Commit Graph

454 Commits

Author SHA1 Message Date
David Anderson
a64d9ec66a Internalize and memoize FrameInfo structures (bug 501398, r=brendan). 2009-09-29 14:33:43 -07:00
Graydon Hoare
ce8212d362 Bug 518744 - TM: mark and rewind dataAlloc, r=gal.
--HG--
extra : rebase_source : fd9eac329fc70879ad5b45a267152937fed693fb
2009-09-25 17:20:01 -07:00
L. David Baron
7e046dd800 Add assertions that fire when a JSAutoRequest, JSAutoSuspendRequest, JSAutoLocalRootScope, JSAutoTempValueRooter, JSAutoTempIdRooter, JSAutoIdArray, JSAutoEnumStateRooter, or JSAutoResolveFlags is used as a temporary. (Bug 518633) r=igor 2009-09-27 23:17:47 -07:00
Luke Wagner
a79d78e239 Fix ridiculous manual merge error. (r=graydon) 2009-09-24 12:40:13 -07:00
Luke Wagner
668372111c Bug 517973 - remove deep abort, fix blacklisting bug 2009-09-23 18:21:41 -07:00
Graydon Hoare
f0b694edca Bug 517083 - Fix performance regression, r=dvander.
--HG--
extra : rebase_source : 57e132434da8c652c41ffa889aa2002eeeb4698c
2009-09-23 15:12:58 -07:00
David Mandelin
99fb23d064 Bug 511695: Fix JS warnings in MSVC8, r=brendan 2009-09-23 11:16:30 -07:00
Graydon Hoare
ccc6ad4904 Bug 517083 - TM: introduce a temp allocator for allocations during recording and compilation, r=gal. 2009-09-22 12:36:20 -07:00
Jeff Walden
553915decf Bug 307791 - Implement ES5's Object.keys(O). r=jorendorff 2009-09-08 15:59:14 -07:00
Igor Bukanov
8da9fad53a bug 512046 - avoid maintaing a per-runtime iterator list. r=brendan 2009-09-19 13:40:43 +04:00
Jason Orendorff
b74384b5c8 Backed out changeset 3f508cfdfa36 (bug 500431) due to tinderbox orangeness 2009-09-18 08:55:01 -05:00
Jason Orendorff
b040e5c724 Bug 500431 - Encapsulate the property cache using C++ best practices - Part 2, refactoring. r=Waldo. 2009-09-17 18:24:38 -05:00
Andreas Gal
55697d8ebf Merge. 2009-09-17 15:44:30 -07:00
Andreas Gal
38b0ee0297 Backed out changeset 5f449dffdff5 (regressed tp4 private bytes, and we are not sure what the hell that means). 2009-09-17 15:40:37 -07:00
Andreas Gal
4b2f3150ab Use one single GC heap chunk, avoiding frequent mmap and malloc calls (508707, r=igor,brendan). 2009-09-15 14:29:01 -07:00
Graydon Hoare
d0efa663da Bug 503424 - Additional followup to fix more profound build-breakage, r=me. 2009-09-15 17:07:33 -07:00
Graydon Hoare
70f3343672 Bug 503424 - Followup to fix windows build breakage, r=gal.
--HG--
extra : rebase_source : 338e0b9c4b5c52fbec05074086a80076cf9f8299
2009-09-15 16:31:07 -07:00
Julian Seward
b214130415 Bug 503424 - Add built-in support for compiled-trace and -guard profiling, r=graydon. 2009-09-15 15:05:53 -07:00
Brendan Eich
eaabbf253c Bug 516075 - Move shape into JSObjectMap from JSScope, const-ipate and use LIR_ldc* to get to it (r=jorendorff). 2009-09-14 20:29:46 -04:00
Andreas Gal
72c92b196f Use a single code cache for the JIT and the RegExp compiler (515852, r=dmandelin). 2009-09-12 23:04:27 -07:00
Luke Wagner
d72c89d060 Backed out changeset b866396faae4 2009-09-10 16:44:01 -07:00
Luke Wagner
e9357856b4 Bug 506410 - bring template lovin' to hash tables (r=jorendorff) 2009-09-10 16:16:29 -07:00
Andreas Gal
bd58a78c21 Set nativeVp immediately before the native call (514999, r=mrbkap). 2009-09-09 17:38:07 -07:00
Brendan Eich
809ce29270 Cope with null tvr string refs (515000 followup, r=self). 2009-09-07 21:08:31 -07:00
Brendan Eich
7a412f40b3 JSTempValueRooter JSTVU_SINGLE over-constrains GC-thing against static allocation (515000, r=gal). 2009-09-07 00:35:27 -07:00
Gregor Wagner
2e1104b40f Cache result of Number2String (bug 513530, r=brendan). 2009-09-04 16:28:30 -07:00
Luke Wagner
5c14309cd1 Bug 511750 - factor template utilities into js::tl (r=brendan)
--HG--
extra : rebase_source : a5a396811268e71af8e56c66dab491f3081bc20a
2009-09-01 18:46:19 -07:00
Andreas Gal
030a8114b2 Backed out changeset 9b6b17a275ec (bug 506125). 2009-08-25 14:42:42 -07:00
Andreas Gal
f808b9ca98 Memory-pressure based GC scheduler (506125, r=igor,jwalden). 2009-08-24 18:24:52 -07:00
Graydon Hoare
503c025d90 Bug 511309 - Delete class Fragmento and all uses of it, r=dvander.
--HG--
extra : rebase_source : 47047589288ac0758e0068307b5a8f393db97606
2009-08-20 17:22:47 -07:00
Graydon Hoare
