When the result type is v2i64/v2f64 and the index element size is i32, the index vector has two unused elements making the type v4i32. The mask VT should match the number of memory accesses that will be made.
This is consistent with the isel patterns used for the target independent gather/scatter intrinsic.
llvm-svn: 350687
For stack frames on the size of a register in x86, a code size optimization
emits "push rax/eax" instead of "sub" for stack allocation. For example:
foo:
.cfi_startproc
BB#0:
pushq %rax
Ltmp0:
.cfi_def_cfa_offset 16
...
.cfi_endproc
However, we are falling back to DWARF in this case because we cannot
encode %rax as a saved register.
This requirement is wrong, since we don't care about the contents of
%rax, it is the equivalent of a sub.
In order to specify that we care about the contents of %rax, we would
need a .cfi_offset %rax, <offset>.
It's also overzealous in the case where there are pushes for callee saved
registers followed by a "push rax/eax" instead of "sub", in which case we should
also be able to encode the callee saved regs and everything else using compact
unwind.
Patch authored by Bruno Cardoso Lopes.
Differential Revision: https://reviews.llvm.org/D13793
llvm-svn: 350623
Summary: AVX512VBMI2 supports a funnel shift by immediate and a funnel shift by a variable vector.
Reviewers: spatel, RKSimon
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D56361
llvm-svn: 350498
There are no test changes here in the existing cost model
regression tests because integer add/sub have a default
legal cost of 1 already. This would break, however, if
we custom lower those ops because the default cost model
assumes that custom-lowered ops are more expensive.
This is similar to the change in rL350403. See discussion
in D56011 for more details. When we enhance that patch to
handle integer ops, we need this cost model change to avoid
unintended diffs here from the custom lowering.
llvm-svn: 350496
This was here because out and in instructions allow '(%dx)' even though its not a memory reference. To handle this we build a special operand for the DX register reference before we get to the call to CheckBaseRegAndIndexRegAndScale. So we no longer need this special case.
llvm-svn: 350483
This is especially helpful on targets without avx512bw since we don't have a good way to convert from v16i8/v32i8 to v16i1/v32i1 for the truncate anyway. If we're just going to convert it to a GPR we might as well use pmovmskb to accomplish both.
llvm-svn: 350480
We don't need to require the first operand to be an integer because we already said it was the same type as the result which we also constrained to an integer.
llvm-svn: 350455
The 1st try for this was at rL350369, but it caused IR-level diffs because
our cost models differentiate custom vs. legal/promote lowering. So that was
reverted at rL350373. The cost models were fixed independently at rL350403,
so this is effectively the same patch as last time.
Original commit message:
This would show up if we fix horizontal reductions to narrow as they go along,
but it's an improvement for size and/or Jaguar (fast-hops) independent of that.
We need to do this late to not interfere with other pattern matching of larger
horizontal sequences.
We can extend this to integer ops in a follow-up patch.
Differential Revision: https://reviews.llvm.org/D56011
llvm-svn: 350421
Noticed in D56011 - handle the case that scalar fp ops are quicker on P3 than P4
Add the other costs so that we're not relying on the default "is legal/custom" cost logic.
llvm-svn: 350403
Doing this late so we will prefer to fold the AND into a masked comparison first. That can be better for the live range of the mask register.
Differential Revision: https://reviews.llvm.org/D56246
llvm-svn: 350374
This would show up if we fix horizontal reductions to narrow as they go along,
but it's an improvement for size and/or Jaguar (fast-hops) independent of that.
We need to do this late to not interfere with other pattern matching of larger
horizontal sequences.
We can extend this to integer ops in a follow-up patch.
Differential Revision: https://reviews.llvm.org/D56011
llvm-svn: 350369
As noted in PR39973 and D55558:
https://bugs.llvm.org/show_bug.cgi?id=39973
...this is a partial implementation of a fold that we do as an IR canonicalization in instcombine:
// extelt (binop X, Y), Index --> binop (extelt X, Index), (extelt Y, Index)
We want to have this in the DAG too because as we can see in some of the test diffs (reductions),
the pattern may not be visible in IR.
Given that this is already an IR canonicalization, any backend that would prefer a vector op over
a scalar op is expected to already have the reverse transform in DAG lowering (not sure if that's
a realistic expectation though). The transform is limited with a TLI hook because there's an
existing transform in CodeGenPrepare that tries to do the opposite transform.
Differential Revision: https://reviews.llvm.org/D55722
llvm-svn: 350354
INC/DEC are pretty much the same as ADD/SUB except that they don't update the C flag.
This patch removes the special nodes and just pattern matches from ADD/SUB during isel if the C flag isn't being used.
I had to avoid selecting DEC is the result isn't used. This will become a SUB immediate which will turned into a CMP later by optimizeCompareInstr. This lead to the one test change where we use a CMP instead of a DEC for an overflow intrinsic since we only checked the flag.
This also exposed a hole in our RMW flag matching use of hasNoCarryFlagUses. Our root node for the match is a store and there's no guarantee that all the flag users have been selected yet. So hasNoCarryFlagUses needs to check copyToReg and machine opcodes, but it also needs to check for the pre-match SETCC, SETCC_CARRY, BRCOND, and CMOV opcodes.
Differential Revision: https://reviews.llvm.org/D55975
llvm-svn: 350245
Peek through shift modulo masks while matching double shift patterns.
I was hoping to delay this until I could remove the X86 code with generic funnel shift matching (PR40081) but this will do for now.
Differential Revision: https://reviews.llvm.org/D56199
llvm-svn: 350222
All of these use custom isel so we can pretty easily detect the differences in the custom code in X86ISelDAGToDAG. The ISD opcodes just need to express the desired semantics not the details of how they would be selected by isel. So unifying them lets us remove the special casing from lowering.
llvm-svn: 350206
These require a different X86ISD node to be created than i16/i32/i64. I guess no one wanted to add the special code for that except in LowerXALUO. But now LowerXALUO, LowerSELECT, and LowerBRCOND all use a common helper function so they all share the special code.
Unfortunately, there are no test changes because we seem to correct the miss in a DAG combine later. I did verify it manually using test cases from xmulo.ll
llvm-svn: 350205
This makes it easier to keep the LowerBRCOND and LowerSELECT code in sync with LowerXALUO so they always pick the same operation for overflowing instructions.
This is inspired by the helper functions used by ARM and AArch64 for the same purpose.
