D117829 added the generic "__builtin_reduce_mul" which we can use to replace the x86 specific integer mul reduction builtins - internally these were mapping to the same intrinsic already so there are no test changes required.
Differential Revision: https://reviews.llvm.org/D125222
When extracting the first lane of a predicate created using the
llvm.get.active.lane.mask intrinsic, it should give the same codegen as
when the predicate is created using the llvm.aarch64.sve.whilelo
intrinsic, since get.active.lane.mask is lowered to whilelo. This patch
ensures the codegen is the same by recognizing
llvm.get.active.lane.mask as a flag-setting operation in this case.
Differential Revision: https://reviews.llvm.org/D125215
Previously the EXPECT_AVAILABLE macros would rebuild the code at each marked
point, by expanding the cases textually.
There were often lots, and it's nice to have lots!
This reduces total unittest time by ~10% on my machine.
I did have to sacrifice a little apply() coverage in AddUsingTests (was calling
expandCases directly, which was otherwise unused), but we have
EXPECT_AVAILABLE tests covering that, I don't think there's real risk here.
Differential Revision: https://reviews.llvm.org/D125109
This is a clever cross-cutting sanity test for clang's arg parsing I suppose.
But clangd creates thousands of invocations, ~all with identical trivial
arguments, and problems with these would be caught by clang's tests.
This overhead accounts for 10% of total unittest time!
Differential Revision: https://reviews.llvm.org/D125169
These aren't needed. With them the generated predefines buffer is 13KB.
For every TestTU, we must:
- generate the buffer (3 times: parsing preamble, scanning preamble, main file)
- parse the buffer (again 3 times)
- serialize all the macros it defines in the PCH
- compress the buffer itself to write it into the PCH
- decompress it from the PCH
Avoiding this reduces unit test time by ~25%.
Differential Revision: https://reviews.llvm.org/D125172
The output buffer has a 'back' member, which returns NUL when you try
it with an empty buffer. But there are no use cases that need that
additional functionality. This makes the 'back' member behave more
like STL containers' back members. (It still returns a value, not a
reference.)
Reviewed By: dblaikie
Differential Revision: https://reviews.llvm.org/D123201
Similar to the existing bitwise reduction builtins, this lowers to a llvm.vector.reduce.mul intrinsic call.
For other reductions, we've tried to share builtins for float/integer vectors, but the fmul reduction intrinsic also take a starting value argument and can either do unordered or serialized, but not reduction-trees as specified for the builtins. However we address fmul support this shouldn't affect the integer case.
Differential Revision: https://reviews.llvm.org/D117829
There's many instances in clang tidy checks where owning strings are used when we already have a stable string from the options, so using a StringRef makes much more sense.
Reviewed By: aaron.ballman
Differential Revision: https://reviews.llvm.org/D124341
See [[ https://github.com/llvm/llvm-project/issues/55040 | issue 55040 ]] where static members of classes declared in the anonymous namespace are incorrectly returned as member fields from lldb::SBType::GetFieldAtIndex(). It appears that attrs.member_byte_offset contains a sentinel value for members that don't have a DW_AT_data_member_location.
Reviewed By: labath
Differential Revision: https://reviews.llvm.org/D124409
Converts to SVBool are already considered as a nop, if they
are converting an operand from a ptrue or a cmp, because
they zero the extra predicate lanes by construction.
This patch adds 2 similar cases:
- The wide cmp, which were not directly recognized by the test
for other forms of cmp
- Splats of 1, which will be generated as ptrue, and as such
will also zero the extra predicate lines.
Reviewed By: paulwalker-arm, peterwaller-arm
Differential Revision: https://reviews.llvm.org/D124908
libm doesn't have overloads for the small types, so promote them to a
bigger type and use the f32 function.
Differential Revision: https://reviews.llvm.org/D125093
IIUC, the purpose of CopyUniqueClassMethodTypes is to link together
class definitions in two compile units so that we only have a single
definition of a class. It does this by adding entries to the die_to_type
and die_to_decl_ctx maps.
However, the direction of the linking seems to be reversed. It is taking
entries from the class that has not yet been parsed, and copying them to
the class which has been parsed already -- i.e., it is a very
complicated no-op.
Changing the linking order allows us to revert the changes in D13224
(while keeping the associated test case passing), and is sufficient to
fix PR54761, which was caused by an undesired interaction with that
patch.
Differential Revision: https://reviews.llvm.org/D124370
This isn't a configuration that we unfortunately can add to
the CI practically at the moment, but I do run the tests
sporadically offline in this configuration.
Differential Revision: https://reviews.llvm.org/D124993
The initial support for the Ampere1 mistakenly signalled support for
the MTE feature. However, the core does not include the optional MTE
functionality.