ebe9c3f20b Bug 511237 - Remove primary use of Fragmento from jsregexp, r=dmandelin. 2009-08-18 15:32:54 -07:00
Jason Orendorff
5418886640 Bug 511418 - static-analysis error in jsobj.cpp:4257: cannot access JS_REQUIRES_STACK variable JSContext::fp and another trivial error in jstracer.cpp. r=gal.
--HG--
extra : rebase_source : 9ee93b34fbf656437932a4f3c8b373a11911cffd
2009-08-20 14:13:21 -05:00
Andreas Gal
9a78cb5d5f Use thread-local RNG for Math.random() (511328, r=shaver,waldo). 2009-08-19 15:23:54 -07:00
Andreas Gal
c438241025 Backout bug 506125. 2009-08-18 17:30:05 -07:00
Andreas Gal
57285268f7 Schedule GC based on process memory pressure (506125, r=igor,waldo). 2009-08-18 16:38:02 -07:00
Andreas Gal
ebc6f01967 Add an API to notify the JS engine that we are about to destroy the runtime (511252, r=brendan). 2009-08-18 16:31:20 -07:00
Jason Orendorff
b3aba1ee4c Delete unused JSBUILTIN_ constants. No bug. r=gal. 2009-08-17 17:36:36 -05:00
Andreas Gal
5d82186f20 Don't flush the code cache every time we GC (506117, r=brendan). 2009-08-17 14:50:57 -07:00
Jason Orendorff
8384b1e239 Bug 509098 - Remove JS_HAS_LVALUE_RETURN support. r=brendan. 2009-08-13 16:18:33 -05:00
Andreas Gal
fee752ae24 Cleaning code cache flush handling (510136, r=jorendorff). 2009-08-13 08:35:35 -07:00
Graydon Hoare
b8154cc364 Bug 504462 - Merge tamarin's CodeAlloc into tracemonkey, r=gal. 2009-07-15 16:50:01 -07:00
Andreas Gal
9c938dd547 Backout bug 506125 due to talos failures. 2009-08-04 18:42:53 -07:00
Andreas Gal
06ad033ff2 Memory-pressure based GC scheduler (506125, r=igor). 2009-08-04 14:58:21 -07:00
Jason Orendorff
2df7b71611 Bug 507665 - Avoid imacros for JSOP_GETELEM and JSOP_CALLELEM. r=gal. 2009-08-03 11:38:28 -05:00
Graydon Hoare
2e92fcbbe3 Backed out changeset d0be9284ea67 2009-07-30 14:28:26 -07:00
Graydon Hoare
caf3859643 Bug 504462 - Merge tamarin's CodeAlloc into tracemonkey, r=gal. 2009-07-15 16:50:01 -07:00
Andreas Gal
e7f2714ea1 When finalizing, deallocate memory in a separate thread (505612, r=bent,brendan,waldo). 2009-07-27 21:10:12 -07:00
Jason Orendorff
a22e8a6032 Bug 505932 - Shape regeneration does not touch most empty scopes. r=brendan. 2009-07-27 16:29:02 -05:00
Igor Bukanov
bfbac69d3f bug 506243 - eliminate GC thread-local lists pools. r=brendan 2009-07-26 18:24:39 +02:00
Brendan Eich
b4a10de944 Bug 488731 - Avoid shape regeneration and property cache purge during the GC (r=mrbkap). 2009-07-24 06:55:28 -07:00
Robert Sayre
6968ea2205 Merge backout. 2009-07-24 00:02:02 -04:00
Robert Sayre
9e0ade7287 Backed out changeset b3d459d23452 2009-07-24 00:01:37 -04:00
Robert Sayre
3722b48ae2 Merge backout. 2009-07-23 22:08:05 -04:00
Robert Sayre
d71c6818a0 Backed out changeset 047b9102dddb 2009-07-23 22:06:31 -04:00
Brendan Eich
cf4a2c7921 Avoid shape regeneration and property cache purging from the GC unless the shape generator overflows (488731, r=jorendorff). 2009-07-23 17:59:49 -07:00
Andreas Gal
eb711234db Use a background thread to free() memory while sweeping (505612, r=brendan,bent). 2009-07-23 15:12:34 -07:00
Andreas Gal
241532b53a Inline malloc accounting functions (504219, r=jwalden). 2009-07-16 18:36:19 -07:00
Andreas Gal
88cf1d87c6 Avoid integer math for GC trigger factor calculation in allocation path, take 2 (503463, r=dmandelin). 2009-07-14 17:03:20 -07:00
Peter Van der Beken
cf914cb140 Backed out changeset 2073d5aae8b6 (Avoid integer math for GC trigger factor calculation in allocation path (bug 503463)). 2009-07-14 11:49:15 +02:00
Andreas Gal
0c6942b658 Avoid integer math for GC trigger factor calculation in allocation path (503463, r=dmandelin). 2009-07-13 13:24:25 -07:00
Brendan Eich
f60779aef5 Bug 503343 - Change JSScope macros to methods. r=jorendorff 2009-07-09 15:27:21 -05:00
Robert Sayre
878a89a073 Merge mozilla-central to tracemonkey. 2009-07-06 12:20:54 -04:00
Luke Wagner
145fbf91c2 Bug 200505 - Optimization of jsref array_join_sub() function. r=waldo 2009-06-30 20:19:42 -04:00
Robert Sayre
3f97d5e6f5 Backed out changeset b2256abf53c0 2009-06-30 12:14:09 -07:00
Luke Wagner
a6311cf8b5 Bug 200505 - Optimization of jsref array_join_sub() function. r=waldo 2009-06-30 11:29:43 -07:00
Jason Orendorff
