Граф коммитов

240635 Коммитов

Автор SHA1 Сообщение Дата
Nicholas Nethercote 7258795aa4 Bug 639420 - Speed up the scanner ten ways, part 6. r=brendan. 2011-03-16 15:17:55 -07:00
Nicholas Nethercote 8d79a77d17 Bug 639420 - Speed up the scanner ten ways, part 5. r=brendan. 2011-03-16 15:17:38 -07:00
Nicholas Nethercote 67ab711320 Bug 639420 - Speed up the scanner ten ways, part 4. r=brendan. 2011-03-16 15:17:21 -07:00
Nicholas Nethercote 862b0d3b9b Bug 639420 - Speed up the scanner ten ways, part 3. r=brendan. 2011-03-16 15:17:05 -07:00
Nicholas Nethercote 3a2df9994f Bug 639420 - Speed up the scanner ten ways, part 2. r=brendan. 2011-03-16 15:16:21 -07:00
Nicholas Nethercote be665996a0 Bug 639420 - Speed up the scanner ten ways, part 1. r=brendan. 2011-03-16 15:16:08 -07:00
Bob Clary 1e7a57b9f5 Bug 610185 - JavaScript Tests - temporarily mark ecma/Date/15.9.5.34-1.js as random due to DST orange, r=jorendorff. 2011-03-16 14:39:30 -07:00
Igor Bukanov e378f67446 Bug 633219 - replacing resolve hashtable with a linked list. r=luke 2011-03-07 23:00:00 +01:00
Michael Wu ee48d4040e Bug 632253 - Move filename check to JS_XDRScript, r=igor 2011-03-15 18:01:00 -07:00
Michael Wu 20eca7a69e Bug 632253 - Move magic check to JS_XDRScript, r=igor 2011-03-15 18:00:58 -07:00
Luke Wagner 0994a69030 Bug 640852 - Don't try to avoid marking atoms in ropes and dependent strings (r=gwagner)
--HG--
extra : rebase_source : 5cd329574101889ba97b512ad464b23f1e2e693e
2011-03-14 13:55:58 -07:00
Luke Wagner f7f121db6c Bug 640274 - rm {JS_,js_,}GetExternalStringGCType (r=brendan)
--HG--
extra : rebase_source : 1a2b1df8c7ed9cdbf0455e6e8c117382cb1fb2bb
2011-03-14 13:55:43 -07:00
Luke Wagner 52ac799f59 Bug 640294 - Remove constant conditional from js_AtomizeString (r=gal)
--HG--
extra : rebase_source : c0eaaa662857ef420002f7446fe82132f12d21f9
2011-03-14 13:55:24 -07:00
Jim Blandy dac293078a Bug 554955: Give blocks and call objects unique shapes when they have parents that may be extended with new bindings. r=jorendorff
The comments for js::Bindings::extensibleParents explain why this is necessary.

AssertValidPropertyCacheHit should have been catching this bug, but for
reasons I don't understand, it is restricted from checking this case. This
patch extends it to assert when the bug is detected.

I've gathered the infallible parts of the initialization for Call objects
and cloned block objects into their own functions.
2011-03-15 12:18:36 -07:00
Jim Blandy 615c9163b1 Bug 554955: Make JSObject::setMap not pretend to take a const shape. r=jorendorff 2011-03-15 12:18:36 -07:00
Jim Blandy 427569d433 Bug 554955: Introduce a typed accessor for retrieving functions from JSFunctionBoxes. r=brendan 2011-03-15 12:18:36 -07:00
Jim Blandy 1b3b57d96b Bug 554955: Fix comment for JSOP_NEW. no r. 2011-03-15 12:18:36 -07:00
Jim Blandy 812d2a14ed Bug 554955: Revert test and assertion changes from bug 633890. r=jorendorff
At present, AssertValidPropertyCacheHit doesn't catch most property cache
failures. I would like to fix it, but doing so makes
jit-test/tests/basic/bug633890.js fail with a C++ assertion, not a JS shell
assertion, which jit-test/jit_test.py can't handle.

I think it's important to have AssertValidPropertyCacheHit doing its job;
it would have caught problems like bug 554955 earlier. Since bug633890.js
is a known failure (bug 458271), deleting it doesn't reduce coverage.

