Summary:
These instructions zero the non-scalar part of the lower 128-bits which makes them different than the FMA3 instructions which pass through the non-scalar part of the lower 128-bits.
I've only added fmadd because we should be able to derive all other variants using operand negation in the intrinsic header like we do for AVX512.
I think there are still some missed negate folding opportunities with the FMA4 instructions in light of this behavior difference that I hadn't noticed before.
I've split the tests so that we can use different intrinsics for scalar testing between the two. I just copied the tests split the RUN lines and changed out the scalar intrinsics.
fma4-fneg-combine.ll is a new test to make sure we negate the fma4 intrinsics correctly though there are a couple TODOs in it.
Reviewers: RKSimon, spatel
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D39851
llvm-svn: 318984
Summary:
This adds a new fast gather feature bit to cover all CPUs that support fast gather that we can use independent of whether the AVX512 feature is enabled. I'm only using this new bit to qualify AVX2 codegen. AVX512 is still implicitly assuming fast gather to keep tests working and to match the scatter behavior.
Test command lines have been added for these two cases.
Reviewers: magabari, delena, RKSimon, zvi
Reviewed By: RKSimon
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D40282
llvm-svn: 318983
v4i32 isn't a legal type with sse1 only and would end up getting scalarized otherwise.
This isn't completely ideal as it doesn't handle cases like v8i32 that would get split to v4i32. But it at least helps with code written using the clang intrinsic header.
llvm-svn: 318967
This optimization can occur after type legalization and emit a vselect with v4i32 type. But that type is not legal with sse1. This ultimately gets scalarized by the second type legalization that runs after vector op legalization, but that's really intended to handle the scalar types that might be introduced by legalizing vector ops.
For now just stop this from happening by disabling the optimization with sse1.
llvm-svn: 318965
This patch fixes an issue where microMIPS ASE flag is not set
when a function has micromips attribute or when .set micromips
directive is used.
Differential Revision: https://reviews.llvm.org/D40316
llvm-svn: 318948
The NewCC variable is calculated outside of the loop that processes jcc/setcc/cmovcc instructions. If we invert it during the loop it can cause an incorrect value to be used by a later iteration. Instead only read it during the loop and use a new variable to store the possibly inverted value.
Fixes PR35399.
llvm-svn: 318934
(V)PHMINPOSUW determines the UMIN element in an v8i16 input, with suitable bit flipping it can also be used for SMAX/SMIN/UMAX cases as well.
This patch matches vXi16 SMAX/SMIN/UMAX/UMIN horizontal reductions and reduces the input down to a v8i16 vector before calling (V)PHMINPOSUW.
A later patch will use this for v16i8 reductions as well (PR32841).
Differential Revision: https://reviews.llvm.org/D39729
llvm-svn: 318917
TableGen already generates code for selecting a G_FDIV, so we only need
to add a test.
For the legalizer and reg bank select, we do the same thing as for the
other floating point binary operations: either mark as legal if we have
a FP unit or lower to a libcall, and map to the floating point
registers.
llvm-svn: 318915
TableGen already generates code for selecting a G_FMUL, so we only need
to add a test for that part.
For the legalizer and reg bank select, we do the same thing as the other
floating point binary operators: either mark as legal if we have a FP
unit or lower to a libcall, and map to the floating point registers.
llvm-svn: 318910
The MIPS delay slot filler converts delay slot branches into compact
forms for the MIPS ISAs which support them. For branches that compare
(in)equality with with zero, it converts them into branches with implict
zero register operands. These branches have a slightly greater range
than normal two register operands branches.
Changing the branches at this point in the pipeline offers the long
branch pass the ability to mark better judgements if a long branch
sequence is required.
Reviewers: atanasyan
Differential Revision: https://reviews.llvm.org/D40314
llvm-svn: 318908
Change LowerBUILD_VECTOR to use those functions. This commit will tempora-
rily affect constant vector generation (it will generate constant-extended
values instead of non-extended combines), but the code for the general case
should be better. The constant selection part will be fixed later.
llvm-svn: 318877
SITargetLowering::LowerCall uses dummy pointer info for byval argument, which causes
flat load instead of buffer load.
This patch fixes that.
Differential Revision: https://reviews.llvm.org/D40040
llvm-svn: 318844
Summary:
This bug seems to have gone unnoticed because critical cases with LDS
instructions are eliminated by the peephole optimizer.
