20638c2e61
PM4 buffer handling made a virtual member of commandprocessor, place the implementation/declaration into reusable macro files. this is probably the biggest boost here. Optimized SET_CONSTANT/ LOAD_CONSTANT pm4 ops based on the register range they start writing at, this was also a nice boost Expose X64 extension flags to code outside of x64 backend, so we can detect and use things like avx512, xop, avx2, etc in normal code Add freelists for HIR structures to try to reduce the number of last level cache misses during optimization (currently disabled... fixme later) Analyzed PGO feedback and reordered branches, uninlined functions, moved code out into different functions based on info from it in the PM4 functions, this gave like a 2% boost at best. Added support for the db16cyc opcode, which is used often in xb360 spinlocks. before it was just being translated to nop, now on x64 we translate it to _mm_pause but may change that in the future to reduce cpu time wasted texture util - all our divisors were powers of 2, instead we look up a shift. this made texture scaling slightly faster, more so on intel processors which seem to be worse at int divs. GetGuestTextureLayout is now a little faster, although it is still one of the heaviest functions in the emulator when scaling is on. xe_unlikely_mutex was not a good choice for the guest clock lock, (running theory) on intel processors another thread may take a significant time to update the clock? maybe because of the uint64 division? really not sure, but switched it to xe_mutex. This fixed audio stutter that i had introduced to 1 or 2 games, fixed performance on that n64 rare game with the monkeys. Took another crack at DMA implementation, another failure. Instead of passing as a parameter, keep the ringbuffer reader as the first member of commandprocessor so it can be accessed through this Added macro for noalias Applied noalias to Memory::LookupHeap. This reduced the size of the executable by 7 kb. Reworked kernel shim template, this shaved like 100kb off the exe and eliminated the indirect calls from the shim to the actual implementation. We still unconditionally generate string representations of kernel calls though :(, unless it is kHighFrequency Add nvapi extensions support, currently unused. Will use CPUVISIBLE memory at some point Inserted prefetches in a few places based on feedback from vtune. Add native implementation of SHA int8 if all elements are the same Vectorized comparisons for SetViewport, SetScissorRect Vectorized ranged comparisons for WriteRegister Add XE_MSVC_ASSUME Move FormatInfo::name out of the structure, instead look up the name in a different table. Debug related data and critical runtime data are best kept apart Templated UpdateSystemConstantValues based on ROV/RTV and primitive_polygonal Add ArchFloatMask functions, these are for storing the results of floating point comparisons without doing costly float->int pipeline transfers (vucomiss/setb) Use floatmasks in UpdateSystemConstantValues for checking if dirty, only transfer to int at end of function. Instead of dirty |= (x == y) in UpdateSystemConstantValues, now we do dirty_u32 |= (x^y). if any of them are not equal, dirty_u32 will be nz, else if theyre all equal it will be zero. This is more friendly to register renaming and the lack of dependencies on EFLAGS lets the compiler reorder better Add PrefetchSamplerParameters to D3D12TextureCache use PrefetchSamplerParameters in UpdateBindings to eliminate cache misses that vtune detected Add PrefetchTextureBinding to D3D12TextureCache Prefetch texture bindings to get rid of more misses vtune detected (more accesses out of order with random strides) Rewrote DMAC, still terrible though and have disabled it for now. Replace tiny memcmp of 6 U64 in render_target_cache with inline loop, msvc fails to make it a loop and instead does a thunk to their memcmp function, which is optimized for larger sizes PrefetchTextureBinding in AreActiveTextureSRVKeysUpToDate Replace memcmp calls for pipelinedescription with handwritten cmp Directly write some registers that dont have special handling in PM4 functions Changed EstimateMaxY to try to eliminate mispredictions that vtune was reporting, msvc ended up turning the changed code into a series of blends in ExecutePacketType3_EVENT_WRITE_EXT, instead of writing extents to an array on the stack and then doing xe_copy_and_swap_16 of the data to its dest, pre-swap each constant and then store those. msvc manages to unroll that into wider stores stop logging XE_SWAP every time we receive XE_SWAP, stop logging the start and end of each viz query Prefetch watch nodes in FireWatches based on feedback from vtune Removed dead code from texture_info.cc NOINLINE on GpuSwap, PGO builds did it so we should too.
1606 lines
52 KiB
C++
1606 lines
52 KiB
C++
/**
|
|
******************************************************************************
|
|
* Xenia : Xbox 360 Emulator Research Project *
|
|
******************************************************************************
|
|
* Copyright 2022 Ben Vanik. All rights reserved. *
|
|
* Released under the BSD license - see LICENSE in the root for more details. *
|
|
******************************************************************************
|
|
*/
|
|
|
|
#include "xenia/cpu/backend/x64/x64_emitter.h"
|
|
|
|
#include <stddef.h>
|
|
|
|
#include <climits>
|
|
#include <cstring>
|
|
|
|
#include "third_party/fmt/include/fmt/format.h"
|
|
#include "xenia/base/assert.h"
|
|
#include "xenia/base/atomic.h"
|
|
#include "xenia/base/debugging.h"
|
|
#include "xenia/base/literals.h"
|
|
#include "xenia/base/logging.h"
|
|
#include "xenia/base/math.h"
|
|
#include "xenia/base/memory.h"
|
|
#include "xenia/base/profiling.h"
|
|
#include "xenia/base/vec128.h"
|
|
#include "xenia/cpu/backend/x64/x64_backend.h"
|
|
#include "xenia/cpu/backend/x64/x64_code_cache.h"
|
|
#include "xenia/cpu/backend/x64/x64_function.h"
|
|
#include "xenia/cpu/backend/x64/x64_sequences.h"
|
|
#include "xenia/cpu/backend/x64/x64_stack_layout.h"
|
|
#include "xenia/cpu/cpu_flags.h"
|
|
#include "xenia/cpu/function.h"
|
|
#include "xenia/cpu/function_debug_info.h"
|
|
#include "xenia/cpu/hir/instr.h"
|
|
#include "xenia/cpu/hir/opcodes.h"
|
|
#include "xenia/cpu/hir/value.h"
|
|
#include "xenia/cpu/processor.h"
|
|
#include "xenia/cpu/symbol.h"
|
|
#include "xenia/cpu/thread_state.h"
|
|
|
|
DEFINE_bool(debugprint_trap_log, false,
|
|
"Log debugprint traps to the active debugger", "CPU");
|
|
DEFINE_bool(ignore_undefined_externs, true,
|
|
"Don't exit when an undefined extern is called.", "CPU");
|
|
DEFINE_bool(emit_source_annotations, false,
|
|
"Add extra movs and nops to make disassembly easier to read.",
|
|
"CPU");
|
|
DEFINE_bool(resolve_rel32_guest_calls, true,
|
|
"Experimental optimization, directly call already resolved "
|
|
"functions via x86 rel32 call/jmp",
|
|
"CPU");
|
|
|
|
DEFINE_bool(enable_incorrect_roundingmode_behavior, false,
|
|
"Disables the FPU/VMX MXCSR sharing workaround, potentially "
|
|
"causing incorrect rounding behavior and denormal handling in VMX "
|
|
"code. The workaround may cause reduced CPU performance but is a "
|
|
"more accurate emulation",
|
|
"x64");
|
|
DEFINE_uint32(align_all_basic_blocks, 0,
|
|
"Aligns the start of all basic blocks to N bytes. Only specify a "
|
|
"power of 2, 16 is the recommended value. Results in larger "
|
|
"icache usage, but potentially faster loops",
|
|
"x64");
|
|
#if XE_X64_PROFILER_AVAILABLE == 1
|
|
DEFINE_bool(instrument_call_times, false,
|
|
"Compute time taken for functions, for profiling guest code",
|
|
"x64");
|
|
#endif
|
|
namespace xe {
|
|
namespace cpu {
|
|
namespace backend {
|
|
namespace x64 {
|
|
|
|
using xe::cpu::hir::HIRBuilder;
|
|
using xe::cpu::hir::Instr;
|
|
using namespace xe::literals;
|
|
|
|
static const size_t kMaxCodeSize = 1_MiB;
|
|
|
|
static const size_t kStashOffset = 32;
|
|
// static const size_t kStashOffsetHigh = 32 + 32;
|
|
|
|
const uint32_t X64Emitter::gpr_reg_map_[X64Emitter::GPR_COUNT] = {
|
|
Xbyak::Operand::RBX, Xbyak::Operand::R10, Xbyak::Operand::R11,
|
|
