Permalink
Comparing changes
Choose two branches to see what’s changed or to start a new pull request.
If you need to, you can also or
learn more about diff comparisons.
Open a pull request
Create a new pull request by comparing changes across two branches. If you need to, you can also .
Learn more about diff comparisons here.
...
- 19 commits
- 86 files changed
- 7 contributors
Commits on Aug 26, 2024
-
Merged: [wasm] Lower kMaxCanonicalTypes again
This reverts part of 2b43121, because there are still some ValueType uses of canonicalized type indices left. So for now we allow a maximum of 1'000'000 canonicalized types. Fixed: 360533914 (cherry picked from commit 79f3f12) Change-Id: I8578d4b77233e7f3815dbf5ec6335f2c4cef963a Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5813422 Commit-Queue: Matthias Liedtke <[email protected]> Auto-Submit: Jakob Kummerow <[email protected]> Commit-Queue: Jakob Kummerow <[email protected]> Reviewed-by: Matthias Liedtke <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#44} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
-
Revert ScopeInfo reuse work on the 12.8 branch
It's still broken despite earlier attempts to fix it with merges. This reverts the following commits: "[compiler] Reuse ScopeInfos from previous compiles" (9cbd134) "[parser] Set max info id for class initializers" (7e06e46) "[parsing] Disable scopeinfo sharing for hidden scopes" (3f66bff) "[compiler] Put reuse_scope_infos behind a disabled flag" (9e36f73) "[parsing] better disambiguate with scopes" (479fdc1) "[parser] Fix eval handling in scope info reuse" (774d5a6) "[streaming] Use UniqueIdInScript instead of StartPosition" (eeba676) "[parsing] Reuse scope infos held alive through eval" (cc09cc9) "[parser] hidden non-catch scopes can safely cause reuse" (f3a13e2) "[compiler] Skip scope infos on "new scripts"" (3155058) "[scopeinfo] Don't try to reuse the native context's script scope info" (7d4bea5) Fixed: 361549672 Change-Id: I2c66e0d7c1655e5ceb7abd4a59f42e84b468e250 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5811749 Reviewed-by: Leszek Swirski <[email protected]> Commit-Queue: Jakob Kummerow <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#46} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Commits on Aug 30, 2024
-
Merged: [wasm] Don't catch uncatchable exceptions in the JSPI wrapper
... And forward the exception to the parent stack instead. [email protected] Fixed: 361717714 (cherry picked from commit 9495e79) Change-Id: I8124ae6d310aa59c98623eaebcfcabffb6902a69 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5828824 Reviewed-by: Jakob Kummerow <[email protected]> Commit-Queue: Thibaud Michaud <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#48} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Commits on Sep 3, 2024
-
Merged: [top-level-await] fix async evaluation with errored async parent
Updates a CHECK to support EVALUATING-ASYNC modules. Updates GatherAvailableAncestors to check kErrored before calling AsyncEvaluationOrdinalCompare to prevent hitting DCHECK. Bug: 361615555 Bug: 361603701 Bug: 361611791 (cherry picked from commit 5af53ec) Change-Id: Ib4917468ca911dc2bfc6f5579846e0dd17ee41b8 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5833740 Reviewed-by: Deepti Gandluri <[email protected]> Commit-Queue: Deepti Gandluri <[email protected]> Auto-Submit: Shu-yu Guo <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#52} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Commits on Sep 4, 2024
-
Merged: [wasm] Skip WasmJSFunctions in js-to-wasm wrapper tier-up
When a js-to-wasm wrapper tiers up, we also set the newly compiled wrapper as the target for other exports that have the same signature. This assumed that all exports have type WasmExportedFunction, but they can also have type WasmJSFunction in the case of a re-exported WebAssembly.Function import. [email protected] Fixed: 362539773 (cherry picked from commit 7860c96) Change-Id: Id685ae73a84121701729178a3a6f28dbc39648df Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5835724 Reviewed-by: Clemens Backes <[email protected]> Commit-Queue: Thibaud Michaud <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#54} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
-
Merged: [compiler][wasm] Fix TypeGuard on br_on_cast with null type
Similar pattern to https://crrev.com/c/5826664 . (cherry picked from commit c42a07f) Fixed: 361862737 Change-Id: I59c3cea5d572423aa06583a1448bf7486213a233 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5829361 Reviewed-by: Jakob Kummerow <[email protected]> Auto-Submit: Matthias Liedtke <[email protected]> Commit-Queue: Matthias Liedtke <[email protected]> Commit-Queue: Jakob Kummerow <[email protected]> Cr-Original-Commit-Position: refs/heads/main@{#95899} Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5833495 Reviewed-by: Manos Koukoutos <[email protected]> Commit-Queue: Eva Herencsárová <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#56} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
-
Merged: [compiler][wasm] Fix endless loop in WasmTyper
The graph builder was applying a TypeGuard into the wrong control chain. In this case the br_on_non_null implementation added a TypeGuard for the non-null type before performing the actual branch. The br_on_non_null target is anyways going to merge controls with some other branch and Turboshaft is hopefully making all this code oboslete soon, so the easiest and safest fix is to drop the TypeGuard completely. (cherry picked from commit 18e6776) Fixed: 361862737 Change-Id: Ifc1c34ab726576b861d3d5dc6f6a9d20e93d4af0 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5826664 Auto-Submit: Matthias Liedtke <[email protected]> Reviewed-by: Jakob Kummerow <[email protected]> Commit-Queue: Jakob Kummerow <[email protected]> Cr-Original-Commit-Position: refs/heads/main@{#95897} Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5833494 Reviewed-by: Manos Koukoutos <[email protected]> Commit-Queue: Eva Herencsárová <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#57} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Commits on Sep 5, 2024
-
Merged: [maglev] Check value before inlining Object builtin
Fixed: 363983041 (cherry picked from commit dc03a6d) Change-Id: Ib3a4532d57192ac3952e0f80d57a2c1dcb20f204 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5837273 Reviewed-by: Jakob Linke <[email protected]> Commit-Queue: Igor Sheludko <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#59} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Commits on Sep 6, 2024
-
Merged: [maglev] Record phi use as smi when eliding a write barrier
We can only elide the write barrier under the assumption that the phi's type will remain a Smi. If we don't record its tagged usage at this bytecode offset, the phi representation selector might later change its representation to int32. It will then emit a `Int32ToNumber` before `StoreTaggedFieldNoWriteBarrier`. This is incorrect because it could allocate a heap number, potentially promoting it to old space during a GC and requiring a write barrier. Fixed: 362784006 (cherry picked from commit 165dd87) Change-Id: I7ccfb250847eb8cd3ad0e72081663476ada27558 Reviewed-on: https://chromium-review.googlesource.com/c/v8/v8/+/5840209 Reviewed-by: Michael Lippautz <[email protected]> Commit-Queue: Patrick Thier <[email protected]> Cr-Commit-Position: refs/branch-heads/12.8@{#61} Cr-Branched-From: 70cbb39-refs/heads/12.8.374@{#1} Cr-Branched-From: 451b63e-refs/heads/main@{#95151}
Loading
This comparison is taking too long to generate.
Unfortunately it looks like we can’t render this comparison for you right now. It might be too big, or there might be something weird with your repository.
You can try running this command locally to see the comparison on your machine:
git diff 8a8a71d...6f774f9