Commit graph

1,057 commits

Author SHA1 Message Date
Moonchild
c07ea665ea Issue #2838 - Use ToIndex in TypedArray CTORs
Also adjusts index handling to be more accurate/spec compliant.
2025-08-08 09:28:38 +08:00
Moonchild
3f297457d8 [js] Align br_table limits with V8/Gecko. 2025-07-25 07:31:43 +08:00
Andy
abb891ff26 Issue #2106 - Follow-up: Un-prefix DOM Attribute MozUserSelect 2025-07-08 10:00:32 +08:00
Moonchild
71ba67cfbe No issue - Don't support modulo math space in ExtractLinearSum. 2025-05-27 23:18:24 +08:00
Moonchild
01ebabde5b No issue - Fix promise combinator function state. 2025-05-27 23:18:09 +08:00
Brian Smith
e8b3077d4f No Issue - Fixes for building with LLVM 19 included with FreeBSD 13.5. Fix a conflict with libc++ 19 and the old Mozilla (re)alloc macros. LLVM 18+ does not allow std::char_traits<unsigned char> so avoid it. https://bugzilla.mozilla.org/show_bug.cgi?id=1849070 Partial NSS upgrade to replace ByteString with a class. https://bugzilla.mozilla.org/show_bug.cgi?id=1851092 2025-05-14 14:20:07 +08:00
Moonchild
d7f01911b8 Issue #2731 - Fix crashes on unknown BigInt types in JIT
This only works around the unknown type crashes but doesn't actually fix
attempts at using 64-bit BigInts in Ion.
Resolves #2731
2025-04-30 22:58:27 +08:00
Brian Smith
59977aa358 No issue - Various updates for Mac buildability with XCode 16.3 2025-04-18 10:30:02 +08:00
Martok
7d3dc5a2f2 Issue #2692 - Follow-Up: fix compilation in debug 2025-03-27 09:04:21 +08:00
Moonchild
c31cda6c74 No issue - Fix incorrect js message ref in js shell. 2025-03-27 09:04:00 +08:00
Moonchild
788cc4acb2 No issue - Add range checking to SharedArrayBuffer references.
This avoids a potential crash when juggling SAB allocations.
2025-03-27 09:02:33 +08:00
Moonchild
5fd304507d Issue #2692 - Part 7: De-globalize {*}CompileOptions.
Make all users of the various *CompileOptions classes #include "js/CompileOptions.h"
so that nothing but that file has to know about these classes having a
JS_PUBLIC_API on them, that would have to be present in forward-declarations.
2025-03-13 23:03:43 +08:00
Moonchild
2e4620b5af Issue #2692 - Part 6: Don't #include js/SourceBufferHolder.h in jsapi.h.
Instead, require users to do so. This is a minor translation-unit size
improvement for anyone who never has to use SourceBufferHolder other than
by reference.
2025-03-13 23:01:46 +08:00
Moonchild
40ed1b10f4 Issue #2692 - Part 5: Split out compilation and evaluation APIs into its own header. 2025-03-13 22:59:38 +08:00
Moonchild
307621db50 Issue #2692 - Part 4: Split out off-thread compilation API to its own header. 2025-03-13 22:59:23 +08:00
Moonchild
eca185aa79 Issue #2692 - Part 3: Split out JS compiled script transcoding to its own header. 2025-03-13 22:59:08 +08:00
Moonchild
565fb4b05a Issue #2692 - Part 2: Split out JS SourceBufferHolder to its own header. 2025-03-13 22:58:50 +08:00
Moonchild
9436bfa175 Issue #2692 - Part 1: Split out JS CompileOptions to its own header. 2025-03-13 22:58:36 +08:00
Moonchild
43a2299480 Disable potentially unsafe attempts at recovering JIT operations. 2025-03-07 21:49:32 +08:00
Shadow
37dcef684a PR #2695 - Simplify block-value popping in Ion compilation 2025-03-07 21:43:39 +08:00
Moonchild
3bac3ab0b5 No issue - avoid potential underflow in StructuredClone.
Prevents potential loop-around if there's bogus internal Map data.
2025-02-18 19:10:57 +08:00
Moonchild
7163430d1a Null-check inputs for String and BigInt
Crash fix for null crash when feeding bogus strings or BigInts to
StructuredClone.
2025-02-06 09:26:42 +08:00
Moonchild
38daeffb9b Issue #2678 - Avoid QIing for NoteXPCOMRoot.
This callback is only used in very limited ways, so just require that the
caller passes in the canonical supports pointer plus the participant.

