Bug 251667

Summary:	Implement a more accurate currentStackPointer().
Product:	WebKit	Reporter:	Mark Lam <mark.lam>
Component:	JavaScriptCore	Assignee:	Mark Lam <mark.lam>
Status:	RESOLVED FIXED
Severity:	Normal	CC:	berto, bugs-noreply, Hironori.Fujii, mcatanzaro, webkit-bug-importer
Priority:	P2	Keywords:	InRadar
Version:	WebKit Nightly Build
Hardware:	Unspecified
OS:	Unspecified
See Also:	https://bugs.webkit.org/show_bug.cgi?id=252196

Description Mark Lam 2023-02-02 22:49:30 PST

currentStackPointer() was already accurate for some ports (the ones that support inline assembly), but for others and when using a Debug build, we fall back to a C runtime function that only gives an approximate value.  This patch implements a currentStackPointer() that is accurate for all currently support ports even when using a Debug build.

Comment 1 Radar WebKit Bug Importer 2023-02-02 22:49:54 PST

<rdar://problem/104993866>

Comment 2 Mark Lam 2023-02-02 23:00:33 PST

Pull request: https://github.com/WebKit/WebKit/pull/9585

Comment 3 EWS 2023-02-03 08:34:11 PST

Committed 259818@main (7dbed20e73b0): <https://commits.webkit.org/259818@main>

Reviewed commits have been landed. Closing PR #9585 and removing active labels.

Comment 4 Fujii Hironori 2023-02-05 18:11:26 PST

Bug 251770 – [Win] REGRESSION(259818@main) error LNK2019: unresolved external symbol currentStackPointer

Comment 5 Michael Catanzaro 2023-02-12 05:56:19 PST

OK, this will break Fedora due to missing required architectures ppc64le and s390x, and will likely be worse for Debian since they have tons of architectures. I'll look closer next week, but ideally we would fall back to generic/inaccurate implementation rather than failing to build.

Comment 6 Michael Catanzaro 2023-02-13 14:39:37 PST

I found a simple typo causing the build failure. Fix incoming in bug #252196.