This is the mail archive of the libc-alpha@sourceware.org mailing list for the glibc project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Remaining ia64 bits in libc


It looks like the following ia64 bits are still in libc (but have a 
standard way in which ports targets do such things entirely within ports, 
so it's not appropriate for them to be in libc when the rest of the port 
was removed):

nptl/sysdeps/ia64 (directory)
nptl/sysdeps/unix/sysv/linux/ia64 (directory)
scripts/data/c++-types-ia64-linux-gnu.data

References to ia64 in: configure.in (ports targets have their own 
preconfigure scripts), elf/tls-macros.h (ports targets have their own 
versions of this header; also has alpha code), elf/tst-auditmod1.c (ports 
targets have tst-audit.h ... actually, it appears removal of alpha code 
from this file was forgotten as well), nptl/shlib-versions, shlib-versions 
(ports architectures have their own shlib-versions files ... it appears 
there are stray alpha and cris versions still in the libc copies as well 
as ia64), sysdeps/generic/ldsodefs.h (again, stray alpha references), 
sysdeps/unix/sysv/linux/configure.in (stray references for a lot of ports 
architectures: alpha, arm, hppa, m68k including a "m68*"), 
sysdeps/unix/sysv/linux/kernel-features.h.

(There are other __ia64__ conditionals as well, but some are in files that 
are shared with GCC, or may not have a good approach for 
architecture-specific definitions in ports; the above are all cases where 
there is a well-established approach for a port to have its own 
definitions - cases where I think it would be better for libc targets to 
follow the same approach as ports targets rather than having architecture 
conditionals in common files at all.)

-- 
Joseph S. Myers
joseph@codesourcery.com


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]