG_TYPE_I32) { + tcg_debug_assert(s->addr_type <= TCG_TYPE_REG); + if (TCG_TARGET_REG_BITS == 32) { /* - * 32-bit host with 32-bit guest: zero-extend the guest address + * 32-bit host (and thus 32-bit guest): zero-extend the guest address * to 64-bits for the helper by storing the low part. Later, * after we have processed the register inputs, we will load a * zero for the high part. */ tcg_out_helper_add_mov(mov, loc + HOST_BIG_ENDIAN, TCG_TYPE_I32, TCG_TYPE_I32, - ldst->addrlo_reg, -1); + ldst->addr_reg, -1); next_arg += 2; nmov += 1; } else { n = tcg_out_helper_add_mov(mov, loc, TCG_TYPE_I64, s->addr_type, - ldst->addrlo_reg, ldst->addrhi_reg); + ldst->addr_reg, -1); next_arg += n; nmov += n; } @@ -XXX,XX +XXX,XX @@ static void tcg_out_st_helper_args(TCGContext *s, const TCGLabelQemuLdst *ldst, g_assert_not_reached(); } - if (TCG_TARGET_REG_BITS == 32 && s->addr_type == TCG_TYPE_I32) { + if (TCG_TARGET_REG_BITS == 32) { /* Zero extend the address by loading a zero for the high part. */ loc = &info->in[1 + !HOST_BIG_ENDIAN]; tcg_out_helper_load_imm(s, loc->arg_slot, TCG_TYPE_I32, 0, parm); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; mask_type = (s->page_bits + s->tlb_dyn_max_bits > 32 ? TCG_TYPE_I64 : TCG_TYPE_I32); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; /* tst addr, #mask */ tcg_out_logicali(s, I3404_ANDSI, 0, TCG_REG_XZR, addr_reg, a_mask); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* Load cpu->neg.tlb.f[mmu_idx].{mask,table} into {r0,r1}. */ QEMU_BUILD_BUG_ON(offsetof(CPUTLBDescFast, mask) != 0); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* We are expecting alignment to max out at 7 */ tcg_debug_assert(a_mask <= 0xff); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; if (TCG_TARGET_REG_BITS == 64) { ttype = s->addr_type; @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* jne slow_path */ jcc = tcg_out_cmp(s, TCG_COND_TSTNE, addr, a_mask, true, false); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP0, TCG_AREG0, mask_ofs); tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, TCG_AREG0, table_ofs); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; /* * Without micro-architecture details, we don't know which of 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */ tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP0, TCG_AREG0, mask_off); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* We are expecting a_bits to max out at 7, much lower than ANDI. */ tcg_debug_assert(a_bits < 16); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* Load tlb_mask[mmu_idx] and tlb_table[mmu_idx]. */ tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP1, TCG_AREG0, mask_off); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr; + ldst->addr_reg = addr; /* We are expecting a_bits to max out at 7, much lower than ANDI. */ tcg_debug_assert(a_bits < 16); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, TCGReg *pbase, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; init_setting_vtype(s); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, TCGReg *pbase, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; init_setting_vtype(s); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; tcg_out_sh64(s, RSY_SRLG, TCG_TMP0, addr_reg, TCG_REG_NONE, s->page_bits - CPU_TLB_ENTRY_BITS); @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; tcg_debug_assert(a_mask <= 0xffff); tcg_out_insn(s, RI, TMLL, addr_reg, a_mask); 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; ldst->label_ptr[0] = s->code_ptr; /* bne,pn %[xi]cc, label0 */ @@ -XXX,XX +XXX,XX @@ static TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, ldst = new_ldst_label(s); ldst->is_ld = is_ld; ldst->oi = oi; - ldst->addrlo_reg = addr_reg; + ldst->addr_reg = addr_reg; ldst->label_ptr[0] = s->code_ptr; /* bne,pn %icc, label0 */ -- 2.43.0 The declaration uses uint64_t for addr. Fixes: 595cd9ce2ec ("plugins: add plugin API to read guest memory") Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- plugins/api.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/plugins/api.c b/plugins/api.c index XXXXXXX..XXXXXXX 100644 --- a/plugins/api.c +++ b/plugins/api.c @@ -XXX,XX +XXX,XX @@ GArray *qemu_plugin_get_registers(void) return create_register_handles(regs); } -bool qemu_plugin_read_memory_vaddr(vaddr addr, GByteArray *data, size_t len) +bool qemu_plugin_read_memory_vaddr(uint64_t addr, GByteArray *data, size_t len) { g_assert(current_cpu); -- 2.43.0 The declarations use vaddr for size. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- accel/tcg/cputlb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 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 @@ void tlb_set_page_full(CPUState *cpu, int mmu_idx, void tlb_set_page_with_attrs(CPUState *cpu, vaddr addr, hwaddr paddr, MemTxAttrs attrs, int prot, - int mmu_idx, uint64_t size) + int mmu_idx, vaddr size) { CPUTLBEntryFull full = { .phys_addr = paddr, @@ -XXX,XX +XXX,XX @@ void tlb_set_page_with_attrs(CPUState *cpu, vaddr addr, void tlb_set_page(CPUState *cpu, vaddr addr, hwaddr paddr, int prot, - int mmu_idx, uint64_t size) + int mmu_idx, vaddr size) { tlb_set_page_with_attrs(cpu, addr, paddr, MEMTXATTRS_UNSPECIFIED, prot, mmu_idx, size); -- 2.43.0 DisasContextBase.pc_next has type vaddr; use the correct log format. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- target/loongarch/tcg/translate.c | 2 +- target/loongarch/tcg/insn_trans/trans_atomic.c.inc | 2 +- 2 files changed, 2 insertions(+), 2 deletions(-) diff --git a/target/loongarch/tcg/translate.c b/target/loongarch/tcg/translate.c index XXXXXXX..XXXXXXX 100644 --- a/target/loongarch/tcg/translate.c +++ b/target/loongarch/tcg/translate.c @@ -XXX,XX +XXX,XX @@ static void loongarch_tr_translate_insn(DisasContextBase *dcbase, CPUState *cs) if (!decode(ctx, ctx->opcode)) { qemu_log_mask(LOG_UNIMP, "Error: unknown opcode. " - TARGET_FMT_lx ": 0x%x\n", + "0x%" VADDR_PRIx ": 0x%x\n", ctx->base.pc_next, ctx->opcode); generate_exception(ctx, EXCCODE_INE); } diff --git a/target/loongarch/tcg/insn_trans/trans_atomic.c.inc b/target/loongarch/tcg/insn_trans/trans_atomic.c.inc index XXXXXXX..XXXXXXX 100644 --- a/target/loongarch/tcg/insn_trans/trans_atomic.c.inc +++ b/target/loongarch/tcg/insn_trans/trans_atomic.c.inc @@ -XXX,XX +XXX,XX @@ static bool gen_am(DisasContext *ctx, arg_rrr *a, if (a->rd != 0 && (a->rj == a->rd || a->rk == a->rd)) { qemu_log_mask(LOG_GUEST_ERROR, "Warning: source register overlaps destination register" - "in atomic insn at pc=0x" TARGET_FMT_lx "\n", + "in atomic insn at pc=0x%" VADDR_PRIx "\n", ctx->base.pc_next - 4); return false; } -- 2.43.0 Since we no longer support 64-bit guests on 32-bit hosts, we can use a 32-bit type on a 32-bit host. