It also shows an unnecessary pshufb/broadcast being used - the original pshufb mask only requested the lowest byte.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295046 91177308-0d34-0410-b5e6-96231b3b80d8
This adds MXCSR to the set of recognized registers for X86 targets and updates the instructions that read or write it. I do not intend for all of the various floating point instructions that implicitly use the control bits or update the status bits of this register to ever have that usage modeled by default. However, when constrained floating point modes (such as strict FP exception status modeling or dynamic rounding modes) are enabled, implicit use/def information for MXCSR will be added to those instructions.
Until those additional updates are made this should cause (almost?) no functional changes. Theoretically, this will prevent instructions like LDMXCSR and STMXCSR from being moved past one another, but that should be prevented anyway and I haven't found a case where it is happening now.
Differential Revision: https://reviews.llvm.org/D29903
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@295004 91177308-0d34-0410-b5e6-96231b3b80d8
Backends don't support this yet. They would have to move to the swifterror
register before the tail call to make sure it is live-in to the call.
rdar://30495920
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294982 91177308-0d34-0410-b5e6-96231b3b80d8
This is consistent with what we do for GlobalISel. That way, it is easy
to see whether or not FastISel is able to fully select a function.
At some point we may want to switch that to an optimization remark.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294970 91177308-0d34-0410-b5e6-96231b3b80d8
We now detect that both the extract and insert indices are non-zero and convert to a shuffle. This will be lowered as a blend for 256-bit vectors or as a vshuf operations for 512-bit vectors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294931 91177308-0d34-0410-b5e6-96231b3b80d8
This gives more parallelism opportunities for AVX-512 when dealing with 128-bit extracts from 512-bit vectors.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294930 91177308-0d34-0410-b5e6-96231b3b80d8
The bug was introduced with:
https://reviews.llvm.org/rL294863
...and manifests as a selection failure in x86, but that's actually
another bug. This fix prevents wrong codegen with -0.0, but in the
more common case when we have NSZ and NNAN (-ffast-math), we should
still be able to fold this setcc/compare.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294924 91177308-0d34-0410-b5e6-96231b3b80d8
Initial 256-bit vector support - 512-bit support requires extra checks for AVX512BW support (PMOVZXBW) that will be handled in a future patch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294896 91177308-0d34-0410-b5e6-96231b3b80d8
Generalize VSEXT/VZEXT constant folding to work with any target constant bits source not just BUILD_VECTOR .
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294873 91177308-0d34-0410-b5e6-96231b3b80d8
I don't know if anything other than x86 vectors is affected by this change, but this may allow
us to remove target-specific intrinsics for blendv* (vector selects). The simplification arises
from the fact that blendv* instructions only use the sign-bit when deciding which vector element
to choose for the destination vector. The mechanism to fold VSELECT into SHRUNKBLEND nodes already
exists in x86 lowering; this demanded bits change just enables the transform to fire more often.
The original motivation starts with a bug for DSE of masked stores that seems completely unrelated,
but I've explained the likely steps in this series here:
https://llvm.org/bugs/show_bug.cgi?id=11210
Differential Revision: https://reviews.llvm.org/D29687
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294863 91177308-0d34-0410-b5e6-96231b3b80d8
All commutations confirmed to give identical results - note PFMAX/PFMIN do not
PFSUB<->PFSUBR should be commutable as well
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294846 91177308-0d34-0410-b5e6-96231b3b80d8
Seems the execution dependency pass likes to use FP instructions when most of the consuming code is integer if a vextractf128 instruction produced the register. Without AVX2 we don't have the corresponding integer instruction available.
This patch suppresses the domain on these instructions to GenericDomain if AVX2 is not supported so that they are ignored by domain fixing. If AVX2 is supported we'll report the correct domain and allow them to switch between integer and fp.
Overall I think this produces better results in the modified test cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294824 91177308-0d34-0410-b5e6-96231b3b80d8
The recommit includes some changes of testcases. No functional change to the patch.
In RateRegister of existing LSR, if a formula contains a Reg which is a SCEVAddRecExpr,
and this SCEVAddRecExpr's loop is an outerloop, the formula will be marked as Loser
and dropped.
Suppose we have an IR that %for.body is outerloop and %for.body2 is innerloop. LSR only
handle inner loop now so only %for.body2 will be handled.
Using the logic above, formula like
reg(%array) + reg({1,+, %size}<%for.body>) + 1*reg({0,+,1}<%for.body2>) will be dropped
no matter what because reg({1,+, %size}<%for.body>) is a SCEVAddRecExpr type reg related
with outerloop. Only formula like
reg(%array) + 1*reg({{1,+, %size}<%for.body>,+,1}<nuw><nsw><%for.body2>) will be kept
because the SCEVAddRecExpr related with outerloop is folded into the initial value of the
SCEVAddRecExpr related with current loop.
But in some cases, we do need to share the basic induction variable
reg{0 ,+, 1}<%for.body2> among LSR Uses to reduce the final total number of induction
variables used by LSR, so we don't want to drop the formula like
reg(%array) + reg({1,+, %size}<%for.body>) + 1*reg({0,+,1}<%for.body2>) unconditionally.
From the existing comment, it tries to avoid considering multiple level loops at the same time.
However, existing LSR only handles innermost loop, so for any SCEVAddRecExpr with a loop other
than current loop, it is an invariant and will be simple to handle, and the formula doesn't have
to be dropped.
Differential Revision: https://reviews.llvm.org/D26429
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294814 91177308-0d34-0410-b5e6-96231b3b80d8
Since r274013, we've been looking through bitcasts on broadcast inputs.
In the scalar-folding case (from a load, build_vector, or sc2vec),
the input type didn't matter, as we'd simply bitcast the resulting
scalar back.
However, when broadcasting a 128-bit-lane-aligned element, we create an
EXTRACT_SUBVECTOR. Use proper types, by creating an extract_subvector
of the original input type.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294774 91177308-0d34-0410-b5e6-96231b3b80d8
Tests to demonstrate throughput-latency decision between div and rcp on faster hardware such as Haswell
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294750 91177308-0d34-0410-b5e6-96231b3b80d8
The patch comes in 2 parts:
1 - it makes use of the SelectionDAG::NewNodesMustHaveLegalTypes flag to tell when it can safely constant fold illegal types.
2 - it correctly resets SelectionDAG::NewNodesMustHaveLegalTypes at the start of each call to SelectionDAGISel::CodeGenAndEmitDAG so all the pre-legalization stages can make use of it - not just the first basic block that gets handled.
Fix for PR30760
Differential Revision: https://reviews.llvm.org/D29568
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294749 91177308-0d34-0410-b5e6-96231b3b80d8
In some cases we call getTargetConstantBitsFromNode for nodes that haven't been lowered from BUILD_VECTOR yet
Note: We're getting very close to being able to move most of the constant extraction code from getTargetShuffleMaskIndices into getTargetConstantBitsFromNode
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294746 91177308-0d34-0410-b5e6-96231b3b80d8
This requires that we communicate to X86InstrInfo::optimizeCompareInstr
that the second operand is neither a register nor an immediate. The way we
do that is by setting CmpMask to zero.
Note that there were already instructions where the second operand was not a
register nor an immediate, namely X86::SUB*rm, so also set CmpMask to zero
for those instructions. This seems like a latent bug, but I was unable to
trigger it.
Differential Revision: https://reviews.llvm.org/D28621
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@294634 91177308-0d34-0410-b5e6-96231b3b80d8
Simon Pilgrim