Fix perf regression from Miri Machine trait changes #62264

RalfJung · 2019-07-01T08:24:02Z

Maybe this fixes the perf regression that #62003 seemingly introduced?

Cc @nnethercote

rust-highfive · 2019-07-01T08:24:06Z

r? @zackmdavis

(rust_highfive has picked a reviewer for you, use r? to override)

RalfJung · 2019-07-01T08:24:19Z

@bors try

bors · 2019-07-01T08:24:32Z

⌛ Trying commit 8157856098a7038232d52a993bb945f7eafa0add with merge a36397fa11d1d3c42a1ac182d0159fa152cc005d...

bors · 2019-07-01T12:24:32Z

💥 Test timed out

RalfJung · 2019-07-01T12:28:24Z

The build succeeded but the timeout went off... @rust-lang/infra should the timeout be increased?

Let's see if we can time this anyway.
@rust-timer build a36397fa11d1d3c42a1ac182d0159fa152cc005d

rust-timer · 2019-07-01T12:28:25Z

Success: Queued a36397fa11d1d3c42a1ac182d0159fa152cc005d with parent 5748825, comparison URL.

rust-timer · 2019-07-01T12:28:49Z

Success: Queued a36397fa11d1d3c42a1ac182d0159fa152cc005d with parent 5748825, comparison URL.

mati865 · 2019-07-01T12:33:38Z

I think try builds will timeout until #62247 is merged.

rust-timer · 2019-07-01T15:35:43Z

Finished benchmarking try commit a36397fa11d1d3c42a1ac182d0159fa152cc005d, comparison URL.

RalfJung · 2019-07-01T15:37:51Z

That doesn't seem to make a difference. I guess it's time to revert parts of that PR to figure out where it is coming from.

(Please don't merge!)

RalfJung · 2019-07-01T15:41:02Z

@bors try

bors · 2019-07-01T15:41:14Z

⌛ Trying commit 269866b with merge 32136d3...

@nnethercote

[WIP] Debug perf regression from Miri Machine trait changes Maybe this fixes the perf regression that #62003 seemingly introduced? Cc @nnethercote

RalfJung · 2019-07-01T19:07:45Z

@rust-timer build 32136d3

rust-timer · 2019-07-01T19:07:47Z

Success: Queued 32136d3 with parent 765eebf, comparison URL.

rust-timer · 2019-07-01T23:14:42Z

Finished benchmarking try commit 32136d3, comparison URL.

RalfJung · 2019-07-02T09:10:12Z

Yeah, this confirms the perf regression is fixed by reverting the entire PR.

@nnethercote can you explain that? I am completely mystified. That just makes no sense at all. And without any hypothesis what is even causing this, it is very hard to do anything about it.^^

Next run: reverting those parts of the PR that we did just "for consistency", but don't technically need.

@bors try

bors · 2019-07-02T09:10:23Z

⌛ Trying commit 4c8da8b268372d1229bcdda40d2387f22af59860 with merge b33ad710d9c74907741387cf9a85c8e5b95b7e52...

RalfJung · 2019-07-02T09:12:29Z

Ah never mind, I found it.

Is there a way to abort the try build?

kennytm · 2019-07-02T09:42:30Z

@bors try- retry

bors · 2019-07-02T10:36:14Z

⌛ Trying commit ffc6670 with merge ed4ad9c...

@nnethercote

Fix perf regression from Miri Machine trait changes Maybe this fixes the perf regression that #62003 seemingly introduced? Cc @nnethercote

bors · 2019-07-02T13:33:14Z

☀️ Try build successful - checks-azure, checks-travis
Build commit: ed4ad9c

RalfJung · 2019-07-02T13:45:05Z

@rust-timer build ed4ad9c

rust-timer · 2019-07-02T13:45:07Z

Success: Queued ed4ad9c with parent 99abdfa, comparison URL.

rust-timer · 2019-07-02T17:05:42Z

Finished benchmarking try commit ed4ad9c, comparison URL.

