| Commit message (Collapse) | Author | Age | Files | Lines |
| ... | |
| |
|
|
| |
llvm-svn: 292764
|
| |
|
|
| |
llvm-svn: 292763
|
| |
|
|
|
|
| |
The colon is important.
llvm-svn: 292761
|
| |
|
|
|
|
| |
https://llvm.org/bugs/show_bug.cgi?id=31672
llvm-svn: 292758
|
| |
|
|
|
|
| |
Fixes PR31714.
llvm-svn: 292713
|
| |
|
|
|
|
| |
operation.
llvm-svn: 292712
|
| |
|
|
|
|
| |
Test-only change.
llvm-svn: 292690
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
By enhancing value tracking, we allow an existing min/max canonicalization to
kick in and improve codegen for several targets that have min/max instructions.
Unfortunately, recognizing min/max in value tracking may cause us to hit
a hack in InstCombiner::visitICmpInst() more often:
http://lists.llvm.org/pipermail/llvm-dev/2017-January/109340.html
...but I'm hoping we can remove that soon.
Correctness proofs based on Alive:
Name: smaxmin
Pre: C1 < C2
%cmp2 = icmp slt i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp slt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %min
=>
%cmp2 = icmp slt i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp sgt i8 %min, C1
%r = select i1 %cmp1, i8 %min, i8 C1
Name: sminmax
Pre: C1 > C2
%cmp2 = icmp sgt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp sgt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %max
=>
%cmp2 = icmp sgt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp slt i8 %max, C1
%r = select i1 %cmp1, i8 %max, i8 C1
----------------------------------------
Optimization: smaxmin
Done: 1
Optimization is correct!
----------------------------------------
Optimization: sminmax
Done: 1
Optimization is correct!
Name: umaxmin
Pre: C1 u< C2
%cmp2 = icmp ult i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp ult i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %min
=>
%cmp2 = icmp ult i8 %x, C2
%min = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp ugt i8 %min, C1
%r = select i1 %cmp1, i8 %min, i8 C1
Name: uminmax
Pre: C1 u> C2
%cmp2 = icmp ugt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp3 = icmp ugt i8 %x, C1
%r = select i1 %cmp3, i8 C1, i8 %max
=>
%cmp2 = icmp ugt i8 %x, C2
%max = select i1 %cmp2, i8 %x, i8 C2
%cmp1 = icmp ult i8 %max, C1
%r = select i1 %cmp1, i8 %max, i8 C1
----------------------------------------
Optimization: umaxmin
Done: 1
Optimization is correct!
----------------------------------------
Optimization: uminmax
Done: 1
Optimization is correct!
llvm-svn: 292660
|
| |
|
|
|
|
|
|
|
|
|
| |
Add DUMMY_CHAIN SDNode to denote stores of interest
Bugzilla: https://llvm.org/bugs/show_bug.cgi?id=28915
Bugzilla: https://llvm.org/bugs/show_bug.cgi?id=30411
Differential Revision: https://reviews.llvm.org/D27964
llvm-svn: 292651
|
| |
|
|
|
|
|
|
|
| |
WebAssembly varargs functions use a significantly different ABI than
non-varargs functions, and the current code in
WebAssemblyFixFunctionBitcasts doesn't handle that difference. For now,
just avoid creating wrapper functions in the presence of varargs.
llvm-svn: 292645
|
| |
|
|
| |
llvm-svn: 292640
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Kill flags need to be updated correctly when moving stores up/down to
form store pair instructions.
Those invalid flags have been ignored before but as of r290014 they are
recognized when using -mllvm -verify-machineinstrs.
Also simplifies test/CodeGen/AArch64/ldst-opt-dbg-limit.mir, renames it
to ldst-opt.mir test and adds a new tests for this change.
Differential Revision: https://reviews.llvm.org/D28875
llvm-svn: 292625
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The recommit fixes a bug related with live interval update after the partial
redundent copy is moved.
The original patch is to solve the performance problem described in PR27827.
Register coalescing sometimes cannot remove a copy because of interference.
But if we can find a reverse copy in one of the predecessor block of the copy,
the copy is partially redundent and we may remove the copy partially by moving
it to the predecessor block without the reverse copy.
Differential Revision: https://reviews.llvm.org/D28585
llvm-svn: 292621
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
We also want to optimise tests like this: return a*b == 0. The MULS
instruction is flag setting, so we don't need the CMP instruction but can
instead branch on the result of the MULS. The generated instructions sequence
for this example was: MULS, MOVS, MOVS, CMP. The MOVS instruction load the
boolean values resulting from the select instruction, but these MOVS
instructions are flag setting and were thus preventing this optimisation. Now
we first reorder and move the MULS to before the CMP and generate sequence
MOVS, MOVS, MULS, CMP so that the optimisation could trigger. Reordering of the
MULS and MOVS is safe to do because the subsequent MOVS instructions just set
the CPSR register and don't use it, i.e. the CPSR is dead.
