This adds Intel's Knights Mill CPU to valid CPU names for the backend. For now its an alias of "knl", but ultimately we need to support AVX5124FMAPS and AVX5124VNNIW instruction sets for it.
Differential Revision: https://reviews.llvm.org/D38811
llvm-svn: 315722
There's no advantage to using these instructions when they aren't masked. This enables some additional execution domain switching without needing to update the table.
llvm-svn: 315674
Legalization of fp128 assumes things that we should have asserts for,
so that's another potential improvement.
Differential Revision: https://reviews.llvm.org/D38771
llvm-svn: 315485
Eg:
insert v4i32 V, (v2i16 X), 2 --> shuffle v8i16 V', X', {0,1,2,3,8,9,6,7}
This is a generalization of the IR fold in D38316 to handle insertion into a non-undef vector.
We may want to abandon that one if we can't find value in squashing the more specific pattern sooner.
We're using the existing legal shuffle target hook to avoid AVX512 horror with vXi1 shuffles.
There may be room for improvement in the shuffle lowering here, but that would be follow-up work.
Differential Revision: https://reviews.llvm.org/D38388
llvm-svn: 315460
Adding this test files now so after another commit that will add a new pattern for
TESTM and TESTNM instructions will show the improvemnts that have been done.
Change-Id: If3908b7f91897d764053312365a2bc1de78b291d
llvm-svn: 315443
This patch ensures that the rule:
fold (zext (load x)) -> (zext (truncate (zextload x)))
propagates the SDLoc of the load to the zextload.
<rdar://problem/33755881>
llvm-svn: 315340
NFC.
Updated 6 regression tests to differentiate between HASWELL and SKYLAKE scheduling information.
The fix is in preparation of a patch to update the information of the Skylake Client scheduling to include the appropriate load and store latencies.
Reviewers: zvi, RKSimon
Differential Revision: https://reviews.llvm.org/D38685
Change-Id: Ifc6b98d9eaf266913698f24c766fd994fc977555
llvm-svn: 315291
Summary:
This suppresses the generation of .Lcfi labels in our textual assembler.
It was annoying that this generated cascading .Lcfi labels:
llc foo.ll -o - | llvm-mc | llvm-mc
After three trips through MCAsmStreamer, we'd have three labels in the
output when none are necessary. We should only bother creating the
labels and frame data when making a real object file.
This supercedes D38605, which moved the entire .seh_ implementation into
MCObjectStreamer.
This has the advantage that we do more checking when emitting textual
assembly, as a minor efficiency cost. Outputting textual assembly is not
performance critical, so this shouldn't matter.
Reviewers: majnemer, MatzeB
Subscribers: qcolombet, nemanjai, javed.absar, eraman, hiraditya, JDevlieghere, llvm-commits
Differential Revision: https://reviews.llvm.org/D38638
llvm-svn: 315259
We end up creating COPY's that are either truncating/extending and this
should be illegal.
https://reviews.llvm.org/D37640
Patch for X86 and ARM by igorb, rovka
llvm-svn: 315240
Summary:
On behalf of julia.koval@intel.com
The patch transforms canonical version of unsigned saturation, which is sub(max(a,b),a) or sub(a,min(a,b)) to special psubus insturuction on targets, which support it(8bit and 16bit uints).
umax(a,b) - b -> subus(a,b)
a - umin(a,b) -> subus(a,b)
There is also extra case handled, when right part of sub is 32 bit and can be truncated, using UMIN(this transformation was discussed in https://reviews.llvm.org/D25987).
The example of special case code:
```
void foo(unsigned short *p, int max, int n) {
int i;
unsigned m;
for (i = 0; i < n; i++) {
m = *--p;
*p = (unsigned short)(m >= max ? m-max : 0);
}
}
```
Max in this example is truncated to max_short value, if it is greater than m, or just truncated to 16 bit, if it is not. It is vaid transformation, because if max > max_short, result of the expression will be zero.
Here is the table of types, I try to support, special case items are bold:
| Size | 128 | 256 | 512
| ----- | ----- | ----- | -----
| i8 | v16i8 | v32i8 | v64i8
| i16 | v8i16 | v16i16 | v32i16
| i32 | | **v8i32** | **v16i32**
| i64 | | | **v8i64**
Reviewers: zvi, spatel, DavidKreitzer, RKSimon
Reviewed By: zvi
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D37534
llvm-svn: 315237
We believe that despite AMD's documentation, that they really do support all 32 comparision predicates under AVX.
Differential Revision: https://reviews.llvm.org/D38609
llvm-svn: 315201
Summary:
We currently disable some converting of shuffles to MOVSS/MOVSD during legalization if SSE41 is enabled. But later during shuffle combining we go back to prefering MOVSS/MOVSD.
Additionally we have patterns that look for BLENDIs to detect scalar arithmetic operations. I believe due to the combining using MOVSS/MOVSD these are unnecessary.
Interestingly, we still codegen blend instructions even though lowering/isel emit movss/movsd instructions. Turns out machine CSE commutes them to blend, and then commuting those blends back into blends that are equivalent to the original movss/movsd.
This patch fixes the inconsistency in legalization to prefer MOVSS/MOVSD. The one test change was caused by this change. The problem is that we have integer types and are mostly selecting integer instructions except for the shufps. This shufps forced the execution domain, but the vpblendw couldn't have its domain changed with a naive instruction swap. We could fix this by special casing VPBLENDW based on the immediate to widen the element type.
The rest of the patch is removing all the excess scalar patterns.
Long term we should probably add isel patterns to make MOVSS/MOVSD emit blends directly instead of relying on the double commute. We may also want to consider emitting movss/movsd for optsize. I also wonder if we should still use the VEX encoded blendi instructions even with AVX512. Blends have better throughput, and that may outweigh the register constraint.
Reviewers: RKSimon, zvi
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38023
llvm-svn: 315181
Adding the scheduling information for the SkylakeServer (SKX) target.
This patch adds the instruction scheduling information for the SkylakeServer (SKX) architecture target by adding the file X86SchedSkylakeServer.td located under the X86 Target.
We used the scheduling information retrieved from the Skylake architects in order to create the file.
The scheduling information includes latency, number of micro-Ops and used ports by each SKL instruction.
The patch continues the scheduling replacement and insertion effort started with the SNB target in r310792, the HSW target in r311879 and the SkylakeClient (SKL) target in rL313613.