Xbyak::Operand::R12, Xbyak::Operand::R13, Xbyak::Operand::R14,
|
|
Xbyak::Operand::R15,
|
|
};
|
|
|
|
const uint32_t X64Emitter::xmm_reg_map_[X64Emitter::XMM_COUNT] = {
|
|
4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,
|
|
};
|
|
|
|
X64Emitter::X64Emitter(X64Backend* backend, XbyakAllocator* allocator)
|
|
: CodeGenerator(kMaxCodeSize, Xbyak::AutoGrow, allocator),
|
|
processor_(backend->processor()),
|
|
backend_(backend),
|
|
code_cache_(backend->code_cache()),
|
|
allocator_(allocator) {
|
|
if (!cpu_.has(Xbyak::util::Cpu::tAVX)) {
|
|
xe::FatalError(
|
|
"Your CPU does not support AVX, which is required by Xenia. See the "
|
|
"FAQ for system requirements at https://xenia.jp");
|
|
return;
|
|
}
|
|
#if 1
|
|
feature_flags_ = amd64::GetFeatureFlags();
|
|
#else
|
|
#define TEST_EMIT_FEATURE(emit, ext) \
|
|
if ((cvars::x64_extension_mask & emit) == emit) { \
|
|
feature_flags_ |= (cpu_.has(ext) ? emit : 0); \
|
|
}
|
|
|
|
TEST_EMIT_FEATURE(kX64EmitAVX2, Xbyak::util::Cpu::tAVX2);
|
|
TEST_EMIT_FEATURE(kX64EmitFMA, Xbyak::util::Cpu::tFMA);
|
|
TEST_EMIT_FEATURE(kX64EmitLZCNT, Xbyak::util::Cpu::tLZCNT);
|
|
TEST_EMIT_FEATURE(kX64EmitBMI1, Xbyak::util::Cpu::tBMI1);
|
|
TEST_EMIT_FEATURE(kX64EmitBMI2, Xbyak::util::Cpu::tBMI2);
|
|
TEST_EMIT_FEATURE(kX64EmitMovbe, Xbyak::util::Cpu::tMOVBE);
|
|
TEST_EMIT_FEATURE(kX64EmitGFNI, Xbyak::util::Cpu::tGFNI);
|
|
TEST_EMIT_FEATURE(kX64EmitAVX512F, Xbyak::util::Cpu::tAVX512F);
|
|
TEST_EMIT_FEATURE(kX64EmitAVX512VL, Xbyak::util::Cpu::tAVX512VL);
|
|
TEST_EMIT_FEATURE(kX64EmitAVX512BW, Xbyak::util::Cpu::tAVX512BW);
|
|
TEST_EMIT_FEATURE(kX64EmitAVX512DQ, Xbyak::util::Cpu::tAVX512DQ);
|
|
TEST_EMIT_FEATURE(kX64EmitAVX512VBMI, Xbyak::util::Cpu::tAVX512VBMI);
|
|
TEST_EMIT_FEATURE(kX64EmitPrefetchW, Xbyak::util::Cpu::tPREFETCHW);
|
|
#undef TEST_EMIT_FEATURE
|
|
/*
|
|
fix for xbyak bug/omission, amd cpus are never checked for lzcnt. fixed in
|
|
latest version of xbyak
|
|
*/
|
|
unsigned int data[4];
|
|
Xbyak::util::Cpu::getCpuid(0x80000001, data);
|
|
unsigned amd_flags = data[2];
|
|
if (amd_flags & (1U << 5)) {
|
|
if ((cvars::x64_extension_mask & kX64EmitLZCNT) == kX64EmitLZCNT) {
|
|
feature_flags_ |= kX64EmitLZCNT;
|
|
}
|
|
}
|
|
// todo: although not reported by cpuid, zen 1 and zen+ also have fma4
|
|
if (amd_flags & (1U << 16)) {
|
|
if ((cvars::x64_extension_mask & kX64EmitFMA4) == kX64EmitFMA4) {
|
|
feature_flags_ |= kX64EmitFMA4;
|
|
}
|
|
}
|
|
if (amd_flags & (1U << 21)) {
|
|
if ((cvars::x64_extension_mask & kX64EmitTBM) == kX64EmitTBM) {
|
|
feature_flags_ |= kX64EmitTBM;
|
|
}
|
|
}
|
|
if (amd_flags & (1U << 11)) {
|
|
if ((cvars::x64_extension_mask & kX64EmitXOP) == kX64EmitXOP) {
|
|
feature_flags_ |= kX64EmitXOP;
|
|
XELOGCPU("Cpu support XOP!\n\n");
|
|
}
|
|
}
|
|
if (cpu_.has(Xbyak::util::Cpu::tAMD)) {
|
|
bool is_zennish = cpu_.displayFamily >= 0x17;
|
|
/*
|
|
chrispy: according to agner's tables, all amd architectures that
|
|
we support (ones with avx) have the same timings for
|
|
jrcxz/loop/loope/loopne as for other jmps
|
|
*/
|
|
feature_flags_ |= kX64FastJrcx;
|
|
feature_flags_ |= kX64FastLoop;
|
|
if (is_zennish) {
|
|
// ik that i heard somewhere that this is the case for zen, but i need to
|
|
// verify. cant find my original source for that.
|
|
// todo: ask agner?
|
|
feature_flags_ |= kX64FlagsIndependentVars;
|
|
}
|
|
}
|
|
#endif
|
|
may_use_membase32_as_zero_reg_ =
|
|
static_cast<uint32_t>(reinterpret_cast<uintptr_t>(
|
|
processor()->memory()->virtual_membase())) == 0;
|
|
}
|
|
|
|
X64Emitter::~X64Emitter() = default;
|
|
|
|
bool X64Emitter::Emit(GuestFunction* function, HIRBuilder* builder,
|
|
uint32_t debug_info_flags, FunctionDebugInfo* debug_info,
|
|
void** out_code_address, size_t* out_code_size,
|
|
std::vector<SourceMapEntry>* out_source_map) {
|
|
SCOPE_profile_cpu_f("cpu");
|
|
guest_module_ = dynamic_cast<XexModule*>(function->module());
|
|
current_guest_function_ = function->address();
|
|
// Reset.
|
|
debug_info_ = debug_info;
|
|
debug_info_flags_ = debug_info_flags;
|
|
trace_data_ = &function->trace_data();
|
|
source_map_arena_.Reset();
|
|
|
|
// Fill the generator with code.
|
|
EmitFunctionInfo func_info = {};
|
|
if (!Emit(builder, func_info)) {
|
|
return false;
|
|
}
|
|
|
|
// Copy the final code to the cache and relocate it.
|
|
*out_code_size = getSize();
|
|
*out_code_address = Emplace(func_info, function);
|
|
|
|
// Stash source map.
|
|
source_map_arena_.CloneContents(out_source_map);
|
|
|
|
return true;
|
|
}
|
|
#pragma pack(push, 1)
|
|
struct RGCEmitted {
|
|
uint8_t ff_;
|
|
uint32_t rgcid_;
|
|
};
|
|
#pragma pack(pop)
|
|
|
|
#if 0
|
|
void X64Emitter::InjectCallAddresses(void* new_execute_address) {
|
|
for (auto&& callsite : call_sites_) {
|
|
RGCEmitted* hunter = (RGCEmitted*)new_execute_address;
|
|
while (hunter->ff_ != 0xFF || hunter->rgcid_ != callsite.offset_) {
|
|
hunter =
|
|
reinterpret_cast<RGCEmitted*>(reinterpret_cast<char*>(hunter) + 1);
|
|
}
|
|
|
|
hunter->ff_ = callsite.is_jump_ ? 0xE9 : 0xE8;
|
|
hunter->rgcid_ =
|
|
static_cast<uint32_t>(static_cast<intptr_t>(callsite.destination_) -
|
|
reinterpret_cast<intptr_t>(hunter + 1));
|
|
}
|
|
}
|
|
|
|
#else
|
|
void X64Emitter::InjectCallAddresses(void* new_execute_address) {
|
|
#if 0
|
|
RGCEmitted* hunter = (RGCEmitted*)new_execute_address;
|
|
|
|
std::map<uint32_t, ResolvableGuestCall*> id_to_rgc{};
|
|
|
|
for (auto&& callsite : call_sites_) {
|
|
id_to_rgc[callsite.offset_] = &callsite;
|
|
}
|
|
#else
|
|
RGCEmitted* hunter = (RGCEmitted*)new_execute_address;
|
|
for (auto&& callsite : call_sites_) {
|
|
while (hunter->ff_ != 0xFF || hunter->rgcid_ != callsite.offset_) {
|
|
hunter =
|
|
reinterpret_cast<RGCEmitted*>(reinterpret_cast<char*>(hunter) + 1);
|
|
}
|
|
|
|
hunter->ff_ = callsite.is_jump_ ? 0xE9 : 0xE8;
|
|
hunter->rgcid_ =
|
|
static_cast<uint32_t>(static_cast<intptr_t>(callsite.destination_) -
|
|
reinterpret_cast<intptr_t>(hunter + 1));
|
|
}
|
|
#endif
|
|
}
|
|
#endif
|
|
void* X64Emitter::Emplace(const EmitFunctionInfo& func_info,
|
|
GuestFunction* function) {
|
|
// To avoid changing xbyak, we do a switcharoo here.
|
|
// top_ points to the Xbyak buffer, and since we are in AutoGrow mode
|
|
// it has pending relocations. We copy the top_ to our buffer, swap the
|
|
// pointer, relocate, then return the original scratch pointer for use.
|
|
// top_ is used by Xbyak's ready() as both write base pointer and the absolute
|
|
// address base, which would not work on platforms not supporting writable
|
|
// executable memory, but Xenia doesn't use absolute label addresses in the
|
|
// generated code.
|
|
uint8_t* old_address = top_;
|
|
void* new_execute_address;
|
|
void* new_write_address;
|
|
assert_true(func_info.code_size.total == size_);
|
|
if (function) {
|
|
code_cache_->PlaceGuestCode(function->address(), top_, func_info, function,
|
|
new_execute_address, new_write_address);
|
|
|
|
if (cvars::resolve_rel32_guest_calls) {
|
|
InjectCallAddresses(new_execute_address);
|
|
}
|
|
} else {
|
|
code_cache_->PlaceHostCode(0, top_, func_info, new_execute_address,
|
|
new_write_address);
|
|
}
|
|
top_ = reinterpret_cast<uint8_t*>(new_write_address);
|
|
ready();
|
|
top_ = old_address;
|
|
reset();
|
|
call_sites_.clear();
|
|
tail_code_.clear();
|
|
for (auto&& cached_label : label_cache_) {
|
|
delete cached_label;
|
|
}
|
|
label_cache_.clear();
|
|
return new_execute_address;
|
|
}
|
|
|
|
bool X64Emitter::Emit(HIRBuilder* builder, EmitFunctionInfo& func_info) {
|
|
Xbyak::Label epilog_label;
|
|
epilog_label_ = &epilog_label;
|
|
|
|
// Calculate stack size. We need to align things to their natural sizes.
|
|
// This could be much better (sort by type/etc).
|
|
auto locals = builder->locals();
|
|
size_t stack_offset = StackLayout::GUEST_STACK_SIZE;
|
|
for (auto it = locals.begin(); it != locals.end(); ++it) {
|
|
auto slot = *it;
|
|
size_t type_size = GetTypeSize(slot->type);
|
|
|
|
// Align to natural size.
|
|
stack_offset = xe::align(stack_offset, type_size);
|
|
slot->set_constant((uint32_t)stack_offset);
|
|
stack_offset += type_size;
|
|
}
|
|
|
|
// Ensure 16b alignment.
