instance_id stringlengths 11 13 | project stringclasses 2
values | bug_id stringlengths 7 9 | image_name stringlengths 27 29 | work_dir stringclasses 2
values | verification_binary stringclasses 7
values | command_options stringlengths 0 165 | target_source_files listlengths 1 10 | target_subdir listlengths 1 3 | target_vulnerability_type stringclasses 24
values | error_type stringclasses 4
values | description stringlengths 103 3.13k | vrp stringclasses 17
values | fix_patches listlengths 1 6 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
v8__454927471 | v8 | 454927471 | hwiwonlee/v8.x86_64:454927471 | /src/v8 | out/x64.asan/d8 | --allow-natives-syntax --sandbox-testing --disable-in-process-stack-traces | [
"src/maglev/maglev-ir.cc",
"src/builtins/js-trampoline-assembler.cc"
] | [
"src/maglev",
"src/builtins"
] | Stack corruption | SANDBOX_VIOLATION | Maglev can directly call a disabled builtin with a stale dispatch handle, producing stack imbalance and sandbox violation. | 22000 | [
"From 40d2c7e1787000809f7c4268fe2a95f8a62a198f Mon Sep 17 00:00:00 2001\nFrom: Krishna Ravishankar <krishna.ravi732@gmail.com>\nDate: Fri, 24 Oct 2025 21:59:23 -0400\nSubject: [PATCH] [sandbox] SBXCHECK builtin is enabled in\n CallKnownJSFunction::GenerateCode\n\n---\n src/maglev/maglev-ir.cc | 3 ++-\n 1 file chang... |
v8__457989902 | v8 | 457989902 | hwiwonlee/v8.x86_64:457989902 | /src/v8 | out/x64.asan/d8 | --allow-natives-syntax | [
"src/maglev/maglev-graph-builder.cc"
] | [
"src/maglev"
] | Hole leak | DCHECK | A hole leak vulnerability exists in the Maglev compiler's inlined Array constructor for double elements kinds, introduced in commit 006fd6637119ca51e48e33c600cb173957b76712. When Maglev inlines `new Array(x0, x1, ...)` with PACKED_DOUBLE_ELEMENTS, it calls `ConvertForField()` for the kFloat64 field type, which internal... | none | [
"From a972bc44e32acb2ed019326a15df10013c1038dd Mon Sep 17 00:00:00 2001\nFrom: Jakob Linke <jgruber@chromium.org>\nDate: Mon, 10 Nov 2025 10:26:57 +0100\nSubject: [PATCH] [maglev] Temporarily disable double array ctor optimization\n\nFixed: 457866804\nChange-Id: I6c7a9330a5af5a9958ef4c0fc8b3e3e32d168921\nReviewed-o... |
v8__472139305 | v8 | 472139305 | hwiwonlee/v8.x86_64:472139305 | /src/v8 | out/x64.asan/d8 | --sandbox-testing --expose-gc | [
"src/objects/js-number-format.cc"
] | [
"src/objects"
] | Use-after-free | SANDBOX_VIOLATION | A use-after-free (UAF) vulnerability exists in V8's JSNumberFormat implementation that allows attackers to execute arbitrary code or cause denial of service. The vulnerability is triggered when calling Intl.NumberFormat.prototype.format(), formatToParts(), formatRange(), or formatRangeToParts() with a malicious object ... | none | [
"From bdc8f396b7d81e6521ee395970935edede39349a Mon Sep 17 00:00:00 2001\nFrom: Frank Tang <ftang@chromium.org>\nDate: Thu, 29 Jan 2026 12:53:09 -0800\nSubject: [PATCH] [intl] Fix Sandbox Bypass: Use-After-Free in ICU\n NumberFormatter\n\nBug: 472139305\nChange-Id: I48e159312a2a56e6ddb6b0d7234a4df37ac62e20\nReviewe... |
sm__1791520 | sm | 1791520 | hwiwonlee/sm.x86_64:1791520 | /src/gecko-dev | /out/js | --fuzzing-safe --no-threads --ion-eager | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Missing GC keep-alive on TypedArray across BigInt allocation in Ion JIT-compiled BigInt64Array element ops. CVE-2022-42928 (sec-high, MFSA 2022-46, Firefox 107). MLoadUnboxedScalar / MAtomicTypedArrayElementBinop for Scalar::BigInt64 returned via CreateBigIntFromInt64 which nursery-allocates a fresh BigInt. The Ion-com... | [
"From c224760a76fa4c1f59d4bb82d3e69d74503dfc83 Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Thu, 6 Oct 2022 18:22:12 +0000\nSubject: [PATCH] Bug 1791520: Add some keep alive annotations. r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/... | |
sm__1791975 | sm | 1791975 | hwiwonlee/sm.x86_64:1791975 | /src/gecko-dev | /out/js | --fuzzing-safe --no-threads --no-baseline --no-ion | [
"js/src/gc/GC.cpp",
"js/src/vm/Realm-inl.h",
"js/src/vm/Realm.cpp",
"js/src/vm/Realm.h"
] | [
"js/src/gc",
"js/src/vm"
] | Use-after-free | ASAN_CRASH | Realm allocated during incremental GC swept as dead before it could be marked. CVE-2022-45406 (sec-high, MFSA 2022-47, Firefox 107). Compartment::sweepRealms in js/src/gc/GC.cpp assumed every realm in the compartment existed at the start of the current incremental collection. A newGlobal/Reflect.parse/OOM path can allo... | [
"From 4cd3dc8813f171cb6a8cb41088fd97eaa5909249 Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Mon, 17 Oct 2022 17:09:07 +0000\nSubject: [PATCH] Bug 1791975 - Don't sweep realms that were allocated during\n incremental GC r=jandem\n\nWhen marking a BaseShape we mark its global, and we re... | |
sm__1796901 | sm | 1796901 | hwiwonlee/sm.x86_64:1796901 | /src/gecko-dev | /out/js | --fuzzing-safe --no-threads --no-baseline --no-ion | [
"js/src/gc/GC.cpp"
] | [
"js/src/gc"
] | Use-after-free | ASAN_CRASH | Realm incremental-marking state cleared at the end of GC instead of the start, leaving stale realm allocation state visible to the next collection. CVE-2022-45409 (sec-high, MFSA 2022-47, Firefox 107). GCRuntime stored per-realm flags such as wasAllocatedDuringGC to protect realms created during an incremental collecti... | [
"From 041774db08880cca1ab10e807d0d71718964dffe Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Wed, 26 Oct 2022 11:09:39 +0000\nSubject: [PATCH] Bug 1796901 - Clear realm incremental marking state at the\n start of GC rather than at the end r=jandem\n\nGC can be aborted in several states... | |
sm__1804626 | sm | 1804626 | hwiwonlee/sm.x86_64:1804626 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --ion-warmup-threshold=0 --baseline-eager | [
"js/src/jit/CodeGenerator.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Ion MIR shape-tracing regression - missing GC edge from Ion code to a JS Shape, detected by endVerifyPreBarriers. Bug 1804626 is a use-after-free class bug (csectype-uaf): the regressing Ion optimization in js/src/jit/CodeGenerator.cpp assumed shapes were always tenured and did not need to be traced through MIR operati... | [
"From 11a4da715e43daf8839a8220d1097fc489075f6f Mon Sep 17 00:00:00 2001\nFrom: Doug Thayer <dothayer@mozilla.com>\nDate: Thu, 8 Dec 2022 22:07:54 +0000\nSubject: [PATCH] Bug 1804626 - Backed out changeset 77591550134c r=iain\n\nIt looks like shapes are not being traced through MIR ops, instead relying on\nthe fact ... | |
sm__1810711 | sm | 1810711 | hwiwonlee/sm.x86_64:1810711 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/vm/ErrorObject.cpp"
] | [
"js/src/vm"
] | Cross-compartment violation | ASAN_CRASH | FindErrorInstanceOrPrototype — Cross-Compartment Proxy getPrototypeOf Trap Returns Foreign Object, CheckedUnwrapStatic Bypassed on Prototype Chain Advance. CVE-2023-25735 (sec-high, MFSA 2023-05, Firefox 110). js::FindErrorInstanceOrPrototype in js/src/vm/ErrorObject.cpp is the helper that backs the `Error.prototype.st... | [
"From 5461a647546a79ff479804263a3b7ef114eaf0f7 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Tue, 24 Jan 2023 00:27:33 +0000\nSubject: [PATCH] Bug 1810711: Refactor FindErrorInstanceOrPrototype r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D167105\n---\n js... | |
sm__1814899 | sm | 1814899 | hwiwonlee/sm.x86_64:1814899 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/CodeGenerator.cpp",
"js/src/jit/shared/CodeGenerator-shared.cpp"
] | [
"js/src/jit",
"js/src/jit/shared"
] | Incorrect code generation | ASAN_CRASH | Ion CodeGenerator OsiSpace — Missing ensureOsiSpace() Before Call Sites + Erroneous lastOsiPointOffset_ Advance Lets Consecutive OSI Patches Overlap. CVE-2023-25751 (sec-high, MFSA 2023-09, Firefox 111). Ion's on-stack invalidation (OSI) mechanism lets the GC atomically replace a live jitcode call return address with a... | [
"From d307cab06d56b4be62796b083b16feb86a47829c Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Mon, 13 Feb 2023 18:33:10 +0000\nSubject: [PATCH] Bug 1814899: Ensure more OsiSpace r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D169274\n---\n js/src/jit/CodeGene... | |
sm__1820543 | sm | 1820543 | hwiwonlee/sm.x86_64:1820543 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/gc/Compacting.cpp"
] | [
"js/src/gc"
] | Use-after-free | ASAN_CRASH | Heap use-after-free in SpiderMonkey GC compaction. CVE-2023-29535 (sec-high, MFSA 2023-13, Firefox 112). During GCRuntime::updatePointersToRelocatedCells, FinalizationRegistry observer edges were updated before the weak maps they depend on. When compaction relocated a WeakMap entry, the observer traversal still held th... | [
"From 533d5be1269d680dbef53e0116a917f1a7485e9d Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Wed, 15 Mar 2023 08:23:53 +0000\nSubject: [PATCH] Bug 1820543 - Update weak maps before finalization observer\n edges when compacting r=sfink\n\nThe latter relies on being able to access the fo... | |
sm__1821959 | sm | 1821959 | hwiwonlee/sm.x86_64:1821959 | /src/gecko-dev | /out/js | [