RalfJung · 2019-07-02T17:11:28Z

Perf looks good, confirms my assumption that this fixes the regression. This is ready for review.

nnethercote · 2019-07-02T21:30:37Z

Thank you for hunting this down, RalfJung! For compiler perf, fixing regressions is less glamorous but just as important as writing new optimizations.

bors · 2019-07-04T03:19:21Z

☔ The latest upstream changes (presumably #62355) made this pull request unmergeable. Please resolve the merge conflicts.

…ooks This is needed to avoid doing unnecessary global alloc_map lookups

RalfJung · 2019-07-04T08:48:58Z

Rebased to fix conflict.

zackmdavis · 2019-07-05T03:10:59Z

@bors r+

bors · 2019-07-05T03:11:00Z

📌 Commit 52e6f85 has been approved by zackmdavis

@nnethercote

Fix perf regression from Miri Machine trait changes Maybe this fixes the perf regression that rust-lang#62003 seemingly introduced? Cc @nnethercote

@ghost

Rollup of 8 pull requests Successful merges: - #60260 (Add support for UWP targets) - #62151 (Update linked OpenSSL version) - #62245 (Miri engine: support extra function (pointer) values) - #62257 (forward read_c_str method from Memory to Alloc) - #62264 (Fix perf regression from Miri Machine trait changes) - #62296 (request at least ptr-size alignment from posix_memalign) - #62329 (Remove support for 1-token lookahead from the lexer) - #62377 (Add test for ICE #62375) Failed merges: r? @ghost

@nnethercote

Fix perf regression from Miri Machine trait changes Maybe this fixes the perf regression that rust-lang#62003 seemingly introduced? Cc @nnethercote

@ghost

Rollup of 7 pull requests Successful merges: - #62151 (Update linked OpenSSL version) - #62245 (Miri engine: support extra function (pointer) values) - #62257 (forward read_c_str method from Memory to Alloc) - #62264 (Fix perf regression from Miri Machine trait changes) - #62296 (request at least ptr-size alignment from posix_memalign) - #62329 (Remove support for 1-token lookahead from the lexer) - #62377 (Add test for ICE #62375) Failed merges: r? @ghost

rust-highfive assigned zackmdavis Jul 1, 2019

rust-highfive added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Jul 1, 2019

RalfJung changed the title ~~more inlining for force_bits~~ [WIP] Debug perf regression from Miri Machine trait changes Jul 1, 2019

RalfJung force-pushed the inline-forcing branch from 8157856 to 269866b Compare July 1, 2019 15:40

bors added a commit that referenced this pull request Jul 1, 2019

Auto merge of #62264 - RalfJung:inline-forcing, r=<try>

32136d3

[WIP] Debug perf regression from Miri Machine trait changes Maybe this fixes the perf regression that #62003 seemingly introduced? Cc @nnethercote

RalfJung force-pushed the inline-forcing branch from 269866b to 4a8f3d9 Compare July 1, 2019 19:08

RalfJung force-pushed the inline-forcing branch from 4a8f3d9 to 4c8da8b Compare July 2, 2019 09:02

RalfJung changed the title ~~[WIP] Debug perf regression from Miri Machine trait changes~~ Fix perf regression from Miri Machine trait changes Jul 2, 2019

RalfJung force-pushed the inline-forcing branch from 4c8da8b to ffc6670 Compare July 2, 2019 10:34

bors added a commit that referenced this pull request Jul 2, 2019

Auto merge of #62264 - RalfJung:inline-forcing, r=<try>

ed4ad9c

Fix perf regression from Miri Machine trait changes Maybe this fixes the perf regression that #62003 seemingly introduced? Cc @nnethercote

RalfJung force-pushed the inline-forcing branch from ffc6670 to 4ebddd8 Compare July 2, 2019 17:05

RalfJung mentioned this pull request Jul 2, 2019

try builds: include a copyable version of the full commit SHA in comment rust-lang/homu#53