Please expect some performance fluctuations due to code alignment effects.
Reviewers: zvi, RKSimon, craig.topper, chandlerc, aymanmu
Differential Revision: https://reviews.llvm.org/D38443
Change-Id: I5c228fcc09e9e5a99b6116e62b356c4f9b971185
llvm-svn: 315175
Recognise cases when we can merge the shuffles with their horizontal (HADD/HSUB/PACK) instruction inputs.
Replaces an older implementation which performed some of this during lowering, expanding an existing target shuffle combine stage instead.
Differential Revision: https://reviews.llvm.org/D38506
llvm-svn: 315150
Patch to fix ternlog instructions with a folded
broadcast. The broadcast decorator, e.g. {1toX}, was missing.
Differential Revision: https://reviews.llvm.org/D38649
llvm-svn: 315122
The code which lowers BUILD_VECTOR of consecutive loads into a single vector
load doesn't update chains properly. As a result the vector load can be
reordered with the store to the same location.
The current code in EltsFromConsecutiveLoads only updates the chain following
the first load. The fix is to update the chains following all the loads
comprising the vector.
This is a fix for PR10114.
Reviewed By: niravd
Differential Revision: https://reviews.llvm.org/D38547
llvm-svn: 314988
It broke the Chromium / SQLite build; see PR34830.
> Summary:
> 1/ Operand folding during complex pattern matching for LEAs has been
> extended, such that it promotes Scale to accommodate similar operand
> appearing in the DAG.
> e.g.
> T1 = A + B
> T2 = T1 + 10
> T3 = T2 + A
> For above DAG rooted at T3, X86AddressMode will no look like
> Base = B , Index = A , Scale = 2 , Disp = 10
>
> 2/ During OptimizeLEAPass down the pipeline factorization is now performed over LEAs
> so that if there is an opportunity then complex LEAs (having 3 operands)
> could be factored out.
> e.g.
> leal 1(%rax,%rcx,1), %rdx
> leal 1(%rax,%rcx,2), %rcx
> will be factored as following
> leal 1(%rax,%rcx,1), %rdx
> leal (%rdx,%rcx) , %edx
>
> 3/ Aggressive operand folding for AM based selection for LEAs is sensitive to loops,
> thus avoiding creation of any complex LEAs within a loop.
>
> Reviewers: lsaba, RKSimon, craig.topper, qcolombet, jmolloy
>
> Reviewed By: lsaba
>
> Subscribers: jmolloy, spatel, igorb, llvm-commits
>
> Differential Revision: https://reviews.llvm.org/D35014
llvm-svn: 314919
This patch redefines the MOVSS/MOVSD instructions to take VR128 as its second input. This allows the MOVSS/SD->BLEND commute to work without requiring a COPY to be inserted.
This should fix PR33079
Overall this looks to be an improvement in the generated code. I haven't checked the EXPENSIVE_CHECKS build but I'll do that and update with results.
Differential Revision: https://reviews.llvm.org/D38449
llvm-svn: 314914
Summary:
1/ Operand folding during complex pattern matching for LEAs has been
extended, such that it promotes Scale to accommodate similar operand
appearing in the DAG.
e.g.
T1 = A + B
T2 = T1 + 10
T3 = T2 + A
For above DAG rooted at T3, X86AddressMode will no look like
Base = B , Index = A , Scale = 2 , Disp = 10
2/ During OptimizeLEAPass down the pipeline factorization is now performed over LEAs
so that if there is an opportunity then complex LEAs (having 3 operands)
could be factored out.
e.g.
leal 1(%rax,%rcx,1), %rdx
leal 1(%rax,%rcx,2), %rcx
will be factored as following
leal 1(%rax,%rcx,1), %rdx
leal (%rdx,%rcx) , %edx
3/ Aggressive operand folding for AM based selection for LEAs is sensitive to loops,
thus avoiding creation of any complex LEAs within a loop.
Reviewers: lsaba, RKSimon, craig.topper, qcolombet, jmolloy
Reviewed By: lsaba
Subscribers: jmolloy, spatel, igorb, llvm-commits
Differential Revision: https://reviews.llvm.org/D35014
llvm-svn: 314886
The previous version didn't work if the jump table base address didn't
fit in 32 bit, since it was encoded as an immediate offset. And in case
the jump table is encoded as 32 bit label differences, we need to
load and add them to the table base first.
This solves the first half of the issues mentioned in PR34720.
Also fix some of the errors pointed out by -verify-machineinstrs, by
using GR32_NOSPRegClass.
Differential Revision: https://reviews.llvm.org/D38333
llvm-svn: 314876
If the upper bits of a truncation shuffle patterns have at least the minimum number of sign/zero bits on their inputs then we can safely use PACKSS/PACKUS as shuffles.
Partial fix for https://bugs.llvm.org/show_bug.cgi?id=34773
Differential Revision: https://reviews.llvm.org/D38472
llvm-svn: 314788
This makes sure the LSDA pointer isn't truncated to 32 bit.
Make LowerINTRINSIC_WO_CHAIN a member function instead of a static
function, so that it can use the getGlobalWrapperKind method.
This solves the second half of the issues mentioned in PR34720.
Differential Revision: https://reviews.llvm.org/D38343
llvm-svn: 314767
Issues addressed since original review:
- Avoid bug in regalloc greedy/machine verifier when forwarding to use
in an instruction that re-defines the same virtual register.
- Fixed bug when forwarding to use in EarlyClobber instruction slot.
- Fixed incorrect forwarding to register definitions that showed up in
explicit_uses() iterator (e.g. in INLINEASM).
- Moved removal of dead instructions found by
LiveIntervals::shrinkToUses() outside of loop iterating over
instructions to avoid instructions being deleted while pointed to by
iterator.
- Fixed ARMLoadStoreOptimizer bug exposed by this change in r311907.
- The pass no longer forwards COPYs to physical register uses, since
doing so can break code that implicitly relies on the physical
register number of the use.
- The pass no longer forwards COPYs to undef uses, since doing so
can break the machine verifier by creating LiveRanges that don't
end on a use (since the undef operand is not considered a use).
[MachineCopyPropagation] Extend pass to do COPY source forwarding
This change extends MachineCopyPropagation to do COPY source forwarding.
