This patch teaches the DAGCombiner how to combine shuffles according to rules:
shuffle(shuffle(A, Undef, M0), B, M1) -> shuffle(B, A, M2)
shuffle(shuffle(A, B, M0), B, M1) -> shuffle(B, A, M2)
shuffle(shuffle(A, B, M0), A, M1) -> shuffle(B, A, M2)
llvm-svn: 222090
Updated X86TargetLowering::isShuffleMaskLegal to match SHUFP masks with commuted inputs and PSHUFD masks that reference the second input.
As part of this I've refactored isPSHUFDMask to work in a more general manner and allow it to match against either the first or second input vector.
Differential Revision: http://reviews.llvm.org/D6287
llvm-svn: 222087
getTargetConstant should only be used when you can guarantee the instruction
selected will be able to cope with the raw value. BUILD_VECTOR is rather too
generic for this so we should use getConstant instead. In that case, an
instruction can still consume the constant, but if it doesn't it'll be
materialised through its own round of ISel.
Should fix PR21352.
llvm-svn: 221961
in-lane shuffles that aren't always handled well by the current vector
shuffle lowering.
No functionality change yet, that will follow in a subsequent commit.
llvm-svn: 221938
between splitting a vector into 128-bit lanes and recombining them vs.
decomposing things into single-input shuffles and a final blend.
This handles a large number of cases in AVX1 where the cross-lane
shuffles would be much more expensive to represent even though we end up
with a fast blend at the root. Instead, we can do a better job of
shuffling in a single lane and then inserting it into the other lanes.
This fixes the remaining bits of Halide's regression captured in PR21281
for AVX1. However, the bug persists in AVX2 because I've made this
change reasonably conservative. The cases where it makes sense in AVX2
to split into 128-bit lanes are much more rare because we can often do
full permutations across all elements of the 256-bit vector. However,
the particular test case in PR21281 is an example of one of the rare
cases where it is *always* better to work in a single 128-bit lane. I'm
going to try to teach the logic to detect and form the good code even in
AVX2 next, but it will need to use a separate heuristic.
Finally, there is one pesky regression here where we previously would
craftily use vpermilps in AVX1 to shuffle both high and low halves at
the same time. We no longer pull that off, and not for any really good
reason. Ultimately, I think this is just another missing nuance to the
selection heuristic that I'll try to add in afterward, but this change
already seems strictly worth doing considering the magnitude of the
improvements in common matrix math shuffle patterns.
As always, please let me know if this causes a surprising regression for
you.
llvm-svn: 221861
re-combining shuffles because nothing was available in the wider vector
type.
The key observation (which I've put in the comments for future
maintainers) is that at this point, no further combining is really
possible. And so even though these shuffles trivially could be combined,
we need to actually do that as we produce them when producing them this
late in the lowering.
This fixes another (huge) part of the Halide vector shuffle regressions.
As it happens, this was already well covered by the tests, but I hadn't
noticed how bad some of these got. The specific patterns that turn
directly into unpckl/h patterns were occurring *many* times in common
vector processing code.
There are still more problems here sadly, but trying to incrementally
tease them apart and it looks like this is the core of the problem in
the splitting logic.
There is some chance of regression here, you can see it in the test
changes. Specifically, where we stop forming pshufb in some cases, it is
possible that pshufb was in fact faster. Intel "says" that pshufb is
slower than the instruction sequences replacing it.
llvm-svn: 221852
Prior to this patch the TypePromotionHelper was promoting only sign extensions.
Supporting zero extensions changes:
- How constants are extended.
- How sign extensions, zero extensions, and truncate are composed together.
- How the type of the extended operation is recorded. Now we need to know the
kind of the extension as well as its type.
Each change is fairly small, unlike the diff.
Most of the diff are comments/variable renaming to say "extension" instead of
"sign extension".
The performance improvements on the test suite are within the noise.
Related to <rdar://problem/18310086>.
llvm-svn: 221851
This is a follow-on to r221706 and r221731 and discussed in more detail in PR21385.
This patch also loosens the testcase checking for btver2. We know that the "1.0" will be loaded, but
we can't tell exactly when, so replace the CHECK-NEXT specifiers with plain CHECKs. The CHECK-NEXT
sequence relied on a quirk of post-RA-scheduling that may change independently of anything in these tests.
llvm-svn: 221819
cases from Halide folks. This initial step was extracted from
a prototype change by Clay Wood to try and address regressions found
with Halide and the new vector shuffle lowering.
llvm-svn: 221779
This commit adds a new pass that can inject checks before indirect calls to
make sure that these calls target known locations. It supports three types of
checks and, at compile time, it can take the name of a custom function to call
when an indirect call check fails. The default failure function ignores the
error and continues.
This pass incidentally moves the function JumpInstrTables::transformType from
private to public and makes it static (with a new argument that specifies the
table type to use); this is so that the CFI code can transform function types
at call sites to determine which jump-instruction table to use for the check at
that site.
Also, this removes support for jumptables in ARM, pending further performance
analysis and discussion.
Review: http://reviews.llvm.org/D4167
llvm-svn: 221708
This is a first step for generating SSE rcp instructions for reciprocal
calcs when fast-math allows it. This is very similar to the rsqrt optimization
enabled in D5658 ( http://reviews.llvm.org/rL220570 ).
For now, be conservative and only enable this for AMD btver2 where performance
improves significantly both in terms of latency and throughput.
We may never enable this codegen for Intel Core* chips because the divider circuits
are just too fast. On SandyBridge, divss can be as fast as 10 cycles versus the 21
cycle critical path for the rcp + mul + sub + mul + add estimate.
Follow-on patches may allow configuration of the number of Newton-Raphson refinement
steps, add AVX512 support, and enable the optimization for more chips.
More background here: http://llvm.org/bugs/show_bug.cgi?id=21385
Differential Revision: http://reviews.llvm.org/D6175
llvm-svn: 221706
The ISel lowering for global TLS access in PIC mode was creating a pseudo
instruction that is later expanded to a call, but the code was not
setting the hasCalls flag in the MachineFrameInfo alongside the adjustsStack
flag. This caused some functions to be mistakenly recognized as leaf functions,
and this in turn affected the decision to eliminate the frame pointer.
With the fix, hasCalls is properly set and the leaf frame pointer is correctly
preserved.
llvm-svn: 221695
This fixes an issue with matching trunc -> assertsext -> zext on x86-64, which would not zero the high 32-bits. See PR20494 for details.
Recommitting - This time, with a hopefully working test.
Differential Revision: http://reviews.llvm.org/D6128
llvm-svn: 221672
AVX2 is available.
According to IACA, the new lowering has a throughput of 8 cycles instead of 13
with the previous one.
Althought this lowering kicks in some SPECs benchmarks, the performance
improvement was within the noise.
Correctness testing has been done for the whole range of uint32_t with the
following program:
uint4 v = (uint4) {0,1,2,3};
uint32_t i;
//Check correctness over entire range for uint4 -> float4 conversion
for( i = 0; i < 1U << (32-2); i++ )
{
float4 t = test(v);
float4 c = correct(v);
if( 0xf != _mm_movemask_ps( t == c ))
{
printf( "Error @ %vx: %vf vs. %vf\n", v, c, t);
return -1;
}
v += 4;
}
Where "correct" is the old lowering and "test" the new one.
The patch adds a test case for the two custom lowering instruction.
It also modifies the vector cost model, which is why cast.ll and uitofp.ll are
modified.
2009-02-26-MachineLICMBug.ll is also modified because we now hoist 7
instructions instead of 4 (3 more constant loads).
rdar://problem/18153096>
llvm-svn: 221657
This fixes an issue with matching trunc -> assertsext -> zext on x86-64, which would not zero the high 32-bits.
See PR20494 for details.
Differential Revision: http://reviews.llvm.org/D6128
llvm-svn: 221626
Fixed an issue with the (v)cvttps2dq and (v)cvttpd2dq instructions being incorrectly put in the 2 source operand folding tables instead of the 1 source operand and added the missing SSE/AVX versions.
Also added missing (v)cvtps2dq and (v)cvtpd2dq instructions to the folding tables.
Differential Revision: http://reviews.llvm.org/D6001
llvm-svn: 221489
On 32 bit windows we use label differences and .set does not suppress
rolocations, a combination that was not used before r220256.