"js/src/vm/JSObject.cpp"
] | [
"js/src/vm"
] | Invalid free | ASAN_CRASH | JSObject::swap transplant invalid free - shifted dense elements use getElementsHeader() instead of the allocation base. In js/src/vm/JSObject.cpp, the transplant path computed the dynamic-elements allocation from getElementsHeader(); after Array.prototype.shift this can be an interior pointer. The GC background free pa... | [
"From abeb53ae551913d99778601be86b1ff1b195420c Mon Sep 17 00:00:00 2001\nFrom: Ted Campbell <tcampbell@mozilla.com>\nDate: Tue, 28 Mar 2023 15:37:11 +0000\nSubject: [PATCH] Bug 1821959 - Use unshiftedElementsHeader when transplanting\n objects. r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla... | ||
sm__1827073 | sm | 1827073 | hwiwonlee/sm.x86_64:1827073 | /src/gecko-dev | /out/js | --baseline-warmup-threshold=10 --fuzzing-safe | [
"js/src/jit/JSJitFrameIter.h",
"js/src/jit/JitFrames.cpp",
"js/src/jit/RematerializedFrame.cpp",
"js/src/vm/FrameIter-inl.h",
"js/src/vm/ArgumentsObject.cpp"
] | [
"js/src/jit",
"js/src/vm"
] | Out-of-bounds write | ASAN_CRASH | InlineFrameIterator::unaliasedForEachActual — Callback Invoked nformals Times Instead of nactuals → GCVector infallibleAppend Overflow. Out-of-bounds write vulnerability (csectype-bounds, sec-high) in InlineFrameIterator::unaliasedForEachActual in js/src/jit/JSJitFrameIter.h. When the number of actual arguments (nactua... | [
"From af7bb7c7c1031ae7fe82f6aa63b9e1e757af7e1a Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Wed, 12 Apr 2023 04:51:21 +0000\nSubject: [PATCH] Bug 1827073 part 1 - Remove unused ReadFrame_Formals. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla.com/D175112\n---\n j... | |
sm__1834711 | sm | 1834711 | hwiwonlee/sm.x86_64:1834711 | /src/gecko-dev | /out/js | [
"js/src/proxy/CrossCompartmentWrapper.cpp",
"js/src/proxy/DeadObjectProxy.cpp",
"js/src/proxy/DeadObjectProxy.h",
"js/src/vm/Compartment.cpp",
"js/src/vm/JSObject.cpp",
"js/src/vm/JSObject.h"
] | [
"js/src/proxy",
"js/src/vm"
] | Debug assertion failure | ASAN_CRASH | DeadObjectProxy lifecycle mismatch - NewDeadProxyObject did not consistently set ObjectFlag::BackgroundFinalize after nuking cross-compartment wrappers. The vulnerable path spans js/src/proxy/DeadObjectProxy.cpp, DeadObjectProxy.h, CrossCompartmentWrapper.cpp, and the transplant/remap logic in js/src/vm/Compartment.cpp... | [
"From 5432788739532336e444d0f4dbe88c6941f54f60 Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Tue, 6 Jun 2023 08:37:02 +0000\nSubject: [PATCH] Bug 1834711 - Set background finalized flag for dead object\n proxes created after nuking all CCWs r=jandem\n\nThe background finalized flag was... | ||
sm__1841119 | sm | 1841119 | hwiwonlee/sm.x86_64:1841119 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --wasm-function-references --wasm-gc | [
"js/src/wasm/WasmBaselineCompile.cpp",
"js/src/wasm/WasmIonCompile.cpp",
"js/src/wasm/WasmJS.cpp",
"js/src/wasm/WasmModuleTypes.cpp",
"js/src/wasm/WasmModuleTypes.h",
"js/src/wasm/WasmSerialize.cpp",
"js/src/wasm/WasmTable.cpp",
"js/src/wasm/WasmTable.h",
"js/src/wasm/WasmTypeDef.h",
"js/src/wasm/... | [
"js/src/wasm"
] | Use-after-free | ASAN_CRASH | WASM-GC Exported Object Lifecycle — TypeDef RecGroup Freed Before WasmTableObject/WasmGlobalObject/WasmTagObject Finalizer Reads RefType::hierarchy(). Bug 1841119 (sec-high). In the WebAssembly GC implementation, exported tables, globals, and tags hold references to TypeDef objects stored in RecGroups (recursion groups... | [
"From 9115ac056da8285348457b5c206636c280636cbb Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Fri, 8 Sep 2023 17:56:47 +0000\nSubject: [PATCH] Bug 1841119 - wasm: Keep recursion group alive for globals,\n tables, and tags. r=yury\n\nGlobals, tables, and tags can refer to specific TypeDef object... | |
sm__1842617 | sm | 1842617 | hwiwonlee/sm.x86_64:1842617 | /src/gecko-dev | /out/js | --fuzzing-safe --cpu-count=2 --ion-offthread-compile=off --baseline-eager | [
"js/src/jit/MacroAssembler.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | MacroAssembler branchTestObjShapeList — Missing Empty-List Check on Post-GC Shape Folding Allows Stale-Shape-Pointer Match and SetDynamicSlot Corruption. Bug 1842617 (sec-high). In js/src/jit/MacroAssembler.cpp, the function branchTestObjShapeList generates a do-while loop that unconditionally reads the first element o... | [
"From 3450c044be862b04778daa8a3535efb3782255e1 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Wed, 12 Jul 2023 16:39:14 +0000\nSubject: [PATCH] Bug 1842617: Check for empty shape list in\n branchTestObjShapeList r=jonco\n\nDifferential Revision: https://phabricator.services.mozilla.com/D... | |
sm__1851569 | sm | 1851569 | hwiwonlee/sm.x86_64:1851569 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --wasm-gc | [
"js/src/jit/MacroAssembler.cpp",
"js/src/wasm/WasmBaselineCompile.cpp",
"js/src/wasm/WasmIonCompile.cpp",
"js/src/wasm/WasmValType.cpp",
"js/src/wasm/WasmValidate.cpp"
] | [
"js/src/jit",
"js/src/wasm"
] | Type confusion | ASAN_CRASH | Wasm-GC i31ref cast type confusion - the JIT and wasm validation paths kept an obsolete assumption that values subtype-compatible with eqref are heap GC objects. The fix touches js/src/jit/MacroAssembler.cpp plus wasm compiler/validation code in WasmBaselineCompile.cpp, WasmIonCompile.cpp, WasmValType.cpp, and WasmVali... | [
"From a637208c750c031c65e5c162660b0970786a4399 Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Wed, 13 Sep 2023 20:57:56 +0000\nSubject: [PATCH] Bug 1851569: Add i31 to ref.cast test suite. r=jseward\n\nThis thoroughly tests i31's interaction with other types, and fixes a couple subtle bug... | |
sm__1852218 | sm | 1852218 | hwiwonlee/sm.x86_64:1852218 | /src/gecko-dev | /out/js | --fuzzing-safe --no-threads --disable-oom-functions --baseline-eager --ion-warmup-threshold=0 | [
"js/src/jit/JitHints.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | JitHints::addIonHint — mozilla::HashTable remove() Called With Live AddPtr → shrinkIfUnderloaded Invalidates Cached Pointer / Bucket Metadata Corruption. CVE-2023-5172 (sec-high). In js/src/jit/JitHints.cpp, the function addIonHint() accepts an AddPtr (a live hash table lookup handle) and, when the hint map exceeds Ion... | [
"From fd2d87c96285a5eb430ae45e28a4a2f351e95293 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Mon, 11 Sep 2023 22:23:58 +0000\nSubject: [PATCH] Bug 1852218: Remove old hints after adding new hints\n r=dpalmeiro\n\nDifferential Revision: https://phabricator.services.mozilla.com/D187807\n-... | |
sm__1854068 | sm | 1854068 | hwiwonlee/sm.x86_64:1854068 | /src/gecko-dev | /out/js | --fuzzing-safe --cpu-count=2 --ion-offthread-compile=off --ion-warmup-threshold=10 --baseline-eager | [
"js/src/jit/JitFrames.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Ion safepoint tracing omission - wasm::AnyRef register spills were not traced in Ion JS frames. In js/src/jit/JitFrames.cpp, TraceIonJSFrame handled normal GC slots, Value spills, and wasm AnyRef stack slots, but missed wasm AnyRef values spilled in general registers around out-of-line calls. If a GC occurred while JS ... | [
"From 48114f50e75e0637cfd2e63f14868ad1fbaeff14 Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Tue, 26 Sep 2023 16:20:39 +0000\nSubject: [PATCH] Bug 1854068 - wasm: Trace AnyRef spill slots in Ion frames.\n r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D188845\n---\... | |
sm__1862473 | sm | 1862473 | hwiwonlee/sm.x86_64:1862473 | /src/gecko-dev | /out/js | --wasm-tail-calls | [
"js/src/jit/shared/CodeGenerator-shared.cpp"
] | [
"js/src/jit/shared"
] | Stack corruption | ASAN_CRASH | Ion wasm tail-call stack alignment bug - local and temporary slots for wasm tail calls were packed without enforcing WasmStackAlignment. The vulnerable Ion frame-layout code in js/src/jit/shared/CodeGenerator-shared.cpp could move the stack pointer too far during return_call/return_call_indirect, letting callee stack a... | [
"From 5e55c7effe8cd125c0404ed9b00193de388eee3e Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Wed, 8 Nov 2023 16:56:44 +0000\nSubject: [PATCH] Bug 1862473 - Align local slots for wasm tail calls. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D192895\n---\n ... | |
sm__1871089 | sm | 1871089 | hwiwonlee/sm.x86_64:1871089 | /src/gecko-dev | /out/js | --fuzzing-safe --fast-warmup --gc-zeal=14 | [
"js/src/jit/BaselineCacheIRCompiler.cpp",
"js/src/jit/JitScript.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | BaselineCacheIRCompiler Constructor IC — Stale ICStubReg Reload After GC Active-Stub Cloning Discards Spilled Pointer. CVE-2024-0744 (sec-high). A use-after-free vulnerability exists in SpiderMonkey's baseline JIT IC (Inline Cache) stub handling in js/src/jit/BaselineCacheIRCompiler.cpp. When the baseline IC inlines a ... | [