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/vaddr.h | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) diff --git a/include/exec/vaddr.h b/include/exec/vaddr.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/vaddr.h +++ b/include/exec/vaddr.h @@ -XXX,XX +XXX,XX @@ /** * vaddr: * Type wide enough to contain any #target_ulong virtual address. + * We do not support 64-bit guest on 32-host and detect at configure time. + * Therefore, a host pointer width will always fit a guest pointer. */ -typedef uint64_t vaddr; -#define VADDR_PRId PRId64 -#define VADDR_PRIu PRIu64 -#define VADDR_PRIo PRIo64 -#define VADDR_PRIx PRIx64 -#define VADDR_PRIX PRIX64 -#define VADDR_MAX UINT64_MAX +typedef uintptr_t vaddr; +#define VADDR_PRId PRIdPTR +#define VADDR_PRIu PRIuPTR +#define VADDR_PRIo PRIoPTR +#define VADDR_PRIx PRIxPTR +#define VADDR_PRIX PRIXPTR +#define VADDR_MAX UINTPTR_MAX #endif -- 2.43.0 Since we no longer support 64-bit guests on 32-bit hosts, we can use a 32-bit type on a 32-bit host. This shrinks the size of the structure to 16 bytes on a 32-bit host. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/exec/tlb-common.h | 10 +++++----- accel/tcg/cputlb.c | 21 ++++----------------- tcg/arm/tcg-target.c.inc | 1 - tcg/mips/tcg-target.c.inc | 12 +++++------- tcg/ppc/tcg-target.c.inc | 21 +++++---------------- 5 files changed, 19 insertions(+), 46 deletions(-) diff --git a/include/exec/tlb-common.h b/include/exec/tlb-common.h index XXXXXXX..XXXXXXX 100644 --- a/include/exec/tlb-common.h +++ b/include/exec/tlb-common.h @@ -XXX,XX +XXX,XX @@ #ifndef EXEC_TLB_COMMON_H #define EXEC_TLB_COMMON_H 1 -#define CPU_TLB_ENTRY_BITS 5 +#define CPU_TLB_ENTRY_BITS (HOST_LONG_BITS == 32 ? 4 : 5) /* Minimalized TLB entry for use by TCG fast path. */ typedef union CPUTLBEntry { struct { - uint64_t addr_read; - uint64_t addr_write; - uint64_t addr_code; + uintptr_t addr_read; + uintptr_t addr_write; + uintptr_t addr_code; /* * Addend to virtual address to get host address. IO accesses * use the corresponding iotlb value. @@ -XXX,XX +XXX,XX @@ typedef union CPUTLBEntry { * Padding to get a power of two size, as well as index * access to addr_{read,write,code}. */ - uint64_t addr_idx[(1 << CPU_TLB_ENTRY_BITS) / sizeof(uint64_t)]; + uintptr_t addr_idx[(1 << CPU_TLB_ENTRY_BITS) / sizeof(uintptr_t)]; } CPUTLBEntry; QEMU_BUILD_BUG_ON(sizeof(CPUTLBEntry) != (1 << CPU_TLB_ENTRY_BITS)); 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 inline uint64_t tlb_read_idx(const CPUTLBEntry *entry, { /* Do not rearrange the CPUTLBEntry structure members. */ QEMU_BUILD_BUG_ON(offsetof(CPUTLBEntry, addr_read) != - MMU_DATA_LOAD * sizeof(uint64_t)); + MMU_DATA_LOAD * sizeof(uintptr_t)); QEMU_BUILD_BUG_ON(offsetof(CPUTLBEntry, addr_write) != - MMU_DATA_STORE * sizeof(uint64_t)); + MMU_DATA_STORE * sizeof(uintptr_t)); QEMU_BUILD_BUG_ON(offsetof(CPUTLBEntry, addr_code) != - MMU_INST_FETCH * sizeof(uint64_t)); + MMU_INST_FETCH * sizeof(uintptr_t)); -#if TARGET_LONG_BITS == 32 - /* Use qatomic_read, in case of addr_write; only care about low bits. */ - const uint32_t *ptr = (uint32_t *)&entry->addr_idx[access_type]; - ptr += HOST_BIG_ENDIAN; - return qatomic_read(ptr); -#else - const uint64_t *ptr = &entry->addr_idx[access_type]; + const uintptr_t *ptr = &entry->addr_idx[access_type]; /* ofs might correspond to .addr_write, so use qatomic_read */ return qatomic_read(ptr); -#endif } static inline uint64_t tlb_addr_write(const CPUTLBEntry *entry) @@ -XXX,XX +XXX,XX @@ static void tlb_reset_dirty_range_locked(CPUTLBEntry *tlb_entry, addr &= TARGET_PAGE_MASK; addr += tlb_entry->addend; if ((addr - start) < length) { -#if TARGET_LONG_BITS == 32 - uint32_t *ptr_write = (uint32_t *)&tlb_entry->addr_write; - ptr_write += HOST_BIG_ENDIAN; - qatomic_set(ptr_write, *ptr_write | TLB_NOTDIRTY); -#else qatomic_set(&tlb_entry->addr_write, tlb_entry->addr_write | TLB_NOTDIRTY); -#endif } } } 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, * Add the tlb_table pointer, creating the CPUTLBEntry address in R1. * Load the tlb comparator into R2 and the fast path addend into R1. */ - QEMU_BUILD_BUG_ON(HOST_BIG_ENDIAN); if (cmp_off == 0) { tcg_out_ld32_rwb(s, COND_AL, TCG_REG_R2, TCG_REG_R1, TCG_REG_R0); } else { 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, /* Add the tlb_table pointer, creating the CPUTLBEntry address. */ tcg_out_opc_reg(s, ALIAS_PADD, TCG_TMP3, TCG_TMP3, TCG_TMP1); - if (TCG_TARGET_REG_BITS == 32 || addr_type == TCG_TYPE_I32) { - /* Load the (low half) tlb comparator. */ + /* Load the tlb comparator. */ + if (TCG_TARGET_REG_BITS == 64 && addr_type == TCG_TYPE_I32) { tcg_out_ld(s, TCG_TYPE_I32, TCG_TMP0, TCG_TMP3, cmp_off + HOST_BIG_ENDIAN * 4); } else { - tcg_out_ld(s, TCG_TYPE_I64, TCG_TMP0, TCG_TMP3, cmp_off); + tcg_out_ld(s, TCG_TYPE_REG, TCG_TMP0, TCG_TMP3, cmp_off); } - if (TCG_TARGET_REG_BITS == 64 || addr_type == TCG_TYPE_I32) { - /* Load the tlb addend for the fast path. */ - tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP3, TCG_TMP3, add_off); - } + /* Load the tlb addend for the fast path. */ + tcg_out_ld(s, TCG_TYPE_PTR, TCG_TMP3, TCG_TMP3, add_off); /* * Mask the page bits, keeping the alignment bits to compare against. 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 TCGLabelQemuLdst *prepare_host_addr(TCGContext *s, HostAddress *h, tcg_out32(s, AND | SAB(TCG_REG_TMP1, TCG_REG_TMP1, TCG_REG_R0)); /* - * Load the (low part) TLB comparator into TMP2. + * Load the TLB comparator into TMP2. * For 64-bit host, always load the entire 64-bit slot for simplicity. * We will ignore the high bits with tcg_out_cmp(..., addr_type). */ - if (TCG_TARGET_REG_BITS == 64) { - if (cmp_off == 0) { - tcg_out32(s, LDUX | TAB(TCG_REG_TMP2, - TCG_REG_TMP1, TCG_REG_TMP2)); - } else { - tcg_out32(s, ADD | TAB(TCG_REG_TMP1, - TCG_REG_TMP1, TCG_REG_TMP2)); - tcg_out_ld(s, TCG_TYPE_I64, TCG_REG_TMP2, - TCG_REG_TMP1, cmp_off); - } - } else if (cmp_off == 0 && !HOST_BIG_ENDIAN) { - tcg_out32(s, LWZUX | TAB(TCG_REG_TMP2, - TCG_REG_TMP1, TCG_REG_TMP2)); + if (cmp_off == 0) { + tcg_out32(s, (TCG_TARGET_REG_BITS == 64 ? LDUX : LWZUX) + | TAB(TCG_REG_TMP2, TCG_REG_TMP1, TCG_REG_TMP2)); } else { tcg_out32(s, ADD | TAB(TCG_REG_TMP1, TCG_REG_TMP1, TCG_REG_TMP2)); - tcg_out_ld(s, TCG_TYPE_I32, TCG_REG_TMP2, TCG_REG_TMP1, - cmp_off + 4 * HOST_BIG_ENDIAN); + tcg_out_ld(s, TCG_TYPE_PTR, TCG_REG_TMP2, TCG_REG_TMP1, cmp_off); } /* -- 2.43.0 For loongarch, mips, riscv and sparc, a zero register is available all the time. For aarch64, register index 31 depends on context: sometimes it is the stack pointer, and sometimes it is the zero register. Introduce a new general-purpose constraint which maps 0 to TCG_REG_ZERO, if defined. This differs from existing constant constraints in that const_arg[*] is recorded as false, indicating that the value is in a register. