This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: RFC: Optimize out i386/x86-64 JUMP_SLOT relocation
- From: Roland McGrath <roland at hack dot frob dot com>
- To: "H.J. Lu" <hjl dot tools at gmail dot com>
- Cc: binutils at sourceware dot org, GNU C Library <libc-alpha at sourceware dot org>
- Date: Mon, 24 Nov 2014 13:02:23 -0800 (PST)
- Subject: Re: RFC: Optimize out i386/x86-64 JUMP_SLOT relocation
- Authentication-results: sourceware.org; auth=none
- References: <20141123150719 dot GA2587 at intel dot com>
We could change the test to grok that case. We just need to find a way to
change it such that we can be confident that the internal calls in the
module are actually going via that GOT slot even though there is no JMPREL
reloc for it. We're already assuming that the presence of any PLT entry
for the symbol name means all calls are using the PLT. So it is probably
just as safe to assume that the presence of a GOT reloc for the symbol name
means calls are using that slot.