Reapply the patch with getObjectSize() replaced by getAllocSize().
The former will also look through calls that return their argument,
and we'll end up placing dereferenceable attributes on intrinsics
like llvm.launder.invariant.group. While this isn't wrong, it also
doesn't seem to be particularly useful. For now, use getAllocSize()
instead, which sticks closer to the original behavior of this code.
-----
This code is just interested in the allocsize, not any other
allocator properties.
Default getAllocSize() to use the trivial mapper. Also switch
from using std::function to function_ref.
Furthermore, update the doc comment to point out a subtle difference
between getAllocSize() and getObjectSize(): The latter may also
return something for calls that return their argument (via "returned"
attribute or special intrinsics like invariant groups).
Fetching allocation data for calls is a rather hot operation, and
TLI lookups are slow. We can greatly reduce the number of calls
for which TLI is queried by checking that they return a pointer
value first, as this is a requirement for allocation functions
anyway.
Currently, clang may meet an infinite loop in a very tricky case when it
iterates the default args. This patch tries to fix this by adding a
`fixed` check.
Tests that use `--mlir-pass-statistics-display=` from MLIR require the
following condition to hold: (extracted from LLVM's Statistics.h):
```
#define LLVM_ENABLE_STATS 1
```
This is normally enforced with `REQUIRES: asserts`. This patch updates
relevant Flang tests accordingly.
For "Release" builds (with assertions disabled), the affected tests will
be failing without this change.
Differential Revision: https://reviews.llvm.org/D130185
Shape can be memref of index type, so memref::LoadOp result need to be converted into llvm type.
Differential Revision: https://reviews.llvm.org/D129965
These headers used to be guarded only on PowerPC64 Linux or FreeBSD, but
they can also be enabled for AIX OS target since it's big-endian ready.
Reviewed By: shchenz
Differential Revision: https://reviews.llvm.org/D129461
re-land fixes an unwanted interaction with module-map modules, seen in
Greendragon testing.
This provides updates to
[class.mfct]:
Pre C++20 [class.mfct]p2:
A member function may be defined (8.4) in its class definition, in
which case it is an inline member function (7.1.2)
Post C++20 [class.mfct]p1:
If a member function is attached to the global module and is defined
in its class definition, it is inline.
and
[class.friend]:
Pre-C++20 [class.friend]p5
A function can be defined in a friend declaration of a
class . . . . Such a function is implicitly inline.
Post C++20 [class.friend]p7
Such a function is implicitly an inline function if it is attached
to the global module.
We add the output of implicit-inline to the TextNodeDumper, and amend
a couple of existing tests to account for this, plus add tests for the
cases covered above.
Differential Revision: https://reviews.llvm.org/D129045
Replace iterators of the outermost loop with region arguments of the innermost
one. The changes avoid later `bufferization` passes to insert allocation within
the body of the innermost loop.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D130083
This commit reduces the size of the emitted rebase sections by
generating the REBASE_OPCODE_DO_REBASE_ADD_ADDR_ULEB and
REBASE_OPCODE_DO_REBASE_ULEB_TIMES_SKIPPING_ULEB opcodes.
With this change, chromium_framework's rebase section is a 40% smaller
197 kilobytes, down from the previous 320 kB. That is 6 kB smaller than
what ld64 produces for the same input.
Performance figures from my M1 Mac mini:
x before
+ after
N Min Max Median Avg Stddev
x 10 4.2269349 4.3300061 4.2689675 4.2690016 0.031151669
+ 10 4.219331 4.2914009 4.2398136 4.2448277 0.023817308
No difference proven at 95.0% confidence
Differential Revision: https://reviews.llvm.org/D130180
- Test cases for arch only has 16-bit instruction such as ck801/ck802 need
compile with -mattr=+btst16
- Fix the GPR copy instruction with MOV16 for 16-bit only arch.
