| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Modify GenerateConstantOffsetsImpl to create offsets that can be used
by indexed addressing modes. If formulae can be generated which
result in the constant offset being the same size as the recurrence,
we can generate a pre-indexed access. This allows the pointer to be
updated via the single pre-indexed access so that (hopefully) no
add/subs are required to update it for the next iteration. For small
cores, this can significantly improve performance DSP-like loops.
Differential Revision: https://reviews.llvm.org/D55373
llvm-svn: 353403
|
| |
|
|
|
|
|
|
|
|
| |
We should canonicalize to one of these forms,
and compare-with-zero could be more conducive
to follow-on transforms. This also leads to
generally better codegen as shown in PR40611:
https://bugs.llvm.org/show_bug.cgi?id=40611
llvm-svn: 353313
|
| |
|
|
|
|
| |
Lots of unrelated diffs here from the newer version of the script.
llvm-svn: 353312
|
| |
|
|
| |
llvm-svn: 353276
|
| |
|
|
|
|
|
|
|
| |
Resumes that are not reachable from a cleanup landing pad are considered
to be unreachable. It’s not safe to split them out.
rdar://47808235
llvm-svn: 353242
|
| |
|
|
|
|
|
|
|
|
| |
As discussed in D53037, this can lead to worse codegen, and we
don't generally expect the backend to be able to optimize
arbitrary shuffles. If there's only one use of the 1st shuffle,
that means it's getting removed, so that should always be
safe.
llvm-svn: 353235
|
| |
|
|
|
|
|
| |
As discussed in D53037, this transform can cause codegen problems
if the 1st shuffle has multiple uses.
llvm-svn: 353233
|
| |
|
|
| |
llvm-svn: 353179
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch improves code generation for some AArch64 ACLE intrinsics. It adds
support to CGP to duplicate and sink operands to their user, if they can be
folded into a target instruction, like zexts and sub into usubl. It adds a
TargetLowering hook shouldSinkOperands, which looks at the operands of
instructions to see if sinking is profitable.
I decided to add a new target hook, as for the sinking to be profitable,
at least on AArch64, we have to look at multiple operands of an
instruction, instead of looking at the users of a zext for example.
The sinking is done in CGP, because it works around an instruction
selection limitation. If instruction selection is not limited to a
single basic block, this patch should not be needed any longer.
Alternatively this could be done in the LoopSink pass, which tries to
undo LICM for instructions in blocks that are not executed frequently.
Note that we do not force the operands to sink to have a single user,
because we duplicate them before sinking. Therefore this is only
desirable if they really can be done for free. Additionally we could
consider the impact on live ranges later on.
This should fix https://bugs.llvm.org/show_bug.cgi?id=40025.
As for performance, we have internal code that uses intrinsics and can
be speed up by 10% by this change.
Reviewers: SjoerdMeijer, t.p.northover, samparker, efriedma, RKSimon, spatel
Reviewed By: samparker
Differential Revision: https://reviews.llvm.org/D57377
llvm-svn: 353152
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
When LSR first adds SCEVs to BaseRegs, it only does it if `isZero()` has
returned false. In the end, in invocation of `InsertFormula`, it asserts that
all values there are still not zero constants. However between these two
points, it makes some transformations, in particular extends them to wider
type.
SCEV does not give us guarantee that if `S` is not a constant zero, then
`sext(S)` is also not a constant zero. It might have missed some optimizing
transforms when it was calculating `S` and then made them when it took `sext`.
For example, it may happen if previously optimizing transforms were limited
by depth or somehow else.
This patch adds a bailout when we may end up with a zero SCEV after extension.
Differential Revision: https://reviews.llvm.org/D57565
Reviewed By: samparker
llvm-svn: 353136
|
| |
|
|
|
|
| |
This reverts accidental commit ff5527718d5d3b9966f6e8948866c0dc15ffcf3c.
llvm-svn: 353118
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
It seems that the run time for Windows has changed and supports more math
functions than before. Since LLVM requires at least VS2015, I assume that
this is the run time that would be redistributed with programs built with
Clang. Thus, I based this update on the header file `math.h` that
accompanies it.
This patch addresses the PR40541. Unfortunately, I have no access to a
Windows development environment to validate it.
llvm-svn: 353114
|
| |
|
|
|
|
| |
Inverted operand 0 and operand 1.
llvm-svn: 353106
|
| |
|
|
| |
llvm-svn: 353103
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
The fix added in r352904 is not quite correct, or rather misleading:
1. When the texfailctrl (TFC) argument was non-constant, the fix assumed
non-TFE/LWE, which is incorrect.