This change also extends the MachineCopyPropagation pass to be able to
be run during register allocation, after physical registers have been
assigned, but before the virtual registers have been re-written, which
allows it to remove virtual register COPY LiveIntervals that become dead
through the forwarding of all of their uses.
llvm-svn: 314729
I continue to support different VF interleaved and in this pass for this patch,
I added the vf64 stride3 support for both load and store.
I also added support fot the stride4 store.
Reviewers:
1. zvi
2. dorit
3. igorb
4. guyblank
Differential Revision: https://reviews.llvm.org/D37687
Change-Id: I3d238efedf217d1768b348d710de1efa2f19d27b
llvm-svn: 314651
Summary:
Intel documentation shows the memory operand as the first operand. But we currently treat it as the second operand. Conceptually the order doesn't matter since it doesn't write memory. We have aliases to parse with the operands in either order and the isel matching is commutable.
For the register®ister form order does matter for the assembly parser. PR22995 was previously filed and fixed by changing the register®ister form from MRMSrcReg to MRMDestReg to match gas. Ideally the memory form should match by using MRMDestMem.
I believe this supercedes D38025 which was trying to switch the register®ister form back to pre-PR22995.
Reviewers: aymanmus, RKSimon, zvi
Reviewed By: aymanmus
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38120
llvm-svn: 314639
NFC.
Added code gen regression tests for avx512 instructions scheduling called avx512-schedule.ll and
avx512-shuffle-schedule.ll.
This patch is in preparation of a larger patch of adding all SKX instruction scheduling and therefore
the scheduling for the avx512 instructions are still missing.
Reviewers: zvi, delena, RKSimon, igorb
Differential Revision: https://reviews.llvm.org/D38035
Change-Id: I792762763127a921b9e13684b58af03646536533
llvm-svn: 314594
Implemented by splitting into two v32i8 mulhu/mulhs and concatenating the results.
Differential Revision: https://reviews.llvm.org/D38307
llvm-svn: 314584
Added additional tests for vector multiplications with multipliers that are:
* powers of 2 displaced by 1,
* product of a power of 2 displaced by one with another power of 2.
Patch by @pacxx (Michael Haidl)
Differential Revision: https://reviews.llvm.org/D38350
llvm-svn: 314504
Summary:
X86ISelDAGToDAG tries to analyze ANDs compared with 0 to optimize to narrower immediates using subregisters.
I don't think we should be optimizing to 16-bit test instructions. It goes against our normal behavior of promoting i16 operations to i32. It only saves one byte due to the need to add a 0x66 prefix. I think it would also be subject to a length changing prefix penalty in the decoders on Intel CPUs.
Reviewers: RKSimon, zvi, spatel
Reviewed By: spatel
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38273
llvm-svn: 314474
If we have BWI, we can truncate in a much simpler way by using vpmovwb. This even works without VLX by using the wider zmm->ymm truncate with a subvector extract.
Differential Revision: https://reviews.llvm.org/D38375
llvm-svn: 314457
Summary: If we have BWI instructions we can widen to v32i16 to do the multiply instead of splitting.
Reviewers: RKSimon, spatel, zvi
Reviewed By: zvi
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D38305
llvm-svn: 314432
Previously we were using one of the subvector indices twice. The included test case causes an assert without this change.
Thanks to Simon Pilgrim for catching this.
llvm-svn: 314429
This commit allows the outliner to avoid saving and restoring the link register
on AArch64 when it is dead within an entire class of candidates.
This introduces changes to the way the outliner interfaces with the target.
For example, the target now interfaces with the outliner using a
MachineOutlinerInfo struct rather than by using getOutliningCallOverhead and
getOutliningFrameOverhead.
This also improves several comments on the outliner's cost model.
https://reviews.llvm.org/D36721
llvm-svn: 314341
Summary:
According to https://gcc.gnu.org/wiki/SplitStacks, the linker expects a zero-sized .note.GNU-split-stack section if split-stack is used (and also .note.GNU-no-split-stack section if it also contains non-split-stack functions), so it can handle the cases where a split-stack function calls non-split-stack function.
This change adds the sections if needed.
Fixes PR #34670.
Reviewers: thanm, rnk, luqmana
Reviewed By: rnk
Subscribers: llvm-commits
Patch by Cherry Zhang <cherryyz@google.com>
Differential Revision: https://reviews.llvm.org/D38051
llvm-svn: 314335
In some cases the result psadbw is smaller than the type of the add that started the match. Currently in these cases we are using a smaller add and inserting the result.
If we instead combine the psadbw with zeros and use the full size add we can take advantage of implicit zeroing we get if we emit a narrower move before the add.
In a future patch, I want to make isel aware that the psadbw itself already zeroed the upper bits and remove the move entirely.
Differential Revision: https://reviews.llvm.org/D37453
llvm-svn: 314331
NFC.
Updated 8 regression tests to use -mattr instead of -mcpu flag as follows:
-mcpu=knl --> -mattr=+avx512f
-mcpu=skx --> -mattr=+avx512f,+avx512bw,+avx512vl,+avx512dq
The updates are as part of the preparation of a large commit to add all instruction scheduling for the SKX target.
Reviewers: delena, zvi, RKSimon
Differential Revision: https://reviews.llvm.org/D38222
Change-Id: I2381c9b5bb75ecacfca017243c22d054f6eddd14
llvm-svn: 314306
As commented on D37849 and rL313547, AVX1 targets were missing a chance to use vmovmskpd for v4f64/v4i64 results for bool vector bitcasts
llvm-svn: 314293
This is necessary, but not sufficient, for having working SJLJ exception
handling on x86_64.
Differential Revision: https://reviews.llvm.org/D38254
llvm-svn: 314277
The callsite value is already stored indexed from 0 in
the _Unwind_Context struct. When accessed via the functions
_Unwind_GetIP and _Unwind_SetIP, the value is indexed from 1,
but those functions handle the offseting. When reading directly
from the struct here, we shouldn't subtract 1.
This matches the code generated by the ARM target, where SJLJ
exception handling is used by default on iOS.
This makes clang-built object files for 32 bit x86 mingw work when
linked with libgcc/libstdc++.
Differential Revision: https://reviews.llvm.org/D38251
llvm-svn: 314276
This matches the types of the struct members defined in
lib/CodeGen/SjLjEHPrepare.cpp, and the definition of this struct in libgcc.