"From 6d17e93756e872eb2c3a3ddbccb80ad94c5bbed9 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 4 Jan 2024 22:54:14 +0000\nSubject: [PATCH] Bug 1871089 - Load ICStub from the frame instead of storing\n it separately. r=iain\n\nDifferential Revision: https://phabricator.services.mozill... | |
sm__1879237 | sm | 1879237 | hwiwonlee/sm.x86_64:1879237 | /src/gecko-dev | /out/js | --wasm-memory-control --wasm-compiler=baseline --wasm-exnref | [
"js/src/wasm/WasmBaselineCompile.cpp"
] | [
"js/src/wasm"
] | Incorrect code generation | ASAN_CRASH | WASM Baseline br_on_cast Block Params — Spurious needResultRegs Spill After Shuffle Breaks isMem==onStack Invariant. CVE-2024-2606 (sec-high). In the WebAssembly baseline compiler, the br_on_cast instruction with block parameters triggers incorrect register allocation in js::wasm::BaseCompiler::shuffleStackResultsBefor... | [
"From 9a3cef41437423bf9e833e1e127349080e27945c Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Tue, 27 Feb 2024 15:54:19 +0000\nSubject: [PATCH] Bug 1879237: Consolidate wasm cast code. r=rhunt\n\nThe code for tracking register allocation in wasm casts was getting very\nunwieldy. This patc... | |
sm__1880719 | sm | 1880719 | hwiwonlee/sm.x86_64:1880719 | /src/gecko-dev | /out/js | --wasm-memory-control --wasm-compiler=optimizing --wasm-exnref | [
"js/src/jit/CodeGenerator.cpp"
] | [
"js/src/jit"
] | Integer overflow | ASAN_CRASH | Wasm-GC array allocation integer overflow - Ion code generation used a wrapped element-storage byte count for wasm array allocation. In js/src/jit/CodeGenerator.cpp, visitWasmNewArrayObject computed numElements * elementSize for inline allocation and did not correctly reject overflowing products before allocation. A no... | [
"From cd226ea7e9d033ff1fb50bbfe82eefcea9d2f239 Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Tue, 27 Feb 2024 15:24:35 +0000\nSubject: [PATCH] Bug 1880719: Fix incorrect array size calculation. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D202466\n---\n js/s... | |
sm__1882751 | sm | 1882751 | hwiwonlee/sm.x86_64:1882751 | /src/gecko-dev | /out/js | --wasm-compiler=baseline | [
"js/src/jit/MacroAssembler.cpp",
"js/src/wasm/WasmGcObject.h"
] | [
"js/src/jit",
"js/src/wasm"
] | Integer overflow | ASAN_CRASH | Wasm-GC inline array allocation integer overflow - MacroAssembler::wasmArrayAllocCheckLength used a signed-overflow test for numElements * elementSize, allowing unsigned-overflowing lengths to enter the fast allocation path. The vulnerable fast-path check in js/src/jit/MacroAssembler.cpp interacted with wasm array sizi... | [
"From d6d6aababd5b2aa1d622fe4f56a1e60c167c5547 Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Wed, 6 Mar 2024 22:40:41 +0000\nSubject: [PATCH] Bug 1882751: Replace runtime array length checks with a\n lookup. r=rhunt\n\nWe can reduce branching in our array allocation code by doing a singl... | |
sm__1883542 | sm | 1883542 | hwiwonlee/sm.x86_64:1883542 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/CacheIR.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | Baseline CacheIR GetBoundName IC returned the wrong `this` object across a security boundary. CVE-2024-3852 (sec-high, MFSA 2024-18, Firefox 125). When a getter was installed on the global-lexical binding and invoked through the Baseline GetBoundName inline cache, the IC exposed the bare inner global object as `this` i... | [
"From 86083f35c4dc832565464cca79f6224821a985e0 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Thu, 21 Mar 2024 15:45:59 +0000\nSubject: [PATCH] Bug 1883542: Simplify GetBoundName ICs r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D204130\n---\n js/src/jit/Cac... | |
sm__1884427 | sm | 1884427 | hwiwonlee/sm.x86_64:1884427 | /src/gecko-dev | /out/js | --fuzzing-safe --no-threads --no-baseline --no-ion | [
"js/src/vm/Realm.cpp"
] | [
"js/src/vm"
] | Use-after-free | ASAN_CRASH | Realm initialization lifetime bug - a Realm being initialized at the start of GC was not marked to be kept alive. In js/src/vm/Realm.cpp, failed global creation during an out-of-memory path could destroy a Realm while a live BaseShape still referenced the Realm's global. Later GC tracing followed that dangling Realm/gl... | [
"From 85c8dad6f466d7cf5c45444126ec3440e8c15aba Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Thu, 21 Mar 2024 16:24:25 +0000\nSubject: [PATCH] Bug 1884427 - Don't destroy realms that were being\n initialised at the start of GC r=jandem\n\nThe problem is that when global object creation... | |
sm__1884552 | sm | 1884552 | hwiwonlee/sm.x86_64:1884552 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | Ion MaybeFoldDiamondConditionBlock — Successor Count Check Admits Degenerate MTableSwitch, UpdateTestSuccessors toGoto() Cast Dereferences Switch Jump Table as Block Pointer. CVE-2024-3854 (sec-high). A type confusion and out-of-bounds read vulnerability exists in SpiderMonkey's Ion JIT compiler in the MaybeFoldDiamond... | [
"From fa7f622001a7afe1cd78a2dc4111eca29dd7e6fc Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Thu, 21 Mar 2024 15:52:27 +0000\nSubject: [PATCH] Bug 1884552: Refactor IsDiamondPattern r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D204546\n---\n js/src/jit/Ion... | |
sm__1884887 | sm | 1884887 | hwiwonlee/sm.x86_64:1884887 | /src/gecko-dev | /out/js | --no-threads --wasm-compiler=baseline | [
"js/src/wasm/WasmInitExpr.cpp",
"js/src/wasm/WasmInstance.cpp",
"js/src/wasm/WasmJS.cpp",
"js/src/wasm/WasmJS.h",
"js/src/wasm/WasmModule.cpp",
"js/src/wasm/WasmValue.h"
] | [
"js/src/wasm"
] | Use-after-free | ASAN_CRASH | Wasm exported-global type lifetime bug - literal global initialization and subtyping could leave an exported global pointing at a TypeDef/RecGroup that was later freed. The fix spans js/src/wasm/WasmInitExpr.cpp, WasmInstance.cpp, WasmJS.cpp/WasmJS.h, WasmModule.cpp, and WasmValue.h, cleaning up global initialization s... | [
"From 22ea154bc960b7810f5e316d179746222b476d05 Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Fri, 22 Mar 2024 19:11:19 +0000\nSubject: [PATCH] Bug 1884887 - wasm: Cleanup global initialization logic.\n r=yury\n\nDifferential Revision: https://phabricator.services.mozilla.com/D204378\n---\n js/... | |
sm__1885775 | sm | 1885775 | hwiwonlee/sm.x86_64:1885775 | /src/gecko-dev | /out/js | --enable-symbols-as-weakmap-keys --fuzzing-safe | [
"js/src/gc/WeakMap-inl.h"
] | [
"js/src/gc"
] | Use-after-free | ASAN_CRASH | WeakMap+Symbol cross-zone use-after-free. sec-high (csectype-uaf). When a JS::Symbol is used as a WeakMap key, WeakMap::addImplicitEdges failed to add a sweep-group edge from the symbol's zone to the WeakMap's zone — the same handling already applied for JSObject keys was missing for symbols. Across multiple zones the ... | [
"From 26125066c9f657eae22718920085779bb586b72b Mon Sep 17 00:00:00 2001\nFrom: Yoshi Cheng-Hao Huang <allstars.chh@gmail.com>\nDate: Tue, 26 Mar 2024 14:08:52 +0000\nSubject: [PATCH] Bug 1885775 - addSweepGroupEdgeTo Weakmap's zone if the key\n is of a JS::Symbol. r=jonco\n\nDifferential Revision: https://phabricat... | |
sm__1885828 | sm | 1885828 | hwiwonlee/sm.x86_64:1885828 | /src/gecko-dev | /out/js | --fast-warmup --fuzzing-safe | [
"js/src/jit/MIR.h",
"js/src/jit/MIROps.yaml"
] | [
"js/src/jit"
] | Out-of-bounds read | ASAN_CRASH | Ion MIR MSubstr Movable Flag Regression — LICM Hoists String.slice SubstringKernel Above Negative-Index Guard → OOB Read. CVE-2024-3855 (sec-high). An out-of-bounds read vulnerability exists in SpiderMonkey's Ion JIT compiler due to the MSubstr MIR instruction being incorrectly marked as movable, allowing Loop Invarian... | [
"From 691384c30df255fd95ffd6e332c4d6df6a36f2a6 Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Thu, 21 Mar 2024 16:48:43 +0000\nSubject: [PATCH] Bug 1885828: Make MSubstr non-movable. r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D20488... | |
sm__1885829 | sm | 1885829 | hwiwonlee/sm.x86_64:1885829 | /src/gecko-dev | /out/js | [
"js/src/jit/CodeGenerator.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Ion wasm array-data pointer safepoint bug - MWasmLoadField could keep an ArrayDataPointer live across a GC-triggering helper call without recording it in the safepoint. The vulnerable code generation in js/src/jit/CodeGenerator.cpp let createArrayIL trigger a nursery collection while the array-data pointer was only a r... | [
"From ca67e9c7017484ebfe722e19a3f2028143072449 Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Fri, 29 Mar 2024 14:41:19 +0000\nSubject: [PATCH] Bug 1885829 - Track ArrayDataPointer at safepoints.\n r=bvisness,rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D205... | ||
sm__1886683 | sm | 1886683 | hwiwonlee/sm.x86_64:1886683 | /src/gecko-dev | /out/js | --fast-warmup --ion-check-range-analysis --ion-extra-checks --fuzzing-safe --disable-oom-functions | [
"js/src/jit/JitFrames.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | JitFrames TraceThisAndArguments — Truncates Trace Range to nactuals, Missing WasmValueBox Tracing for Rectifier-Padded Formal externref Slots. CVE-2024-3857 (sec-high). js::jit::TraceThisAndArguments in js/src/jit/JitFrames.cpp is the JSJitToWasm frame tracer used during minor/compacting GC to mark the arguments live f... | [