However, equivalent situations arise with buffer loads and stores
as well, so this fixes regressions since r317751 ("AMDGPU: Merge
S_BUFFER_LOAD_DWORD_IMM into x2, x4").
Fixes at least:
KHR-GL45.shader_storage_buffer_object.basic-operations-case1-cs
KHR-GL45.cull_distance.functional
piglit tes-input-gl_ClipDistance.shader_test
... and probably more
Change-Id: I0e371536288eb8e6afeaa241a185266fd45d129d
Reviewers: arsenm, mareko, rampitec
Subscribers: kzhuravl, wdng, yaxunl, dstuttard, tpr, t-tye, llvm-commits
Differential Revision: https://reviews.llvm.org/D40303
llvm-svn: 318829
Since i1 is a legal type, this:
NumBytes = Op1->getMemoryVT().getSizeInBits() >> 3;
is wrong and should be instead
NumBytes = Op0->getMemoryVT().getStoreSize();
There seems to be more places where this should be fixed outside DAGCombiner.
Review: Hal Finkel
https://bugs.llvm.org/show_bug.cgi?id=35366
llvm-svn: 318824
This partially reverts r298851. The the underlying issue is that we don't
currently model the dependency between mrs (read system register) and
msr (write system register) instructions.
Something like the below should never be reordered:
msr TPIDR_EL0, x0 ;; set thread pointer
mrs x8, TPIDR_EL0 ;; read thread pointer
but was being reordered after r298851. The functional part of the patch
that wasn't reverted needed to remain in place in order to not break
r299462.
PR35317
llvm-svn: 318788
The obvious approach of defining a pattern like the one below actually doesn't
work:
`def : Pat<(i32 0), (i32 X0)>;`
As was noted when Lanai made this change (https://reviews.llvm.org/rL288215),
attempting to handle the constant 0 in tablegen leads to assertions due to a
physical register being used where a virtual register is expected.
llvm-svn: 318738
Previous patches primarily ensured that codegen was possible for the standard
RISC-V instructions. However, there are a number of IR inputs that wouldn't be
appropriately lowered. This patch both adds test cases and supports lowering
for a number of these cases:
* Improved sext/zext/trunc support
* Support for setcc variants that don't map directly to RISC-V instructions
* Lowering mul, and hence support for external symbols
* addc, adde, subc, sube
* mulhs, srem, mulhu, urem, udiv, sdiv
* {srl,sra,shl}_parts
* brind
* br_jt
* bswap, ctlz, cttz, ctpop
* rotl, rotr
* BlockAddress operands
Differential Revision: https://reviews.llvm.org/D29938
llvm-svn: 318737
Although ISD::SELECT_CC is a more natural match for RISCVISD::SELECT_CC (and
ultimately the integer RISC-V conditional branch instructions), we choose to
expand ISD::SELECT_CC and lower ISD::SELECT. The appropriate compare+branch
will be created in the case where an ISD::SELECT condition value is created by
an ISD::SETCC node, which operates on XLen types. Other datatypes such as
floating point don't have conditional branch instructions, and lowering
ISD::SELECT allows more flexibility for handling these cases.
Differential Revision: https://reviews.llvm.org/D29937
llvm-svn: 318735
DAGTypeLegalizer::SplitInteger uses default pointer size as shift amount constant type,
which causes less performant ISA in amdgcn---amdgiz target since the default pointer
type is i64 whereas the desired shift amount type is i32.
This patch fixes that by using TLI.getScalarShiftAmountTy in DAGTypeLegalizer::SplitInteger.
The X86 change is necessary since splitting i512 requires shifting amount of 256, which
cannot be held by i8.
Differential Revision: https://reviews.llvm.org/D40148
llvm-svn: 318727
Normally this would be cleaned up by promoting the condition operand next. But in the attached case we promoted the result from v2i48 to v2i64 and the condition from v2i1 to v2i48. Then we tried to "promote" the v2i48 condition back to v2i1 because that's what the SetCC result type for v2i64 is on X86 with VLX. But promote is either a NOP or SIGN_EXTEND and this would need a truncation.
With the change here we now get the SetCC type of v2i1 when we're handling the result promotion and the operand no longer needs to be promoted itself.
Fixes PR35272.
llvm-svn: 318706
Summary:
while investigating performance degradation of imagick benchmark
there were found inefficient pattern for UINT_TO_FP conversion.
