As pointed out in post-commit review, this can legally be called
on instructions that are not inserted into basic blocks,
so don't blindly assume that there is basic block.
Parameters were in a different order in the header and in the implementation.
Fix surrounding comments a bit.
Differential Revision: https://reviews.llvm.org/D86691
llvm::isa<>() and llvm::isa_and_not_null<>() template functions recently became
variadic. Unfortunately this causes crashes in case of isa_and_not_null<>()
and incorrect behavior in isa<>(). This patch fixes this issue.
Differential Revision: https://reviews.llvm.org/D85728
The successfulness of a dynamic cast depends only on the C++ class, not the pointer or reference. Thus if *A is a *B, then &A is a &B,
const *A is a const *B etc. This patch changes DynamicCastInfo to store
and check the cast between the unqualified pointed/referenced types.
It also removes e.g. SubstTemplateTypeParmType from both the pointer
and the pointed type.
Differential Revision: https://reviews.llvm.org/D85752
value-profile-load.test needs adjustment with a mutator change in
bb54bcf849, which reverted as of now, but will be
recommitted after landing this patch.
This patch makes value-profile-load.test more friendly to (and aware of) the
current value profiling strategy, which is based on the hamming as well as the
absolute distance. To this end, this patch adjusts the set of input values that
trigger an expected crash. More specifically, this patch now uses a single value
0x01effffe as a crashing input, because this value is close to values like
{0x1ffffff, 0xffffff, ...}, which are very likely to be added to the corpus per
the current hamming- and absolute-distance-based value profiling strategy. Note
that previously the crashing input values were {1234567 * {1, 2, ...}, s.t. <
INT_MAX}.
Every byte in the chosen value 0x01effeef is intentionally different; this was
to make it harder to find the value without the intermediate inputs added to the
corpus by the value profiling strategy.
Also note that LoadTest.cpp now uses a narrower condition (Size != 8) for
initial pruning of inputs, effectively preventing libFuzzer from generating
inputs longer than necessary and spending time on mutating such long inputs in
the corpus - a functionality not meant to be tested by this specific test.
Differential Revision: https://reviews.llvm.org/D86247
If we query an AA with `Attributor::getAAFor` in `AbstractAttribute::manifest`, the AA may be updated.
This patch makes use of the phase flag in Attributor, and handle `getAAFor` behavior according to the flag.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86635
The prior diff that introduced `addAffineIfOpDomain` missed appending
constraints from the ifOp domain. This revision fixes this problem.
Differential Revision: https://reviews.llvm.org/D86421
Many OpenMP Clang tests do not RUN for version 4.5 and the default
version. This second patch in the series handles test cases which
require updation in CHECK lines along with adding RUN lines for
the default version. It involves updating line number of pragmas.
Reviewed By: ABataev
Differential Revision: https://reviews.llvm.org/D85150
Adding a conversion pattern for the parallel Operation. This will
help the conversion of parallel operation with standard dialect to
parallel operation with llvm dialect. The type conversion of the block
arguments in a parallel region are controlled by the pattern for the
parallel Operation. Without this pattern, a parallel Operation with
block arguments cannot be converted from standard to LLVM dialect.
Other OpenMP operations without regions are marked as legal. When
translation of OpenMP operations with regions are added then patterns
for these operations can also be added.
Also uses all the standard to llvm patterns. Patterns of other dialects
can be added later if needed.
Reviewed By: rriddle
Differential Revision: https://reviews.llvm.org/D86273
It seems to be failing on some Google Buildbots.
This diff also includes a minor fix for the install name of one of
libSystem's re-exports. I don't think it's the cause of the test
failure, though. The wrong install name just meant that the symbol
lookup failure would still happen, but it would have been caused by the
re-export not being found, instead of the arch failing to match.
Differential Revision: https://reviews.llvm.org/D86728
This fixes a mismatched visibility attribute on the call operator in
addition to making the code clearer. Given this is a simple lambda
in essence, the intent has always been to give it inline visibility.
This patch adjusts the following ARM/AArch64 LLVM IR intrinsics:
- neon_bfmmla
- neon_bfmlalb
- neon_bfmlalt
so that they take and return bf16 and float types. Previously these
intrinsics used <8 x i8> and <4 x i8> vectors (a rudiment from
implementation lacking bf16 IR type).
