AMD has it's own implementation of vector calls.
New vector calls are introduced in the library for exp10, log10, sincos and finite asin/acos
Please refer [https://github.com/amd/aocl-libm-ose]
---------
Co-authored-by: Rohit Aggarwal <Rohit.Aggarwal@amd.com>
This computes a structural hash while allowing for selective ignoring of
certain operands based on a custom function that is provided. Instead of
a single hash value, it now returns FunctionHashInfo which includes a
hash value, an instruction mapping, and a map to track the operand
location and its corresponding hash value that is ignored.
Depends on https://github.com/llvm/llvm-project/pull/112621.
This is a patch for
https://discourse.llvm.org/t/rfc-global-function-merging/82608.
This is a recommit of #107120 . The original PR was approved but failed
buildbot. The newly added tests should only be run for compilers that
support the ARM target. This has been resolved by adding a config file
for these tests.
- Pass optimizes memcpy's by padding out destinations and sources to a
full word to make ARM backend generate full word loads instead of
loading a single byte (ldrb) and/or half word (ldrh). Only pads
destination when it's a stack allocated constant size array and source
when it's constant string. Heuristic to decide whether to pad or not
is very basic and could be improved to allow more examples to be
padded.
- Pass works at the midend level
This is part of the effort to support for enabling plugins on windows by
adding better support for building llvm as a DLL. The export macros used
here were added in #96630
Since shared library symbols aren't deduplicated across multiple
libraries on windows like Linux we have to manually explicitly import
and export `Any::TypeId` template instantiations for the uses of
`llvm::Any` in the LLVM codebase to support LLVM Windows shared library
builds.
This change ensures that external code, including LLVM's own tests, can
use PassManager callbacks when LLVM is built as a DLL.
I also removed the only use of llvm::Any for LoopNest that only existed
in debug code and there also doesn't seem to be any code creating
`Any<LoopNest>`
With the introduction of the nusw flag in GEPNoWrapFlags, it should be
safe to weaken the check in LoopAccessAnalysis to just check the nusw
flag on the GEP, instead of inbounds.
isNoWrap has exactly one caller which handles Assume = true separately,
but too conservatively. Instead, pass Assume to isNoWrap, so it is
threaded into getPtrStride, which has the correct handling for the
Assume flag. Also note that the Stride == 1 check in isNoWrap is
incorrect: getPtrStride returns Strides == 1 or -1, except when
isNoWrapAddRec or Assume are true, assuming ShouldCheckWrap is true; we
can include the case of -1 Stride, and when isNoWrapAddRec is true. With
this change, passing Assume = true to getPtrStride could return a
non-unit stride, and we correctly handle that case as well.
Based on this RFC:
https://discourse.llvm.org/t/rfc-allow-the-scalarizer-pass-to-scalarize-vectors-returned-in-structs/82306
LLVM intrinsics do not support out params. To get around this limitation
implementers will make intrinsics return structs to capture a return
type and an out param. This implementation detail should not impact
scalarization since these cases should be elementwise operations.
## Three changes are needed.
- The CallInst visitor needs to be updated to handle Structs
- A new visitor is needed for `ExtractValue` instructions
- finsh needs to be update to handle structs so that insert elements are
properly propogated.
## Testing changes
- Add support for `llvm.frexp`
- Add support for `llvm.dx.splitdouble`
fixes https://github.com/llvm/llvm-project/issues/111437
GEP offsets have sext_or_trunc semantics. We were already doing
this for the outer-most GEP, but not for the inner ones.
I believe one of the sanitizer buildbot failures was due to this,
but I did not manage to reproduce the issue or come up with a
test case. Usually the problematic case will already be folded
away due to index type canonicalization.
This patch adds basic support for `scalbln, scalblnf, scalblnl, scalbn,
scalbnf, scalbnl`. Constant folding support will be submitted in a
subsequent patch.
Related issue: <#112631>
The -enable-memprof-indirect-call-support meant to guard the recently
added memprof ICP support was not used in enough places. Specifically,
it was not checked in mayHaveMemprofSummary, which is called from the
ThinLTO backend applyImports. This led to failures when checking the
callsite records, as we incorrectly expected records for indirect calls.
Fix the option to be checked in all necessary locations, and add
testing.
- Pass optimizes memcpy's by padding out destinations and sources to a
full word to make backend generate full word loads instead of loading a
single byte (ldrb) and/or half word (ldrh). Only pads destination when
it's a stack allocated constant size array and source when it's constant
array. Heuristic to decide whether to pad or not is very basic and could
be improved to allow more examples to be padded.
- Pass works within GlobalOpt but is disabled by default on all targets
except ARM.
This fixes all the places that hit the new assertion added in
https://github.com/llvm/llvm-project/pull/106524 in tests. That is,
cases where the value passed to the APInt constructor is not an N-bit
signed/unsigned integer, where N is the bit width and signedness is
determined by the isSigned flag.