52fe5a97ba Bug 497618 - Change JSString macros to methods. r=Waldo. 2009-06-10 20:29:44 -05:00
Andreas Gal
ec254bec6f Support tracing instanceof operator (498351, r=jorendorff). 2009-06-15 16:44:30 +01:00
Johnny Stenback
a511964116 Fixing bug 442399. Remove LiveConnect from the tree. r=joshmoz@gmail.com, bclary@bclary.com, sr=brendan@mozilla.org 2009-06-30 15:55:16 -07:00
Brendan Eich
0416e1e679 Bug 492355 - Suspected Txul regression from JS engine changes (r=igor). 2009-05-13 17:11:23 -07:00
Graydon Hoare
2a50d488f5 Bug 489615 - TM: Store recording attempts in a long-lived hashtable rather than fragments, r=brendan. 2009-05-07 15:49:44 -07:00
jorendorff
93c6d226d3 Record all calls to native functions (487134, r=gal, brendan). 2009-05-05 14:26:06 -07:00
Igor Bukanov
35116060a2 bug 489501 - consolidating
js_SetProto must mutate the prototype chain. r=brendan
2009-05-01 13:48:27 -07:00
Graydon Hoare
bb4177dccd Bug 490044 - TM: Add deep-bailing write barrier to global shape change code, r=brendan. 2009-04-24 17:47:32 -07:00
Andreas Gal
a51f22dbdf Need an API exposed to control code cache size (474497, r=bent,brendan, sr=mrbkap). 2009-04-22 16:52:59 -07:00
Jason Orendorff
4ea515fa8b Bug 488874 - Change tm.onTrace to tm.tracecx. r=brendan.
--HG--
extra : rebase_source : bf683bee5e518cf5bfea8113c17685b8301d72c9
2009-04-22 15:07:55 -05:00
jorendorff
e1b8077e65 Leaving outermost request should js_LeaveTrace (480301, r=brendan). 2009-04-20 18:22:00 -07:00
Igor Bukanov
4ad2e764fa bug 488414 - js_GenerateShape just schedules, not run, the GC. r=brendan,gal 2009-04-20 12:10:29 +02:00
Igor Bukanov
fd49b19773 Backed out changeset f4662701526b (bug 488414) to fix !JS_THREADSAFE compilation errors 2009-04-20 11:59:36 +02:00
Igor Bukanov
56547f8087 bug 488414 - js_GenerateShape just schedules, not run, the GC. r=brendan,gal 2009-04-20 10:17:28 +02:00
Igor Bukanov
907592f765 bug 477627 - avoiding deadlocks in ClaimTitle. r=brendan 2009-04-16 00:01:24 +02:00
Jason Orendorff
4dd9ad09ef Bug 487845 - TM: After deep-bailing, we can lirbuf->rewind() and then return to a dead code page. r=gal.
--HG--
extra : rebase_source : b7de71b526d45c7caed1982fba80c65763a0a9c6
2009-04-15 09:47:02 -05:00
Andreas Gal
c11974bb60 Backed out changeset 4c157cfe2289 (bug 487845). 2009-04-14 21:28:40 -07:00
Jason Orendorff
0616f91d10 Bug 487845 - TM: After deep-bailing, we can lirbuf->rewind() and then return to a dead code page. r=gal. 2009-04-14 08:45:37 -05:00
Blake Kaplan
1c5e6afeac Bug 487684 - Store the last trace PC to pass into the decompiler. r=igor 2009-04-10 14:06:28 -07:00
Jason Orendorff
b5d4ba583e Bug 487676 - Nesting deep-aborting trace calls don't work. r=gal.
--HG--
extra : rebase_source : 436164485bd5e981e5ae8997eed2c371f3ecd77b
2009-04-09 18:07:00 -05:00
Brendan Eich
3f4a58e4f0 Bug 487271 - Crash [@ js_Invoke ], and missing google-maps background, at padmapper.com (r=mrbkap). 2009-04-08 13:14:02 -07:00
Jason Orendorff
a613c0d579 Backout changeset 143e997c858e (bug 484693) because it caused crashes on Mac tinderboxen. 2009-04-07 17:07:11 -05:00
Graydon Hoare
163d74068e Bug 484693 - Remove uses of alloca, r=gal. 2009-04-07 13:47:36 -07:00
Brendan Eich
962b187fe5 upvar2, aka the big one take 2 (452598, r=mrbkap). 2009-04-05 21:17:22 -07:00
Andreas Gal
3c19b8b7fc Merge. 2009-04-04 10:15:04 -07:00
Andreas Gal
706120c0fb Backed out changeset 972c44aa9d1f (bug 452598). 2009-04-04 10:14:52 -07:00
Brendan Eich
7485baa00c Merge. 2009-04-04 10:07:44 +01:00
Brendan Eich
5f4d9ff474 upvar2, aka the big one (452598, r=mrbkap). 2009-04-04 10:05:49 +01:00
Andreas Gal
8787072e96 Backed out changeset b512be855093 (bug 484693). See bug for details. 2009-04-04 01:06:04 -07:00
Graydon Hoare
a2d4d5a779 Bug 484693 - Remove uses of alloca, r=gal. 2009-04-03 16:30:46 -07:00
Jim Blandy
02252a653d Bug 482743: Fix up bytecode execution tracing. Allow tracing to file. r=igor
js_TraceOpcode: Remember the last bytecode we traced explicitly,
instead of subtracting 'len' from regs.pc, which isn't reliable.
Decline to trace values in script prologues (between 'code' and
'main').  Decline to walk off the bottom of the stack when the 'last
bytecode' is misleading.  Flush the stream after each bytecode.