2. Regardless, this code path cannot even be hit for correct
TFE/LWE-enabled calls, because those return a struct. Added
a test case for those for completeness.
Change-Id: I92d314dbc67a2670f6d7adaab765ef45f56a49cf
Reviewers: hliao, dstuttard, arsenm
Subscribers: kzhuravl, jvesely, wdng, yaxunl, tpr, t-tye, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D57681
llvm-svn: 353097
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
LoopVectorize adds llvm.loop.isvectorized, but leaves
llvm.loop.vectorize.enable. Do not consider such a loop for user-forced
vectorization since vectorization already happened -- by prioritizing
llvm.loop.isvectorized except for TM_SuppressedByUser.
Fixes http://llvm.org/PR40546
Differential Revision: https://reviews.llvm.org/D57542
llvm-svn: 353082
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is no-functional-change-intended although there could
be intermediate variations caused by a difference in the
debug info produced by setting that from the builder's
insertion point.
I'm updating the IR test file associated with this code just
to show that the naming differences from using the builder
are visible.
The motivation for adding a helper function is that we are
likely to extend this code to deal with other overflow ops.
llvm-svn: 353056
|
| |
|
|
| |
llvm-svn: 353037
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary: PR40564
Reviewers: aprantl, rnk
Subscribers: llvm-commits, hiraditya
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D57629
llvm-svn: 353007
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
There are 2 changes visible here:
1. There's no reason to limit this transform based on number
of condition registers. That diff allows PPC to produce
slightly better (dot-instructions should be generally good)
code.
Note: someone that cares about PPC codegen might want to
look closer at that output because it seems like we could
still improve this.
2. We (probably?) should not bother trying to form uaddo (or
other overflow ops) when there's no target support for such
an op. This goes beyond checking whether the op is expanded
because both PPC and AArch64 show better codegen for standard
types regardless of whether the op is legal/custom.
llvm-svn: 353001
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is the most important uaddo problem mentioned in PR31754:
https://bugs.llvm.org/show_bug.cgi?id=31754
...but that was overcome in x86 codegen with D57637.
That patch also corrects the inc vs. add regressions seen with the previous attempt at this.
Still, we want to make this matcher complete, so we can potentially canonicalize the pattern
even if it's an 'add 1' operation.
Pattern matching, however, shouldn't assume that we have canonicalized IR, so we match 4
commuted variants of uaddo.
There's also a test with a crazy type to show that the existing CGP transform based on this
matcher is not limited by target legality checks.
I'm not sure if the Hexagon diff means the test is no longer testing what it intended to
test, but that should be solvable in a follow-up.
Differential Revision: https://reviews.llvm.org/D57516
llvm-svn: 352998
|
| |
|
|
|
|
|
|
| |
The test specifiies the triple, so it needs to be in the
x86 directory in case a bot has been configured without
the x86 target.
llvm-svn: 352992
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
x, log2(pow(2.0,x)) == x
Summary: This patch enables folding following instructions under -ffast-math flag: log10(pow(10.0,x)) -> x, log2(pow(2.0,x)) -> x
Reviewers: hfinkel, spatel, efriedma, craig.topper, zvi, majnemer, lebedev.ri
Reviewed By: spatel, lebedev.ri
Subscribers: lebedev.ri, llvm-commits
Differential Revision: https://reviews.llvm.org/D41940
llvm-svn: 352981
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If there is only a single available value, all uses must be dominated by
the single value and there is no need to search for a reaching
definition.
This drastically speeds up LCSSA in some cases. For the test case
from PR37202, it speeds up LCSSA construction by 4 times.
