Writing zeros to shadow (including checking for existing zero) is now ~2x
faster on one example.
Reviewed By: morehouse
Differential Revision: https://reviews.llvm.org/D110733
Add MCDwarfLineStr class to the public API.
Note that MCDwarfLineTableHeader::Emit(), takes MCDwarfLineStr as
an Optional<> parameter making it impossible to use the API if the class
is not publicly defined.
Reviewed By: alexander-shaposhnikov
Differential Revision: https://reviews.llvm.org/D109412
This patch removes the uneccessary mf/mtvsr generated in conjunction
with xscvdpsxws/xscvdpuxws.
Differential revision: https://reviews.llvm.org/D109902
Fold the transformational intrinsic function FINDLOC() for
all combinations of optional arguments and data types.
Differential Revision: https://reviews.llvm.org/D110757
This is NFCI because the pattern with 2 left-shifts should get
folded independently by smaller folds.
The motivation is to refine this block to avoid infinite loops
seen with D110170.
Add methods to appropriately extend KnownBits/ConstantRange there,
same as with APInt. Also clean up the known bits handling by
actually doing that extension rather than checking ZExtBits. This
doesn't matter now, but becomes relevant once truncation is
involved.
A followup to D110201.
For example, we'd set OptimizationRemarkMissed's Regex to '.*' when
encountering -Rpass. Normally this doesn't actually affect remarks we
emit because in clang::ProcessWarningOptions() we'll separately look at
all -R arguments and turn on/off corresponding diagnostic groups.
However, this is reproducible with -round-trip-args.
Reviewed By: JamesNagurne
Differential Revision: https://reviews.llvm.org/D110673
The algorithm used to fold SQRT has some holes that
led to test failures; debug and add more tests.
Differential Revision: https://reviews.llvm.org/D110744
When clang crashes, it writes a standalone source file and shell script
to reproduce the crash.
The Driver used to set `Mode = CPPMode` in generateCompilationDiagnostics()
to force preprocessing mode. This has the side effect of making
IsCLMode() return false, which in turn meant Clang::AddClangCLArgs()
didn't get called when creating the standalone source file, which meant
the stand-alone file was preprocessed with the gcc driver's defaults
In particular, exceptions default to on with the gcc driver, but to
off with the cl driver. The .sh script did use the original command
line, so in the reproducer for a clang-cl crash, the standalone source
file could contain exception-using code after preprocessing that the
compiler invocation in the shell script would then complain about.
This patch removes the `Mode = CPPMode;` line and instead additionally
checks for `CCGenDiagnostics` in most places that check `CCCIsCPP().
This also matches the strategy Clang::ConstructJob() uses to add
-frewrite-includes for creating the standalone source file for a crash
report.
Fixes PR52007.
Differential Revision: https://reviews.llvm.org/D110783
Call Driver::getFinalPhase() instead of duplicating it.
https://reviews.llvm.org/D65993 added the duplication, then
02e35832c3 maded it more obviously a copy of getFinalPhase().
The only difference is that getCompilationPhases() used to use
LastPhase / IfsMerge where getFinalPhase() used Link. Adapt
getFinalPhase() to return IfsMerge when needed.
No intentional behavior change.
Differential Revision: https://reviews.llvm.org/D110770
Previously, the definitions of __cxa_terminate_handler and __cxa_unexpected_handler
(and their set_xxx_handler functions) were grouped together, but the
definition of __cxa_new_handler wasn't. This commit simply moves those
to the same file to treat all handlers consistently.
Support the extension intrinsic subroutines EXIT([status]) and ABORT()
in the intrinsic table and runtime support library. Lowering remains
to be done.
Differential Revision: https://reviews.llvm.org/D110741
For convolution, the input window dimension's access affine map
is of the form `(d0 * s0 + d1)`, where `d0`/`d1` is the output/
filter window dimension, and `s0` is the stride.
When tiling, https://reviews.llvm.org/D109267 changed how the
way dimensions are acquired. Instead of directly querying using
`*.dim` ops on the original convolution op, we now get it by
applying the access affine map to the loop upper bounds. This
is fine for dimensions having single-dimension affine maps,
like matmul, but not for convolution input. It will cause
incorrect compuation and out of bound. A concrete example, say
we have 1x225x225x3 (NHWC) input, 3x3x3x32 (HWCF) filter, and
1x112x112x3 (NHWC) output with stride 2, (112 * 2 + 3) would be
227, which is different from the correct input window dimension
size 225.
Instead, we should first calculate the max indices for each loop,
and apply the affine map to them, and then plus one to get the
dimension size. Note this makes no difference for matmul-like
ops given they will have `d0 - 1 + 1` effectively.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D110849
It is now very simple and can go right into the header
allowing optimizer to combine callers, such as isVGPRClass
and similar.
It does not need anything from the TRI itself anymore, so
make it static class member along with the callers.