Differential Revision: https://reviews.llvm.org/D38248
llvm-svn: 314275
This patch expands the support of lowerInterleavedStore to {8|16|32}x8i stride 3.
LLVM creates suboptimal shuffle code-gen for AVX2. In overall, this patch is a specific fix for the pattern (Strid=3 VF={8|16|32}) .
This patch is part two of two patches and it covers the store (interlevaed) side.
The patch goal is to optimize the following sequence:
a0 a1 a2 a3 a4 a5 a6 a7
b0 b1 b2 b3 b4 b5 b6 b7
c0 c1 c2 c3 c4 c5 c6 c7
into
a0 b0 c0 a1 b1 c1 a2 b2
c2 a3 b3 c3 a4 b4 c4 a5
b5 c5 a6 b6 c6 a7 b7 c7
Reviewers:
zvi
guyblank
dorit
Ayal
Differential Revision: https://reviews.llvm.org/D37117
Change-Id: I56ced8bcbea809a37654060771911ade20246ccc
llvm-svn: 314234
Removing X86 broadcast(f/i)32x2 intrinsics from llvm.
Adding autoUpgrade support.
Moving matching tests from avx512dq-intrinsics.ll to avx512dq-intrinsics-upgrade.ll and from avx512dqvl-intrinsics.ll to avx512dqvl-intrinsics-upgrade.ll.
Differential Revision: https://reviews.llvm.org/D38220
llvm-svn: 314195
R12 is used for the SwiftError parameter. It is no longer a CSR as it
is used for transfer the SwiftError, and the caller must preserve it if
they need to.
llvm-svn: 314165
As far as I know SUBREG_TO_REG is stating that the upper bits are 0. But if we are just converting the GR32 with no checks, then we have no reason to say the upper bits are 0.
I don't really know how to test this today since I can't find anything that looks that closely at SUBREG_TO_REG. The test changes here seems to be some perturbance of register allocation.
Differential Revision: https://reviews.llvm.org/D38001
llvm-svn: 314152
This teach simplifyDemandedBits to handle constant splat vector shifts.
This required changing some uses of getZExtValue to getLimitedValue since we can't rely on legalization using getShiftAmountTy for the shift amount.
I believe there may have been a bug in the ((X << C1) >>u ShAmt) handling where we didn't check if the inner shift was too large. I've fixed that here.
I had to add new patterns to ARM because the zext/sext the patterns were trying to look for got turned into an any_extend with this patch. Happy to split that out too, but not sure how to test without this change.
Differential Revision: https://reviews.llvm.org/D37665
llvm-svn: 314139
Summary:
This code iterates the 'Orders' vector in parallel with the DbgValue
list, emitting all DBG_VALUEs that occurred between the last IR order
insertion point and the next insertion point. This assumes the
SDDbgValue list is sorted in IR order, which it usually is. However, it
is not sorted when a node with a debug value is replaced with another
one. When this happens, TransferDbgValues is called, and the new value
is added to the end of the list.
The problem can be solved by stably sorting the list by IR order.
Reviewers: aprantl, Ka-Ka
Reviewed By: aprantl
Subscribers: MatzeB, hiraditya, llvm-commits
Differential Revision: https://reviews.llvm.org/D38197
llvm-svn: 314114
This patch expands the support of lowerInterleavedStore to 8x8i stride 4.
LLVM creates suboptimal shuffle code-gen for AVX2.
In overall, this patch is a specific fix for the pattern (Strid=4 VF=8) and we plan to include more patterns in the future.
The patch goal is to optimize the following sequence:
At the end of the computation, we have xmm2, xmm0, xmm12 and xmm3 holding
each 8 chars:
c0, c1, , c7
m0, m1, , m7
y0, y1, , y7
k0, k1, ., k7
And these need to be transposed/interleaved and stored like so:
c0 m0 y0 k0 c1 m1 y1 k1 c2 m2 y2 k2 c3 m3 y3 k3 ....
Reviewers
DavidKreitzer
Farhana
zvi
igorb
guyblank
RKSimon
Ayal
Differential Revision: https://reviews.llvm.org/D36058
Change-Id: I3cc5c2ca5d6318901c192a4428493b99ef424c32
llvm-svn: 314109
This required changing the ISD opcode for these instructions to have the commutable operands first and the addend last. This way tablegen can autogenerate the additional patterns for us.
llvm-svn: 314083
This patch acts as a reverse to combineBitcastvxi1 - bitcasting a scalar integer to a boolean vector and extending it 'in place' to the requested legal type.
Currently this doesn't handle AVX512 at all - but the current mask register approach is lacking for some cases.
Differential Revision: https://reviews.llvm.org/D35320
llvm-svn: 314076
We use a v16i32/v16f32 compare instead and truncate the result. We already did this for the unmasked version, but were missing the version with 'and'.
llvm-svn: 314072
This is a follow-up from D38181 (r314023). We have to put 64-bit
constants into a register using a separate instruction, so we
should try harder to avoid that.
From what I see, we're not likely to encounter this pattern in the
DAG because the upstream setcc combines from this don't (usually?)
produce this pattern. If we fix that, then this will become more
relevant. Since the cost of handling this case is just loosening
the predicate of the existing fold, we might as well do it now.
llvm-svn: 314064
The (non-)obvious win comes from saving 3 bytes by using the 0x83 'and' opcode variant instead of 0x81.
There are also better improvements based on known-bits that allow us to eliminate the mask entirely.
As noted, this could be extended. There are potentially other wins from always shifting first, but doing
that reveals a tangle of problems in other pattern matching. We do this transform generically in
instcombine, but we often have icmp IR that doesn't match that pattern, so we must account for this
in the backend.
Differential Revision: https://reviews.llvm.org/D38181
llvm-svn: 314023
Combine CMOV[i16]<-[SIGN,ZERO,ANY]_EXTEND to [i32,i64] into CMOV[i32,i64].