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- include/tcg/tcg.h | 3 ++- tcg/aarch64/tcg-target.h | 2 ++ tcg/loongarch64/tcg-target.h | 2 ++ tcg/mips/tcg-target.h | 2 ++ tcg/riscv/tcg-target.h | 2 ++ tcg/sparc64/tcg-target.h | 3 ++- tcg/tcg.c | 29 ++++++++++++++++++++++------- docs/devel/tcg-ops.rst | 4 +++- 8 files changed, 37 insertions(+), 10 deletions(-) diff --git a/include/tcg/tcg.h b/include/tcg/tcg.h index XXXXXXX..XXXXXXX 100644 --- a/include/tcg/tcg.h +++ b/include/tcg/tcg.h @@ -XXX,XX +XXX,XX @@ void tb_target_set_jmp_target(const TranslationBlock *, int, void tcg_set_frame(TCGContext *s, TCGReg reg, intptr_t start, intptr_t size); -#define TCG_CT_CONST 1 /* any constant of register size */ +#define TCG_CT_CONST 1 /* any constant of register size */ +#define TCG_CT_REG_ZERO 2 /* zero, in TCG_REG_ZERO */ typedef struct TCGArgConstraint { unsigned ct : 16; diff --git a/tcg/aarch64/tcg-target.h b/tcg/aarch64/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/aarch64/tcg-target.h +++ b/tcg/aarch64/tcg-target.h @@ -XXX,XX +XXX,XX @@ typedef enum { TCG_AREG0 = TCG_REG_X19, } TCGReg; +#define TCG_REG_ZERO TCG_REG_XZR + #define TCG_TARGET_NB_REGS 64 #endif /* AARCH64_TCG_TARGET_H */ diff --git a/tcg/loongarch64/tcg-target.h b/tcg/loongarch64/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target.h +++ b/tcg/loongarch64/tcg-target.h @@ -XXX,XX +XXX,XX @@ typedef enum { TCG_VEC_TMP0 = TCG_REG_V23, } TCGReg; +#define TCG_REG_ZERO TCG_REG_ZERO + #endif /* LOONGARCH_TCG_TARGET_H */ diff --git a/tcg/mips/tcg-target.h b/tcg/mips/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/mips/tcg-target.h +++ b/tcg/mips/tcg-target.h @@ -XXX,XX +XXX,XX @@ typedef enum { TCG_AREG0 = TCG_REG_S8, } TCGReg; +#define TCG_REG_ZERO TCG_REG_ZERO + #endif 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 @@ typedef enum { TCG_REG_TMP2 = TCG_REG_T4, } TCGReg; +#define TCG_REG_ZERO TCG_REG_ZERO + #endif diff --git a/tcg/sparc64/tcg-target.h b/tcg/sparc64/tcg-target.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/sparc64/tcg-target.h +++ b/tcg/sparc64/tcg-target.h @@ -XXX,XX +XXX,XX @@ typedef enum { TCG_REG_I7, } TCGReg; -#define TCG_AREG0 TCG_REG_I0 +#define TCG_AREG0 TCG_REG_I0 +#define TCG_REG_ZERO TCG_REG_G0 #endif 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 process_constraint_sets(void) case 'i': args_ct[i].ct |= TCG_CT_CONST; break; +#ifdef TCG_REG_ZERO + case 'z': + args_ct[i].ct |= TCG_CT_REG_ZERO; + break; +#endif /* Include all of the target-specific constraints. */ @@ -XXX,XX +XXX,XX @@ static void tcg_reg_alloc_op(TCGContext *s, const TCGOp *op) arg_ct = &args_ct[i]; ts = arg_temp(arg); - if (ts->val_type == TEMP_VAL_CONST - && tcg_target_const_match(ts->val, arg_ct->ct, ts->type, - op_cond, TCGOP_VECE(op))) { - /* constant is OK for instruction */ - const_args[i] = 1; - new_args[i] = ts->val; - continue; + if (ts->val_type == TEMP_VAL_CONST) { +#ifdef TCG_REG_ZERO + if (ts->val == 0 && (arg_ct->ct & TCG_CT_REG_ZERO)) { + /* Hardware zero register: indicate register via non-const. */ + const_args[i] = 0; + new_args[i] = TCG_REG_ZERO; + continue; + } +#endif + + if (tcg_target_const_match(ts->val, arg_ct->ct, ts->type, + op_cond, TCGOP_VECE(op))) { + /* constant is OK for instruction */ + const_args[i] = 1; + new_args[i] = ts->val; + continue; + } } reg = ts->reg; diff --git a/docs/devel/tcg-ops.rst b/docs/devel/tcg-ops.rst index XXXXXXX..XXXXXXX 100644 --- a/docs/devel/tcg-ops.rst +++ b/docs/devel/tcg-ops.rst @@ -XXX,XX +XXX,XX @@ operation uses a constant input constraint which does not allow all constants, it must also accept registers in order to have a fallback. The constraint '``i``' is defined generically to accept any constant. The constraint '``r``' is not defined generically, but is consistently -used by each backend to indicate all registers. +used by each backend to indicate all registers. If ``TCG_REG_ZERO`` +is defined by the backend, the constraint '``z``' is defined generically +to map constant 0 to the hardware zero register. The movi_i32 and movi_i64 operations must accept any constants. -- 2.43.0 Note that 'Z' is still used for addsub2. Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/aarch64/tcg-target-con-set.h | 12 ++++----- tcg/aarch64/tcg-target.c.inc | 46 ++++++++++++++------------------ 2 files changed, 26 insertions(+), 32 deletions(-) diff --git a/tcg/aarch64/tcg-target-con-set.h b/tcg/aarch64/tcg-target-con-set.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/aarch64/tcg-target-con-set.h +++ b/tcg/aarch64/tcg-target-con-set.h @@ -XXX,XX +XXX,XX @@ */ C_O0_I1(r) C_O0_I2(r, rC) -C_O0_I2(rZ, r) +C_O0_I2(rz, r) C_O0_I2(w, r) -C_O0_I3(rZ, rZ, r) +C_O0_I3(rz, rz, r) C_O1_I1(r, r) C_O1_I1(w, r) C_O1_I1(w, w) C_O1_I1(w, wr) -C_O1_I2(r, 0, rZ) +C_O1_I2(r, 0, rz) C_O1_I2(r, r, r) C_O1_I2(r, r, rA) C_O1_I2(r, r, rAL) C_O1_I2(r, r, rC) C_O1_I2(r, r, ri) C_O1_I2(r, r, rL) -C_O1_I2(r, rZ, rZ) +C_O1_I2(r, rz, rz) C_O1_I2(w, 0, w) C_O1_I2(w, w, w) C_O1_I2(w, w, wN) C_O1_I2(w, w, wO) C_O1_I2(w, w, wZ) C_O1_I3(w, w, w, w) -C_O1_I4(r, r, rC, rZ, rZ) +C_O1_I4(r, r, rC, rz, rz) C_O2_I1(r, r, r) -C_O2_I4(r, r, rZ, rZ, rA, rMZ) +C_O2_I4(r, r, rz, rz, rA, rMZ) 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_op(TCGContext *s, TCGOpcode opc, TCGType ext, TCGArg a2 = args[2]; int c2 = const_args[2]; - /* Some operands are defined with "rZ" constraint, a register or - the zero register. These need not actually test args[I] == 0. */ -#define REG0(I) (const_args[I] ? TCG_REG_XZR : (TCGReg)args[I]) - switch (opc) { case INDEX_op_goto_ptr: tcg_out_insn(s, 3207, BR, a0); @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, case INDEX_op_st8_i32: case INDEX_op_st8_i64: - tcg_out_ldst(s, I3312_STRB, REG0(0), a1, a2, 0); + tcg_out_ldst(s, I3312_STRB, a0, a1, a2, 0); break; case INDEX_op_st16_i32: case INDEX_op_st16_i64: - tcg_out_ldst(s, I3312_STRH, REG0(0), a1, a2, 1); + tcg_out_ldst(s, I3312_STRH, a0, a1, a2, 1); break; case INDEX_op_st_i32: case INDEX_op_st32_i64: - tcg_out_ldst(s, I3312_STRW, REG0(0), a1, a2, 2); + tcg_out_ldst(s, I3312_STRW, a0, a1, a2, 2); break; case INDEX_op_st_i64: - tcg_out_ldst(s, I3312_STRX, REG0(0), a1, a2, 3); + tcg_out_ldst(s, I3312_STRX, a0, a1, a2, 3); break; case INDEX_op_add_i32: @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, /* FALLTHRU */ case INDEX_op_movcond_i64: tcg_out_cmp(s, ext, args[5], a1, a2, c2); - tcg_out_insn(s, 3506, CSEL, ext, a0, REG0(3), REG0(4), args[5]); + tcg_out_insn(s, 3506, CSEL, ext, a0, args[3], args[4], args[5]); break; case INDEX_op_qemu_ld_i32: @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, break; case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - tcg_out_qemu_st(s, REG0(0), a1, a2, ext); + tcg_out_qemu_st(s, a0, a1, a2, ext); break; case