"From d4b2ea13efebadcfbbb79baf70804f27d835d81a Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Tue, 26 Mar 2024 15:00:42 +0000\nSubject: [PATCH] Bug 1886683 - Simplify tracing of arguments in\n TraceThisAndArguments. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla.co... | |
sm__1886849 | sm | 1886849 | hwiwonlee/sm.x86_64:1886849 | /src/gecko-dev | /out/js | --no-threads --spectre-mitigations=off | [
"js/src/jit/RangeAnalysis.cpp",
"js/src/jit/MIROps.yaml"
] | [
"js/src/jit"
] | Incorrect JIT optimization | ASAN_CRASH | Ion RangeAnalysis MObjectKeysLength::computeRange — Upper Bound Fixed at NativeObject::MAX_SLOTS_COUNT Ignores TypedArray Indexed-Property Keys (CVE-2024-29943, Pwn2Own 2024). CVE-2024-29943 (Pwn2Own 2024). MObjectKeysLength::computeRange in js/src/jit/RangeAnalysis.cpp:1824 incorrectly computes the range of Object.key... | [
"From 81806e7ccec7dde41e37c9891592a6e39ce46380 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Thu, 21 Mar 2024 18:48:46 +0000\nSubject: [PATCH] Bug 1886849: Remove MObjectKeysLength::computeRange r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D205375\n---\n j... | |
sm__1888614 | sm | 1888614 | hwiwonlee/sm.x86_64:1888614 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/JSJitFrameIter.cpp",
"js/src/jit/JSJitFrameIter.h",
"js/src/jit/JitFrames.cpp",
"js/src/vm/FrameIter.cpp"
] | [
"js/src/jit",
"js/src/vm"
] | Cross-compartment violation | ASAN_CRASH | JIT exception unwinding realm mismatch - trampoline native frames did not restore the caller realm when unwinding an exception. The fix changes js/src/jit/JSJitFrameIter.cpp, JSJitFrameIter.h, JitFrames.cpp, and js/src/vm/FrameIter.cpp so frame iteration recognizes TrampolineNative frames and recovers their callee/real... | [
"From 7a6d8fd713576aafe1923dbec51a46a42fb60fdf Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Wed, 3 Apr 2024 09:27:15 +0000\nSubject: [PATCH] Bug 1888614 - Fix exception handler to restore realm for\n trampoline native frames too. r=iain\n\nDifferential Revision: https://phabricator.ser... | |
sm__1888892 | sm | 1888892 | hwiwonlee/sm.x86_64:1888892 | /src/gecko-dev | /out/js | --fuzzing-safe --gc-zeal=10 | [
"js/src/jit/CacheIRCompiler.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | CacheIRCompiler TraceWeakBaselineStubFrame — Ignores TraceWeakCacheIRStub False Return, Active Sparse-GetElement Stub Retains Dead Array.prototype Shape After Proxy Proto Swap. CVE-2024-3858 (sec-high). A wild pointer dereference vulnerability exists in SpiderMonkey's Baseline JIT CacheIR stub tracing logic in js/src/j... | [
"From a334c67b34d158501e072910310a63db72a94401 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 4 Apr 2024 13:31:40 +0000\nSubject: [PATCH] Bug 1888892 - Trace all fields in TraceWeakCacheIRStub.\n r=jonco\n\nDifferential Revision: https://phabricator.services.mozilla.com/D206481\n---... | |
sm__1889317 | sm | 1889317 | hwiwonlee/sm.x86_64:1889317 | /src/gecko-dev | /out/js | [
"js/src/jit/MIR.h"
] | [
"js/src/jit"
] | Incorrect JIT optimization | ASAN_CRASH | Ion wasm subtype-check optimization bug - MWasmRefIsSubtypeOfAbstract and MWasmRefIsSubtypeOfConcrete in js/src/jit/MIR.h advertised AliasSet::None, so GVN/LICM could treat subtype checks as effect-free and hoist or merge them across ref.cast trap boundaries. The generated check then ran at a program point where the dy... | [
"From 6f0d23973709845703363ef7033b1c3de3ba7581 Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Thu, 11 Apr 2024 16:44:58 +0000\nSubject: [PATCH] Bug 1889317: Disable GVN for wasm subtype checks. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D206967\n---\n js/sr... | ||
sm__1895086 | sm | 1895086 | hwiwonlee/sm.x86_64:1895086 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/vm/JSObject.cpp"
] | [
"js/src/vm"
] | Use-after-free | ASAN_CRASH | JSObject::swap Transplant — setIsUsedAsPrototype Triggers GC Outside SuppressGC Region Before Pre-Write Barrier → Shape getObjectClass UAF. CVE-2024-5688 (sec-high). A use-after-free vulnerability exists in SpiderMonkey's object transplant mechanism in js/src/vm/JSObject.cpp. During JSObject::swap, which is used by the... | [
"From ada64fca6f4e88ab59b45f6e866523a3c460c511 Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Wed, 15 May 2024 08:20:36 +0000\nSubject: [PATCH] Bug 1895086 - Suppress GC during JSObject::swap r=jandem\n\nWe already suppress GC for part of this, but not for the part where we call\nJSObje... | |
sm__1901411 | sm | 1901411 | hwiwonlee/sm.x86_64:1901411 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/vm/AsyncIteration.cpp",
"js/src/vm/AsyncIteration.h",
"js/src/vm/List-inl.h"
] | [
"js/src/vm"
] | Type confusion | ASAN_CRASH | AsyncGeneratorDrainQueue — Object.prototype.then Getter Re-enters AsyncGeneratorAwaitReturn, Mutating State From 'completed' Back to 'awaiting-return' Between Queue Operations (cross-engine spec bug, CVE-2024-7652). CVE-2024-7652 (sec-high). A spec-level bug in the ECMAScript AsyncGenerator protocol causes type confusi... | [
"From 907e0ce2a1aef3f8cb3e9de2bdda3f8b74dc0a84 Mon Sep 17 00:00:00 2001\nFrom: Tooru Fujisawa <arai_a@mac.com>\nDate: Wed, 19 Jun 2024 16:31:26 +0000\nSubject: [PATCH] Bug 1901411 - Part 1: Partially revert\n AsyncGeneratorDrainQueue to AsyncGeneratorResumeNext. r=mgaudet,jandem\n\nDifferential Revision: https://ph... | |
sm__1902983 | sm | 1902983 | hwiwonlee/sm.x86_64:1902983 | /src/gecko-dev | /out/js | --fast-warmup --gc-zeal=21,100 --fuzzing-safe | [
"js/src/jit/JSJitFrameIter.cpp",
"js/src/jit/JSJitFrameIter.h"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Wasm bailout-data use-after-free in JSJitFrameIter when Ion calls a Wasm function directly via FrameType::Exit. JSJitFrameIter held a BailoutInfo* pointing into stack memory tied to Wasm frames; after iterating past the Wasm frames the memory was reused, leaving the iterator pointing at freed bailout data. Follow-up to... | [
"From d5df70357db3d244669ed3147467e5b7ade0f51c Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Wed, 19 Jun 2024 12:50:57 +0000\nSubject: [PATCH] Bug 1902983 - Don't use bailout data after iterating Wasm\n frames. r=iain\n\nThis is similar to bug 1900523, but the fix there was incomplete b... | |
sm__1903041 | sm | 1903041 | hwiwonlee/sm.x86_64:1903041 | /src/gecko-dev | /out/js | [
"js/src/wasm/WasmTypeDef.h",
"js/src/wasm/WasmValidate.cpp"
] | [
"js/src/wasm"
] | Type confusion | ASAN_CRASH | Wasm-GC subtype validation type confusion - recursive group validation compared struct fields before some SuperTypeVectors were initialized. In js/src/wasm/WasmTypeDef.h and WasmValidate.cpp, the fast subtype path could compare null/uninitialized STV pointers and conclude that incompatible wasm types were subtypes. Deb... | [
"From c6786f1019e59269e01c81e9e9d79947f199870a Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Thu, 11 Jul 2024 13:52:19 +0000\nSubject: [PATCH] Bug 1903041: Use a linear search in more situations. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D214989\n---\n js... | ||
sm__1903219 | sm | 1903219 | hwiwonlee/sm.x86_64:1903219 | /src/gecko-dev | /out/js | --setpref=wasm_js_string_builtins=true | [
"js/src/wasm/WasmBuiltinModule.cpp"
] | [
"js/src/wasm"
] | Type confusion | ASAN_CRASH | Wasm builtin-function table entry type confusion - js-string-builtins references were treated like ordinary wasm functions when filling funcref tables. The vulnerable path is in js/src/wasm/WasmBuiltinModule.cpp: builtin function references did not have a normal CodeRange::funcCheckedCallEntry, so the engine could stor... | [
"From f6b32f7328122f4b5c7e79be66825ec06796c06a Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Wed, 26 Jun 2024 03:53:20 +0000\nSubject: [PATCH] Bug 1903219 - Allow wasm builtin functions to be used in ref.\n r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D21... | |
sm__1904644 | sm | 1904644 | hwiwonlee/sm.x86_64:1904644 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/jit/JitFrames.cpp",
"js/src/jit/x64/MacroAssembler-x64.cpp",
"js/src/wasm/WasmBuiltins.cpp",
"js/src/wasm/WasmBuiltins.h",
"js/src/wasm/WasmStubs.cpp",
"js/src/wasm/WasmStubs.h"
] | [
"js/src/jit",
"js/src/wasm"
] | Use-after-free | ASAN_CRASH | JS-to-wasm exception handling use-after-free - JIT exception unwinding did not restore the wasm instance register before jumping to a wasm catch handler. In js/src/jit/JitFrames.cpp, the exception path for a JS call that throws into wasm left the catch handler running with a poisoned/stale instance register. Debug buil... | [
"From b44f14fb5cc6c96aa085eeb436541587ed4084d1 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 18 Jul 2024 10:51:40 +0000\nSubject: [PATCH] Bug 1904644 - Share more exception handling code. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D214959\n---\n js/s... | |
