We would like to start pushing -mcpu=generic towards enabling the set of
features that improves performance for some CPUs, without hurting any
others. A blend of the performance options hopefully beneficial to all
CPUs. The largest part of that is enabling in-order scheduling using the
Cortex-A55 schedule model. This is similar to the Arm backend change
from eecb353d0e25ba which made -mcpu=generic perform in-order scheduling
using the cortex-a8 schedule model.
The idea is that in-order cpu's require the most help in instruction
scheduling, whereas out-of-order cpus can for the most part out-of-order
schedule around different codegen. Our benchmarking suggests that
hypothesis holds. When running on an in-order core this improved
performance by 3.8% geomean on a set of DSP workloads, 2% geomean on
some other embedded benchmark and between 1% and 1.8% on a set of
singlecore and multicore workloads, all running on a Cortex-A55 cluster.
On an out-of-order cpu the results are a lot more noisy but show flat
performance or an improvement. On the set of DSP and embedded
benchmarks, run on a Cortex-A78 there was a very noisy 1% speed
improvement. Using the most detailed results I could find, SPEC2006 runs
on a Neoverse N1 show a small increase in instruction count (+0.127%),
but a decrease in cycle counts (-0.155%, on average). The instruction
count is very low noise, the cycle count is more noisy with a 0.15%
decrease not being significant. SPEC2k17 shows a small decrease (-0.2%)
in instruction count leading to a -0.296% decrease in cycle count. These
results are within noise margins but tend to show a small improvement in
general.
When specifying an Apple target, clang will set "-target-cpu apple-a7"
on the command line, so should not be affected by this change when
running from clang. This also doesn't enable more runtime unrolling like
-mcpu=cortex-a55 does, only changing the schedule used.
A lot of existing tests have updated. This is a summary of the important
differences:
- Most changes are the same instructions in a different order.
- Sometimes this leads to very minor inefficiencies, such as requiring
an extra mov to move variables into r0/v0 for the return value of a test
function.
- misched-fusion.ll was no longer fusing the pairs of instructions it
should, as per D110561. I've changed the schedule used in the test
for now.
- neon-mla-mls.ll now uses "mul; sub" as opposed to "neg; mla" due to
the different latencies. This seems fine to me.
- Some SVE tests do not always remove movprfx where they did before due
to different register allocation giving different destructive forms.
- The tests argument-blocks-array-of-struct.ll and arm64-windows-calls.ll
produce two LDR where they previously produced an LDP due to
store-pair-suppress kicking in.
- arm64-ldp.ll and arm64-neon-copy.ll are missing pre/postinc on LPD.
- Some tests such as arm64-neon-mul-div.ll and
ragreedy-local-interval-cost.ll have more, less or just different
spilling.
- In aarch64_generated_funcs.ll.generated.expected one part of the
function is no longer outlined. Interestingly if I switch this to use
any other scheduled even less is outlined.
Some of these are expected to happen, such as differences in outlining
or register spilling. There will be places where these result in worse
codegen, places where they are better, with the SPEC instruction counts
suggesting it is not a decrease overall, on average.
Differential Revision: https://reviews.llvm.org/D110830
The motivating case for this is a long way from here:
https://bugs.llvm.org/show_bug.cgi?id=43146
...but I think this is where we have to start.
We need to canonicalize/optimize sequences of shift and logic to ease
pattern matching for things like bswap and improve perf in general.
But without the artificial limit of '!LegalTypes' (early combining),
there are a lot of test diffs, and not all are good.
In the minimal tests added for this proposal, x86 should have better
throughput in all cases. AArch64 is neutral for scalar tests because
it can fold shifts into bitwise logic ops.
There are 3 shift opcodes and 3 logic opcodes for a total of 9 possible patterns:
https://rise4fun.com/Alive/VlIhttps://rise4fun.com/Alive/n1mhttps://rise4fun.com/Alive/1Vn
Differential Revision: https://reviews.llvm.org/D67021
llvm-svn: 370617
If anything called the recursive isKnownNeverNaN/computeKnownBits/ComputeNumSignBits/SimplifyDemandedBits/SimplifyMultipleUseDemandedBits with an incorrect depth then we could continue to recurse if we'd already exceeded the depth limit.
This replaces the limit check (Depth == 6) with a (Depth >= 6) to make sure that we don't circumvent it.