|
|
stack_offset -= StackLayout::GUEST_STACK_SIZE;
|
|
stack_offset = xe::align(stack_offset, static_cast<size_t>(16));
|
|
|
|
struct _code_offsets {
|
|
size_t prolog;
|
|
size_t prolog_stack_alloc;
|
|
size_t body;
|
|
size_t epilog;
|
|
size_t tail;
|
|
} code_offsets = {};
|
|
|
|
code_offsets.prolog = getSize();
|
|
|
|
// Function prolog.
|
|
// Must be 16b aligned.
|
|
// Windows is very strict about the form of this and the epilog:
|
|
// https://docs.microsoft.com/en-us/cpp/build/prolog-and-epilog?view=vs-2017
|
|
// IMPORTANT: any changes to the prolog must be kept in sync with
|
|
// X64CodeCache, which dynamically generates exception information.
|
|
// Adding or changing anything here must be matched!
|
|
const size_t stack_size = StackLayout::GUEST_STACK_SIZE + stack_offset;
|
|
assert_true((stack_size + 8) % 16 == 0);
|
|
func_info.stack_size = stack_size;
|
|
stack_size_ = stack_size;
|
|
|
|
sub(rsp, (uint32_t)stack_size);
|
|
|
|
code_offsets.prolog_stack_alloc = getSize();
|
|
code_offsets.body = getSize();
|
|
xor_(eax, eax);
|
|
/*
|
|
* chrispy: removed this, it serves no purpose
|
|
mov(qword[rsp + StackLayout::GUEST_CTX_HOME], GetContextReg());
|
|
*/
|
|
|
|
mov(qword[rsp + StackLayout::GUEST_RET_ADDR], rcx);
|
|
|
|
mov(qword[rsp + StackLayout::GUEST_CALL_RET_ADDR], rax); // 0
|
|
|
|
#if XE_X64_PROFILER_AVAILABLE == 1
|
|
if (cvars::instrument_call_times) {
|
|
mov(rdx, 0x7ffe0014); // load pointer to kusershared systemtime
|
|
mov(rdx, qword[rdx]);
|
|
mov(qword[rsp + StackLayout::GUEST_PROFILER_START],
|
|
rdx); // save time for end of function
|
|
}
|
|
#endif
|
|
// Safe now to do some tracing.
|
|
if (debug_info_flags_ & DebugInfoFlags::kDebugInfoTraceFunctions) {
|
|
// We require 32-bit addresses.
|
|
assert_true(uint64_t(trace_data_->header()) < UINT_MAX);
|
|
auto trace_header = trace_data_->header();
|
|
|
|
// Call count.
|
|
lock();
|
|
inc(qword[low_address(&trace_header->function_call_count)]);
|
|
|
|
// Get call history slot.
|
|
static_assert(FunctionTraceData::kFunctionCallerHistoryCount == 4,
|
|
"bitmask depends on count");
|
|
mov(rax, qword[low_address(&trace_header->function_call_count)]);
|
|
and_(rax, 0b00000011);
|
|
|
|
// Record call history value into slot (guest addr in RDX).
|
|
mov(dword[Xbyak::RegExp(uint32_t(uint64_t(
|
|
low_address(&trace_header->function_caller_history)))) +
|
|
rax * 4],
|
|
edx);
|
|
|
|
// Calling thread. Load ax with thread ID.
|
|
EmitGetCurrentThreadId();
|
|
lock();
|
|
bts(qword[low_address(&trace_header->function_thread_use)], rax);
|
|
}
|
|
|
|
// Load membase.
|
|
/*
|
|
* chrispy: removed this, as long as we load it in HostToGuestThunk we can
|
|
count on no other code modifying it. mov(GetMembaseReg(),
|
|
qword[GetContextReg() + offsetof(ppc::PPCContext, virtual_membase)]);
|
|
*/
|
|
// Body.
|
|
auto block = builder->first_block();
|
|
while (block) {
|
|
ForgetMxcsrMode(); // at start of block, mxcsr mode is undefined
|
|
|
|
// Mark block labels.
|
|
auto label = block->label_head;
|
|
while (label) {
|
|
L(label->name);
|
|
label = label->next;
|
|
}
|
|
|
|
if (cvars::align_all_basic_blocks) {
|
|
align(cvars::align_all_basic_blocks, true);
|
|
}
|
|
// Process instructions.
|
|
const Instr* instr = block->instr_head;
|
|
while (instr) {
|
|
const Instr* new_tail = instr;
|
|
if (!SelectSequence(this, instr, &new_tail)) {
|
|
// No sequence found!
|
|
// NOTE: If you encounter this after adding a new instruction, do a full
|
|
// rebuild!
|
|
assert_always();
|
|
XELOGE("Unable to process HIR opcode {}", GetOpcodeName(instr->opcode));
|
|
break;
|
|
}
|
|
instr = new_tail;
|
|
}
|
|
|
|
block = block->next;
|
|
}
|
|
|
|
// Function epilog.
|
|
L(epilog_label);
|
|
epilog_label_ = nullptr;
|
|
EmitTraceUserCallReturn();
|
|
/*
|
|
* chrispy: removed this, it serves no purpose
|
|
mov(GetContextReg(), qword[rsp + StackLayout::GUEST_CTX_HOME]);
|
|
*/
|
|
code_offsets.epilog = getSize();
|
|
EmitProfilerEpilogue();
|
|
|
|
add(rsp, (uint32_t)stack_size);
|
|
ret();
|
|
// todo: do some kind of sorting by alignment?
|
|
for (auto&& tail_item : tail_code_) {
|
|
if (tail_item.alignment) {
|
|
align(tail_item.alignment);
|
|
}
|
|
tail_item.func(*this, tail_item.label);
|
|
}
|
|
|
|
code_offsets.tail = getSize();
|
|
|
|
if (cvars::emit_source_annotations) {
|
|
nop(5);
|
|
}
|
|
|
|
assert_zero(code_offsets.prolog);
|
|
func_info.code_size.total = getSize();
|
|
func_info.code_size.prolog = code_offsets.body - code_offsets.prolog;
|
|
func_info.code_size.body = code_offsets.epilog - code_offsets.body;
|
|
func_info.code_size.epilog = code_offsets.tail - code_offsets.epilog;
|
|
func_info.code_size.tail = getSize() - code_offsets.tail;
|
|
func_info.prolog_stack_alloc_offset =
|
|
code_offsets.prolog_stack_alloc - code_offsets.prolog;
|
|
|
|
return true;
|
|
}
|
|
// dont use rax, we do this in tail call handling
|
|
void X64Emitter::EmitProfilerEpilogue() {
|
|
#if XE_X64_PROFILER_AVAILABLE == 1
|
|
if (cvars::instrument_call_times) {
|
|
uint64_t* profiler_entry =
|
|
backend()->GetProfilerRecordForFunction(current_guest_function_);
|
|
mov(ecx, 0x7ffe0014);
|
|
mov(rdx, qword[rcx]);
|
|
mov(rbx, (uintptr_t)profiler_entry);
|
|
sub(rdx, qword[rsp + StackLayout::GUEST_PROFILER_START]);
|
|
|
|
// atomic add our time to the profiler entry
|
|
// this could be atomic free if we had per thread profile counts, and on a
|
|
// threads exit we lock and sum up to the global counts, which would make
|
|
// this a few cycles less intrusive, but its good enough for now
|
|
// actually... lets just try without atomics lol
|
|
// lock();
|
|
add(qword[rbx], rdx);
|
|
}
|
|
#endif
|
|
}
|
|
|
|
void X64Emitter::MarkSourceOffset(const Instr* i) {
|
|
auto entry = source_map_arena_.Alloc<SourceMapEntry>();
|
|
entry->guest_address = static_cast<uint32_t>(i->src1.offset);
|
|
entry->hir_offset = uint32_t(i->block->ordinal << 16) | i->ordinal;
|
|
entry->code_offset = static_cast<uint32_t>(getSize());
|
|
|
|
if (cvars::emit_source_annotations) {
|
|
nop(2);
|
|
mov(eax, entry->guest_address);
|
|
nop(2);
|
|
}
|
|
|
|
if (debug_info_flags_ & DebugInfoFlags::kDebugInfoTraceFunctionCoverage) {
|
|
uint32_t instruction_index =
|
|
(entry->guest_address - trace_data_->start_address()) / 4;
|
|
lock();
|
|
inc(qword[low_address(trace_data_->instruction_execute_counts() +
|
|
instruction_index * 8)]);
|
|
}
|
|
}
|
|
|
|
void X64Emitter::EmitGetCurrentThreadId() {
|
|
// rsi must point to context. We could fetch from the stack if needed.
|
|
mov(ax, word[GetContextReg() + offsetof(ppc::PPCContext, thread_id)]);
|
|
}
|
|
|
|
void X64Emitter::EmitTraceUserCallReturn() {}
|
|
|
|
void X64Emitter::DebugBreak() {
|
|
// TODO(benvanik): notify debugger.
|
|
db(0xCC);
|
|
}
|
|
|
|
uint64_t TrapDebugPrint(void* raw_context, uint64_t address) {
|
|
auto thread_state =
|
|
reinterpret_cast<ppc::PPCContext_s*>(raw_context)->thread_state;
|
|
uint32_t str_ptr = uint32_t(thread_state->context()->r[3]);
|
|
// uint16_t str_len = uint16_t(thread_state->context()->r[4]);
|
|
auto str = thread_state->memory()->TranslateVirtual<const char*>(str_ptr);
|
|
// TODO(benvanik): truncate to length?