Use the TRACE_OPCODE macro in both threaded and non-threaded
interpreters.  Take care to make threaded and non-threaded
interpreters produce the same traces.

In the shell's 'tracing' function, use JS_ValueToBoolean to recognize
all sorts of booleans, and treat a string as the name of a file to
write the trace to.
2009-04-01 08:50:57 -07:00
Andreas Gal
78ced670dd specialize code for a specific global object, not just a global object shape (480905, r=graydon). 2009-03-31 11:49:44 -07:00
Igor Bukanov
a8094a7903 bug 437325 - JSThread is no longer shared between JSRuntime instances. r=brendan 2009-03-24 13:07:35 +01:00
Igor Bukanov
f28d80b499 Backed out changeset e117c22cc1d1 - the landed patch for bug 437325 has a shutdown leak. 2009-03-22 10:15:27 +01:00
Igor Bukanov
a61176553e bug 437325 - JSThread is no longer shared between JSRuntime instances. r=brendan 2009-03-22 10:07:14 +01:00
Andreas Gal
4017bf1be3 str_match should use js_GetCurrentBytecodePC (484308, r=mrbkap). 2009-03-20 17:07:30 -07:00
Jim Blandy
debdb32ef3 Bug 480132: Clone lexical blocks only when needed. r=igor
Terminology:

A "script block" is an object of class Block allocated by the byte
compiler and associated with a script.  Script blocks are never
modified, and may be used as a prototype for a "closure block":

A "closure block" is an object of class Block that holds variables
that have been closed over (although we actually leave the variables
on the stack until we leave their dynamic scope).  A closure block is
a clone of a script block (its prototype is a script block).

Adjust the meanings of fp->blockChain and fp->scopeChain:

  fp->blockChain is always the innermost script block in whose static
  scope we're executing.

  fp->scopeChain is the current scope chain, including 'call' objects
  and closure blocks for those function calls and blocks in whose
  static scope we are currently executing, and 'with' objects for with
  statements; the chain is typically terminated by a global object.
  However, as an optimization, the young end of the chain omits block
  objects we have not yet needed to clone.

Closures need fully reified scope chains, so have js_GetScopeChain
reify any closure blocks missing from the young end of fp->scopeChain
by cloning script blocks as needed from fp->blockChain.  Thus, if we
never actually close over a particular block, we never place a closure
block for it on fp->scopeChain.

Have JSOP_ENTERBLOCK and JSOP_LEAVEBLOCK always keep fp->blockChain
current.  When JSOP_LEAVEBLOCK pops a block from fp->blockChain that
has been cloned on fp->scopeChain, pop fp->scopeChain as well.

Remove the JSFRAME_POP_BLOCKS flag, as it is no longer needed.

Ensure that the JIT won't have to create closure blocks or call
js_PutBlockObject; it can't handle those things yet.  Note our current
script block when we begin recording.  Abort recording if we leave
that block; we can't tell in advance whether it will need to be "put"
in future trace invocations.  Leave trace if we call js_GetScopeChain
while in the static scope of lexical blocks.  Remove JIT tests based
on JSFRAME_POP_BLOCKS.

Verify that generators capture the correct value for blockChain.

Add a constructor to JSAutoTempValueRooter for rooting JSObject
pointers.
2009-03-16 09:55:06 -07:00
Jim Blandy
74847db9cd Back out 480132: orange on Linux (http://tinderbox.mozilla.org/showlog.cgi?log=TraceMonkey/1237017748.1237029269.2403.gz&fulltext=1) 2009-03-14 08:19:58 -07:00
Jim Blandy
bed3611385 Bug 480132: Clone lexical blocks only when needed. r=igor
Terminology:

A "script block" is an object of class Block allocated by the byte
compiler and associated with a script.  Script blocks are never
modified, and may be used as a prototype for a "closure block":

A "closure block" is an object of class Block that holds variables
that have been closed over (although we actually leave the variables
on the stack until we leave their dynamic scope).  A closure block is
a clone of a script block (its prototype is a script block).

Adjust the meanings of fp->blockChain and fp->scopeChain:

  fp->blockChain is always the innermost script block in whose static
  scope we're executing.

  fp->scopeChain is the current scope chain, including 'call' objects
  and closure blocks for those function calls and blocks in whose
  static scope we are currently executing, and 'with' objects for with
  statements; the chain is typically terminated by a global object.
  However, as an optimization, the young end of the chain omits block
  objects we have not yet needed to clone.