The neon_vbfdot[q] intrinsics are adjusted similarly. This change
required some additional selection patterns for vbfdot itself and
also for vector shuffles (in a previous patch) because of SelectionDAG
transformations kicking in and mangling the original code.
This patch makes the generated IR cleaner (less useless bitcasts are
produced), but it does not affect the final assembly.
Reviewed By: dmgreen
Differential Revision: https://reviews.llvm.org/D86146
Original Commit Message:
After the commit r368987 (rG643adb55769e) was landed, the frame record (FP and LR register)
may be placed in the middle of a stack frame if a function has both callee-saved
general-purpose registers and floating point registers. This will break the stack unwinders
that simply walk through the frame records (based on the guarantee from AAPCS64
"The Frame Pointer" section). This commit fixes the problem by adding the frame record offset.
Patch By: logan
Differential Revision: D70800
The NPM doesn't support call-site alwaysinline as described in the comments.
Also make NPM runs more similar to legacy PM runs.
Reviewed By: ychen, asbirlea
Differential Revision: https://reviews.llvm.org/D86663
https://reviews.llvm.org/D83833
Patch adds two new GICombinerRules for G_SELECT. The rules include:
combining selects with undef comparisons into their first selectee value,
and to combine away selects with constant comparisons. Patch additionally
adds a new combiner test for the AArch64 target to test these new G_SELECT
combiner rules and the existing select_same_val combiner rule.
Patch by mkitzan
Always make lldb-argdumper a dependency of liblldb. Currently it is only
a dependency of the python swig target because of the relative symlink
in the python resource directory. That means that the dependency won't
be there when LLDB_ENABLE_PYTHON is disabled.
Differential revision: https://reviews.llvm.org/D86722
Add a new flag that indicates which stage in the process we are in.
This flag is introduced for handling behavior of `getAAFor` according to the stage. (discussed in D86635)
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D86678
https://reviews.llvm.org/D86676
Sometimes we can have the following code
x:gpr(s32) = G_OP
Say we build G_OP2 to the same x and then delete the previous instruction. Using something like
Register X = ...;
auto NewMIB = CSEBuilder.buildOp2(X, ... args);
Currently there's a mismatch in how NewMIB is profiled and inserted into the CSEMap (ie it doesn't consider register bank/register class along with type).Unify the profiling by refactoring and calling the common method.
This was found by turning on the CSEInfo::verify in at the end of each of our GISel passes which turns inconsistent state/non determinism in CSEing into crashes which likely usually indicates missing calls to Observer on mutations (the most common case). Here non determinism usually means not cseing sometimes, but almost never about producing incorrect code.
Also this patch adds this verification at the end of the combiners as well.
When joining the legal parts of vector arguments into its original value
during the lower of Formal Arguments in SelectionDAGBuilder, the Calling
Convention information was not being propagated for the handling of each
individual parts. The same did not happen when lowering calls, causing a
mismatch.
This patch fixes the issue by properly propagating the Calling
Convention details.
This fixes Bugzilla #47001.
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D86715
See RFC for background:
http://lists.llvm.org/pipermail/llvm-dev/2020-June/142744.html
Note that the runtime changes will be sent separately (hopefully this
week, need to add some tests).
This patch includes the LLVM pass to instrument memory accesses with
either inline sequences to increment the access count in the shadow
location, or alternatively to call into the runtime. It also changes
calls to memset/memcpy/memmove to the equivalent runtime version.
The pass is modeled on the address sanitizer pass.
The clang changes add the driver option to invoke the new pass, and to
link with the upcoming heap profiling runtime libraries.
Currently there is no attempt to optimize the instrumentation, e.g. to
aggregate updates to the same memory allocation. That will be
implemented as follow on work.
Differential Revision: https://reviews.llvm.org/D85948
Apparently, we don't do this, neither in EarlyCSE, nor in InstSimplify,
nor in (old) GVN, but do in NewGVN and SimplifyCFG of all places..
While i could teach EarlyCSE how to hash PHI nodes,
we can't really do much (anything?) even if we find two identical
PHI nodes in different basic blocks, same-BB case is the interesting one,
and if we teach InstSimplify about it (which is what i wanted originally,
https://reviews.llvm.org/D86530), we get EarlyCSE support for free.