Update the target parser to not include MTE for Ampere1.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D125191
This patch implements a for a target specific optimization that replaces
the cmp and csel from cttz with an and mask.
Differential Revision: https://reviews.llvm.org/D123782
Fixes https://github.com/llvm/llvm-project/issues/54522.
This fixes regression introduced in 5e5efd8a91.
Before the culprit commit, macros in WhitespaceSensitiveMacros were correctly formatted even if their closing parenthesis weren't followed by semicolon (or, to be precise, when they were followed by a newline).
That commit changed the type of the macro token type from TT_UntouchableMacroFunc to TT_FunctionLikeOrFreestandingMacro.
Correct formatting (with `WhitespaceSensitiveMacros = ['FOO']`):
```
FOO(1+2)
FOO(1+2);
```
Regressed formatting:
```
FOO(1 + 2)
FOO(1+2);
```
Reviewed By: HazardyKnusperkeks, owenpan, ksyx
Differential Revision: https://reviews.llvm.org/D123676
This prevents an infinite loop from D123801, where code trying to reduce
the total number of bitcasts, but also handling constants, could create
the opposite transform. Prevent the transform in these case to let the
bitcast of a constant transform naturally.
Fixes#55345
When processing an entry-stmt in name resolution, attrs_ was
reset before SetBindNameOn was called, causing the symbol to lose
the binding label information.
Differential Revision: https://reviews.llvm.org/D125097
The per-callsite size threshold used today to drive preinline decision is based on hotness/coldness cutoff. The default setup is for callsites with a sample count above the hotness cutoff (99%), a 1500 size threshold is used. Any callsite below 99.99% coldness cutoff uses a zero threshold. This has a couple issues:
1. While both cutoffs and size thoresholds are configurable, different applications may need different setups, making a universal setup impractical.
2. The callsites between hotness cutoff and coldness cutoff are not considered as inline candidates, which could be a missing opportunity.
3. Hot callsites always use the same threshold. In reality we may want a bigger threshold for hotter callsites.
In this change we are introducing a linear threshold regardless of hot/cold cutoffs. Given a sample space, a threshold is computed for a callsite based on the position of that callsite sample in the whole space. With that we no longer need to define what's hot or cold. Callsites with different hotness will get a different threshold. This should overcome the above three issues.
I have seen good results with a universal default setup for two of our internal services.
For one service, 0.2% to 0.5% perf improvement over a baseline with a previous default setup, on-par code size.
For the second service, 0.5% to 0.8% perf improvement over a baseline with a previous default setup, 0.2% code size increase; on-par performance and code size with a baseline that is with a carefully tuned cutoff to cover enough hot functions.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D125023
Adds missing logic in the lowering from NvGPU to NVVM to support fp32
(in an accumulator operand) and tf32 (in multiplicand operand) types.
Fixes logic in one of the helper functions for converting the result
of a mma.sync operation with multiple 8x256bit output tiles, which is
the case for f32 outputs.
Differential Revision: https://reviews.llvm.org/D124533
As Fortran 2018 5.2.2 states, a program shall consist of exactly one
main program. Add this semantic check.
Reviewed By: klausler
Differential Revision: https://reviews.llvm.org/D125186
These are all microcoded/multi-pipe nightmares on Ryzen, but we shouldn't just be using the WriteMicrocoded class which is for REALLY bad microcoded nightmares - instead use the same approximate latencies as znver2 (Agner and uops.info both suggest similar values) - and make sure we use the FPU defs for both
Fixes#53242
\operation ... \endoperation are not valid doxygen commands and cause issues when -Wdocumentation is enabled (Issue #35297)
This patch proposes to replace them with \code{.operation} ... \endcode blocks so that the pseudo-code is correctly retained in any documentation and downstream can use the ".operation" type for its own formatting.
Differential Revision: https://reviews.llvm.org/D125170
Given a commutative reduction leading from a shuffle, the order of the
lanes on the shuffle are not important for the result. This means we can
reorder the shuffle to something simpler, which we try shuffling the
first vector lanes first. This was D123494.
The new shuffle may not be profitable though, and if it is not we can
try the folding of select shuffles from D123911. This, with some
adjustment as the output lane ordering is now unimportant, can allow the
final shuffle to simplify given the inputs to the patterns from D123911.
Where as each transformation on their own are not profitable, the
combination is.
We can only support a single shuffle when called from reductions, but we
are able to sort the ReconstructMask, potentially allowing it to
simplify to an identity or concat mask.
Differential Revision: https://reviews.llvm.org/D125086
Another step toward enabling full -Wsystem-headers testing across all x86 headers
Fix a number of cases where the arg / return value signedness doesn't match the C/C++ intrinsic.
So far I've just added explicit casts as necessary, but we might want to address some of the mismatches directly
Differential Revision: https://reviews.llvm.org/D125164
