Before D94153 this threshold was in a pre-scaled units.
After D94153 inlining threshold multiplier is not applied
to this portion of the threshold anymore. Restore the
threshold by applying the multiplier.
Differential Revision: https://reviews.llvm.org/D98362
Add printf-style alternate form flag to prefix hex number with 0x when
present. This works on both empty numeric expression (e.g. variable
definition from input) and when matching a numeric expression. The
syntax is as follows:
[[#%#<precision specifier><format specifier>, ...]
where <precision specifier> and <format specifier> are optional and ...
can be a variable definition or not with an empty expression or not.
This feature was requested in https://reviews.llvm.org/D81144#2075532
for llvm/test/MC/ELF/gen-dwarf64.s
Reviewed By: jdenny
Differential Revision: https://reviews.llvm.org/D97845
- This patch adds in support for the ordinary HLASM comment syntax asm
statements (Reference - Chapter 7, Comment Statements, Ordinary Comment
Statements)
- In brief, the ordinary comment syntax if used, must begin with the "*"
character
- To achieve this, this patch makes use of the CommentString attribute
provided in the base MCAsmInfo class
- In the SystemZMCAsmInfo class, the CommentString attribute was set to
"*" based on the assembler dialect
- Furthermore, a new attribute RestrictCommentString, is provided to only
treat a string as a comment if it appears at the start of the asm
statement. Example: "jo *-4" is valid in HLASM (jump back 4 bytes from
current point - similar to jo -4 in gnu asm) and we don't want "*-4" to
be treated as a comment.
- RFC for HLASM Parser support implementation: https://lists.llvm.org/pipermail/llvm-dev/2021-January/147686.html
Reviewed By: scott.linder, Kai
Differential Revision: https://reviews.llvm.org/D97703
This also briefly tests a larger set of architectures than the more
exhaustive functionality tests for AArch64 and x86.
As requested in D88785
Reviewed By: RKSimon
Differential Revision: https://reviews.llvm.org/D98339
byval arguments need to be assumed writable. Only implicitly stack
passed arguments which aren't addressable in the IR can be assumed
immutable.
Mips is still broken since for some reason its doing its own thing
with the ValueHandlers (and x86 doesn't actually handle byval
arguments now, although some of the code is there).
This was essentially ignoring byval and treating them as a pointer
argument which needed to be loaded from. This should copy the frame
index value to the virtual register, not insert a load from the frame
index into the pointer value.
For AMDGPU, this was producing a load from the byval pointer argument,
to a pointer used for the byval arguments. I do not understand how
AArch64 managed to work before since it appears to be similarly
broken.
We could also change the ValueHandler API to avoid the extra copy from
the frame index, since currently it returns a new register.
I believe there is still an issue with outgoing byval arguments. These
should have a copy inserted in case the callee decided to overwrite
the memory.
This patch fixes a crash when trying to get a scalar value using
VPTransformState::get() for uniform induction values or truncated
induction values. IVs and truncated IVs can be uniform and the updated
code accounts for that, fixing the crash.
This should fix
https://bugs.chromium.org/p/oss-fuzz/issues/detail?id=31981
The test is reduced from a C source example in:
https://llvm.org/PR49541
It's possible that the test could be reduced further or
the predicate generalized further, but it seems to require
a few ingredients (including the "late" SimplifyCFG options
on the RUN line) to fall into the infinite-loop trap.
During finalization the debug object is registered with the target. Materialization must wait for this process to finish. Otherwise we might start running code before the debugger finished processing the corresponding debug info.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D98417
This broke the check-profile tests on Mac, see comment on the code
review.
> This is no longer needed, we can add __llvm_profile_runtime directly
> to llvm.compiler.used or llvm.used to achieve the same effect.
>
> Differential Revision: https://reviews.llvm.org/D98325
This reverts commit c7712087cbb505d324e1149fa224f607c91a8c6a.
Also reverting the dependent follow-up commit:
Revert "[InstrProfiling] Generate runtime hook for ELF platforms"
> When using -fprofile-list to selectively apply instrumentation only
> to certain files or functions, we may end up with a binary that doesn't
> have any counters in the case where no files were selected. However,
> because on Linux and Fuchsia, we pass -u__llvm_profile_runtime, the
> runtime would still be pulled in and incur some non-trivial overhead,
> especially in the case when the continuous or runtime counter relocation
> mode is being used. A better way would be to pull in the profile runtime
> only when needed by declaring the __llvm_profile_runtime symbol in the
> translation unit only when needed.
