Commit Graph

28665 Commits

Author SHA1 Message Date
Nicolas Pierron
ccf9338ab8 Bug 909544 - IonMonkey: Check VM Call sanity in visit functions. r=jandem 2013-08-29 10:30:35 -07:00
Jan de Mooij
b865ad7c58 Bug 909401 - Fix MacroAssembler::tagValue on x86/ARM to not assert if the input register is the same as the output type register. r=nbp 2013-08-29 17:39:22 +02:00
Bobby Holley
e5550b2f93 Bug 908019 - Split up watchdog tests to run in parallel. r=mrbkap
This speeds up local xpcshell-test runs by ~6 seconds.
2013-08-29 09:32:25 -07:00
Brian Hackett
359dbd34fe Bug 908699 - Allow generating parse errors and warnings when off the main thread, r=billm. 2013-08-29 09:56:05 -06:00
Jon Coppeard
8bab73cb34 Bug 906387 - Rehash hash tables if necessary when rekeying during minor GC r=terrence 2013-08-29 10:27:50 +01:00
Jan de Mooij
30de7f9dde Bug 910240 part 2 - Assert IonBuilder sets the Folded flag if needed. r=bhackett 2013-08-29 10:08:22 +02:00
Jan de Mooij
a537c05378 Bug 910240 part 1 - Add some missing setFoldedUnchecked calls. r=bhackett 2013-08-29 10:08:16 +02:00
Nicholas Nethercote
9e88f47219 Bug 910109 (part 2) - Make jswrapper.h not depend on jsapi.h. r=luke.
--HG--
extra : rebase_source : 138e93b074691e9da0feab9cd24e149f4d5edf6a
2013-08-28 17:24:34 -07:00
Nicholas Nethercote
57b389b877 Bug 910109 (part 1) - Make jsproxy.h not depend on jsapi.h. r=luke.
--HG--
extra : rebase_source : bc2fa911e3a7feaafb778b756994e126257585a7
2013-08-28 17:20:24 -07:00
Benjamin Bouvier
e2296fbc66 Bug 900125: Tests for Math.fround; r=waldo 2013-08-28 22:00:31 -07:00
Benjamin Bouvier
f31a47ba7a Bug 900125: Add Math.fround to the interpreter; r=sstangl 2013-08-28 22:00:11 -07:00
Boris Zbarsky
ec93636aa6 Bug 903419 part 2. Report unhandled rejections in promises. r=smaug,bholley,luke 2013-08-29 00:30:06 -04:00
Boris Zbarsky
7891c84fba Bug 903419 part 1. Change js::GetErrorTypeName to take a JSRuntime, not a JSContext. r=luke 2013-08-29 00:30:06 -04:00
Eric Faust
fe5e567440 Bug 905999 - Part 3: Only generate Array.length stubs in idempotent caches with accurate TI. (r=bhackett)
Relanding after erroneous backout on a CLOSED TREE.
2013-08-28 16:12:59 -07:00
Eric Faust
13fed2fb18 Bug 905999 - Part 2: Store cache location data in IonScript for runtime lookup. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
712658a5dc Bug 905999 - Part 1: Get accurate location data for idempotent GetPropertyCaches. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
956d2bbd96 Bug 905999 - Part 0: Implement InlineConcatList for union-based algorithms. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
7f18306e49 Bug 909989 - Part 2: Use DataPtr in visit*IC instead of direct pointers. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
31bef2d0f9 Bug 909989 - Part 1: Implement DataPtr to refer to objects in runtimeData_. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
373dcf1704 Bug 909989 - Part 0: Use runtimeData_ offsets instead of cacheList_ offsets to refer to ICs. (r=nbp) 2013-08-28 16:12:59 -07:00
Ryan VanderMeulen
94f3ba03e7 Backed out 7 changesets (bug 905999, bug 909989) for Linux tp crashes on a CLOSED TREE.
Backed out changeset 3bb7a1c950a3 (bug 905999)
Backed out changeset a95e00dbefea (bug 905999)
Backed out changeset f66ece725f98 (bug 905999)
Backed out changeset f7779522f6f6 (bug 905999)
Backed out changeset 91b14e27101a (bug 909989)
Backed out changeset f1842e91034d (bug 909989)
Backed out changeset ccc30b7c2277 (bug 909989)
2013-08-28 21:53:28 -04:00
Ryan VanderMeulen
7f252b9c76 Backed out 2 changesets (bug 909447, bug 909441) for jit-test failures on a CLOSED TREE.
Backed out changeset 3300c750cd53 (bug 909447)
Backed out changeset b1b242a80ef7 (bug 909441)
2013-08-28 20:18:00 -04:00
Benjamin Bouvier
cc27f21f1b Bug 908001: Reuse default block for non cases in giant sparse switch tables in Odin; r=mjrosenb 2013-08-28 17:03:29 -07:00
Eric Faust
6c7534ccaf Bug 905999 - Part 3: Only generate Array.length stubs in idempotent caches with accurate TI. (r=bhackett) 2013-08-28 16:12:59 -07:00
Eric Faust
6eb7fe8594 Bug 905999 - Part 2: Store cache location data in IonScript for runtime lookup. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
c1d40f9171 Bug 905999 - Part 1: Get accurate location data for idempotent GetPropertyCaches. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
b1da7f3c84 Bug 905999: Part 0: Implement InlineConcatList for union-based algorithms. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
a68f6d7a57 Bug 909989 - Part 2: Use DataPtr in visit*IC instead of direct pointers. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
fc9ab6f8be Bug 909989 - Part 1: Implement DataPtr to refer to objects in runtimeData_. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
f0a5a7010c Bug 909989 - Part 0: Use runtimeData_ offsets instead of cacheList_ offsets to refer to ICs. (r=nbp) 2013-08-28 16:12:59 -07:00
Eric Faust
257deada77 Bug 908507 - Remove unused scriptList() from IonScript and MIRGraph. (r=nbp) 2013-08-28 16:12:58 -07:00
Eric Faust
61a2982620 Bug 909601 - Don't optimize common accessors on objects with {get,set}Generic hooks. (r=jandem) 2013-08-28 16:12:58 -07:00
Gary Kwong
725c93a76e Bug 909447 - Test. r=bholley DONTBUILD 2013-08-28 16:03:18 -07:00
Gary Kwong
bebe91bf29 Bug 909441 - Test. r=bholley 2013-08-28 16:02:51 -07:00
Shu-yu Guo
23f945b1ef Bug 905986 - Refactor adjusting phi types. (r=jandem) 2013-08-28 15:15:08 -07:00
Bobby Holley
a899450297 Bug 732665 - Tests. r=luke 2013-08-28 11:04:16 -07:00
Bobby Holley
756469507b Bug 732665 - Use separate stack depths for XPConnect JS. r=luke 2013-08-28 11:04:16 -07:00
Bobby Holley
83137c28bf Bug 732665 - Clean up and clarify the stack sizing logic. r=luke
The only change in behavior here is that we give the osx check higher precedence
than the DEBUG check, meaning that we end up with 7MB stacks on osx debug rather
than 4MB stacks. But this seems strictly better given the logic involved here.
2013-08-28 11:04:16 -07:00
Bobby Holley
f817aef987 Bug 732665 - Introduce a mechanism for specifying different stack limits for system, trusted, and untrusted code. r=luke
We kind of cop out for ion and PJS stuff. We should at least have a good story
for ion, but it probably needs to be done with someone with more jit familiarity
than I have.

