If the incoming block to a phi node is an EH pad, then we will
materialize into an EH pad, which is not supposed to happen. To fix
this, I added a check to see if incoming block of a phi node is an EH
pad before using it as the insertion point.
Differential Revision: https://reviews.llvm.org/D95019
Fuchsia's zxtest has a slightly different death test definition, and
this macro makes our death test API work on Fuchsia.
Reviewed By: sivachandra
Differential Revision: https://reviews.llvm.org/D95648
Given a shuffle(vqdmulh(shuffle, shuffle), we can flatter the shuffles
out if they become an identity mask. This can come up during lane
interleaving, when we do that better.
Differential Revision: https://reviews.llvm.org/D94034
Some benchmarks regress with non-trivial unswitching, so add an option
to opt-out of performing non-trivial unswitching while investigating.
Reviewed By: asbirlea
Differential Revision: https://reviews.llvm.org/D95796
This is consistent with the VEX version. It also fixes a sorting
issue in the matching table that caused the EVEX version to be
prioritized over VEX in intel syntax.
Fixes issue [2] from PR48991.
The constant trunc/ext may not be the optimal pre-condition,
but I think that handles the common cases.
Example of Alive2 proof:
https://alive2.llvm.org/ce/z/sREeLC
This is another step towards canonicalizing to the intrinsics.
Narrowing was identified as source of potential regression for
abs(), so we need to handle this for min/max - see:
https://llvm.org/PR48816
If this is not enough, we could process intrinsics in
the trunc-driven matching in canEvaluateTruncated().
Some libcs define non-standard FE_* macros and include them in
FE_ALL_EXCEPT. This change adjusts the fenv tests so that the
non-standard FE_* macros do not interfere when compiled with
fenv.h from another libc.
Reviewed By: sivachandra
Differential Revision: https://reviews.llvm.org/D95650
This patch is a follow up to D94821 to ensure the correct behavior of the
general directive structure checker.
This patch add the generation of the Enter function declaration for clauses in
the TableGen backend.
This helps to ensure each clauses declared in the TableGen file has at least
a basic check.
Reviewed By: kiranchandramohan
Differential Revision: https://reviews.llvm.org/D95108
FileCheck is required for OpenMP tests. The current detection can fail
if building OpenMP in-tree when user sets `LLVM_INSTALL_TOOLCHAIN_ONLY=ON`. As a
result, CMake will raise an error and the compilation will be broken. This patch
fixed the issue. When `FileCheck` is not a target, tests will just be skipped.
Reviewed By: jdoerfert, JonChesterfield
Differential Revision: https://reviews.llvm.org/D95689
Implements:
- LWG3149 DefaultConstructible should require default initialization
Implements parts of:
- P0898R3 Standard Library Concepts
- P1754 Rename concepts to standard_case for C++20, while we still can
Depends on D91986
Reviewed By: ldionne, #libc
Differential Revision: https://reviews.llvm.org/D93461
When we have a zeroext parameter, emit G_ASSERT_ZEXT.
Add a check that we actually emit it.
This is a 0.1% code size win on CTMark/7zip and CTMark/consumer-typeset at -Os.
Differential Revision: https://reviews.llvm.org/D95567
Part of the gold test added in 1487747e99
relies on more recent fixes to gold that fix the plugin behavior with
--export-dynamic-symbol and --dynamic-list. Extract those parts of the
new test into a v1.16 test.
A follow up patch will add support for commuting operands or
changing opcode to vfmacc and friends.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D95662
When replacing the dst reg with the src reg, we need to make sure that we
propagate the dst reg's register class through to the src.
Otherwise, we aren't meeting the requirements for G_ASSERT_ZEXT, and so the
verifier will fail.
Differential Revision: https://reviews.llvm.org/D95708
Rather than materializing the 0xffff immediate for the AND, use
a shift left to remove the upper bits and then shift in zeros
from the right.
This pattern occurs when type legalizing an i16 right shift.
I've implemented this with custom selection code for a number of
reasons. I've limited this to the AND having a single use. We need
to compensate for SimplifyDemandedBits altering the AND mask. I'm
using *W opcodes on RV64. We may want to generlize this in the
future. For all these reason it seemed easiest to do it this way.
Reviewed By: luismarques
Differential Revision: https://reviews.llvm.org/D95774
Instead of using ConstraintSystem::negate when adding new constraints,
flip the condition in IR.
The main advantage is that EQ predicates can be represented by 2
constraints, which makes negating based on the constraint tricky. The IR
condition can easily negated.