>
> This approach was already used prior to 9a041a75221ca, but we changed it
> to always generate the __llvm_profile_runtime due to a TAPI limitation.
> Since TAPI is only used on Mach-O platforms, we could use the early
> emission of __llvm_profile_runtime there, and on other platforms we
> could change back to the earlier approach where the symbol is generated
> later only when needed. We can stop passing -u__llvm_profile_runtime to
> the linker on Linux and Fuchsia since the generated undefined symbol in
> each translation unit that needed it serves the same purpose.
>
> Differential Revision: https://reviews.llvm.org/D98061
This reverts commit 87fd09b25f8892e07b7ba11525baa9c3ec3e5d3f.
As readnone function they become movable and LICM can hoist them
out of a loop. As a result in LCSSA form phi node of type token
is created. No one is ready that GCRelocate first operand is phi node
but expects to be token.
GVN test were also updated, it seems it does not do what is expected.
Test for LICM is also added.
This reverts commit f352463ade6e49c3b0275f296d9190d828b7630b.
This patch optimizes the codegen for INSERT_VECTOR_ELT in various ways.
Primarily, it removes the use of vslidedown during lowering, and the
vector element is inserted entirely using vslideup with a custom VL and
slide index.
Additionally, lowering of i64-element vectors on RV32 has been optimized
in several ways. When the 64-bit value to insert is the same as the
sign-extension of the lower 32-bits, the codegen can follow the regular
path. When this is not possible, a new sequence of two i32 vslide1up
instructions is used to get the vector element into a vector. This
sequence was suggested by @craig.topper. From there, the value is slid
into the final position for more consistent lowering across RV32 and
RV64.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D98250
Add simplification of smul.fix and smul.fix.sat according to
X * 0 -> 0
X * undef -> 0
X * (1 << scale) -> X
This includes the commuted patterns and splatted vectors.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D98299
Do constant folding according to
posion * C -> poison
C * poison -> poison
undef * C -> 0
C * undef -> 0
for smul_fix and smul_fix_sat intrinsics (for any scale).
Reviewed By: nikic, aqjune, nagisa
Differential Revision: https://reviews.llvm.org/D98410
Since D86233 we have `mustprogress` which, in combination with
`readonly`, implies `willreturn`. The idea is that every side-effect
has to be modeled as a "write". Consequently, `readonly` means there
is no side-effect, and `mustprogress` guarantees that we cannot "loop"
forever without side-effect.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D94125
The update_test_checks script can now check for global symbols and is able
to handle them properly when they differ across prefixes, e.g.,
attribute #0 might be different in different runs.
This patch simply updates all the Attributor tests with the new script.
Reviewed By: sstefan1
Differential Revision: https://reviews.llvm.org/D97906
Nested `omp [begin|end] declare variant` inherit the selectors from
surrounding `omp (begin|end) declare variant` constructs. To stop such
propagation the user can add the `disable_selector_propagation` to the
`extension` set in the `implementation` selector.
Reviewed By: tianshilei1992
Differential Revision: https://reviews.llvm.org/D95765
If we have nested declare variant context, it doesn't make sense to
inherit the match extension from the parent. Instead, just skip it.
Reviewed By: JonChesterfield
Differential Revision: https://reviews.llvm.org/D95764
This allows to check for various globals (metadata/attributes/...) and
also resolves problems with globals (metadata/attributes/...) being
reused across different prefixes.
Reviewed By: sstefan1
Differential Revision: https://reviews.llvm.org/D94741
Summary: This is a minor patch to add names for the debug line prologue, as a follow-up of D95998.
Reviewed By: dblaikie, ikudrin, shchenz
Differential Revision: https://reviews.llvm.org/D98383
We don't support any other shuffles currently.
This changes the bswap/bitreverse tests that check for this in
their expansion code. Previously we expanded a byte swapping
shuffle through memory. Now we're scalarizing and doing bit
operations on scalars to swap bytes.
In the future we can probably use vrgather.vx to do a byte swap
shuffle.
Currently, BPF backend does not support all variants of
atomic_load_{add,and,or,xor}, atomic_swap and atomic_cmp_swap
For example, it only supports 32bit (with alu32 mode) and 64bit
operations for atomic_load_{and,or,xor}, atomic_swap and
atomic_cmp_swap. Due to historical reason, atomic_load_add is
always supported with 32bit and 64bit.