The test change is because LowerSELECT was not in sync with LowerXALUO with regard to INC/DEC for SADDO/SSUBO.
llvm-svn: 350198
This seems to be getting in the way more than its helping. This does mean we stop scalarizing some cases, but I'm not convinced the scalarization was really better.
Some of the changes to vsel-cmp-load.ll are a regression but D56156 should fix it.
llvm-svn: 350159
This allows us to sign extend to v4i32 first. And then share that extension to implement the final steps to v4i64 using a pcmpgt and punpckl and punpckh.
We already do something similar for SIGN_EXTEND with -x86-experimental-vector-widening-legalization.
llvm-svn: 350158
This was tricking us into making these operations and then letting them get scalarized later. But I can't prove that the scalarized version is actually better.
llvm-svn: 350141
Previously we emitted a multiply and some masking that was supposed to matched to PMULUDQ, but the masking could sometimes be removed before we got a chance to match it. So instead just emit the PMULUDQ directly.
Remove the DAG combine that was added when the ReplaceNodeResults code was originally added. Add a new DAG combine to avoid regressions in shrink_vmul.ll
Some of the shrink_vmul.ll test cases now pick PMULUDQ instead of PMADDWD/PMULLD, but I think this should be an improvement on most CPUs.
I think all of this can go away if/when we switch to -x86-experimental-vector-widening-legalization
llvm-svn: 350134
Create PMULDQ/PMULUDQ as long as the number of elements is a power of 2.
This seems to give some improvements in our ability to use SimplifyDemandedBits.
llvm-svn: 350084
Make each of the helper functions only return their comparison node and the condition code. Leave X86ISD::SETCC creation to the LowerSETCC function itself.
Looking into whether we can use this code directly in BRCOND and SELECT lowering instead of going through LowerSETCC which creates an X86ISD::SETCC node we need to look through.
llvm-svn: 350082
Only one of the 3 callers of LowerAndToBT need the SETCC node. Two of them have to look through it to find the operands they really need. Instead create it after the one call that needs it.
LowerAndToBT now returns both the BT node and the X86 specific condition code separately.
llvm-svn: 350081
This is an alternative to what I attempted in D56057.
GetDemandedBits is a special version of SimplifyDemandedBits that allows simplifications even when the operand has other uses. GetDemandedBits will only do simplifications that allow a node to be bypassed. It won't create new nodes or alter any of the other users.
I had to add support for bypassing SIGN_EXTEND_INREG to GetDemandedBits.
Based on a patch that Simon Pilgrim sent me in email.
Fixes PR40142.
llvm-svn: 350059
Remove the TESTmr isel patterns and add another postprocessing combine for TESTrr+ANDrm->TESTmr. We already have a postprocessing combine for TESTrr+ANDrr->TESTrr. With this we can give ANDN a chance to match first. And clean it up during post processing if we ended up with just a regular AND.
This is another step towards my plan to gut EmitTest and do more flag handling during isel matching or by using optimizeCompare.
llvm-svn: 350038
The missed load folding noticed in D55898 is visible independent of that change
either with an adjusted IR pattern to start or with AVX2/AVX512 (where the build
vector becomes a broadcast first; movddup is not produced until we get into isel
via tablegen patterns).
Differential Revision: https://reviews.llvm.org/D55936
llvm-svn: 350005
Summary:
Added a pair of APIs for encoding/decoding the 3 components of a DWARF discriminator described in http://lists.llvm.org/pipermail/llvm-dev/2016-October/106532.html: the base discriminator, the duplication factor (useful in profile-guided optimization) and the copy index (used to identify copies of code in cases like loop unrolling)
The encoding packs 3 unsigned values in 32 bits. This CL addresses 2 issues:
- communicates overflow back to the user
- supports encoding all 3 components together. Current APIs assume a sequencing of events. For example, creating a new discriminator based on an existing one by changing the base discriminator was not supported.
Reviewers: davidxl, danielcdh, wmi, dblaikie
Reviewed By: dblaikie
Subscribers: zzheng, dmgreen, aprantl, JDevlieghere, llvm-commits
Differential Revision: https://reviews.llvm.org/D55681
llvm-svn: 349973
This fixes the patterns that have or/and as a root. 'and' is handled differently since thy usually have a CMP wrapped around them.
I had to look for uses of the CF flag because all these nodes have non-standard CF flag behavior. A real or/xor would always clear CF. In practice we shouldn't be using the CF flag from these nodes as far as I know.
Differential Revision: https://reviews.llvm.org/D55813
llvm-svn: 349962
The BEXTR instruction documents the SF bit as undefined.
The TBM BEXTR instruction has the same issue, but I'm not sure how to test it. With the control being an immediate we can determine the sign bit is 0 or the BEXTR would have been removed.
Fixes PR40060
Differential Revision: https://reviews.llvm.org/D55807
llvm-svn: 349956
This is admittedly a narrow fix for the problem:
https://bugs.llvm.org/show_bug.cgi?id=37502
...but as the XOP restriction shows, it's a maze to get this right.
In the motivating example, note that we have movddup before SSE4.1 and
again with AVX2. That's because insertps isn't available pre-SSE41 and
vbroadcast is (more generally) available with AVX2 (and the splat is
reduced to movddup via isel pattern).
Differential Revision: https://reviews.llvm.org/D55898
llvm-svn: 349937
This shortens the switches in X86ISelDAGToDAG.cpp to only need to check condition code instead of a list of opcodes.
This also fixes a bug where the memory forms of SETcc were missing from hasNoCarryFlagUses.
llvm-svn: 349868
Summary:
This allows expanding {7,11,13,14,15,21,22,23,25,26,27,28,29,30,31}-byte memcmp
in just two loads on X86. These were previously calling memcmp.
Reviewers: spatel, gchatelet
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D55263
llvm-svn: 349731
The (cmp (and X, Y) 0) pattern is greedy and ends up forming a TESTrr and consuming the and when it might be better to use one of the BMI/TBM like BLSR or BLSI.
This patch moves removes the pattern from isel and adds a post processing check to combine TESTrr+ANDrr into just a TESTrr. With this patch we are able to select the BMI/TBM instructions, but we'll also emit a TESTrr when the result is compared to 0. In many cases the peephole pass will be able to use optimizeCompareInstr to remove the TEST, but its probably not perfect.