sm__1908631 | sm | 1908631 | hwiwonlee/sm.x86_64:1908631 | /src/gecko-dev | /out/js | --wasm-compiler=baseline --fuzzing-safe | [
"js/src/wasm/WasmBaselineCompile.cpp"
] | [
"js/src/wasm"
] | Out-of-bounds read | ASAN_CRASH | Wasm baseline compiler missing dead-code check in BaseCompiler::emitTableFill. Sec-high (csectype-bounds). When table.fill appears in dead code (after an unreachable), the baseline compiler still consumed operands from the wasm value stack, violating the numval <= stk_.length() invariant and tripping MOZ_RELEASE_ASSERT... | [
"From 5a54e992206bf047493b1081886b364fe21837d7 Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Mon, 22 Jul 2024 21:04:37 +0000\nSubject: [PATCH] Bug 1908631: Add missing dead code check in table.fill.\n r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D217310\n---... | |
sm__1911909 | sm | 1911909 | hwiwonlee/sm.x86_64:1911909 | /src/gecko-dev | /out/js | [
"js/src/wasm/WasmTypeDef.h",
"js/src/wasm/WasmTypeDef.cpp",
"js/src/wasm/WasmValidate.cpp"
] | [
"js/src/wasm"
] | Type confusion | ASAN_CRASH | WasmTypeDef ArrayType::canBeSubtypeOf — Unconditionally Returns True When Mutability Differs, Admits anyref↔i64 / struct(i64) Subtyping For addrof/fakeobj/Arbitrary R/W (CVE-2024-8385). CVE-2024-8385 (sec-high). js::wasm::ArrayType::canBeSubtypeOf() always returns true when array mutability differs between subtype and ... | [
"From 7a57964c63f520aef9d0c46731d94fd55680e81f Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Thu, 8 Aug 2024 20:49:44 +0000\nSubject: [PATCH] Bug 1911909 - wasm: Refactor structs and arrays to more\n closely match the spec. r=bvisness\n\nThe spec has the notion of a 'FieldType' which both stru... | ||
sm__1912715 | sm | 1912715 | hwiwonlee/sm.x86_64:1912715 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/gc/Cell.h",
"js/src/vm/NativeObject.h",
"js/src/vm/NativeObject.cpp",
"js/src/vm/Interpreter-inl.h",
"js/src/vm/Interpreter.cpp",
"js/src/vm/EnvironmentObject.cpp"
] | [
"js/src/gc",
"js/src/vm"
] | Type confusion | ASAN_CRASH | Interpreter FetchName with_LookupProperty — Stale PropertyResult After Symbol.unscopables Getter Self-Redefines to Data Property → Cell::as<GetterSetter>() Type Confusion (CVE-2024-8381). CVE-2024-8381. Type confusion vulnerability (csectype-wildptr, sec-high) in the 'with' statement environment lookup path, rooted in ... | [
"From fab7e5c28e628ddc2b873a723838562c9b41205e Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Tue, 20 Aug 2024 17:55:37 +0000\nSubject: [PATCH] Bug 1912715: Simplify with-env handling in FetchName r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D219198\n---\n ... | |
sm__1914009 | sm | 1914009 | hwiwonlee/sm.x86_64:1914009 | /src/gecko-dev | /out/js | --wasm-compiler=baseline --fuzzing-safe | [
"js/src/wasm/WasmBaselineCompile.cpp"
] | [
"js/src/wasm"
] | Stack corruption | ASAN_CRASH | Wasm baseline emitReturnCallRef mishandled multi-value stack results for return_call_ref tail calls. CVE-2024-9402 / MFSA 2024-50 cluster (sec-high, Firefox 130/131). In js/src/wasm/WasmBaselineCompile.cpp, BaseCompiler::emitReturnCallRef lowered a tail call through a function reference as if the callee produced a sing... | [
"From d942bcd55a236074e4fd0601a103e8c252e56899 Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Thu, 22 Aug 2024 14:23:03 +0000\nSubject: [PATCH] Bug 1914009 - [wasm] Properly handle return_call_ref results\n in baseline. r=rhunt\n\nDifferential Revision: https://phabricator.services.moz... | |
sm__1914475 | sm | 1914475 | hwiwonlee/sm.x86_64:1914475 | /src/gecko-dev | /out/js | [
"js/src/jit/CodeGenerator.cpp",
"js/src/jit/Lowering.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Ion wasm call_ref safepoint omission - cross-instance function-reference calls needed an adjunct safepoint but not all LWasmCall nodes received one. The vulnerable lowering/codegen path in js/src/jit/Lowering.cpp and CodeGenerator.cpp could leave nursery values in wasm call descriptors untracked across a GC triggered b... | [
"From a85fc72269124b76f985b0b84b46bde92e5bd552 Mon Sep 17 00:00:00 2001\nFrom: Julian Seward <jseward@acm.org>\nDate: Wed, 4 Sep 2024 06:26:52 +0000\nSubject: [PATCH] Bug 1914475. r=yury.\n\nDifferential Revision: https://phabricator.services.mozilla.com/D220254\n---\n js/src/jit/CodeGenerator.cpp | 3 ++-\n js/src... | ||
sm__1917807 | sm | 1917807 | hwiwonlee/sm.x86_64:1917807 | /src/gecko-dev | /out/js | --wasm-compiler=ion --blinterp-eager --fuzzing-safe | [
"js/src/wasm/WasmFrameIter.cpp",
"js/src/wasm/WasmStubs.cpp",
"js/src/jit/JitFrames.cpp"
] | [
"js/src/wasm",
"js/src/jit"
] | Stack corruption | ASAN_CRASH | Wasm JIT entry frame footer alignment mismatch for return_call_indirect tail calls. CVE-2024-9403 (sec-high, MFSA 2024-50, Firefox 131). GenerateJitEntry in js/src/wasm/WasmStubs.cpp reserved the JIT-to-wasm exit footer using the wrong parameter-area alignment for cases where the entered function later performed return... | [
"From 16781a13b5ecce172cf06a272f1cee629c9afee2 Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Fri, 13 Sep 2024 17:07:58 +0000\nSubject: [PATCH] Bug 1917807 - Fix GenerateJitEntry call parameters alignment.\n r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D2... | |
sm__1919246 | sm | 1919246 | hwiwonlee/sm.x86_64:1919246 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Incorrect JIT optimization | ASAN_CRASH | Ion OSR-loop optimization graph corruption - an optional optimization for loops containing OSR loops produced inconsistent predecessor/backedge structure. In js/src/jit/IonAnalysis.cpp, the pass could drop or mismatch phi inputs for inner-loop predecessors; debug builds hit AssertGraphCoherency/pred->isLoopBackedge ass... | [
"From 3b81464999d7697966fcd3807dd2ebd89f30339b Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Tue, 17 Sep 2024 16:29:02 +0000\nSubject: [PATCH] Bug 1919246 - Back out D221953 and add tests. r=mgaudet\n\nFor now back out D221953 because it was an optional change to improve codegen a bit\n... | |
sm__1926235 | sm | 1926235 | hwiwonlee/sm.x86_64:1926235 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --blinterp-eager | [
"js/src/jit/CacheIR.cpp"
] | [
"js/src/jit"
] | Integer truncation | ASAN_CRASH | CacheIR argument-slot index truncation - optimized argument access did not reject call frames with more than 255 arguments. The vulnerable code in js/src/jit/CacheIR.cpp emitted loadArgumentFixedSlot with an 8-bit fixed-slot index; large arities truncated the slot index and loaded from the wrong stack location. Debug b... | [
"From 2120c788217455ea67c1b6f31f084b51e6d17fdf Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Tue, 22 Oct 2024 15:11:44 +0000\nSubject: [PATCH] Bug 1926235: Add missing argument length checks. r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.... | |
sm__1929623 | sm | 1929623 | hwiwonlee/sm.x86_64:1929623 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/vm/Modules.cpp"
] | [
"js/src/vm"
] | Cross-compartment violation | ASAN_CRASH | SyntheticModuleEvaluate — Cross-Compartment JSON Module Passes Result MutableHandle<Value> to AsyncFunctionReturned Without AutoRealm (Compartment Mismatch → UAF). CVE-2025-0240 (sec-high). A use-after-free vulnerability exists in SpiderMonkey's synthetic (JSON) module evaluation in js/src/vm/Modules.cpp. In SyntheticM... | [
"From 62ea7950e9268e5b58d1dc6e4de6016ab719e11f Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Fri, 6 Dec 2024 09:28:16 +0000\nSubject: [PATCH] Bug 1929623 - The result of evaluating a JSON module should\n be |undefined| r=jandem\n\nDifferential Revision: https://phabricator.services.moz... | |
sm__1933023 | sm | 1933023 | hwiwonlee/sm.x86_64:1933023 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/builtin/intl/Segmenter.cpp",
"js/src/builtin/intl/Segmenter.h"
] | [
"js/src/builtin/intl"
] | Type confusion | ASAN_CRASH | Intl.Segmenter Latin1/two-byte char buffer type confusion after string atomization. CVE-2025-0241 (sec-moderate, MFSA 2025-01, Firefox 134). SegmentsObject cached the underlying JS string's Latin1 bit at construction time via getString()->hasLatin1Chars(). Bug 1881995's atom-ref string mechanism allows a two-byte JSStr... | [
"From 6a1bdb80cb57384cd915ae2e27d3363a13ecb6f9 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Wed, 27 Nov 2024 12:16:40 +0000\nSubject: [PATCH] Bug 1933023 - Check whether the segment's chars are Latin1\n instead of the JS string. r=anba\n\nDifferential Revision: https://phabricator.serv... | |
sm__1934423 | sm | 1934423 | hwiwonlee/sm.x86_64:1934423 | /src/gecko-dev | /out/js | --fuzzing-safe --cpu-count=2 --ion-offthread-compile=off --enable-explicit-resource-management --baseline-eager | [
"js/src/jit/BaselineCodeGen.cpp",
"js/src/jit/CodeGenerator.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Baseline/Ion TakeDisposeCapability — Missing Pre-Write Barrier When Clearing Environment Slot After Copying DisposeCapability Array to Stack (Incremental-GC SAB Invariant Violation). Bug 1934423 (sec-high). The Explicit Resource Management proposal adds `using`/`await using` declarations, whose teardown uses a per-bloc... | [
