Currently, `llvm_bb_addr_map` sections are generated per section names because we use
the `LinkedToSymbol` argument of getELFSection. This will cause the address map tables of functions
grouped into the same section when `-function-sections=true -unique-section-names=false` which is not
the intended behaviour. This patch lets the unique id of every `.text` section propagate to the associated
`.llvm_bb_addr_map` section.
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D92113
The new MachO lld just grew support for response files in D92149, so let
the clang driver use it.
Differential Revision: https://reviews.llvm.org/D92399
This is yet another attempt at providing support for epilogue
vectorization following discussions raised in RFC http://llvm.1065342.n5.nabble.com/llvm-dev-Proposal-RFC-Epilog-loop-vectorization-tt106322.html#none
and reviews D30247 and D88819.
Similar to D88819, this patch achieve epilogue vectorization by
executing a single vplan twice: once on the main loop and a second
time on the epilogue loop (using a different VF). However it's able
to handle more loops, and generates more optimal control flow for
cases where the trip count is too small to execute any code in vector
form.
Reviewed By: SjoerdMeijer
Differential Revision: https://reviews.llvm.org/D89566
If an object file has an undefined foo@v1, we emit a dynamic symbol foo.
This is incorrect if at runtime a shared object provides the non-default version foo@v1
(the undefined foo may bind to foo@@v2, for example).
GNU ld issues an error for this case, even if foo@v1 is undefined weak
(https://sourceware.org/bugzilla/show_bug.cgi?id=3351). This behavior makes
sense because to represent an undefined foo@v1, we have to construct a Verneed
entry. However, without knowing the defining filename, we cannot construct a
Verneed entry (Verneed::vn_file is unavailable).
This patch implements the error.
Depends on D92258
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D92260
This is a straightforward port of MemCpyOpt to MemorySSA following
the approach of D26739. MemDep queries are replaced with MSSA queries
without changing the overall structure of the pass. Some care has
to be taken to account for differences between these APIs
(MemDep also returns reads, MSSA doesn't).
Differential Revision: https://reviews.llvm.org/D89207
The symbol resolution rules for versioned symbols are:
* foo@@v1 (default version) resolves both undefined foo and foo@v1
* foo@v1 (non-default version) resolves undefined foo@v1
Note, foo@@v1 must be defined (the assembler errors if attempting to
create an undefined foo@@v1).
For defined foo@@v1 in a shared object, we call `SymbolTable::addSymbol` twice,
one for foo and the other for foo@v1. We don't do the same for object files, so
foo@@v1 defined in one object file incorrectly does not resolve a foo@v1
reference in another object file.
This patch fixes the issue by reusing the --wrap code to redirect symbols in
object files. This has to be done after processing input files because
foo and foo@v1 are two separate symbols if we haven't seen foo@@v1.
Add a helper `Symbol::getVersionSuffix` to retrieve the optional trailing
`@...` or `@@...` from the possibly truncated symbol name.
Depends on D92258
Reviewed By: jhenderson
Differential Revision: https://reviews.llvm.org/D92259
Test the symbol resolution related to
* defined foo@@v1 and foo@v1 in object files/shared objects
* undefined foo@v1
* weak foo@@v1 and foo@v1
* visibility
* interaction with --wrap.
Reviewed By: grimar
Differential Revision: https://reviews.llvm.org/D92258
We were not correctly splitting a blocks for chains of length 1.
Before that change, additional instructions for blocks in chains of
length 1 were not split off from the block before removing (this was
done correctly for chains of longer size).
If this first block contained an instruction referenced elsewhere,
deleting the block, would result in invalidation of the produced value.
This caused a miscompile which motivated D92297 (before D17993,
nonnull and dereferenceable attributed were not added so MergeICmps were
not triggered.) The new test gep-references-bb.ll demonstrate the issue.
The regression was introduced in
rG0efadbbcdeb82f5c14f38fbc2826107063ca48b2.
This supersedes D92364.
Test case by MaskRay (Fangrui Song).
Differential Revision: https://reviews.llvm.org/D92375
Fix hip test failures that were introduced by
previous changes to hip-toolchain-rdc-static-lib.hip
test. The .*lld.* is matching a longer string than
expected.
Differential Revision: https://reviews.llvm.org/D92342
Without FMF, we lower these intrinsics into something like this:
vmaxsd %xmm0, %xmm1, %xmm2
vcmpunordsd %xmm0, %xmm0, %xmm0
vblendvpd %xmm0, %xmm1, %xmm2, %xmm0
But if we can ignore NANs, the single min/max instruction is enough
because there is no need to fix up the x86 logic that corresponds to
X > Y ? X : Y.
