If a loop has a constant trip count known to be a multiple of MaxVF (times user
UF), LV infers that no tail will be generated for any chosen VF. This relies on
the chosen VF's being powers of 2 bound by MaxVF, and assumes MaxVF is a power
of 2. Make sure the latter holds, in particular when MaxVF is set by a memory
dependence distance which may not be a power of 2.
Differential Revision: https://reviews.llvm.org/D80491
Fixes a build issue with libc++ configured with _LIBCPP_RAW_ITERATORS (ADL not effective)
```
llvm/lib/CodeGen/TargetLoweringObjectFileImpl.cpp:1602:3: error: no matching function for call to 'transform'
transform(HexString.begin(), HexString.end(), HexString.begin(), tolower);
^~~~~~~~~
```
Reviewed By: MaskRay
Differential Revision: https://reviews.llvm.org/D80475
If we're extracting an upper subvector from a broadcast we're better off extracting the lowest subvector instead as it avoids an actual extract instruction and might help SimplifyDemandedVectorElts further simplify the code.
EarlyCSE was added with D75145, but the motivating test is
not regressed by removing the extra pass now. That might be
because VectorCombine altered the way it processes instructions,
or it might be from (re)moving VectorCombine in the pipeline.
The extra round of EarlyCSE appears to cost approximately
0.26% in compile-time as discussed in D80236, so we need some
evidence to justify its inclusion here, but we do not have
that (yet).
I suspect that between SLP and VectorCombine, we are creating
patterns that InstCombine and/or codegen are not prepared for,
but we will need to reduce those examples and include them as
PhaseOrdering and/or test-suite benchmarks.
Currently we unconditionally get the first lane of the condition
operand, even if we later use the full vector condition. This can result
in some unnecessary instructions being generated.
Suggested as follow-up in D80219.
This initial version only peeks through cases where we just demand the sign bit of an ashr shift, but we could generalize this further depending on how many sign bits we already have.
The pr18014.ll case is a minor annoyance - we've failed to to move the psrad/paddd after the blendvps which would have avoided the extra move, but we have still increased the ILP.
Replace TargetMachine.h include with forward declaration and CodeGen.h include in AMDGPU.h.
Exposes a couple of implicit dependencies that require additional forward declarations/includes.
VPWidenSelectRecipe already contains a VPUser, but it is not used. This
patch updates the code related to VPWidenSelectRecipe to use VPUser for
its operands.
Reviewers: Ayal, gilr, rengolin
Reviewed By: gilr
Differential Revision: https://reviews.llvm.org/D80219
For the 'inverse shift', we currently always perform a subtraction of the original (masked) shift amount.
But for the case where we are handling power-of-2 type widths, we can replace:
(sub bw-1, (and amt, bw-1) ) -> (and (xor amt, bw-1), bw-1) -> (and ~amt, bw-1)
This allows x86 shifts to fold away the and-mask.
Followup to D77301 + D80466.
http://volta.cs.utah.edu:8080/z/Nod0Gr
Differential Revision: https://reviews.llvm.org/D80489
On X86 (AVX1/AVX2), non-boolean masked loads only demand the sign bit of the mask, we already do the equivalent for masked stores.
Annoyingly I can't easily handle this inside TargetLowering::SimplifyDemandedBits as this is an x86 specific case for a generic node.
Differential Revision: https://reviews.llvm.org/D80478
This adds the family/model returned by CPUID for some Intel
Comet Lake CPUs. Instruction set and tuning wise these are
the same as "skylake".
These are not in the Intel SDM yet, but these should be correct.
This is a preliminary patch before I deal with the xor+and issue raised in D77301.
We get much better code for i8/i16 funnel shifts by concatenating the operands together and performing the shift as a double width type, it avoids repeated use of the shift amount and partial registers.
fshl(x,y,z) -> (((zext(x) << bw) | zext(y)) << (z & (bw-1))) >> bw.
fshr(x,y,z) -> (((zext(x) << bw) | zext(y)) >> (z & (bw-1))) >> bw.
Alive2: http://volta.cs.utah.edu:8080/z/CZx7Cn
This doesn't do as well for i32 cases on x86_64 (the xor+and followup patch is much better) so I haven't bothered with that.
