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
Add a comment when there is a shifted value,
add x9, x0, #291, lsl #12 ; =1191936
but not when the immediate value is unshifted,
subs x9, x0, #256 ; =256
when the comment adds nothing additional to the reader.
Differential Revision: https://reviews.llvm.org/D107196
Based ontop of D104598, which is a NFCI-ish refactoring.
Here, a restriction, that only empty blocks can be merged, is lifted.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D104597
The motivation here is that MachineBlockPlacement relies on analyzeBranch to remove branches to fallthrough blocks when the branch is not fully analyzeable. With the introduction of the FAULTING_OP psuedo for implicit null checking (see D87861), this case becomes important. Note that it's hard to otherwise exercise this path as BranchFolding handle's any fully analyzeable branch sequence without using this interface.
p.s. For anyone who saw my comment in the original review, what I thought was an issue in BranchFolding originally turned out to simply be a bug in my patch. (Now fixed.)
Differential Revision: https://reviews.llvm.org/D88035
This change enables the generic implicit null transformation for the AArch64 target. As background for those unfamiliar with our implicit null check support:
An implicit null check is the use of a signal handler to catch and redirect to a handler a null pointer. Specifically, it's replacing an explicit conditional branch with such a redirect. This is only done for very cold branches under frontend control w/appropriate metadata.
FAULTING_OP is used to wrap the faulting instruction. It is modelled as being a conditional branch to reflect the fact it can transfer control in the CFG.
FAULTING_OP does not need to be an analyzable branch to achieve it's purpose. (Or at least, that's the x86 model. I find this slightly questionable.)
When lowering to MC, we convert the FAULTING_OP back into the actual instruction, record the labels, and lower the original instruction.
As can be seen in the test changes, currently the AArch64 backend does not eliminate the unconditional branch to the fallthrough block. I've tried two approaches, neither of which worked. I plan to return to this in a separate change set once I've wrapped my head around the interactions a bit better. (X86 handles this via AllowModify on analyzeBranch, but adding the obvious code causing BranchFolding to crash. I haven't yet figured out if it's a latent bug in BranchFolding, or something I'm doing wrong.)
Differential Revision: https://reviews.llvm.org/D87851
