This is the mail archive of the
mailing list for the GCC project.
Re: [gnu] Darwin libgcc/assert fix, uncontroversial piece (need sh-linux testing)
Kaz Kojima <firstname.lastname@example.org> writes:
> Zack Weinberg <email@example.com> wrote:
>> 3) Delete config/sh/libgcc-std.ver. Create config/sh/libgcc-excl.ver,
>> and change config/sh/t-linux to use it. (Perhaps the exclusions
>> should be in config/sh/libgcc-glibc.ver instead of creating a new
>> file. Thoughts?)
>> This has been bootstrapped, with no regressions, on i686-linux and
>> powerpc-darwin. I do not have the ability to test sh-linux; I would
>> appreciate help on that score. I will not check this patch in without
>> hearing from SH maintainers first.
> I'm only an sh-linux maintainer but I'd like to reply because the file
> config/sh/libgcc-std.ver is used only with linux ports.
> I prefer libgcc-excl.ver rather than to push it into libgcc-glibc.ver
> because it's unrelated to glibc.
Fine with me. I speculated that it might be appropriate to combine
the two because sh/t-linux is the only fragment that uses either, but
that is arguably an accident of no one else having done SH shared
> I've tested your patch on x86 cross sh4-unknown-linux-gnu and found
> no problems. I've also started a native bootstrap on
> sh4-unknown-linux-gnu which will take a couple of hours, though I
> believe that the tests on the cross build are enough for this case.
Thank you. I will check the patch in later tonight.