Use a larger length/offset for rmeta tables

[saethlin]

A potential fix for #112934
#113542 is a much better implementation

r? @ghost

[saethlin]

@bors try @rust-timer queue

[bors]

⌛ Trying commit eb908cebfdcc3098b2107cb7cbf4b9168232a52f with merge b75185b7c3847331e3a074a6d6ac4fb8d14e7835...

[bors]

☀️ Try build successful - checks-actions
Build commit: b75185b7c3847331e3a074a6d6ac4fb8d14e7835 (b75185b7c3847331e3a074a6d6ac4fb8d14e7835)

[rust-timer]

Finished benchmarking commit (b75185b7c3847331e3a074a6d6ac4fb8d14e7835): comparison URL.

Overall result: ✅ improvements - no 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.

@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
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.5%	[-1.5%, -1.5%]	1
All ❌✅ (primary)	-	-	0

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)	3.6%	[1.2%, 6.9%]	189
Regressions ❌ (secondary)	3.5%	[0.9%, 5.7%]	115
Improvements ✅ (primary)	-0.5%	[-0.5%, -0.5%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.5%	[-0.5%, 6.9%]	190

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)	0.5%	[0.5%, 0.5%]	1
Regressions ❌ (secondary)	-	-	0
Improvements ✅ (primary)	-0.6%	[-0.6%, -0.6%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	-0.0%	[-0.6%, 0.5%]	2

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)	12.3%	[0.9%, 31.6%]	129
Regressions ❌ (secondary)	13.2%	[0.0%, 43.8%]	84
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	12.3%	[0.9%, 31.6%]	129

Bootstrap: 660.993s -> 662.494s (0.23%)

[saethlin]

It looks like the size of rlibs grows between 1% and 30%. That seems a bit steep. Based on the fact that icount/cycles/time are perf-neutral I'm going to try to do some varint encoding.

[saethlin]

Of course, varint encoding isn't possible because this is a table format. So instead of a varint encoding, let's try just bumping up number of bytes we encode. 5 bytes should be enough for a while.

[saethlin]

@bors try @rust-timer queue

[bors]

⌛ Trying commit 6bb51e7 with merge 905eceaa29971ec6f56586b96233d97fb49b37d8...

[bors]

☀️ Try build successful - checks-actions
Build commit: 905eceaa29971ec6f56586b96233d97fb49b37d8 (905eceaa29971ec6f56586b96233d97fb49b37d8)

[bors]

☀️ Try build successful - checks-actions
Build commit: 905eceaa29971ec6f56586b96233d97fb49b37d8 (905eceaa29971ec6f56586b96233d97fb49b37d8)

[rust-timer]

Finished benchmarking commit (905eceaa29971ec6f56586b96233d97fb49b37d8): 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)	0.4%	[0.4%, 0.5%]	5
Regressions ❌ (secondary)	1.1%	[1.0%, 1.1%]	4
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-1.7%	[-1.7%, -1.7%]	1
All ❌✅ (primary)	0.4%	[0.4%, 0.5%]	5

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)	1.2%	[0.5%, 8.6%]	27
Regressions ❌ (secondary)	2.1%	[2.1%, 2.1%]	1
Improvements ✅ (primary)	-1.5%	[-1.9%, -1.1%]	2
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	1.1%	[-1.9%, 8.6%]	29

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)	1.0%	[0.5%, 2.1%]	3
Regressions ❌ (secondary)	3.0%	[2.1%, 3.7%]	4
Improvements ✅ (primary)	-0.4%	[-0.4%, -0.4%]	1
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	0.7%	[-0.4%, 2.1%]	4

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)	3.2%	[0.2%, 7.9%]	125
Regressions ❌ (secondary)	3.7%	[0.1%, 10.9%]	75
Improvements ✅ (primary)	-	-	0
Improvements ✅ (secondary)	-	-	0
All ❌✅ (primary)	3.2%	[0.2%, 7.9%]	125

Bootstrap: 661.654s -> 660.885s (-0.12%)

[saethlin]

That looks a lot more acceptable to me. We still have some regressions, which is inevitable given the fact that this changing the size of the elements of a table, but they are less.

Now I think we just need a regression test.

[Kobzol]

Here are the metadata regressions.

[saethlin]

Closing in favor of #113542