This probably won't affect performance much.
2025-01-20 21:07:45 +08:00
Moonchild
32da3f4de4 Issue #2678 - Remove the DOM proxy expando object hashtable in xpconnect. 2025-01-20 21:06:59 +08:00
Moonchild
6c6ab6de1a [js] Avoid potential crash if GC cell is no longer valid. 2024-11-29 20:50:26 +08:00
Moonchild
7f086f216c Issue #2653 - Part 3: Remove support for packaged apps from asmjscache 2024-11-21 21:02:19 +08:00
Moonchild
1878228db7 Issue #2305 Part 4: Check for failure when getting RegExp match result template
Null deref crash fix.
2024-11-13 23:24:27 +08:00
Martok
a382c2aa66 Issue #2305 Part 3c: Merge repeated loops over MatchPairs 2024-11-13 23:23:59 +08:00
Martok
6dbe2de6db Issue #2305 Part 3b: Create .indices and .indices.groups 2024-11-13 23:23:27 +08:00
Moonchild
2d8696c4fe Issue #2305 Part 3a: Pass ResultTemplateKind through CreateRegExpMatchResult
This sets up the MatchResult with the proper structure to contain indices
when necessary.
2024-11-13 23:23:00 +08:00
Moonchild
845d36d582 Issue #2305 Part 2: Support .indices in match result template objects
Before match indices, all match results had the same shape. (If there are
no named capture groups, `.groups` exists but is undefined.) Match results
with `.indices` have a distinct shape, so we need a distinct template
object. We also need a template object for `.indices` itself.
2024-11-13 23:21:30 +08:00
Moonchild
47ef1701b3 Issue #2305 Part 1: Add support for hasIndices to RegExpFlags (/d)
Adds recognition and parsing of the /d RegExp flag.
Drive-by correction: add missing "s" for dotAll to RegExpObject::toString.
Tag #1284 and #2008
2024-11-13 23:20:49 +08:00
Moonchild
8bbcb50b62 Revert "Issue #2645 - Fix type confusion for ResumeKind"
This reverts commit 12e9db2481d2b542e4f32f9b5ba63a4c72b874d5.
2024-10-31 21:46:55 +08:00
Moonchild
e665b30629 Issue #2645 - Return null if getName() PC is invalid. 2024-10-24 07:37:47 +08:00
Moonchild
19c4996881 Issue #2645 - Fix type confusion for ResumeKind
It used a mix of uint8_t and uint16_t. Set it to uint16_t.
Also don't allow non-typeset to simplify stubs used in compound opcodes.
2024-10-24 07:36:44 +08:00
Steve Fink
9a6e3bc30d [js] Disallow deserializing structured clone buffers with transferables more than once. 2024-10-03 14:35:12 +08:00
Martin Husemann
b1fd3b72cf Make NetBSD/powerpc use the powerpc atomic ops.
NetBSD/powerpc gcc does not predefine __ppc__, but only __powerpc__,
so add that to the preprocessor ifdef cascade to select atomic
ops.
2024-09-22 20:45:12 +08:00
Moonchild
efad094b02 Make inaccurate GC stats reporting non-fatal in debug builds.
Not sure why this was fatal as it only hinders being able to report memory
issues to a debugging dev :P
2024-09-22 20:44:32 +08:00
FranklinDM
97a40adc75 Issue #2251 - Expose ghost window unlinking to release builds. 2024-09-22 20:43:48 +08:00
Moonchild
2b4e577d8e No issue - Fix asserts in debug.
Hashbang parser changes allow more than one EOL which screws up these
assertions' premise. Just remove them as the preconditions are no longer
valid.
2024-09-22 20:43:23 +08:00
Brian Smith
27e66e3f82 No Issue - Fix building for Apple Silicon Mac with Xcode 16. Due to the following LLVM change labels need to be outside .cfi_start/endprocs: https://reviews.llvm.org/D155245 Solution pointed out here is to move the label above .cfi_startproc: https://github.com/libffi/libffi/issues/807 2024-09-22 20:42:53 +08:00
Nia Alarie
dd6526daae js: Support for the MPROTECT security feature (PaX and NetBSD)
On such platforms, remapping memory that was once writable
to executable is forbidden unless the initial mmap() is declared
to change in such a way using the PROT_MPROTECT macro.
2024-09-13 10:02:50 +08:00
Moonchild
ec8793a094 [js xpconnect] Fix error handling in xpc::CreateSandboxObject 2024-09-06 09:17:32 +08:00
Moonchild
8f31479968 [js] Simplify with-env handling in FetchName 2024-09-06 09:16:51 +08:00
Moonchild
7cc9551e79 Issue #2559 - Implement hashbang grammar
This adds special-case #! comment handling at the top level, treating
any hashbang line as a single-line comment in the parser.
This only applies to Eval(), module script or global context.
2024-09-06 09:15:51 +08:00
Moonchild
ef1f142cde Issue #2531 - Adjust rounding in toExponential (and kin) to round up at midpoint.
Resolves #2531
2024-07-02 22:08:14 +08:00
Moonchild
f4b6cee0cc Issue #2543 - adjust toFixed precision range according to spec update.
As pointed out in the ecma-262 spec change discussion, the behavior is
inconsistent for negative numbers because the exponential fallback assumes
all the digits will be displayed. The negative support was motivated just
by consistency, without any particular concrete use cases, and it's not
supported in other browsers, so this patch removes negatives.

Resolves #2543
2024-07-02 22:07:55 +08:00
Moonchild
40c9102d6d [js gc] Suppress GC during JSObject::swap
We already suppressed GC for part of this, but we may sweep before the pre-write
barrier is established.
The simplest and safest thing is to suppress GC for the whole method.
2024-06-13 11:30:25 +08:00
Martok
2128ac642e Issue #2519 - Use parser error reporter to report missing module exports 2024-06-04 10:16:36 +08:00
Moonchild
dc52303799 [js] Simplify tracing of arguments in MarkThisAndArguments. 2024-04-19 09:34:45 +08:00