|
|
XELOGD("(DebugPrint) {}", str);
|
|
|
|
if (cvars::debugprint_trap_log) {
|
|
debugging::DebugPrint("(DebugPrint) {}", str);
|
|
}
|
|
|
|
return 0;
|
|
}
|
|
|
|
uint64_t TrapDebugBreak(void* raw_context, uint64_t address) {
|
|
auto thread_state =
|
|
reinterpret_cast<ppc::PPCContext_s*>(raw_context)->thread_state;
|
|
XELOGE("tw/td forced trap hit! This should be a crash!");
|
|
if (cvars::break_on_debugbreak) {
|
|
xe::debugging::Break();
|
|
}
|
|
return 0;
|
|
}
|
|
|
|
void X64Emitter::Trap(uint16_t trap_type) {
|
|
switch (trap_type) {
|
|
case 20:
|
|
case 26:
|
|
// 0x0FE00014 is a 'debug print' where r3 = buffer r4 = length
|
|
CallNative(TrapDebugPrint, 0);
|
|
break;
|
|
case 0:
|
|
case 22:
|
|
// Always trap?
|
|
// TODO(benvanik): post software interrupt to debugger.
|
|
CallNative(TrapDebugBreak, 0);
|
|
break;
|
|
case 25:
|
|
// ?
|
|
break;
|
|
default:
|
|
XELOGW("Unknown trap type {}", trap_type);
|
|
db(0xCC);
|
|
break;
|
|
}
|
|
}
|
|
|
|
void X64Emitter::UnimplementedInstr(const hir::Instr* i) {
|
|
// TODO(benvanik): notify debugger.
|
|
db(0xCC);
|
|
assert_always();
|
|
}
|
|
|
|
// This is used by the X64ThunkEmitter's ResolveFunctionThunk.
|
|
uint64_t ResolveFunction(void* raw_context, uint64_t target_address) {
|
|
auto thread_state =
|
|
reinterpret_cast<ppc::PPCContext_s*>(raw_context)->thread_state;
|
|
|
|
// TODO(benvanik): required?
|
|
assert_not_zero(target_address);
|
|
|
|
auto fn = thread_state->processor()->ResolveFunction(
|
|
static_cast<uint32_t>(target_address));
|
|
assert_not_null(fn);
|
|
auto x64_fn = static_cast<X64Function*>(fn);
|
|
uint64_t addr = reinterpret_cast<uint64_t>(x64_fn->machine_code());
|
|
|
|
return addr;
|
|
}
|
|
|
|
void X64Emitter::Call(const hir::Instr* instr, GuestFunction* function) {
|
|
assert_not_null(function);
|
|
ForgetMxcsrMode();
|
|
auto fn = static_cast<X64Function*>(function);
|
|
// Resolve address to the function to call and store in rax.
|
|
|
|
if (cvars::resolve_rel32_guest_calls && fn->machine_code()) {
|
|
ResolvableGuestCall rgc;
|
|
rgc.destination_ = uint32_t(uint64_t(fn->machine_code()));
|
|
rgc.offset_ = current_rgc_id_;
|
|
current_rgc_id_++;
|
|
|
|
if (!(instr->flags & hir::CALL_TAIL)) {
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_CALL_RET_ADDR]);
|
|
|
|
db(0xFF);
|
|
rgc.is_jump_ = false;
|
|
|
|
dd(rgc.offset_);
|
|
|
|
} else {
|
|
// tail call
|
|
EmitTraceUserCallReturn();
|
|
|
|
rgc.is_jump_ = true;
|
|
// Pass the callers return address over.
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_RET_ADDR]);
|
|
|
|
add(rsp, static_cast<uint32_t>(stack_size()));
|
|
db(0xFF);
|
|
dd(rgc.offset_);
|
|
}
|
|
call_sites_.push_back(rgc);
|
|
return;
|
|
}
|
|
|
|
if (fn->machine_code()) {
|
|
// TODO(benvanik): is it worth it to do this? It removes the need for
|
|
// a ResolveFunction call, but makes the table less useful.
|
|
assert_zero(uint64_t(fn->machine_code()) & 0xFFFFFFFF00000000);
|
|
// todo: this should be changed so that we can actually do a call to
|
|
// fn->machine_code. the code will be emitted near us, so 32 bit rel jmp
|
|
// should be possible
|
|
mov(eax, uint32_t(uint64_t(fn->machine_code())));
|
|
} else if (code_cache_->has_indirection_table()) {
|
|
// Load the pointer to the indirection table maintained in X64CodeCache.
|
|
// The target dword will either contain the address of the generated code
|
|
// or a thunk to ResolveAddress.
|
|
mov(ebx, function->address());
|
|
mov(eax, dword[ebx]);
|
|
} else {
|
|
// Old-style resolve.
|
|
// Not too important because indirection table is almost always available.
|
|
// TODO: Overwrite the call-site with a straight call.
|
|
CallNative(&ResolveFunction, function->address());
|
|
}
|
|
|
|
// Actually jump/call to rax.
|
|
if (instr->flags & hir::CALL_TAIL) {
|
|
// Since we skip the prolog we need to mark the return here.
|
|
EmitTraceUserCallReturn();
|
|
EmitProfilerEpilogue();
|
|
// Pass the callers return address over.
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_RET_ADDR]);
|
|
|
|
add(rsp, static_cast<uint32_t>(stack_size()));
|
|
jmp(rax);
|
|
} else {
|
|
// Return address is from the previous SET_RETURN_ADDRESS.
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_CALL_RET_ADDR]);
|
|
|
|
call(rax);
|
|
}
|
|
}
|
|
|
|
void X64Emitter::CallIndirect(const hir::Instr* instr,
|
|
const Xbyak::Reg64& reg) {
|
|
ForgetMxcsrMode();
|
|
// Check if return.
|
|
if (instr->flags & hir::CALL_POSSIBLE_RETURN) {
|
|
cmp(reg.cvt32(), dword[rsp + StackLayout::GUEST_RET_ADDR]);
|
|
je(epilog_label(), CodeGenerator::T_NEAR);
|
|
}
|
|
|
|
// Load the pointer to the indirection table maintained in X64CodeCache.
|
|
// The target dword will either contain the address of the generated code
|
|
// or a thunk to ResolveAddress.
|
|
if (code_cache_->has_indirection_table()) {
|
|
if (reg.cvt32() != ebx) {
|
|
mov(ebx, reg.cvt32());
|
|
}
|
|
mov(eax, dword[ebx]);
|
|
} else {
|
|
// Old-style resolve.
|
|
// Not too important because indirection table is almost always available.
|
|
mov(edx, reg.cvt32());
|
|
mov(rax, reinterpret_cast<uint64_t>(ResolveFunction));
|
|
mov(rcx, GetContextReg());
|
|
call(rax);
|
|
}
|
|
|
|
// Actually jump/call to rax.
|
|
if (instr->flags & hir::CALL_TAIL) {
|
|
// Since we skip the prolog we need to mark the return here.
|
|
EmitTraceUserCallReturn();
|
|
EmitProfilerEpilogue();
|
|
// Pass the callers return address over.
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_RET_ADDR]);
|
|
|
|
add(rsp, static_cast<uint32_t>(stack_size()));
|
|
jmp(rax);
|
|
} else {
|
|
// Return address is from the previous SET_RETURN_ADDRESS.
|
|
mov(rcx, qword[rsp + StackLayout::GUEST_CALL_RET_ADDR]);
|
|
|
|
call(rax);
|
|
}
|
|
}
|
|
|
|
uint64_t UndefinedCallExtern(void* raw_context, uint64_t function_ptr) {
|
|
auto function = reinterpret_cast<Function*>(function_ptr);
|
|
if (!cvars::ignore_undefined_externs) {
|
|
xe::FatalError(fmt::format("undefined extern call to {:08X} {}",
|
|
function->address(), function->name().c_str()));
|
|
} else {
|
|
XELOGE("undefined extern call to {:08X} {}", function->address(),
|
|
function->name());
|
|
}
|
|
return 0;
|
|
}
|
|
void X64Emitter::CallExtern(const hir::Instr* instr, const Function* function) {
|
|
ForgetMxcsrMode();
|
|
bool undefined = true;
|
|
if (function->behavior() == Function::Behavior::kBuiltin) {
|
|
auto builtin_function = static_cast<const BuiltinFunction*>(function);
|
|
if (builtin_function->handler()) {
|
|
undefined = false;
|
|
// rcx = target function
|
|
// rdx = arg0
|
|
// r8 = arg1
|
|
// r9 = arg2
|
|
mov(rcx, reinterpret_cast<uint64_t>(builtin_function->handler()));
|
|
mov(rdx, reinterpret_cast<uint64_t>(builtin_function->arg0()));
|
|
mov(r8, reinterpret_cast<uint64_t>(builtin_function->arg1()));
|
|
call(backend()->guest_to_host_thunk());
|
|
// rax = host return
|
|
}
|
|
} else if (function->behavior() == Function::Behavior::kExtern) {
|
|
auto extern_function = static_cast<const GuestFunction*>(function);
|
|
if (extern_function->extern_handler()) {
|
|
undefined = false;
|
|
// rcx = target function
|
|
// rdx = arg0
|
|
// r8 = arg1
|
|
// r9 = arg2
|
|
mov(rcx, reinterpret_cast<uint64_t>(extern_function->extern_handler()));
|
|
mov(rdx,
|
|
qword[GetContextReg() + offsetof(ppc::PPCContext, kernel_state)]);
|
|
call(backend()->guest_to_host_thunk());
|
|
// rax = host return
|
|
}
|
|
}
|
|
if (undefined) {
|
|
CallNative(UndefinedCallExtern, reinterpret_cast<uint64_t>(function));
|
|
}
|
|
}
|
|
|
|
void X64Emitter::CallNative(void* fn) { CallNativeSafe(fn); }
|
|
|
|
void X64Emitter::CallNative(uint64_t (*fn)(void* raw_context)) {
|
|
CallNativeSafe(reinterpret_cast<void*>(fn));
|
|
}
|
|
|
|
void X64Emitter::CallNative(uint64_t (*fn)(void* raw_context, uint64_t arg0)) {
|
|
CallNativeSafe(reinterpret_cast<void*>(fn));
|
|
}
|
|
|
|
void X64Emitter::CallNative(uint64_t (*fn)(void* raw_context, uint64_t arg0),
|
|
uint64_t arg0) {
|
|
mov(GetNativeParam(0), arg0);
|
|
CallNativeSafe(reinterpret_cast<void*>(fn));
|
|
}
|
|
|
|
void X64Emitter::CallNativeSafe(void* fn) {
|
|
// rcx = target function
|
|
// rdx = arg0
|
|
// r8 = arg1
|
|
// r9 = arg2
|
|
mov(rcx, reinterpret_cast<uint64_t>(fn));
|
|
call(backend()->guest_to_host_thunk());
|
|
// rax = host return
|
|
}
|
|
|
|
void X64Emitter::SetReturnAddress(uint64_t value) {
|
|
mov(rax, value);
|
|
mov(qword[rsp + StackLayout::GUEST_CALL_RET_ADDR], rax);
|
|
}
|
|
|
|
Xbyak::Reg64 X64Emitter::GetNativeParam(uint32_t param) {
|
|
if (param == 0)
|
|
return rdx;
|
|
else if (param == 1)
|
|
return r8;
|
|
else if (param == 2)
|
|
return r9;
|
|
|
|
assert_always();
|
|
return r9;
|
|
}
|
|
|
|
// Important: If you change these, you must update the thunks in x64_backend.cc!
