llvm-project.git/llvm/lib/CodeGen/CodeGenPrepare.cpp, branch users/boomanaiden154/main.lit-remove-t-from-tests Unnamed repository; edit this file 'description' to name the repository. [CodeGenPrepare] Bail out of usubo creation if sub's parent is not the same as the comparison (#160358) 2025-09-25T13:55:01+00:00 AZero13 gfunni234@gmail.com 2025-09-25T13:55:01+00:00 09bdbfd9d1a7b10b3b91c86672ca2994dc997c82 We match uadd's behavior here. Codegen comparison: https://godbolt.org/z/x8j4EhGno 
We match uadd's behavior here.

Codegen comparison: https://godbolt.org/z/x8j4EhGno
 [CodeGenPrepare] Consider target memory intrinics as memory use (#159638) 2025-09-19T21:18:21+00:00 Jeffrey Byrnes jeffrey.byrnes@amd.com 2025-09-19T21:18:21+00:00 d8a4c61fe4c11b16a496a089e2962473c6b1b183 When deciding to sink address instructions into their uses, we check if it is profitable to do so. The profitability check is based on the types of uses of this address instruction -- if there are users which are not memory instructions, then do not fold. However, this profitability check wasn't considering target intrinsics, which may be loads / stores. This adds some logic to handle target memory intrinsics. 
When deciding to sink address instructions into their uses, we check if
it is profitable to do so. The profitability check is based on the types
of uses of this address instruction -- if there are users which are not
memory instructions, then do not fold.

However, this profitability check wasn't considering target intrinsics,
which may be loads / stores.

This adds some logic to handle target memory intrinsics.
 Re-apply "[NFCI][Globals] In GlobalObjects::setSectionPrefix, do conditional update if existing prefix is not equivalent to the new one. Returns whether prefix changed." (#159161) 2025-09-16T20:33:29+00:00 Mingming Liu mingmingl@google.com 2025-09-16T20:33:29+00:00 8b3c91c4fb1b8efaccb71894684e4fb16b0e8945 This is a reland of https://github.com/llvm/llvm-project/pull/158460 Test failures are gone once I undo the changes in codegenprepare. 
This is a reland of https://github.com/llvm/llvm-project/pull/158460

Test failures are gone once I undo the changes in codegenprepare.
 Revert "[NFCI][Globals] In GlobalObjects::setSectionPrefix, do conditional update if existing prefix is not equivalent to the new one. Returns whether prefix changed." (#159159) 2025-09-16T19:51:54+00:00 Mingming Liu mingmingl@google.com 2025-09-16T19:51:54+00:00 9277bcd1ab83fbfecfaaa5e15f3821808c5a6e38 Reverts llvm/llvm-project#158460 due to buildbot failures 
Reverts llvm/llvm-project#158460 due to buildbot failures
 [NFCI][Globals] In GlobalObjects::setSectionPrefix, do conditional update if existing prefix is not equivalent to the new one. Returns whether prefix changed. (#158460) 2025-09-16T19:01:21+00:00 Mingming Liu mingmingl@google.com 2025-09-16T19:01:21+00:00 027bccc4692923d0f1ba3d4d970071f747c2255c Before this change, `setSectionPrefix` overwrites existing section prefix with new one unconditionally. After this change, `setSectionPrefix` checks for equivalences, updates conditionally and returns whether an update happens. Update the existing callers to make use of the return value. [PR 155337](https://github.com/llvm/llvm-project/pull/155337/files#diff-cc0c67ac89807f4453f0cfea9164944a4650cd6873a468a0f907e7158818eae9) is a motivating use case whether the 'update' semantic is needed. 
Before this change, `setSectionPrefix` overwrites existing section
prefix with new one unconditionally.

After this change, `setSectionPrefix` checks for equivalences, updates
conditionally and returns whether an update happens.

