Age | Commit message (Collapse) | Author |
|
As experimental backends, I didn't have them configured to build in my local
build config.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321696 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Currently it's not possible to access MCSubtargetInfo from a TgtMCAsmBackend.
D20830 threaded an MCSubtargetInfo reference through
MCAsmBackend::relaxInstruction, but this isn't the only function that would
benefit from access. This patch removes the Triple and CPUString arguments
from createMCAsmBackend and replaces them with MCSubtargetInfo.
This patch just changes the interface without making any intentional
functional changes. Once in, several cleanups are possible:
* Get rid of the awkward MCSubtargetInfo handling in ARMAsmBackend
* Support 16-bit instructions when valid in MipsAsmBackend::writeNopData
* Get rid of the CPU string parsing in X86AsmBackend and just use a SubtargetFeature for HasNopl
* Emit 16-bit nops in RISCVAsmBackend::writeNopData if the compressed instruction set extension is enabled (see D41221)
This change initially exposed PR35686, which has since been resolved in r321026.
Differential Revision: https://reviews.llvm.org/D41349
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321692 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Differential Revision: https://reviews.llvm.org/D40524
Change-Id: Ie3a405b28503ceae999f5f3ba07a68fa733a2400
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321674 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
(PR33325)
This is an extension of D31156 with the goal that we'll allow memcmp() == 0 expansion
for x86 to use 2 pairs of loads per block.
The memcmp expansion pass (formerly part of CGP) will generate this kind of pattern
with oversized integer compares, so we want to transform these into x86-specific vector
nodes before legalization splits things into scalar chunks.
See PR33325 for more details:
https://bugs.llvm.org/show_bug.cgi?id=33325
Differential Revision: https://reviews.llvm.org/D41618
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321656 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Tests updated to explicitly use fast-isel at -O0 instead of implicitly.
This change also allows an explicit -fast-isel option to override an
implicitly enabled global-isel. Otherwise -fast-isel would have no effect at -O0.
Differential Revision: https://reviews.llvm.org/D41362
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321655 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321650 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Summary:
isReg() in AArch64AsmParser.cpp is a bit of a misnomer, and would be better named 'isScalarReg()' instead.
Patch [1/3] in a series to add operand constraint checks for SVE's predicated ADD/SUB.
Reviewers: rengolin, mcrosier, evandro, fhahn, echristo
Reviewed By: fhahn
Subscribers: aemerson, javed.absar, llvm-commits, kristof.beyls
Differential Revision: https://reviews.llvm.org/D41445
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321646 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321644 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Differential Revision: https://reviews.llvm.org/D41339
Patch by Shiva Chen.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321643 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
XLenVT in LowerFormalArguments is used only in an assert.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321642 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321632 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
The custom lowering was just doing the same thing promotion would do.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321630 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
when using setOperationAction Promote for INT_TO_FP and FP_TO_INT
Currently the promotion for these ignores the normal getTypeToPromoteTo and instead just tries to double the element width. This is because the default behavior of getTypeToPromote to just adds 1 to the SimpleVT, which has the affect of increasing the element count while keeping the scalar size the same.
If multiple steps are required to get to a legal operation type, int_to_fp will be promoted multiple times. And fp_to_int will keep trying wider types in a loop until it finds one that works.
getTypeToPromoteTo does have the ability to query a promotion map to get the type and not do the increasing behavior. It seems better to just let the target specify the promotion type in the map explicitly instead of letting the legalizer iterate via widening.
FWIW, it's worth I think for any other vector operations that need to be promoted, we have to specify the type explicitly because the default behavior of getTypeToPromote isn't useful for vectors. The other types of promotion already require either the element count is constant or the total vector width is constant, but neither happens by incrementing the SimpleVT enum.
Differential Revision: https://reviews.llvm.org/D40664
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321629 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
sign bits.
If the input is all sign bits then the LSB through MSB are all the same so we don't need to be move the LSB to the MSB.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321617 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
registers to implement 128/256-bit operations without VLX.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321613 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
false input being zero.
We can use zmm move with zero masking for this. We already had patterns for using a masked move, but we didn't check for the zero masking case separately.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321612 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
vector to v8i1 pre-legalize.
The CONCAT_VECTORS will be lowered to INSERT_SUBVECTOR later. In the modified cases this seems to be enough to trick a later DAG combine into running in a different order than allows the ANDs to be removed.
I'll admit this is a bit of a hack that happens to work, but using CONCAT_VECTORS is more consistent with other legalization code anyway.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321611 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
As it has a scalar source we don't treat it as a target shuffle so needs special handling.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321610 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Don't combine buildvector(binop(),binop(),binop(),binop()) -> binop(buildvector(), buildvector()) if its a splat - keep the binop scalar and just splat the result to avoid large vector constants.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321607 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
legalization sees the i4 and changes to load/store.
Same for v2i1 and i2.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321602 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
legalization sees the i4 and changes to load/store.
Same for i2 and v2i1.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321601 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
don't have DQI.
We end up using an i8 load via an isel pattern from v8i1 anyway. This just makes it more explicit. This seems to improve codgen in some cases and I'd like to kill off some of the load patterns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321598 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This is better handled by a DAG combine if its not already being done. No lit tests fail from the removal of these patterns.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321597 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
I don't think anything would actually expect the other bits to be zero.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321596 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321595 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Use getMemBasePlusOffset and calculate proper pointer info and alignment for the second store.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321594 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321585 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
If the callee and caller use different calling convensions, we cannot apply TCO if the callee requires arguments on stack; e.g. C calling convention and Fast CC use the same registers for parameter passing, but the stack offset is not necessarily same.