|
|
Xbyak::Reg64 X64Emitter::GetContextReg() const { return rsi; }
|
|
Xbyak::Reg64 X64Emitter::GetMembaseReg() const { return rdi; }
|
|
|
|
void X64Emitter::ReloadMembase() {
|
|
mov(GetMembaseReg(),
|
|
qword[GetContextReg() +
|
|
offsetof(ppc::PPCContext, virtual_membase)]); // membase
|
|
}
|
|
|
|
// Len Assembly Byte Sequence
|
|
// ============================================================================
|
|
// 1b NOP 90H
|
|
// 2b 66 NOP 66 90H
|
|
// 3b NOP DWORD ptr [EAX] 0F 1F 00H
|
|
// 4b NOP DWORD ptr [EAX + 00H] 0F 1F 40 00H
|
|
// 5b NOP DWORD ptr [EAX + EAX*1 + 00H] 0F 1F 44 00 00H
|
|
// 6b 66 NOP DWORD ptr [EAX + EAX*1 + 00H] 66 0F 1F 44 00 00H
|
|
// 7b NOP DWORD ptr [EAX + 00000000H] 0F 1F 80 00 00 00 00H
|
|
// 8b NOP DWORD ptr [EAX + EAX*1 + 00000000H] 0F 1F 84 00 00 00 00 00H
|
|
// 9b 66 NOP DWORD ptr [EAX + EAX*1 + 00000000H] 66 0F 1F 84 00 00 00 00 00H
|
|
void X64Emitter::nop(size_t length) {
|
|
for (size_t i = 0; i < length; ++i) {
|
|
db(0x90);
|
|
}
|
|
}
|
|
|
|
bool X64Emitter::ConstantFitsIn32Reg(uint64_t v) {
|
|
if ((v & ~0x7FFFFFFF) == 0) {
|
|
// Fits under 31 bits, so just load using normal mov.
|
|
return true;
|
|
} else if ((v & ~0x7FFFFFFF) == ~0x7FFFFFFF) {
|
|
// Negative number that fits in 32bits.
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
/*
|
|
WARNING: do not use any regs here, addr is often produced by
|
|
ComputeAddressOffset, which may use rax/rdx/rcx in its addr expression
|
|
*/
|
|
void X64Emitter::MovMem64(const Xbyak::RegExp& addr, uint64_t v) {
|
|
uint32_t lowpart = static_cast<uint32_t>(v);
|
|
uint32_t highpart = static_cast<uint32_t>(v >> 32);
|
|
// check whether the constant coincidentally collides with our membase
|
|
if (v == (uintptr_t)processor()->memory()->virtual_membase()) {
|
|
mov(qword[addr], GetMembaseReg());
|
|
} else if ((v & ~0x7FFFFFFF) == 0) {
|
|
// Fits under 31 bits, so just load using normal mov.
|
|
|
|
mov(qword[addr], v);
|
|
} else if ((v & ~0x7FFFFFFF) == ~0x7FFFFFFF) {
|
|
// Negative number that fits in 32bits.
|
|
mov(qword[addr], v);
|
|
} else if (!highpart) {
|
|
// All high bits are zero. It'd be nice if we had a way to load a 32bit
|
|
// immediate without sign extending!
|
|
// TODO(benvanik): this is super common, find a better way.
|
|
if (lowpart == 0 && CanUseMembaseLow32As0()) {
|
|
mov(dword[addr], GetMembaseReg().cvt32());
|
|
} else {
|
|
mov(dword[addr], static_cast<uint32_t>(v));
|
|
}
|
|
if (CanUseMembaseLow32As0()) {
|
|
mov(dword[addr + 4], GetMembaseReg().cvt32());
|
|
} else {
|
|
mov(dword[addr + 4], 0);
|
|
}
|
|
} else {
|
|
// 64bit number that needs double movs.
|
|
|
|
if (lowpart == 0 && CanUseMembaseLow32As0()) {
|
|
mov(dword[addr], GetMembaseReg().cvt32());
|
|
} else {
|
|
mov(dword[addr], lowpart);
|
|
}
|
|
if (highpart == 0 && CanUseMembaseLow32As0()) {
|
|
mov(dword[addr + 4], GetMembaseReg().cvt32());
|
|
} else {
|
|
mov(dword[addr + 4], highpart);
|
|
}
|
|
}
|
|
}
|
|
static inline vec128_t v128_setr_bytes(unsigned char v0, unsigned char v1,
|
|
unsigned char v2, unsigned char v3,
|
|
unsigned char v4, unsigned char v5,
|
|
unsigned char v6, unsigned char v7,
|
|
unsigned char v8, unsigned char v9,
|
|
unsigned char v10, unsigned char v11,
|
|
unsigned char v12, unsigned char v13,
|
|
unsigned char v14, unsigned char v15) {
|
|
vec128_t result;
|
|
|
|
result.u8[0] = v0;
|
|
result.u8[1] = v1;
|
|
result.u8[2] = v2;
|
|
result.u8[3] = v3;
|
|
result.u8[4] = v4;
|
|
result.u8[5] = v5;
|
|
result.u8[6] = v6;
|
|
result.u8[7] = v7;
|
|
result.u8[8] = v8;
|
|
result.u8[9] = v9;
|
|
result.u8[10] = v10;
|
|
result.u8[11] = v11;
|
|
result.u8[12] = v12;
|
|
result.u8[13] = v13;
|
|
result.u8[14] = v14;
|
|
|
|
result.u8[15] = v15;
|
|
return result;
|
|
}
|
|
|
|
static const vec128_t xmm_consts[] = {
|
|
/* XMMZero */ vec128f(0.0f),
|
|
/* XMMByteSwapMask */
|
|
vec128i(0x00010203u, 0x04050607u, 0x08090A0Bu, 0x0C0D0E0Fu),
|
|
/* XMMOne */ vec128f(1.0f),
|
|
/* XMMOnePD */ vec128d(1.0),
|
|
/* XMMNegativeOne */ vec128f(-1.0f, -1.0f, -1.0f, -1.0f),
|
|
/* XMMFFFF */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu),
|
|
/* XMMMaskX16Y16 */
|
|
vec128i(0x0000FFFFu, 0xFFFF0000u, 0x00000000u, 0x00000000u),
|
|
/* XMMFlipX16Y16 */
|
|
vec128i(0x00008000u, 0x00000000u, 0x00000000u, 0x00000000u),
|
|
/* XMMFixX16Y16 */ vec128f(-32768.0f, 0.0f, 0.0f, 0.0f),
|
|
/* XMMNormalizeX16Y16 */
|
|
vec128f(1.0f / 32767.0f, 1.0f / (32767.0f * 65536.0f), 0.0f, 0.0f),
|
|
/* XMM0001 */ vec128f(0.0f, 0.0f, 0.0f, 1.0f),
|
|
/* XMM3301 */ vec128f(3.0f, 3.0f, 0.0f, 1.0f),
|
|
/* XMM3331 */ vec128f(3.0f, 3.0f, 3.0f, 1.0f),
|
|
/* XMM3333 */ vec128f(3.0f, 3.0f, 3.0f, 3.0f),
|
|
/* XMMSignMaskPS */
|
|
vec128i(0x80000000u, 0x80000000u, 0x80000000u, 0x80000000u),
|
|
/* XMMSignMaskPD */
|
|
vec128i(0x00000000u, 0x80000000u, 0x00000000u, 0x80000000u),
|
|
/* XMMAbsMaskPS */
|
|
vec128i(0x7FFFFFFFu, 0x7FFFFFFFu, 0x7FFFFFFFu, 0x7FFFFFFFu),
|
|
/* XMMAbsMaskPD */
|
|
vec128i(0xFFFFFFFFu, 0x7FFFFFFFu, 0xFFFFFFFFu, 0x7FFFFFFFu),
|
|
|
|
/* XMMByteOrderMask */
|
|
vec128i(0x01000302u, 0x05040706u, 0x09080B0Au, 0x0D0C0F0Eu),
|
|
/* XMMPermuteControl15 */ vec128b(15),
|
|
/* XMMPermuteByteMask */ vec128b(0x1F),
|
|
/* XMMPackD3DCOLORSat */ vec128i(0x404000FFu),
|
|
/* XMMPackD3DCOLOR */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu, 0x0C000408u),
|
|
/* XMMUnpackD3DCOLOR */
|
|
vec128i(0xFFFFFF0Eu, 0xFFFFFF0Du, 0xFFFFFF0Cu, 0xFFFFFF0Fu),
|
|
/* XMMPackFLOAT16_2 */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu, 0x01000302u),
|
|
/* XMMUnpackFLOAT16_2 */
|
|
vec128i(0x0D0C0F0Eu, 0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu),
|
|
/* XMMPackFLOAT16_4 */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0x01000302u, 0x05040706u),
|
|
/* XMMUnpackFLOAT16_4 */
|
|
vec128i(0x09080B0Au, 0x0D0C0F0Eu, 0xFFFFFFFFu, 0xFFFFFFFFu),
|
|
/* XMMPackSHORT_Min */ vec128i(0x403F8001u),
|
|
/* XMMPackSHORT_Max */ vec128i(0x40407FFFu),
|
|
