Skip to content

[RISC-V] On Stack Replacement #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
sirntar merged 6 commits into from
Dec 18, 2023
Merged

[RISC-V] On Stack Replacement #2

Show file tree
Hide file tree
Changes from 1 commit
Commits
Show all changes
6 commits
Select commit Hold shift + click to select a range
71844c6
[RISC-V] Implement On Stack Replacement
sirntar Dec 7, 2023
377e6f4
[RISC-V] Sketch implementation of genAllocLclFrame
sirntar Dec 11, 2023
9bd763b
[RISC-V] Implement memory page stack probing
sirntar Dec 11, 2023
ea9555d
[RISC-V] OSR cleanup
sirntar Dec 12, 2023
7e999bf
[RISC-V] Improve genAllocLclFrame implementation
sirntar Dec 13, 2023
07feeaf
[RISC-V] Improve support for non standard memory pages and code clarity
sirntar Dec 15, 2023
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Next Next commit
[RISC-V] Implement On Stack Replacement
  • Loading branch information
@sirntar
sirntar authored Dec 7, 2023
commit 71844c65e9da81e11bf3e3a0f531e74660d3c5a1
4 changes: 2 additions & 2 deletions src/coreclr/clrdefinitions.cmake
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -184,9 +184,9 @@ endif(FEATURE_ENABLE_NO_ADDRESS_SPACE_RANDOMIZATION)
add_definitions(-DFEATURE_SVR_GC)
add_definitions(-DFEATURE_SYMDIFF)
add_compile_definitions(FEATURE_TIERED_COMPILATION)
if (CLR_CMAKE_TARGET_ARCH_AMD64 OR CLR_CMAKE_TARGET_ARCH_ARM64 OR CLR_CMAKE_TARGET_ARCH_LOONGARCH64)
if (CLR_CMAKE_TARGET_ARCH_AMD64 OR CLR_CMAKE_TARGET_ARCH_ARM64 OR CLR_CMAKE_TARGET_ARCH_LOONGARCH64 OR CLR_CMAKE_TARGET_ARCH_RISCV64)
add_compile_definitions(FEATURE_ON_STACK_REPLACEMENT)
endif (CLR_CMAKE_TARGET_ARCH_AMD64 OR CLR_CMAKE_TARGET_ARCH_ARM64 OR CLR_CMAKE_TARGET_ARCH_LOONGARCH64)
endif (CLR_CMAKE_TARGET_ARCH_AMD64 OR CLR_CMAKE_TARGET_ARCH_ARM64 OR CLR_CMAKE_TARGET_ARCH_LOONGARCH64 OR CLR_CMAKE_TARGET_ARCH_RISCV64)
add_compile_definitions(FEATURE_PGO)
if (CLR_CMAKE_TARGET_WIN32)
add_definitions(-DFEATURE_TYPEEQUIVALENCE)
Expand Down
6 changes: 3 additions & 3 deletions src/coreclr/inc/clrconfigvalues.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -570,11 +570,11 @@ RETAIL_CONFIG_DWORD_INFO(INTERNAL_HillClimbing_GainExponent,
#endif // _DEBUG
RETAIL_CONFIG_DWORD_INFO(EXTERNAL_TieredCompilation, W("TieredCompilation"), 1, "Enables tiered compilation")
RETAIL_CONFIG_DWORD_INFO(EXTERNAL_TC_QuickJit, W("TC_QuickJit"), 1, "For methods that would be jitted, enable using quick JIT when appropriate.")
#if defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#if defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
RETAIL_CONFIG_DWORD_INFO(UNSUPPORTED_TC_QuickJitForLoops, W("TC_QuickJitForLoops"), 1, "When quick JIT is enabled, quick JIT may also be used for methods that contain loops.")
#else // !(defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64))
#else // !(defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)) || defined(TARGET_RISCV64)
RETAIL_CONFIG_DWORD_INFO(UNSUPPORTED_TC_QuickJitForLoops, W("TC_QuickJitForLoops"), 0, "When quick JIT is enabled, quick JIT may also be used for methods that contain loops.")
#endif // defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#endif // defined(TARGET_AMD64) || defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
RETAIL_CONFIG_DWORD_INFO(EXTERNAL_TC_AggressiveTiering, W("TC_AggressiveTiering"), 0, "Transition through tiers aggressively.")
RETAIL_CONFIG_DWORD_INFO(EXTERNAL_TC_CallCountThreshold, W("TC_CallCountThreshold"), TC_CallCountThreshold, "Number of times a method must be called in tier 0 after which it is promoted to the next tier.")
RETAIL_CONFIG_DWORD_INFO(EXTERNAL_TC_CallCountingDelayMs, W("TC_CallCountingDelayMs"), TC_CallCountingDelayMs, "A perpetual delay in milliseconds that is applied to call counting in tier 0 and jitting at higher tiers, while there is startup-like activity.")
Expand Down
8 changes: 4 additions & 4 deletions src/coreclr/jit/codegen.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -262,7 +262,7 @@ class CodeGen final : public CodeGenInterface
void genFnPrologCalleeRegArgs(regNumber xtraReg, bool* pXtraRegClobbered, RegState* regState);
#endif
void genEnregisterIncomingStackArgs();
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
void genEnregisterOSRArgsAndLocals(regNumber initReg, bool* pInitRegZeroed);
#else
void genEnregisterOSRArgsAndLocals();
Expand Down Expand Up @@ -447,11 +447,11 @@ class CodeGen final : public CodeGenInterface
regMaskTP fiSaveRegs; // Set of callee-saved registers saved in the funclet prolog (includes RA)
int fiFunction_CallerSP_to_FP_delta; // Delta between caller SP and the frame pointer in the parent function
// (negative)
int fiSP_to_FPRA_save_delta; // FP/RA register save offset from SP (positive)
int fiSP_to_CalleeSaved_delta; // CalleeSaved register save offset from SP (positive)
int fiCalleeSavedPadding; // CalleeSaved offset padding (positive)
int fiSP_to_PSP_slot_delta; // PSP slot offset from SP (positive)
int fiCallerSP_to_PSP_slot_delta; // PSP slot offset from Caller SP (negative)
int fiFrameType; // Funclet frame types are numbered. See genFuncletProlog() for details.
int fiSpDelta1; // Stack pointer delta 1 (negative)
int fiSpDelta; // Stack pointer delta (negative)
};