That pattern causes RAW hazard in assembly code. Specifically,
uitofp IR operator results in poor assembler :
st %i0, [%fp - 952]
ldd [%fp - 952], %f0
it stores 32-bit integer register into memory location and then
loads 64-bit floating point data from that location.
That is exactly RAW hazard case. To optimize that case it is
possible to use SPISD::ITOF and SPISD::XTOF for conversion from
integer to floating point data type and to use ISD::BITCAST to
copy from integer register into floating point register.
The fix is to write custom UINT_TO_FP pattern using SPISD::ITOF,
SPISD::XTOF, ISD::BITCAST.
Patch by Alexey Lapshin
Reviewers: fedor.sergeev, jyknight, dcederman, lero_chris
Reviewed By: jyknight
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D36875
llvm-svn: 318704
Commit b5cbc7760ab8 ("[bpf] allow direct and indirect calls")
allowed more than one function in the bpf program, and
commit 114353884415 ("bpf: fix a bug in trunc-op optimization")
fixed a bug in trunc-op optimization which only showed up
with more than one function in the bpf program.
This patch added a test case for trunc-op optimization
for bpf programs with two functions. Reverting commit
"bpf: fix a bug in trunc-op optimization" will cause
failure for this test case.
Signed-off-by: Yonghong Song <yhs@fb.com>
llvm-svn: 318695
MachineSink attempts to place instructions near the basic blocks where
they are needed. Once an instruction has been sunk, its location
relative to other instructions is no longer consistent with the
original source code. In order to ensure correct single-stepping and
profiling, the debug location for sunk instructions is either merged
with the insertion point or erased if the target successor block is
empty.
Patch by Matthew Voss!
Differential Revision: https://reviews.llvm.org/D39933
llvm-svn: 318679
The instructions addis,addi, bl are used to calculate the address of TLS thread
local variables. These TLS access code sequences are generated repeatedly every
time the thread local variable is accessed. By communicating to Machine CSE that
X2 is guaranteed to have the same value within the same function call (so called
Caller Preserved Physical Register), the redundant TLS access code sequences are
cleaned up.
Differential Revision: https://reviews.llvm.org/D39173
llvm-svn: 318661
Manually update test r600.amdgpu-alias-analysis.ll for amdgiz environment
since it cannot be done by script.
The two pointers are swapped in the output because PrintResults in
AliasAnalysisEvaluator.cpp sorts the strings obtained from printAsOperand
before printing them.
Differential Revision: https://reviews.llvm.org/D40131
llvm-svn: 318660
Move the hazard scheduling pass to after the long branch pass, as the
long branch pass can create forbiddden slot hazards. Rather than complicating
the implementation of the long branch pass to handle forbidden slot hazards,
just reorder the passes.
llvm-svn: 318657
The VSX versions have the advantage of a full 64-register target whereas the FP
ones have the advantage of lower latency and higher throughput. So what we’re
after is using the faster instructions in low register pressure situations and
using the larger register file in high register pressure situations.
The heuristic chooses between the following 7 pairs of instructions.
PPC::LXSSPX vs PPC::LFSX
PPC::LXSDX vs PPC::LFDX
PPC::STXSSPX vs PPC::STFSX
PPC::STXSDX vs PPC::STFDX
PPC::LXSIWAX vs PPC::LFIWAX
PPC::LXSIWZX vs PPC::LFIWZX
PPC::STXSIWX vs PPC::STFIWX
Differential Revision: https://reviews.llvm.org/D38486
llvm-svn: 318651
Summary:
This patch fixes an issue so that the right alias is printed when the instruction has tied operands. It checks the number of operands in the resulting instruction as opposed to the alias, and then skips over tied operands that should not be printed in the alias.
This allows to generate the preferred assembly syntax for the AArch64 'ins' instruction, which should always be displayed as 'mov' according to the ARM Architecture Reference Manual. Several unit tests have changed as a result, but only to reflect the preferred disassembly. Some other InstAlias patterns (movk/bic/orr) needed a slight adjustment to stop them becoming the default and breaking other unit tests.
Please note that the patch is mostly the same as https://reviews.llvm.org/D29219 which was reverted because of an issue found when running TableGen with the Address Sanitizer. That issue has been addressed in this iteration of the patch.
Reviewers: rengolin, stoklund, huntergr, SjoerdMeijer, rovka
Reviewed By: rengolin, SjoerdMeijer
Subscribers: fhahn, aemerson, javed.absar, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D40030
llvm-svn: 318650
Coby Tayree