One example of where it is useful is:
before (20 bytes)
<foo>:
test $0x1,%dil
mov $0x307e,%ax
mov $0xffff,%cx
cmovne %ax,%cx
movzwl %cx,%eax
retq
after (18 bytes)
<foo>:
test $0x1,%dil
mov $0x307e,%ecx
mov $0xffff,%eax
cmovne %ecx,%eax
retq
Reviewers: craig.topper, aaboud, spatel, RKSimon, zvi
Reviewed By: spatel
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D36711
llvm-svn: 313982
The Swift CC is identical to Win64 CC with the exception of swift error
being passed in r12 which is a CSR. However, since this calling
convention is only used in swift -> swift code, it does not impact
interoperability and can be treated entirely as Win64 CC. We would
previously incorrectly lower the frame setup as we did not treat the
frame as conforming to Win64 specifications.
llvm-svn: 313813
Add support for passing SwiftError through a register on the Windows x64
calling convention. This allows the use of swifterror attributes on
parameters which is used by the swift front end for the `Error`
parameter. This partially enables building the swift standard library
for Windows x86_64.
llvm-svn: 313791
This version of the patch fixes an off-by-one error causing PR34596. We
do not need to use std::next(BlockIter) when calling updateDepths, as
BlockIter already points to the next element.
Original commit message:
> For large basic blocks with lots of combinable instructions, the
> MachineTraceMetrics computations in MachineCombiner can dominate the compile
> time, as computing the trace information is quadratic in the number of
> instructions in a BB and it's relevant successors/predecessors.
> In most cases, knowing the instruction depth should be enough to make
> combination decisions. As we already iterate over all instructions in a basic
> block, the instruction depth can be computed incrementally. This reduces the
> cost of machine-combine drastically in cases where lots of instructions
> are combined. The major drawback is that AFAIK, computing the critical path
> length cannot be done incrementally. Therefore we only compute
> instruction depths incrementally, for basic blocks with more
> instructions than inc_threshold. The -machine-combiner-inc-threshold
> option can be used to set the threshold and allows for easier
> experimenting and checking if using incremental updates for all basic
> blocks has any impact on the performance.
>
> Reviewers: sanjoy, Gerolf, MatzeB, efriedma, fhahn
>
> Reviewed By: fhahn
>
> Subscribers: kiranchandramohan, javed.absar, efriedma, llvm-commits
>
> Differential Revision: https://reviews.llvm.org/D36619
llvm-svn: 313751
This patch adds the instruction scheduling information for the SkylakeClient (SKL) architecture target by adding the file X86SchedSkylakeClient.td located under the X86 Target.
We used the scheduling information retrieved from the Skylake architects in order to create the file.
The scheduling information includes latency, number of micro-Ops and used ports by each SKL instruction.
The patch continues the scheduling replacement and insertion effort started with the SNB target in r307529 and r310792 and for HSW in r311879.
Please expect some performance fluctuations due to code alignment effects.
Reviewers: craig.topper, zvi, chandlerc, igorb, aymanmus, RKSimon, delena
Differential Revision: https://reviews.llvm.org/D37294
llvm-svn: 313613
If we have an AssertZext of a truncated value that has already been AssertZext'ed,
we can assert on the wider source op to improve the zext-y knowledge:
assert (trunc (assert X, i8) to iN), i1 --> trunc (assert X, i1) to iN
This moves a fold from being Mips-specific to general combining, and x86 shows
improvements.
Differential Revision: https://reviews.llvm.org/D37017
llvm-svn: 313577
rL310710 allowed store merging to occur after legalization to catch stores that are created late,
but this exposes a logic hole seen in PR34217:
https://bugs.llvm.org/show_bug.cgi?id=34217
We will miss merging stores if the target lowers vector extracts into target-specific operations.
This patch allows store merging to occur both before and after legalization if the target chooses
to get maximum merging.
I don't think the potential regressions in the other tests are relevant. The tests are for
correctness of weird IR constructs rather than perf tests, and I think those are still correct.
Differential Revision: https://reviews.llvm.org/D37987
llvm-svn: 313564
The AssertZext we might see in this case is only giving information about the lower 32 bits. It isn't providing information about the upper 32 bits. So we should emit a zext.
This fixes PR28540.
Differential Revision: https://reviews.llvm.org/D37729
llvm-svn: 313563
For cases where we are BITCASTing to vectors of smaller elements, then if the entire source was a splatted sign (src's NumSignBits == SrcBitWidth) we can say that the dst's NumSignBit == DstBitWidth, as we're just splitting those sign bits across multiple elements.
We could generalize this but at the moment the only use case I have is to peek through bitcasts to vector comparison results.
Differential Revision: https://reviews.llvm.org/D37849
llvm-svn: 313543
The shuffle combining and lowerVectorShuffleAsLanePermuteAndBlend were both still trying to use VPERM2XF128 for unary shuffles when AVX2 is enabled. VPERM2X128 takes two inputs meaning when we use it for a unary shuffle one of those inputs is left undefined creating a false dependency on whatever register gets allocated there.
If we have VPERMQ/PD we should prefer those since they only have a single input.
Differential Revision: https://reviews.llvm.org/D37947
llvm-svn: 313542
Summary:
Subregister liveness tracking is not implemented for X86 backend, so
sometimes the whole super register is said to be live, when only a
subregister is really live. That might happen if the def and the use
are located in different MBBs, see added fixup-bw-isnt.mir test.
However, using knowledge of the specific instructions handled by the
bw-fixup-pass we can get more precise liveness information which this
change does.
Reviewers: MatzeB, DavidKreitzer, ab, andrew.w.kaylor, craig.topper
Reviewed By: craig.topper
Subscribers: n.bozhenov, myatsina, llvm-commits, hiraditya
Patch by Andrei Elovikov <andrei.elovikov@intel.com>
Differential Revision: https://reviews.llvm.org/D37559
llvm-svn: 313524
This allows vector-sized store merging of constants in DAGCombiner using the existing code in MergeConsecutiveStores().
All of the twisted logic that decides exactly what vector operations are legal and fast for each particular CPU are
handled separately in there using the appropriate hooks.
For the motivating tests in merge-store-constants.ll, we already produce the same vector code in IR via the SLP vectorizer.
So this is just providing a backend backstop for code that doesn't go through that pass (-O1). More details in PR24449:
https://bugs.llvm.org/show_bug.cgi?id=24449 (this change should be the last step to resolve that bug)
Differential Revision: https://reviews.llvm.org/D37451
llvm-svn: 313458
We just need to toggle bits 1 and 5 of the immediate and swap the sources. The peephole pass could trigger commuting/folding for this later, but its easy enough to fix in isel.