This fixes PR21497.
llvm-svn: 221456
Example:
define <4 x i32> @test(<4 x i32> %a, <4 x i32> %b) {
%shuffle = shufflevector <4 x i32> %a, <4 x i32> %b, <4 x i32> <i32 4, i32 5, i32 6, i32 3>
ret <4 x i32> %shuffle
}
Before llc (-mattr=+sse4.1), produced the following assembly instruction:
pblendw $4294967103, %xmm1, %xmm0
After
pblendw $63, %xmm1, %xmm0
llvm-svn: 221455
condition to match a blend.
This prevents optimizations that work on VSELECT to perform invalid
transformations. Indeed, the optimized condition does not match the vector
boolean content that is expected and bad things may happen.
This patch yields the exact same code on the whole test-suite + specs (-O3 and
-O3 -march=core-avx2), it improves one test case (vector-blend.ll) and fixes a
bug reduced in vselect-avx.ll.
<rdar://problem/18819506>
llvm-svn: 221429
Added missing memory folding for the (V)CVTDQ2PS instructions - we can safely fold these (but not the (V)CVTDQ2PD versions which have a register/memory size discrepancy in the source operand). I've added a test case demonstrating that stack folding now works.
Differential Revision: http://reviews.llvm.org/D5981
llvm-svn: 221407
Summary:
X86FastISel::fastMaterializeAlloca was incorrectly conditioning its
opcode selection on subtarget bitness rather than pointer size.
Differential Revision: http://reviews.llvm.org/D6136
llvm-svn: 221386
This patch improves the folding of vector AND nodes into blend operations for
targets that feature SSE4.1. A vector AND node where one of the operands is
a constant build_vector with elements that are either zero or all-ones can be
converted into a blend.
This allows for example to simplify the following code:
define <4 x i32> @test(<4 x i32> %A, <4 x i32> %B) {
%1 = and <4 x i32> %A, <i32 0, i32 0, i32 0, i32 -1>
%2 = and <4 x i32> %B, <i32 -1, i32 -1, i32 -1, i32 0>
%3 = or <4 x i32> %1, %2
ret <4 x i32> %3
}
Before this patch llc (-mcpu=corei7) generated:
andps LCPI1_0(%rip), %xmm0, %xmm0
andps LCPI1_1(%rip), %xmm1, %xmm1
orps %xmm1, %xmm0, %xmm0
retq
With this patch we generate a single 'vpblendw'.
llvm-svn: 221343
Patch to allow (v)blendps, (v)blendpd, (v)pblendw and vpblendd instructions to be commuted - swaps the src registers and inverts the blend mask.
This is primarily to improve memory folding (see new tests), but it also improves the quality of shuffles (see modified tests).
Differential Revision: http://reviews.llvm.org/D6015
llvm-svn: 221313
For 8-bit divrems where the remainder is used, we used to generate:
divb %sil
shrw $8, %ax
movzbl %al, %eax
That was to avoid an H-reg access, which is problematic mainly because
it isn't possible in REX-prefixed instructions.
This patch optimizes that to:
divb %sil
movzbl %ah, %eax
To do that, we explicitly extend AH, and extract the L-subreg in the
resulting register. The extension is done using the NOREX variants of
MOVZX. To support signed operations, MOVSX_NOREX is also added.
Further, this introduces a new SDNode type, [us]divrem_ext_hreg, which is
then lowered to a sequence containing a single zext (rather than 2).
Differential Revision: http://reviews.llvm.org/D6064
llvm-svn: 221176
call DAGCombiner. But we ran into a case (on Windows) where the
calling convention causes argument lowering to bail out of fast-isel,
and we end up in CodeGenAndEmitDAG() which does run DAGCombiner.
So, we need to make DAGCombiner check for 'optnone' after all.
Commit includes the test that found this, plus another one that got
missed in the original optnone work.
llvm-svn: 221168
"[x86] Simplify vector selection if condition value type matches vselect value type and true value is all ones or false value is all zeros."
llvm-svn: 221028
This transformation worked if selector is produced by SETCC, however SETCC is needed only if we consider to swap operands. So I replaced SETCC check for this case.
Added tests for vselect of <X x i1> values.
llvm-svn: 220777
Ffter commit at rev219046 512-bit broadcasts lowering become non-optimal. Most of tests on broadcasting and embedded broadcasting were changed and they doesn’t produce efficient code.
Example below is from commit changes (it’s the first test from test/CodeGen/X86/avx512-vbroadcast.ll):
define <16 x i32> @_inreg16xi32(i32 %a) {
; CHECK-LABEL: _inreg16xi32:
; CHECK: ## BB#0:
-; CHECK-NEXT: vpbroadcastd %edi, %zmm0
+; CHECK-NEXT: vmovd %edi, %xmm0
+; CHECK-NEXT: vpbroadcastd %xmm0, %ymm0
+; CHECK-NEXT: vinserti64x4 $1, %ymm0, %zmm0, %zmm0
; CHECK-NEXT: retq
%b = insertelement <16 x i32> undef, i32 %a, i32 0
%c = shufflevector <16 x i32> %b, <16 x i32> undef, <16 x i32> zeroinitializer
ret <16 x i32> %c
}
Here, 256-bit broadcast was generated instead of 512-bit one.
In this patch
1) I added vector-shuffle lowering through broadcasts
2) Removed asserts and branches likes because this is incorrect
- assert(Subtarget->hasDQI() && "We can only lower v8i64 with AVX-512-DQI");
3) Fixed lowering tests
llvm-svn: 220774
This is a Microsoft calling convention that supports both x86 and x86_64
subtargets. It passes vector and floating point arguments in XMM0-XMM5,
and passes them indirectly once they are consumed.
Homogenous vector aggregates of up to four elements can be passed in
sequential vector registers, but this part is not implemented in LLVM
and will be handled in Clang.
On 32-bit x86, it is similar to fastcall in that it uses ecx:edx as
integer register parameters and is callee cleanup. On x86_64, it
delegates to the normal win64 calling convention.
Reviewers: majnemer
Differential Revision: http://reviews.llvm.org/D5943
llvm-svn: 220745
For a call to not return in to the stackmap shadow, the shadow must end with the call.
To do this, we must insert any required nops *before* the call, and not after it.
llvm-svn: 220728
To avoid emitting too many nops, a stackmap shadow can include emitted instructions in the shadow, but these must not include branch targets.
A return from a call should count as a branch target as patching over the instructions after the call would lead to incorrect behaviour for threads currently making that call, when they return.
llvm-svn: 220710
Minor patch to fix an issue in XFormVExtractWithShuffleIntoLoad where a load is unary shuffled, then bitcast (to a type with the same number of elements) before extracting an element.
An undef was created for the second shuffle operand using the original (post-bitcasted) vector type instead of the pre-bitcasted type like the rest of the shuffle node - this was then causing an assertion on the different types later on inside SelectionDAG::getVectorShuffle.
Differential Revision: http://reviews.llvm.org/D5917
llvm-svn: 220592
This is a first step for generating SSE rsqrt instructions for
reciprocal square root calcs when fast-math is allowed.
For now, be conservative and only enable this for AMD btver2
where performance improves significantly - for example, 29%
on llvm/projects/test-suite/SingleSource/Benchmarks/BenchmarkGame/n-body.c
(if we convert the data type to single-precision float).
This patch adds a two constant version of the Newton-Raphson
refinement algorithm to DAGCombiner that can be selected by any target
via a parameter returned by getRsqrtEstimate()..
See PR20900 for more details:
http://llvm.org/bugs/show_bug.cgi?id=20900
Differential Revision: http://reviews.llvm.org/D5658
llvm-svn: 220570
x86's CMPXCHG -> EFLAGS consumer wasn't being recorded as a real EFLAGS
dependency because it was represented by a pair of CopyFromReg(EFLAGS) ->
CopyToReg(EFLAGS) nodes. ScheduleDAG was expecting the source to be an
implicit-def on the instruction, where the result numbers in the DAG and the
Uses list in TableGen matched up precisely.
The Copy notation seems much more robust, so this patch extends ScheduleDAG
rather than refactoring x86.
Should fix PR20376.
llvm-svn: 220529
Currently, @llvm.smul.with.overflow.i8 expands to 9 instructions, where
3 are really needed.
This adds X86ISD::UMUL8/SMUL8 SD nodes, and custom lowers them to
MUL8/IMUL8 + SETO.
i8 is a special case because there is no two/three operand variants of
(I)MUL8, so the first operand and return value need to go in AL/AX.