Merged

RalfJung added 4 commits July 4, 2019 10:10

Machine: make self-like parameters come first

8ad28cd

Go back to just passing MemoryExtra to the machine-level allocation h…

127610b

…ooks This is needed to avoid doing unnecessary global alloc_map lookups

more inlining

956a3ef

organize methods a bit better

52e6f85

RalfJung force-pushed the inline-forcing branch from 4ebddd8 to 52e6f85 Compare July 4, 2019 08:41

bors added S-waiting-on-bors Status: Waiting on bors to run and complete tests. Bors will change the label on completion. and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Jul 5, 2019

Centril mentioned this pull request Jul 5, 2019

Rollup of 8 pull requests #62424

Closed

Centril mentioned this pull request Jul 6, 2019

Rollup of 7 pull requests #62428

Merged

bors merged commit 52e6f85 into rust-lang:master Jul 6, 2019

RalfJung deleted the inline-forcing branch July 6, 2019 08:05

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Fix perf regression from Miri Machine trait changes #62264

Fix perf regression from Miri Machine trait changes #62264

RalfJung commented Jul 1, 2019

rust-highfive commented Jul 1, 2019

RalfJung commented Jul 1, 2019

bors commented Jul 1, 2019

bors commented Jul 1, 2019

RalfJung commented Jul 1, 2019 •

edited

Loading

rust-timer commented Jul 1, 2019

rust-timer commented Jul 1, 2019

mati865 commented Jul 1, 2019

rust-timer commented Jul 1, 2019

RalfJung commented Jul 1, 2019 •

edited

Loading

RalfJung commented Jul 1, 2019

bors commented Jul 1, 2019

RalfJung commented Jul 1, 2019

rust-timer commented Jul 1, 2019

rust-timer commented Jul 1, 2019

RalfJung commented Jul 2, 2019

bors commented Jul 2, 2019

RalfJung commented Jul 2, 2019 •

edited

Loading

kennytm commented Jul 2, 2019

bors commented Jul 2, 2019

bors commented Jul 2, 2019

RalfJung commented Jul 2, 2019

rust-timer commented Jul 2, 2019

rust-timer commented Jul 2, 2019

RalfJung commented Jul 2, 2019

nnethercote commented Jul 2, 2019

bors commented Jul 4, 2019

RalfJung commented Jul 4, 2019

zackmdavis commented Jul 5, 2019

bors commented Jul 5, 2019

Fix perf regression from Miri Machine trait changes #62264

Fix perf regression from Miri Machine trait changes #62264

Conversation

RalfJung commented Jul 1, 2019

rust-highfive commented Jul 1, 2019

RalfJung commented Jul 1, 2019

bors commented Jul 1, 2019

bors commented Jul 1, 2019

RalfJung commented Jul 1, 2019 • edited Loading

rust-timer commented Jul 1, 2019

rust-timer commented Jul 1, 2019

mati865 commented Jul 1, 2019

rust-timer commented Jul 1, 2019

RalfJung commented Jul 1, 2019 • edited Loading

RalfJung commented Jul 1, 2019

bors commented Jul 1, 2019

RalfJung commented Jul 1, 2019

rust-timer commented Jul 1, 2019

rust-timer commented Jul 1, 2019

RalfJung commented Jul 2, 2019

bors commented Jul 2, 2019

RalfJung commented Jul 2, 2019 • edited Loading

kennytm commented Jul 2, 2019

bors commented Jul 2, 2019

bors commented Jul 2, 2019

RalfJung commented Jul 2, 2019

rust-timer commented Jul 2, 2019

rust-timer commented Jul 2, 2019

RalfJung commented Jul 2, 2019

nnethercote commented Jul 2, 2019

bors commented Jul 4, 2019

RalfJung commented Jul 4, 2019

zackmdavis commented Jul 5, 2019

bors commented Jul 5, 2019

RalfJung commented Jul 1, 2019 •

edited

Loading

RalfJung commented Jul 1, 2019 •

edited

Loading

RalfJung commented Jul 2, 2019 •

edited

Loading