/* XMMPackSHORT_2 */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0xFFFFFFFFu, 0x01000504u),
|
|
/* XMMPackSHORT_4 */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0x01000504u, 0x09080D0Cu),
|
|
/* XMMUnpackSHORT_2 */
|
|
vec128i(0xFFFF0F0Eu, 0xFFFF0D0Cu, 0xFFFFFFFFu, 0xFFFFFFFFu),
|
|
/* XMMUnpackSHORT_4 */
|
|
vec128i(0xFFFF0B0Au, 0xFFFF0908u, 0xFFFF0F0Eu, 0xFFFF0D0Cu),
|
|
/* XMMUnpackSHORT_Overflow */ vec128i(0x403F8000u),
|
|
/* XMMPackUINT_2101010_MinUnpacked */
|
|
vec128i(0x403FFE01u, 0x403FFE01u, 0x403FFE01u, 0x40400000u),
|
|
/* XMMPackUINT_2101010_MaxUnpacked */
|
|
vec128i(0x404001FFu, 0x404001FFu, 0x404001FFu, 0x40400003u),
|
|
/* XMMPackUINT_2101010_MaskUnpacked */
|
|
vec128i(0x3FFu, 0x3FFu, 0x3FFu, 0x3u),
|
|
/* XMMPackUINT_2101010_MaskPacked */
|
|
vec128i(0x3FFu, 0x3FFu << 10, 0x3FFu << 20, 0x3u << 30),
|
|
/* XMMPackUINT_2101010_Shift */ vec128i(0, 10, 20, 30),
|
|
/* XMMUnpackUINT_2101010_Overflow */ vec128i(0x403FFE00u),
|
|
/* XMMPackULONG_4202020_MinUnpacked */
|
|
vec128i(0x40380001u, 0x40380001u, 0x40380001u, 0x40400000u),
|
|
/* XMMPackULONG_4202020_MaxUnpacked */
|
|
vec128i(0x4047FFFFu, 0x4047FFFFu, 0x4047FFFFu, 0x4040000Fu),
|
|
/* XMMPackULONG_4202020_MaskUnpacked */
|
|
vec128i(0xFFFFFu, 0xFFFFFu, 0xFFFFFu, 0xFu),
|
|
/* XMMPackULONG_4202020_PermuteXZ */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0x0A0908FFu, 0xFF020100u),
|
|
/* XMMPackULONG_4202020_PermuteYW */
|
|
vec128i(0xFFFFFFFFu, 0xFFFFFFFFu, 0x0CFFFF06u, 0x0504FFFFu),
|
|
/* XMMUnpackULONG_4202020_Permute */
|
|
vec128i(0xFF0E0D0Cu, 0xFF0B0A09u, 0xFF080F0Eu, 0xFFFFFF0Bu),
|
|
/* XMMUnpackULONG_4202020_Overflow */ vec128i(0x40380000u),
|
|
/* XMMOneOver255 */ vec128f(1.0f / 255.0f),
|
|
/* XMMMaskEvenPI16 */
|
|
vec128i(0x0000FFFFu, 0x0000FFFFu, 0x0000FFFFu, 0x0000FFFFu),
|
|
/* XMMShiftMaskEvenPI16 */
|
|
vec128i(0x0000000Fu, 0x0000000Fu, 0x0000000Fu, 0x0000000Fu),
|
|
/* XMMShiftMaskPS */
|
|
vec128i(0x0000001Fu, 0x0000001Fu, 0x0000001Fu, 0x0000001Fu),
|
|
/* XMMShiftByteMask */
|
|
vec128i(0x000000FFu, 0x000000FFu, 0x000000FFu, 0x000000FFu),
|
|
/* XMMSwapWordMask */
|
|
vec128i(0x03030303u, 0x03030303u, 0x03030303u, 0x03030303u),
|
|
/* XMMUnsignedDwordMax */
|
|
vec128i(0xFFFFFFFFu, 0x00000000u, 0xFFFFFFFFu, 0x00000000u),
|
|
/* XMM255 */ vec128f(255.0f),
|
|
/* XMMPI32 */ vec128i(32),
|
|
/* XMMSignMaskI8 */
|
|
vec128i(0x80808080u, 0x80808080u, 0x80808080u, 0x80808080u),
|
|
/* XMMSignMaskI16 */
|
|
vec128i(0x80008000u, 0x80008000u, 0x80008000u, 0x80008000u),
|
|
/* XMMSignMaskI32 */
|
|
vec128i(0x80000000u, 0x80000000u, 0x80000000u, 0x80000000u),
|
|
/* XMMSignMaskF32 */
|
|
vec128i(0x80000000u, 0x80000000u, 0x80000000u, 0x80000000u),
|
|
/* XMMShortMinPS */ vec128f(SHRT_MIN),
|
|
/* XMMShortMaxPS */ vec128f(SHRT_MAX),
|
|
/* XMMIntMin */ vec128i(INT_MIN),
|
|
/* XMMIntMax */ vec128i(INT_MAX),
|
|
/* XMMIntMaxPD */ vec128d(INT_MAX),
|
|
/* XMMPosIntMinPS */ vec128f((float)0x80000000u),
|
|
/* XMMQNaN */ vec128i(0x7FC00000u),
|
|
/* XMMInt127 */ vec128i(0x7Fu),
|
|
/* XMM2To32 */ vec128f(0x1.0p32f),
|
|
/* XMMFloatInf */ vec128i(0x7f800000),
|
|
|
|
/* XMMIntsToBytes*/
|
|
v128_setr_bytes(0, 4, 8, 12, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80, 0x80,
|
|
0x80, 0x80, 0x80, 0x80),
|
|
/*XMMShortsToBytes*/
|
|
v128_setr_bytes(0, 2, 4, 6, 8, 10, 12, 14, 0x80, 0x80, 0x80, 0x80, 0x80,
|
|
0x80, 0x80, 0x80),
|
|
/*XMMLVSLTableBase*/
|
|
vec128b(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15),
|
|
/*XMMLVSRTableBase*/
|
|
vec128b(16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31),
|
|
/* XMMSingleDenormalMask */
|
|
vec128i(0x7f800000),
|
|
/* XMMThreeFloatMask */
|
|
vec128i(~0U, ~0U, ~0U, 0U),
|
|
/*
|
|
XMMF16UnpackLCPI2
|
|
*/
|
|
|
|
vec128i(0x38000000),
|
|
/*
|
|
XMMF16UnpackLCPI3
|
|
*/
|
|
vec128q(0x7fe000007fe000ULL),
|
|
|
|
/* XMMF16PackLCPI0*/
|
|
vec128i(0x8000000),
|
|
/*XMMF16PackLCPI2*/
|
|
vec128i(0x47ffe000),
|
|
/*XMMF16PackLCPI3*/
|
|
vec128i(0xc7800000),
|
|
/*XMMF16PackLCPI4
|
|
*/
|
|
vec128i(0xf7fdfff),
|
|
/*XMMF16PackLCPI5*/
|
|
vec128i(0x7fff),
|
|
/*
|
|
XMMF16PackLCPI6
|
|
*/
|
|
vec128i(0x8000),
|
|
/* XMMXOPByteShiftMask,*/
|
|
vec128b(7),
|
|
/*XMMXOPWordShiftMask*/
|
|
vec128s(15),
|
|
/*XMMXOPDwordShiftMask*/
|
|
vec128i(31),
|
|
/*XMMLVLShuffle*/
|
|
v128_setr_bytes(3, 2, 1, 0, 7, 6, 5, 4, 11, 10, 9, 8, 15, 14, 13, 12),
|
|
/*XMMLVRCmp16*/
|
|
vec128b(16),
|
|
/*XMMSTVLShuffle*/
|
|
v128_setr_bytes(0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15),
|
|
/* XMMSTVRSwapMask*/
|
|
vec128b((uint8_t)0x83)};
|
|
|
|
void* X64Emitter::FindByteConstantOffset(unsigned bytevalue) {
|
|
for (auto& vec : xmm_consts) {
|
|
for (auto& u8 : vec.u8) {
|
|
if (u8 == bytevalue) {
|
|
return reinterpret_cast<void*>(backend_->emitter_data() +
|
|
(&u8 - &xmm_consts[0].u8[0]));
|
|
}
|
|
}
|
|
}
|
|
return nullptr;
|
|
}
|
|
void* X64Emitter::FindWordConstantOffset(unsigned wordvalue) {
|
|
for (auto& vec : xmm_consts) {
|
|
for (auto& u16 : vec.u16) {
|
|
if (u16 == wordvalue) {
|
|
return reinterpret_cast<void*>(backend_->emitter_data() +
|
|
((&u16 - &xmm_consts[0].u16[0]) * 2));
|
|
}
|
|
}
|
|
}
|
|
return nullptr;
|
|
}
|
|
void* X64Emitter::FindDwordConstantOffset(unsigned dwordvalue) {
|
|
for (auto& vec : xmm_consts) {
|
|
for (auto& u32 : vec.u32) {
|
|
if (u32 == dwordvalue) {
|
|
return reinterpret_cast<void*>(backend_->emitter_data() +
|
|
((&u32 - &xmm_consts[0].u32[0]) * 4));
|
|
}
|
|
}
|
|
}
|
|
return nullptr;
|
|
}
|
|
void* X64Emitter::FindQwordConstantOffset(uint64_t qwordvalue) {
|
|
for (auto& vec : xmm_consts) {
|
|
for (auto& u64 : vec.u64) {
|
|
if (u64 == qwordvalue) {
|
|
return reinterpret_cast<void*>(backend_->emitter_data() +
|
|
((&u64 - &xmm_consts[0].u64[0]) * 8));
|
|
}
|
|
}
|
|
}
|
|
return nullptr;
|
|
}
|
|
// First location to try and place constants.
|
|
static const uintptr_t kConstDataLocation = 0x20000000;
|
|
static const uintptr_t kConstDataSize = sizeof(xmm_consts);
|
|
|
|
// Increment the location by this amount for every allocation failure.