"From 9ff87bbb1f97706895070ab1c8cc68fff4997cb1 Mon Sep 17 00:00:00 2001\nFrom: Debadree Chatterjee <debadree333@gmail.com>\nDate: Thu, 5 Dec 2024 18:19:19 +0000\nSubject: [PATCH] Bug 1934423 - Emit pre-barrier for TakeDisposeCapability\n opcode in baseline and ion. r=arai,iain\n\nDifferential Revision: https://phab... | |
sm__1942648 | sm | 1942648 | hwiwonlee/sm.x86_64:1942648 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --fast-warmup | [
"js/src/jit/CodeGenerator.cpp",
"js/src/jit/LIROps.yaml",
"js/src/jit/Lowering.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | Ion MSign lowering type confusion - Int32-input/Double-output Math.sign lowering could allocate the result in the wrong register class. The fix spans js/src/jit/Lowering.cpp, LIROps.yaml, and CodeGenerator.cpp, adding an explicit LIR/codegen path for the Int32-to-Double MSign case. Before the fix, lowering could use a ... | [
"From 5b061cdc4d40d44988dc61aa941cfbd98e31791f Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Wed, 22 Jan 2025 10:15:06 +0000\nSubject: [PATCH] Bug 1942648: Support MSign with Int32 input and Double\n output. r=jandem\n\nDifferential Revision: https://phabricator.s... | |
sm__1942881 | sm | 1942881 | hwiwonlee/sm.x86_64:1942881 | /src/gecko-dev | /out/js | --fuzzing-safe --fast-warmup | [
"js/src/jit/MIROps.yaml",
"js/src/jit/Recover.cpp",
"js/src/jit/Recover.h"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Ion MRegExpMatcher recovery bug - MRegExpMatcher in js/src/jit/MIROps.yaml was marked can_recover even though bailout recovery invokes RRegExpMatcher::recover in js/src/jit/Recover.cpp, which calls RegExpMatcherRaw and can trigger GC while cx->suppressGC is set. The debug/ASAN crash is a suppressed-GC assertion during ... | [
"From 1dc7af44dc9d04a81404d7f717d359642dd612c2 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Mon, 3 Feb 2025 21:49:54 +0000\nSubject: [PATCH] Bug 1942881: Remove RRegExpMatcher r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D236071\n---\n js/src/jit/MIROps.y... | |
sm__1945318 | sm | 1945318 | hwiwonlee/sm.x86_64:1945318 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --setpref=experimental.error_capture_stack_trace | [
"js/src/vm/SavedStacks.cpp"
] | [
"js/src/vm"
] | Out-of-bounds read | ASAN_CRASH | SavedStacks::insertFrames — Error.captureStackTrace In Suspended Async Generator Produces Empty Saved Stack, Reparenting Indexes Empty mozilla::Vector. Bug 1945318 (sec-high). When Error.captureStackTrace is called inside an async generator that has yielded, the SavedStacks::insertFrames function attempts to reparent t... | [
"From 0995847692570d46684b778af517a08c09748184 Mon Sep 17 00:00:00 2001\nFrom: Matthew Gaudet <mgaudet@mozilla.com>\nDate: Tue, 4 Feb 2025 16:21:20 +0000\nSubject: [PATCH] Bug 1945318 - Don't try to reparent an empty stack r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D236587\n---\n js... | |
sm__1946004 | sm | 1946004 | hwiwonlee/sm.x86_64:1946004 | /src/gecko-dev | /out/js | [
"js/src/jit/x64/MacroAssembler-x64.cpp",
"js/src/wasm/WasmStubs.cpp"
] | [
"js/src/jit/x64",
"js/src/wasm"
] | Uninitialized memory read | ASAN_CRASH | Wasm GenerateJitEntry x64 — i32 Return From Ion-Compiled Wasm Boxed via boxNonDouble Without Upper-32-Bit Zero-Extension → Stale rax Bits Leak Into JS::Value Tag (CVE-2025-1933). CVE-2025-1933 (sec-high). When a WebAssembly Ion-compiled function stores an i32 stack result value and returns to a Wasm-Baseline caller, on... | [
"From 7a1bf5ae72cdef17ac65f39c26e0f4c665500252 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 13 Feb 2025 12:21:29 +0000\nSubject: [PATCH] Bug 1946004 - Widen i32 return values in GenerateJitEntry.\n r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D237530\... | ||
sm__1952215 | sm | 1952215 | hwiwonlee/sm.x86_64:1952215 | /src/gecko-dev | /out/js | --no-threads --differential-testing --setpref=wasm_memory_control=true --setpref=wasm_js_string_builtins=true | [
"js/src/wasm/WasmCode.cpp"
] | [
"js/src/wasm"
] | Control-flow integrity violation | ASAN_CRASH | Wasm lazy-entry stub control-flow violation - tiering/lazy stubs computed the callee pointer incorrectly for exported wasm calls. The vulnerable logic in js/src/wasm/WasmCode.cpp could branch into the middle of a code segment or helper stub rather than the intended wasm entry point. That misdirected control flow is cau... | [
"From 4472b790da8e82d9ed5919aedbaec8b72c7b9b8f Mon Sep 17 00:00:00 2001\nFrom: Julien Pages <jpages@mozilla.com>\nDate: Wed, 12 Mar 2025 16:16:56 +0000\nSubject: [PATCH] Bug 1952215 - wasm: Fix calleePtr for lazy stubs. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D240945\n---\n js/src... | |
sm__1954042 | sm | 1954042 | hwiwonlee/sm.x86_64:1954042 | /src/gecko-dev | /out/js | --fuzzing-safe --cpu-count=2 --ion-offthread-compile=off | [
"js/src/wasm/WasmBCClass.h",
"js/src/wasm/WasmBaselineCompile.cpp"
] | [
"js/src/wasm"
] | Out-of-bounds write | ASAN_CRASH | WasmBaselineCompile struct.new — Missing OOM Check After readAllocSiteIndex, Post-OOM Vector Index Lands One Past End (Adjacent Heap Write During Function Discard). Bug 1954042 (sec-high). During wasm baseline compilation of struct.new, a missing OOM check after readAllocSiteIndex() causes the compiler to generate an i... | [
"From 33cf2bc78a72536179af5366d73df2751e1afcd0 Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Fri, 14 Mar 2025 19:28:29 +0000\nSubject: [PATCH] Bug 1954042 - Check OOM condition after readAllocSiteIndex().\n r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D24... | |
sm__1966612 | sm | 1966612 | hwiwonlee/sm.x86_64:1966612 | /src/gecko-dev | /out/js | [
"js/src/builtin/Promise.cpp"
] | [
"js/src/builtin"
] | Out-of-bounds write | ASAN_CRASH | Promise.allSettled — Separate alreadyCalled State Per onFulfilled/onRejected Function Allows Paired Callback Re-Entry After Result Array Exposed → OOB Write (CVE-2025-4918, Pwn2Own 2025). CVE-2025-4918 (Pwn2Own 2025, sec-critical). A subtle bug in the Promise.allSettled implementation allows an out-of-bounds write on t... | [
"From d4f9495ebcf3295ea5a1b2c45bf84a790e7afbd8 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Fri, 16 May 2025 11:52:12 -0400\nSubject: [PATCH] Bug 1966612 - Fix promise combinator function state.\n r=arai,a=dsmith\n\nDifferential Revision: https://phabricator.services.mozilla.com/D24978... | ||
sm__1966614 | sm | 1966614 | hwiwonlee/sm.x86_64:1966614 | /src/gecko-dev | /out/js-release | --fuzzing-safe --ion-offthread-compile=off --spectre-mitigations=off | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Incorrect JIT optimization | RUNTIME_CRASH | Ion IonAnalysis ExtractLinearSum — Wrapping uint32 Constant Folding Under MathSpace::Modulo (bitor-Truncated Index) Lets Redundant-Bounds-Check Elimination Drop Guards For Out-of-Range Dominated Accesses (CVE-2025-4919, Pwn2Own 2025). CVE-2025-4919 (Pwn2Own 2025). An incorrect bounds check elimination vulnerability exi... | [
"From e51daa6f33122c66c59a9dc8080807ba1c003131 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Sat, 17 May 2025 08:17:13 -0400\nSubject: [PATCH] Bug 1966614 - Don't support modulo math space in\n ExtractLinearSum. r=iain, a=dsmith\n\nDifferential Revision: https://phabricator.services.moz... | |
sm__1968423 | sm | 1968423 | hwiwonlee/sm.x86_64:1968423 | /src/gecko-dev | /out/js | [
"js/src/jit/CodeGenerator.cpp",
"js/src/jit/LIROps.yaml",
"js/src/jit/Lowering.cpp",
"js/src/wasm/WasmIonCompile.cpp"
] | [
"js/src/jit",
"js/src/wasm"
] | Uninitialized memory read | ASAN_CRASH | Ion/baseline wasm stack-result ABI mismatch - 32-bit wasm stack results were stored as 32 bits by Ion and later read as full pointer-width values by baseline. The fix spans js/src/jit/CodeGenerator.cpp, LIROps.yaml, Lowering.cpp, and js/src/wasm/WasmIonCompile.cpp to simplify and type the Ion stack-result handling corr... | [
"From 8c2be2db0ab32915b756d1c46988e17eb70feaae Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Tue, 8 Jul 2025 16:34:45 +0000\nSubject: [PATCH] Bug 1968423 - wasm: Simplify stack result handling in Ion.\n r=jseward\n\nIon stores stack results that are refs using MWasmStoreRef, and this\nis only ... | ||
sm__1970095 | sm | 1970095 | hwiwonlee/sm.x86_64:1970095 | /src/gecko-dev | /out/js | [
"js/src/builtin/OrderedHashTableObject.h"
] | [
"js/src/builtin"
] | Integer truncation | ASAN_CRASH | OrderedHashTableObject Rehash — size_t Allocation Size Truncated to uint32_t in AllocNurseryOrMallocBuffer, memcpy Writes Past Under-Allocated Buffer (CVE-2025-49710). CVE-2025-49710 (sec-high). The OrderedHashTableObject code uses size_t for buffer size calculations but calls AllocNurseryOrMallocBuffer which takes a u... | [
"From 4f27a916218de6a299db8d00d48ad5e91db9832d Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 5 Jun 2025 16:36:42 +0000\nSubject: [PATCH] Bug 1970095 - Clean up some Map/Set code. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla.com/D252581\n---\n js/src/builtin... | ||