This patch also recommit r319218 "[PowerPC] Allow tail calls of fastcc functions from C CallingConv functions." by @sfertile since the problem reported in r320106 should be fixed.
Differential Revision: https://reviews.llvm.org/D40893
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321579 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
natively.
We should only be creating natively supported kshifts now.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321577 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
This allows us to remove some isel patterns.
This is mostly NFC, but we now use KSHIFTB instead of KSHIFTW with DQI.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321576 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321566 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Initially, if the `c` constraint applied to the wrong data type that
causes LLVM to assert. This commit replaces the assert by an error
message.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321565 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Also fixes using the wrong memory type for some
intrinsics when custom lowering them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321557 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Atomics still have hasSideEffects set on them because
of the mess that is the memory properties.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321556 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Currently all images are lowered to have a single
image PseudoSourceValue. Image stores happen to have
overly strict mayLoad/mayStore/hasSideEffects flags
set on them, so this happens to work. When these
are fixed to be correct, the scheduler breaks
this because the identical PSVs are assumed to
be the same address. These need to be unique
to the image resource value.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321555 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
As noted in PR34686, we are relying on a PSHUFD+PSHUFLW+PSHUFHW shuffle chain for most general vXi16 unary shuffles.
This patch checks for simpler PSHUFLW+PSHUFD and PSHUFHW+PSHUFD cases beforehand, building on some existing code that just handled splat shuffles.
By doing so we also prevent premature use of PSHUFB shuffles which can be slower and require the creation/loading of constant shuffle masks.
We now have the 'fast-variable-shuffle' option for hardware that prefers combining 2 or more shuffles to VPSHUFB etc.
Differential Revision: https://reviews.llvm.org/D38318
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321553 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
See bug 35730: https://bugs.llvm.org/show_bug.cgi?id=35730
Differential Revision: https://reviews.llvm.org/D41598
Reviewers: vpykhtin, artem.tamazov, arsenm
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321552 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Revision 320791 introduced a pass that transforms reg+reg instructions to
reg+imm if they're fed by "load immediate". However, it didn't
handle out-of-range shifts correctly as reported in PR35688.
This patch fixes that and therefore the PR.
Furthermore, there was undefined behaviour in the patch where the RHS of an
initialization expression was 32 bits and constant `1` was shifted left 32
bits. This was fixed by ensuring the RHS is 64 bits just like the LHS.
Differential Revision: https://reviews.llvm.org/D41369
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321551 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
and punpckldq.
Differential Revision: https://reviews.llvm.org/D41595
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321549 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
narrower extend.
Previously we used an extend from v8i1 to v8i32/v8i64. Then extracted to the final width. But if we have VLX we should extract first. This way we don't end up with an overly large extend.
This allows us to use vcmpeq to make all ones for the sign extend when DQI isn't available. Otherwise we get a VPTERNLOG.
If we make v2i1/v4i1 legal like proposed in D41560, we could always do this and rely on the lowering of the extend to widen when necessary.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321538 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321537 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
-Use MinAlign instead of std::min.
-Use SelectionDAG::getMemBasePlusOffset.
-Apply offset to the pointer info for the second load/store created.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321536 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
LowerZERO_EXTEND_Mask/LowerSIGN_EXTEND_Mask.
The truncate will be lowered X86ISD::VTRUNC later.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321534 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
The custom lowering already widens the result type to 512-bits if VLX isn't supported.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321533 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
The exception handler thunk needs to reference the LSDA of the parent
function, which won't be emitted if it's available_externally.
Fixes PR35736. ThinLTO ends up producing available_externally functions
that use _CxxFrameHandler3.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321532 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Some output changes from uppercase hex to lowercase hex, no other functionality change intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321526 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
If there are 17 or more leading zeros to the v4i32 elements, then we can use PMADD for the integer multiply when PMULLD is unavailable or slow.
The 17 bits need to be zero as the PMADDWD performs a v8i16 signed-mul-extend + pairwise-add - the upper 16 so we're adding a zero pair and the 17th bit so we don't incorrectly sign extend.
Differential Revision: https://reviews.llvm.org/D41484
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321516 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
Per Table 1-1 in October 2017 edition of Intel® Architecture Instruction Set Extensions and Future Features
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321501 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
My original implementation ran as a DAG combine post type legalization, but it turns out we don't run that DAG combine step if type legalization didn't change anything. Attempts to make the combine run before type legalization as well hit other issues.
So just do it in LowerMUL where we can catch more cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321496 91177308-0d34-0410-b5e6-96231b3b80d8
|
|
r319980 added new patterns to the machine combiner for transforming (fsub (fmul
x y) z) into (fmla (fneg z) x y). That is, fsub's where the first source
operand is an fmul are transformed. We previously only matched the case where
the second source operand of an fsub was an fmul, transforming (fsub z (fmul x
y)) into (fmls z x y). Now, if we have an fsub where both source operands are
fmuls, both of the above patterns are applicable.
However, the order in which we add the patterns to the list of candidates
determines the transformation that takes place, since only the first pattern
that matches will be used. This patch changes the order these two patterns are
added to the list of candidates such that we prefer the case where the second
source operand is an fmul (the fmls case), rather than the other one (the
fmla/fneg case). When both source operands are fmuls, this ordering results in
fewer instructions.
Differential Revision: https://reviews.llvm.org/D41587
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@321491 91177308-0d34-0410-b5e6-96231b3b80d8
|