Closures need fully reified scope chains, so have js_GetScopeChain
reify any closure blocks missing from the young end of fp->scopeChain
by cloning script blocks as needed from fp->blockChain.  Thus, if we
never actually close over a particular block, we never place a closure
block for it on fp->scopeChain.

Have JSOP_ENTERBLOCK and JSOP_LEAVEBLOCK always keep fp->blockChain
current.  When JSOP_LEAVEBLOCK pops a block from fp->blockChain that
has been cloned on fp->scopeChain, pop fp->scopeChain as well.

Remove the JSFRAME_POP_BLOCKS flag, as it is no longer needed.

Ensure that the JIT won't have to create closure blocks or call
js_PutBlockObject; it can't handle those things yet.  Note our current
script block when we begin recording.  Abort recording if we leave
that block; we can't tell in advance whether it will need to be "put"
in future trace invocations.  Leave trace if we call js_GetScopeChain
while in the static scope of lexical blocks.  Remove JIT tests based
on JSFRAME_POP_BLOCKS.

Verify that generators capture the correct value for blockChain.

Add a constructor to JSAutoTempValueRooter for rooting JSObject
pointers.
2009-03-14 00:58:27 -07:00
Jim Blandy
d72dff994b Back out 480132 fix; static analysis red 2009-03-14 00:57:54 -07:00
Jim Blandy
4778b96a85 Bug 480132: Clone lexical blocks only when needed. r=igor
Terminology:

A "script block" is an object of class Block allocated by the byte
compiler and associated with a script.  Script blocks are never
modified, and may be used as a prototype for a "closure block":

A "closure block" is an object of class Block that holds variables
that have been closed over (although we actually leave the variables
on the stack until we leave their dynamic scope).  A closure block is
a clone of a script block (its prototype is a script block).

Adjust the meanings of fp->blockChain and fp->scopeChain:

  fp->blockChain is always the innermost script block in whose static
  scope we're executing.

  fp->scopeChain is the current scope chain, including 'call' objects
  and closure blocks for those function calls and blocks in whose
  static scope we are currently executing, and 'with' objects for with
  statements; the chain is typically terminated by a global object.
  However, as an optimization, the young end of the chain omits block
  objects we have not yet needed to clone.

Closures need fully reified scope chains, so have js_GetScopeChain
reify any closure blocks missing from the young end of fp->scopeChain
by cloning script blocks as needed from fp->blockChain.  Thus, if we
never actually close over a particular block, we never place a closure
block for it on fp->scopeChain.

Have JSOP_ENTERBLOCK and JSOP_LEAVEBLOCK always keep fp->blockChain
current.  When JSOP_LEAVEBLOCK pops a block from fp->blockChain that
has been cloned on fp->scopeChain, pop fp->scopeChain as well.

Remove the JSFRAME_POP_BLOCKS flag, as it is no longer needed.

Ensure that the JIT won't have to create closure blocks or call
js_PutBlockObject; it can't handle those things yet.  Note our current
script block when we begin recording.  Abort recording if we leave
that block; we can't tell in advance whether it will need to be "put"
in future trace invocations.  Abort recording if we call
js_GetScopeChain while in the static scope of lexical blocks.  Remove
JIT tests based on JSFRAME_POP_BLOCKS.

Verify that generators capture the correct value for blockChain.