INDEX_op_qemu_ld_i128: tcg_out_qemu_ldst_i128(s, a0, a1, a2, args[3], true); break; case INDEX_op_qemu_st_i128: - tcg_out_qemu_ldst_i128(s, REG0(0), REG0(1), a2, args[3], false); + tcg_out_qemu_ldst_i128(s, a0, a1, a2, args[3], false); break; case INDEX_op_bswap64_i64: @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, case INDEX_op_deposit_i64: case INDEX_op_deposit_i32: - tcg_out_dep(s, ext, a0, REG0(2), args[3], args[4]); + tcg_out_dep(s, ext, a0, a2, args[3], args[4]); break; case INDEX_op_extract_i64: @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, case INDEX_op_extract2_i64: case INDEX_op_extract2_i32: - tcg_out_extr(s, ext, a0, REG0(2), REG0(1), args[3]); + tcg_out_extr(s, ext, a0, a2, a1, args[3]); break; case INDEX_op_add2_i32: - tcg_out_addsub2(s, TCG_TYPE_I32, a0, a1, REG0(2), REG0(3), + tcg_out_addsub2(s, TCG_TYPE_I32, a0, a1, a2, args[3], (int32_t)args[4], args[5], const_args[4], const_args[5], false); break; case INDEX_op_add2_i64: - tcg_out_addsub2(s, TCG_TYPE_I64, a0, a1, REG0(2), REG0(3), args[4], + tcg_out_addsub2(s, TCG_TYPE_I64, a0, a1, a2, args[3], args[4], args[5], const_args[4], const_args[5], false); break; case INDEX_op_sub2_i32: - tcg_out_addsub2(s, TCG_TYPE_I32, a0, a1, REG0(2), REG0(3), + tcg_out_addsub2(s, TCG_TYPE_I32, a0, a1, a2, args[3], (int32_t)args[4], args[5], const_args[4], const_args[5], true); break; case INDEX_op_sub2_i64: - tcg_out_addsub2(s, TCG_TYPE_I64, a0, a1, REG0(2), REG0(3), args[4], + tcg_out_addsub2(s, TCG_TYPE_I64, a0, a1, a2, args[3], args[4], args[5], const_args[4], const_args[5], true); break; @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType ext, default: g_assert_not_reached(); } - -#undef REG0 } static void tcg_out_vec_op(TCGContext *s, TCGOpcode opc, @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st16_i64: case INDEX_op_st32_i64: case INDEX_op_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_add_i32: case INDEX_op_add_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_movcond_i32: case INDEX_op_movcond_i64: - return C_O1_I4(r, r, rC, rZ, rZ); + return C_O1_I4(r, r, rC, rz, rz); case INDEX_op_qemu_ld_i32: case INDEX_op_qemu_ld_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) return C_O2_I1(r, r, r); case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_qemu_st_i128: - return C_O0_I3(rZ, rZ, r); + return C_O0_I3(rz, rz, r); case INDEX_op_deposit_i32: case INDEX_op_deposit_i64: - return C_O1_I2(r, 0, rZ); + return C_O1_I2(r, 0, rz); case INDEX_op_extract2_i32: case INDEX_op_extract2_i64: - return C_O1_I2(r, rZ, rZ); + return C_O1_I2(r, rz, rz); case INDEX_op_add2_i32: case INDEX_op_add2_i64: case INDEX_op_sub2_i32: case INDEX_op_sub2_i64: - return C_O2_I4(r, r, rZ, rZ, rA, rMZ); + return C_O2_I4(r, r, rz, rz, rA, rMZ); case INDEX_op_add_vec: case INDEX_op_sub_vec: -- 2.43.0 Replace target-specific 'Z' with generic 'z'. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/loongarch64/tcg-target-con-set.h | 15 ++++++------- tcg/loongarch64/tcg-target-con-str.h | 1 - tcg/loongarch64/tcg-target.c.inc | 32 ++++++++++++---------------- 3 files changed, 21 insertions(+), 27 deletions(-) diff --git a/tcg/loongarch64/tcg-target-con-set.h b/tcg/loongarch64/tcg-target-con-set.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target-con-set.h +++ b/tcg/loongarch64/tcg-target-con-set.h @@ -XXX,XX +XXX,XX @@ * tcg-target-con-str.h; the constraint combination is inclusive or. */ C_O0_I1(r) -C_O0_I2(rZ, r) -C_O0_I2(rZ, rZ) +C_O0_I2(rz, r) +C_O0_I2(rz, rz) C_O0_I2(w, r) C_O0_I3(r, r, r) C_O1_I1(r, r) @@ -XXX,XX +XXX,XX @@ C_O1_I2(r, r, rI) C_O1_I2(r, r, rJ) C_O1_I2(r, r, rU) C_O1_I2(r, r, rW) -C_O1_I2(r, r, rZ) -C_O1_I2(r, 0, rZ) -C_O1_I2(r, rZ, ri) -C_O1_I2(r, rZ, rJ) -C_O1_I2(r, rZ, rZ) +C_O1_I2(r, 0, rz) +C_O1_I2(r, rz, ri) +C_O1_I2(r, rz, rJ) +C_O1_I2(r, rz, rz) C_O1_I2(w, w, w) C_O1_I2(w, w, wM) C_O1_I2(w, w, wA) C_O1_I3(w, w, w, w) -C_O1_I4(r, rZ, rJ, rZ, rZ) +C_O1_I4(r, rz, rJ, rz, rz) C_N2_I1(r, r, r) diff --git a/tcg/loongarch64/tcg-target-con-str.h b/tcg/loongarch64/tcg-target-con-str.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target-con-str.h +++ b/tcg/loongarch64/tcg-target-con-str.h @@ -XXX,XX +XXX,XX @@ REGS('w', ALL_VECTOR_REGS) CONST('I', TCG_CT_CONST_S12) CONST('J', TCG_CT_CONST_S32) CONST('U', TCG_CT_CONST_U12) -CONST('Z', TCG_CT_CONST_ZERO) CONST('C', TCG_CT_CONST_C12) CONST('W', TCG_CT_CONST_WSZ) CONST('M', TCG_CT_CONST_VCMP) 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_target_call_oarg_reg(TCGCallReturnKind kind, int slot) #define TCG_GUEST_BASE_REG TCG_REG_S1 -#define TCG_CT_CONST_ZERO 0x100 -#define TCG_CT_CONST_S12 0x200 -#define TCG_CT_CONST_S32 0x400 -#define TCG_CT_CONST_U12 0x800 -#define TCG_CT_CONST_C12 0x1000 -#define TCG_CT_CONST_WSZ 0x2000 -#define TCG_CT_CONST_VCMP 0x4000 -#define TCG_CT_CONST_VADD 0x8000 +#define TCG_CT_CONST_S12 0x100 +#define TCG_CT_CONST_S32 0x200 +#define TCG_CT_CONST_U12 0x400 +#define TCG_CT_CONST_C12 0x800 +#define TCG_CT_CONST_WSZ 0x1000 +#define TCG_CT_CONST_VCMP 0x2000 +#define TCG_CT_CONST_VADD 0x4000 #define ALL_GENERAL_REGS MAKE_64BIT_MASK(0, 32) #define ALL_VECTOR_REGS MAKE_64BIT_MASK(32, 32) @@ -XXX,XX +XXX,XX @@ static bool tcg_target_const_match(int64_t val, int ct, if (ct & TCG_CT_CONST) { return true; } - if ((ct & TCG_CT_CONST_ZERO) && val == 0) { - return true; - } if ((ct & TCG_CT_CONST_S12) && val == sextreg(val, 0, 12)) { return true; } @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st_i64: case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_qemu_ld_i128: return C_N2_I1(r, r, r); @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_brcond_i32: case INDEX_op_brcond_i64: - return C_O0_I2(rZ, rZ); + return C_O0_I2(rz, rz); case INDEX_op_ext8s_i32: case INDEX_op_ext8s_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_deposit_i32: case INDEX_op_deposit_i64: /* Must deposit into the same register as input */ - return C_O1_I2(r, 0, rZ); + return C_O1_I2(r, 0, rz); case INDEX_op_sub_i32: case INDEX_op_setcond_i32: - return C_O1_I2(r, rZ, ri); + return C_O1_I2(r, rz, ri); case INDEX_op_sub_i64: case INDEX_op_setcond_i64: - return C_O1_I2(r, rZ, rJ); + return C_O1_I2(r, rz, rJ); case INDEX_op_mul_i32: case INDEX_op_mul_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_rem_i64: case INDEX_op_remu_i32: case INDEX_op_remu_i64: - return C_O1_I2(r, rZ, rZ); + return C_O1_I2(r, rz, rz); case INDEX_op_movcond_i32: case INDEX_op_movcond_i64: - return C_O1_I4(r, rZ, rJ, rZ, rZ); + return C_O1_I4(r, rz, rJ, rz, rz); case INDEX_op_ld_vec: case INDEX_op_dupm_vec: -- 2.43.0 Replace target-specific 'Z' with generic 'z'. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/mips/tcg-target-con-set.h | 26 ++++++++++----------- tcg/mips/tcg-target-con-str.h | 1 - tcg/mips/tcg-target.c.inc | 44 ++++++++++++++--------------------- 3 files changed, 31 insertions(+), 40 deletions(-) diff --git a/tcg/mips/tcg-target-con-set.h b/tcg/mips/tcg-target-con-set.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/mips/tcg-target-con-set.h +++ b/tcg/mips/tcg-target-con-set.h @@ -XXX,XX +XXX,XX @@ * tcg-target-con-str.h; the constraint combination is inclusive or. */ C_O0_I1(r) -C_O0_I2(rZ, r) -C_O0_I2(rZ, rZ) -C_O0_I3(rZ, r, r) -C_O0_I3(rZ, rZ, r) -C_O0_I4(rZ, rZ, rZ, rZ) -C_O0_I4(rZ, rZ, r, r) +C_O0_I2(rz, r) +C_O0_I2(rz, rz) +C_O0_I3(rz, r, r) +C_O0_I3(rz, rz, r) +C_O0_I4(rz, rz, rz, rz) +C_O0_I4(rz, rz, r, r) C_O1_I1(r, r) -C_O1_I2(r, 0, rZ) +C_O1_I2(r, 0, rz) C_O1_I2(r, r, r) C_O1_I2(r, r, ri) C_O1_I2(r, r, rI) C_O1_I2(r, r, rIK) C_O1_I2(r, r, rJ) -C_O1_I2(r, r, rWZ) -C_O1_I2(r, rZ, rN) -C_O1_I2(r, rZ, rZ) -C_O1_I4(r, rZ, rZ, rZ, 0) -C_O1_I4(r, rZ, rZ, rZ, rZ) +C_O1_I2(r, r, rzW) +C_O1_I2(r, rz, rN) +C_O1_I2(r, rz, rz) +C_O1_I4(r, rz, rz, rz, 0) +C_O1_I4(r, rz, rz, rz, rz) C_O2_I1(r, r, r) C_O2_I2(r, r, r, r) -C_O2_I4(r, r, rZ, rZ, rN, rN) +C_O2_I4(r, r, rz, rz, rN, rN) diff --git a/tcg/mips/tcg-target-con-str.h b/tcg/mips/tcg-target-con-str.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/mips/tcg-target-con-str.h +++ b/tcg/mips/tcg-target-con-str.h @@ -XXX,XX +XXX,XX @@ CONST('J', TCG_CT_CONST_S16) CONST('K', TCG_CT_CONST_P2M1) CONST('N', TCG_CT_CONST_N16) CONST('W', TCG_CT_CONST_WSZ) -CONST('Z', TCG_CT_CONST_ZERO) 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 patch_reloc(tcg_insn_unit *code_ptr, int type, g_assert_not_reached(); } -#define TCG_CT_CONST_ZERO 0x100 -#define TCG_CT_CONST_U16 0x200 /* Unsigned 16-bit: 0 - 0xffff. */ -#define TCG_CT_CONST_S16 0x400 /* Signed 16-bit: -32768 - 32767 */ -#define TCG_CT_CONST_P2M1 0x800 /* Power of 2 minus 1. */ -#define TCG_CT_CONST_N16 0x1000 /* "Negatable" 16-bit: -32767 - 32767 */ -#define TCG_CT_CONST_WSZ 0x2000 /* word size */ +#define TCG_CT_CONST_U16 0x100 /* Unsigned 16-bit: 0 - 0xffff. */ +#define TCG_CT_CONST_S16 0x200 /* Signed 16-bit: -32768 - 32767 */ +#define TCG_CT_CONST_P2M1 0x400 /* Power of 2 minus 1. */ +#define TCG_CT_CONST_N16 0x800 /* "Negatable" 16-bit: -32767 - 32767 */ +#define TCG_CT_CONST_WSZ 0x1000 /* word size */ #define ALL_GENERAL_REGS 0xffffffffu @@ -XXX,XX +XXX,XX @@ static bool tcg_target_const_match(int64_t val, int ct, { if (ct & TCG_CT_CONST) { return 1; - } else if ((ct & TCG_CT_CONST_ZERO) && val == 0) { - return 1; } else if ((ct & TCG_CT_CONST_U16) && val == (uint16_t)val) { return 1; } else if ((ct & TCG_CT_CONST_S16) && val == (int16_t)val) { @@ -XXX,XX +XXX,XX @@ static void tcg_out_op(TCGContext *s, TCGOpcode opc, TCGType type, TCGArg a0, a1, a2; int c2; - /* - * Note that many operands use the constraint set "rZ". - * We make use of the fact that 0 is the ZERO register, - * and hence such cases need not check for const_args. - */ a0 = args[0]; a1 = args[1]; a2 = args[2]; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st16_i64: case INDEX_op_st32_i64: case INDEX_op_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_add_i32: case INDEX_op_add_i64: return C_O1_I2(r, r, rJ); case INDEX_op_sub_i32: case INDEX_op_sub_i64: - return C_O1_I2(r, rZ, rN); + return C_O1_I2(r, rz, rN); case INDEX_op_mul_i32: case INDEX_op_mulsh_i32: case INDEX_op_muluh_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_remu_i64: case INDEX_op_nor_i64: case INDEX_op_setcond_i64: - return C_O1_I2(r, rZ, rZ); + return C_O1_I2(r, rz, rz); case INDEX_op_muls2_i32: case INDEX_op_mulu2_i32: case INDEX_op_muls2_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) return C_O1_I2(r, r, ri); case INDEX_op_clz_i32: case INDEX_op_clz_i64: - return C_O1_I2(r, r, rWZ); + return C_O1_I2(r, r, rzW); case INDEX_op_deposit_i32: case INDEX_op_deposit_i64: - return C_O1_I2(r, 0, rZ); + return C_O1_I2(r, 0, rz); case INDEX_op_brcond_i32: case INDEX_op_brcond_i64: - return C_O0_I2(rZ, rZ); + return C_O0_I2(rz, rz); case INDEX_op_movcond_i32: case INDEX_op_movcond_i64: return (use_mips32r6_instructions - ? C_O1_I4(r, rZ, rZ, rZ, rZ) - : C_O1_I4(r, rZ, rZ, rZ, 0)); + ? C_O1_I4(r, rz, rz, rz, rz) + : C_O1_I4(r, rz, rz, rz, 0)); case INDEX_op_add2_i32: case INDEX_op_sub2_i32: - return C_O2_I4(r, r, rZ, rZ, rN, rN); + return C_O2_I4(r, r, rz, rz, rN, rN); case INDEX_op_setcond2_i32: - return C_O1_I4(r, rZ, rZ, rZ, rZ); + return C_O1_I4(r, rz, rz, rz, rz); case INDEX_op_brcond2_i32: - return C_O0_I4(rZ, rZ, rZ, rZ); + return C_O0_I4(rz, rz, rz, rz); case INDEX_op_qemu_ld_i32: return C_O1_I1(r, r); case INDEX_op_qemu_st_i32: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_qemu_ld_i64: return TCG_TARGET_REG_BITS == 64 ? C_O1_I1(r, r) : C_O2_I1(r, r, r); case INDEX_op_qemu_st_i64: - return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(rZ, r) : C_O0_I3(rZ, rZ, r); + return TCG_TARGET_REG_BITS == 64 ? C_O0_I2(rz, r) : C_O0_I3(rz, rz, r); default: return C_NotImplemented; -- 2.43.0 Replace target-specific 'Z' with generic 'z'. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/riscv/tcg-target-con-set.h | 10 +++++----- tcg/riscv/tcg-target-con-str.h | 1 - tcg/riscv/tcg-target.c.inc | 28 ++++++++++++---------------- 3 files changed, 17 insertions(+), 22 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 @@ * tcg-target-con-str.h; the constraint combination is inclusive or. */ C_O0_I1(r) -C_O0_I2(rZ, r) -C_O0_I2(rZ, rZ) +C_O0_I2(rz, r) +C_O0_I2(rz, rz) C_O1_I1(r, r) C_O1_I2(r, r, ri) C_O1_I2(r, r, rI) C_O1_I2(r, r, rJ) -C_O1_I2(r, rZ, rN) -C_O1_I2(r, rZ, rZ) +C_O1_I2(r, rz, rN) +C_O1_I2(r, rz, rz) C_N1_I2(r, r, rM) C_O1_I4(r, r, rI, rM, rM) -C_O2_I4(r, r, rZ, rZ, rM, rM) +C_O2_I4(r, r, rz, rz, rM, rM) C_O0_I2(v, r) C_O1_I1(v, r) C_O1_I1(v, v) diff --git a/tcg/riscv/tcg-target-con-str.h b/tcg/riscv/tcg-target-con-str.