This is the mail archive of the
libc-alpha@sourceware.org
mailing list for the glibc project.
Re: Another GLIBC build error with GCC6
- From: Richard Henderson <rth at twiddle dot net>
- To: Paul Eggert <eggert at cs dot ucla dot edu>, GNU C Library <libc-alpha at sourceware dot org>
- Cc: sellcey at imgtec dot com
- Date: Tue, 4 Aug 2015 22:04:02 -0700
- Subject: Re: Another GLIBC build error with GCC6
- Authentication-results: sourceware.org; auth=none
- References: <55C12065 dot 8080101 at twiddle dot net> <55C15492 dot 4090703 at cs dot ucla dot edu>
On 08/04/2015 05:10 PM, Paul Eggert wrote:
On 08/04/2015 01:28 PM, Richard Henderson wrote:
I believe you can just change L to UL and it should all Just Work.
Yes, though it's still confusing code. A better approach is in:
https://sourceware.org/ml/libc-alpha/2015-07/msg00713.html
Sure.
Probably a formulation like #define DT_EXTRATAGIDX(tag) ((tag) - (DT_HIPROC -
DT_EXTRANUM)) would be a lot better. And understandable, from a "what the
hell is this trying to do" standpoint. r~
Right idea, though that particular formulation looks backwards -- see Andreas's
suggestion as codified in:
https://sourceware.org/ml/libc-alpha/2015-07/msg00742.html
It isn't backward, in that it retains the current indexing.
If reversing the indexing is considered ok, then that's (sightly) better. But
even then, using DT_FILTER isn't best, whereas DT_HIPROC is.
r~