Bas Schouten
88b755bca9
Bug 719776 - Part 2: Make MOZ_CRASH and MOZ_ASSERT expand inline. r=waldo
2012-04-12 02:03:07 +02:00
Brian Hackett
02a516051e
Use right compartment iterator when profiling scripts, bug 744611. r=billm
2012-04-11 15:54:55 -07:00
Terrence Cole
59e9f32d2b
Bug 743868 - Fix barriers in SupressDeletedProperties; r=billm
...
Right now, we simply stop tracing through deleted properties. This causes us to
delete things that are still in the write buffer. We need to, in addition, set
the reference to NULL to ensure that the write buffer doesn't dereference these
dead entries.
--HG--
extra : rebase_source : add220d5fb47e0e3ffd767c9b5ebd6172df07671
2012-04-09 16:52:50 -07:00
Masatoshi Kimura
d72b323902
Bug 743049 - Part 1: Add a friend API to get an error type name from a number. r=luke
2012-04-11 17:55:22 -04:00
Masatoshi Kimura
0cd58c8a6f
Bug 730161 - Replace IDBDatabaseException and FileException with DOMException. r=sicking
2012-04-11 17:55:21 -04:00
Terrence Cole
04d84ae8bf
No bug - Warning fix for 91357:fa24b215d49e; r=luke
...
--HG--
extra : rebase_source : 9d48a4a09cb54d711a24da41576ffc24d44a3cb4
2012-04-11 10:32:50 -07:00
Terrence Cole
306f897d42
Bug 743896 - Fix failure misattribution in jstest runner; r=dmandelin
...
In the unix task runner, when we kill tasks, we should be checking and returning
bool if we kill so that we then run waitpid in timeout mode. We weren't, so
tasks were never failing with timeout, only with failure. It turns out that
this is not safe anyway: it will race against signal delivery. We're better off
just removing this hack and letting the timeout cutoff be fuzzy.
--HG--
extra : rebase_source : cd44838e67cf541fd5741406e8c3f02ad7530f16
2012-04-11 10:02:33 -07:00
Nathan Froyd
301dbca285
Bug 743988 - unconditionally link PROGRAMS and SIMPLE_PROGRAMS with CXX; r=ted
2012-04-10 10:16:26 -04:00
Jan de Mooij
c172af7ba1
Bug 742393 - Don't inline array.concat if the array prototype has indexed properties. r=bhackett
2012-04-11 18:49:42 +02:00
Phil Ringnalda
dc765d4953
Back out 4b82f26f8593:2fcf9cbedd48 (bug 743393, bug 719776) for bustage
2012-04-11 09:06:40 -07:00
Bas Schouten
8ecf21d31c
Bug 719776 - Part 1 - Followup: Fix one more case of assuming ASSERT expands to an expression. r=bustage
2012-04-11 17:18:30 +02:00
Bas Schouten
4458bdfe56
Bug 719776 - Part 2: Inline MOZ_Assert and MOZ_Crash. r=waldo
2012-04-11 16:55:31 +02:00
David Rajchenbach-Teller
108eebc31e
Bug 742384 - CDataFinalizer.dispose now returns a value. r=jorendorff
2012-04-11 10:46:19 +02:00
David Rajchenbach-Teller
1ff0a12dd6
Bug 720771 - Implementing finalization for CData values. r=jorendorff
2012-04-05 15:06:25 +02:00
Mike Hommey
451bafe5f3
Bug 644608 - Implement full dependencies for expandlibs. r=ted
...
* * *
Bug 644608 - Fixups
2012-03-31 09:24:39 +02:00
Nicholas Nethercote
31d7713366
Bug 739512: Patch 9: move JSScript::debug into a table. r=jorendorff.
2012-04-10 18:03:01 -07:00
Nicholas Nethercote
091eef05a5
Bug 739512 - Patch 8: move sourceMap into a table. r=jorendorff.
...
--HG--
extra : rebase_source : a986cf630b84c9b531e6179afce57b0e8c69d1ff
2012-04-09 19:56:04 -07:00
Andrew McCreight
c3b5a510a8
Bug 744272. disable test_watchpoints due to bogus failures. rs=billm
2012-04-10 18:43:22 -07:00
Nicholas Nethercote
70a553b78b
Bug 739512 - Patch 7: move scriptCounts into a table (attempt 2). r=bhackett.
2012-04-10 18:41:42 -07:00
Nicholas Nethercote
8bdc9b2a07
Backed out changeset fdd10e0e6e5e (bug 739512) due to 32-bit bustage.
2012-04-10 18:36:08 -07:00
Nicholas Nethercote
12cb5ac08b
Bug 739512 - Patch 7: move scriptCounts into a table. r=bhackett.
2012-04-09 19:50:06 -07:00
Jeff Walden
0457887cdb
Bug 743878 - Move uint8_clamped and associated machinery into ObjectImpl.h, so that the new object representation can use it as well. r=sfink
2012-04-09 15:14:59 -07:00
Terrence Cole
8c91faa124
Backed out changeset a43eb4a2d2fd
2012-04-10 17:03:25 -07:00
Terrence Cole
23e2dd5dc5
Bug 744192 - Assert that Vectors do not contain implicitly postbarriered types; r=luke
...