Also, we can't write patterns for these instructions: TableGen refuses
patterns where output operands don't match SDNode results. In this case,
instructions where the output operand is an implicitly defined register.
A related special case (and FIXME) exists for MUL8 (X86InstrArith.td):
// FIXME: Used for 8-bit mul, ignore result upper 8 bits.
// This probably ought to be moved to a def : Pat<> if the
// syntax can be accepted.
[(set AL, (mul AL, GR8:$src)), (implicit EFLAGS)]
Ideally, these go away with UMUL8, but we still need to improve TableGen
support of implicit operands in patterns.
Before this change:
movsbl %sil, %eax
movsbl %dil, %ecx
imull %eax, %ecx
movb %cl, %al
sarb $7, %al
movzbl %al, %eax
movzbl %ch, %esi
cmpl %eax, %esi
setne %al
After:
movb %dil, %al
imulb %sil
seto %al
Also, remove a made-redundant testcase for PR19858, and enable more FastISel
ALU-overflow tests for SelectionDAG too.
Differential Revision: http://reviews.llvm.org/D5809
llvm-svn: 220516
This reverts commit r214981.
I'm not sure what I was thinking when I wrote this. Testing with MSVC
shows that this function is mangled to '@f@8':
int __fastcall f(int a, int b);
llvm-svn: 220492
Every target we support has support for assembly that looks like
a = b - c
.long a
What is special about MachO is that the above combination suppresses the
production of a relocation.
With this change we avoid producing the intermediary labels when they don't
add any value.
llvm-svn: 220256
X86 code to lower VSELECT messed a bit with the bits set in the mask of VSELECT
when it knows it can be lowered into BLEND. Indeed, only the high bits need to be
set for those and it optimizes those accordingly.
However, when the mask is a compile time constant, the lowering will be handled
by the generic optimizer and those modifications will generate bad code in the
generic optimizer.
This patch fixes that by preventing the optimization if the VSELECT will be
handled by the generic optimizer.
<rdar://problem/18675020>
llvm-svn: 220242
This patch improves support for commutative instructions in the x86 memory folding implementation by attempting to fold a commuted version of the instruction if the original folding fails - if that folding fails as well the instruction is 're-commuted' back to its original order before returning.
Updated version of r219584 (reverted in r219595) - the commutation attempt now explicitly ensures that neither of the commuted source operands are tied to the destination operand / register, which was the source of all the regressions that occurred with the original patch attempt.
Added additional regression test case provided by Joerg Sonnenberger.
Differential Revision: http://reviews.llvm.org/D5818
llvm-svn: 220239
TL;DR: Indexing maps with [] creates missing entries.
The long version:
When selecting lifetime intrinsics, we index the *static* alloca map with the AllocaInst we find for that lifetime. Trouble is, we don't first check to see if this is a dynamic alloca.
On the attached example, this causes a dynamic alloca to create an entry in the static map, and returns 0 (the default) as the frame index for that lifetime. 0 was used for the frame index of the stack protector, which given that it now has a lifetime, is coloured, and merged with other stack slots.
PEI would later trigger an assert because it expects the stack protector to not be dead.
This fix ensures that we only get frame indices for static allocas, ie, those in the map. Dynamic ones are effectively dropped, which is suboptimal, but at least isn't completely broken.
rdar://problem/18672951
llvm-svn: 220099
When the input to a store instruction was a zero vector, the backend
always selected a normal vector store regardless of the non-temporal
hint. This is fixed by this patch.
This fixes PR19370.
llvm-svn: 220054
In AVX512f we support 64x2 and 32x8 inserts via matching them to 32x4 and 64x4
respectively. These are matched by "Alt" Pat<>'s (Alt stands for alternative
VTs).
Since DQ has native support for these intructions, I peeled off the non-"Alt"
part of the baseclass into vinsert_for_size_no_alt. The DQ instructions are
derived from this multiclass. The "Alt" Pat<>'s are disabled with DQ.
Fixes <rdar://problem/18426089>
llvm-svn: 219874
Summary:
Fixes a FIXME in MachineSinking. Instead of using the simple heuristics in
isPostDominatedBy, use the real MachinePostDominatorTree and MachineLoopInfo.
The old heuristics caused instructions to sink unnecessarily, and might create
register pressure.
This is the second try of the fix. The first one (D4814) caused a performance
regression due to failing to sink instructions out of loops (PR21115). This
patch fixes PR21115 by sinking an instruction from a deeper loop to a shallower
one regardless of whether the target block post-dominates the source.
Thanks Alexey Volkov for reporting PR21115!
Test Plan:
Added a NVPTX codegen test to verify that our change prevents the backend from
over-sinking. It also shows the unnecessary register pressure caused by
over-sinking.
Added an X86 test to verify we can sink instructions out of loops regardless of
the dominance relationship. This test is reduced from Alexey's test in PR21115.
Updated an affected test in X86.
Also ran SPEC CINT2006 and llvm-test-suite for compilation time and runtime
performance. Results are attached separately in the review thread.
Reviewers: Jiangning, resistor, hfinkel
Reviewed By: hfinkel
Subscribers: hfinkel, bruno, volkalexey, llvm-commits, meheff, eliben, jholewinski
Differential Revision: http://reviews.llvm.org/D5633
llvm-svn: 219773
Patch to provide shuffle decodes and asm comments for the sse pslldq/psrldq SSE2/AVX2 byte shift instructions.
Differential Revision: http://reviews.llvm.org/D5598
llvm-svn: 219738
This patch improves support for commutative instructions in the x86 memory folding implementation by attempting to fold a commuted version of the instruction if the original folding fails - if that folding fails as well the instruction is 're-commuted' back to its original order before returning.
This mainly helps the stack inliner better fold reloads of 3 (or more) operand instructions (VEX encoded SSE etc.) but by performing this in the lowest foldMemoryOperandImpl implementation it also replaces the X86InstrInfo::optimizeLoadInstr version and is now used by FastISel too.
Differential Revision: http://reviews.llvm.org/D5701
llvm-svn: 219584
Fixes a logic error in the MachineScheduler found by Steve Montgomery (and
confirmed by Andy). This has gone unfixed for months because the fix has been
found to introduce some small performance regressions. However, Andy has
recommended that, at this point, we fix this to avoid further dependence on the
incorrect behavior (and then follow-up separately on any regressions), and I
agree.
Fixes PR18883.
llvm-svn: 219512
This adds the Pat<>'s for the intrinsics. These are necessary because we
don't lower these intrinsics to SDNodes but match them directly. See the
rational in the previous commit.
llvm-svn: 219362
Summary:
I had forgotten to check for NotSlowIncDec in the patterns that can generate
inc/dec for the above pattern (added in D4796).
This currently applies to Atom Silvermont, KNL and SKX.
Test Plan: New checks on atomic_mi.ll
Reviewers: jfb, nadav
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5677
llvm-svn: 219336
Summary:
Fix pr21099
The pseudocode of what we were doing (spread through two functions) was:
if (operand.doesNotFitIn32Bits())
Opc.initializeWithFoo();
if (operand < 0)
operand = -operand;
if (operand.doesFitIn8Bits())
Opc.initializeWithBar();
else if (operand.doesFitIn32Bits())
Opc.initializeWithBlah();
doStuff(Opc);
So for operand == INT32_MIN, Opc was never initialized because the operand changes
from fitting in 32 bits to not fitting, causing the various bugs/error messages
noted by pr21099.
This patch adds an extra test at the beginning for this case, and an
llvm_unreachable to have better error message if the operand ends up
not fitting in 32-bits at the end.
Test Plan: new test + make check
Reviewers: jfb
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5655
llvm-svn: 219257
The patch's author points out that, despite the function's documentation,
getSetCCResultType is only used to get the SETCC result type (with one
here-removed problematic exception). In one case, getSetCCResultType was being
used to get the predicate type to use for a SELECT node, and then
SIGN_EXTENDing (or truncating) to get the input predicate to match that type.
Unfortunately, this was happening inside visitSIGN_EXTEND, and creating new
SIGN_EXTEND nodes was causing an infinite loop. In addition, this behavior was
wrong if a target was not using ZeroOrNegativeOneBooleanContent. Lastly, the
extension/truncation seems unnecessary here: SELECT is defined as:
Select(COND, TRUEVAL, FALSEVAL). If the type of the boolean COND is not i1
then the high bits must conform to getBooleanContents.