Differential Revision: https://reviews.llvm.org/D55870
llvm-svn: 349661
Fixes https://bugs.llvm.org/show_bug.cgi?id=38743
The function removeRedundantBlockingStores is supposed to remove any blocking stores contained in each other in lockingStoresDispSizeMap.
But it currently looks only at the previous one, which will miss some cases that result in assert.
This patch refine the function to check all previous layouts until find the uncontained one. So all redundant stores will be removed.
Patch by Pengfei Wang
Differential Revision: https://reviews.llvm.org/D55642
llvm-svn: 349660
We already had BSF here as part of __builtin_ffs improvements and I was just wondering yesterday whether we should have BSR there.
This addresses one issue from PR40090.
llvm-svn: 349531
Migrate the X86 backend from X86ISD opcodes ADDS and SUBS to generic
ISD opcodes SADDSAT and SSUBSAT. This also improves scodegen for
@llvm.sadd.sat() and @llvm.ssub.sat() intrinsics.
This is a followup to D55787 and part of PR40056.
Differential Revision: https://reviews.llvm.org/D55833
llvm-svn: 349520
InstCombine seems to canonicalize or PSUB patter into a max with the cosntant and an add with an inverse of the constant.
This patch recognizes this pattern and turns it into PSUBUS. Future work could improve undef element handling.
Fixes some of PR40053
Differential Revision: https://reviews.llvm.org/D55780
llvm-svn: 349519
Replace the X86ISD opcodes ADDUS and SUBUS with generic ISD opcodes
UADDSAT and USUBSAT. As a side-effect, this also makes codegen for
the @llvm.uadd.sat and @llvm.usub.sat intrinsics reasonable.
This only replaces use in the X86 backend, and does not move any of
the ADDUS/SUBUS X86 specific combines into generic codegen.
Differential Revision: https://reviews.llvm.org/D55787
llvm-svn: 349481
(VSRAI (VSHLI X, C1), C1) --> X iff NumSignBits(X) > C1
This works better as part of SimplifyDemandedBits than part of the general combine.
llvm-svn: 349462
This fold was incredibly specific - replace with a SimplifyDemandedBits fold to remove a VSRAI if only the original sign bit is demanded (its guaranteed to stay the same).
Test change is merely a rescheduling.
llvm-svn: 349459
Convert VSRAI to VSRLI is the sign bit is known zero and improve KnownBits output for all shift instruction.
Fixes the poor codegen comments in D55768.
llvm-svn: 349407
This allows a TEST to be used and can be combined with any AND that may already exist as an input to the shift.
This was already done in EmitTest, but was easily tricked by multiple uses because the setcc might be used by multiple instructions. Once the SETCC and users are legalized then we can look for the shift to be used by a single CMP, but the CMP itself can have multiple users.
This appears to fix the case in PR39968.
llvm-svn: 349385
This is an initial patch to add the necessary support for a DemandedElts argument to SimplifyDemandedBits, more closely matching computeKnownBits and to help improve vector codegen.
I've added only a small amount of the changes necessary to get at least one test to update - a lot more can be done but I'd like to add these methodically with proper test coverage, at the same time the hope is to slowly move some/all of SimplifyDemandedVectorElts into SimplifyDemandedBits as well.
Differential Revision: https://reviews.llvm.org/D55768
llvm-svn: 349374
We keep a few iterators into the basic block we're selecting while
performing FastISel. Usually this is fine, but occasionally code wants
to remove already-emitted instructions. When this happens we have to be
careful to update those iterators so they're not pointint at dangling
memory.
llvm-svn: 349365
I'd like to try to move a lot of the flag matching out of EmitTest and push it to isel or isel preprocessing. This is a step towards that.
The test-shrink-bug.ll changie is an improvement because we are no longer interfering with test shrink handling in isel.
The pr34137.ll change is a regression, but the IR came from -O0 and was not reduced by InstCombine. So it contains a lot of redundancies like duplicate loads that made it combine poorly.
llvm-svn: 349315
Use consistent rules for when to lower to SHLD/SHRD for slow machines - fixes a weird issue where funnel shift gets expanded but then X86ISelLowering's combineOr sees the optsize and combines to SHLD/SHRD, but now with the modulo amount guard......
llvm-svn: 349285
The only caller of this turns CMP with 0 into TEST. CMP with 0 and TEST both set OF to 0 so we should have no issues with instructions that only use OF.
Though I don't think there's any reason we would read just OF after a compare with 0 anyway. So this probably isn't an observable change.
llvm-svn: 349223
hasNoCarryFlagUses hardcoded that the flag result is 1 and used that to filter which uses were of interest. hasNoSignedComparisonUses just assumes the only result is flags and checks whether any user of the node is a CopyToReg instruction.
After this patch we now do a result number check in both and rely on the caller to provide the result number.
This shouldn't change behavior it was just an odd difference between the two functions that I noticed.
llvm-svn: 349222
Summary:
If the setcc already has the target desired type we can reach the getSetCC/getSExtOrTrunc after the MatchingVecType check with the exact same types as the nodes we started with. This causes those causes VsetCC to be CSEd to N0 and the getSExtOrTrunc will CSE to N. When we return N, the caller will think that meant we called CombineTo and did our own worklist management. But that's not what happened. This prevents target hooks from being called for the node.
To fix this, I've now returned SDValue if the setcc is already the desired type. But to avoid some regressions in X86 I've had to disable one of the target combines that wasn't being reached before in the case of a (sext (setcc)). If we get vector widening legalization enabled that entire function will be deleted anyway so hopefully this is only for the short term.
Reviewers: RKSimon, spatel
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D55459
llvm-svn: 349137
This requires the two callers to manifest a 0 to make EmitCmp call EmitTest.
I'm looking into changing how we combine TEST and flag setting instructions to not be part of lowering. And instead be part of DAG combine or isel. Which will mean EmitTest will probably become gutted and maybe disappear entirely.
llvm-svn: 349094
Summary:
Macros are expanded on a single line. In case of large expansions,
with sufficiently many instructions with memory operands (and when
-fdebug-info-for-profiling is requested), we may be unable to generate
new base discriminator values - new values overflow (base
discriminators may not be larger than 2^12).
This CL warns instead of asserting in such a case. A subsequent CL
will add APIs to check for overflow before creating new debug info.
