Normally, given that the DA results are kept consistent over the selection DAG, uniform comparisons get selected to S_CMP_* but divergent to V_CMP_*. Sometimes, for the sake of efficiency, SSA subgraphs may be converted to VALU to avoid repeatedly copying data back and forth. Hence we have to be able to sustain the correctness passing the i1 from VALU to SALU context and vice versa.
VALU operations only process the active lanes of the VGPR and ignore inactive ones.
Active lanes correspond to 1 bit in the EXEC mask register.
SALU represents i1 as just one bit but VALU as 64bits: 0/1 and 0/(0xffffffffffffffff & EXEC) respectively.
SALU uses one-bit conditional flag SCC but VALU - VCC that is a pair of 32-bit SGPRs
To expose SCC to the VALU context we need to convert the one-bit boolean value to the appropriate 64bit.
To return back to the SALU context we need to do the opposite.
To correctly convert 64bit VALU boolean to either 0 or 1 we need to filter out the bits corresponding to the inactive lanes.
Reviewed By: piotr
Differential Revision: https://reviews.llvm.org/D109900
When adding alias.scope and noalias metadata to a memcpy function,
the alias.scope and noalias metadata from the operands are merged.
The rule for merging alias.scope is to take the intersection of
the domains and the union of the scopes within those domains.
The rule for merging noalias is to take the intersection.
The bug is that AMDGPULowerModuleLDS was using concatenation for
both alias.scope and noalias. For example, when f1 and f2 are added
to the LDS structure and there is a memcpy(f2, f1, sizeof(f1)).
Then, concatenation creates noalias metadata for the memcpy that
includes both {f1, f2}. That means that the memcpy is assumed
not to alias a prior load of f2, which enables the optimizer to
remove a load of f2 that occurs after mempcy.
The function MDNode::getmostGenericAliasScope defines the semantics
for alias.scope. There is a function, combineMetadata in Local.cpp,
that uses intersect for noalias.
Differential Revision: https://reviews.llvm.org/D110049
Parallel regions are outlined as functions with capture variables explicitly generated as distinct parameters in the function's argument list. That complicates the fork_call interface in the OpenMP runtime: (1) the fork_call is variadic since there is a variable number of arguments to forward to the outlined function, (2) wrapping/unwrapping arguments happens in the OpenMP runtime, which is sub-optimal, has been a source of ABI bugs, and has a hardcoded limit (16) in the number of arguments, (3) forwarded arguments must cast to pointer types, which complicates debugging. This patch avoids those issues by aggregating captured arguments in a struct to pass to the fork_call.
Reviewed By: jdoerfert, jhuber6
Differential Revision: https://reviews.llvm.org/D102107
This patch adds a prefixed load pattern involving v2f32 fpext v2f64, where we
are dealing with a value with an offset that fits into a 34-bit signed immediate.
A reduced test case is also added to patch that tests the pattern, in which the
pattern is tested in the big endian CHECKs of the newly added test.
Differential Revision: https://reviews.llvm.org/D109887
This patch fixes the crash found by PR51614:
whenever doing tail folding, interleave groups must be considered under mask.
Another fix D108900 follows for targets that support masked loads and stores:
when *deciding* to vectorize with masked interleave groups, check if the access
is reverse - which is currently not supported; rather than (only) asserting when
computing cost and generating code.
Differential Revision: https://reviews.llvm.org/D108891
This requires a minor change to CodeGenPrepare to ensure that
shouldSinkOperands will be called for And.
Reviewed By: frasercrmck
Differential Revision: https://reviews.llvm.org/D110106
Trying to use the DIA SDK reader only to fail with "DIA SDK wasn't enabled"
isn't very useful. The native PDB reader is missing some stuff, but it's still
better than nothing.
Reduces number of lldb-check-shell test failures with LLVM_ENABLE_DIA_SDK=NO
from 27 to 15.
Differential Revision: https://reviews.llvm.org/D110172
This implements the generic std.format.spec framework for all types.
The Unicode support will be added in a separate patch.
Implements parts of:
- P0645 Text Formatting
Completes:
- LWG-3242 std::format: missing rules for arg-id in width and precision
- P1892 Extended locale-specific presentation specifiers for std::format
Reviewed By: #libc, ldionne, vitaut
Differential Revision: https://reviews.llvm.org/D103368
We already have pow(x, y) * pow(x, z) -> pow(x, y + z) transformation, but we are missing same transformation for powi (power is integer).
Requires reassoc.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D109954
This patch changes the signature of the load and store vector pair
builtins to match their documentation. The type of the `signed long long`
argument is changed to `signed long`. This patch also changes existing testcases
to match the signature change.
Reviewed By: lei, Conanap
Differential Revision: https://reviews.llvm.org/D109996
The current code writes the pc-table at the process startup,
which may happen before the common_flags() are initialized.
