This is the mail archive of the libc-hacker@cygnus.com 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]

some 2.0.110 reports


Hi,

2.0.110 doesn't compile on Sparc.  There is an undefined reference to 
__syscall_fork, presumably from the vfork code.  fork.S seems to define 
__libc_fork instead.  I guess one of the sparc hackers needs to sort out what 
has to happen with vfork.

On ARM, as Rod mentioned, all is OK bar one test failure which is due to an 
GCC bug.  In fact the code in vfscanf dealing with %hu formats is being 
miscompiled and this manifests itself in the timezone problem.  (The bug, if 
anybody is interested, is that if a secondary input reload requires a scratch 
register GCC can sometimes try to use the same register for the scratch as the 
final result is going into.  The alpha port has a workaround for this which 
essentially means it asks for two scratch registers and then uses the one that 
doesn't conflict.  Maybe we can use something similar.)

p.




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