So here we remove this use of getSetCCResultType and update
getSetCCResultType's documentation to reflect its actual uses.
Patch by deadal nix!
llvm-svn: 219141
This trades a (register-renamer-friendly) movaps for a floating point
/ integer domain cross. That is a very bad trade, even on architectures
where domain crossing is relatively fast. On any chip where there is
even a cycle stall, this is a Very Bad Idea. It doesn't even seem likely
to cause a spill to be introduced because the reason for the copy is to
destructively shuffle in place.
Thanks to Ben Kramer for fixing a bug in this code that my new shuffle
lowering exposed and highlighting that perhaps it should just go away.
=]
llvm-svn: 219090
that are unused.
This allows the combiner to delete math feeding shuffles where the math
isn't actually necessary. This improves some of the vperm2x128 tests
that regressed when the vector shuffle lowering started actually
generating vperm instructions rather than forcibly decomposing them.
Sadly, this isn't enough to get this *really* right because we still
form a completely unnecessary permutation. To fix that, we also need to
fold shuffles which just rearrange concatenated or inserted subvectors.
llvm-svn: 219086
new vector shuffle lowering.
This is loosely based on a patch by Marius Wachtler to the PR (thanks!).
I refactored it a bi to use std::count_if and a mutable array ref but
the core idea was exactly right. I also added some direct testing of
this case.
I believe PR21137 is now the only remaining regression.
llvm-svn: 219081
shuffles using AVX and AVX2 instructions. This fixes PR21138, one of the
few remaining regressions impacting benchmarks from the new vector
shuffle lowering.
You may note that it "regresses" many of the vperm2x128 test cases --
these were actually "improved" by the naive lowering that the new
shuffle lowering previously did. This regression gave me fits. I had
this patch ready-to-go about an hour after flipping the switch but
wasn't sure how to have the best of both worlds here and thought the
correct solution might be a completely different approach to lowering
these vector shuffles.
I'm now convinced this is the correct lowering and the missed
optimizations shown in vperm2x128 are actually due to missing
target-independent DAG combines. I've even written most of the needed
DAG combine and will submit it shortly, but this part is ready and
should help some real-world benchmarks out.
llvm-svn: 219079
Update the entire regression test suite for the new shuffles. Remove
most of the old testing which was devoted to the old shuffle lowering
path and is no longer relevant really. Also remove a few other random
tests that only really exercised shuffles and only incidently or without
any interesting aspects to them.
Benchmarking that I have done shows a few small regressions with this on
LNT, zero measurable regressions on real, large applications, and for
several benchmarks where the loop vectorizer fires in the hot path it
shows 5% to 40% improvements for SSE2 and SSE3 code running on Sandy
Bridge machines. Running on AMD machines shows even more dramatic
improvements.
When using newer ISA vector extensions the gains are much more modest,
but the code is still better on the whole. There are a few regressions
being tracked (PR21137, PR21138, PR21139) but by and large this is
expected to be a win for x86 generated code performance.
It is also more correct than the code it replaces. I have fuzz tested
this extensively with ISA extensions up through AVX2 and found no
crashes or miscompiles (yet...). The old lowering had a few miscompiles
and crashers after a somewhat smaller amount of fuzz testing.
There is one significant area where the new code path lags behind and
that is in AVX-512 support. However, there was *extremely little*
support for that already and so this isn't a significant step backwards
and the new framework will probably make it easier to implement lowering
that uses the full power of AVX-512's table-based shuffle+blend (IMO).
Many thanks to Quentin, Andrea, Robert, and others for benchmarking
assistance. Thanks to Adam and others for help with AVX-512. Thanks to
Hal, Eric, and *many* others for answering my incessant questions about
how the backend actually works. =]
I will leave the old code path in the tree until the 3 PRs above are at
least resolved to folks' satisfaction. Then I will rip it (and 1000s of
lines of code) out. =] I don't expect this flag to stay around for very
long. It may not survive next week.
llvm-svn: 219046
the various ways in which blends can be used to do vector element
insertion for lowering with the scalar math instruction forms that
effectively re-blend with the high elements after performing the
operation.
This then allows me to bail on the element insertion lowering path when
we have SSE4.1 and are going to be doing a normal blend, which in turn
restores the last of the blends lost from the new vector shuffle
lowering when I got it to prioritize insertion in other cases (for
example when we don't *have* a blend instruction).
Without the patterns, using blends here would have regressed
sse-scalar-fp-arith.ll *completely* with the new vector shuffle
lowering. For completeness, I've added RUN-lines with the new lowering
here. This is somewhat superfluous as I'm about to flip the default, but
hey, it shows that this actually significantly changed behavior.
The patterns I've added are just ridiculously repetative. Suggestions on
making them better very much welcome. In particular, handling the
commuted form of the v2f64 patterns is somewhat obnoxious.
llvm-svn: 219033
perform a load to use blendps rather than movss when it is available.
For non-loads, blendps is *much* faster. It can execute on two ports in
Sandy Bridge and Ivy Bridge, and *three* ports on Haswell. This fixes
one of the "regressions" from aggressively taking the "insertion" path
in the new vector shuffle lowering.
This does highlight one problem with blendps -- it isn't commuted as
heavily as it should be. That's future work though.
llvm-svn: 219022
This reverts commit r218918, effectively reapplying r218914 after fixing
an Ocaml bindings test and an Asan crash. The root cause of the latter
was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a
PR to investigate who requires the loose check (and why).
Original commit message follows.
--
This patch addresses the first stage of PR17891 by folding constant
arguments together into a single MDString. Integers are stringified and
a `\0` character is used as a separator.
Part of PR17891.
Note: I've attached my testcases upgrade scripts to the PR. If I've
just broken your out-of-tree testcases, they might help.
llvm-svn: 219010
In the X86 backend, matching an address is initiated by the 'addr' complex
pattern and its friends. During this process we may reassociate and-of-shift
into shift-of-and (FoldMaskedShiftToScaledMask) to allow folding of the
shift into the scale of the address.
However as demonstrated by the testcase, this can trigger CSE of not only the
shift and the AND which the code is prepared for but also the underlying load
node. In the testcase this node is sitting in the RecordedNode and MatchScope
data structures of the matcher and becomes a deleted node upon CSE. Returning
from the complex pattern function, we try to access it again hitting an assert
because the node is no longer a load even though this was checked before.
Now obviously changing the DAG this late is bending the rules but I think it
makes sense somewhat. Outside of addresses we prefer and-of-shift because it
may lead to smaller immediates (FoldMaskAndShiftToScale is an even better
example because it create a non-canonical node). We currently don't recognize
addresses during DAGCombiner where arguably this canonicalization should be
performed. On the other hand, having this in the matcher allows us to cover
all the cases where an address can be used in an instruction.
I've also talked a little bit to Dan Gohman on llvm-dev who added the RAUW for
the new shift node in FoldMaskedShiftToScaledMask. This RAUW is responsible
for initiating the recursive CSE on users
(http://lists.cs.uiuc.edu/pipermail/llvmdev/2014-September/076903.html) but it
is not strictly necessary since the shift is hooked into the visited user. Of
course it's safer to keep the DAG consistent at all times (e.g. for accurate
number of uses, etc.).
So rather than changing the fundamentals, I've decided to continue along the
previous patches and detect the CSE. This patch installs a very targeted
DAGUpdateListener for the duration of a complex-pattern match and updates the
matching state accordingly. (Previous patches used HandleSDNode to detect the
CSE but that's not practical here). The listener is only installed on X86.
I tested that there is no measurable overhead due to this while running
through the spec2k BC files with llc. The only thing we pay for is the
creation of the listener. The callback never ever triggers in spec2k since
this is a corner case.
Fixes rdar://problem/18206171
llvm-svn: 219009
and MOVSD nodes for single element vector inserts.
This is particularly important because a number of patterns in the
backend detect these patterns and leverage them to simplify things. It
also fixes quite a few of the insertion bad code examples. However, it
regresses a specific area: when available, blendps and blendpd are
*dramatically* faster than movss and movsd respectively. But it doesn't
really work to form the blend logic first because the blends *aren't* as
crazy efficient when the data is coming from memory anyways, and thus
will have a movss or movsd regardless. Also, doing that would block
a bunch of the patterns that this is designed to hit.