Move writing to the process end.
This is consistent with how we write the counters and avoids the problem with the uninitalized flags.
Add prints if verbosity>=1.
Reviewed By: kostik
Differential Revision: https://reviews.llvm.org/D110119
isValidAssumeForContext can provide better results with access to the
dominator tree in some cases. This patch adjusts computeConstantRange to
allow passing through a dominator tree.
The use VectorCombine is updated to pass through the DT to enable
additional scalarization.
Note that similar APIs like computeKnownBits already accept optional dominator
tree arguments.
Reviewed By: lebedev.ri
Differential Revision: https://reviews.llvm.org/D110175
When using instructions which have a MetadataAsValue argument
(e.g. some target-specific intrinsics) MD canonicalization strips
internal MDNodes with a single ConstantAsMetadata child. That
prevented IRTranslator from the proper translation of such a calls.
Compute the tiled producer slice dimensions directly starting from the consumer not using the producer at all.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D110147
Add a helper method to check if an index vector contains a permutation of its indices. Additionally, refactor applyPermutationToVector to take int64_t.
Reviewed By: nicolasvasilache
Differential Revision: https://reviews.llvm.org/D110135
It was previously assumed that tensor.insert_slice should be bufferized first in a greedy fashion to avoid out-of-place bufferization of the large tensor. This heuristic does not hold upon further inspection.
This CL removes the special handling of such ops and adds a test that exhibits better behavior and appears in real use cases.
The only test adversely affected is an artificial test which results in a returned memref: this pattern is not allowed by comprehensive bufferization in real scenarios anyway and the offending test is deleted.
Differential Revision: https://reviews.llvm.org/D110072
Previously, comprehensive bufferize would consider all aliasing reads and writes to
the result buffer and matching operand. This resulted in spurious dependences
being considered and resulted in too many unnecessary copies.
Instead, this revision revisits the gathering of read and write alias sets.
This results in fewer alloc and copies.
An exhaustive test cases is added that considers all possible permutations of
`matmul(extract_slice(fill), extract_slice(fill), ...)`.
Optimize (add (mul x, c0), c1) -> (ADDI (MUL (ADDI, c1/c0), c0), c1%c0),
if c1/c0 and c1%c0 are simm12, while c1 is not.
Optimize (add (mul x, c0), c1) -> (MUL (ADDI, c1/c0), c0),
if c1%c0 is zero, and c1/c0 is simm12 while c1 is not.
Reviewed By: craig.topper
Differential Revision: https://reviews.llvm.org/D108607
Helper function `getDefaultOpenCLPointeeAddrSpace()` introduced to
`ASTContext` class. It returns default OpenCL address space
depending on language version and enabled features. If generic
address space is supported, the helper function returns value
`LangAS::opencl_generic`. Otherwise, value `LangAS::opencl_private`
is returned. Code refactoring changes performed in several suitable
places.
Differential Revision: https://reviews.llvm.org/D109874
This patch allows sinking an instruction which can have multiple uses in a
single user. We were previously over-restrictive by looking for exactly one use,
rather than one user.
Also added an API for retrieving a unique undroppable user.
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D109700
Add documentation of unbundling of heterogeneous device archives to
create device specific archives, as introduced by D93525. Also, add
documentation for supported text file formats.
Reviewed By: yaxunl
Differential Revision: https://reviews.llvm.org/D110083
The updated list is based on the output of
cmake -G Ninja -S llvm -B build -DLLVM_ENABLE_PROJECTS='foo'.
Differential Revision: https://reviews.llvm.org/D110124
One of the two inputs of the Shufflevector is often a placeholder.
Previously, there were cases where the placeholder was undef, and there were cases where it was poison.
I added these constructors to create a placeholder consistently.
Changing to use the newly added constructor will be written in a separate patch.
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D110146
SystemZ adds the EXRL target instructions in the end of each file. This must
be done before debug info emission since that may end the text section, and
therefore this is now done in emitConstantPools() (instead of in
emitEndOfAsmFile).
Review: Ulrich Weigand
Differential Revision: https://reviews.llvm.org/D109513
Currently we set thr->tctx after OnStarted callback
taking thread registry mutex again and searching for the context.
But OnStarted already runs under the thread registry mutex
and has access to the context, so set it in the OnStarted.
This makes code simpler and faster.
Depends on D110132.
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D110133
Thread state functions are split into 2 parts:
tsan entry function (e.g. ThreadStart) and thread registry
state change callback (e.g. OnStart). Currently these
pairs of functions are located far from each other and
in reverse order. This makes it hard to read and follow the logic.
Reorder the code so that OnFoo directly follows ThreadFoo.
No other code changes.
Reviewed By: melver
Differential Revision: https://reviews.llvm.org/D110132
alex-t