This patch, on top of D120048 <https://reviews.llvm.org/D120048>, supports
GetTls on Solaris 11.3 and Illumos that lack `dlpi_tls_modid`. It's the
same method originally used in D91605 <https://reviews.llvm.org/D91605>,
but integrated into `GetStaticTlsBoundary`.
Tested on `amd64-pc-solaris2.11`, `sparcv9-sun-solaris2.11`, and
`x86_64-pc-linux-gnu`.
Differential Revision: https://reviews.llvm.org/D120059
Vector fptosi_sat and fptoui_sat were being expanded by unrolling the
vector operation. This doesn't work for scalable vector, so this patch
adds a call to TLI.expandFP_TO_INT_SAT if the vector is scalable.
Scalable tests are added for AArch64 and RISCV. Some of the AArch64
fptoi_sat operations should be legal, but that will be handled in
another patch.
Differential Revision: https://reviews.llvm.org/D130028
Replace iterators of the outermost loop with region arguments of the innermost
one. The changes avoid later `bufferization` passes to insert allocation within
the body of the innermost loop.
Reviewed By: mravishankar
Differential Revision: https://reviews.llvm.org/D130083
There is a problem in loop cache analysis that the types of SCEV variables
`Coeff` and `ElemSize` in function `isConsecutive()` may not match. The
mismatch would cause SCEV failures when `Coeff` is multiplied with `ElemSize`.
The fix in this patch is to extend the type of both `Coeff` and `ElemSize` to
whichever is wider in those two variables. As a clean-up, duplicate calculations
of `Stride` in `computeRefCost()` is then removed.
Reviewed By: Meinersbur, #loopoptwg
Differential Revision: https://reviews.llvm.org/D128877
This patch makes use of OMPIRBuilder support for codegen of taskgroup
construct in clang.
Depends on D128203
Reviewed By: Meinersbur
Differential Revision: https://reviews.llvm.org/D129992
(srl (and X, 1<<C), C) is the form we receive for testing bit C.
An earlier combine removed the setcc so it wasn't there to match
when we created the SELECT_CC. This doesn't happen for BR_CC because
generic DAG combine rebuilds the setcc if it is used by BRCOND.
We can shift X left by XLen-1-C to put the bit to be tested in the
MSB, and use a signed compare with 0 to test the MSB.
sanitizer_platform_limits_posix.h defines `__sanitizer_XDR ` if `SANITIZER_LINUX && !SANITIZER_ANDROID`, but sanitizer_platform_limits_posix.cpp tries to check it if `HAVE_RPC_XDR_H`. This coincidentally works because macOS has a broken <rpc/xdr.h> which causes `HAVE_RPC_XDR_H` to be 0, but if <rpc/xdr.h> is fixed then clang fails to compile on macOS. Restore the platform checks so that <rpc/xdr.h> can be fixed on macOS.
Reviewed By: vitalybuka
Differential Revision: https://reviews.llvm.org/D130060
The `tileAndFuseLinalgOps` is a legacy approach for tiling + fusion of
Linalg operations. Since it was also intended to work on operations
with buffer operands, this method had fairly complex logic to make
sure tile and fuse was correct even with side-effecting linalg ops.
While complex, it still wasnt robust enough. This patch deprecates
this method and thereby deprecating the tiling + fusion method for ops
with buffer semantics. Note that the core transformation to do fusion
of a producer with a tiled consumer still exists. The deprecation here
only removes methods that auto-magically tried to tile and fuse
correctly in presence of side-effects.
The `tileAndFuseLinalgOps` also works with operations with tensor
semantics. There are at least two other ways the same functionality
exists.
1) The `tileConsumerAndFuseProducers` method. This does a similar
transformation, but using a slightly different logic to
automatically figure out the legal tile + fuse code. Note that this
is also to be deprecated soon.
2) The prefered way uses the `TilingInterface` for tile + fuse, and
relies on the caller to set the tiling options correctly to ensure
that the generated code is correct.