The parsing of I/O units uses look-ahead to discriminate between
keywords, variables and expressions as part of distinguishing internal
from external I/O. The look-ahead was inaccurate for variables that
appear as the initial parts of expressions.
Differential Revision: https://reviews.llvm.org/D95743
This would assert with amdgpu-spill-sgpr-to-vgpr disabled when trying to
spill the FP.
Fixes: SWDEV-262704
Reviewed By: RamNalamothu
Differential Revision: https://reviews.llvm.org/D95768
This fixes an apparent oversight in D91156, where the symbol was defined
without the leading underscore, then the visibility was later declared with it.
rdar://73364185
Differential revision: https://reviews.llvm.org/D95639
Summary:
One option for the LIBOMPTARGET_INFO environment variable is to print the current status of the device's data mappings. These are a shared resource among threads so this needs to be protected when using multiple streams.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D95786
This requires a second index query for refs to overrides, as the refs
call doesn't tell you which ref points at which symbol.
Differential Revision: https://reviews.llvm.org/D95451
Under the softfp calling convention, we are often left with
VMOVRRD(extract(bitcast(build_vector(a, b, c, d)))) for the return value
of the function. These can be simplified to a,b or c,d directly,
depending on the value of the extract.
Big endian is a little different because the bitcast switches the lanes
around, meaning we end up with b,a or d,c.
Differential Revision: https://reviews.llvm.org/D94989
Unfortunately this treats overrides declarations as declarations, not as
references. I don't plan to land this until I have a fix for that issue.
Differential Revision: https://reviews.llvm.org/D95450
If we instantiate self-referenced anonymous records in foreach and
multiclass, the NAME value will point to incorrect record. It's because
anonymous name is resolved too early.
This patch adds AnonymousNameInit to represent an anonymous record name.
When instantiating an anonymous record, it will update the referred name.
Differential Revision: https://reviews.llvm.org/D95309
Add deleted volatile copy-assignment operator in the most derived atomic
to fix the Bug 41784. The root cause: there is an `operator=(T) volatile`
that has better match than the deleted copy-assignment operator of the base
class when `this` is `volatile`. The compiler sees that right operand of
the assignment operator can be converted to `T` and chooses that path
without taking into account the deleted copy-assignment operator of the
base class.
The current behavior on libstdc++ is different from what we have in libc++.
On the same test compilation fails with libstdc++. Proof: https://godbolt.org/z/nebPYd
(everything is the same except the -stdlib option).
I choose the way with explicit definition of copy-assignment for atomic
in the most derived class. But probably we can fix that by moving
`operator=(T)` overloads to the base class from both specializations.
At first glance, it shouldn't break anything.
Differential Revision: https://reviews.llvm.org/D90968
The AArch64 DAG combine added by D90945 & D91433 extends the index
of a scalable masked gather or scatter to i32 if necessary.
This patch removes the combine and instead adds shouldExtendGSIndex, which
is used by visitMaskedGather/Scatter in SelectionDAGBuilder to query whether
the index should be extended before calling getMaskedGather/Scatter.
Reviewed By: david-arm
Differential Revision: https://reviews.llvm.org/D94525
This patch skips the test for the SBTarget::IsLoaded method on windows
since the logic is different.
Differential Revision: https://reviews.llvm.org/D95686
Signed-off-by: Med Ismail Bennani <medismail.bennani@gmail.com>
D90687 introduced a crash:
llvm::LoopVectorizationCostModel::computeMaxVF(llvm::ElementCount, unsigned int):
Assertion `WideningDecisions.empty() && Uniforms.empty() && Scalars.empty() &&
"No decisions should have been taken at this point"' failed.
when compiling the following C code:
typedef struct {
char a;
} b;
b *c;
int d, e;
int f() {
int g = 0;
for (; d; d++) {
e = 0;
for (; e < c[d].a; e++)
g++;
}
return g;
}
with:
clang -Os -target hexagon -mhvx -fvectorize -mv67 testcase.c -S -o -
This occurred since prior to D90687 computeFeasibleMaxVF would only be
called in computeMaxVF when a scalar epilogue was allowed, but now it's
always called. This causes the assert above since computeFeasibleMaxVF
collects all viable VFs larger than the default MaxVF, and for each VF
calculates the register usage which results in analysis being done the
assert above guards against. This can occur in computeFeasibleMaxVF if
TTI.shouldMaximizeVectorBandwidth and this target hook is implemented in
the hexagon backend to always return true.
Reported by @iajbar.
Reviewed By: fhahn
Differential Revision: https://reviews.llvm.org/D94869
Wouter van Oortmerssen