|
|
static const uintptr_t kConstDataIncrement = 0x00001000;
|
|
|
|
// This function places constant data that is used by the emitter later on.
|
|
// Only called once and used by multiple instances of the emitter.
|
|
//
|
|
// TODO(DrChat): This should be placed in the code cache with the code, but
|
|
// doing so requires RIP-relative addressing, which is difficult to support
|
|
// given the current setup.
|
|
uintptr_t X64Emitter::PlaceConstData() {
|
|
uint8_t* ptr = reinterpret_cast<uint8_t*>(kConstDataLocation);
|
|
void* mem = nullptr;
|
|
while (!mem) {
|
|
mem = memory::AllocFixed(
|
|
ptr, xe::round_up(kConstDataSize, memory::page_size()),
|
|
memory::AllocationType::kReserveCommit, memory::PageAccess::kReadWrite);
|
|
|
|
ptr += kConstDataIncrement;
|
|
}
|
|
|
|
// The pointer must not be greater than 31 bits.
|
|
assert_zero(reinterpret_cast<uintptr_t>(mem) & ~0x7FFFFFFF);
|
|
std::memcpy(mem, xmm_consts, sizeof(xmm_consts));
|
|
memory::Protect(mem, kConstDataSize, memory::PageAccess::kReadOnly, nullptr);
|
|
|
|
return reinterpret_cast<uintptr_t>(mem);
|
|
}
|
|
|
|
void X64Emitter::FreeConstData(uintptr_t data) {
|
|
memory::DeallocFixed(reinterpret_cast<void*>(data), 0,
|
|
memory::DeallocationType::kRelease);
|
|
}
|
|
|
|
Xbyak::Address X64Emitter::GetXmmConstPtr(XmmConst id) {
|
|
// Load through fixed constant table setup by PlaceConstData.
|
|
// It's important that the pointer is not signed, as it will be sign-extended.
|
|
return ptr[reinterpret_cast<void*>(backend_->emitter_data() +
|
|
sizeof(vec128_t) * id)];
|
|
}
|
|
// Implies possible StashXmm(0, ...)!
|
|
void X64Emitter::LoadConstantXmm(Xbyak::Xmm dest, const vec128_t& v) {
|
|
// https://www.agner.org/optimize/optimizing_assembly.pdf
|
|
// 13.4 Generating constants
|
|
if (!v.low && !v.high) {
|
|
// 0000...
|
|
vpxor(dest, dest);
|
|
} else if (v.low == ~uint64_t(0) && v.high == ~uint64_t(0)) {
|
|
// 1111...
|
|
vpcmpeqb(dest, dest);
|
|
} else {
|
|
for (size_t i = 0; i < (kConstDataSize / sizeof(vec128_t)); ++i) {
|
|
if (xmm_consts[i] == v) {
|
|
vmovapd(dest, GetXmmConstPtr((XmmConst)i));
|
|
return;
|
|
}
|
|
}
|
|
if (IsFeatureEnabled(kX64EmitAVX2)) {
|
|
bool all_equal_bytes = true;
|
|
|
|
unsigned firstbyte = v.u8[0];
|
|
for (unsigned i = 1; i < 16; ++i) {
|
|
if (v.u8[i] != firstbyte) {
|
|
all_equal_bytes = false;
|
|
break;
|
|
}
|
|
}
|
|
|
|
if (all_equal_bytes) {
|
|
void* bval = FindByteConstantOffset(firstbyte);
|
|
|
|
if (bval) {
|
|
vpbroadcastb(dest, byte[bval]);
|
|
return;
|
|
}
|
|
// didnt find existing mem with the value
|
|
mov(byte[rsp + kStashOffset], firstbyte);
|
|
vpbroadcastb(dest, byte[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
|
|
bool all_equal_words = true;
|
|
unsigned firstword = v.u16[0];
|
|
for (unsigned i = 1; i < 8; ++i) {
|
|
if (v.u16[i] != firstword) {
|
|
all_equal_words = false;
|
|
break;
|
|
}
|
|
}
|
|
if (all_equal_words) {
|
|
void* wval = FindWordConstantOffset(firstword);
|
|
if (wval) {
|
|
vpbroadcastw(dest, word[wval]);
|
|
return;
|
|
}
|
|
// didnt find existing mem with the value
|
|
mov(word[rsp + kStashOffset], firstword);
|
|
vpbroadcastw(dest, word[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
|
|
bool all_equal_dwords = true;
|
|
unsigned firstdword = v.u32[0];
|
|
for (unsigned i = 1; i < 4; ++i) {
|
|
if (v.u32[i] != firstdword) {
|
|
all_equal_dwords = false;
|
|
break;
|
|
}
|
|
}
|
|
if (all_equal_dwords) {
|
|
void* dwval = FindDwordConstantOffset(firstdword);
|
|
if (dwval) {
|
|
vpbroadcastd(dest, dword[dwval]);
|
|
return;
|
|
}
|
|
mov(dword[rsp + kStashOffset], firstdword);
|
|
vpbroadcastd(dest, dword[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
|
|
bool all_equal_qwords = v.low == v.high;
|
|
|
|
if (all_equal_qwords) {
|
|
void* qwval = FindQwordConstantOffset(v.low);
|
|
if (qwval) {
|
|
vpbroadcastq(dest, qword[qwval]);
|
|
return;
|
|
}
|
|
MovMem64(rsp + kStashOffset, v.low);
|
|
vpbroadcastq(dest, qword[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
}
|
|
|
|
for (auto& vec : xmm_consts) {
|
|
if (vec.low == v.low && vec.high == v.high) {
|
|
vmovdqa(dest,
|
|
ptr[reinterpret_cast<void*>(backend_->emitter_data() +
|
|
((&vec - &xmm_consts[0]) * 16))]);
|
|
return;
|
|
}
|
|
}
|
|
|
|
if (v.high == 0 && v.low == ~0ULL) {
|
|
vpcmpeqb(dest, dest);
|
|
movq(dest, dest);
|
|
return;
|
|
}
|
|
if (v.high == 0) {
|
|
if ((v.low & 0xFFFFFFFF) == v.low) {
|
|
mov(dword[rsp + kStashOffset], static_cast<unsigned>(v.low));
|
|
movd(dest, dword[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
MovMem64(rsp + kStashOffset, v.low);
|
|
movq(dest, qword[rsp + kStashOffset]);
|
|
return;
|
|
}
|
|
|
|
// TODO(benvanik): see what other common values are.
|
|
// TODO(benvanik): build constant table - 99% are reused.
|
|
MovMem64(rsp + kStashOffset, v.low);
|
|
MovMem64(rsp + kStashOffset + 8, v.high);
|
|
vmovdqa(dest, ptr[rsp + kStashOffset]);
|
|
}
|
|
}
|
|
|
|
void X64Emitter::LoadConstantXmm(Xbyak::Xmm dest, float v) {
|
|
union {
|
|
float f;
|
|
uint32_t i;
|
|
} x = {v};
|
|
if (!x.i) {
|
|
// +0.0f (but not -0.0f because it may be used to flip the sign via xor).
|
|
vxorps(dest, dest);
|
|
} else if (x.i == ~uint32_t(0)) {
|
|
// 1111...
|
|
vcmpeqss(dest, dest);
|
|
} else {
|
|
unsigned raw_bits = *reinterpret_cast<unsigned*>(&v);
|
|
|
|
for (size_t i = 0; i < (kConstDataSize / sizeof(vec128_t)); ++i) {
|
|
if (xmm_consts[i].u32[0] == raw_bits) {
|
|
vmovss(dest, GetXmmConstPtr((XmmConst)i));
|
|
return;
|
|
}
|
|
}
|
|
// TODO(benvanik): see what other common values are.
|
|
// TODO(benvanik): build constant table - 99% are reused.
|
|
mov(eax, x.i);
|
|
vmovd(dest, eax);
|
|
}
|
|
}
|
|
|
|
void X64Emitter::LoadConstantXmm(Xbyak::Xmm dest, double v) {
|
|
union {
|
|
double d;
|
|
uint64_t i;
|
|
} x = {v};
|
|
if (!x.i) {
|
|
// +0.0 (but not -0.0 because it may be used to flip the sign via xor).
|
|
vxorpd(dest, dest);
|
|
} else if (x.i == ~uint64_t(0)) {
|
|
// 1111...
|
|
vcmpeqpd(dest, dest);
|
|
} else {
|
|
uint64_t raw_bits = *reinterpret_cast<uint64_t*>(&v);
|
|
|
|
for (size_t i = 0; i < (kConstDataSize / sizeof(vec128_t)); ++i) {
|
|
if (xmm_consts[i].u64[0] == raw_bits) {
|
|
vmovsd(dest, GetXmmConstPtr((XmmConst)i));
|
|
return;
|
|
}
|
|
}
|
|
// TODO(benvanik): see what other common values are.
|
|
// TODO(benvanik): build constant table - 99% are reused.