See https://bugs.llvm.org/show_bug.cgi?id=39890
Reviewers: davidxl, wmi, gbedwell
Reviewed By: davidxl
Subscribers: aprantl, llvm-commits
Differential Revision: https://reviews.llvm.org/D55643
llvm-svn: 349075
There's still a couple of minor SimplifyDemandedElts regressions in some of the shift amount splats that will be fixed in future patches.
llvm-svn: 349052
Move existing rotation expansion code into TargetLowering and set it up for vectors as well.
Ideally this would share more of the funnel shift expansion, but we handle the shift amount modulo quite differently at the moment.
Begun removing x86 vector rotate custom lowering to use the expansion.
llvm-svn: 349025
MULX has somewhat improved register allocation constraints compared to the legacy MUL instruction. Both output registers are encoded instead of fixed to EAX/EDX, but EDX is used as input. It also doesn't touch flags. Unfortunately, the encoding is longer.
Prefering it whenever BMI2 is enabled is probably not optimal. Choosing it should somehow be a function of register allocation constraints like converting adds to three address. gcc and icc definitely don't pick MULX by default. Not sure what if any rules they have for using it.
Differential Revision: https://reviews.llvm.org/D55565
llvm-svn: 348975
I'm hoping we can just replace SETCC_CARRY with SBB. This is another step towards that.
I've explicitly used zero as the input to the setcc to avoid a false dependency that we've had with the SETCC_CARRY. I changed one of the patterns that used NEG to instead use an explicit compare with 0 on the LHS. We needed the zero anyway to avoid the false dependency. The negate would clobber its input register. By using a CMP we can avoid that which could be useful.
Differential Revision: https://reviews.llvm.org/D55414
llvm-svn: 348959
This patch introduces a generic function to determine whether a given vector type is known to be a splat value for the specified demanded elements, recursing up the DAG looking for BUILD_VECTOR or VECTOR_SHUFFLE splat patterns.
It also keeps track of the elements that are known to be UNDEF - it returns true if all the demanded elements are UNDEF (as this may be useful under some circumstances), so this needs to be handled by the caller.
A wrapper variant is also provided that doesn't take the DemandedElts or UndefElts arguments for cases where we just want to know if the SDValue is a splat or not (with/without UNDEFS).
I had hoped to completely remove the X86 local version of this function, but I'm seeing some regressions in shift/rotate codegen that will take a little longer to fix and I hope to get this in sooner so I can continue work on PR38243 which needs more capable splat detection.
Differential Revision: https://reviews.llvm.org/D55426
llvm-svn: 348953
This extends the code that handles 16-bit add promotion to form LEA to also allow 8-bit adds.
That allows us to combine add ops with register moves and save some instructions. This is
another step towards allowing add truncation in generic DAGCombiner (see D54640).
Differential Revision: https://reviews.llvm.org/D55494
llvm-svn: 348946
Summary:
When doing X86CondBrFolding::analyzeCompare, it will meet the SUB32ri instruction as below to use the global address for its operand,
%733:gr32 = SUB32ri %62:gr32(tied-def 0), @img2buf_normal, implicit-def $eflags
JNE_1 %bb.41, implicit $eflags
so the assertion "assert(MI.getOperand(ValueIndex).isImm() && "Expecting Imm operand")" is not correct and change the assert to if make X86CondBrFolding::analyzeCompare return false as not finding the compare for this
Patch by Jianping Chen
Reviewers: smaslov, LuoYuanke, liutianle, Jianping
Reviewed By: Jianping
Subscribers: lebedev.ri, llvm-commits
Differential Revision: https://reviews.llvm.org/D54250
llvm-svn: 348853
As discussed in D55494, we want to extend this to handle 8-bit
ops too, but that could be extended further to enable this on
32-bit systems too.
llvm-svn: 348851
As discussed in:
D55494
...this code has been disabled/dead for a long time (the code references
Athlon and Pentium 4), and there's almost no chance that it will be used
given the last decade of uarch evolution. Also, in SDAG we promote 16-bit
ops to 32-bit, so there's almost no way to test this code any more.
llvm-svn: 348845
This patch restricts the capability of G_MERGE_VALUES, and uses the new
G_BUILD_VECTOR and G_CONCAT_VECTORS opcodes instead in the appropriate places.
This patch also includes AArch64 support for selecting G_BUILD_VECTOR of <4 x s32>
and <2 x s64> vectors.
Differential Revisions: https://reviews.llvm.org/D53629
llvm-svn: 348788
This should really be generalized to allow increment and/or
we should replace it by using ISD::matchUnaryPredicate().
See D55515 for context.
llvm-svn: 348776
Fixes https://bugs.llvm.org/show_bug.cgi?id=39926.
The size of the first copy was computed as
std::abs(std::abs(LdDisp2) - std::abs(LdDisp1)), which results in
skipped bytes if the signs of LdDisp2 and LdDisp1 differ. As far as
I can see, this should just be LdDisp2 - LdDisp1. The case where
LdDisp1 > LdDisp2 is already handled in the code above, in which case
LdDisp2 is set to LdDisp1 and this subtraction will evaluate to
Size1 = 0, which is the correct value to skip an overlapping copy.
Differential Revision: https://reviews.llvm.org/D55485
llvm-svn: 348750
Both intrinsics do the exact same thing so we really only need one.
Earlier in the 8.0 cycle we changed the signature of this intrinsic without renaming it. But it looks difficult to get the autoupgrade code to allow me to merge the intrinsics and change the signature at the same time. So I've renamed the intrinsic slightly for the new merged intrinsic. I'm skipping autoupgrading from the previous new to 8.0 signature. I've also renamed the subborrow for consistency.
llvm-svn: 348737
Previously we had to take the carry in and add -1 to it to set the carry flag so we could use it with ADC/SBB. But if we know its 0 then we don't need to bother.
This should go a long way towards fixing PR24545.
llvm-svn: 348727
The dependency was added in r213995 in response to r213986 which did make
X86/Utils depend on IR, but r256680 later removed that dependency again.
llvm-svn: 348724
The existing code tries to handle an undef operand while transforming an add to an LEA,
but it's incomplete because we will crash on the i16 test with the debug output shown below.
It's better to just give up instead. Really, GlobalIsel should have folded these before we
could get into trouble.