If user used an unsupported atomic operation, currently,
codegen selectiondag cannot find bpf support and will issue
a fatal error. This is not user friendly as user may mistakenly
think this is a compiler bug.
This patch added Custom rule for unsupported atomic operations
and will emit better error message during ReplaceNodeResults()
callback. The following is an example output.
$ cat t.c
short sync(short *p) {
return __sync_val_compare_and_swap (p, 2, 3);
}
$ clang -target bpf -O2 -g -c t.c
t.c:2:11: error: Unsupported atomic operations, please use 64 bit version
return __sync_val_compare_and_swap (p, 2, 3);
^
fatal error: error in backend: Cannot select: t19: i64,ch =
AtomicCmpSwap<(load store seq_cst seq_cst 2 on %ir.p)> t0, t2,
Constant:i64<2>, Constant:i64<3>, t.c:2:11
t2: i64,ch = CopyFromReg t0, Register:i64 %0
t1: i64 = Register %0
t11: i64 = Constant<2>
t10: i64 = Constant<3>
In function: sync
PLEASE submit a bug report ...
Fatal error will still happen since we did not really do proper
lowering for these unsupported atomic operations. But we do get
a much better error message.
Differential Revision: https://reviews.llvm.org/D98471
As llvm.amdgcn.kill is lowered to a terminator it can cause
else branch annotations to end up in the wrong block.
Do not annotate conditionals as else branches where there is
a kill to avoid this.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D97427
The "Inputs" subdirectory is used for all files read by the test, not
only those used as input to the execution - so even though this file is
used as a golden reference for the output of the test, it's still an
input to the test execution (it is read in the process of executing the
test).
From what I can tell, the loop inside applyExternalSymbolRelocations()
used to call getSymbolAddress(). After the JITSymbolResolver interface
redesign, the functionality has changed, and the loop should no longer
trigger repopulation of ExternalSymbolRelocations. If that's the case,
there is no need to update the loop iterator manually, and
ExternalSymbolRelocations can be cleared at once. This way, when there
are many external symbols in the program, the function runs much faster.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D97531
This instruction is only valid on 2D MSAA and 2D MSAA Array
surfaces. Remove intrinsic support for other dimension types,
and block assembly for unsupported dimensions.
Reviewed By: foad
Differential Revision: https://reviews.llvm.org/D98397
We have amdgpu_gfx functions that have high register pressure. If
we do not reserve VGPR for SGPR spill, we will fall into the path
to spill the SGPR to memory, which does not only have correctness issue,
but also have really bad performance.
I don't know why there is the check for hasStackObjects(), in our case,
we don't have stack objects at the time of finalizeLowering(). So just
remove the check that we always reserve a VGPR for possible SGPR spill
in non-entry functions.
Reviewed by: arsenm
Differential Revision: https://reviews.llvm.org/D98345
I met some code generation behavior change when I tried to remove
the hasStackObject() check when reserving VGPR for SGPR spill.
For example, the function `callee_no_stack_no_fp_elim_all` in the lit
test file `callee-frame-setup.ll`.
The generated code changed from:
```
s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
s_mov_b32 s4, s33
s_mov_b32 s33, s32
s_mov_b32 s33, s4
s_setpc_b64 s[30:31]
```
into something like:
```
s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0)
v_writelane_b32 v63, s33, 0
s_mov_b32 s33, s32
v_readlane_b32 s33, v63, 0
s_setpc_b64 s[30:31]
```
I think we still prefer the old version where only scalar instructions are needed.
The idea here is free the reserved VGPR if no SGPR spills. So we will very likely
to use a free SGPR for fp/sp spill.
Reviewed by: arsenm
Differential Revision: https://reviews.llvm.org/D98344
This patch extends the matrix spec to allow matrix-by-scalar division.
Originally support for `/` was left out to avoid ambiguity for the
matrix-matrix version of `/`, which could either be elementwise or
specified as matrix multiplication M1 * (1/M2).
For the matrix-scalar version, no ambiguity exists; `*` is also
an elementwise operation in that case. Matrix-by-scalar division
is commonly supported by systems including Matlab, Mathematica
or NumPy.
Reviewed By: rjmccall
Differential Revision: https://reviews.llvm.org/D97857