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target-con-str.h +++ b/tcg/riscv/tcg-target-con-str.h @@ -XXX,XX +XXX,XX @@ CONST('K', TCG_CT_CONST_S5) CONST('L', TCG_CT_CONST_CMP_VI) CONST('N', TCG_CT_CONST_N12) CONST('M', TCG_CT_CONST_M12) -CONST('Z', TCG_CT_CONST_ZERO) 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) return TCG_REG_A0 + slot; } -#define TCG_CT_CONST_ZERO 0x100 -#define TCG_CT_CONST_S12 0x200 -#define TCG_CT_CONST_N12 0x400 -#define TCG_CT_CONST_M12 0x800 -#define TCG_CT_CONST_J12 0x1000 -#define TCG_CT_CONST_S5 0x2000 -#define TCG_CT_CONST_CMP_VI 0x4000 +#define TCG_CT_CONST_S12 0x100 +#define TCG_CT_CONST_N12 0x200 +#define TCG_CT_CONST_M12 0x400 +#define TCG_CT_CONST_J12 0x800 +#define TCG_CT_CONST_S5 0x1000 +#define TCG_CT_CONST_CMP_VI 0x2000 #define ALL_GENERAL_REGS MAKE_64BIT_MASK(0, 32) #define ALL_VECTOR_REGS MAKE_64BIT_MASK(32, 32) @@ -XXX,XX +XXX,XX @@ static bool tcg_target_const_match(int64_t val, int ct, if (ct & TCG_CT_CONST) { return 1; } - if ((ct & TCG_CT_CONST_ZERO) && val == 0) { - return 1; - } if (type >= TCG_TYPE_V64) { /* Val is replicated by VECE; extract the highest element. */ val >>= (-8 << vece) & 63; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st16_i64: case INDEX_op_st32_i64: case INDEX_op_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_add_i32: case INDEX_op_and_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_sub_i32: case INDEX_op_sub_i64: - return C_O1_I2(r, rZ, rN); + return C_O1_I2(r, rz, rN); case INDEX_op_mul_i32: case INDEX_op_mulsh_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_divu_i64: case INDEX_op_rem_i64: case INDEX_op_remu_i64: - return C_O1_I2(r, rZ, rZ); + return C_O1_I2(r, rz, rz); case INDEX_op_shl_i32: case INDEX_op_shr_i32: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_brcond_i32: case INDEX_op_brcond_i64: - return C_O0_I2(rZ, rZ); + return C_O0_I2(rz, rz); case INDEX_op_movcond_i32: case INDEX_op_movcond_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_add2_i64: case INDEX_op_sub2_i32: case INDEX_op_sub2_i64: - return C_O2_I4(r, r, rZ, rZ, rM, rM); + return C_O2_I4(r, r, rz, rz, rM, rM); case INDEX_op_qemu_ld_i32: case INDEX_op_qemu_ld_i64: return C_O1_I1(r, r); case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_st_vec: return C_O0_I2(v, r); -- 2.43.0 Replace target-specific 'Z' with generic 'z'. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/sparc64/tcg-target-con-set.h | 12 ++++++------ tcg/sparc64/tcg-target-con-str.h | 1 - tcg/sparc64/tcg-target.c.inc | 17 +++++++---------- 3 files changed, 13 insertions(+), 17 deletions(-) diff --git a/tcg/sparc64/tcg-target-con-set.h b/tcg/sparc64/tcg-target-con-set.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/sparc64/tcg-target-con-set.h +++ b/tcg/sparc64/tcg-target-con-set.h @@ -XXX,XX +XXX,XX @@ * tcg-target-con-str.h; the constraint combination is inclusive or. */ C_O0_I1(r) -C_O0_I2(rZ, r) -C_O0_I2(rZ, rJ) +C_O0_I2(rz, r) +C_O0_I2(rz, rJ) C_O1_I1(r, r) C_O1_I2(r, r, r) -C_O1_I2(r, rZ, rJ) -C_O1_I4(r, rZ, rJ, rI, 0) -C_O2_I2(r, r, rZ, rJ) -C_O2_I4(r, r, rZ, rZ, rJ, rJ) +C_O1_I2(r, rz, rJ) +C_O1_I4(r, rz, rJ, rI, 0) +C_O2_I2(r, r, rz, rJ) +C_O2_I4(r, r, rz, rz, rJ, rJ) diff --git a/tcg/sparc64/tcg-target-con-str.h b/tcg/sparc64/tcg-target-con-str.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/sparc64/tcg-target-con-str.h +++ b/tcg/sparc64/tcg-target-con-str.h @@ -XXX,XX +XXX,XX @@ REGS('r', ALL_GENERAL_REGS) */ CONST('I', TCG_CT_CONST_S11) CONST('J', TCG_CT_CONST_S13) -CONST('Z', TCG_CT_CONST_ZERO) 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 char * const tcg_target_reg_names[TCG_TARGET_NB_REGS] = { #define TCG_CT_CONST_S11 0x100 #define TCG_CT_CONST_S13 0x200 -#define TCG_CT_CONST_ZERO 0x400 #define ALL_GENERAL_REGS MAKE_64BIT_MASK(0, 32) @@ -XXX,XX +XXX,XX @@ static bool tcg_target_const_match(int64_t val, int ct, val = (int32_t)val; } - if ((ct & TCG_CT_CONST_ZERO) && val == 0) { - return 1; - } else if ((ct & TCG_CT_CONST_S11) && check_fit_tl(val, 11)) { + if ((ct & TCG_CT_CONST_S11) && check_fit_tl(val, 11)) { return 1; } else if ((ct & TCG_CT_CONST_S13) && check_fit_tl(val, 13)) { return 1; @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_st_i64: case INDEX_op_qemu_st_i32: case INDEX_op_qemu_st_i64: - return C_O0_I2(rZ, r); + return C_O0_I2(rz, r); case INDEX_op_add_i32: case INDEX_op_add_i64: @@ -XXX,XX +XXX,XX @@ tcg_target_op_def(TCGOpcode op, TCGType type, unsigned flags) case INDEX_op_setcond_i64: case INDEX_op_negsetcond_i32: case INDEX_op_negsetcond_i64: - return C_O1_I2(r, rZ, rJ); + return C_O1_I2(r, rz, rJ); case INDEX_op_brcond_i32: case INDEX_op_brcond_i64: - return C_O0_I2(rZ, rJ); + return C_O0_I2(rz, rJ); case INDEX_op_movcond_i32: case INDEX_op_movcond_i64: - return C_O1_I4(r, rZ, rJ, rI, 0); + return C_O1_I4(r, rz, rJ, rI, 0); case INDEX_op_add2_i32: case INDEX_op_add2_i64: case INDEX_op_sub2_i32: case INDEX_op_sub2_i64: - return C_O2_I4(r, r, rZ, rZ, rJ, rJ); + return C_O2_I4(r, r, rz, rz, rJ, rJ); case INDEX_op_mulu2_i32: case INDEX_op_muls2_i32: - return C_O2_I2(r, r, rZ, rJ); + return C_O2_I2(r, r, rz, rJ); case INDEX_op_muluh_i64: return C_O1_I2(r, r, r); -- 2.43.0 From: Fabiano Rosas <farosas@suse.de> When complying with the alignment requested in the ELF and unmapping the excess reservation, having align_end not aligned to the guest page causes the unmap to be rejected by the alignment check at target_munmap and later brk adjustments hit an EEXIST. Fix by aligning the start of region to be unmapped. Fixes: c81d1fafa6 ("linux-user: Honor elf alignment when placing images") Resolves: https://gitlab.com/qemu-project/qemu/-/issues/1913 Signed-off-by: Fabiano Rosas <farosas@suse.de> [rth: Align load_end as well.] Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20250213143558.10504-1-farosas@suse.de> --- linux-user/elfload.c | 4 ++-- 1 file changed, 2 insertions(+), 2 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 void load_elf_image(const char *image_name, const ImageSource *src, if (align_size != reserve_size) { abi_ulong align_addr = ROUND_UP(load_addr, align); - abi_ulong align_end = align_addr + reserve_size; - abi_ulong load_end = load_addr + align_size; + abi_ulong align_end = TARGET_PAGE_ALIGN(align_addr + reserve_size); + abi_ulong load_end = TARGET_PAGE_ALIGN(load_addr + align_size); if (align_addr != load_addr) { target_munmap(load_addr, align_addr - load_addr); -- 2.43.0 From: Andreas Schwab <schwab@suse.de> SA_RESTORER and the associated sa_restorer field of struct sigaction are an obsolete feature, not expected to be used by future architectures. They are also absent on RISC-V, LoongArch, Hexagon and OpenRISC, but defined due to their use of generic/signal.h. This leads to corrupted data and out-of-bounds accesses. Move the definition of TARGET_SA_RESTORER out of generic/signal.h into the target_signal.h files that need it. Note that m68k has the sa_restorer field, but does not use it and does not define SA_RESTORER. Reported-by: Thomas Weißschuh <thomas@t-8ch.de> Signed-off-by: Andreas Schwab <schwab@suse.de> Reviewed-by: Thomas Weißschuh <thomas@t-8ch.de> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <mvmed060xc9.fsf@suse.de> --- linux-user/aarch64/target_signal.h | 2 ++ linux-user/arm/target_signal.h | 2 ++ linux-user/generic/signal.h | 1 - linux-user/i386/target_signal.h | 2 ++ linux-user/m68k/target_signal.h | 1 + linux-user/microblaze/target_signal.h | 2 ++ linux-user/ppc/target_signal.h | 2 ++ linux-user/s390x/target_signal.h | 2 ++ linux-user/sh4/target_signal.h | 2 ++ linux-user/x86_64/target_signal.h | 2 ++ linux-user/xtensa/target_signal.h | 2 ++ 11 files changed, 19 insertions(+), 1 deletion(-) diff --git a/linux-user/aarch64/target_signal.h b/linux-user/aarch64/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/aarch64/target_signal.h +++ b/linux-user/aarch64/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_SEGV_MTEAERR 8 /* Asynchronous ARM MTE error */ #define TARGET_SEGV_MTESERR 9 /* Synchronous ARM MTE exception */ diff --git a/linux-user/arm/target_signal.h b/linux-user/arm/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/arm/target_signal.h +++ b/linux-user/arm/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SETUP_FRAME #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 diff --git a/linux-user/generic/signal.h b/linux-user/generic/signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/generic/signal.h +++ b/linux-user/generic/signal.h @@ -XXX,XX +XXX,XX @@ #define TARGET_SA_RESTART 0x10000000 #define TARGET_SA_NODEFER 0x40000000 #define TARGET_SA_RESETHAND 0x80000000 -#define TARGET_SA_RESTORER 0x04000000 #define TARGET_SIGHUP 1 #define TARGET_SIGINT 2 diff --git a/linux-user/i386/target_signal.h b/linux-user/i386/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/i386/target_signal.h +++ b/linux-user/i386/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SETUP_FRAME #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 diff --git a/linux-user/m68k/target_signal.h b/linux-user/m68k/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/m68k/target_signal.h +++ b/linux-user/m68k/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_ARCH_HAS_SA_RESTORER 1 #define TARGET_ARCH_HAS_SETUP_FRAME #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 diff --git a/linux-user/microblaze/target_signal.h b/linux-user/microblaze/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/microblaze/target_signal.h +++ b/linux-user/microblaze/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 #endif /* MICROBLAZE_TARGET_SIGNAL_H */ diff --git a/linux-user/ppc/target_signal.h b/linux-user/ppc/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/ppc/target_signal.h +++ b/linux-user/ppc/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #if !defined(TARGET_PPC64) #define TARGET_ARCH_HAS_SETUP_FRAME #endif diff --git a/linux-user/s390x/target_signal.h b/linux-user/s390x/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/s390x/target_signal.h +++ b/linux-user/s390x/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SETUP_FRAME #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 diff --git a/linux-user/sh4/target_signal.h b/linux-user/sh4/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/sh4/target_signal.h +++ b/linux-user/sh4/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SETUP_FRAME #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 diff --git a/linux-user/x86_64/target_signal.h b/linux-user/x86_64/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/x86_64/target_signal.h +++ b/linux-user/x86_64/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + /* For x86_64, use of SA_RESTORER is mandatory. */ #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 0 diff --git a/linux-user/xtensa/target_signal.h b/linux-user/xtensa/target_signal.h index XXXXXXX..XXXXXXX 100644 --- a/linux-user/xtensa/target_signal.h +++ b/linux-user/xtensa/target_signal.h @@ -XXX,XX +XXX,XX @@ #include "../generic/signal.h" +#define TARGET_SA_RESTORER 0x04000000 + #define TARGET_ARCH_HAS_SIGTRAMP_PAGE 1 #endif -- 2.43.0 From: Mikael Szreder <git@miszr.win> A bug was introduced in commit 0bba7572d40d which causes the fdtox and fqtox instructions to incorrectly select the destination registers. More information and a test program can be found in issue #2802. Cc: qemu-stable@nongnu.org Fixes: 0bba7572d40d ("target/sparc: Perform DFPREG/QFPREG in decodetree") Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2802 Signed-off-by: Mikael Szreder <git@miszr.win> Acked-by: Artyom Tarasenko <atar4qemu@gmail.com> [rth: Squash patches together, since the second fixes a typo in the first.] Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20250205090333.19626-3-git@miszr.win> --- target/sparc/insns.decode | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/target/sparc/insns.decode b/target/sparc/insns.decode index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/insns.decode +++ b/target/sparc/insns.decode @@ -XXX,XX +XXX,XX @@ FdMULq 10 ..... 110100 ..... 0 0110 1110 ..... @q_d_d FNHADDs 10 ..... 110100 ..... 0 0111 0001 ..... @r_r_r FNHADDd 10 ..... 110100 ..... 0 0111 0010 ..... @d_d_d FNsMULd 10 ..... 110100 ..... 0 0111 1001 ..... @d_r_r -FsTOx 10 ..... 110100 00000 0 1000 0001 ..... @r_r2 -FdTOx 10 ..... 110100 00000 0 1000 0010 ..... @r_d2 -FqTOx 10 ..... 110100 00000 0 1000 0011 ..... @r_q2 -FxTOs 10 ..... 110100 00000 0 1000 0100 ..... @r_r2 -FxTOd 10 ..... 110100 00000 0 1000 1000 ..... @d_r2 -FxTOq 10 ..... 110100 00000 0 1000 1100 ..... @q_r2 +FsTOx 10 ..... 110100 00000 0 1000 0001 ..... @d_r2 +FdTOx 10 ..... 110100 00000 0 1000 0010 ..... @d_d2 +FqTOx 10 ..... 110100 00000 0 1000 0011 ..... @d_q2 +FxTOs 10 ..... 110100 00000 0 1000 0100 ..... @r_d2 +FxTOd 10 ..... 110100 00000 0 1000 1000 ..... @d_d2 +FxTOq 10 ..... 110100 00000 0 1000 1100 ..... @q_d2 FiTOs 10 ..... 110100 00000 0 1100 0100 ..... @r_r2 FdTOs 10 ..... 110100 00000 0 1100 0110 ..... @r_d2 FqTOs 10 ..... 110100 00000 0 1100 0111 ..... @r_q2 -- 2.43.0 From: Mikael Szreder <git@miszr.win> The gdbstub implementation for the Sparc architecture would incorrectly calculate the the floating point register offset. This resulted in, for example, registers f32 and f34 to point to the same value. The issue was caused by the confusion between even register numbers and even register indexes. For example, the register index of f32 is 64 and f34 is 65. Cc: qemu-stable@nongnu.org Fixes: 30038fd81808 ("target-sparc: Change fpr representation to doubles.") Signed-off-by: Mikael Szreder <git@miszr.win> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20250214070343.11501-1-git@miszr.win> --- target/sparc/gdbstub.c | 18 ++++++++++++++---- 1 file changed, 14 insertions(+), 4 deletions(-) diff --git a/target/sparc/gdbstub.c b/target/sparc/gdbstub.c index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/gdbstub.c +++ b/target/sparc/gdbstub.c @@ -XXX,XX +XXX,XX @@ int sparc_cpu_gdb_read_register(CPUState *cs, GByteArray *mem_buf, int n) } } if (n < 80) { - /* f32-f62 (double width, even numbers only) */ - return gdb_get_reg64(mem_buf, env->fpr[(n - 32) / 2].ll); + /* f32-f62 (16 double width registers, even register numbers only) + * n == 64: f32 : env->fpr[16] + * n == 65: f34 : env->fpr[17] + * etc... + * n == 79: f62 : env->fpr[31] + */ + return gdb_get_reg64(mem_buf, env->fpr[(n - 64) + 16].ll); } switch (n) { case 80: @@ -XXX,XX +XXX,XX @@ int