FuncletFrameInfoDsc genFuncletInfo;
Expand Down
30 changes: 21 additions & 9 deletions src/coreclr/jit/codegencommon.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4730,7 +4730,7 @@ void CodeGen::genZeroInitFrame(int untrLclHi, int untrLclLo, regNumber initReg,
// initReg -- scratch register to use if needed
// pInitRegZeroed -- [IN,OUT] if init reg is zero (on entry/exit)
//
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
void CodeGen::genEnregisterOSRArgsAndLocals(regNumber initReg, bool* pInitRegZeroed)
#else
void CodeGen::genEnregisterOSRArgsAndLocals()
Expand Down Expand Up @@ -4871,7 +4871,7 @@ void CodeGen::genEnregisterOSRArgsAndLocals()

GetEmitter()->emitIns_R_AR(ins_Load(lclTyp), size, varDsc->GetRegNum(), genFramePointerReg(), offset);

#elif defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#elif defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)

// Patchpoint offset is from top of Tier0 frame
//
Expand Down Expand Up @@ -4903,7 +4903,7 @@ void CodeGen::genEnregisterOSRArgsAndLocals()

genInstrWithConstant(ins_Load(lclTyp), size, varDsc->GetRegNum(), genFramePointerReg(), offset, initReg);
*pInitRegZeroed = false;
#endif
#endif // TARGET_ARM64 || TARGET_LOONGARCH64 || TARGET_RISCV64
}
}

Expand Down Expand Up @@ -5510,7 +5510,7 @@ void CodeGen::genFnProlog()
psiBegProlog();
}

#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
// For arm64 OSR, emit a "phantom prolog" to account for the actions taken
// in the tier0 frame that impact FP and SP on entry to the OSR method.
//
Expand All @@ -5525,7 +5525,7 @@ void CodeGen::genFnProlog()
// SP is tier0 method's SP.
compiler->unwindAllocStack(tier0FrameSize);
}
#endif // defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#endif // defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)

#ifdef DEBUG

Expand Down Expand Up @@ -5855,13 +5855,25 @@ void CodeGen::genFnProlog()
{
initReg = REG_SCRATCH;
}
#elif defined(TARGET_RISCV64)
// For RISC-V64 OSR root frames, we may need a scratch register for large
// offset addresses. Use a register that won't be allocated.
if (isRoot && compiler->opts.IsOSR())
{
initReg = REG_SCRATCH; // REG_T0
}
#endif

#ifndef TARGET_LOONGARCH64
#if !defined(TARGET_LOONGARCH64) && !defined(TARGET_RISCV64)
// For LoongArch64's OSR root frames, we may need a scratch register for large
// offset addresses. But this does not conflict with the REG_PINVOKE_FRAME.
//
// RISC-V64's OSR root frames are similar to LoongArch64's. In this case
// REG_SCRATCH also shouldn't conflict with REG_PINVOKE_FRAME, even if
// technically they are the same register - REG_T0.
//
noway_assert(!compiler->compMethodRequiresPInvokeFrame() || (initReg != REG_PINVOKE_FRAME));
#endif
#endif // !TARGET_LOONGARCH64 && !TARGET_RISCV64

#if defined(TARGET_AMD64)
// If we are a varargs call, in order to set up the arguments correctly this
Expand Down Expand Up @@ -6172,7 +6184,7 @@ void CodeGen::genFnProlog()
// Otherwise we'll do some of these fetches twice.
//
CLANG_FORMAT_COMMENT_ANCHOR;
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64)
#if defined(TARGET_ARM64) || defined(TARGET_LOONGARCH64) || defined(TARGET_RISCV64)
genEnregisterOSRArgsAndLocals(initReg, &initRegZeroed);
#else
genEnregisterOSRArgsAndLocals();
Expand Down Expand Up @@ -6224,7 +6236,7 @@ void CodeGen::genFnProlog()
assignIncomingRegisterArgs(&floatRegState);
#else
assignIncomingRegisterArgs(&intRegState);
#endif
#endif // TARGET_ARM64 || TARGET_LOONGARCH64 || TARGET_RISCV64

#endif // TARGET_LOONGARCH64 || TARGET_RISCV64

Expand Down
Loading