Differential Revision: https://reviews.llvm.org/D27990
llvm-svn: 292608
|
| |
|
|
|
|
| |
intrinsic. This could break if any future optimizations taken advantage of the undef.
llvm-svn: 292585
|
| |
|
|
| |
llvm-svn: 292567
|
| |
|
|
|
|
|
| |
It's incorrect to ignore the higher bits of the integer source.
Teach the legalizer how to widen it.
llvm-svn: 292563
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Inline spiller can decide to move a spill as early as possible in the basic block.
It will skip phis and label, but we also need to make sure it skips instructions
in the basic block prologue which restore exec mask.
Added isPositionLike callback in TargetInstrInfo to detect instructions which
shall be skipped in addition to common phis, labels etc.
Differential Revision: https://reviews.llvm.org/D27997
llvm-svn: 292554
|
| |
|
|
|
|
|
|
|
| |
Big functions with large vreg # are quite unwieldy to update.
Change it to have one function per test (it does increase boilerplate,
but makes the core hopefully more readable and maintanable).
llvm-svn: 292552
|
| |
|
|
|
|
|
|
|
|
|
| |
Big functions with large vreg # are quite unwieldy to update. This test
also relied on legal s8 operations which we're considering removing.
Change it to have one function per test (it does increase boilerplate,
but makes the core hopefully more readable and maintanable), and use
100% legal operations throughout.
llvm-svn: 292551
|
| |
|
|
|
|
|
|
| |
This completes r292321 by adding support for generic registers, e.g.:
%2:_(s32) = G_ADD %0, %1
llvm-svn: 292550
|
| |
|
|
|
|
|
| |
We can't currently handle "calls" to inlineasm strings so it's better to let
the DAG handle it than generate rubbish.
llvm-svn: 292540
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This patch improves the knownbits logic for unsigned integer min/max opcodes.
For UMIN we know that the result will have the maximum of the inputs' known leading zero bits in the result, similarly for UMAX the maximum of the inputs' leading one bits.
This is particularly useful for simplifying clamping patterns,. e.g. as SSE doesn't have a uitofp instruction we want to use sitofp instead where possible and for that we need to confirm that the top bit is not set.
Differential Revision: https://reviews.llvm.org/D28853
llvm-svn: 292528
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
such problem earlier
Summary:
Emission of XRay table was occasionally disabled for Arm32, but this bug was not then detected because earlier (also by mistake) testing of XRay was occasionally disabled on 32-bit Arm targets. This patch should fix that problem and detect such problems in the future.
This patch is one of a series, see also
- https://reviews.llvm.org/D28623
Reviewers: rengolin, dberris
Reviewed By: dberris
Subscribers: llvm-commits, aemerson, rengolin, dberris, iid_iunknown
Differential Revision: https://reviews.llvm.org/D28624
llvm-svn: 292516
|
| |
|
|
|
|
|
|
| |
been extended
As discussed on D28219 - it is profitable to combine trunc(binop (s/zext(x), s/zext(y)) to binop(trunc(s/zext(x)), trunc(s/zext(y))) assuming the trunc(ext()) will simplify further
llvm-svn: 292493
|
| |
|
|
|
|
|
|
| |
already been extended
As discussed on D28219 - it is profitable to combine trunc(binop (s/zext(x), s/zext(y)) to binop(trunc(s/zext(x)), trunc(s/zext(y))) assuming the trunc(ext()) will simplify further
llvm-svn: 292487
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
The SDNodeOrder is saved in the IROrder field in the SDNode, and this
field may affects scheduling. Thus, letting dbg.value/declare increase
the order numbers may in turn affect scheduling.
Because of this change we also need to update the code deciding when
dbg values should be output, in ScheduleDAGSDNodes.cpp/ProcessSDDbgValues.
Dbg values now have the same order as the SDNode they are connected to,
not the following orders.
Test cases provided by Florian Hahn.
Reviewers: bogner, aprantl, sunfish, atrick
Reviewed By: atrick
Subscribers: fhahn, probinson, andreadb, llvm-commits, MatzeB
Differential Revision: https://reviews.llvm.org/D25318
llvm-svn: 292485
|
| |
|
|
|
|
| |
Differential Revision: https://reviews.llvm.org/D28805
llvm-svn: 292481
|
| |
|
|
|
|
|
| |
A test case that crached is added to avx512-trunc.ll.
(PR31589)
llvm-svn: 292479
|
| |
|
|
| |
llvm-svn: 292476
|
| |
|
|
|
|
|
|
|
|
| |
to broadcast a 128-bit subvector into a 512-bit vector. We'd be better off using something like SHUFF32X4.
