optimize zipping over array iterators

[the8472]

Fixes #115339 (somewhat)

the new assembly:

zip_arrays:
        .cfi_startproc
        vmovups (%rdx), %ymm0
        leaq    32(%rsi), %rcx
        vxorps  %xmm1, %xmm1, %xmm1
        vmovups %xmm1, -24(%rsp)
        movq    $0, -8(%rsp)
        movq    %rsi, -88(%rsp)
        movq    %rdi, %rax
        movq    %rcx, -80(%rsp)
        vmovups %ymm0, -72(%rsp)
        movq    $0, -40(%rsp)
        movq    $32, -32(%rsp)
        movq    -24(%rsp), %rcx
        vmovups (%rsi,%rcx), %ymm0
        vorps   -72(%rsp,%rcx), %ymm0, %ymm0
        vmovups %ymm0, (%rsi,%rcx)
        vmovups (%rsi), %ymm0
        vmovups %ymm0, (%rdi)
        vzeroupper
        retq

This is still longer than the slice version given in the issue but at least it eliminates the terrible vpextrb/orb chain. I guess this is due to excessive memcpys again (haven't looked at the llvmir)?

The TrustedLen specialization is a drive-by change since I had to do something for the default impl anyway to be able to specialize the TrustedRandomAccessNoCoerce impl.

[the8472]

@bors try @rust-timer queue

[bors]

⌛ Trying commit 0580b27 with merge 3effcd54b6ab31955a45d85d1ca90746310d294b...

[bors]

☀️ Try build successful - checks-actions
Build commit: 3effcd54b6ab31955a45d85d1ca90746310d294b (3effcd54b6ab31955a45d85d1ca90746310d294b)

[rust-timer]

Finished benchmarking commit (3effcd54b6ab31955a45d85d1ca90746310d294b): comparison URL.

Overall result: ❌✅ regressions and improvements - ACTION NEEDED

Benchmarking this pull request likely means that it is perf-sensitive, so we're automatically marking it as not fit for rolling up. While you can manually mark this PR as fit for rollup, we strongly recommend not doing so since this PR may lead to changes in compiler perf.

Next Steps: If you can justify the regressions found in this try perf run, please indicate this with @rustbot label: +perf-regression-triaged along with sufficient written justification. If you cannot justify the regressions please fix the regressions and do another perf run. If the next run shows neutral or positive results, the label will be automatically removed.

@bors rollup=never
@rustbot label: -S-waiting-on-perf +perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.7%	[0.7%, 0.7%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.7%	[-0.7%, -0.6%]	6
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.5%	[-0.7%, 0.7%]	7

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	2.7%	[2.7%, 2.7%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-4.7%	[-6.6%, -2.8%]	2
Improvements ✅ (secondary)	-3.5%	[-4.7%, -1.4%]	4
All ❌✅ (primary)	-2.2%	[-6.6%, 2.7%]	3

Cycles

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	1.1%	[1.1%, 1.1%]	1
Regressions ❌ (secondary)	2.2%	[2.2%, 2.2%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.1%	[1.1%, 1.1%]	1

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.1%	[0.0%, 0.5%]	8
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.0%	[-0.1%, -0.0%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[-0.1%, 0.5%]	10

Bootstrap: 629.273s -> 628.851s (-0.07%)
Artifact size: 316.72 MiB -> 316.72 MiB (0.00%)

[scottmcm]

Oh, of course, random access to an array iterator makes a ton of sense 👍

[scottmcm]

Hmm, the LLVM IR for the array version in nightly has load <32 x i8>, which makes me wonder whether #111999 is involved here somehow.

Why that would make the codegen so much worse is unclear to me, though. Does the same thing happen for different lengths? Or is it just 32 that goes bad? That happens to be exactly the somewhat-arbitrary threshold in that PR.

Regardless, I think doing this to enable random access for array::IntoIter seems like a good idea, so
@bors r+

[bors]

📌 Commit 0580b27 has been approved by scottmcm

It is now in the queue for this repository.

[bors]

⌛ Testing commit 0580b27 with merge d603d0f...

[bors]

💔 Test failed - checks-actions

[scottmcm]

@bors r-

Hmm, I wonder if the upstream changes mentioned in #115339 (comment) will obviate this at all...

[the8472]

Hmm, I wonder if the upstream changes mentioned in #115339 (comment) will obviate this at all...

I don't know but guess that random access could still be beneficial for more complex combinations of array iterators + zip.

[the8472]

@rustbot ready

I think the issue was that nopt builds disable optimizations for tests too if an optimization level hasn't been set explicitly

[scottmcm]

@bors r+ rollup=never

[bors]

📌 Commit b018ad3 has been approved by scottmcm

It is now in the queue for this repository.

[bors]

⌛ Testing commit b018ad3 with merge 0d410be...

[bors]

☀️ Test successful - checks-actions
Approved by: scottmcm
Pushing 0d410be to master...

[rust-timer]

Finished benchmarking commit (0d410be): comparison URL.

Overall result: ✅ improvements - no action needed

@rustbot label: -perf-regression

Instruction count

This is a highly reliable metric that was used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	-	-	0
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.3%	[-0.4%, -0.2%]	3
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.3%	[-0.4%, -0.2%]	3

Max RSS (memory usage)

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	3.7%	[1.1%, 8.5%]	4
Regressions ❌ (secondary)	4.5%	[4.5%, 4.5%]	1
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-2.8%	[-4.0%, -1.6%]	3
All ❌✅ (primary)	3.7%	[1.1%, 8.5%]	4

Cycles

This benchmark run did not return any relevant results for this metric.

Binary size

Results

This is a less reliable metric that may be of interest but was not used to determine the overall result at the top of this comment.

	mean	range	count
Regressions ❌ (primary)	0.1%	[0.0%, 0.2%]	8
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.1%	[-0.2%, -0.1%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.1%	[-0.2%, 0.2%]	10

Bootstrap: 624.559s -> 624.729s (0.03%)
Artifact size: 305.55 MiB -> 305.51 MiB (-0.01%)