This reverts commit e9c49901a43f5b16c3df416460b7e4dbdd24ce03.
Current AMDGPURegBankSelect does nothing different then RegBankSelect.
Revert to using generic RegBankSelect in preparation for adding new
regbankselect passes. New AMDGPURegBankSelect, that will use uniformity
analysis for regbank select decisions, will not subclass RegBankSelect.
Revert regression tests to use regbankselect since amdgpu-regbankselect
will be used by new pass and behavior will be different.
This patch adds YAML-based deserialization for MemProf profile.
It's been painful to write tests for MemProf passes because we do not
have a text format for the MemProf profile. We would write a test
case in C++, run it for a binary MemProf profile, and then finally run
a test written in LLVM IR with the binary profile.
This patch paves the way toward YAML-based MemProf profile.
Specifically, it adds new class YAMLMemProfReader derived from
MemProfReader. For now, it only adds a function to parse StringRef
pointing to YAML data. Subseqeunt patches will wire it to
llvm-profdata and read from a file.
The field names are based on various printYAML functions in MemProf.h.
I'm not aiming for compatibility with the format used in printYAML,
but I don't see a point in changing the field names.
This iteration works around the unavailability of
ScalarTraits<uintptr_t> on macOS.
The issue with slow compile-time was caused by an assert in
AArch64RegisterInfo.cpp. The assert invokes 'checkAllSuperRegsMarked'
after adding all the reserved registers. This call gets very expensive
after adding the _HI registers due to the way the function searches
in the 'Exception' list, which is expected to be a small list but isn't
(the patch added 190 _HI regs).
It was possible to rewrite the code in such a way that the _HI registers
are marked as reserved after the check. This makes the problem go away
entirely and restores compile-time to what it was before (tested for
`check-runtimes`, which previously showed a ~5x slowdown).
This reverts commits:
1434d2ab215e3ea9c5f34689d056edd3d4423a78
2704647fb7986673b89cef1def729e3b022e2607
This reverts commit c00e53208db638c35499fc80b555f8e14baa35f0.
It looks like this breaks building LLVM on macOS and some other
platform/compiler combos
https://lab.llvm.org/buildbot/#/builders/23/builds/5252https://green.lab.llvm.org/job/llvm.org/job/clang-san-iossim/5356/console
In file included from /Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/lib/ProfileData/MemProfReader.cpp:34:
In file included from /Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/ProfileData/MemProfReader.h:24:
In file included from /Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/ProfileData/InstrProfReader.h:22:
In file included from /Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/ProfileData/InstrProfCorrelator.h:21:
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/Support/YAMLTraits.h:1173:36: error: implicit instantiation of undefined template 'llvm::yaml::MissingTrait<unsigned long>'
char missing_yaml_trait_for_type[sizeof(MissingTrait<T>)];
^
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/Support/YAMLTraits.h:961:7: note: in instantiation of function template specialization 'llvm::yaml::yamlize<unsigned long>' requested here
yamlize(*this, Val, Required, Ctx);
^
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/Support/YAMLTraits.h:883:11: note: in instantiation of function template specialization 'llvm::yaml::IO::processKey<unsigned long, llvm::yaml::EmptyContext>' requested here
this->processKey(Key, Val, true, Ctx);
^
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/ProfileData/MIBEntryDef.inc:55:1: note: in instantiation of function template specialization 'llvm::yaml::IO::mapRequired<unsigned long>' requested here
MIBEntryDef(AccessHistogram = 27, AccessHistogram, uintptr_t)
^
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/lib/ProfileData/MemProfReader.cpp:77:8: note: expanded from macro 'MIBEntryDef'
Io.mapRequired(KeyStr.str().c_str(), MIB.Name); \
^
/Users/ec2-user/jenkins/workspace/llvm.org/clang-san-iossim/llvm-project/llvm/include/llvm/Support/YAMLTraits.h:310:8: note: template is declared here
struct MissingTrait;
^
1 error generated.