We probably want to make other adjustments for FP intrinsics with FMF
to account for specialized codegen (for example, FSQRT).
Differential Revision: https://reviews.llvm.org/D92337
We already expand select and select_cc in codegenprepare, but they can
still be generated under some situations. Explicitly mark them as expand
to ensure they are not produced, leading to a failure to select the
nodes.
Differential Revision: https://reviews.llvm.org/D92373
icmp is the preferred spelling in IR because icmp analysis is
expected to be better than any other analysis. This should
lead to more follow-on folding potential.
It's difficult to say exactly what we should do in codegen to
compensate. For example on AArch64, which of these is preferred:
sub w8, w0, w1
lsr w0, w8, #31
vs:
cmp w0, w1
cset w0, lt
If there are perf regressions, then we should deal with those in
codegen on a case-by-case basis.
A possible motivating example for better optimization is shown in:
https://llvm.org/PR43198 but that will require other transforms
before anything changes there.
Alive proof:
https://rise4fun.com/Alive/o4E
Name: sign-bit splat
Pre: C1 == (width(%x) - 1)
%s = sub nsw %x, %y
%r = ashr %s, C1
=>
%c = icmp slt %x, %y
%r = sext %c
Name: sign-bit LSB
Pre: C1 == (width(%x) - 1)
%s = sub nsw %x, %y
%r = lshr %s, C1
=>
%c = icmp slt %x, %y
%r = zext %c
* Un-inline the test.
* Use expect_expr everywhere and also check all involved types.
* Clang-format the test sources.
* Explain what we're actually testing with the 'C' and 'D' templates.
* Split out the non-template-parameter-pack part of the test into its own small test.
All three use readFile() for their argument so their argument file is
already copied to the tar, but we weren't rewriting the argument to
point to the path used in the tar file.
No test because the change is trivial (several other flags in
createResponseFile() also aren't tested, likely for the same reason.)
Differential Revision: https://reviews.llvm.org/D92356
If the shift amount was undef for some lane, the shift amount in opposite
shift is irrelevant for that lane, and the new shift amount for that lane
can be undef.
Added UNLIKELY hint to one-time or rarely executed branches.
This improves performance of the library on some tasking benchmarks.
Differential Revision: https://reviews.llvm.org/D92322
If the shift amount was undef for some lane, the shift amount in opposite
shift is irrelevant for that lane, and the new shift amount for that lane
can be undef.
There is no correctness need for that, and since we allow live-out
uses, this could theoretically happen, because currently nothing
will move the cond to right before the branch in those tests.
But regardless, lifting that restriction even makes the transform
easier to understand.
This makes the transform happen in 81 more cases (+0.55%)
)
Move the X86 VSELECT->UADDSAT fold to DAGCombiner - there's nothing target specific about these folds.
The SSE42 test diffs are relatively benign - its avoiding an extra constant load in exchange for an extra xor operation - there are extra register moves, which is annoying as all those operations should commute them away.
Differential Revision: https://reviews.llvm.org/D91876
The restriction on pointer-to-pointer kernel arguments has been
relaxed in OpenCL 2.0. Apply the same address space restrictions for
pointer argument types to the inner pointer types.
Differential Revision: https://reviews.llvm.org/D92091
In the following loop the dependence distance is 2 and can only be
vectorized if the vector length is no larger than this.
void foo(int *a, int *b, int N) {
#pragma clang loop vectorize(enable) vectorize_width(4)
for (int i=0; i<N; ++i) {
a[i + 2] = a[i] + b[i];
}
}
However, when specifying a VF of 4 via a loop hint this loop is
vectorized. According to [1][2], loop hints are ignored if the
optimization is not safe to apply.
This patch introduces a check to bail of vectorization if the user
specified VF is greater than the maximum feasible VF, unless explicitly
forced with '-force-vector-width=X'.
[1] https://llvm.org/docs/LangRef.html#llvm-loop-vectorize-and-llvm-loop-interleave
[2] https://clang.llvm.org/docs/LanguageExtensions.html#extensions-for-loop-hint-optimizations
Reviewed By: sdesmalen, fhahn, Meinersbur
Differential Revision: https://reviews.llvm.org/D90687
This patch replaces the attribute `unsigned VF` in the class
IntrinsicCostAttributes by `ElementCount VF`.
This is a non-functional change to help upcoming patches to compute the cost
model for scalable vector inside this class.
Differential Revision: https://reviews.llvm.org/D91532
PreferedType were not set when parsing compound literals, hence
designated initializers were not available as code completion suggestions.
This patch sets the preferedtype to parsed type for the following initializer
list.
Fixes https://github.com/clangd/clangd/issues/142.
Differential Revision: https://reviews.llvm.org/D92370
Zahira Ammarguellat