# Machine code for function add_undef_i16: NoPHIs, TracksLiveness, Legalized, RegBankSelected, Selected
bb.0 (%ir-block.0):
liveins: $edi
%1:gr32 = COPY killed $edi
%0:gr16 = COPY %1.sub_16bit:gr32
%5:gr64_nosp = IMPLICIT_DEF
%5.sub_16bit:gr64_nosp = COPY %0:gr16
%6:gr64_nosp = IMPLICIT_DEF
%6.sub_16bit:gr64_nosp = COPY %2:gr16
%4:gr32 = LEA64_32r killed %5:gr64_nosp, 1, killed %6:gr64_nosp, 0, $noreg
%3:gr16 = COPY killed %4.sub_16bit:gr32
$ax = COPY killed %3:gr16
RET 0, implicit killed $ax
# End machine code for function add_undef_i16.
*** Bad machine code: Reading virtual register without a def ***
- function: add_undef_i16
- basic block: %bb.0 (0x7fe6cd83d940)
- instruction: %6.sub_16bit:gr64_nosp = COPY %2:gr16
- operand 1: %2:gr16
LLVM ERROR: Found 1 machine code errors.
Differential Revision: https://reviews.llvm.org/D54710
llvm-svn: 348722
Extension to rL348617, turns out llvm-exegesis doesn't need to match the perf counter name against a scheduler model resource name - so I've added a few more counters that I could find in the libpfm4 source code (and fix a typo in the knl/knm retired_uops counter - which uses 'all' instead of 'any').
llvm-svn: 348721
To make X86CondBrFoldingPass can be run with --run-pass option, this can test one wrong assertion on analyzeCompare function for SUB32ri when its operand is not imm
Patch by Jianping Chen
Differential Revision: https://reviews.llvm.org/D55412
llvm-svn: 348620
This patch attempts to improve pfm perf counter coverage for all the x86 CPUs that libpfm4 supports.
Intel/AMD CPU families tend to share names for cycle/uops counters so even if they don't have a scheduler model yet they can at least use the default values (checked against the libpfm4 source code).
The remaining CPUs (where their port/pipe resource counters are known) I've tried to add to the existing model mappings.
These are untested but don't represent a regression to current llvm-exegesis behaviour for these CPUs.
Differential Revision: https://reviews.llvm.org/D55432
llvm-svn: 348617
Adds fatal errors for any target that does not support the Tiny or Kernel
codemodels by rejigging the getEffectiveCodeModel calls.
Differential Revision: https://reviews.llvm.org/D50141
llvm-svn: 348585
This addresses a FIXME and avoids depending on an isel pattern match I think. I've remove the isel patterns too since he have no lit tests left that cover them. Hopefully that really means they are unused.
I'm trying to decide if we need SETCC_CARRY. This removes one of its usages.
Differential Revision: https://reviews.llvm.org/D55355
llvm-svn: 348536
Initial step towards making the function more generic (and probably move into SelectionDAG).
This is necessary to avoid massive codegen bloat for PR38243 (Add modulo rotate support to LowerRotate).
llvm-svn: 348498
Whenever we effectively take the address of a basic block we need to
manually update that basic block to reflect that fact or later passes
such as tail duplication and tail merging can break the invariants of
the code. =/ Sadly, there doesn't appear to be any good way of
automating this or even writing a reasonable assert to catch it early.
The change seems trivially and obviously correct, but sadly the only
really good test case I have is 1000s of basic blocks. I've tried
directly writing a test case that happens to make tail duplication do
something that crashes later on, but this appears to require an
*amazingly* complex set of conditions that I've not yet reproduced.
The change is technically covered by the tests because we mark the
blocks as having their address taken, but that doesn't really count as
properly testing the functionality.
llvm-svn: 348374
Prep work for PR38243 - mainly adding comments on where we need to add modulo support (doing so at the moment causes massive codegen regressions).
I've also consistently added support for modulo folding for uniform constants (although at the moment we have no way to trigger this) and removed the old assertions.
llvm-svn: 348366
This is an initial patch to add a minimum level of support for funnel shifts to the SelectionDAG and to begin wiring it up to the X86 SHLD/SHRD instructions.
Some partial legalization code has been added to handle the case for 'SlowSHLD' where we want to expand instead and I've added a few DAG combines so we don't get regressions from the existing DAG builder expansion code.
Differential Revision: https://reviews.llvm.org/D54698
llvm-svn: 348353
PR17686 demonstrates that for some targets FP exceptions can fire in cases where the FP_TO_UINT is expanded using a FP_TO_SINT instruction.
The existing code converts both the inrange and outofrange cases using FP_TO_SINT and then selects the result, this patch changes this for 'strict' cases to pre-select the FP_TO_SINT input and the offset adjustment.
The X87 cases don't need the strict flag but generates much nicer code with it....
Differential Revision: https://reviews.llvm.org/D53794
llvm-svn: 348251
We only needed this because it provided really aggressive constant folding even through constant pool entries created from build_vectors. The main case was for vXi8 MULH legalization which was happening as part of legalize DAG instead of as part of legalize vector ops. Now its part of vector op legalization and we've added special handling for build vectors of all constants there. This has removed the need for this code on the list tests we have.
llvm-svn: 348237
This is the smallest vector enhancement I could find to D54640.
Here, we're allowing narrowing to only legal vector ops because we'll see
regressions without that. All of the test diffs are wins from what I can tell.
With AVX/AVX512, we can shrink ymm/zmm ops to xmm.
x86 vector multiplies are the problem case that we're avoiding due to the
patchwork ISA, and it's not clear to me if we can dance around those
regressions using TLI hooks or if we need preliminary patches to plug those
holes.
Differential Revision: https://reviews.llvm.org/D55126
llvm-svn: 348195
Summary:
We need to unpackl and unpackh the operands to use two vXi16 multiplies. Previously it looks like the low unpack would get constant folded at least in the 128-bit case after shuffle lowering turned the unpackl into ZERO_EXTEND_VECTOR_INREG and X86 custom DAG combined it. The same doesn't happen for the high half. So we'd load a constant and then shuffle it. But the low half would just be loaded and used by the multiply directly.
After this patch we now end up with a constant pool entry for the low and high unpacks separately with no shuffle operations.
This is a step towards removing custom constant folding for ZERO_EXTEND_VECTOR_INREG/SIGN_EXTEND_VECTOR_INREG in the X86 backend.