The fixes either set the correct value for isSigned, set the
implicitTrunc flag, or perform more calculations inside APInt.
Note that the assertion is currently still disabled by default, so this
patch is mostly NFC.
Today, InstCombine can fold fcmp+select patterns to minnum/maxnum
intrinsics when the nnan and nsz flags are set. The ordering of the
operands in both the fcmp and select instructions is important for the
folding to occur.
maxnum patterns:
1. (a op b) ? a : b -> maxnum(a, b), where op is one of {ogt, oge}
2. (a op b) ? b : a -> maxnum(a, b), where op is one of {ule, ult}
The second pattern is supposed to make the order of the operands in the
select instruction irrelevant. However, the pattern matching code uses
the CmpInst::getInversePredicate method to invert the comparison
predicate. This method doesn't take into account the fast-math flags,
which can lead missing the folding opportunity.
The patch extends the pattern matching code to handle unordered fcmp
instructions. This allows the folding to occur even when the select
instruction has the operands in the inverse order.
New maxnum patterns:
1. (a op b) ? a : b -> maxnum(a, b), where op is one of {ugt, uge}
2. (a op b) ? b : a -> maxnum(a, b), where op is one of {ole, olt}
The same changes are applied to the minnum intrinsic.
foldSelectEquivalence currently doesn't support GVN-like replacements on
vector types. Put in the checks for potentially lane-crossing
operations, and lift the limitation.
Reuse llvm::isTriviallyVectorizable in llvm::isNotCrossLaneOperation, in
order to get it to handle more intrinsics.
Alive2 proofs for changed tests: https://alive2.llvm.org/ce/z/XSV_GT
Factor out and unify common code from InstSimplify and InstCombine that
partially guard against cross-lane vector operations into
llvm::isNotCrossLaneOperation in ValueTracking.
Alive2 proofs for changed tests: https://alive2.llvm.org/ce/z/68H4ka
A call to a function that has this attribute is not a source of
divergence, as used by UniformityAnalysis. That allows a front-end to
use known-name calls as an instruction extension mechanism (e.g.
https://github.com/GPUOpen-Drivers/llvm-dialects ) without such a call
being a source of divergence.
This patch enables support for cloning in indirect callsites.
This is done by synthesizing callsite records for each virtual call
target from the profile metadata. In the thin link all the synthesized
records for a particular indirect callsite initially share the same
context node, but support is added to partition the callsites and
outgoing edges based on the callee function, creating a separate node
for each target.
In the LTO backend, when cloning is needed we first perform indirect
call promotion, then change the target of the new direct call to the
desired clone.
Note this is ThinLTO-specific, since for regular LTO indirect call
promotion should have already occurred.
Currently we will not be able to inline a large function even if it only
has one live use because the inline cost is still very high after
applying `LastCallToStaticBonus`, which is a constant. This could
significantly impact the performance because CSR spill is very
expensive.
This PR adds a new function `getInliningLastCallToStaticBonus` to TTI to
allow targets to customize this value.
Fixes SWDEV-471398.
There are a number of places where we call getSmallConstantMaxTripCount
without passing a vector of predicates:
getSmallBestKnownTC
isIndvarOverflowCheckKnownFalse
computeMaxVF
isMoreProfitable
I've changed all of these to now pass in a predicate vector so that
we get the benefit of making better vectorisation choices when we
know the max trip count for loops that require SCEV predicate checks.
I've tried to add tests that cover all the cases affected by these
changes.
Porting to TTI provides direct access to the instruction cost model,
which can enable instruction cost based sinking without introducing code
duplication.
createMIBNode does not modify MIBCallStack, so we can take ArrayRef
instead.
While I am at it, this patch changes the type of MIBPayload to
SmallVector. We put at most three elements, so we can avoid a heap
allocation.
Extend decomposeBitTestICmp() to handle cases where the resulting
comparison is of the form `icmp (X & Mask) pred C` with non-zero
`C`. Add a flag to allow code to opt-in to this behavior and use it in
the "log op of icmp" fold infrastructure.
This addresses regressions from #97289.
Proofs: https://alive2.llvm.org/ce/z/hUhdbU
Allow AllowEphemerals in isValidAssumeForContext, as the CxtI might
be the producer of the pointer in the bundle. At the moment, align
assumptions aren't optimized away.
This allows using the assumption in the computeKnownBits call in
getConstantMultipleImpl.
We could extend the computeKnownBits API to allow callers to specify if
ephemerals are allowed, if the info from computeKnownBitsFromContext is
used to remove alignment assumptions.
PR: https://github.com/llvm/llvm-project/pull/108632
Canonicalize le to lt predicate by adjusting the constant by one.
As such, we only have to handle slt and ult.
This will make it easier to handle additional cases.
There is a stray break statement in the recurrence-handling code in
computeKnownBitsFromOperator, that seems to be unintended. Strip this
statement so that we have the opportunity to go through the rest of
phi-handling code, and refine KnownBits further.