This causes a couple of regressions as a mixture of calls (SimplifyMultipleUseDemandedBits + combineX86ShufflesRecursively) were calling with depths that were already over the limit. I've fixed SimplifyMultipleUseDemandedBits to not do this. combineX86ShufflesRecursively is trickier as we get a lot of regressions if we reduce its own limit from 8 to 6 (it also starts at Depth == 1 instead of Depth == 0 like the others....) - I'll see what I can do in future patches.
llvm-svn: 367171
This patch introduces the DAG version of SimplifyMultipleUseDemandedBits, which attempts to peek through ops (mainly and/or/xor so far) that don't contribute to the demandedbits/elts of a node - which means we can do this even in cases where we have multiple uses of an op, which normally requires us to demanded all bits/elts. The intention is to remove a similar instruction - SelectionDAG::GetDemandedBits - once SimplifyMultipleUseDemandedBits has matured.
The InstCombine version of SimplifyMultipleUseDemandedBits can constant fold which I haven't added here yet, and so far I've only wired this up to some basic binops (and/or/xor/add/sub/mul) to demonstrate its use.
We do see a couple of regressions that need to be addressed:
AMDGPU unsigned dot product codegen retains an AND mask (for ZERO_EXTEND) that it previously removed (but otherwise the dotproduct codegen is a lot better).
X86/AVX2 has poor handling of vector ANY_EXTEND/ANY_EXTEND_VECTOR_INREG - it prematurely gets converted to ZERO_EXTEND_VECTOR_INREG.
The code owners have confirmed its ok for these cases to fixed up in future patches.
Differential Revision: https://reviews.llvm.org/D63281
llvm-svn: 366799
ARM seems to prefer that long literals be formed from their little end in
order to promote the fusion of the instrs pairs MOV/MOVK and MOVK/MOVK on
Cortex A57 and others (v. "Cortex A57 Software Optimisation Guide", section
4.14).
Differential revision: https://reviews.llvm.org/D28697
llvm-svn: 292422
The backend has been around for years, it's pretty ridiculous that we can't
even use the preferred form for printing "MOV" aliases. Unfortunately, TableGen
can't handle the complex predicates when printing so it's a bunch of nasty C++.
Oh well.
llvm-svn: 272865
We were assuming all SBFX-like operations would have the shl/asr form, but often
when the field being extracted is an i8 or i16, we end up with a
SIGN_EXTEND_INREG acting on a shift instead.
This is a port of r213754 from ARM to AArch64.
llvm-svn: 271677
If and only if the value being inserted sets only known zero bits.
This combine transforms things like
and w8, w0, #0xfffffff0
movz w9, #5
orr w0, w8, w9
into
movz w8, #5
bfxil w0, w8, #0, #4
The combine is tuned to make sure we always reduce the number of instructions.
We avoid churning code for what is expected to be performance neutral changes
(e.g., converted AND+OR to OR+BFI).
Differential Revision: http://reviews.llvm.org/D20387
llvm-svn: 270846
Mask0Imm and ~Mask1Imm must be equivalent and one of the MaskImms is a shifted
mask (e.g., 0x000ffff0). Both 'and's must have a single use.
This changes code like:
and w8, w0, #0xffff000f
and w9, w1, #0x0000fff0
orr w0, w9, w8
into
lsr w8, w1, #4
bfi w0, w8, #4, #12
llvm-svn: 270063
For narrow stores (e.g., strb, srth) we know the upper bits of the register are
unused/not useful. In some cases we can use this information to eliminate
unnecessary instructions.
For example, without this patch we generate (from the 2nd test case):
ldr w8, [x0]
and w8, w8, #0xfff0
bfxil w8, w2, #16, #4
strh w8, [x1]
and after the patch the 'and' is removed:
ldr w8, [x0]
bfxil w8, w2, #16, #4
strh w8, [x1]
ret
During the lowering of the bitfield insert instruction the 'and' is eliminated
because we know the upper 16-bits that are masked off are unused and the lower
4-bits that are masked off are overwritten by the insert itself. Therefore, the
'and' is unnecessary.
Differential Revision: http://reviews.llvm.org/D20175
llvm-svn: 269226
Summary:
When not useful bits, BitWidth becomes 0 and APInt will not be happy.
See https://llvm.org/bugs/show_bug.cgi?id=25571
We can just mark the operand as IMPLICIT_DEF is none bits of it is used.
Reviewers: t.p.northover, jmolloy
Subscribers: gberry, jmolloy, mgrang, aemerson, llvm-commits, rengolin
Differential Revision: http://reviews.llvm.org/D14803
llvm-svn: 254440
Summary:
For bitfield insert OR matching, check both operands for larger pattern
first before checking for smaller pattern.
Add pattern for unsigned bitfield insert-in-zero done with SHL+AND.
Resolves PR21631.
Reviewers: jmolloy, t.p.northover
Subscribers: aemerson, rengolin, llvm-commits, mcrosier
Differential Revision: http://reviews.llvm.org/D12908
llvm-svn: 248006
Essentially the same as the GEP change in r230786.