So my plan is to go into the patterns for lowering MOVSS and MOVSD and
lower them via blends when available. However that's a pretty invasive
restructuring so it will need to be a follow-up patch.
I have already gone into the patterns to lower MOVSS and MOVSD from
memory using MOVLPD, etc. Without that, several of the test cases
I already have regress.
llvm-svn: 218985
I got them quite wrong when updating it and had the SSE4.1 run checked
for SSE2 and the SSE2 run checked for SSE4.1. I think everything was
actually generic SSE, but this still seems good to fix. While here,
hoist the triple into the IR and make the flag set a bit more direct in
what it is trying to test.
llvm-svn: 218978
lowering to match VZEXT_MOVL patterns.
I hadn't realized that these had sufficient pattern smarts in the
backend to lower zext-ing from the low element of a vector without it
being a scalar_to_vector node. They do, and this is how to match a bunch
of patterns for movq, movss, etc.
There is a weird propensity to end up using pshufd to place the element
afterward even though it means domain crossing (or rather, to use
xorps+movss to zext the element rather than movq) but that's an
orthogonal problem with VZEXT_MOVL that someone should probably look at.
llvm-svn: 218977
vector to a zero vector for the v2 cases and fix the v4 integer cases to
actually blend from a vector.
There are already seprate tests for the case of inserting from a scalar.
These cases cover a lot of the regressions I've seen in the regression
test suite for the new vector shuffle lowering and specifically cover
the reported lack of using various zext-ing instruction patterns. My
next patch should fix a big chunk of this, but wanted to get a nice
baseline for these patterns in the test cases first.
llvm-svn: 218976
element types to form illegal vector types.
I've added a special SSE1 test case here that makes sure we don't break
this going forward.
llvm-svn: 218974
testing that we generated divps and divss but not in a very systematic
way. There are other tests for widening binary operations already that
make these unnecessary.
The second one seems mostly about testing Atom as well as normal X86,
but despite the comment claiming it is testing a different instruction
sequence, it then tests for exactly the same div instruction sequence!
(The sequence of instructions is actually quite different on Atom, but
not the sequence of div instructions....)
And then it has an "execution" test that simply isn't run? Very strange.
Anyways, none of this is really needed so clean this up.
llvm-svn: 218972
intergrated much more fully into some logical part of the backend to
really understand what it is trying to accomplish and how to update it.
I suspect it no longer holds enough value to be worth having.
llvm-svn: 218950
shufle switch.
I nuked a win64 config from one test as it doesn't really make sense to
cover that ABI specially for generic v2f32 tests...
llvm-svn: 218948
test cases that will change with the new vector shuffle lowering. This
gives us a nice baseline for deltas against. I've checked and removed
the cases where there were weird register usage being pinned down, and
all of these are extremely pin-pointed tests so fully checking them
seems very appropriate.
llvm-svn: 218941
tighter, more strict FileCheck assertions. Some of these I really like
as they show case exactly what instruction sequences come out of these
microscopic functionality tests.
llvm-svn: 218936
baseline for updates from the new vector shuffle lowering.
I've inspected the results here, and I couldn't find any register
allocation decisions where there should be any realistic way to register
allocate things differently. The closest was the imul test case. If you
see something here you'd like register number variables on, just shout
and I'll add them.
llvm-svn: 218935
need to be updated for the new vector shuffle lowering.
After talking to Adam Nemet, Tim Northover, etc., it seems that testing
MC encodings in the same suite as the basic codegen isn't the right
approach. Instead, we're going to want dedicated MC tests for the
encodings. These encodings are starting to get in my way so I wanted to
cut them out early. The total set of instructions that should have
encoding tests added is:
vpaddd
vsqrtss
vsqrtsd
vmovlhps
vmovhlps
valignq
vbroadcastss
Not too many parts of these tests were even using this. =]
llvm-svn: 218932
Do not eliminate the frame pointer if there is a stackmap or patchpoint in the
function. All stackmap references should be FP relative.
This fixes PR21107.
llvm-svn: 218920
This patch addresses the first stage of PR17891 by folding constant
arguments together into a single MDString. Integers are stringified and
a `\0` character is used as a separator.
Part of PR17891.
Note: I've attached my testcases upgrade scripts to the PR. If I've
just broken your out-of-tree testcases, they might help.
llvm-svn: 218914
elements as well as integer elements in order to form simpler shuffle
patterns.
This is the primary reason why we were failing to match some of the
2-and-2 floating point shuffles such as PR21140. Even after fixing this
we need to support some extra patterns in the backend in order to match
the resulting X86ISD::UNPCKL nodes into the correct instructions. This
commit should fix PR21140 and includes more comprehensive testing of
insertion patterns in v4 shuffles.
Not all of the added tests are beautiful. For example, we don't have
clever instructions to insert-via-load in the integer domain. There are
also some places where we aren't sufficiently cunning with our use of
movq and movd, but that's future work.
llvm-svn: 218911
floating point and integer domains.
Merge the AVX2 test into it and add an extra RUN line. Generate clean
FileCheck statements with my script. Remove the now merged AVX2 tests.
llvm-svn: 218903
This file isn't really doing anything useful. Many of the tests that
seem to be combined are also repeats from other test files. Many of the
other tests, despite the comment that they should be combined into
a single shuffle... well... aren't combined into a single shuffle.
=/
llvm-svn: 218862
least seem *slightly* more interesting test wise, although given how
spotily we actually combine anything, I remain somewhat suspicious.
llvm-svn: 218861
checks for all the ISA variants.
If the SSE2 checks here terrify you, good. This is (in large part) the
kind of amazingly bad code that is holding LLVM back when vectorizing on
older ISAs.
At the same time, these tests seem increasingly dubious to me. There are
a very large number of tests and it isn't clear that they are
systematically covering a specific set of functionality. Anyways,
I don't want to reduce testing during the transition, I just want to
consolidate it to where it is easier to manage.
llvm-svn: 218860
file.
Some of these really don't make sense to test -- we're testing for the
*lack* of combining two shuffles into one, presumably because the two
would generate better shuffles in the end. But if you look at the
generated code shown here, in many cases the generated code is, frankly,
terrible. Or we combine any two generated shuffles back into a single
instruction! I've left a FIXME to revisit these decisions.
llvm-svn: 218859
and use the new grouped FileCheck patterns to match them.
No interesting changes yet, but this test is now in proper form to have
the other shuffle combining tests merged into it.
llvm-svn: 218857
The test has to do with DAG combines, and so it doesn't need the new
vector shuffle lowering to be effective. Also, it has a nice in-IR
triple string which we should really be using rather than command line
flags (unless it varies form RUN-line to RUN-line). Finally, I much
prefer letting LLVM synthesize the correct datalayout string from the
triple rather than baking one in here that will just become stale.
llvm-svn: 218856
generic DAG combining of shuffles relevant to x86.
My plan is to fold a bunch of the other DAG combining test cases into
this one, while converting them to use the nice new FileCheck assertion
syntax.
llvm-svn: 218855
a bare-metal triple and have nice BB labels, etc.
No significant change here, just tidying up to have a consistent set of
OS-agnostic vector functionality here.
llvm-svn: 218854
matching and lowering 64-bit insertions.
The first problem was that we weren't looking through bitcasts to
discover that we *could* lower as insertions. Once fixed, we in turn
weren't looking through bitcasts to discover that we could fold a load
into the lowering. Once fixed, we weren't forming a SCALAR_TO_VECTOR
node around the inserted element and instead were passing a scalar to
a DAG node that expected a vector. It turns out there are some patterns
that will "lower" this into the correct asm, but the rest of the X86
backend is very unhappy with such antics.
This should fix a few more edge case regressions I've spotted going
through the regression test suite to enable the new vector shuffle
lowering.
llvm-svn: 218839
Negative FABS of either a scalar or vector should be handled the same way
on x86 with SSE/AVX: a single OR instruction of the FP operand with a
constant to light up the sign bit(s).
http://llvm.org/bugs/show_bug.cgi?id=20578
Differential Revision: http://reviews.llvm.org/D5201
llvm-svn: 218822
test file.
This old test had a bunch of functions that were never even checked. =/
The only thing it really did was to make sure that we did something
reasonable in 32-bit mode with SSE4.1. Adding another run line to the
main vector-sext.ll test seems a better way to do that.
llvm-svn: 218810
of architectures: SSE2, SSSE3, SSE4.1, AVX, and AVX2.