sparc_cpu_gdb_write_register(CPUState *cs, uint8_t *mem_buf, int n) } return 4; } else if (n < 80) { - /* f32-f62 (double width, even numbers only) */ - env->fpr[(n - 32) / 2].ll = tmp; + /* f32-f62 (16 double width registers, even register numbers only) + * n == 64: f32 : env->fpr[16] + * n == 65: f34 : env->fpr[17] + * etc... + * n == 79: f62 : env->fpr[31] + */ + env->fpr[(n - 64) + 16].ll = tmp; } else { switch (n) { case 80: -- 2.43.0 From: Artyom Tarasenko <atar4qemu@gmail.com> Fake access to PCR Performance Control Register and PIC Performance Instrumentation Counter. Ignore writes in privileged mode, and return 0 on reads. This allows booting Tribblix, MilaX and v9os under Niagara target. Signed-off-by: Artyom Tarasenko <atar4qemu@gmail.com> Reviewed-by: Richard Henderson <richard.henderson@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> Message-ID: <20250209211248.50383-1-atar4qemu@gmail.com> --- target/sparc/translate.c | 19 +++++++++++++++++++ target/sparc/insns.decode | 7 ++++++- 2 files changed, 25 insertions(+), 1 deletion(-) 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 TCGv do_rd_leon3_config(DisasContext *dc, TCGv dst) TRANS(RDASR17, ASR17, do_rd_special, true, a->rd, do_rd_leon3_config) +static TCGv do_rdpic(DisasContext *dc, TCGv dst) +{ + return tcg_constant_tl(0); +} + +TRANS(RDPIC, HYPV, do_rd_special, supervisor(dc), a->rd, do_rdpic) + + static TCGv do_rdccr(DisasContext *dc, TCGv dst) { gen_helper_rdccr(dst, tcg_env); @@ -XXX,XX +XXX,XX @@ static void do_wrfprs(DisasContext *dc, TCGv src) TRANS(WRFPRS, 64, do_wr_special, a, true, do_wrfprs) +static bool do_priv_nop(DisasContext *dc, bool priv) +{ + if (!priv) { + return raise_priv(dc); + } + return advance_pc(dc); +} + +TRANS(WRPCR, HYPV, do_priv_nop, supervisor(dc)) +TRANS(WRPIC, HYPV, do_priv_nop, supervisor(dc)) + static void do_wrgsr(DisasContext *dc, TCGv src) { gen_trap_ifnofpu(dc); diff --git a/target/sparc/insns.decode b/target/sparc/insns.decode index XXXXXXX..XXXXXXX 100644 --- a/target/sparc/insns.decode +++ b/target/sparc/insns.decode @@ -XXX,XX +XXX,XX @@ CALL 01 i:s30 RDTICK 10 rd:5 101000 00100 0 0000000000000 RDPC 10 rd:5 101000 00101 0 0000000000000 RDFPRS 10 rd:5 101000 00110 0 0000000000000 - RDASR17 10 rd:5 101000 10001 0 0000000000000 + { + RDASR17 10 rd:5 101000 10001 0 0000000000000 + RDPIC 10 rd:5 101000 10001 0 0000000000000 + } RDGSR 10 rd:5 101000 10011 0 0000000000000 RDSOFTINT 10 rd:5 101000 10110 0 0000000000000 RDTICK_CMPR 10 rd:5 101000 10111 0 0000000000000 @@ -XXX,XX +XXX,XX @@ CALL 01 i:s30 WRCCR 10 00010 110000 ..... . ............. @n_r_ri WRASI 10 00011 110000 ..... . ............. @n_r_ri WRFPRS 10 00110 110000 ..... . ............. @n_r_ri + WRPCR 10 10000 110000 01000 0 0000000000000 + WRPIC 10 10001 110000 01000 0 0000000000000 { WRGSR 10 10011 110000 ..... . ............. @n_r_ri WRPOWERDOWN 10 10011 110000 ..... . ............. @n_r_ri -- 2.43.0 Eliminate code repetition by using the appropriate helpers. Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/i386/tcg-target.c.inc | 65 +++++---------------------------------- 1 file changed, 8 insertions(+), 57 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_brcond2(TCGContext *s, const TCGArg *args, tcg_out_brcond(s, 0, cond, args[1], args[3], const_args[3], label_this, small); break; + case TCG_COND_NE: case TCG_COND_TSTNE: tcg_out_brcond(s, 0, cond, args[0], args[2], const_args[2], @@ -XXX,XX +XXX,XX @@ static void tcg_out_brcond2(TCGContext *s, const TCGArg *args, tcg_out_brcond(s, 0, cond, args[1], args[3], const_args[3], label_this, small); break; - case TCG_COND_LT: - tcg_out_brcond(s, 0, TCG_COND_LT, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_LTU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_LE: - tcg_out_brcond(s, 0, TCG_COND_LT, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_LEU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_GT: - tcg_out_brcond(s, 0, TCG_COND_GT, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_GTU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_GE: - tcg_out_brcond(s, 0, TCG_COND_GT, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_GEU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_LTU: - tcg_out_brcond(s, 0, TCG_COND_LTU, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_LTU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_LEU: - tcg_out_brcond(s, 0, TCG_COND_LTU, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_LEU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_GTU: - tcg_out_brcond(s, 0, TCG_COND_GTU, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_GTU, args[0], args[2], const_args[2], - label_this, small); - break; - case TCG_COND_GEU: - tcg_out_brcond(s, 0, TCG_COND_GTU, args[1], args[3], const_args[3], - label_this, small); - tcg_out_jxx(s, JCC_JNE, label_next, 1); - tcg_out_brcond(s, 0, TCG_COND_GEU, args[0], args[2], const_args[2], - label_this, small); - break; + default: - g_assert_not_reached(); + tcg_out_brcond(s, 0, tcg_high_cond(cond), args[1], + args[3], const_args[3], label_this, small); + tcg_out_jxx(s, JCC_JNE, label_next, 1); + tcg_out_brcond(s, 0, tcg_unsigned_cond(cond), args[0], + args[2], const_args[2], label_this, small); + break; } tcg_out_label(s, label_next); } -- 2.43.0 These defines never should have been added as they were never used. Only 32-bit hosts may have these opcodes and they have them unconditionally. Fixes: 6cb14e4de29 ("tcg/loongarch64: Add the tcg-target.h file") Fixes: fb1f70f3685 ("tcg/riscv: Add the tcg-target.h file") Acked-by: Alistair Francis <alistair.francis@wdc.com> Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org> Signed-off-by: Richard Henderson <richard.henderson@linaro.org> --- tcg/loongarch64/tcg-target-has.h | 2 -- tcg/riscv/tcg-target-has.h | 2 -- 2 files changed, 4 deletions(-) diff --git a/tcg/loongarch64/tcg-target-has.h b/tcg/loongarch64/tcg-target-has.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/loongarch64/tcg-target-has.h +++ b/tcg/loongarch64/tcg-target-has.h @@ -XXX,XX +XXX,XX @@ #define TCG_TARGET_HAS_clz_i32 1 #define TCG_TARGET_HAS_ctz_i32 1 #define TCG_TARGET_HAS_ctpop_i32 0 -#define TCG_TARGET_HAS_brcond2 0 -#define TCG_TARGET_HAS_setcond2 0 #define TCG_TARGET_HAS_qemu_st8_i32 0 /* 64-bit operations */ diff --git a/tcg/riscv/tcg-target-has.h b/tcg/riscv/tcg-target-has.h index XXXXXXX..XXXXXXX 100644 --- a/tcg/riscv/tcg-target-has.h +++ b/tcg/riscv/tcg-target-has.h @@ -XXX,XX +XXX,XX @@ #define TCG_TARGET_HAS_clz_i32 (cpuinfo & CPUINFO_ZBB) #define TCG_TARGET_HAS_ctz_i32 (cpuinfo & CPUINFO_ZBB) #define TCG_TARGET_HAS_ctpop_i32 (cpuinfo & CPUINFO_ZBB) -#define TCG_TARGET_HAS_brcond2 1 -#define TCG_TARGET_HAS_setcond2 1 #define TCG_TARGET_HAS_qemu_st8_i32 0 #define TCG_TARGET_HAS_negsetcond_i64 1 -- 2.43.0