Time-passes without this patch for test case from PR37202:
Total Execution Time: 29.9285 seconds (29.9276 wall clock)
---User Time--- --System Time-- --User+System-- ---Wall Time--- --- Name ---
5.2786 ( 17.7%) 0.0021 ( 1.2%) 5.2806 ( 17.6%) 5.2808 ( 17.6%) Unswitch loops
4.3739 ( 14.7%) 0.0303 ( 18.1%) 4.4042 ( 14.7%) 4.4042 ( 14.7%) Loop-Closed SSA Form Pass
4.2658 ( 14.3%) 0.0192 ( 11.5%) 4.2850 ( 14.3%) 4.2851 ( 14.3%) Loop-Closed SSA Form Pass #2
2.2307 ( 7.5%) 0.0013 ( 0.8%) 2.2320 ( 7.5%) 2.2318 ( 7.5%) Loop Invariant Code Motion
2.0888 ( 7.0%) 0.0012 ( 0.7%) 2.0900 ( 7.0%) 2.0897 ( 7.0%) Unroll loops
1.6761 ( 5.6%) 0.0013 ( 0.8%) 1.6774 ( 5.6%) 1.6774 ( 5.6%) Value Propagation
1.3686 ( 4.6%) 0.0029 ( 1.8%) 1.3716 ( 4.6%) 1.3714 ( 4.6%) Induction Variable Simplification
1.1457 ( 3.8%) 0.0010 ( 0.6%) 1.1468 ( 3.8%) 1.1468 ( 3.8%) Loop-Closed SSA Form Pass #4
1.1384 ( 3.8%) 0.0005 ( 0.3%) 1.1389 ( 3.8%) 1.1389 ( 3.8%) Loop-Closed SSA Form Pass #6
1.1360 ( 3.8%) 0.0027 ( 1.6%) 1.1387 ( 3.8%) 1.1387 ( 3.8%) Loop-Closed SSA Form Pass #5
1.1331 ( 3.8%) 0.0010 ( 0.6%) 1.1341 ( 3.8%) 1.1340 ( 3.8%) Loop-Closed SSA Form Pass #3
Time passes with this patch
Total Execution Time: 19.2802 seconds (19.2813 wall clock)
---User Time--- --System Time-- --User+System-- ---Wall Time--- --- Name ---
4.4234 ( 23.2%) 0.0038 ( 2.0%) 4.4272 ( 23.0%) 4.4273 ( 23.0%) Unswitch loops
2.3828 ( 12.5%) 0.0020 ( 1.1%) 2.3848 ( 12.4%) 2.3847 ( 12.4%) Unroll loops
1.8714 ( 9.8%) 0.0020 ( 1.1%) 1.8734 ( 9.7%) 1.8735 ( 9.7%) Loop Invariant Code Motion
1.7973 ( 9.4%) 0.0022 ( 1.2%) 1.7995 ( 9.3%) 1.8003 ( 9.3%) Value Propagation
1.4010 ( 7.3%) 0.0033 ( 1.8%) 1.4043 ( 7.3%) 1.4044 ( 7.3%) Induction Variable Simplification
0.9978 ( 5.2%) 0.0244 ( 13.1%) 1.0222 ( 5.3%) 1.0224 ( 5.3%) Loop-Closed SSA Form Pass #2
0.9611 ( 5.0%) 0.0257 ( 13.8%) 0.9868 ( 5.1%) 0.9868 ( 5.1%) Loop-Closed SSA Form Pass
0.5856 ( 3.1%) 0.0015 ( 0.8%) 0.5871 ( 3.0%) 0.5869 ( 3.0%) Unroll loops #2
0.4132 ( 2.2%) 0.0012 ( 0.7%) 0.4145 ( 2.1%) 0.4143 ( 2.1%) Loop Invariant Code Motion #3
Reviewers: efriedma, davide, mzolotukhin
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D57033
llvm-svn: 352960
|
| |
|
|
| |
llvm-svn: 352935
|
| |
|
|
| |
llvm-svn: 352932
|
| |
|
|
|
|
| |
Run checks for Win32 as well.
llvm-svn: 352917
|
| |
|
|
|
|
| |
Run checks for Win64 as well.
llvm-svn: 352908
|
| |
|
|
|
|
|
|
| |
- If that operand is not ConstantInt, skip enabling TFE/LWE.
Differential Revision: https://reviews.llvm.org/D57539
llvm-svn: 352904
|
| |
|
|
| |
llvm-svn: 352895
|
| |
|
|
|
|
| |
Add checks for Win64 to existing cases.
llvm-svn: 352892
|
| |
|
|
| |
llvm-svn: 352886
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
If we can reduce the x86-specific intrinsic to the generic op, it allows existing
simplifications and value tracking folds. AFAICT, this always results in identical
x86 codegen in the non-reduced case...which should be true because we semi-generically
(too aggressively IMO) convert to llvm.uadd.with.overflow in CGP, so the DAG/isel must
already combine/lower this intrinsic as expected.
This isn't quite what was requested in:
https://bugs.llvm.org/show_bug.cgi?id=40486
...but we want to have these kinds of folds early for efficiency and to enable greater
simplifications. For the case in the bug report where we have:
_addcarry_u64(0, ahi, 0, &ahi)
...this gets completely simplified away in IR.
Differential Revision: https://reviews.llvm.org/D57453
llvm-svn: 352870
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
InlineCost's isInlineViable() is changed to return InlineResult
instead of bool. This provides messages for failure reasons and
allows to get more specific messages for cases where callsites
are not viable for inlining.