Unfortunately, this exposses the absolute horror of the code we generate
for many of these patterns. Anyone wanting to familiarize themselves
with the x86 backend and improve performance could do a lot of good
sitting down and making these test cases not look so terrible. While the
new vector shuffle code I'm working on well help some, it won't fix all
of the crimes here.
llvm-svn: 218807
These tests are far and away the best sext and zext tests we have for
vectors. I'm going to merge the other similar tests into them and expand
the ISA coverage.
llvm-svn: 218800
script to make them nice and predictable. This will ease updating them
for the new vector shuffle lowering and seeing the delta if any.
llvm-svn: 218795
avx-sext.ll using my new script.
Also add an AVX2 mode to this test.
Part of cleaning up the test suite before enabling the new vector
shuffle lowering. This also highlights some of the abysmal failures of
the old shuffle lowering. Check out those 'pinsrw' and 'pextrw'
sequences!
llvm-svn: 218794
argument of the llvm.dbg.declare/llvm.dbg.value intrinsics.
Previously, DIVariable was a variable-length field that has an optional
reference to a Metadata array consisting of a variable number of
complex address expressions. In the case of OpPiece expressions this is
wasting a lot of storage in IR, because when an aggregate type is, e.g.,
SROA'd into all of its n individual members, the IR will contain n copies
of the DIVariable, all alike, only differing in the complex address
reference at the end.
By making the complex address into an extra argument of the
dbg.value/dbg.declare intrinsics, all of the pieces can reference the
same variable and the complex address expressions can be uniqued across
the CU, too.
Down the road, this will allow us to move other flags, such as
"indirection" out of the DIVariable, too.
The new intrinsics look like this:
declare void @llvm.dbg.declare(metadata %storage, metadata %var, metadata %expr)
declare void @llvm.dbg.value(metadata %storage, i64 %offset, metadata %var, metadata %expr)
This patch adds a new LLVM-local tag to DIExpressions, so we can detect
and pretty-print DIExpression metadata nodes.
What this patch doesn't do:
This patch does not touch the "Indirect" field in DIVariable; but moving
that into the expression would be a natural next step.
http://reviews.llvm.org/D4919
rdar://problem/17994491
Thanks to dblaikie and dexonsmith for reviewing this patch!
Note: I accidentally committed a bogus older version of this patch previously.
llvm-svn: 218787
argument of the llvm.dbg.declare/llvm.dbg.value intrinsics.
Previously, DIVariable was a variable-length field that has an optional
reference to a Metadata array consisting of a variable number of
complex address expressions. In the case of OpPiece expressions this is
wasting a lot of storage in IR, because when an aggregate type is, e.g.,
SROA'd into all of its n individual members, the IR will contain n copies
of the DIVariable, all alike, only differing in the complex address
reference at the end.
By making the complex address into an extra argument of the
dbg.value/dbg.declare intrinsics, all of the pieces can reference the
same variable and the complex address expressions can be uniqued across
the CU, too.
Down the road, this will allow us to move other flags, such as
"indirection" out of the DIVariable, too.
The new intrinsics look like this:
declare void @llvm.dbg.declare(metadata %storage, metadata %var, metadata %expr)
declare void @llvm.dbg.value(metadata %storage, i64 %offset, metadata %var, metadata %expr)
This patch adds a new LLVM-local tag to DIExpressions, so we can detect
and pretty-print DIExpression metadata nodes.
What this patch doesn't do:
This patch does not touch the "Indirect" field in DIVariable; but moving
that into the expression would be a natural next step.
http://reviews.llvm.org/D4919
rdar://problem/17994491
Thanks to dblaikie and dexonsmith for reviewing this patch!
llvm-svn: 218778
that keep cropping up in the regression test suite.
This also addresses one of the issues raised on the mailing list with
failing to form 'movsd' in as many cases as we realistically should.
There will be corresponding patches forthcoming for v4f32 at least. This
was a lot of fuss for a relatively small gain, but all the fuss was on
my end trying different ways of holding the pieces of the x86 fragment
patterns *just right*. Now that it works, the code is reasonably simple.
In the new test cases I'm adding here, v2i64 sticks out as just plain
horrible. I've not come up with any great ideas here other than that it
would be nice to recognize when we're *going* to take a domain crossing
hit and cross earlier to get the decent instructions. At least with AVX
it is slightly less silly....
llvm-svn: 218756
in exposing the scalar value to the broadcast DAG fragment so that we
can catch even reloads and fold them into the broadcast.
This is somewhat magical I'm afraid but seems to work. It is also what
the old lowering did, and I've switched an old test to run both
lowerings demonstrating that we get the same result.
Unlike the old code, I'm not lowering f32 or f64 scalars through this
path when we only have AVX1. The target patterns include pretty heinous
code to re-cast those as shuffles when the scalar happens to not be
spilled because AVX1 provides no broadcast mechanism from registers
what-so-ever. This is terribly brittle. I'd much rather go through our
generic lowering code to get this. If needed, we can add a peephole to
get even more opportunities to broadcast-from-spill-slots that are
exposed post-RA, but my suspicion is this just doesn't matter that much.
llvm-svn: 218734
the same speed as pshufd but we can fold loads into the pmovzx
instructions.
This fixes some regressions that came up in the regression test suite
for the new vector shuffle lowering.
llvm-svn: 218733
VPBROADCAST.
This has the somewhat expected pervasive impact. I don't know why
I forgot about this. Everything seems good with lots of significant
improvements in the tests.
llvm-svn: 218724
cases.
While clearly we don't need the AVX vector width, these ISA extensions
often cause us to select different instructions and we should cover them
even with the narrow vector width.
Also, while here, nuke the stress_test2 contents. There is no reason to
try to FileCheck this entire body when it is mostly a test for
successfully surviving the code generator.
llvm-svn: 218710
shuffle tests to match that used in the script I posted and now used
consistently in 128-bit tests.
Nothing interesting changing here, just using the label name as the
FileCheck label and a slightly more general comment marker consumption
strategy.
llvm-svn: 218709
updating script so that they are more thorough and consistent.
Specific fixes here include:
- Actually test VEX-encoded AVX mnemonics.
- Actually use an SSE 4.1 run to test SSE 4.1 features!
- Correctly check instructions sequences from the start of the function.
- Elide the shuffle operands and comment designator in a consistent way.
- Test all of the architectures instead of just the ones I was motivated
to manually author.
I've gone back through and fixed up any egregious issues I spotted. Let
me know if I missed something you really dislike.
One downside to this is that we're now not as diligently using FileCheck
variables for registers. I would be much more concerned with this if we
had larger register usage, but there just aren't that interesting of
register choices here and most of the registers are constrained by the
ABI. Ultimately, I don't think this is likely to be the maintenance
burden for these tests and updating them again should be staright
forward.
llvm-svn: 218707
Fixed lowering of this intrinsics in case when mask is v2i1 and v4i1.
Now cmp intrinsics lower in the following way:
(i8 (int_x86_avx512_mask_pcmpeq_q_128
(v2i64 %a), (v2i64 %b), (i8 %mask))) ->
(i8 (bitcast
(v8i1 (insert_subvector undef,
(v2i1 (and (PCMPEQM %a, %b),
(extract_subvector
(v8i1 (bitcast %mask)), 0))), 0))))
llvm-svn: 218669
a flawed direction and causing miscompiles. Read on for details.
Fundamentally, the premise of this patch series was to map
VECTOR_SHUFFLE DAG nodes into VSELECT DAG nodes for all blends because
we are going to *have* to lower to VSELECT nodes for some blends to
trigger the instruction selection patterns of variable blend
instructions. This doesn't actually work out so well.
In order to match performance with the existing VECTOR_SHUFFLE
lowering code, we would need to re-slice the blend in order to fit it
into either the integer or floating point blends available on the ISA.
When coming from VECTOR_SHUFFLE (or other vNi1 style VSELECT sources)
this works well because the X86 backend ensures that these types of
operands to VSELECT get sign extended into '-1' and '0' for true and
false, allowing us to re-slice the bits in whatever granularity without
changing semantics.
However, if the VSELECT condition comes from some other source, for
example code lowering vector comparisons, it will likely only have the
required bit set -- the high bit. We can't blindly slice up this style
of VSELECT. Reid found some code using Halide that triggers this and I'm
hopeful to eventually get a test case, but I don't need it to understand
why this is A Bad Idea.
