1346269 - Crash in js::jit::AutoFlushICache::AutoFlushICache

Status: RESOLVED FIXED

(Core :: JavaScript Engine: JIT, defect, P1)

Description

[Julian Seward [:jseward]]

This bug was filed from the Socorro interface and is 
report bp-02ae372f-45d3-46ac-8b4b-e32cd2170309.
=============================================================

This is the #2 crash in the Android nightly of 20170309110217.

Comment 1

[Benjamin Bouvier [:bbouvier] (inactive)]

wasm::DeserializeModule on the stack.

Comment 2

[Luke Wagner [:luke]]

Ah, so we're deserializing a module off the main thread and the build-id mismatches so we're recompiling, and deep inside someone is assuming we're on a JSContext's main thread so TlsContext.get() is non-null, but we're on some IDB thread so it is indeed null.  The whole point of this AFC is just to suppress icache flushing; let's see if there's a better way to do that that doesn't involve a JSContext*.

Comment 3

[Luke Wagner [:luke]]

Attachment: fix-afc

The simplest solution I found is just to make the suppression explicit as a parameter.

Comment 4

[Benjamin Bouvier [:bbouvier] (inactive)]

Comment on attachment 8846059 [details] [diff] [review]
fix-afc

Review of attachment 8846059 [details] [diff] [review]:
-----------------------------------------------------------------

Thanks.

Comment 5

[Pulsebot]

Pushed by [e-mail]:
https://hg.mozilla.org/integration/mozilla-inbound/rev/195ba401a2c6
Baldr: remove dependency on TlsContext in ModuleGenerator::finish (r=bbouvier)

Comment 6

[Hannes Verschore [:h4writer]]

What is the version this was introduced? We should backport this.

Comment 7

[Wes Kocher (:KWierso) (Not reading bugmail; email directly if needed)]

https://hg.mozilla.org/mozilla-central/rev/195ba401a2c6

Comment 8

[Ryan VanderMeulen [:RyanVM]]

crash-stats shows crashes on 52 for this (all Android, though). Given that this patch touches non-ARM code as well, I'm wondering if it's worth backporting to ESR52 as well. There are some Tier 3 ARM & Android builds off that branch IIRC.

Comment 9

[Benjamin Bouvier [:bbouvier] (inactive)]

esr52 hasn't shipped wasm (bug 1342440), so no need to backport there. Other platforms are worth it, though. Thanks!

Comment 10

[Luke Wagner [:luke]]

Comment on attachment 8846059 [details] [diff] [review]
fix-afc

Approval Request Comment
[Feature/Bug causing the regression]: 1342060
[User impact if declined]: crash in FF Android when deserializing wasm module after build-id change
[Is this code covered by automated tests?]: yes
[Has the fix been verified in Nightly?]: yes
[Needs manual test from QE? If yes, steps to reproduce]: no
[List of other uplifts needed for the feature/fix]: none
[Is the change risky?]: no
[Why is the change risky/not risky?]: pretty simple change, just adding/propagating an additional parameter

Comment 11

[Julien Cristau [:jcristau]]

Comment on attachment 8846059 [details] [diff] [review]
fix-afc

fix a crash in wasm, beta53+, aurora54+

Comment 12

[Gerry Chang [:gchang]]

Comment on attachment 8846059 [details] [diff] [review]
fix-afc

Fix a crash. Beta53+ & Aurora54+.

Comment 13

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-aurora/rev/c33b6585c80d

Comment 14

[Ryan VanderMeulen [:RyanVM]]

https://hg.mozilla.org/releases/mozilla-beta/rev/8d5b69356034

Comment 15

[Andrei Vaida [:avaida]]

Setting qe-verify- based on Luke's assessment on manual testing needs (see Comment 10) and the fact that this fix has automated coverage.