Differential Revision: https://reviews.llvm.org/D110762
This avoids creating tons of make_unique template instantiations. And we
only create a unique_ptr of the actual pass concept type, rather than
creating a unique_ptr of the pass model subclass then casting it to the
pass concept type.
This reduces the work spent compiling PassBuilder.cpp from 83M -> 73M
instructions according to perf stat.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D110784
We create many instantiations of PassManager::addPass() in
PassBuilder.cpp. vector::emplace_back() and make_unique() are both
templated and would have many instantiations based on the number of
times we instantiate addPass(). Now we directly construct the
unique_ptr with the type as the actual unique_ptr type in the vector we
are adding it to, so we only have one unique_ptr constructor
instantiation across all addPass() instantiations and only the
non-templated push_back().
This makes PassBuilder.cpp slightly faster to build.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D110775
The expansion for these was updated in https://reviews.llvm.org/D47927 but the cost model was not adjusted.
I believe the cost model was also incorrect for the old expansion.
The expansion prior to D47927 used 3 icmps using LHS, RHS, and Result
to calculate theirs signs. Then 2 icmps to compare the signs. Followed
by an And. The previous cost model was using 3 icmps and 2 selects.
Digging back through git blame, those 2 selects in the cost model used to
be 2 icmps, but were changed in https://reviews.llvm.org/D90681
Differential Revision: https://reviews.llvm.org/D110739
* This could have been removed some time ago as it only had one op left in it, which is redundant with the new approach.
* `matmul_i8_i8_i32` (the remaining op) can be trivially replaced by `matmul`, which natively supports mixed precision.
Differential Revision: https://reviews.llvm.org/D110792
This patch improves the effectiveness of BDCE's debug info salvaging
by processing the instructions in reverse order and delaying
dropAllReferences until after debug info salvaging. This allows
salvaging of entire chains of deleted instructions!
Previously we would remove all references from an instruction, which
would make it impossible to use that instruction to salvage a later
instruction in the instruction stream, because its operands were
already removed.
This reapplies the previous patch with a fix for a use-after-free.
Differential Revision: https://reviews.llvm.org/D110568
Fine tuning after couple of upstreamed patches.
Rename the lenParams varaiables in buidlers to typeparams.
This patch is part of the upstreaming effort from fir-dev branch.
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D110821
Co-authored-by: Valentin Clement <clementval@gmail.com>
This is a more complex pattern than we handled with the
initial patch for PR51245:
D110342 / 09e71c367a
We could extend the logic matching to allow a setcc as
one operand and an extract of vector setcc (or even an
arbitrary bool?) as the other.
Add pinned attributes and speicifc builders.
The pinned attribute helps mark those allocas in OpenMP regions that should not
be hoisted out by an alloca hoisting pass.
This patch is part of the upstreaming effort from fir-dev branch.
Coming from PR: https://github.com/flang-compiler/f18-llvm-project/pull/1065
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D110815
Co-authored-by: Valentin Clement <clementval@gmail.com>
Apple's libc++ has a few differences with the LLVM libc++, and it is
necessary to use a custom configuration file to test it properly.
Differential Revision: https://reviews.llvm.org/D110777
After rG452714f8f8037ff37f9358317651d1652e231db2, the Function `F` retrieved in LoopPredication is not used.
Remove this unused variable to stop some buildbots (ASAN, clang-ppc) from failing.
This is analogous to D86156 (which preserves "lossy" BFI in loop
passes). Lossy means that the analysis preserved may not be up to date
with regards to new blocks that are added in loop passes, but BPI will
not contain stale pointers to basic blocks that are deleted by the loop
passes.
This is achieved through BasicBlockCallbackVH in BPI, which calls
eraseBlock that updates the data structures in BPI whenever a basic
block is deleted.
This patch does not have any changes in the upstream pipeline, since
none of the loop passes in the pipeline use BPI currently.
However, since BPI wasn't previously preserved in loop passes, the loop
predication pass was invoking BPI *on the entire
function* every time it ran in an LPM. This caused massive compile time
in our downstream LPM invocation which contained loop predication.
See updated test with an invocation of a loop-pipeline containing loop
predication and -debug-pass turned ON.
Reviewed-By: asbirlea, modimo
Differential Revision: https://reviews.llvm.org/D110438
Move fir.freemem and fir.store to the Memory SSA operations sections.
Move parser, printer and verifier of fir.store to the .cpp file.
This patch is part of the upstreaming effort from fir-dev branch.
Co-authored-by: Jean Perier <jperier@nvidia.com>
Co-authored-by: Eric Schweitz <eschweitz@nvidia.com>
Reviewed By: jeanPerier
Differential Revision: https://reviews.llvm.org/D110816
Previously, the dialect was exposed for linking and pass management purposes,
but we did not generate op classes for it. Generate them.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D110819
This adds a few more unpredicated intrinsics to sink splats to, in order
to create more qr instruction variants. Notably this includes
saddsat/uaddsat but also some of the unpredicated mve intrinsics.
Differential Revision: https://reviews.llvm.org/D110333
Andrew Browne