If the subvector comes from a load, we convert to SUBV_BROADCAST and use a broadcast instruction. But if there is no load we keep the inserts. I think we should create the SUBV_BROADCAST even without the load and let isel use the fallback patterns that are used if the load can't be folded. This will use the SHUFF32X4 or similar instruction for the 128-bit into 512-bit case and a single insert for 128 into 256 or 256 into 512.
This should be fixed so subvector broadcast intrinsics can be replaced with native IR since some of those currently lower directly to SHUFF32X4.
llvm-svn: 292475
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
Currently we expand and scalarize these operations, but I think we should be able to implement ADD/SUB with KXOR and MUL with KAND.
We already do this for scalar i1 operations so I just extended it to vectors of i1.
Reviewers: zvi, delena
Reviewed By: delena
Subscribers: guyblank, llvm-commits
Differential Revision: https://reviews.llvm.org/D28888
llvm-svn: 292474
|
| |
|
|
|
|
|
|
|
|
|
|
| |
For -(x + y) -> (-x) + (-y), if x == -y, this would
change the result from -0.0 to 0.0. Since the fma/fmad
combine is an extension of this problem it also
applies there.
fmul should be fine, and I don't think any of the unary
operators or conversions should be a problem either.
llvm-svn: 292473
|
| |
|
|
|
|
|
|
| |
They seem to produce nonsense results when used.
This should be applied to the release branch.
llvm-svn: 292472
|
| |
|
|
|
|
| |
subvector broadcasts that can't fold the load.
llvm-svn: 292466
|
| |
|
|
|
|
| |
demonstates that we don't fold the loads to use a broadcast instruction.
llvm-svn: 292465
|
| |
|
|
| |
llvm-svn: 292461
|
| |
|
|
| |
llvm-svn: 292459
|
| |
|
|
|
|
|
|
|
| |
There's no neg.f16 instruction, so negation has to
be done via subtraction from zero.
Differential Revision: https://reviews.llvm.org/D28876
llvm-svn: 292452
|
| |
|
|
| |
llvm-svn: 292446
|
| |
|
|
| |
llvm-svn: 292445
|
| |
|
|
|
|
|
|
|
| |
r291670 doesn't crash on the original testcase from PR31589,
but it crashes on a slightly more complex one.
PR31589 has the new reproducer.
llvm-svn: 292444
|
| |
|
|
|
|
|
|
|
|
|
| |
ARM seems to prefer that long literals be formed from their little end in
order to promote the fusion of the instrs pairs MOV/MOVK and MOVK/MOVK on
Cortex A57 and others (v. "Cortex A57 Software Optimisation Guide", section
4.14).
Differential revision: https://reviews.llvm.org/D28697
llvm-svn: 292422
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Limit register coalescer by not allowing it to artificially increase
size of registers beyond dword. Such super-registers are in fact
register sequences and not distinct HW registers.
With more super-regs we would need to allocate adjacent registers
and constraint regalloc more than needed. Moreover, our super
registers are overlapping. For instance we have VGPR0_VGPR1_VGPR2,
VGPR1_VGPR2_VGPR3, VGPR2_VGPR3_VGPR4 etc, which complicates registers
allocation even more, resulting in excessive spilling.
Differential Revision: https://reviews.llvm.org/D28782
llvm-svn: 292413
|
| |
|
|
|
|
|
| |
Legalize stores of types that are too wide by breaking them up into
sequences of smaller stores.
llvm-svn: 292412
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Non-prevailing weak/linkonce odr symbols will be dropped by ThinLTO to
available_externally when possible. If they had an initializer in the
global_ctors list, a comdat group was being created. This code
already had logic to skip available_externally defs, but now the
EliminateAvailableExternally pass will drop these symbols to
declarations earlier. Change the check to skip all declarations for
linker (which includes available_externally along with declarations).
Reviewers: mehdi_amini
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D28737
llvm-svn: 292408
|
| |
|
|
| |
llvm-svn: 292363
|
| |
|
|
|
|
| |
combineSRA doesn't detect sign bits splats that it does itself so just use -1 as the demanded input so that its already splatted
llvm-svn: 292361
|
| |
|
|
|
|
|
|
|
|
|
| |
This patch improves the mul instruction combine function (combineMul)
by adding new layer of logic.
In this patch, we are adding the ability to fold (mul x, -((1 << c) -1))
or (mul x, -((1 << c) +1)) into (neg(X << c) -x) or (neg((x << c) + x) respective.
Differential Revision: https://reviews.llvm.org/D28232
llvm-svn: 292358
|
| |
|
|
|
|
|
|
|
|
|
| |
identify such problem earlier"
This reverts commit r292210, as it broke the Thumb buldbot with:
clang-5.0: error: the clang compiler does not support '-fxray-instrument
on thumbv7-unknown-linux-gnueabihf'.
llvm-svn: 292357
|