There is another aspect that makes this approach flawed. When in
VECTOR_SHUFFLE form, we have very distilled information that represents
the *constant* blend mask. Converting back to a VSELECT form actually
can lose this information, and so I think now that it is better to treat
this as VECTOR_SHUFFLE until the very last moment and only use VSELECT
nodes for instruction selection purposes.
My plan is to:
1) Clean up and formalize the target pre-legalization DAG combine that
converts a VSELECT with a constant condition operand into
a VECTOR_SHUFFLE.
2) Remove any fancy lowering from VSELECT during *legalization* relying
entirely on the DAG combine to catch cases where we can match to an
immediate-controlled blend instruction.
One additional step that I'm not planning on but would be interested in
others' opinions on: we could add an X86ISD::VSELECT or X86ISD::BLENDV
which encodes a fully legalized VSELECT node. Then it would be easy to
write isel patterns only in terms of this to ensure VECTOR_SHUFFLE
legalization only ever forms the fully legalized construct and we can't
cycle between it and VSELECT combining.
llvm-svn: 218658
nodes, and rely exclusively on its logic. This removes a ton of
duplication from the blend lowering and centralizes it in one place.
One downside is that it requires a bunch of hacks to make this work with
the current legalization framework. We have to manually speculate one
aspect of legalizing VSELECT nodes to get everything to work nicely
because the existing legalization framework isn't *actually* bottom-up.
The other grossness is that we somewhat duplicate the analysis of
constant blends. I'm on the fence here. If reviewers thing this would
look better with VSELECT when it has constant operands dumping over tho
VECTOR_SHUFFLE, we could go that way. But it would be a substantial
change because currently all of the actual blend instructions are
matched via patterns in the TD files based around VSELECT nodes (despite
them not being perfect fits for that). Suggestions welcome, but at least
this removes the rampant duplication in the backend.
llvm-svn: 218600
X86 target-specific DAG combining that tried to convert VSELECT nodes
into VECTOR_SHUFFLE nodes that it "knew" would lower into
immediate-controlled blend nodes.
Turns out, we have perfectly good lowering of all these VSELECT nodes,
and indeed that lowering already knows how to handle lowering through
BLENDI to immediate-controlled blend nodes. The code just wasn't getting
used much because this thing forced the world to go through the vector
shuffle lowering. Yuck.
This also exposes that I was too aggressive in avoiding domain crossing
in v218588 with that lowering -- when the other option is to expand into
two 128-bit vectors, it is worth domain crossing. Restore that behavior
now that we have nice tests covering it.
The test updates here fall into two camps. One is where previously we
ended up with an unsigned encoding of the blend operand and now we get
a signed encoding. In most of those places there were elaborate comments
explaining exactly what these operands really mean. Rather than that,
just switch these tests to use the nicely decoded comments that make it
obvious that the final shuffle matches.
The other updates are just removing pointless domain crossing by
blending integers with PBLENDW rather than BLENDPS.
llvm-svn: 218589
AVX-512.
There is no interesting logic yet. Everything ends up eventually
delegating to the generic code to split the vector and shuffle the
halves. Interestingly, that logic does a significantly better job of
lowering all of these types than the generic vector expansion code does.
Mostly, it lets most of the cases fall back to nice AVX2 code rather
than all the way back to SSE code paths.
Step 2 of basic AVX-512 support in the new vector shuffle lowering. Next
up will be to incrementally add direct support for the basic instruction
set to each type (adding tests first).
llvm-svn: 218585
vectors.
Someone will need to build the AVX512 lowering, which should follow
AVX1 and AVX2 *very* closely for AVX512F and AVX512BW resp. I've added
a dummy test which is a port of the v8f32 and v8i32 tests from AVX and
AVX2 to v8f64 and v8i64 tests for AVX512F and AVX512BW. Hopefully this
is enough information for someone to implement proper lowering here. If
not, I'll be happy to help, but right now the AVX-512 support isn't
a priority for me.
llvm-svn: 218583
lowerings.
This was hopelessly broken. First, the x86 backend wants '-1' to be the
element value representing true in a boolean vector, and second the
operand order for VSELECT is backwards from the actual x86 instructions.
To make matters worse, the backend is just using '-1' as the true value
to get the high bit to be set. It doesn't actually symbolically map the
'-1' to anything. But on x86 this isn't quite how it works: there *only*
the high bit is relevant. As a consequence weird non-'-1' values like
0x80 actually "work" once you flip the operands to be backwards.
Anyways, thanks to Hal for helping me sort out what these *should* be.
llvm-svn: 218582
new vector shuffle target DAG combines -- it helps to actually test for
the value you want rather than just using an integer in a boolean
context.
Have I mentioned that I loathe implicit conversions recently? :: sigh ::
llvm-svn: 218576
of widening masks.
We can't widen a zeroing mask unless both elements that would be merged
are either zeroed or undef. This is the only way to widen a mask if it
has a zeroed element.
Also clean up the code here by ordering the checks in a more logical way
and by using the symoblic values for undef and zero. I'm actually torn
on using the symbolic values because the existing code is littered with
the assumption that -1 is undef, and moreover that entries '< 0' are the
special entries. While that works with the values given to these
constants, using the symbolic constants actually makes it a bit more
opaque why this is the case.
llvm-svn: 218575
and in the target shuffle combining when trying to widen vector
elements.
Previously only one of these was correct, and we didn't correctly
propagate zeroing target shuffle masks (which have a different sentinel
value from undef in non- target shuffle masks now). This isn't just
a missed optimization, this caused us to drop zeroing shuffles on the
floor and miscompile code. The added test case is one example of that.
There are other fixes to the test suite as a consequence of this as well
as restoring the undef elements in some of the masks that were lost when
I brought sanity to the actual *value* of the undef and zero sentinels.
I've also just cleaned up some of the PSHUFD and PSHUFLW and PSHUFHW
combining code, but that code really needs to go. It was a nice initial
attempt, but it isn't very principled and the recursive shuffle combiner
is much more powerful.
llvm-svn: 218562
to significantly more sane sentinels. Notably, everywhere else in the
backend's representation of shuffles uses '-1' to represent undef. The
target shuffle masks really shouldn't diverge from that, especially as
in a few places they are manipulated by shared code.
This causes us to lose some undef lanes in various test masks. I want to
get these back, but technically it isn't invalid and there are a *lot*
of bugs here so I want to try to establish a saner baseline for fixing
some of the bugs by aligning the specific senitnel values used.
llvm-svn: 218561
that managed to elude all of my fuzz testing historically. =/
Something changed to allow this code path to actually be exercised and
it was doing bad things. It is especially heavily exercised by the
patterns that emerge when doing AVX shuffles that end up lowered through
the 128-bit code path.
llvm-svn: 218540
layer of tie-breaking sorting, it really helps to check that you're in
a tie first. =] Otherwise the whole thing cycles infinitely. Test case
added, another one found through fuzz testing.
llvm-svn: 218523
AVX support.
New test cases included. Note that none of the existing test cases
covered these buggy code paths. =/ Also, it is clear from this that
SHUFPS and SHUFPD are the most bug prone shuffle instructions in x86. =[
These were all detected by fuzz-testing. (I <3 fuzz testing.)
llvm-svn: 218522
Machine Sink uses loop depth information to select between successors BBs to
sink machine instructions into, where BBs within smaller loop depths are
preferable. This patch adds support for choosing between successors by using
profile information from BlockFrequencyInfo instead, whenever the information
is available.
Tested it under SPEC2006 train (average of 30 runs for each program); ~1.5%
execution speedup in average on x86-64 darwin.
<rdar://problem/18021659>
llvm-svn: 218472
Summary:
I originally tried doing this specifically for X86 in the backend in D5091,
but it was rather brittle and generally running too late to be general.
Furthermore, other targets may want to implement similar optimizations.
So I reimplemented it at the IR-level, fitting it into AtomicExpandPass
as it interacts with that pass (which could not be cleanly done before
at the backend level).
This optimization relies on a new target hook, which is only used by X86
for now, as the correctness of the optimization on other targets remains
an open question. If it is found correct on other targets, it should be
trivial to enable for them.
Details of the optimization are discussed in D5091.
Test Plan: make check-all + a new test
Reviewers: jfb
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5422
llvm-svn: 218455
v4f64 and v8f32 shuffles when they are lane-crossing. We have fully
general lane-crossing permutation functions in AVX2 that make this easy.