sm__1987290 | sm | 1987290 | hwiwonlee/sm.x86_64:1987290 | /src/gecko-dev | /out/js | [
"js/src/builtin/FinalizationRegistryObject.cpp"
] | [
"js/src/builtin"
] | Out-of-bounds read | ASAN_CRASH | FinalizationRegistry.prototype.unregister OOB Argument Read — args[0] Used Without Bounds Check. Bug 1987290 (sec-high, csectype-bounds). The native implementation of FinalizationRegistry.prototype.unregister in js/src/builtin/FinalizationRegistryObject.cpp uses `args[0]` to grab the unregister-token argument instead o... | [
"From 2b8155c2c4c299955d4ae81423992ed56a838aee Mon Sep 17 00:00:00 2001\nFrom: \"Rong \\\"Mantle\\\" Bao\" <webmaster@csmantle.top>\nDate: Mon, 8 Sep 2025 15:03:03 +0000\nSubject: [PATCH] Bug 1987290 - Part 1: Use get() instead of operator[] to get\n arguments in FinalizationRegistryObject::unregister. r=jonco\n\nD... | ||
sm__1987481 | sm | 1987481 | hwiwonlee/sm.x86_64:1987481 | /src/gecko-dev | /out/js | --baseline-warmup-threshold=10 --ion-warmup-threshold=100 --gc-zeal=15 | [
"js/src/jit/MIROps.yaml"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | JIT typed-array element pointer use-after-free - ArrayBufferViewElementsWithOffset was treated as movable/congruent even though its result depends on mutable typed-array backing storage. The fix in js/src/jit/MIROps.yaml removes the movable/congruent/clone declarations for this MIR node. Before the fix, Ion could hoist... | [
"From 4e0b3249af9b343c82cb8bd1f5f2db3ed5f6ef0b Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Thu, 11 Sep 2025 21:22:25 +0000\nSubject: [PATCH] Bug 1987481: Adjust ArrayBufferViewElementsWithOffset. r=iain\n\nDifferential Revision: https://phabricator.services.mozi... | |
sm__1989978 | sm | 1989978 | hwiwonlee/sm.x86_64:1989978 | /src/gecko-dev | /out/js | --no-threads --ion-warmup-threshold=100 | [
"js/src/vm/Iteration.cpp"
] | [
"js/src/vm"
] | Type confusion | ASAN_CRASH | for-in iterator optimization type confusion - optimized iteration allowed writes to non-writable iterator indices to be compiled as direct slot stores. The vulnerable optimization was in js/src/vm/Iteration.cpp. A JSON.rawJSON object could be warmed so a for-in loop overwrote its non-writable rawJSON string slot with a... | [
"From fa91ed58b791e04279b8fbfd6476550dd1fcb27b Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Fri, 26 Sep 2025 20:40:46 +0000\nSubject: [PATCH] Bug 1989978: Don't support unwritable iterator indices\n r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D266023\n--... | |
sm__1992130 | sm | 1992130 | hwiwonlee/sm.x86_64:1992130 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/wasm/WasmGcObject.h"
] | [
"js/src/wasm"
] | Stack buffer overflow | ASAN_CRASH | Stack buffer overflow in SpiderMonkey's WebAssembly GC fromCharCodeArray path. CVE-2025-13016 (sec-high, MFSA 2025-87, Firefox 145). StableWasmArrayObjectElements::StableWasmArrayObjectElements() in js/src/wasm/WasmGcObject.h used std::copy(array->inlineStorage(), array->inlineStorage() + array->numElements_ * sizeof(T... | [
"From 75ef3f79aae344c1d0398787cab666a6a73176c7 Mon Sep 17 00:00:00 2001\nFrom: Yury Delendik <ydelendik@mozilla.com>\nDate: Wed, 15 Oct 2025 20:02:23 +0000\nSubject: [PATCH] Bug 1992130 - Fix arguments types for std::copy. r=rhunt\n\nDifferential Revision: https://phabricator.services.mozilla.com/D267284\n---\n js/... | |
sm__1992902 | sm | 1992902 | hwiwonlee/sm.x86_64:1992902 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off --ion-warmup-threshold=30 | [
"js/src/jit/MIR.h",
"js/src/jit/MIROps.yaml",
"js/src/jit/LIROps.yaml",
"js/src/jit/ScalarReplacement.cpp",
"js/src/jit/WarpCacheIRTranspiler.cpp"
] | [
"js/src/jit"
] | Use-after-free | ASAN_CRASH | Use-after-free in SpiderMonkey IonMonkey caused by incorrect scalar replacement of TypedArray.prototype.subarray(). CVE-2025-13024 (sec-high, MFSA 2025-87, Firefox 145). MTypedArraySubarray was statically marked alias_set=none and can_recover=true, so Ion treated every subarray() call as a pure, hoistable operation wit... | [
"From de49b52ff5f912aa3a7ea7c570ee1aafcde53fa6 Mon Sep 17 00:00:00 2001\nFrom: =?UTF-8?q?Andr=C3=A9=20Bargull?= <andre.bargull@gmail.com>\nDate: Thu, 30 Oct 2025 21:15:52 +0000\nSubject: [PATCH] Bug 1992902: Track scalar replacement for subarray. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla... | |
sm__1998050 | sm | 1998050 | hwiwonlee/sm.x86_64:1998050 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/CacheIR.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | Baseline-JIT type confusion in SpiderMonkey's CacheIR AddSlot attach path when the index is a huge typed-array out-of-bounds value. CVE-2025-14325 (sec-high, MFSA 2025-92, Firefox 146). canAttachAddSlotStub in js/src/jit/CacheIR.cpp previously lacked a check for typed-array OOB indices >= 2^32, so an AddSlot stub was a... | [
"From 49150a9a555a73e21dc20181c204b6003b17c167 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Tue, 11 Nov 2025 16:04:27 +0000\nSubject: [PATCH] Bug 1998050 - Check for typed array index in\n canAttachAddSlotStub. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla.com/D... | |
sm__2003588 | sm | 2003588 | hwiwonlee/sm.x86_64:2003588 | /src/gecko-dev | /out/js | --ion-offthread-compile=off | [
"js/src/proxy/CrossCompartmentWrapper.cpp"
] | [
"js/src/proxy"
] | Cross-compartment violation | ASAN_CRASH | Debugger + cross-compartment wrapper UAF / assertion after nukeAllCCWs. CVE-2026-0884 (MFSA 2026-01). Pre-fix, Compartment::wrap in js/src/proxy/CrossCompartmentWrapper.cpp refused to (re)create CCWs to debugger instances after nukeAllCCWs(), but the Debugger machinery continued to drive debug events (debugger statemen... | [
"From 0264cf850afd91cdd3956066e0b00b982566f684 Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Mon, 8 Dec 2025 13:11:30 +0000\nSubject: [PATCH] Bug 2003588 - Continue to allow creation of CCWs to debugger\n instances after CCWs have been nuked r=jandem\n\nIt turns out that disallowing th... | |
sm__2009303 | sm | 2009303 | hwiwonlee/sm.x86_64:2009303 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/debugger/DebugScript.cpp"
] | [
"js/src/debugger"
] | Use-after-free | ASAN_CRASH | DebugScript Heap-Use-After-Free on About-To-Be-Finalized JSScript. CVE-2026-4701 (sec-moderate, MFSA 2026-20). During a GC sweep SpiderMonkey marks JSScripts as 'about to be finalized' before their destructors actually run. Several DebugScript helpers in js/src/debugger/DebugScript.cpp — DebugScript::get, DebugScript::... | [
"From 0a30c6b6fc37e95465fbf3c325a97f59ede662dd Mon Sep 17 00:00:00 2001\nFrom: Tooru Fujisawa <arai_a@mac.com>\nDate: Tue, 13 Jan 2026 04:48:42 +0000\nSubject: [PATCH] Bug 2009303 - Check the corresponding script before trying to\n get DebugScript. r=jonco\n\nDifferential Revision: https://phabricator.services.mozi... | |
sm__2010940 | sm | 2010940 | hwiwonlee/sm.x86_64:2010940 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/builtin/FinalizationRegistryObject.cpp",
"js/src/builtin/FinalizationRegistryObject.h",
"js/src/gc/FinalizationObservers.cpp"
] | [
"js/src/builtin",
"js/src/gc"
] | Use-after-free | ASAN_CRASH | FinalizationRegistry Dead-Registry Queue Heap-Use-After-Free. Bug 2010940 (embargoed, likely sec-high; CVE pending). js::FinalizationObservers::traceWeakFinalizationRegistryEdges in js/src/gc/FinalizationObservers.cpp walks the per-zone finalization-registry weak map during GC. For each live registry it promotes every ... | [
"From 2e29a3d3d930453174717fd963413492a80b31ec Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Mon, 26 Jan 2026 12:50:32 +0000\nSubject: [PATCH] Bug 2010940 - Clear queued objects for dead finalization\n registries during sweep r=sfink\n\nWhat's happening here is that a FinalizationQueue... | |
sm__2010943 | sm | 2010943 | hwiwonlee/sm.x86_64:2010943 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/jit/JitFrames.cpp",
"js/src/jit/Snapshots.cpp",
"js/src/jit/Snapshots.h",
"js/src/jit/shared/CodeGenerator-shared.cpp"
] | [
"js/src/jit",
"js/src/jit/shared"
] | Out-of-bounds read | ASAN_CRASH | Ion Bailout RValueAllocation — Missing INT64_INT32_STACK Mode, BigInt-From-Int64 Recovery Reads 64 Bits From a 32-Bit Stack Slot → Adjacent-Stack Information Disclosure. CVE-2026-2783 (sec-high). Ion's bailout infrastructure represents the live value in each MIR slot via an `RValueAllocation` mode (`js/src/jit/Snapshot... | [
"From 0ba3af42be7fcfb0773fbf057c5b67277d3aa399 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Tue, 20 Jan 2026 23:44:42 +0000\nSubject: [PATCH] Bug 2010943: Support reading int64 from int32 stack slot\n r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D279595\n... | |
sm__2011069 | sm | 2011069 | hwiwonlee/sm.x86_64:2011069 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/gc/GC.cpp"
] | [
"js/src/gc"
