Summary:
Try to lower a BUILD_VECTOR composed of extract-extract chains that can be
reasoned to be a permutation of a vector by indices in a non-constant vector.
We saw this pattern created by ISPC, which resolts to creating it due to the
requirement that shufflevector's mask operand be a *constant* vector.
I didn't check this but we could possibly use this pattern for lowering the X86 permute
C-instrinsics instead of llvm.x86 instrinsics.
This change can be followed by more improvements:
1. Handle vectors with undef elements.
2. Utilize pshufb and zero-mask-blending to support more effiecient
construction of vectors with constant-0 elements.
3. Use smaller-element vectors of same width, and "interpolate" the indices,
when no native operation available.
Reviewers: RKSimon, craig.topper
Reviewed By: RKSimon
Subscribers: chandlerc, DavidKreitzer
Differential Revision: https://reviews.llvm.org/D39126
llvm-svn: 317463
After ObjectYAML learnt the proper enum names for ARMNT/ARM64
relocations, it no longer accepts the numerical values.
This fixes LLD tests after SVN r317459 in LLVM.
llvm-svn: 317460
This patch, together with a matching clang patch (https://reviews.llvm.org/D38683), implements the lowering of X86 broadcastm intrinsics to IR.
Differential Revision: https://reviews.llvm.org/D38684
Change-Id: I709ac0b34641095397e994c8ff7e15d1315b3540
llvm-svn: 317458
Next step is to use them for the legacy FMA scalar intrinsics as well. This will enable the legacy intrinsics to use EVEX encoded opcodes and the extended registers.
llvm-svn: 317453
Use feature names instead of CPU names.
A future commit will add avx512vl command lines to demonstrate missed use of EVEX instructions.
llvm-svn: 317451
The Traced flag is unnecessary because we only need to set the index
to -1 to mark a symbol for tracing.
Differential Revision: https://reviews.llvm.org/D39672
llvm-svn: 317450
Now that DefinedRegular is the only remaining derived class of
Defined, we can merge the two classes.
Differential Revision: https://reviews.llvm.org/D39667
llvm-svn: 317448
Common symbols are now represented with a DefinedRegular that points
to a BssSection, even during symbol resolution.
Differential Revision: https://reviews.llvm.org/D39666
llvm-svn: 317447
I don't remember why I made shared symbols one type of defined symbols.
Shared symbols aren't undefined, so it could be considered defined, but
categorizing three symbols as:
- defined
- really defined
- shared
- undefined
is not as intuitive as
- defined
- shared
- undefined
to me. So, in this patch, I made a change to stop handling shared
symbols as defined symbols.
Surprisingly, I didn't have to update any tests for this change.
Differential Revision: https://reviews.llvm.org/D39394
llvm-svn: 317446
single-iteration loop
This fixes PR34681. Avoid adding the "Stride == 1" predicate when we know that
Stride >= Trip-Count. Such a predicate will effectively optimize a single
or zero iteration loop, as Trip-Count <= Stride == 1.
Differential Revision: https://reviews.llvm.org/D38785
llvm-svn: 317438
The TR6 document is expected to be publically released around November 15.
This patch does not implement OMPT for libomptarget.
Patch by Simon Convent and Joachim Protze
Differential Revision: https://reviews.llvm.org/D39182
llvm-svn: 317436
This is part of the renaming of data types from OpenMP TR4 to TR6
Patch by Simon Convent
Differential Revision: https://reviews.llvm.org/D39326
llvm-svn: 317435
reverted my changes will be committed later after fixing the failure
This patch contains update of the costs of interleaved loads of v8f32 of stride 3 and 8.
Differential Revision: https://reviews.llvm.org/D39403
llvm-svn: 317433
This patch contains update of the costs of interleaved loads of v8f32 of stride 3 and 8.
Differential Revision: https://reviews.llvm.org/D39403
llvm-svn: 317432
This is an implementation of PR26223.
Currently optimizeMemoryInst optimization tries to fold address computation
if all possible way to get compute the address are of the form
baseGV + base + scale * Index + offset
where scale and offset are constants and baseGV, base and Index are exactly
the same instructions if defined.
The patch extends this optimization to allow different bases. In this case
it tries to find/build a Phi node merging all possible bases and use this Phi node
as a base for sunk address computation. Also it supports Select instruction on
the way.
The main motivation for this scope extension is GCRelocateInst.
If there is a relocation of derived pointer it will be represented as relocation of base + offset.
Also there will be a Phi node merging address computation for relocated derived pointer
and derived pointer itself. If we have a Phi node merging original base and relocated base
and can fold the address computation of derived pointer then we can potentially reduce
the code size and Phi node for derived pointer. The later can have a positive impact to
register allocator.
Reviewers: efriedma, dberlin, mkazantsev, reames, john.brawn
Reviewed By: john.brawn
Subscribers: javed.absar, john.brawn, dneilson, llvm-commits
Differential Revision: https://reviews.llvm.org/D36073
llvm-svn: 317429
That class is used only by LinkerScript.cpp, so we should move it to
that file. Also, it no longer has to be a "factory" class. It can just
be a non-member function.
llvm-svn: 317427
r317396 changed the way how we handle the -defsym option. The option is
now handled using the infrastructure for the linker script.
We used to handle both -defsym and -wrap using the same set of functions
in the symbol table. Now, we don't need to do that.
This patch rewrites the functions so that they become more straightforward.
The new functions directly handle -wrap rather than abstract it.
llvm-svn: 317426
Basically a regression after r316268.
However the diagnostic is correct, but the test coverage is bad.
So just like rL316500, introduce yet more tests,
and adjust the release notes.
See https://bugs.llvm.org/show_bug.cgi?id=35200
llvm-svn: 317421