Add a constructor to JSAutoTempValueRooter for rooting JSObject
pointers.
2009-03-13 23:43:32 -07:00
Robert Sayre
8e061e35da Backout patch for Bug 480132. Debug only nit fix busts static analysis compile. 2009-03-14 02:20:45 -04:00
Jim Blandy
40e8a7d560 Bug 480132. SpiderMonkey clones too many blocks into the heap. r=igor 2009-03-14 02:09:28 -04:00
Igor Bukanov
4b95050fe2 Backed out changeset 57de81309176 - bug 437325 - due to mochitest leaks on tinderbox 2009-03-13 18:36:46 +01:00
Igor Bukanov
3de011b9cd bug 437325 - JSThread is no longer shared between runtimes. r=brendan 2009-03-13 12:36:21 +01:00
Igor Bukanov
75640ecce1 Backed out changeset 4159ebdfe31e to fix some typos in the patch 2009-03-13 12:21:04 +01:00
Igor Bukanov
8eeedd9647 bug 437325 - JSThread is not shared between runtimes. r=brendan 2009-03-13 12:10:34 +01:00
Igor Bukanov
4ccdead266 bug 480700 - no more static asserts in headers. r=brendan 2009-03-05 12:12:50 +01:00
Andreas Gal
8c3223aab5 Browser spuriously sets runtime->anyArrayPrototypeHasElement and makes perf bad (481251, r=mrbkap). 2009-03-03 18:04:15 -08:00
Jason Orendorff
2f464f3348 Bug 479826 - Remove cx->pcHint. r=gal.
--HG--
extra : rebase_source : caf9fef929fe161fad2f6a98b53cb3d00e43f31f
2009-02-23 17:29:40 -06:00
Benjamin Smedberg ext:(%2C%20Jason%20Orendorff%20%3Cjorendorff%40mozilla.com%3E)
d40f9751c8 Bug 476643 - JS_REQUIRES_STACK errors in nsXPCWrappedJSClass::CallMethod. Mozilla code uses some JS-internal APIs and needs to bail off trace before doing so. This shouldn't affect performance now becuase XPConnect methods are not traceable yet. r=mrbkap, sr=mrbkap. 2009-02-18 12:47:16 -06:00
Jason Orendorff
c9b0918474 Bug 477142 - _FAIL builtins need to be GC-safe. r=brendan.
--HG--
extra : rebase_source : 10515f7d1f5a85c1965c812f55d91d3d0f49a28e
2009-02-11 13:33:17 -06:00
Andreas Gal
93eb39269d Substitute operation counting with a watchdog thread (477187, 3nd attempt, r=brendan/mrbkap/jst, sr=brendan/jst). 2009-02-10 14:07:01 -08:00
Andreas Gal
474d0d308f Backout bug 477187. 2009-02-10 09:54:14 -08:00
Andreas Gal
f796bcd60a Substitute operation counting with a watchdog thread (477187, 2nd attempt, r=brendan/mrbkap/jst, sr=brendan/jst). 2009-02-10 03:45:36 -08:00
Andreas Gal
7ea402f38c Merge. 2009-02-09 22:54:41 -08:00
Andreas Gal
2a20c2a6de Backing out bug 477187. Breaks xpcshell. 2009-02-09 22:54:05 -08:00
Andreas Gal
2e46d80547 Merge. 2009-02-09 18:21:15 -08:00
Andreas Gal
cbfbbd3cac Substitute operation counting with a watchdog thread (477187, r=brendan/mrbkap/jst, sr=brendan/jst). 2009-02-09 18:20:50 -08:00
Jeff Walden
31fdce5e81 Bug 467747 - JS version/option unsyncing results in JS modules not being loaded with the very latest JS version, resulting in syntax errors when loading modules that use new JS syntax. r=brendan 2009-02-07 23:23:01 -08:00
Igor Bukanov
61cd533870 bug 476934 - JS_(Set|Clear)ContextThread() must wait bfor the GC. r=brendan 2009-02-07 05:05:32 +01:00
Jason Orendorff
1a135641c5 Bug 462042 - TM: Allow GC with traced machine code on stack. r=brendan. 2009-02-03 18:25:18 -06:00
Jason Orendorff
e4c1fb1cfd Bug 462027 - Bail off trace when reentering interpreter. r=gal. 2009-02-03 18:25:12 -06:00
Jason Orendorff
4e3f2cfd40 Revert to 07be1f190a3d. Revision 5f5c1cd63641 should not have been pushed. 2009-02-03 12:38:44 -06:00
Jason Orendorff
df4d8addef Bug 462042 - TM: Allow GC with traced machine code on stack. r=brendan. 2009-02-02 20:49:53 -06:00
Jason Orendorff
f6ffd1d380 [mq]: bug-462027-v9 2009-02-02 20:42:46 -06:00
Andreas Gal
68e8901a4a Bake the value of fp->imacpc into the trace and report it at recording time when detecting inside JSOP_GETELEM (follow-up for 476238, r=brendan,jorendorff). 2009-02-02 18:21:28 -08:00
Andreas Gal
3682914dfd Leave a hint for GetProperty in the context so it can figure out the current bytecode location without de-optimizing (476238, r=jorendorff). 2009-02-02 17:25:59 -08:00
Jason Orendorff
37f2985bdf Bug 476177 - TM: cx->stackPool must not be accessed on trace. r=brendan.
--HG--
extra : rebase_source : 746d61b54aa38cbe18fe31bce4efcbbad4629678
2009-01-30 17:40:05 -06:00
Graydon Hoare
5219181495 Bug 475474 - Maintain globalSlots per global, not just one per JIT instance, r=danderson. 2009-01-29 15:59:52 -08:00
Jason Orendorff
2ba2f5d3ed Bug 475761 - TM: js_Any_GetProp and friends can reenter. r=brendan. Note that this patch alone does not fix the bug. The rest of the fix comes in bug 462027. 2009-01-29 00:13:03 -06:00
Jason Orendorff