A similar migration script can be used to update test cases, though a few more
test case improvements/changes were required this time around: (r229269-r229278)
import fileinput
import sys
import re
pat = re.compile(r"((?:=|:|^)\s*load (?:atomic )?(?:volatile )?(.*?))(| addrspace\(\d+\) *)\*($| *(?:%|@|null|undef|blockaddress|getelementptr|addrspacecast|bitcast|inttoptr|\[\[[a-zA-Z]|\{\{).*$)")
for line in sys.stdin:
sys.stdout.write(re.sub(pat, r"\1, \2\3*\4", line))
Reviewers: rafael, dexonsmith, grosser
Differential Revision: http://reviews.llvm.org/D7649
llvm-svn: 230794
This commit starts with a "git mv ARM64 AArch64" and continues out
from there, renaming the C++ classes, intrinsics, and other
target-local objects for consistency.
"ARM64" test directories are also moved, and tests that began their
life in ARM64 use an arm64 triple, those from AArch64 use an aarch64
triple. Both should be equivalent though.
This finishes the AArch64 merge, and everyone should feel free to
continue committing as normal now.
llvm-svn: 209577
I'm doing this in two phases for a better "git blame" record. This
commit removes the previous AArch64 backend and redirects all
functionality to ARM64. It also deduplicates test-lines and removes
orphaned AArch64 tests.
The next step will be "git mv ARM64 AArch64" and rewire most of the
tests.
Hopefully LLVM is still functional, though it would be even better if
no-one ever had to care because the rename happens straight
afterwards.
llvm-svn: 209576
The canonical form of the BFM instruction is always one of the more explicit
extract or insert operations, which makes reading output much easier.
llvm-svn: 207752
ARM64 was not producing pure BFI instructions for bitfield insertion
operations, unlike AArch64. The approach had to be a little different (in
ISelDAGToDAG rather than ISelLowering), and the outcomes aren't identical but
hopefully this gives it similar power.
This should address PR19424.
llvm-svn: 207102
This update was done with the following bash script:
find test/CodeGen -name "*.ll" | \
while read NAME; do
echo "$NAME"
if ! grep -q "^; *RUN: *llc.*debug" $NAME; then
TEMP=`mktemp -t temp`
cp $NAME $TEMP
sed -n "s/^define [^@]*@\([A-Za-z0-9_]*\)(.*$/\1/p" < $NAME | \
while read FUNC; do
sed -i '' "s/;\(.*\)\([A-Za-z0-9_-]*\):\( *\)$FUNC: *\$/;\1\2-LABEL:\3$FUNC:/g" $TEMP
done
sed -i '' "s/;\(.*\)-LABEL-LABEL:/;\1-LABEL:/" $TEMP
sed -i '' "s/;\(.*\)-NEXT-LABEL:/;\1-NEXT:/" $TEMP
sed -i '' "s/;\(.*\)-NOT-LABEL:/;\1-NOT:/" $TEMP
sed -i '' "s/;\(.*\)-DAG-LABEL:/;\1-DAG:/" $TEMP
mv $TEMP $NAME
fi
done
llvm-svn: 186280
This was done with the following sed invocation to catch label lines demarking function boundaries:
sed -i '' "s/^;\( *\)\([A-Z0-9_]*\):\( *\)test\([A-Za-z0-9_-]*\):\( *\)$/;\1\2-LABEL:\3test\4:\5/g" test/CodeGen/*/*.ll
which was written conservatively to avoid false positives rather than false negatives. I scanned through all the changes and everything looks correct.
llvm-svn: 186258
Only Linux is supported at the moment, and other platforms quickly fault. As a
result these tests would fail on non-Linux hosts. It may be worth making the
tests more generic again as more platforms are supported.
llvm-svn: 174170
This patch adds support for AArch64 (ARM's 64-bit architecture) to
LLVM in the "experimental" category. Currently, it won't be built
unless requested explicitly.
This initial commit should have support for:
+ Assembly of all scalar (i.e. non-NEON, non-Crypto) instructions
(except the late addition CRC instructions).
+ CodeGen features required for C++03 and C99.
+ Compilation for the "small" memory model: code+static data <
4GB.
+ Absolute and position-independent code.
+ GNU-style (i.e. "__thread") TLS.
+ Debugging information.
The principal omission, currently, is performance tuning.
This patch excludes the NEON support also reviewed due to an outbreak of
batshit insanity in our legal department. That will be committed soon bringing
the changes to precisely what has been approved.
Further reviews would be gratefully received.
llvm-svn: 174054