This patch adds YAML-based deserialization for MemProf profile.
It's been painful to write tests for MemProf passes because we do not
have a text format for the MemProf profile. We would write a test
case in C++, run it for a binary MemProf profile, and then finally run
a test written in LLVM IR with the binary profile.
This patch paves the way toward YAML-based MemProf profile.
Specifically, it adds new class YAMLMemProfReader derived from
MemProfReader. For now, it only adds a function to parse StringRef
pointing to YAML data. Subseqeunt patches will wire it to
llvm-profdata and read from a file.
The field names are based on various printYAML functions in MemProf.h.
I'm not aiming for compatibility with the format used in printYAML,
but I don't see a point in changing the field names.
This patch adds NVVM intrinsics and NVPTX codegen for:
* cp.async.bulk.tensor.reduce.1D -> 5D variants, supporting both Tile
and Im2Col modes.
* These intrinsics optionally support cache_hints as indicated by the
boolean flag argument.
* Lit tests are added for all combinations of these intrinsics in
cp-async-bulk-tensor-reduce.ll.
* The generated PTX is verified with a 12.3 ptxas executable.
* Added docs for these intrinsics in NVPTXUsage.rst file.
PTX Spec reference:
https://docs.nvidia.com/cuda/parallel-thread-execution/#data-movement-and-conversion-instructions-cp-reduce-async-bulk-tensor
Signed-off-by: Durgadoss R <durgadossr@nvidia.com>
These instructions have non-standard use of OPSEL bits to select
dest write byte. The src2_modifiers operand is used without having
its corresponding src2 operand by introducing dummy src2.
OPSEL ASM OPSEL Syntax: opsel:[a,b,c,d]
a & b are meaningless, c & d together decides byte to write in dst reg.
Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
OPSEL ASM Syntax for v_cvt_scalef32_pk_{f|bf}16_fp4 : opsel:[x,y,z]
where, x & y i.e. OPSEL[1 : 0] selects which src_byte to read.
Note: Conventional Inst{13} i.e. OPSEL[2] is ignored in asm syntax.
Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
OPSEL ASM Syntax for v_cvt_scalef32_pk_f32_fp4 : opsel:[x,y,z]
where, x & y i.e. OPSEL[1 : 0] selects which src_byte to read.
OPSEL ASM Syntax for v_cvt_scalef32_pk_fp4_f32 : opsel:[a,b,c,d]
where, c & d i.e. OPSEL[3 : 2] selects which dst_byte to write.
Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
This PR adds support in the gSYM format for call site information and
adds support for loading call sites from a YAML file. The support for
YAML input is mostly for testing purposes - so we have a way to test the
functionality.
Note that this data is not currently used in the gSYM tooling - the
logic to use call sites will be added in a later PR.
The reason why we need call site information in gSYM files is so that we
can support better call stack function disambiguation in the case where
multiple functions have been merged due to optimization (linker ICF).
When resolving a merged function on the callstack, we can use the call
site information of the calling function to narrow down the actual
function that is being called, from the set of all merged functions.
See [this
RFC](https://discourse.llvm.org/t/rfc-extending-gsym-format-with-call-site-information-for-merged-function-disambiguation/80682)
for more details on this change.
This patch adds a default constructor to Frame along with a use in
FrameIdConverter.
The real intent is to facilitate deserialization of YAML-based MemProf
format. Note that the YAML parser default-constructs a struct and
then populates one field at a time.
IndexedMemProfRecord contains a complete package of the MemProf
profile, including frames, call stacks, and records. This patch
replaces the three member variables of MemProfReader with
IndexedMemProfRecord.
This transition significantly simplies both the constructor and the
final "take" method:
MemProfReader(IndexedMemProfData MemProfData)
: MemProfData(std::move(MemProfData)) {}
IndexedMemProfData takeMemProfData() { return std::move(MemProfData); }
OPSEL[1:0] collectively decide which byte to read
from src input.