Since vectors move their memory around outside the control of the GC, it is not
valid to store an implicitly post-barriered item in a Vector.
2012-04-10 16:43:54 -07:00
Luke Wagner
ca9f2eb792
Fix --disable-methodjit bustage (no bug, r=red)
2012-04-10 14:57:15 -07:00
Luke Wagner
53b1f856fe
Bug 740259 - add ALIASEDVAR ops, assert they are used iff the variable is aliased (r=bhackett,waldo,jorendorff,dherman)
2012-03-24 12:29:36 -07:00
Luke Wagner
4a5cf1cd46
Bug 740446 - make 'arguments' more like a normal local binding (r=bhackett)
2012-04-02 08:58:30 -07:00
Luke Wagner
cebbba852e
Bug 740259 - Assert that dynamic binding access is expected (r=bhackett)
2012-03-16 12:02:37 -07:00
Luke Wagner
cf110a1054
Bug 740259 - Tighten assertions around closed-var noting (r=bhackett)
2012-03-28 16:36:57 -07:00
Brian Hackett
cddc63d7af
Treat FILTER and ENDFILTER as unknown ops, bug 740595. r=luke
2012-04-10 12:59:38 -07:00
Brian Hackett
87a73454cf
Remove bogus assert, bug 727476. r=luke
2012-04-10 12:56:59 -07:00
Brian Hackett
0cdfe44fd2
Avoid assert botch in makeLazyType after types have been nuked, bug 734978. r=luke
2012-04-10 12:36:26 -07:00
Brian Hackett
ed990a27c8
Don't botch assert after OOM while constructing object's lazy type, bug 735016. r=luke
2012-04-10 12:27:13 -07:00
Brian Hackett
77930876d9
Bail out from on calling functions if an OOM occurs while during initial bytecode analysis, bug 727341. r=luke
2012-04-10 12:24:46 -07:00
Brian Hackett
0c39a16808
Preserve invariant between saved array and lifetime variables' saved fields on OOM during analysis, bug 735045. r=luke
2012-04-10 12:16:42 -07:00
Brian Hackett
1988d9b71c
Nuking types should be recognized as a recompilation by enclosing monitors, bug 735012. r=dvander
2012-04-10 12:10:04 -07:00
Brian Hackett
ccc0237cf8
Don't botch assert after OOM while linking ICs, bug 727344. r=dvander
2012-04-10 12:07:26 -07:00
Terrence Cole
dfd9f312e5
Bug 744052 - Restructure HashTable::checkOverflow for easier reading; r=luke
...
When I added this function, I just copied the existing logical structure. It
makes more sense here to check overflowed and return early, since we can do this
now.
2012-04-10 09:51:48 -07:00
Ehsan Akhgari
5203026f2f
Merge mozilla-central into mozilla-inbound
2012-04-10 11:12:26 -04:00
Ehsan Akhgari
46c4c63e28
Merge the last PGO-green changeset from mozilla-inbound into mozilla-central
2012-04-10 11:11:37 -04:00
Igor Bukanov
b55c0c0cb1
Bug 743559 - Remove unused JSRuntime::xml_singleton fields. r=:Waldo
...
The patch removes unused anynameObject and functionNamespaceObject fields in JSRuntime.
2012-04-10 11:57:29 +02:00
Terrence Cole
9a178239bc
Bug 743885 - Fix OOM handling regression in HashTable; r=luke
...
Introduced by me in bug 722946.
2012-04-09 19:07:11 -07:00
Kyle Huey
902e133eab
Refix bug 741248 harder. r=me
2012-04-09 17:41:22 -07:00
Brian Hackett
d73b9c4ac3
Don't disable call ICs with the callee frame still on the stack, bug 743088. r=luke
2012-04-09 17:33:42 -07:00
Nicholas Nethercote
04470bd3ad
Bug 742163 (follow-up) - Unbreak --disable-methodjit. r=Ms2ger.
2012-04-09 16:41:27 -07:00
Nicholas Nethercote
a6b25d88ec
Bug 739512 - Patch 2b: rename nClosed{Args,Vars}() as numClosed{Args,Vars}(). r=luke.
2012-04-09 16:37:16 -07:00
Matt Brubeck
4ee87a2a2f
Merge last green changeset from inbound to mozilla-central
2012-04-09 10:05:48 -07:00
Takanori MATSUURA
4a483c2e4c
Bug 743433 - Remove .orig files which were accidentally added by bug 680246. r=khuey
2012-04-07 08:48:23 -04:00
Igor Bukanov
d3018bc6d8
Bug 737364 - part 2 - replace JSContext with JSRuntime in the GC-related API
...
This part replaces the JSContext *cx argument in most GC-related API
with JSRuntime *rt. When possible, the patch removes the code to obtain
a temporary cx just to run the GC.
The patch also removes JS_DestroyContextMaybeGC. That function is not
used in FF code base and its implementation is broken. It requires that
the context has an entered compartment when it is destroyed, which in
turns implies a missing leave compartment call.
2012-03-28 12:13:30 +02:00
Kyle Huey
c06dcdfb58
Fix Bug 743376. r=bz
2012-04-06 17:44:50 -07:00