As proof that (2) is equivalent to the functionality provided by
`tileAndFuseLinalgOps`, relevant tests have been moved to use the
interface, where the test driver sets the tile sizes appropriately to
generate the expected code.
Differential Revision: https://reviews.llvm.org/D129901
The only difference between the combines were the calls to getNode
that include the true/false values for SELECT_CC or the chain
and branch target for BR_CC.
Wrap the rest of the code into a helper that reads LHS, RHS, and
CC and outputs new values and a bool if a new node needs to be
created.
Depends on D129371.
It survived all GCC ASan tests.
Changes are trivial and mostly "borrowed" RISC-V logics, except that a different SHADOW_OFFSET is used.
Reviewed By: SixWeining, MaskRay, XiaodongLoong
Differential Revision: https://reviews.llvm.org/D129418
This patch adds constant folder for LogOp which only supports single and double precision floating-point.
Reviewed By: ftynse
Differential Revision: https://reviews.llvm.org/D130148
If C > 10, this will require a constant to be materialized for the
And. To avoid this, we can shift X left by XLen-1-C bits to put the
tested bit in the MSB, then we can do a signed compare with 0 to
determine if the MSB is 0 or 1. Thanks to @reames for the suggestion.
I've implemented this inside of translateSetCCForBranch which is
called when setcc+brcond or setcc+select is converted to br_cc or
select_cc during lowering. It doesn't make sense to do this for
general setcc since we lack a sgez instruction.
I've tested bit 10, 11, 31, 32, 63 and a couple bits betwen 11 and 31
and between 32 and 63 for both i32 and i64 where applicable. Select
has some deficiencies where we receive (and (srl X, C), 1) instead.
This doesn't happen for br_cc due to the call to rebuildSetCC in the
generic DAGCombiner for brcond. I'll explore improving select in a
future patch.
Reviewed By: reames
Differential Revision: https://reviews.llvm.org/D130203
MapperJITLinkMemoryManager supports executor memory management using any
implementation of MemoryMapper to do the transfer such as InProcessMapper or
SharedMemoryMapper.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D129495
Since 67220c2ad7 empty SPSSequence<char>s deserialize to default-constructed
ArrayRef<char>s, which have a null data field. We need to check for this to
avoid memcpy'ing from a nullptr.
This should fix the bot failure in
https://lab.llvm.org/buildbot/#/builders/85/builds/9323
The heuristic used to determine where the arclite libraries are to be
found was based on the path of the `clang` executable. However, in some
scenarios the `clang` executable is within a toolchain that does not
have arclite. When this happens, derive the arclite paths from the
sysroot option.
This allows Clang to correctly derive the arclite directory in, e.g.,
Swift CI, using similar logic to what the Swift driver has been doing
for several years.
Patched by Doug Gregor.
Reviewed By: keith
Differential Revision: https://reviews.llvm.org/D130205
Flang algebraic simplification pass will run algebraic simplification
rewrite patterns for Math/Complex/etc. dialects. It is enabled
under opt-for-speed optimization levels (i.e. for O1/O2/O3; Os/Oz will not
enable it).
With this change the FIR/MLIR optimization pipeline becomes affected
by the -O* optimization level switches. Until now these switches
only affected the middle-end and back-end.
Differential Revision: https://reviews.llvm.org/D130035
This allows for automatically inserting expected checks for parser and verifier
diagnostics, which simplifies the workflow when building new dialect
constructs or extending existing ones.
Differential Revision: https://reviews.llvm.org/D130152
We were quite conservative when it came to PHI node handling to avoid
recursive reasoning. Now we check more direct if we have seen a PHI
already or not. This allows non-recursive PHI chains to be handled.
This also exposed a bug as we did only model the effect of one loop
traversal. `phi_no_store_3` has been adapted to show how we would have
used `undef` instead of `1` before. With this patch we don't replace
it at all, which is expected as we do not argue about loop iterations
(or alignments).
Dmitry Vyukov