Disable the peephole pass on the main vperm2x128 test so we know we're doing this through isel.
llvm-svn: 313455
I've moved the test cases from the InstCombine optimizations to the backend to keep the coverage we had there. It covered every possible immediate so I've preserved the resulting shuffle mask for each of those immediates.
llvm-svn: 313450
I'm going to autoupgrade these intrinsics in a future commit. This bit will never be set in the resulting output so pre-removing the bit.
llvm-svn: 313434
This caused PR34629: asserts firing when building Chromium. It also broke some
buildbots building test-suite as reported on the commit thread.
> Summary:
> 1/ Operand folding during complex pattern matching for LEAs has been
> extended, such that it promotes Scale to accommodate similar operand
> appearing in the DAG.
> e.g.
> T1 = A + B
> T2 = T1 + 10
> T3 = T2 + A
> For above DAG rooted at T3, X86AddressMode will no look like
> Base = B , Index = A , Scale = 2 , Disp = 10
>
> 2/ During OptimizeLEAPass down the pipeline factorization is now performed over LEAs
> so that if there is an opportunity then complex LEAs (having 3 operands)
> could be factored out.
> e.g.
> leal 1(%rax,%rcx,1), %rdx
> leal 1(%rax,%rcx,2), %rcx
> will be factored as following
> leal 1(%rax,%rcx,1), %rdx
> leal (%rdx,%rcx) , %edx
>
> 3/ Aggressive operand folding for AM based selection for LEAs is sensitive to loops,
> thus avoiding creation of any complex LEAs within a loop.
>
> Reviewers: lsaba, RKSimon, craig.topper, qcolombet
>
> Reviewed By: lsaba
>
> Subscribers: spatel, igorb, llvm-commits
>
> Differential Revision: https://reviews.llvm.org/D35014
llvm-svn: 313376
The early out for AVX2 in lowerV2X128VectorShuffle is positioned in a weird spot below some shuffle mask equivalency checks.
But I think we want to allow VPERMQ for any unary shuffle.
Differential Revision: https://reviews.llvm.org/D37893
llvm-svn: 313373
When handling a v64i1 build vector of constants on 32-bit targets we were creating an illegal i64 constant that we then bitcasted back to v64i1. We need to instead create two 32-bit constants, bitcast them to v32i1 and concat the result. We should also take care to handle the halves being all zeros/ones after the split.
This patch splits the build vector and then recursively lowers the two pieces. This allows us to handle the all ones and all zeros cases with minimal effort. Ideally we'd just do the split and concat, and let lowering get called again on the new nodes, but getNode has special handling for CONCAT_VECTORS that reassembles the pieces back into a single BUILD_VECTOR. Hopefully the two temporary BUILD_VECTORS we had to create to do this that don't get returned don't cause any issues.
Fixes PR34605.
Differential Revision: https://reviews.llvm.org/D37858
llvm-svn: 313366
Currently if we're inserting 0s into the upper elements of a vector register we insert an explicit move of the smaller register to implicitly zero the upper bits. But if we can prove that they are already zero we can skip that. This is based on a similar idea of what we do to avoid emitting explicit zero extends for GR32->GR64.
Unfortunately, this is harder for vector registers because there are several opcodes that don't have VEX equivalent instructions, but can write to XMM registers. Among these are SHA instructions and a MMX->XMM move. Bitcasts can also get in the way.
So for now I'm starting with explicitly allowing only VPMADDWD because we emit zeros in combineLoopMAddPattern. So that is placing extra instruction into the reduction loop.
I'd like to allow PSADBW as well after D37453, but that's currently blocked by a bitcast. We either need to peek through bitcasts or canonicalize insert_subvectors with zeros to remove bitcasts on the value being inserted.
Longer term we should probably have a cleanup pass that removes superfluous zeroing moves even when the producer is in another basic block which is something these isel tricks can't do. See PR32544.
Differential Revision: https://reviews.llvm.org/D37653
llvm-svn: 313365
Summary:
1/ Operand folding during complex pattern matching for LEAs has been
extended, such that it promotes Scale to accommodate similar operand
appearing in the DAG.
e.g.
T1 = A + B
T2 = T1 + 10
T3 = T2 + A
For above DAG rooted at T3, X86AddressMode will no look like
Base = B , Index = A , Scale = 2 , Disp = 10
2/ During OptimizeLEAPass down the pipeline factorization is now performed over LEAs
so that if there is an opportunity then complex LEAs (having 3 operands)
could be factored out.
e.g.
leal 1(%rax,%rcx,1), %rdx
leal 1(%rax,%rcx,2), %rcx
will be factored as following
leal 1(%rax,%rcx,1), %rdx
leal (%rdx,%rcx) , %edx
3/ Aggressive operand folding for AM based selection for LEAs is sensitive to loops,
thus avoiding creation of any complex LEAs within a loop.
Reviewers: lsaba, RKSimon, craig.topper, qcolombet
Reviewed By: lsaba
Subscribers: spatel, igorb, llvm-commits
Differential Revision: https://reviews.llvm.org/D35014
llvm-svn: 313343
We already have a combine for this pattern when the input to shl is add, so we just need to enable the transformation when the input is or.
Original patch by @tstellar
Differential Revision: https://reviews.llvm.org/D19325
llvm-svn: 313251
Summary:
XRay had been assuming that the previous section is the "text" section
of the function when lowering the instrumentation map. Unfortunately
this is not a safe assumption, because we may be coming from lowering
debug type information for the function being lowered.
This fixes an issue with combining -gsplit-dwarf, -generate-type-units,
-debug-compile and -fxray-instrument for sole member functions. When the
split dwarf section is stripped, we're left with references from the
xray_instr_map to the debug section. The change now uses the function's
symbol instead of the previous section's start symbol.
We found the bug while attempting to strip the split debug sections off
an XRay-instrumented object file, which had a peculiar edge-case for
single-function classes where the single function is being lowered.
Because XRay had assocaited the instrumentation map for a function to
the debug types section instead of the function's section, the objcopy
call will fail due to the misplaced reference from the xray_instr_map
section.
Reviewers: pcc, dblaikie, echristo
Subscribers: llvm-commits, aprantl
Differential Revision: https://reviews.llvm.org/D37791
llvm-svn: 313233
This caused PR34596.
> [MachineCombiner] Update instruction depths incrementally for large BBs.
