1028867 - Add a --latin1-strings shell flag

Status: RESOLVED FIXED

(Core :: JavaScript Engine, defect)

Description

[Jan de Mooij [:jandem]]

Would be nice for fuzzing, and we could add it to jit_test.py --tbpl to avoid regressions.

Comment 1

[Gary Kwong [:gkw] [:nth10sd] (NOT official MoCo now)]

Added support in fuzzing rev b7eb5027be93 - this flag will be tested occasionally if the shell supports it.

Comment 2

[Jan de Mooij [:jandem]]

Attachment: Patch

Comment 3

[Jan de Mooij [:jandem]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/b19674e5222c

(In reply to Gary Kwong [:gkw] [:nth10sd] from comment #1)
> Added support in fuzzing rev b7eb5027be93 - this flag will be tested
> occasionally if the shell supports it.

Thanks Gary! There will probably be a few places where we assume 16-bit chars that are not covered by our tests. Hopefully once we fix those there shouldn't be too much bustage.

Comment 4

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

Backed out in https://hg.mozilla.org/integration/mozilla-inbound/rev/dcd6d54edd87 (and part 5 of bug 1028866 in https://hg.mozilla.org/integration/mozilla-inbound/rev/e97f7d9d54d9 ) under suspicion of turning Windows PGO browser-chrome-1 incredibly failure prone.

On the push before part 5 and bug 1028867 landed, bc1 looked like https://tbpl.mozilla.org/?tree=Mozilla-Inbound&jobname=Windows%20XP%2032-bit%20mozilla-inbound%20pgo%20test%20mochitest-browser-chrome-1&rev=7a7f6701c903 and https://tbpl.mozilla.org/?tree=Mozilla-Inbound&jobname=Windows%207%2032-bit%20mozilla-inbound%20pgo%20test%20mochitest-browser-chrome-1&rev=7a7f6701c903

On the push that they landed, bc1 now looks like https://tbpl.mozilla.org/?tree=Mozilla-Inbound&jobname=Windows%20XP%2032-bit%20mozilla-inbound%20pgo%20test%20mochitest-browser-chrome-1&rev=b19674e5222c and https://tbpl.mozilla.org/?tree=Mozilla-Inbound&jobname=Windows%207%2032-bit%20mozilla-inbound%20pgo%20test%20mochitest-browser-chrome-1&rev=b19674e5222c

Comment 5

[Jan de Mooij [:jandem]]

https://hg.mozilla.org/integration/mozilla-inbound/rev/5c88c5b4fe07

Comment 6

[Phil Ringnalda (:philor)]

https://hg.mozilla.org/mozilla-central/rev/5c88c5b4fe07

Comment 7

[Hannes Verschore [:h4writer]]

Haven't looked to the patch, but is seems this introduced a octane-deltablue and kraken-parse-financial on x86 mac os shell builds regression?

Comment 8

[Jan de Mooij [:jandem]]

(In reply to Hannes Verschore [:h4writer] from comment #7)
> Haven't looked to the patch, but is seems this introduced a octane-deltablue
> and kraken-parse-financial on x86 mac os shell builds regression?

The deltablue regression is gone, somehow. parse-financial is only a few ms now; I'll see if I can repro and if there's an easy fix. With this patch EnableLatin1Strings is no longer a constant but has to be tested at runtime, but it's not used on paths where this should matter.