Bug 245874
| Summary: | SIGILL early during WebProcess execution on ARM 32bit | ||
|---|---|---|---|
| Product: | WebKit | Reporter: | Nikolas Zimmermann <zimmermann> |
| Component: | WPE WebKit | Assignee: | Nikolas Zimmermann <zimmermann> |
| Status: | RESOLVED FIXED | ||
| Severity: | Blocker | CC: | bugs-noreply, mikhail, webkit-bug-importer |
| Priority: | P2 | ||
| Version: | WebKit Local Build | ||
| Hardware: | Other | ||
| OS: | Linux | ||
| See Also: | https://bugs.webkit.org/show_bug.cgi?id=245321 | ||
| Bug Depends on: | 245321 | ||
| Bug Blocks: | |||
Nikolas Zimmermann
Since a while I'm hunting a SIGILL during WebProcess execution in early stages (sometimes within BMalloc Scanvenger thread, sometimes in WPE specific WebKitExtensionManager).
I've bisected from beginning of September until now, and found 700ac83f17d8b756e640c6aaf20af02f9a6d08ce (https://bugs.webkit.org/show_bug.cgi?id=245321) to be the culprit.
The HUGE clang-tidy patch that cleans up bmalloc/WTF/JSC/...
I'm splitting it up locally into smaller pieces, only cleaning up bmalloc, only JSC, etc. to find the culprit.
I bet we're seeing a compiler bug with clang10. My build environment is targeting an embedded ARM 32bit device using WPE.
| Attachments | ||
|---|---|---|
| Add attachment proposed patch, testcase, etc. |
Nikolas Zimmermann
Closing as resolved, since the guilty patch was removed (https://github.com/WebKit/WebKit/commit/aabfacb8cf9ca4780d20ce2cca233f7988e994e2
).