6142999fde Bug 468782 - TM: js_FastValueToIterator and js_FastCallIteratorNext can reenter (relanding with a bug fix). r=brendan. Note that this changeset alone does not fix the bug; an upcoming patch in bug 462027 completes the fix. 2009-01-28 09:24:35 -06:00
Andreas Gal
3aa2396acf Merge. 2009-01-27 16:42:29 -08:00
Andreas Gal
3361accf0e Merge. 2009-01-27 16:42:10 -08:00
Andreas Gal
1bbc35daad Backed out changeset d50d3681b94e (attempted re-landing of 474771). 2009-01-27 16:41:43 -08:00
Brendan Eich
60770fa5b5 Bug 454184 - Implement eval caching (r=mrbkap). 2009-01-27 16:40:40 -08:00
Brendan Eich
914f4ef313 Back out due to Mac Tp orange (454184). 2009-01-27 15:52:10 -08:00
Brendan Eich
335791e844 Bug 454184 - Implement eval caching (r=mrbkap). 2009-01-27 13:53:31 -08:00
Andrei Saprykin
7b91b30052 bug 474801 - Checking for MaybeGC conditions when allocating GC things in JS shell. r=igor 2009-01-27 18:21:51 +01:00
Igor Bukanov
65cda4b280 Backed out changeset 39b1c9f21064 - the patch again has triggered the crashtest timeout. 2009-01-27 15:11:47 +01:00
Andreas Gal
f17d7cf056 Set on-trace flag only during trace execution, not recording (474771, r=brendan, patch has failed before, please back out at the earliest sign of trouble). 2009-01-27 01:55:31 -08:00
Andreas Gal
868cf7c72c Backed out changeset ece63b96379b 2009-01-27 00:10:38 -08:00
Jason Orendorff
c0cc74fa99 Bug 468782 - TM: js_FastValueToIterator and js_FastCallIteratorNext can reenter. r=brendan. Note that this changeset alone does not fix the bug; an upcoming patch in bug 462027 completes the fix. 2009-01-27 00:19:58 -06:00
Graydon Hoare
170e5b5921 Bug 473688 - provide a 2-level hash map (global, pc) -> tree, r=gal. 2009-01-26 19:53:19 -08:00
Robert Sayre
b5cd4ca3dc Backout c0b2c82a524e. 2009-01-25 22:36:46 -08:00
Andreas Gal
e481ee3eb8 Only set onTrace flag while running native code, not when recording (474771, r=brendan). 2009-01-25 20:24:25 -08:00
Brendan Eich
c13882a292 Bug 470187 - TM: "Assertion failure: entry->kpc == (jsbytecode*) atoms[index]" with valueOf, regexp (r=jorendorff). 2009-01-23 14:33:42 -08:00
Andreas Gal
184fc63cd1 Backout bug 468782. Confirmed to leak. 2009-01-25 16:38:31 -08:00
Jason Orendorff
c9961fe054 Bug 468782 - TM: js_FastValueToIterator and js_FastCallIteratorNext can reenter. r=brendan. Note that this changeset alone does not fix the bug; an upcoming patch in bug 462027 completes the fix. 2009-01-22 16:39:26 -06:00
Andreas Gal
1d277bc633 Backout all patches since last mozilla-central merge (Thu Jan 22 19:14:02 2009 -500 by sayrer). 2009-01-25 02:58:22 -08:00
Andreas Gal
1baff988de Only set onTrace flag while running native code, not when recording (474771, r=brendan). 2009-01-23 20:22:23 -08:00
Andreas Gal
1b154f72a6 Merge. 2009-01-23 15:15:34 -08:00
Andreas Gal
920ecc2113 Backed out changeset 17663da1b840 (bug 462027). 2009-01-23 15:14:46 -08:00
Brendan Eich
94412bc214 Bug 470187 - TM: "Assertion failure: entry->kpc == (jsbytecode*) atoms[index]" with valueOf, regexp (r=jorendorff). 2009-01-23 14:33:42 -08:00
Jason Orendorff
9ff9140ae0 Bug 468782 - TM: js_FastValueToIterator and js_FastCallIteratorNext can reenter. r=brendan. Note that this changeset alone does not fix the bug; an upcoming patch in bug 462027 completes the fix.
--HG--
extra : rebase_source : 585d590aace159914629a786489b0883d23cf2a8
2009-01-22 16:39:26 -06:00
David Anderson
b3ff178462 Specialize trees to global types, so global type instability does not flush the cache (bug 469044, r=gal,brendan). 2009-01-22 01:45:19 -05:00
Igor Bukanov
348de1aa60 Backed out changeset a0e1d4a2404f - the patch for the bug 474586 causes timeouts in crash tests on Mac and Windows. 2009-01-21 20:22:42 +01:00
Andrei Saprykin
375373bb9b bug 453432 - Checking for MaybeGC conditions when allocating GC things 2009-01-21 16:47:17 +01:00
Igor Bukanov
0d085b6a80 Backed out changeset e74857ea8248 - this caused unit test failures on Mac 2009-01-20 17:11:09 +01:00
Andrei Saprykin
63a490ceb1 bug 453432 - Checking for MaybeGC conditions when allocating GC things 2009-01-20 15:50:39 +01:00
Blake Kaplan
390a28ac59 Bug 473709 - Protect |str| across the call to js_NewRegExp. r=jwalden 2009-01-15 17:54:05 -08:00
Igor Bukanov
c032867e33 bug 472702 - using watchdog thread in js shell to trigger operation callback 2009-01-14 18:23:51 +01:00
Benjamin Smedberg
2e26c94375 Bug 472461 part 2 - jsgc crashes with !JS_TRACER, r=jorendorff for immediate bustage fix, and I'll get ex-post-facto review from Igor 2009-01-08 10:44:16 -05:00
Jason Orendorff
be00149c86 Bug 462021 - TM: Make JSStackFrame reconstitution infallible (part 4 of 4, rename recoveryDoublePool to reservedDoublePool, r=brendan)