Reviewers: RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D55165
llvm-svn: 348159
Summary:
Under -x86-experimental-vector-widening-legalization, fp_to_uint/fp_to_sint with a smaller than 128 bit vector type results are custom type legalized by promoting the result to a 128 bit vector by promoting the elements, inserting an assertzext/assertsext, then truncating back to original type. The truncate will be further legalizdd to a pack shuffle. In the case of a v8i8 result type, we'll end up with a v8i16 fp_to_sint. This will need to be further legalized during vector op legalization by promoting to v8i32 and then truncating again. Under avx2 this produces good code with two pack instructions, but Under avx512 this will result in a truncate instruction and a packuswb instruction. But we should be able to get away with a single truncate instruction.
The other option is to promote all the way to vXi32 result type during the first type legalization. But in some experimentation that seemed to require more work to produce good code for other configurations.
Reviewers: RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D54836
llvm-svn: 348158
Previously this code generated its own extracts and build_vector. But we can use a simpler concat_vectors or scalar_to_vector operation and let type legalization do additional legalization of those operations.
llvm-svn: 348087
The generic legalizer will fall back to a stack spill that uses a truncating store. That store will get expanded into a shuffle and non-truncating store on pre-avx512 targets. Once that happens the stack store/load pair will be combined away leaving behind the shuffle and bitcasts. On avx512 targets the truncating store is legal so doesn't get folded away.
By custom legalizing it we can avoid this churn and maybe produce better code.
llvm-svn: 348085
Summary: With sse4.1 we use two zero_extend_vector_inreg and a pshufd to expand the v16i8 input into two v8i16 vectors for the multiply. That's 3 shuffles to extend one operand. The other operand is usually constant as this is mostly used by division by constant optimization. Pre sse4.1 we use a punpckhbw and a punpcklbw with a zero vector. That's two shuffles and an xor and a copy due to tied register constraints. That seems maybe better than the 3 shuffles. With AVX we avoid the copy so that's obviously better.
Reviewers: spatel, RKSimon
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D55138
llvm-svn: 348079
This reduces the number of shuffle operations that need to be done. The splitting strategy requires the shuffle unit for the extraction and the extension. With the unpack strategy the unpacks accomplish a splitting and extending in one operation.
llvm-svn: 348019
This does require a constant pool load instead of loading an immediate into a gpr, moving to a k register and masking. But its less instructions and more consistent with previous ISAs. It probably opens up more combine opportunities as one of the test cases demonstrates.
llvm-svn: 348018
Previously we emitted a punpcklbw/punpckhbw to move the byte elements into the upper half of 16 bit elements then shifted right by 8 to zero the upper bits. After DAG combine we end up with punpcklbw/punpckhbw into the lower bits with zeros in the uppers bits and no shifts. So just emit that directly.
llvm-svn: 347966
We had a EVT variable capturing the result of getSimpleValueType which returns an MVT. Another place using EVT that could have been MVT. And an 'int' that should be 'unsigned'.
llvm-svn: 347959
Summary:
Suppressed warnings in release builds due to variable used
only in assert statement.
Subscribers: llvm-commits, eraman, mgorny
Differential Revision: https://reviews.llvm.org/D55100
llvm-svn: 347939
I believe we should be legalizing these with the rest of vector binary operations. If any custom lowering is required for these nodes, this will give the DAG combine between LegalizeVectorOps and LegalizeDAG to run on the custom code before constant build_vectors are lowered in LegalizeDAG.
I've moved MULHU/MULHS handling in AArch64 from Lowering to isel. Moving the lowering earlier caused build_vector+extract_subvector simplifications to kick in which made the generated code worse.
Differential Revision: https://reviews.llvm.org/D54276
llvm-svn: 347902
This is another patch for -x86-experimental-vector-widening. This pre widens narrow division by constants so that we can get pass the legal type check in the generic DAG combiner. Otherwise we end up scalarizing.
I've restricted this to splats for now because it was easy to just call DAG.getConstant. Not sure what we should do for non-splat? Increase the element size?Widen the constant vector by padding with 1?
Differential Revision: https://reviews.llvm.org/D54919
llvm-svn: 347898
It causes asserts building BoringSSL. See https://crbug.com/91009#c3 for
repro.
This also reverts the follow-ups:
Revert r347724 "Do not insert prefetches with unsupported memory operands."
Revert r347606 "[X86] Add dependency from X86 to ProfileData after rL347596"
Revert r347607 "Add new passes to X86 pipeline tests"
llvm-svn: 347864
This patch adds the ability to specify via tablegen which processor resources
are load/store queue resources.
A new tablegen class named MemoryQueue can be optionally used to mark resources
that model load/store queues. Information about the load/store queue is
collected at 'CodeGenSchedule' stage, and analyzed by the 'SubtargetEmitter' to
initialize two new fields in struct MCExtraProcessorInfo named `LoadQueueID` and
`StoreQueueID`. Those two fields are identifiers for buffered resources used to
describe the load queue and the store queue.
Field `BufferSize` is interpreted as the number of entries in the queue, while
the number of units is a throughput indicator (i.e. number of available pickers
for loads/stores).
At construction time, LSUnit in llvm-mca checks for the presence of extra
processor information (i.e. MCExtraProcessorInfo) in the scheduling model. If
that information is available, and fields LoadQueueID and StoreQueueID are set
to a value different than zero (i.e. the invalid processor resource index), then
LSUnit initializes its LoadQueue/StoreQueue based on the BufferSize value
declared by the two processor resources.
With this patch, we more accurately track dynamic dispatch stalls caused by the
lack of LS tokens (i.e. load/store queue full). This is also shown by the
differences in two BdVer2 tests. Stalls that were previously classified as
generic SCHEDULER FULL stalls, are not correctly classified either as "load
queue full" or "store queue full".
About the differences in the -scheduler-stats view: those differences are
expected, because entries in the load/store queue are not released at
instruction issue stage. Instead, those are released at instruction executed
stage. This is the main reason why for the modified tests, the load/store
queues gets full before PdEx is full.
Differential Revision: https://reviews.llvm.org/D54957
llvm-svn: 347857
This failed to select (which might be a separate bug) in
X86ISelDAGToDAG because we try to create a select node
that can be simplified away after rL347227.
This change avoids the problem by simplifying the SHRUNKBLEND
node sooner. In the test case, we manage to realize that the
true/false values of the select (SHRUNKBLEND) are the same thing,
so it simplifies away completely.
llvm-svn: 347818
Unlike most cost model functions this code makes a lot of table lookups without using the results from getTypeLegalizationCost. This means 512-bit vectors can be looked up even when the type isn't legal.