Reviewed By: xbolva00, anemet
Differential Revision: https://reviews.llvm.org/D57089
llvm-svn: 352849
|
| |
|
|
|
|
|
|
| |
This ensures that if we make it to the backend w/o lowering widenable_conditions first, that we generate correct code. Doing it in CGP - instead of isel - let's us fold control flow before hitting block local instruction selection.
Differential Revision: https://reviews.llvm.org/D57473
llvm-svn: 352779
|
| |
|
|
|
|
|
|
| |
leader"
Recommit of r352763 with fix for use after free.
llvm-svn: 352770
|
| |
|
|
|
|
|
|
| |
increment-by-one
Missed some regression test updates when testing this.
llvm-svn: 352769
|
| |
|
|
|
|
|
|
|
|
|
|
| |
This reverts commit r352763.
Causing a couple bot failures, root cause pointed to by sanitizer bot:
http://lab.llvm.org:8011/builders/sanitizer-x86_64-linux-fast/builds/28909/steps/annotate/logs/stdio
Use after free. I understand the issue but will revert and test with fix
before recommitting.
llvm-svn: 352768
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is the most important uaddo problem mentioned in PR31754:
https://bugs.llvm.org/show_bug.cgi?id=31754
We were failing to match the canonicalized pattern when it's an 'add 1' operation.
Pattern matching, however, shouldn't assume that we have canonicalized IR, so we
match 4 commuted variants of uaddo.
There's also a test with a crazy type to show that the existing CGP transform
based on this matcher is not limited by target legality checks, but that's a
different problem.
Differential Revision: https://reviews.llvm.org/D57516
llvm-svn: 352766
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Summary:
COFF requires that COMDAT name match that of the leader. When we promote
and rename an internal leader in ThinLTO due to an import, ensure we
subsequently rename the associated COMDAT. Similar to D31963 which did
this during ThinLTO module splitting.
Fixes PR40414.
Reviewers: pcc, inglorion
Subscribers: mehdi_amini, dexonsmith, dmajor, llvm-commits
Differential Revision: https://reviews.llvm.org/D57395
llvm-svn: 352763
|
| |
|
|
| |
llvm-svn: 352762
|
| |
|
|
|
|
|
|
|
|
|
| |
Introduces a pass that provides default lowering strategy for the
`experimental.widenable.condition` intrinsic, replacing all its uses with
`i1 true`.
Differential Revision: https://reviews.llvm.org/D56096
Reviewed By: reames
llvm-svn: 352739
|
| |
|
|
| |
llvm-svn: 352734
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
functions
Summary: This patch enables folding following expressions under -ffast-math flag: exp(X) * exp(Y) -> exp(X + Y), exp2(X) * exp2(Y) -> exp2(X + Y). Motivation: https://bugs.llvm.org/show_bug.cgi?id=35594
Reviewers: hfinkel, spatel, efriedma, lebedev.ri
Reviewed By: spatel, lebedev.ri
Subscribers: lebedev.ri, llvm-commits
Differential Revision: https://reviews.llvm.org/D41342
llvm-svn: 352730
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Currently SCEV attempts to limit transformations so that they do not work with
big SCEVs (that may take almost infinite compile time). But for this, it uses heuristics
such as recursion depth and number of operands, which do not give us a guarantee
that we don't actually have big SCEVs. This situation is still possible, though it is not
likely to happen. However, the bug PR33494 showed a bunch of simple corner case
tests where we still produce huge SCEVs, even not reaching big recursion depth etc.
This patch introduces a concept of 'huge' SCEVs. A SCEV is huge if its expression
size (intoduced in D35989) exceeds some threshold value. We prohibit optimizing
transformations if any of SCEVs we are dealing with is huge. This gives us a reliable
check that we don't spend too much time working with them.
As the next step, we can possibly get rid of old limiting mechanisms, such as recursion
depth thresholds.
Differential Revision: https://reviews.llvm.org/D35990
Reviewed By: reames
llvm-svn: 352728
|
| |
|
|
|
|
|
|
| |
This change reverts r351626.
The changes in r351626 cause quadratic work in several cases. (See r351626 thread on llvm-commits for details.)
llvm-svn: 352722
|
| |
|
|
|
|
| |
Added the checks to the existing cases when the target is Win64.
llvm-svn: 352714
|
| |
|
|
| |
llvm-svn: 352707
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
This is meant to be used with clang's __builtin_dynamic_object_size.
When 'true' is passed to this parameter, the intrinsic has the
potential to be folded into instructions that will be evaluated
at run time. When 'false', the objectsize intrinsic behaviour is
unchanged.
rdar://32212419
Differential revision: https://reviews.llvm.org/D56761
llvm-svn: 352664
|
| |
|
|
| |
llvm-svn: 352662
|