--HG--
extra : rebase_source : d865549cb727aada20fbaa78286de37a97b85bca
2009-01-06 13:39:14 -06:00
Jason Orendorff
b20c028359 Bug 462021 - TM: Make JSStackFrame reconstitution infallible (part 3 of 4, call objects, r=brendan)
--HG--
extra : rebase_source : 14a82b2b08c8b7dbae5e72623685b7c3947df603
2008-12-11 17:53:05 -06:00
Andreas Gal
135f143d10 Backed out changeset adbe8e4b21dc due to tinderbox failures/timeouts (453157). 2009-01-05 16:22:42 -08:00
Andrei Saprykin
9f7c96f36f bug 453157 - watchdog thread as an alternative to operation count. r=myself,mrbkap 2009-01-05 22:09:23 +01:00
Andreas Gal
a59109830e Use a single lirbuf for the tracer and rewind lirbuf during GC (471821, r=danderson). 2009-01-02 10:55:02 -08:00
David Mandelin
90baccb99f Bug 471602 part 1: make regexp compiler use one shared LIR buffer, r=gal 2008-12-31 16:41:33 -08:00
Igor Bukanov
81c3a2b632 Backed out changeset 2d5e6b1c7254 - busted the WINNT 5.1 talos tracemonkey. 2008-12-23 19:24:09 +01:00
Andrei Saprykin
11dac3f234 bug 453157 - watchdog thread as an alternative to operation count. r=igor,mrbkap 2008-12-23 16:12:11 +01:00
Igor Bukanov
136838d60b Backed out changeset 7184e014cd05 - the patch for bug 453157 bursted tgfx test on Windows. 2008-12-21 21:49:06 +01:00
Andrei Saprykin
c68c632137 bug 453157 - watchdog thread as an alternative to operation count 2008-12-21 17:16:09 +01:00
Igor Bukanov
1a1341cdb4 bug 465032 - operationCount as the first field in JSContext. r=brendan 2008-12-18 16:24:34 +01:00
Jason Orendorff
b7fb9bc849 Bug 460865 - Read barrier for cx->fp. r=mrbkap, r=dmandelin.
--HG--
extra : rebase_source : 19963188b2f9f96336ce6ca28dbaefccf3a639b7
2008-12-09 10:38:32 -06:00
David Mandelin
37e098bf53 Bug 464828 fixed: make sure js_ThreadDestructorCB is called on main thread 2008-12-02 13:12:43 -08:00
David Anderson
321ce0dc05 Deep abort recorders outer recorders if we need to flush the JIT cache early (bug 463829, r=brendan,gal). 2008-11-15 18:54:24 -06:00
Andreas Gal
533b7866b1 Add back typedef for JSTraceMonitor (compilation fix for legacy C use of JS engine includes, no bug). 2008-11-07 16:58:36 -08:00
Andreas Gal
e609f0bb95 Cleanup GuardRecord, SideExit, and InterpStruct and extract VM-dependant fields (463313, r=danderson). 2008-11-07 15:23:43 -08:00
David Mandelin
c92c2000cc Bug 461050 WIP: compiling simple regexps, r=danderson,agal 2008-11-04 14:51:51 -08:00
Andreas Gal
25404443ff GC no longer flushes the JIT cache. Instead just make sure the shape of the global object will mismatch next time we try to record or execute code, which in turn will force a flush of the code cache (458288, r=brendan/danderson). 2008-10-28 16:22:08 -07:00
David Anderson
bf1a78f4fe Implemented multiple type specialized trees per entry point (bug 450833, r=gal) 2008-10-27 20:42:49 -07:00
David Anderson
590d3d2443 Fixed crash from JIT cache flushes when js_Interpret was holding TraceRecorders (bug 458288, r=brendan). 2008-10-24 10:47:59 -07:00
Blake Kaplan
76e8be0e68 Bug 457310 - Don't potentially run code with an exception still on cx. r+sr=jst 2008-10-13 15:53:33 -07:00
Blake Kaplan
54cdbe81ec Merge mozilla-central -> tracemonkey 2008-10-13 11:05:05 -07:00
Igor Bukanov
db18f808c9 Bug 408539 - Storing XPCContext inside JSContext 2008-10-11 19:35:39 +02:00
Brendan Eich
19e9638dc3 Bug 457920 - regexp-dna.js and generality want JSOP_GETELEM(dense array, "0") (r=mrbkap/gal). 2008-10-06 22:30:36 -07:00
Andreas Gal
83ea7eb834 Merge. 2008-09-30 16:35:52 -07:00
igor
19701dd91b When replenishing the recovery double pool, detect if a GC happens (follow up work for 456826, r=gal). 2008-09-26 14:16:11 -07:00
Andreas Gal
7bb0a663e6 Use a pre-allocated pool of doubles to make sure we can safely recover in case of OOM or out of doubles (456826, r=brendan). 2008-09-25 19:11:48 -07:00
Jason Orendorff
c35bfc3682 Bug 389034 - JS_SetProperty() ends up resolving w/o JSRESOLVE_ASSIGNING (r=brendan, sr=bzbarsky) 2008-09-25 11:13:31 -05:00
Igor Bukanov
82f9c64e29 bug 378918 - scalable thread-local GC free lists 2008-09-25 15:27:59 +02:00
Robin Bate Boerop
2082181641 Bug 412866 - Reuse regexp arena, original patch by Robin Bate Boerop <moz@shorestreet.com>, refreshed by Ryan VanderMuelen <ryanvm@gmail.com>, r=crowder 2008-09-12 15:11:48 -07:00
Arpad Borsos
9b6f558fee Bug 398946 - Remove JS_STATIC_DLL_CALLBACK and JS_DLL_CALLBACK from the tree; r=(benjamin + bent.mozilla) 2008-09-07 00:21:43 +02:00
Ben Turner
1e5c70827f Bug 451729 - " Allow runtime's security callbacks to be overridden by a context". r=brendan. 2008-09-05 16:24:53 -07:00