[Webkit-unassigned] [Bug 222959] [CMake] Build fails on RISC-V with GCC 11

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=222959

--- Comment #21 from Khem Raj <raj.khem at gmail.com> ---
(In reply to Daniel Kolesa from comment #20)
> (In reply to Khem Raj from comment #13)
> > (In reply to Daniel Kolesa from comment #12)
> > > the test is not correct, the undefined reference was
> > > `__atomic_compare_exchange_1` while the one you're testing is
> > > `__atomic_fetch_add_1`
> > 
> > yes, although I saw that none of 1byte ops were not available as builtins
> > since the test was simple for __atomic_fetch_add_1, I went with it. but
> > perhaps using __atomic_compare_exchange_1 for test would be more accurate.
> 
> well, the reason i'm saying this is mainly that IMO we should be testing for
> the functions that are actually missing from the build
> 
> gcc claims that the hardware support for less-than-4-byte-atomics is partial
> in risc-v, idk what it means in practice though (haven't looked into the
> cross-compiler output more deeply)

right, I have tweaked the test to check for __atomic_compare_exchange_1 in latest patch.

-- 
You are receiving this mail because:
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.webkit.org/pipermail/webkit-unassigned/attachments/20210312/4080feca/attachment-0001.htm>