Cases with constant amounts are more dubious as well so I haven't currently bothered with those - its these kind of 'edge' cases that put me off trying to put this in TargetLowering::expandFunnelShift.
Differential Revision: https://reviews.llvm.org/D80466
Although writing to wzr/xzr is correct since we don't care about the result
of the sub, only the flags, doing so causes tail merge blocks to fail.
Writing to an unused virtual register instead allows the optimization to fire,
improving performance significantly on 256.bzip2.
Differential Revision: https://reviews.llvm.org/D80460
This patch introduces a TargetLowering query, isMulhCheaperThanMulShift.
Currently in DAG Combine, it will transform mulhs/mulhu into a
wider multiply and a shift if the wide multiply is legal.
This TLI function is implemented on 64-bit PowerPC, as it is more desirable to
have multiply-high over multiply + shift for words and doublewords. Having
multiply-high can also aid in further transformations that can be done.
Differential Revision: https://reviews.llvm.org/D78271
This patch updates computeConstantRange to optionally take an assumption
cache as argument and use the available assumptions to limit the range
of the result.
Currently this is limited to assumptions that are comparisons.
Reviewers: reames, nikic, spatel, jdoerfert, lebedev.ri
Reviewed By: nikic
Differential Revision: https://reviews.llvm.org/D76193
Disable pruning of unreachable resumes in the DwarfEHPrepare pass
at optnone. While I expect the pruning itself to be essentially free,
this does require a dominator tree calculation, that is not used for
anything else. Saving this DT construction makes for a 0.4% O0
compile-time improvement.
Differential Revision: https://reviews.llvm.org/D80400
Replace with forward declaration and move dependency down to source files that actually need it.
Both TargetLowering.h and TargetMachine.h are 2 of the most expensive headers (top 10) in the ClangBuildAnalyzer report when building llc.
We have to assume undef could be an snan, which would need quieting so
returning qnan is safer than undef. Also consider strictfp, and don't
care if the result rounded.
This change does not affect the produced binary.
In this patch I assign a technical suffix to each section/fill
(i.e. chunk) name when it is empty. It allows to simplify the code
slightly and improve error messages reported.
In the code we have the section to index mapping, SN2I, which is
globally used. With this change we can use it to map "empty"
names to indexes now, what is helpful.
Differential revision: https://reviews.llvm.org/D79984
Summary:
Added a new IRCanonicalizer pass which aims to transform LLVM modules into
a canonical form by reordering and renaming instructions while preserving the
same semantics. The canonicalizer makes it easier to spot semantic differences
when diffing two modules which have undergone different passes.
Presentation: https://www.youtube.com/watch?v=c9WMijSOEUg
Reviewed by: plotfi
Differential Revision: https://reviews.llvm.org/D66029
When performing codegen at optnone, don't add alias analysis to
the pipeline. We don't need it, but it causes an unnecessary
dominator tree calculation.
I've also moved the module verifier call to the top so that a bunch
of disabled-at-optnone passes group more nicely.
Differential Revision: https://reviews.llvm.org/D80378
If the caller needs to reponsible for making sure the MaybeAlign
has a value, then we should just make the caller convert it to an Align
with operator*.
I explicitly deleted the relational comparison operators that
were being inherited from Optional. It's unclear what the meaning
of two MaybeAligns were one is defined and the other isn't
should be. So make the caller reponsible for defining the behavior.
I left the ==/!= operators from Optional. But now that exposed a
weird quirk that ==/!= between Align and MaybeAlign required the
MaybeAlign to be defined. But now we use the operator== from
Optional that takes an Optional and the Value.
Differential Revision: https://reviews.llvm.org/D80455
Negations are incorrectly added in numerous places and the code just happens to work.
Also fix a missed DW_CFA_def_cfa_offset negation in c693b9c321d5a40d012340619674cf790c9ac86c:
ARMAsmBackendDarwin::generateCompactUnwindEncoding
As noted in D80236, moving the pass in the pipeline exposed this
shortcoming. Extra work to recalculate the alias results showed
up as a compile-time slowdown.
Ayal Zaks