This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: [PATCH 1/2] arm: Implement memcpy ifunc selection in C
On Mon, 9 Oct 2017, Ramana Radhakrishnan wrote:
> > That is, we should fix the default in GCC mainline and then propose glibc
> > changes that do not involve such workarounds. (Arguably the default
> > should be to support the attribute for all architectures, to avoid future
> > issues when new architectures gain binutils support for IFUNCs after the
> > GCC 8 release.)
>
> Is there a patch for this to mainline gcc - my quick searches did not
> find anything ? I see no reason for not defaulting to this on linux /
> glibc configurations in mainline gcc taking care not to break things
> for people who use musl or bionic. We've had ifuncs on aarch32 glibc /
> binutils for a few years now IIRC so I'm happy to review a patch on
> gcc-patches if someone can point me at it or propose a patch ....
My view is that we should default to having the support in GCC for all
architectures with glibc, whether or not binutils/ABI support exists right
now, to avoid unnecessarily requiring a GCC update as well as
binutils/glibc ones to use it when support is added on a new architecture.
But that would be most friendly to users if the assembler then rejects
IFUNC creation in the absence of ABI/linker support.
--
Joseph S. Myers
joseph@codesourcery.com