So i would think this is pretty uncontroversial.
On vanilla llvm test-suite + RawSpeed, this has the following effects:
```
| statistic name | baseline | proposed | Δ | % | \|%\| |
|----------------------------------------------------|-----------|-----------|-------:|---------:|---------:|
| instsimplify.NumPHICSE | 0 | 23779 | 23779 | 0.00% | 0.00% |
| asm-printer.EmittedInsts | 7942328 | 7942392 | 64 | 0.00% | 0.00% |
| assembler.ObjectBytes | 273069192 | 273084704 | 15512 | 0.01% | 0.01% |
| correlated-value-propagation.NumPhis | 18412 | 18539 | 127 | 0.69% | 0.69% |
| early-cse.NumCSE | 2183283 | 2183227 | -56 | 0.00% | 0.00% |
| early-cse.NumSimplify | 550105 | 542090 | -8015 | -1.46% | 1.46% |
| instcombine.NumAggregateReconstructionsSimplified | 73 | 4506 | 4433 | 6072.60% | 6072.60% |
| instcombine.NumCombined | 3640264 | 3664769 | 24505 | 0.67% | 0.67% |
| instcombine.NumDeadInst | 1778193 | 1783183 | 4990 | 0.28% | 0.28% |
| instcount.NumCallInst | 1758401 | 1758799 | 398 | 0.02% | 0.02% |
| instcount.NumInvokeInst | 59478 | 59502 | 24 | 0.04% | 0.04% |
| instcount.NumPHIInst | 330557 | 330533 | -24 | -0.01% | 0.01% |
| instcount.TotalInsts | 8831952 | 8832286 | 334 | 0.00% | 0.00% |
| simplifycfg.NumInvokes | 4300 | 4410 | 110 | 2.56% | 2.56% |
| simplifycfg.NumSimpl | 1019808 | 999607 | -20201 | -1.98% | 1.98% |
```
I.e. it fires ~24k times, causes +110 (+2.56%) more `invoke` -> `call`
transforms, and counter-intuitively results in *more* instructions total.
That being said, the PHI count doesn't decrease that much,
and looking at some examples, it seems at least some of them
were previously getting PHI CSE'd in SimplifyCFG of all places..
I'm adjusting `Instruction::isIdenticalToWhenDefined()` at the same time.
As a comment in `InstCombinerImpl::visitPHINode()` already stated,
there are no guarantees on the ordering of the operands of a PHI node,
so if we just naively compare them, we may false-negatively say that
the nodes are not equal when the only difference is operand order,
which is especially important since the fold is in InstSimplify,
so we can't rely on InstCombine sorting them beforehand.
Fixing this for the general case is costly (geomean +0.02%),
and does not appear to catch anything in test-suite, but for
the same-BB case, it's trivial, so let's fix at least that.
As per http://llvm-compile-time-tracker.com/compare.php?from=04879086b44348cad600a0a1ccbe1f7776cc3cf9&to=82bdedb888b945df1e9f130dd3ac4dd3c96e2925&stat=instructions
this appears to cause geomean +0.03% compile time increase (regression),
but geomean -0.01%..-0.04% code size decrease (improvement).
Fix compilation with -DLIBCXX_BUILD_EXTERNAL_THREAD_LIBRARY when using clang. Now linking target 'cxx_external_threads' with 'cxx-headers'. Fix mismatching visibility for `libcpp_timed_backoff_policy` function in file <__threading_support>.
Reviewed By: #libc, ldionne
Differential Revision: https://reviews.llvm.org/D86598
This patch implements codegen for the 'arm_sve_vector_bits' type
attribute, defined by the Arm C Language Extensions (ACLE) for SVE [1].
The purpose of this attribute is to define vector-length-specific (VLS)
versions of existing vector-length-agnostic (VLA) types.
VLSTs are represented as VectorType in the AST and fixed-length vectors
in the IR everywhere except in function args/return. Implemented in this
patch is codegen support for the following:
* Implicit casting between VLA <-> VLS types.
* Coercion of VLS types in function args/return.
* Mangling of VLS types.