This patch adds a check around the two tables that contain 512-bit types to make sure that neither of the types would be split by type legalization. Meaning 512 bit types are illegal. I wanted to write this in a somewhat generic way that uses type legalization query hooks. But if prefered, I can switch to just using is512BitVector and the subtarget feature.
Differential Revision: https://reviews.llvm.org/D54984
llvm-svn: 347786
This fixes some of scalarization costs reported for sext/zext using avx512bw. This does not fix all scalarization costs being reported. Just the worst.
I've restricted this only to combinations of types that are legal with avx512bw like v32i1/v64i1/v32i16/v64i8 and conversions between vXi1 and vXi8/vXi16 with legal vXi8/vXi16 result types.
Differential Revision: https://reviews.llvm.org/D54979
llvm-svn: 347785
Expansion of SIGN_EXTEND_INREG can create a VSRAI instruction. If there is already a VSRAI after it, we should combine them into a larger VSRAI
Differential Revision: https://reviews.llvm.org/D54959
llvm-svn: 347784
Currently, instructions doing memory accesses through a base operand that is
not a register can not be analyzed using `TII::getMemOpBaseRegImmOfs`.
This means that functions such as `TII::shouldClusterMemOps` will bail
out on instructions using an FI as a base instead of a register.
The goal of this patch is to refactor all this to return a base
operand instead of a base register.
Then in a separate patch, I will add FI support to the mem op clustering
in the MachineScheduler.
Differential Revision: https://reviews.llvm.org/D54846
llvm-svn: 347746
We're already mixing this APInt with other 'unsigned' variables. This allows us to use regular comparison operators instead of needing to use APInt::ult or APInt::uge. And it removes a later conversion from APInt to unsigned.
I might be adding another combine to this function and this will probably simplify the logic required for that.
llvm-svn: 347684
This is skylake-avx512 with the addition of avx512vnni ISA.
Patch by Jianping Chen
Differential Revision: https://reviews.llvm.org/D54785
llvm-svn: 347681
If we fold the bitcast into the store we'll end up creating a truncating store to vXi1 that will get scalarized. Instead allow the bitcast to be turned into a movmsk.
We probably need to do something if the store itself is a vXi1 type, but I'll leave that til a testcase appears.
llvm-svn: 347632
Summary:
Support for profile-driven cache prefetching (X86)
This change is part of a larger system, consisting of a cache prefetches recommender, create_llvm_prof (https://github.com/google/autofdo), and LLVM.
A proof of concept recommender is DynamoRIO's cache miss analyzer. It processes memory access traces obtained from a running binary and identifies patterns in cache misses. Based on them, it produces a csv file with recommendations. The expectation is that, by leveraging such recommendations, we can reduce the amount of clock cycles spent waiting for data from memory. A microbenchmark based on the DynamoRIO analyzer is available as a proof of concept: https://goo.gl/6TM2Xp.
The recommender makes prefetch recommendations in terms of:
* the binary offset of an instruction with a memory operand;
* a delta;
* and a type (nta, t0, t1, t2)
meaning: a prefetch of that type should be inserted right before the instrution at that binary offset, and the prefetch should be for an address delta away from the memory address the instruction will access.
For example:
0x400ab2,64,nta
and assuming the instruction at 0x400ab2 is:
movzbl (%rbx,%rdx,1),%edx
means that the recommender determined it would be beneficial for a prefetchnta instruction to be inserted right before this instruction, as such:
prefetchnta 0x40(%rbx,%rdx,1)
movzbl (%rbx, %rdx, 1), %edx
The workflow for prefetch cache instrumentation is as follows (the proof of concept script details these steps as well):
1. build binary, making sure -gmlt -fdebug-info-for-profiling is passed. The latter option will enable the X86DiscriminateMemOps pass, which ensures instructions with memory operands are uniquely identifiable (this causes ~2% size increase in total binary size due to the additional debug information).
2. collect memory traces, run analysis to obtain recommendations (see above-referenced DynamoRIO demo as a proof of concept).
3. use create_llvm_prof to convert recommendations to reference insertion locations in terms of debug info locations.
4. rebuild binary, using the exact same set of arguments used initially, to which -mllvm -prefetch-hints-file=<file> needs to be added, using the afdo file obtained at step 3.
Note that if sample profiling feedback-driven optimization is also desired, that happens before step 1 above. In this case, the sample profile afdo file that was used to produce the binary at step 1 must also be included in step 4.
The data needed by the compiler in order to identify prefetch insertion points is very similar to what is needed for sample profiles. For this reason, and given that the overall approach (memory tracing-based cache recommendation mechanisms) is under active development, we use the afdo format as a syntax for capturing this information. We avoid confusing semantics with sample profile afdo data by feeding the two types of information to the compiler through separate files and compiler flags. Should the approach prove successful, we can investigate improvements to this encoding mechanism.
Reviewers: davidxl, wmi, craig.topper
Reviewed By: davidxl, wmi, craig.topper
Subscribers: davide, danielcdh, mgorny, aprantl, eraman, JDevlieghere, llvm-commits
Differential Revision: https://reviews.llvm.org/D54052
llvm-svn: 347596
SplitVecOp_TruncateHelper tries to promote the result type while splitting FP_TO_SINT/UINT. It then concatenates the result and introduces a truncate to the original result type. But it does this without inserting the AssertZExt/AssertSExt that the regular result type promotion would insert. Nor does it turn FP_TO_UINT into FP_TO_SINT the way normal result type promotion for these operations does. This is bad on X86 which doesn't support FP_TO_SINT until AVX512.
This patch disables the use of SplitVecOp_TruncateHelper for these operations and just lets normal promotion handle it. I've tweaked a couple things in X86ISelLowering to avoid a few obvious regressions there. I believe all the changes on X86 are improvements. The other targets look neutral.
Differential Revision: https://reviews.llvm.org/D54906
llvm-svn: 347593
Summary:
Add a hook to the GCMetadataPrinter for emitting stack maps in
custom format. The hook will be called at stack map generation
time. The default stack map format is used if there is no hook.
For this to be useful a few data structures and accessors are
exposed from the StackMaps class, so the custom printer can
access the stack map data.
This patch authored by Cherry Zhang <cherryyz@google.com>.
