This is the mail archive of the
libc-alpha@sources.redhat.com
mailing list for the glibc project.
Re: Use of NEED_DL_SYSINFO
On Mon, Apr 21, 2003 at 06:54:22PM -0700, Ulrich Drepper wrote:
> > But everything (else) can be the same between the two! We get passed
> > the pointer via AT_SYSINFO, all the setup stuff in NPTL etc. Having
> > some other symbol seems a bit redundant.
> YOu are contradicting yourself. Evidently it is not redundant,
> otherwise you'd have no problems.
that's why i said everything "else".
> > Maybe ia32 could create
> > another symbol for it's non-generic parts?
>
> You're joking, right? Why should this be done? The symbol was
> introduced for the expressed purpose of using it for x86 with the x86
> semantics. You cannot just hijack the symbol.
because previous to that change it worked fine as a generic
impelmentation, and you have to admit the symbol has a fairly generic
name. differences likely to not be generically compatable could be
easily encased in some processor specific defines.
> Besides, the code using this symbol will be rearranged sometime soon
> anyway. So better hold of your patch until this is handled. Changes
> are that you then can piggy-bag your requirements.
ok so the argument is moot. will wait.
-i