Casting is handled by the CK_BitCast operation, which has been extended
to support the two new vector kinds for fixed-length SVE predicate and
data vectors, where the cast is implemented through memory rather than a
bitcast which is unsupported. Implementing this as a normal bitcast
would require relaxing checks in LLVM to allow bitcasting between
scalable and fixed types. Another option was adding target-specific
intrinsics, although codegen support would need to be added for these
intrinsics. Given this, casting through memory seemed like the best
approach as it's supported today and existing optimisations may remove
unnecessary loads/stores, although there is room for improvement here.
Coercion of VLSTs in function args/return from fixed to scalable is
implemented through the AArch64 ABI in TargetInfo.
The VLA and VLS types are defined by the ACLE to map to the same
machine-level SVE vectors. VLS types are mangled in the same way as:
__SVE_VLS<typename, unsigned>
where the first argument is the underlying variable-length type and the
second argument is the SVE vector length in bits. For example:
#if __ARM_FEATURE_SVE_BITS==512
// Mangled as 9__SVE_VLSIu11__SVInt32_tLj512EE
typedef svint32_t vec __attribute__((arm_sve_vector_bits(512)));
// Mangled as 9__SVE_VLSIu10__SVBool_tLj512EE
typedef svbool_t pred __attribute__((arm_sve_vector_bits(512)));
#endif
The latest ACLE specification (00bet5) does not contain details of this
mangling scheme, it will be specified in the next revision. The
mangling scheme is otherwise defined in the appendices to the Procedure
Call Standard for the Arm Architecture, see [2] for more information.
[1] https://developer.arm.com/documentation/100987/latest
[2] https://github.com/ARM-software/abi-aa/blob/master/aapcs64/aapcs64.rst#appendix-c-mangling
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D85743
This patch optionally replaces the CRT allocator (i.e., malloc and free) with rpmalloc (mixed public domain licence/MIT licence) or snmalloc (MIT licence) or mimalloc (MIT licence). Please note that the source code for these allocators must be available outside of LLVM's tree.
To enable, use `cmake ... -DLLVM_INTEGRATED_CRT_ALLOC=D:/git/rpmalloc -DLLVM_USE_CRT_RELEASE=MT` where `D:/git/rpmalloc` has already been git clone'd from `https://github.com/mjansson/rpmalloc`. The same applies to snmalloc and mimalloc.
When enabled, the allocator will be embeded (statically linked) into the LLVM tools & libraries. This currently only works with the static CRT (/MT), although using the dynamic CRT (/MD) could potentially work as well in the future.
When enabled, this changes the memory stack from:
new/delete -> MS VC++ CRT malloc/free -> HeapAlloc -> VirtualAlloc
to:
new/delete -> {rpmalloc|snmalloc|mimalloc} -> VirtualAlloc
The goal of this patch is to bypass the application's global heap - which is thread-safe thus inducing locking - and instead take advantage of a modern lock-free, thread cache, allocator. On a 6-core Xeon Skylake we observe a 2.5x decrease in execution time when linking a large scale application with LLD and ThinLTO (12 min 20 sec -> 5 min 34 sec), when all hardware threads are being used (using LLD's flag /opt:lldltojobs=all). On a dual 36-core Xeon Skylake with all hardware threads used, we observe a 24x decrease in execution time (1 h 2 min -> 2 min 38 sec) when linking a large application with LLD and ThinLTO. Clang build times also see a decrease in the range 5-10% depending on the configuration.
Differential Revision: https://reviews.llvm.org/D71786
When dealing with dialects that will results in function calls to
external libraries, it is important to be able to handle maps as some
dialects may require mapped data. Before this patch, the detection of
whether normalization can apply or not, operations are compared to an
explicit list of operations (`alloc`, `dealloc`, `return`) or to the
presence of specific operation interfaces (`AffineReadOpInterface`,
`AffineWriteOpInterface`, `AffineDMAStartOp`, or `AffineDMAWaitOp`).
This patch add a trait, `MemRefsNormalizable` to determine if an
operation can have its `memrefs` normalized.
This trait can be used in turn by dialects to assert that such
operations are compatible with normalization of `memrefs` with
nontrivial memory layout specification. An example is given in the
literal tests.
Differential Revision: https://reviews.llvm.org/D86236
Roman Lebedev