] | Race condition | ASAN_CRASH | GCRuntime::setGrayBitsInvalid — Symbol Arena Iteration Races Background Sweeper Mutating the Same Arena List (OOM Incremental Reset Path). CVE-2026-2802 (sec-high). js::gc::GCRuntime::setGrayBitsInvalid (js/src/gc/GC.cpp) is called from js::gc::GCRuntime::resetIncrementalGC when an OOM interrupts an in-progress increme... | [
"From 34384ff7e250e7bdf10717337eb689c87138863c Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Thu, 22 Jan 2026 08:52:57 +0000\nSubject: [PATCH] Bug 2011069 - Wait for any background sweeping to finish\n before marking gray bits invalid r=jandem\n\nMarking gray bits invalid iterates the ... | |
sm__2012018 | sm | 2012018 | hwiwonlee/sm.x86_64:2012018 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/vm/Iteration.cpp"
] | [
"js/src/vm"
] | Use-after-free | ASAN_CRASH | Use-after-free in SpiderMonkey's for..in iteration helper SuppressDeletedProperty (js/src/vm/Iteration.cpp). CVE-2026-2763 (sec-high, MFSA 2026-13, Firefox 148). When a property is deleted during an active for..in enumeration, SuppressDeletedProperty walks the live NativeIterator list to hide that key from in-progress ... | [
"From 2292742e2bba10cc7f56ff2f0f89f29e6f949b04 Mon Sep 17 00:00:00 2001\nFrom: Jan de Mooij <jdemooij@mozilla.com>\nDate: Thu, 29 Jan 2026 12:18:37 +0000\nSubject: [PATCH] Bug 2012018 - Simplify property lookup in\n SuppressDeletedProperty. r=iain\n\nDifferential Revision: https://phabricator.services.mozilla.com/D... | |
sm__2013165 | sm | 2013165 | hwiwonlee/sm.x86_64:2013165 | /src/gecko-dev | /out/js | --fuzzing-safe | [
"js/src/wasm/WasmInstance.cpp"
] | [
"js/src/wasm"
] | Type confusion | ASAN_CRASH | Wasm Import Optimization Type Confusion — MaybeOptimizeFunctionCallBind Accepts Wasm Function as boundThis. CVE-2026-2796 (sec-critical, MFSA 2026-13). js::wasm::MaybeOptimizeFunctionCallBind in js/src/wasm/WasmInstance.cpp:2316 implements the wasm import fast-path for the pattern `Function.prototype.call.bind(targetFn... | [
"From e2acef6711967949cd0825869034165383c482e1 Mon Sep 17 00:00:00 2001\nFrom: Ryan Hunt <rhunt@eqrion.net>\nDate: Thu, 5 Feb 2026 19:10:03 +0000\nSubject: [PATCH] Bug 2013165 - Fix import optimization. r=yury\n\nDifferential Revision: https://phabricator.services.mozilla.com/D281104\n---\n js/src/wasm/WasmInstance... | |
sm__2013543 | sm | 2013543 | hwiwonlee/sm.x86_64:2013543 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Incorrect JIT optimization | ASAN_CRASH | Ion OptimizeIteratorIndices used block-level dominance where instruction-level dominance was required. In js/src/jit/IonAnalysis.cpp, the pass tried to share Object.keys iterator state when another iterator's block dominated the current block. For two Object.keys operations in the same basic block, block()->dominates()... | [
"From bf98f12eb281d60b77984d9497a08209c69ad8d8 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Tue, 3 Feb 2026 21:43:28 +0000\nSubject: [PATCH] Bug 2013543: Strengthen dominance check r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D281543\n---\n js/src/jit-tes... | |
sm__2013549 | sm | 2013549 | hwiwonlee/sm.x86_64:2013549 | /src/gecko-dev | /out/js | --ion-offthread-compile=off | [
"js/src/vm/Scope.cpp",
"js/src/builtin/ModuleObject.cpp",
"js/src/vm/EnvironmentObject.h",
"js/src/vm/Modules.cpp"
] | [
"js/src/vm",
"js/src/builtin"
] | Out-of-bounds read | ASAN_CRASH | Invalid pointer in SpiderMonkey module linking. CVE-2026-2785 (MFSA 2026-15, Firefox 148). For cyclic (parsed) modules, Parser::moduleBody injects a hidden *namespace* binding into the module's scope so that re-exporting an `import * as ns` binding resolves to a slot on the module environment. Synthetic modules (JSON, ... | [
"From c8104bed78889aaebf43316c1ba6acf61bb3a8ba Mon Sep 17 00:00:00 2001\nFrom: Jon Coppeard <jcoppeard@mozilla.com>\nDate: Thu, 5 Feb 2026 17:22:21 +0000\nSubject: [PATCH] Bug 2013549 - Give synthetic module environments a\n *namespace* property the same as for cyclic modules\n r=spidermonkey-reviewers,dminor\n\nMo... | |
sm__2013560 | sm | 2013560 | hwiwonlee/sm.x86_64:2013560 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/debugger/Debugger.cpp"
] | [
"js/src/debugger"
] | Null pointer dereference | ASAN_CRASH | Debugger Force-Return at Generator Yield Opcode Null Generator Object Dereference. CVE-2026-4702 (sec-moderate, MFSA 2026-20). js::Completion::fromJSFramePop in js/src/debugger/Debugger.cpp builds a Completion from the current bytecode op of a popped frame. When the op is JSOp::InitialYield, JSOp::Yield, or JSOp::Await... | [
"From 1bd7ef2aa4bdb9e584d192049cf5cba23a1e8ace Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Fri, 13 Feb 2026 23:41:02 +0000\nSubject: [PATCH] Bug 2013560: Handle force return at yield opcodes r=mgaudet\n\nDifferential Revision: https://phabricator.services.mozilla.com/D283332\n---\n js... | |
sm__2013562 | sm | 2013562 | hwiwonlee/sm.x86_64:2013562 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/builtin/AtomicsObject.cpp"
] | [
"js/src/builtin"
] | Cross-compartment violation | ASAN_CRASH | Atomics.notify cross-realm promise resolution bug - js::atomics_notify_impl in js/src/builtin/AtomicsObject.cpp walks the waiter list from a SharedArrayBuffer and resolves each waitAsync promise without first entering the promise's realm. Waiter lists can contain promises from multiple realms, so Atomics.notify in one ... | [
"From 23cf0b3985c329aae2898bae518a7f5813c132ae Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Mon, 9 Feb 2026 21:40:57 +0000\nSubject: [PATCH] Bug 2013562: Enter correct realm before resolving same-thread\n waitAsync promise r=arai\n\nDifferential Revision: https://phabricator.services.m... | |
sm__2013741 | sm | 2013741 | hwiwonlee/sm.x86_64:2013741 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/wasm/WasmBaselineCompile.cpp"
] | [
"js/src/wasm"
] | Use-after-free | ASAN_CRASH | WASM GC array.fill Missing Pre-Write Barrier on Reference Element Store. CVE-2026-2767 (sec-high, MFSA 2026-15 — Use-after-free in the JavaScript: WebAssembly component). js::wasm::BaseCompiler::emitArrayFill in js/src/wasm/WasmBaselineCompile.cpp:8801 emits the per-element store of a WASM GC `array.fill` instruction b... | [
"From d867a6f69b9f026674f2e846572748818a61a41b Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Mon, 9 Feb 2026 16:45:43 +0000\nSubject: [PATCH] Bug 2013741: Clean up misleading comments in array.fill.\n r=rhunt\n\nThe comments were pulled from a different array method and were slightly\nmi... | |
sm__2019813 | sm | 2019813 | hwiwonlee/sm.x86_64:2019813 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/wasm/WasmValidate.cpp"
] | [
"js/src/wasm"
] | Out-of-bounds read | ASAN_CRASH | WebAssembly Imported Memory importIndex Off-By-One in AddImport. CVE-2026-2785 (sec-moderate). After a refactor that added compact wasm imports, AddImport() in js/src/wasm/WasmValidate.cpp records an imported memory's `importIndex` AFTER calling `moduleMeta->imports.emplaceBack(...)`: `codeMeta->memories.back().importI... | [
"From 9ef71fcf90d12a656c362e4af21e3026bbdcb15f Mon Sep 17 00:00:00 2001\nFrom: Ben Visness <bvisness@mozilla.com>\nDate: Fri, 27 Feb 2026 21:07:35 +0000\nSubject: [PATCH] Bug 2019813: Fix off-by-one in imported memory index. r=rhunt\n\nAfter rearranging the import process for compact imports, the\nimportIndex prope... | |
sm__2023024 | sm | 2023024 | hwiwonlee/sm.x86_64:2023024 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/CacheIR.cpp"
] | [
"js/src/jit"
] | Type confusion | ASAN_CRASH | CacheIR ObjectFuse Attached to Reserved Slot — Stale Cached Value After Engine-Internal Reserved-Slot Write. Bug 2023024 (embargoed, likely sec-high; CVE pending). js::jit::IRGenerator::canOptimizeConstantDataProperty in js/src/jit/CacheIR.cpp:2248 decides whether to attach a CacheIR 'ConstantDataProperty' stub that ca... | [
"From 0d8fa127358980af62ad70abdcba632f0c683a91 Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Mon, 16 Mar 2026 16:07:16 +0000\nSubject: [PATCH] Bug 2023024: Don't use fuses for reserved slots r=jandem\n\nDifferential Revision: https://phabricator.services.mozilla.com/D287740\n---\n js/sr... | |
sm__2029065 | sm | 2029065 | hwiwonlee/sm.x86_64:2029065 | /src/gecko-dev | /out/js | --fuzzing-safe --ion-offthread-compile=off | [
"js/src/jit/IonAnalysis.cpp"
] | [
"js/src/jit"
] | Incorrect JIT optimization | ASAN_CRASH | Ion OptimizeIteratorIndices queried dominance information on freshly-created MIR blocks whose dominator metadata had not been computed yet. This follow-up to Bug 2013543 is rooted in js/src/jit/IonAnalysis.cpp: after the iterator-index optimization split or created blocks, the same pass could call MBasicBlock::dominate... | [
"From 5cfc09e7ce19a151df5651a24ebd250632aac10f Mon Sep 17 00:00:00 2001\nFrom: Iain Ireland <iireland@mozilla.com>\nDate: Fri, 10 Apr 2026 16:08:47 +0000\nSubject: [PATCH] Bug 2029065: Skip new blocks in OptimizeIteratorIndices\n r=jandem\n\nI considered trying to create precise dominator input here, but given the ... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.