Reviewers: thanm, apilipenko, reames
Reviewed By: reames
Subscribers: reames, apilipenko, nemanjai, javed.absar, kbarton, jsji, llvm-commits
Differential Revision: https://reviews.llvm.org/D53892
llvm-svn: 347584
We have these 2 "isDesirable" promotion hooks (I'm not sure why we need both of them, but that's
independent of this patch), and we can adjust them to promote "mul i8 X, C" to i32. Then, all of
our existing LEA and other multiply expansion magic happens as it would for i32 ops.
Some of the test diffs show that we could end up with an actual 32-bit mul instruction here
because we choose not to expand to simpler ops. That instruction could be slower depending on the
subtarget. On the plus side, this means we don't need a separate instruction to load the constant
operand and possibly an extra instruction to move the result. If we need to tune mul i32 further,
we could add a later transform that tries to shrink it back to i8 based on subtarget timing.
I did not bother to duplicate all of the 32-bit test file RUNs and target settings that exist to
test whether LEA expansion is cheap or not. The diffs here assume a default target, so that means
LEA is generally cheap.
Differential Revision: https://reviews.llvm.org/D54803
llvm-svn: 347557
This should likely be adjusted to limit this transform
further, but these diffs should be clear wins.
If we have blendv/conditional move, then we should assume
those are cheap ops. The loads become independent of the
compare, so those can be speculated before we need to use
the values in the blend/mov.
llvm-svn: 347526
These are AVX2 instructions, but have been incorrectly marked in tablegen for a while. This wasn't a problem until r346784 switched the patterns to use target independent ISD opcodes. This made the patterns visible to fast isel.
Fixes PR39733
llvm-svn: 347375
We can't guarantee that demanded bits passing through the vector shuffle won't cause the AND in front of this to be removed. This would prevent the PACKUS from being matched during shuffle lowering.
Unfortunately, this adds a packuswb to one of the vector-reduce-mul.ll tests since we were removing the shuffle via SimplifyDemandedVectorElts. We appear to have similar issues with vpmovwb on the same test case on other targets.
llvm-svn: 347361
Previously we emitted to separate shuffles, one for unpcklbw and one for unpcklwd. Instead emit a single shuffle equivalent to both of the original shuffles. Shuffle lowering seems able to handle it. This avoids a bitcast between the two shuffles which seems helpful to DAG combine.
Remove the custom type legalization for v8i8->v8i32. I had put that in to avoid some almost duplicate punpcklbw instructions I was seeing, but this lowering change seems to fix that. It also fixes some duplicate shuffles seen in vector-sext.ll
llvm-svn: 347348
Pull out getPackDemandedElts demanded elts remapping helper from computeKnownBitsForTargetNode and use in computeKnownBits/ComputeNumSignBits.
llvm-svn: 347303
Previously if V2 was unused we ended up using V1 for both inputs as part of the code that follows the new code. By using lowerVectorShuffleWithUNPCK we keep the undef nature of V2 in the output.
As near as I can tell this makes v16i8 behavior consistent with every other VT now.
This does mean that we give the register allocator freedom to fill in random registers now and create false dependencies. But like I said we're already doing that for other types.
llvm-svn: 347296
getZeroVector produces a specifically canonicalized zero vector, but we can just let DAG legalization take care of it.
The test changes are because MULH lowering happens later than it should and this change gave us the opportunity to constant fold away a multiply during a DAG combine before the build_vector got legalized with a bitcast.
llvm-svn: 347290
We're seeing some issues internally where we sent some intrinsics into the cost model that the getTypeLegalizationCost call fails on, but X86 specific tables don't care about. Our base class implementation takes care of them. We'd just like X86 backend to ignore them.
This patch makes sure the switch returned something X86 cares about and skips the table lookups and type legalization call if not. Probably more efficient too since we don't go scanning the tables for every intrinsic we could possibly see.
Differential Revision: https://reviews.llvm.org/D54711
llvm-svn: 347248
SSE PSHUFB vector ctlz lowering works at the i4 nibble level. As detailed in PR39703, we were masking the lower nibble off but we only actually use it in the case where the upper nibble is known to be zero, making it safe to remove the mask and save an instruction.
Differential Revision: https://reviews.llvm.org/D54707
llvm-svn: 347242
Previously we split the vectors in half to allow the two halves to be any extended then concatenated the results back together.
This patch instead instead extends the v16i8 sse algorithm to extend half of each 128-bit lane using punpcklbw/punpckhbw. Multiplies all the low half lanes and high half lanes together in separate operations. Then merges the half lane results back together using packuswb.
Unfortunately, some of the cases in vector-reduce-mul.ll regress because we aren't narrowing the vector width of the multiplies as we reduce. The splitting was somewhat making up for that before by causing halves to be discarded after the split.
Differential Revision: https://reviews.llvm.org/D54668
llvm-svn: 347240
The shift requires a copy to avoid clobbering a register. Comparing with 0 uses an xor to produce 0 that will be overwritten with the compare results. So still requires 2 instructions, but should be one byte shorter since it doesn't need to encode an immediate.
llvm-svn: 347185
Previously we used an arithmetic shift right by 31, but that requires a copy to preserve the input. So we might as well materialize a zero and compare to it since the comparison will overwrite the register that contains the zeros. This should be one byte shorter.
llvm-svn: 347181
Leave just the v4i8->v4i64 and v8i8->v8i64, but only enable them on pre-sse4.1 targets when 64-bit mode is enabled. In those cases we end up creating sext loads that get scalarized to code that looks better than what we get from loading into a vector register and doing a multiple step sign extend using unpacks and shifts.
llvm-svn: 347180
Pre-SSE4.1 sext_invec for v2i64 is complicated because we don't have a v2i64 sra instruction. So instead we sign extend to i32 using unpack and sra, then copy the elements and do a v4i32 sra to fill with sign bits, then interleave the i32 sign extend and the sign bits. So really we're doing to two sign extends but only using half of the v4i32 intermediate result.
When the result is more than 128 bits, default type legalization would prefer to split the destination type all the way down to v2i64 with shuffles followed by v16i8/v8i16->v2i64 sext_inreg operations. This results in more instructions than necessary because we are only utilizing the lower 2 elements of the v4i32 intermediate result. Instead we can custom split a v4i8/v4i16->v4i64 sign_extend. Then we can sign extend v4i8/v4i16->v4i32 invec producing a full v4i32 result. Create the sign bit vector as a v4i32 then split and interleave with the sign bits using an punpackldq and punpackhdq.
llvm-svn: 347176
Simon Pilgrim