Builtin takes additional imm argument which
represents index (with valid values:[0:3]) of src
byte read. Out of bounds checks will added in next
patch.
OPSEL ASM Syntax: opsel:[x,y,z]
where,
opsel[x] = Inst{11} = src0_modifier{2}
opsel[y] = Inst{12} = src1_modifier{2}
opsel[z] = Inst{14} = src0_modifier{3}
Note: Inst{13} i.e. OPSEL[2] is ignored in
asm syntax and opsel[z] is meaningless
for v_cvt_scalef32_f32_{fp|bf}8
Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
This should act like range.
Previously ConstantRangeList assumed a 64-bit range. Now query from the
actual entries. This also means that the empty range has no bitwidth, so
move asserts to avoid checking the bitwidth of empty ranges.
Two options for clang
-mdiv32: Use div.w[u] and mod.w[u] instructions with input not
sign-extended.
-mno-div32: Do not use div.w[u] and mod.w[u] instructions with input not
sign-extended.
The default is -mno-div32.
Add FIR generation and LLVMIR translation support for mergeable clause
on task construct. If mergeable clause is present on a task, the
relevant flag in `ompt_task_flag_t` is set and passed to
`__kmpc_omp_task_alloc`.
SLEEF math vector library now supports RISC-V target.
Commit: https://github.com/shibatch/sleef/pull/477
This patch enables the use of auto-vectorization with
subsequent replacement by the corresponding SLEEF function.
The encoding of v_dot2c_f32_bf16 opcode is same as v_mac_f32 in gfx90a,
both from gfx9 series. This required a new decoderNameSpace GFX950_DOT.
Co-authored-by: Sirish Pande <Sirish.Pande@amd.com>
Scale packed 16-component single-precision float vectors from
two source inputs using the exponent provided by the third
single-precision float input, then convert the values to a packed
32-component FP6 float value.
Co-authored-by: Pravin Jagtap <Pravin.Jagtap@amd.com>
This looks like a rather weird change, so let me explain why this isn't
as unreasonable as it looks. Let's start with the problem it's solving.
```
define signext i32 @overlap_live_ranges(ptr %arg, i32 signext %arg1) { bb:
%i = icmp eq i32 %arg1, 1
br i1 %i, label %bb2, label %bb5
bb2: ; preds = %bb
%i3 = getelementptr inbounds nuw i8, ptr %arg, i64 4
%i4 = load i32, ptr %i3, align 4
br label %bb5
bb5: ; preds = %bb2, %bb
%i6 = phi i32 [ %i4, %bb2 ], [ 13, %bb ]
ret i32 %i6
}
```
Right now, we codegen this as:
```
li a3, 1
li a2, 13
bne a1, a3, .LBB0_2
lw a2, 4(a0)
.LBB0_2:
mv a0, a2
ret
```
In this example, we have two values which must be assigned to a0 per the
ABI (%arg, and the return value). SelectionDAG ensures that all values
used in a successor phi are defined before exit the predecessor block.
This creates an ADDI to materialize the immediate in the entry block.
Currently, this ADDI is not sunk into the tail block because we'd have
to split a critical edges to do so. Note that if our immediate was
anything large enough to require two instructions we *would* split this
critical edge.
Looking at other targets, we notice that they don't seem to have this
problem. They perform the sinking, and tail duplication that we don't.
Why? Well, it turns out for AArch64 that this is entirely an accident of
the existance of the gpr32all register class. The immediate is
materialized into the gpr32 class, and then copied into the gpr32all
register class. The existance of that copy puts us right back into the
two instruction case noted above.
This change essentially just bypasses this emergent behavior aspect of
the aarch64 behavior, and implements the same "always sink immediates"
behavior for RISCV as well.