This patch deletes that test. A later patch in the series fixes
AssertValidPropertyCacheHit.
2011-03-15 12:18:35 -07:00
Jason Orendorff 8730d5b197 Backed out changeset 39de74c74b20, which is causing the browser to leak windows. 2011-03-14 17:37:04 -05:00
Jason Orendorff 28d3dae7d9 Bug 561359 - Change JSOP_LAMBDA to apply the method optimization deterministically. In particular, it no longer depends on whether enclosing Blocks have been reified. This prevents incorrect behavior and assertions when a JSOP_LAMBDA, JSOP_INITMETHOD pair apply the method optimization once, populating the property cache, but later the same JSOP_LAMBDA instruction does not (under the old code) apply the optimization. With this patch, if JSOP_LAMBDA pushes the uncloned function once, it always will. r=brendan. 2011-03-14 15:54:34 -05:00
Blake Kaplan da297b34c9 Bug 606388 - Re-enable browser_popupUI.js. It is trying to use == to compare an Xray wrapper with a SJOW. r=jst 2011-03-11 12:07:43 -08:00
Bill McCloskey e7eff8b6a0 Bug 639954 - *Really* set cx->compartment to NULL during GC (r=luke) 2011-03-14 10:37:05 -07:00
Jeff Walden 9c2000bde9 Bug 637643 - Tests. r=async.processingjs@yahoo.com
--HG--
extra : rebase_source : 190f79659ffb1305726347310d6c7bb4f4f7d2b1
2011-03-14 10:17:40 -07:00
async.processingjs@yahoo.com 0915122f89 Bug 637643 - A subarray of a typed array subarray not at offset zero doesn't contain the right data. r=jwalden
--HG--
extra : rebase_source : f6fe88592e66101dc065bd82f35e4d2f3fd1cb1a
2011-03-14 10:17:36 -07:00
Andreas Gal 0e5588e8dc Follow-up fix for bug 639729. 2011-03-13 19:20:06 -07:00
Andreas Gal a163a8e485 Inline GC mark path (bug 639729, r=billm). 2011-03-13 18:13:52 -07:00
Bill McCloskey 7d7eaad649 Bug 596338 - Fix --disable-methodjit --disable-tracejit builds (r=ginn.chen) 2011-03-11 15:40:05 -08:00
Brian Hackett 57f7b187fb Trace Parser::emptyCallShape, bug 641103. r=gal 2011-03-11 14:49:11 -08:00
Jeff Walden ff4297e5d0 Correct a URL in tests for bug 614608. r=sparky but DONTBUILD because it's such a small and unimportant change 2011-03-11 13:15:58 -08:00
Bill McCloskey c373416613 Bug 636544 - Fix JavaScript "-m -j" mode (r=dmandelin) 2011-03-11 11:44:57 -08:00
Bill McCloskey c1bbeac8fb Bug 639858 - Bias the profiler toward tracing scripted getters (r=dmandelin) 2011-03-11 10:21:33 -08:00
Nicholas Nethercote 2a84837407 Bug 638627 - Properly handle failures to patch guards due to non-32-bit offsets (TM-specific part). r=gal,dvander. 2011-03-10 16:41:50 -08:00
Nicholas Nethercote 4091ef806d Update nanojit-import-rev stamp. 2011-03-10 16:27:55 -08:00
Nicholas Nethercote c4f33efa34 Bug 638627 - Properly handle failures to patch guards due to non-32-bit offsets (NJ-specific part). r=dvander.
--HG--
extra : convert_revision : b3948560bcbc91ed1e7416d2f9ad495e1c8daca1
2011-03-10 16:16:42 -08:00
Gregor Wagner d27001da51 Bug 638199 - Shrink ArenaHeader (r=wmccloskey) 2011-03-10 15:27:21 -08:00
Nicholas Nethercote 93d1a34c53 merge 2011-03-10 14:32:03 -08:00
Nicholas Nethercote d4a5263fbf Backed out changeset cc7311c09b56 (bug 602397) due to Maemo Talos breakage. 2011-03-10 14:31:34 -08:00
Jacek Caban 79079b108b Bug 633924 - MethodGIT broken on mingw-w64 win64 build r=dvander 2011-03-10 11:32:23 +00:00
Nicholas Nethercote 18383cf18b Bug 602397 - TM: clean up TraceRecorder::alu(). r=wmccloskey. 2011-03-09 15:21:07 -08:00
Brendan Eich d33a68e1e3 Fix assertion botched by function using a previously mentioned name and therefore claiming its placeholder or declared definition; the function must have sane end vs. begin source coordinates for error reporting purposes (640075, r=njn). 2011-03-08 23:51:27 -08:00
Andreas Gal b8b659ffdd Don't shrink object slots during GC (bug 639727, r=bhackett). 2011-03-09 00:53:56 -08:00
Andreas Gal 9b291da73b Sweep compartments when no objects live in them (bug 639270, r=gregor). 2011-03-08 20:58:38 -08:00
Phil Ringnalda 41c0b7c782 Back out 80d8431f209e (bug 639270) for failing to actually compile 2011-03-08 19:59:30 -08:00
Andreas Gal 74d162d185 Sweep compartments when no objects live in them (bug 639270, r=gregor). 2011-03-08 18:10:47 -08:00
Blake Kaplan b55a8920aa Fix some stuff brendan pointed out. 2011-03-08 18:08:59 -08:00
Nicholas Nethercote aab716a96f Bug 638034 - Make scanning safer. r=brendan. 2011-03-08 16:10:51 -08:00
Blake Kaplan cc54206af2 Bug 639501 - Setters can cause the handler to change. r=gal
--HG--
extra : rebase_source : 5cef14e53e2f9032218c725538ba2afbf467191c
2011-03-08 15:05:11 -08:00
Blake Kaplan f28f2573f7 Bug 638026 - Make sure we hold onto the right JS object here. r=jst
--HG--
extra : rebase_source : 5ebebc39f689d0b76e51b2ca79fa432871b1404e
2011-03-08 15:04:59 -08:00
Blake Kaplan 310733626a Bug 639737 - JS_THIS_OBJECT can fail and we need to catch when it does. r=jst
--HG--
extra : rebase_source : 0ecd6f2666c506bbf5837ef14092fcb47c8e5b3b
2011-03-08 15:04:37 -08:00
Jeff Walden 7b9ec4b694 Bug 638997 - Remove the id field from PropDesc: users who needed it should manage the id separately. r=luke
--HG--
extra : rebase_source : eb2152a2259a81bca1d7a4f5521787f9385a3c6f
2011-03-04 17:24:15 -08:00