|
|
mov(rax, x.i);
|
|
vmovq(dest, rax);
|
|
}
|
|
}
|
|
|
|
Xbyak::Address X64Emitter::StashXmm(int index, const Xbyak::Xmm& r) {
|
|
auto addr = ptr[rsp + kStashOffset + (index * 16)];
|
|
vmovups(addr, r);
|
|
return addr;
|
|
}
|
|
|
|
Xbyak::Address X64Emitter::StashConstantXmm(int index, float v) {
|
|
union {
|
|
float f;
|
|
uint32_t i;
|
|
} x = {v};
|
|
auto addr = rsp + kStashOffset + (index * 16);
|
|
MovMem64(addr, x.i);
|
|
MovMem64(addr + 8, 0);
|
|
return ptr[addr];
|
|
}
|
|
|
|
Xbyak::Address X64Emitter::StashConstantXmm(int index, double v) {
|
|
union {
|
|
double d;
|
|
uint64_t i;
|
|
} x = {v};
|
|
auto addr = rsp + kStashOffset + (index * 16);
|
|
MovMem64(addr, x.i);
|
|
MovMem64(addr + 8, 0);
|
|
return ptr[addr];
|
|
}
|
|
|
|
Xbyak::Address X64Emitter::StashConstantXmm(int index, const vec128_t& v) {
|
|
auto addr = rsp + kStashOffset + (index * 16);
|
|
MovMem64(addr, v.low);
|
|
MovMem64(addr + 8, v.high);
|
|
return ptr[addr];
|
|
}
|
|
static bool IsVectorCompare(const Instr* i) {
|
|
hir::Opcode op = i->opcode->num;
|
|
return op >= hir::OPCODE_VECTOR_COMPARE_EQ &&
|
|
op <= hir::OPCODE_VECTOR_COMPARE_UGE;
|
|
}
|
|
|
|
static bool IsFlaggedVectorOp(const Instr* i) {
|
|
if (IsVectorCompare(i)) {
|
|
return true;
|
|
}
|
|
hir::Opcode op = i->opcode->num;
|
|
using namespace hir;
|
|
switch (op) {
|
|
case OPCODE_VECTOR_SUB:
|
|
case OPCODE_VECTOR_ADD:
|
|
case OPCODE_SWIZZLE:
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
static SimdDomain GetDomainForFlaggedVectorOp(const hir::Instr* df) {
|
|
switch (df->flags) { // check what datatype we compared as
|
|
case hir::INT16_TYPE:
|
|
case hir::INT32_TYPE:
|
|
case hir::INT8_TYPE:
|
|
case hir::INT64_TYPE:
|
|
return SimdDomain::INTEGER;
|
|
case hir::FLOAT32_TYPE:
|
|
case hir::FLOAT64_TYPE: // pretty sure float64 doesnt occur with vectors.
|
|
// here for completeness
|
|
return SimdDomain::FLOATING;
|
|
default:
|
|
return SimdDomain::DONTCARE;
|
|
}
|
|
return SimdDomain::DONTCARE;
|
|
}
|
|
// this list is incomplete
|
|
static bool IsDefiniteIntegerDomainOpcode(hir::Opcode opc) {
|
|
using namespace hir;
|
|
switch (opc) {
|
|
case OPCODE_LOAD_VECTOR_SHL:
|
|
case OPCODE_LOAD_VECTOR_SHR:
|
|
case OPCODE_VECTOR_CONVERT_F2I:
|
|
case OPCODE_VECTOR_MIN: // there apparently is no FLOAT32_TYPE for min/maxs
|
|
// flags
|
|
case OPCODE_VECTOR_MAX:
|
|
case OPCODE_VECTOR_SHL:
|
|
case OPCODE_VECTOR_SHR:
|
|
case OPCODE_VECTOR_SHA:
|
|
case OPCODE_VECTOR_ROTATE_LEFT:
|
|
case OPCODE_VECTOR_AVERAGE: // apparently no float32 type for this
|
|
case OPCODE_EXTRACT:
|
|
case OPCODE_INSERT: // apparently no f32 type for these two
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
static bool IsDefiniteFloatingDomainOpcode(hir::Opcode opc) {
|
|
using namespace hir;
|
|
switch (opc) {
|
|
case OPCODE_VECTOR_CONVERT_I2F:
|
|
case OPCODE_VECTOR_DENORMFLUSH:
|
|
case OPCODE_DOT_PRODUCT_3:
|
|
case OPCODE_DOT_PRODUCT_4:
|
|
case OPCODE_LOG2:
|
|
case OPCODE_POW2:
|
|
case OPCODE_RECIP:
|
|
case OPCODE_ROUND:
|
|
case OPCODE_SQRT:
|
|
case OPCODE_MUL:
|
|
case OPCODE_MUL_SUB:
|
|
case OPCODE_MUL_ADD:
|
|
case OPCODE_ABS:
|
|
return true;
|
|
}
|
|
return false;
|
|
}
|
|
|
|
SimdDomain X64Emitter::DeduceSimdDomain(const hir::Value* for_value) {
|
|
hir::Instr* df = for_value->def;
|
|
if (!df) {
|
|
// todo: visit uses to figure out domain
|
|
return SimdDomain::DONTCARE;
|
|
|
|
} else {
|
|
SimdDomain result = SimdDomain::DONTCARE;
|
|
|
|
if (IsFlaggedVectorOp(df)) {
|
|
result = GetDomainForFlaggedVectorOp(df);
|
|
} else if (IsDefiniteIntegerDomainOpcode(df->opcode->num)) {
|
|
result = SimdDomain::INTEGER;
|
|
} else if (IsDefiniteFloatingDomainOpcode(df->opcode->num)) {
|
|
result = SimdDomain::FLOATING;
|
|
}
|
|
|
|
// todo: check if still dontcare, if so, visit uses of the value to figure
|
|
// it out
|
|
return result;
|
|
}
|
|
|
|
return SimdDomain::DONTCARE;
|
|
}
|
|
Xbyak::Address X64Emitter::GetBackendCtxPtr(int offset_in_x64backendctx) const {
|
|
/*
|
|
index context ptr negatively to get to backend ctx field
|
|
*/
|
|
ptrdiff_t delta = (-static_cast<ptrdiff_t>(sizeof(X64BackendContext))) +
|
|
offset_in_x64backendctx;
|
|
return ptr[GetContextReg() + static_cast<int>(delta)];
|
|
}
|
|
Xbyak::Label& X64Emitter::AddToTail(TailEmitCallback callback,
|
|
uint32_t alignment) {
|
|
TailEmitter emitter{};
|
|
emitter.func = std::move(callback);
|
|
emitter.alignment = alignment;
|
|
tail_code_.push_back(std::move(emitter));
|
|
return tail_code_.back().label;
|
|
}
|
|
Xbyak::Label& X64Emitter::NewCachedLabel() {
|
|
Xbyak::Label* tmp = new Xbyak::Label;
|
|
label_cache_.push_back(tmp);
|
|
return *tmp;
|
|
}
|
|
|
|
template <bool switching_to_fpu>
|
|
static void ChangeMxcsrModeDynamicHelper(X64Emitter& e) {
|
|
auto flags = e.GetBackendFlagsPtr();
|
|
if (switching_to_fpu) {
|
|
e.btr(flags, 0); // bit 0 set to 0 = is fpu mode
|
|
} else {
|
|
e.bts(flags, 0); // bit 0 set to 1 = is vmx mode
|
|
}
|
|
Xbyak::Label& come_back = e.NewCachedLabel();
|
|
|
|
Xbyak::Label& reload_bailout =
|
|
e.AddToTail([&come_back](X64Emitter& e, Xbyak::Label& thislabel) {
|
|
e.L(thislabel);
|
|
if (switching_to_fpu) {
|
|
e.LoadFpuMxcsrDirect();
|
|
} else {
|
|
e.LoadVmxMxcsrDirect();
|
|
}
|
|
e.jmp(come_back, X64Emitter::T_NEAR);
|
|
});
|
|
if (switching_to_fpu) {
|
|
e.jc(reload_bailout,
|
|
X64Emitter::T_NEAR); // if carry flag was set, we were VMX mxcsr mode.
|
|
} else {
|
|
e.jnc(
|
|
reload_bailout,
|
|
X64Emitter::T_NEAR); // if carry flag was set, we were VMX mxcsr mode.
|
|
}
|
|
e.L(come_back);
|
|
}
|
|
|
|
bool X64Emitter::ChangeMxcsrMode(MXCSRMode new_mode, bool already_set) {
|
|
if (cvars::enable_incorrect_roundingmode_behavior) {
|
|
return false; // no MXCSR mode handling!
|
|
}
|
|
if (new_mode == mxcsr_mode_) {
|
|
return false;
|
|
}
|
|
assert_true(new_mode != MXCSRMode::Unknown);
|
|
|
|
if (mxcsr_mode_ == MXCSRMode::Unknown) {
|
|
// check the mode dynamically
|
|
mxcsr_mode_ = new_mode;
|
|
if (!already_set) {
|
|
if (new_mode == MXCSRMode::Fpu) {
|
|
ChangeMxcsrModeDynamicHelper<true>(*this);
|
|
} else if (new_mode == MXCSRMode::Vmx) {
|
|
ChangeMxcsrModeDynamicHelper<false>(*this);
|
|
} else {
|
|
assert_unhandled_case(new_mode);
|
|
}
|
|
} else { // even if already set, we still need to update flags to reflect
|
|
// our mode
|
|
if (new_mode == MXCSRMode::Fpu) {
|
|
btr(GetBackendFlagsPtr(), 0);
|
|
} else if (new_mode == MXCSRMode::Vmx) {
|
|
bts(GetBackendFlagsPtr(), 0);
|
|
} else {
|
|
assert_unhandled_case(new_mode);
|
|
}
|
|
}
|
|
} else {
|
|
mxcsr_mode_ = new_mode;
|
|
if (!already_set) {
|
|
if (new_mode == MXCSRMode::Fpu) {
|
|
LoadFpuMxcsrDirect();
|
|
btr(GetBackendFlagsPtr(), 0);
|
|
return true;
|
|
} else if (new_mode == MXCSRMode::Vmx) {
|
|
LoadVmxMxcsrDirect();
|
|
bts(GetBackendFlagsPtr(), 0);
|
|
return true;
|
|
} else {
|
|
assert_unhandled_case(new_mode);
|
|
}
|
|
}
|
|
}
|
|
return false;
|
|
}
|
|
void X64Emitter::LoadFpuMxcsrDirect() {
|
|
vldmxcsr(GetBackendCtxPtr(offsetof(X64BackendContext, mxcsr_fpu)));
|
|
}
|
|
void X64Emitter::LoadVmxMxcsrDirect() {
|
|
vldmxcsr(GetBackendCtxPtr(offsetof(X64BackendContext, mxcsr_vmx)));
|
|
}
|
|
Xbyak::Address X64Emitter::GetBackendFlagsPtr() const {
|
|
Xbyak::Address pt = GetBackendCtxPtr(offsetof(X64BackendContext, flags));
|
|
pt.setBit(32);
|
|
return pt;
|
|
}
|
|
} // namespace x64
|
|
} // namespace backend
|
|
} // namespace cpu
|
|
} // namespace xe
|