Introduces `__builtin_hlsl_buffer_update_counter` clang buildin that is
used to implement the `IncrementCounter` and `DecrementCounter` methods
on `RWStructuredBuffer` and `RasterizerOrderedStructuredBuffer` (see
Note).
The builtin is translated to LLVM intrisic `llvm.dx.bufferUpdateCounter`
or `llvm.spv.bufferUpdateCounter`.
Introduces `BuiltinTypeMethodBuilder` helper in `HLSLExternalSemaSource`
that enables adding methods to builtin types using builder pattern like
this:
```
BuiltinTypeMethodBuilder(Sema, RecordBuilder, "MethodName", ReturnType)
.addParam("param_name", Type, InOutModifier)
.callBuiltin("buildin_name", { BuiltinParams })
.finalizeMethod();
```
Fixes#113513
[First version](llvm/llvm-project#114148) of this PR was reverted
because of build break.
For
add %reg1, name@GOTTPOFF(%rip), %reg2
add name@GOTTPOFF(%rip), %reg1, %reg2
{nf} add %reg1, name@GOTTPOFF(%rip), %reg2
{nf} add name@GOTTPOFF(%rip), %reg1, %reg2
{nf} add name@GOTTPOFF(%rip), %reg
add
`R_X86_64_CODE_6_GOTTPOFF` = 50
if the instruction starts at 6 bytes before the relocation offset. It's
similar to R_X86_64_GOTTPOFF.
Linker can treat `R_X86_64_CODE_6_GOTTPOFF` as `R_X86_64_GOTTPOFF` or
convert the instructions above to
add $name@tpoff, %reg1, %reg2
add $name@tpoff, %reg1, %reg2
{nf} add $name@tpoff, %reg1, %reg2
{nf} add $name@tpoff, %reg1, %reg2
{nf} add $name@tpoff, %reg
if the first byte of the instruction at the relocation `offset - 6` is
`0xd5` (namely, encoded w/REX2 prefix) when possible.
Binutils patch:
5bc71c2a6b
Binutils mailthread:
https://sourceware.org/pipermail/binutils/2024-February/132351.html
ABI discussion:
https://groups.google.com/g/x86-64-abi/c/FhEZjCtDLFw/m/VHDjN4orAgAJ
Blog: https://kanrobert.github.io/rfc/All-about-APX-relocation
This PR adds support seq_cst (sequential consistency) clause for the
flush directive in OpenMP. The seq_cst clause enforces a stricter memory
ordering, ensuring that all threads observe the memory effects of the
flush in the same order, improving consistency in memory operations
across threads.
---------
Co-authored-by: Shashwathi N <nshashwa@pe28vega.hpc.amslabs.hpecorp.net>
Co-authored-by: CHANDRA GHALE <chandra.nitdgp@gmail.com>
CallStackRadixTreeBuilder::build takes the parameter
MemProfFrameIndexes by value, involving copies:
std::optional<const llvm::DenseMap<FrameIdTy, LinearFrameId>>
MemProfFrameIndexes
Then "build" makes another copy of MemProfFrameIndexe and passes it to
encodeCallStack for every call stack, which is painfully slow.
This patch changes the type to a pointer so that we don't have to make
a copy every time we pass the argument.
Without this patch, it takes 553 seconds to run "llvm-profdata merge"
on a large MemProf raw profile. This patch shortenes that down to 67
seconds.
This patch moves the `areInlineCompatible` implementation from multiple
subclasses (`AArch64TTIImpl`, `RISCVTTIImpl`, `WebAssemblyTTIImpl`) to
the base class `BasicTTIImpl`. The new implementation checks whether the
callee's target features are a subset of the caller's, enabling
consistent behavior across targets. Subclasses now simply delegate to
the base implementation, reducing code duplication and improving
maintainability.
This patch removes a dead constructor in AllocationInfo. We used to
use it for MemProf Version 1 deserialization purposes via
MemProfRecord::MemProfRecord.