Part of this also changes exactly when and how these vectors are split
up when we don't have AVX2. This isn't always a win but it usually is
a win, so on the balance I think its better. The primary regressions are
all things that just need to be fixed anyways such as modeling when
a blend can be completely accomplished via VINSERTF128, etc.
Also, this highlights one of the few remaining big features: we do
a really poor job of inserting elements into AVX registers efficiently.
This completes almost all of the big tricks I have in mind for AVX2. The
only things left that I plan to add:
1) element insertion smarts
2) palignr and other fairly specialized lowerings when they happen to
apply
llvm-svn: 218449
256-bit vectors with lane-crossing.
Rather than immediately decomposing to 128-bit vectors, try flipping the
256-bit vector lanes, shuffling them and blending them together. This
reduces our worst case shuffle by a pretty significant margin across the
board.
llvm-svn: 218446
lowering where it only used the mask of the low 128-bit lane rather than
the entire mask.
This allows the new lowering to correctly match the unpack patterns for
v8i32 vectors.
For reference, the reason that we check for the the entire mask rather
than checking the repeated mask is because the repeated masks don't
abide by all of the invariants of normal masks. As a consequence, it is
safer to use the full mask with functions like the generic equivalence
test.
llvm-svn: 218442
lowering.
This completes the basic AVX2 feature support, but there are still some
improvements I'd like to do to really get the last mile of performance
here.
llvm-svn: 218440
I made a mistake in the previous commit and produced the wrong pattern.
Fix that. Also make one more shuffle pattern byte-based rather than
word-based, and add two more blend patterns.
llvm-svn: 218439
shuffles rather than word shuffles.
As you might guess, these were built starting from the word shuffle test
cases and I failed to properly port a bunch of them and left them as
widened word shuffle test cases. We still have a couple of tests that
check our ability to widen shuffles, but now we will test the actual
byte shuffle quite a bit better.
llvm-svn: 218438
missing test cases for it.
Unsurprisingly, without test cases, there were bugs here. Surprisingly,
this bug wasn't caught at compile time. Yep, there is an X86ISD::BLENDV.
It isn't wired to anything. Oops. I'll fix than next.
llvm-svn: 218434
lowering.
This also implements the fancy blend lowering for v16i16 using AVX2 and
teaches the X86 backend to print shuffle masks for 256-bit PSHUFB
and PBLENDW instructions. It also makes the mask decoding correct for
PBLENDW instructions. The yaks, they are legion.
Tests are updated accordingly. There are some missing tests for the
VBLENDVB lowering, but I'll add those in a follow-up as this commit has
accumulated enough cruft already.
llvm-svn: 218430
pool data being loaded into a vector register.
The comments take the form of:
# ymm0 = [a,b,c,d,...]
# xmm1 = <x,y,z...>
The []s are used for generic sequential data and the <>s are used for
specifically ConstantVector loads. Undef elements are printed as the
letter 'u', integers in decimal, and floating point values as floating
point values. Suggestions on improving the formatting or other aspects
of the display are very welcome.
My primary use case for this is to be able to FileCheck test masks
passed to vector shuffle instructions in-register. It isn't fantastic
for that (no decoding special zeroing semantics or other tricks), but it
at least puts the mask onto an instruction line that could reasonably be
checked. I've updated many of the new vector shuffle lowering tests to
leverage this in their test cases so that we're actually checking the
shuffle masks remain as expected.
Before implementing this, I tried a *bunch* of different approaches.
I looked into teaching the MCInstLower code to scan up the basic block
and find a definition of a register used in a shuffle instruction and
then decode that, but this seems incredibly brittle and complex.
I talked to Hal a lot about the "right" way to do this: attach the raw
shuffle mask to the instruction itself in some form of unencoded
operands, and then use that to emit the comments. I still think that's
the optimal solution here, but it proved to be beyond what I'm up for
here. In particular, it seems likely best done by completing the
plumbing of metadata through these layers and attaching the shuffle mask
in metadata which could have fully automatic dropping when encoding an
actual instruction.
llvm-svn: 218377
the native AVX2 instructions.
Note that the test case is really frustrating here because VPERMD
requires the mask to be in the register input and we don't produce
a comment looking through that to the constant pool. I'm going to
attempt to improve this in a subsequent commit, but not sure if I will
succeed.
llvm-svn: 218347
detection. It was incorrectly handling undef lanes by actually treating
an undef lane in the first 128-bit lane as a *numeric* shuffle value.
Fortunately, this almost always DTRT and disabled detecting repeated
patterns. But not always. =/ This patch introduces a much more
principled approach and fixes the miscompiles I spotted by inspection
previously.
llvm-svn: 218346
shuffles using the AVX2 instructions. This is the first step of cutting
in real AVX2 support.
Note that I have spotted at least one bug in the test cases already, but
I suspect it was already present and just is getting surfaced. Will
investigate next.
llvm-svn: 218338
add VPBLENDD to the InstPrinter's comment generation so we get nice
comments everywhere.
Now that we have the nice comments, I can see the bug introduced by
a silly typo in the commit that enabled VPBLENDD, and have fixed it. Yay
tests that are easy to inspect.
llvm-svn: 218335
Summary:
AtomicExpand already had logic for expanding wide loads and stores on LL/SC
architectures, and for expanding wide stores on CmpXchg architectures, but
not for wide loads on CmpXchg architectures. This patch fills this hole,
and makes use of this new feature in the X86 backend.
Only one functionnal change: we now lose the SynchScope attribute.
It is regrettable, but I have another patch that I will submit soon that will
solve this for all of AtomicExpand (it seemed better to split it apart as it
is a different concern).
Test Plan: make check-all (lots of tests for this functionality already exist)
Reviewers: jfb
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D5404
llvm-svn: 218332
VPBLENDD where appropriate even on 128-bit vectors.
According to Agner's tables, this instruction is significantly higher
throughput (can execute on any port) on Haswell chips so we should
aggressively try to form it when available.
Sadly, this loses our delightful shuffle comments. I'll add those back
for VPBLENDD next.
llvm-svn: 218322
undef in the shuffle mask. This shows up when we're printing comments
during lowering and we still have an IR-level constant hanging around
that models undef.
A nice consequence of this is *much* prettier test cases where the undef
lanes actually show up as undef rather than as a particular set of
values. This also allows us to print shuffle comments in cases that use
undef such as the recently added variable VPERMILPS lowering. Now those
test cases have nice shuffle comments attached with their details.
The shuffle lowering for PSHUFB has been augmented to use undef, and the
shuffle combining has been augmented to comprehend it.
llvm-svn: 218301
trick that I missed.
VPERMILPS has a non-immediate memory operand mode that allows it to do
asymetric shuffles in the two 128-bit lanes. Use this rather than two
shuffles and a blend.
However, it turns out the variable shuffle path to VPERMILPS (and
VPERMILPD, although that one offers no functional differenc from the
immediate operand other than variability) wasn't even plumbed through
codegen. Do such plumbing so that we can reasonably emit
a variable-masked VPERMILP instruction. Also plumb basic comment parsing
and printing through so that the tests are reasonable.
There are still a few tests which don't show the shuffle pattern. These
are tests with undef lanes. I'll teach the shuffle decoding and printing
to handle undef mask entries in a follow-up. I've looked at the masks
and they seem reasonable.
llvm-svn: 218300
These are just test cases, no actual code yet. This establishes the
baseline fallback strategy we're starting from on AVX2 and the expected
lowering we use on AVX1.
Also, these test cases are very much generated. I've manually crafted
the specific pattern set that I'm hoping will be useful at exercising
the lowering code, but I've not (and could not) manually verify *all* of
these. I've spot checked and they seem legit to me.
As with the rest of vector shuffling, at a certain point the only really
useful way to check the correctness of this stuff is through fuzz
testing.
llvm-svn: 218267
We generate broadcast instructions on CPUs with AVX2 to load some constant splat vectors.
This patch should preserve all existing behavior with regular optimization levels,
but also use splats whenever possible when optimizing for *size* on any CPU with AVX or AVX2.
The tradeoff is up to 5 extra instruction bytes for the broadcast instruction to save
at least 8 bytes (up to 31 bytes) of constant pool data.
Differential Revision: http://reviews.llvm.org/D5347
llvm-svn: 218263
Bob Wilson