Update the existing callers to make use of the return value. [PR
155337](https://github.com/llvm/llvm-project/pull/155337/files#diff-cc0c67ac89807f4453f0cfea9164944a4650cd6873a468a0f907e7158818eae9)
is a motivating use case whether the 'update' semantic is needed.
 [CodeGen] Remove ExpandInlineAsm hook (#156617) 2025-09-04T07:28:11+00:00 Nikita Popov npopov@redhat.com 2025-09-04T07:28:11+00:00 3f757a39f2855cd06c62a85b8e27fd56fa017e78 This hook replaces inline asm with LLVM intrinsics. It was intended to match inline assembly implementations of bswap in libc headers and replace them more optimizable implementations. At this point, it has outlived its usefulness (see https://github.com/llvm/llvm-project/issues/156571#issuecomment-3247638412), as libc implementations no longer use inline assembly for this purpose. Additionally, it breaks the "black box" property of inline assembly, which some languages like Rust would like to guarantee. Fixes https://github.com/llvm/llvm-project/issues/156571. 
This hook replaces inline asm with LLVM intrinsics. It was intended to
match inline assembly implementations of bswap in libc headers and
replace them more optimizable implementations.

At this point, it has outlived its usefulness (see
https://github.com/llvm/llvm-project/issues/156571#issuecomment-3247638412),
as libc implementations no longer use inline assembly for this purpose.

Additionally, it breaks the "black box" property of inline assembly,
which some languages like Rust would like to guarantee.

Fixes https://github.com/llvm/llvm-project/issues/156571.
 [llvm] Replace SmallSet with SmallPtrSet (NFC) (#154068) 2025-08-18T14:01:29+00:00 Kazu Hirata kazu@google.com 2025-08-18T14:01:29+00:00 07eb7b76928d6873c60859a0339591ed9e0f512a This patch replaces SmallSet<T *, N> with SmallPtrSet<T *, N>. Note that SmallSet.h "redirects" SmallSet to SmallPtrSet for pointer element types: template <typename PointeeType, unsigned N> class SmallSet<PointeeType*, N> : public SmallPtrSet<PointeeType*, N> {}; We only have 140 instances that rely on this "redirection", with the vast majority of them under llvm/. Since relying on the redirection doesn't improve readability, this patch replaces SmallSet with SmallPtrSet for pointer element types. 
This patch replaces SmallSet<T *, N> with SmallPtrSet<T *, N>.  Note
that SmallSet.h "redirects" SmallSet to SmallPtrSet for pointer
element types:

  template <typename PointeeType, unsigned N>
class SmallSet<PointeeType*, N> : public SmallPtrSet<PointeeType*, N>
{};

We only have 140 instances that rely on this "redirection", with the
vast majority of them under llvm/. Since relying on the redirection
doesn't improve readability, this patch replaces SmallSet with
SmallPtrSet for pointer element types.
 [LLVM][CGP] Allow finer control for sinking compares. (#151366) 2025-08-05T10:43:41+00:00 Paul Walker paul.walker@arm.com 2025-08-05T10:43:41+00:00 94d374ab6cd559fa96e96bfb805b73a367bfda08 Compare sinking is selectable based on the result of hasMultipleConditionRegisters. This function is too coarse grained by not taking into account the differences between scalar and vector compares. This PR extends the interface to take an EVT to allow finer control. The new interface is used by AArch64 to disable sinking of scalable vector compares, but with isProfitableToSinkOperands updated to maintain the cases that are specifically tested. 
Compare sinking is selectable based on the result of
hasMultipleConditionRegisters. This function is too coarse grained by
not taking into account the differences between scalar and vector
compares. This PR extends the interface to take an EVT to allow finer
control.
    
The new interface is used by AArch64 to disable sinking of scalable
vector compares, but with isProfitableToSinkOperands updated to maintain
the cases that are specifically tested.
 [X86][APX] Do optimizeMemoryInst for v1X masked load/store (#151331) 2025-07-31T03:52:39+00:00 Phoebe Wang phoebe.wang@intel.com 2025-07-31T03:52:39+00:00 ebf96f93160973ab5b0c886655adde850affc707 Fix redundant LEA: https://godbolt.org/z/34xEYE818 
Fix redundant LEA: https://godbolt.org/z/34xEYE818
 [CodeGenPrepare] Make sure that `AddOffset` is also a loop invariant (#150625) 2025-07-25T16:23:56+00:00 Yingwei Zheng dtcxzyw2333@gmail.com 2025-07-25T16:23:56+00:00 2d0ca09305fcece75e2c501f1ec74aa6eada69a0 Closes https://github.com/llvm/llvm-project/issues/150611. 
Closes https://github.com/llvm/llvm-project/issues/150611.