The uninlined function call in JS_CHECK_RECURSION is unfortunate. We could fix
this either by separating the jsfriendapi JS_CHECK_RECURSION from the internal
one, or moving the appropriate stuff into friend fields.
2013-08-28 11:04:15 -07:00
Bobby Holley
4819690f50 Bug 732665 - Disable a few test on mac. r=tbsaunde,me
These tests trigger very slow infinite recursion in the status quo. With the
existing stack quotas, they always (or usually) hit the native stack limit
before the test times out, causing them to be green on tinderbox. But when
we bump the debug mac stack limit, they end up timing out. We have bugs on
file for the various issues, but let's not let them block this landing.
2013-08-28 11:04:15 -07:00
Luke Wagner
3a9c72568e Bug 900669 - OdinMonkey: store relative offsets in AsmJSModule::{Exit,ExportedFunction} (r=bbouvier) 2013-08-23 17:51:46 -05:00
Luke Wagner
de0a7992dc Bug 900669 - OdinMonkey: un-HeapPtr more AsmJSModule property names (r=terrence) 2013-08-23 17:32:47 -05:00
Luke Wagner
fd668d42f9 Bug 900669 - OdinMonkey: put memcpy-able data in 'pod' fields of AsmJSModule (r=bbouvier) 2013-08-12 18:40:52 -05:00
Boris Zbarsky
be1711a990 Bug 909645 part 4. Don't include ipdl headers in DesktopNotification.h. r=ms2ger,glandium 2013-08-28 00:14:57 -04:00
Nicholas D. Matsakis
ffcd00eaa0 Bug 898347 - Re-enable binary data now that fuzzing is complete (Part 0/6) rs=trivial 2013-08-01 09:38:21 -04:00
Nicholas D. Matsakis
25e6fe76ca Bug 898347 - Allow addenda to be added to binary blocks (Part 6/6) r=bhackett 2013-08-16 17:32:58 -04:00
Nicholas D. Matsakis
c9d3a4090f Bug 898347 - Allow TypeRepresentation objects to be attached to TI objects (Part 5/6) r=bhackett 2013-08-01 09:47:19 -04:00
Nicholas D. Matsakis
e088a95129 Bug 898347 - Generalize the newScript field on type objects into an "addendum" (Part 4/6) r=bhackett 2013-07-20 16:18:53 -04:00
Nicholas D. Matsakis
d57113c33b Bug 898347 - Convert binary data to use TypeRepresentation (Part 3/6) r=sfink 2013-08-12 21:56:24 -04:00
Nicholas D. Matsakis
d414da006b Bug 898347 - Make typed array use type repr constants (Part 2/6) r=sfink 2013-08-12 22:00:50 -04:00