Re: Architecture specific interfaces
Stuart Anderson <anderson@metrolink.com> writes:
> With Matt's help, I was able to get the results of running lsblibchk on
> a coupld of non-IA32 platforms. The purpose of doing this is to try and
> identify things that we have included as being common to all platform,
> but which are really ia32 specific. A couple of interesting things did
> show up
>
> A-64:
> ------
> Checking symbols in /lib/libc.so.6.1
> Didn't find __divdi3 in libc.so.6.1
That's ok.
> Didn't find res_init in libc.so.6.1
Drop res_init.
> Didn't find strtoq in libc.so.6.1
> Didn't find strtouq in libc.so.6.1
> Didn't find wcstoq in libc.so.6.1
> Didn't find wcstouq in libc.so.6.1
Those should be there - you seem to be using an older version of
glibc, this was fixed on 2001-01-03.
> Alpha:
> ------
> Checking symbols in /lib/libc.so.6.1
> Didn't find __divdi3 in libc.so.6.1
> Didn't find imaxdiv in libc.so.6.1
imaxdiv should be there - it's not there because of a bug on alpha, I
just fixed that bug in glibc (fix will be in 2.2.3).
> Checking symbols in /lib/libm.so.6.1
> Didn't find fmal in libm.so.6.1
> Didn't find nexttowardl in libm.so.6.1
>
> First off, I think the libc on my IA-64 box is kindof old, so I suspect that
> the last 5 symbols may have been a result of not being up to date.
>
> __divdi3 is not found on either of these platforms. It looks like the sort
> of thing that might be IA-32 specific, so I have marked it as being unique
> to the IA32.
Ok.
> The Alpha is missing a couple of other symbols: imaxdiv, fmal, and nexttowardl.
> On the IA-64 & IA-32, these are weakly defined. Since they appear on at
> least two other platforms, I don't think that they are IA-32 specific, but
> because they are weakly defined, I wonder if maybe we should just drop these
> interfaces?
Don't drop them, they're required by ISO C99. I'll check with the
other glibc developers why fmal and nexttowardl are missing.
> Does anyone else know anything more specific about the interfaces in question?
>
> Does anyone else have an opinion on wether we should keep them or not?
Andreas
--
Andreas Jaeger
SuSE Labs aj@suse.de
private aj@arthur.inka.de
http://www.suse.de/~aj
Reply to: