Re: go for 2.1

[Andreas Jaeger <aj@arthur.rhein-neckar.de>]

>>>>> Ulrich Drepper writes:

 > Hi,
 > The last (non-technical) problem preventing a glibc 2.1 release is now
 > gone and we can plan the release freely now.
Yeah!;-)

 > I'll try to make a new test release today and given the many geeks who
 > are bored in the upcoming four days without work we might get some
 > results soon so that we can shoot for a release next week or the other
 > week.


 > This is now the time you should speek up if something important is
 > missing.  I don't count the PIC-SHARED thing since this is only
 > renaming a variable and introducing optimizations for MIPS.  Given
 > that MIPS is not really supported anyway (it might compile but isn't
 > really tested) I don't consider it a loss.

 > Here is what I understand the current status is.  Please correct me:

 > - Linux/ix86: compiles, tests ok

 > - Linux/m68k: compiles, tests ok

 > - Linux/Alpha: compiles, tests ok

 > - Linux/PPC: compiles, tests ok

 > - Linux/ArmELF: compiles, tests ok

 > - Hurd/ix86: compiles, tests ok


 > _ Linux/MIPS: ???

In a nutshell:  doesn't compile yet.

I've started on monday building some cross tools for i486-linux ->
mips-linux and encountered quite a number of problems crosscompiling
glibc 2.1.  I hope (but don't know exactly since I'm waiting for
answers) that Ralf has some patches for it - otherwise I'll try to
generate some.  AFAIK the MIPS binutils don't support symbol
versioning yet which makes glibc 2.1 on MIPS IMHO quite useless.  I
don't know yet how far the port is away from successful
compiliation:-(.

 > - Linux/SPARC: ???

 > - Linux/SPARC64: ???


 > Any target I missed?

 > Please also mention if special gcc/egcs/binutils versions are required.

Andreas
-- 
 Andreas Jaeger   aj@arthur.rhein-neckar.de    jaeger@informatik.uni-kl.de
  for pgp-key finger ajaeger@aixd1.rhrk.uni-kl.de