>
> Summary:
> For large basic blocks with lots of combinable instructions, the
> MachineTraceMetrics computations in MachineCombiner can dominate the compile
> time, as computing the trace information is quadratic in the number of
> instructions in a BB and it's relevant successors/predecessors.
>
> In most cases, knowing the instruction depth should be enough to make
> combination decisions. As we already iterate over all instructions in a basic
> block, the instruction depth can be computed incrementally. This reduces the
> cost of machine-combine drastically in cases where lots of instructions
> are combined. The major drawback is that AFAIK, computing the critical path
> length cannot be done incrementally. Therefore we only compute
> instruction depths incrementally, for basic blocks with more
> instructions than inc_threshold. The -machine-combiner-inc-threshold
> option can be used to set the threshold and allows for easier
> experimenting and checking if using incremental updates for all basic
> blocks has any impact on the performance.
>
> Reviewers: sanjoy, Gerolf, MatzeB, efriedma, fhahn
>
> Reviewed By: fhahn
>
> Subscribers: kiranchandramohan, javed.absar, efriedma, llvm-commits
>
> Differential Revision: https://reviews.llvm.org/D36619
llvm-svn: 313213
This is to fix PR34502. After rL311401, the live range of spilled vreg will be
cleared. HoistSpill need to use the live range of the original vreg before splitting
to know the moving range of the spills. The patch saves a copy of live interval for
the spilled vreg inside of HoistSpillHelper.
Differential Revision: https://reviews.llvm.org/D37578
llvm-svn: 313197
NFC.
Replacing -mcpu=skx by -mattr in the run command of the codegen test: avx512-gather-scatter-intrin.ll.
Reviewers: delena
Revision: https://reviews.llvm.org/D37799
llvm-svn: 313144
NFC.
Updating the code gen regression test bmi2-schedule.ll to use the KNL and SKYLAKE prefixes for the run commands that use the knl and Skylake mcpu options.
The fix is in preparation for a large patch of adding all SKL scheduling information.
Reviewers: delena, zvi, RKSimon
Revision: https://reviews.llvm.org/D37796
llvm-svn: 313138
NFC.
Updating the code gen regression test bmi2-schedule.ll to use the KNL and SKYLAKE prefixes for the run commands that use the knl and Skylake mcpu options.
The fix is in preparation for a large patch of adding all SKL scheduling information.
Reviewers: delena, zvi
Revision: https://reviews.llvm.org/D37796
llvm-svn: 313137
Updating codegen test bmi2-schedule.ll to use the SKYLAKE and KNL prefix as preparatipn for an upcoming patch to add all SKL scheduling information.
llvm-svn: 313136
Add explicit mc-encoding checks showing that the AVX512VL ABS intrinsics are actually mapped to EVEX encoding.
This is a pre-commit for a soon to come patch which will lower x86 target specific ABS intrinsics to IR.
Differential Revision: https://reviews.llvm.org/D37688
llvm-svn: 313131
Load with zero-extend and sign-extend from v2i8 to v2i32 is "Legal" since SSE4.1 and may be performed using PMOVZXBD , PMOVSXBD instructions.
llvm-svn: 313121
The masked store instruction only cares about the sign-bit of each mask element,
so the compare s<0 isn't needed.
As noted in PR11210:
https://bugs.llvm.org/show_bug.cgi?id=11210
...fixing this should allow us to eliminate x86-specific masked store intrinsics in IR.
(Although more testing will be needed to confirm that.)
I filed a bug to track improvements for AVX512:
https://bugs.llvm.org/show_bug.cgi?id=34584
Differential Revision: https://reviews.llvm.org/D37446
llvm-svn: 313089
Recognizing this pattern during DAG combine hides information about the 'and' and the shift from other combines. I think it should be recognized at isel so its as late as possible. But it can't be done with table based isel because you need to be able to look at both immediates. This patch moves it to custom isel in X86ISelDAGToDAG.cpp.
This does break a couple tests in tbm_patterns because we are now emitting an and_flag node or (cmp and, 0) that we dont' recognize yet. We already had this problem for several other TBM patterns so I think this fine and we can address of them together.
I've also fixed a bug where the combine to BEXTR was preventing us from using a trick of zero extending AH to handle extracts of bits 15:8. We might still want to use BEXTR if it enables load folding. But honestly I hope we narrowed the load instead before got to isel.
I think we should probably also support matching BEXTR from (srl/srl (and mask << C), C). But that should be a different patch.
Differential Revision: https://reviews.llvm.org/D37592
llvm-svn: 313054
If we allow the OR to be narrowed then the upper bits really are zero and we can't tell if the zeroing movl was removed on purpose.
While here regenerate the test with update_llc_test_checks.py
llvm-svn: 312995
As discussed on llvm-dev in
http://lists.llvm.org/pipermail/llvm-dev/2017-September/117301.html
this changes the command line interface of llvm-dwarfdump to match the
one used by the dwarfdump utility shipping on macOS. In addition to
being shorter to type this format also has the advantage of allowing
more than one section to be specified at the same time.
In a nutshell, with this change
$ llvm-dwarfdump --debug-dump=info
$ llvm-dwarfdump --debug-dump=apple-objc
becomes
$ dwarfdump --debug-info --apple-objc
Differential Revision: https://reviews.llvm.org/D37714
llvm-svn: 312970
Helps improve combineLogicBlendIntoPBLENDV support by allowing us to peek into through PACKSS truncations of vector comparison results.
Differential Revision: https://reviews.llvm.org/D37680
llvm-svn: 312916
Suggested in D37680
Note: had to drop AVX512VL tests as there is an infinite loop in the new tests that needs further investigation (not relevant to D37680).
llvm-svn: 312910
NFC.
Updated 3 Codegen regression tests to use the -mattr flag instead of the -mcpu flags as follows:
Instead of -mcpu=skx use -mattr=+avx512f,+avx512bw,+avx512vl,+avx512dq
Instead of -mcpu=knl use -mattr=+avx512f
Reviewers: delena
Revision: https://reviews.llvm.org/D37674
llvm-svn: 312909
After the split of the Scatter operation, the order of the new instructions is well defined - Lo goes before Hi. Otherwise the semantic of Scatter (from LSB to MSB) is broken.
I'm chaining 2 nodes to prevent reordering.
Differential Revision https://reviews.llvm.org/D37670
llvm-svn: 312894
Summary:
Just because INC/DEC is a little slow on some processors doesn't mean we shouldn't prefer it when optimizing for size.
This appears to match gcc behavior.
Reviewers: chandlerc, zvi, RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D37177
llvm-svn: 312866
cover the bitwise operators.
Nothing really exciting here, this just stamps out the rest of the core
operations that can RMW memory and set flags.
Still not implemented here: ADC, SBB. Those will require more
interesting logic to channel the flags *in*, and I'm not currently
planning to try to tackle that. It might be interesting for someone who
wants to improve our code generation for bignum implementations.
Differential Revision: https://reviews.llvm.org/D37141
llvm-svn: 312768
operands and used flags to support matching immediate operands.
This is a bit trickier than register operands, and we still want to fall
back on a register operands even for things that appear to be
"immediates" when they won't actually select into the operation's
immediate operand. This also requires us to handle things like selecting
`sub` vs. `add` to minimize the number of bits needed to represent the
immediate, and picking the shortest immediate encoding. In order to
that, we in turn need to scan to make sure that CF isn't used as it will
get inverted.
The end result seems very nice though, and we're now generating
optimal instruction sequences for these patterns IMO.
A follow-up patch will further expand this to other operations with RMW
memory operands. But handing `add` and `sub` are useful starting points
to flesh out the machinery and make sure interesting and complex cases
can be handled.
Thanks to Craig Topper who provided a few fixes and improvements to this
patch in addition to the review!
Differential Revision: https://reviews.llvm.org/D37139
llvm-svn: 312764
This patch expands the support of lowerInterleavedload to {8|16|32}x8i stride 3.
LLVM creates suboptimal shuffle code-gen for AVX2. In overall, this patch is a specific fix for the pattern (Strid=3 VF={8|16|32}) and we plan to include the store (deinterleved side).
The patch goal is to optimize the following sequence:
a0 b0 c0 a1 b1 c1 a2 b2
c2 a3 b3 c3 a4 b4 c4 a5
b5 c5 a6 b6 c6 a7 b7 c7
into
a0 a1 a2 a3 a4 a5 a6 a7
b0 b1 b2 b3 b4 b5 b6 b7
c0 c1 c2 c3 c4 c5 c6 c7
Reviewers
1. zvi
2. igor
3. guyblank
4. dorit
5. Ayal
llvm-svn: 312722
Summary:
For large basic blocks with lots of combinable instructions, the
MachineTraceMetrics computations in MachineCombiner can dominate the compile
time, as computing the trace information is quadratic in the number of
instructions in a BB and it's relevant successors/predecessors.
In most cases, knowing the instruction depth should be enough to make
combination decisions. As we already iterate over all instructions in a basic
block, the instruction depth can be computed incrementally. This reduces the
cost of machine-combine drastically in cases where lots of instructions
are combined. The major drawback is that AFAIK, computing the critical path
length cannot be done incrementally. Therefore we only compute
instruction depths incrementally, for basic blocks with more
instructions than inc_threshold. The -machine-combiner-inc-threshold
option can be used to set the threshold and allows for easier
experimenting and checking if using incremental updates for all basic
blocks has any impact on the performance.
Reviewers: sanjoy, Gerolf, MatzeB, efriedma, fhahn
Reviewed By: fhahn
Subscribers: kiranchandramohan, javed.absar, efriedma, llvm-commits
Differential Revision: https://reviews.llvm.org/D36619
llvm-svn: 312719
Summary:
Add patterns for
fptoui <16 x float> to <16 x i8>
fptoui <16 x float> to <16 x i16>
Reviewers: igorb, delena, craig.topper
Reviewed By: craig.topper
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D37505
llvm-svn: 312704
function return the intrinsics's first argument.
llvm.memcpy/memset/memmove return void but they will return the first
argument after they are expanded as libcalls. Now if the parent function
has any return value, llvm.memcpy cannot be turned into tail call after
expansion.
The patch is to handle that case in SelectionDAGBuilder so when caller
function return the same value as the first argument of llvm.memcpy,
tail call is allowed.
Differential Revision: https://reviews.llvm.org/D37406
llvm-svn: 312641
performing a zext of a register.
On the PR there is discussion of how to more effectively handle this,
but this patch prevents us from miscompiling code.
Differential Revision: https://reviews.llvm.org/D37504
llvm-svn: 312620
Summary:
Most instructions in AVX work “in-lane”, that is, each source element is applied only to other
elements of the same lane, thus a cross lane permutation is costly and needs more than one instrution.
AVX2 includes instructions to perform any-to-any permutation of words over a 256-bit register
and vectorized table lookup.
This should also Fix PR34369
Differential Revision: https://reviews.llvm.org/D37388
llvm-svn: 312608
Summary:
This intrinsic represents a label with a list of associated metadata
strings. It is modelled as reading and writing inaccessible memory so
that it won't be removed as dead code. I think the intention is that the
annotation strings should appear at most once in the debug info, so I
marked it noduplicate. We are allowed to inline code with annotations as
long as we strip the annotation, but that can be done later.
Reviewers: majnemer
Subscribers: eraman, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D36904
llvm-svn: 312569
We had already disabled the pattern for SSE4.1 and SSE4.2. But it got re-enabled for AVX and AVX512.
With SSE41 we rely on a separate (v4f32 (X86vzmovl VR128)) pattern to select blendps with a xorps to create zeroess. And a separate (v4f32 (scalar_to_vector FR32X)) to select a COPY_TO_REG_CLASS to move FR32 to VR128
The same thing can happen for AVX with vblendps and those separate patterns already exist.
For AVX512, (v4f32 (X86vzmov VR128)) will select a VMOVSS instruction instead of VBLENDPS due to their not being a EVEX VBLENDPS. This is what we were getting out of the larger pattern anyway. So the larger pattern is unneeded for AVX512 too.
For SSE1-SSSE3 we can rely on (v4f32 (X86vzmov VR128)) selecting a MOVSS similar to AVX512. Again this is what the larger pattern did too.
So the only real change here is that AVX1/2 now properly outputs a VBLENDPS during isel instead of a VMOVSS to match SSE41. Most tests didn't notice because the two address instruction pass knows how to turn VMOVSS into VBLENDPS to get an independent destination register.
llvm-svn: 312564
Craig Topper