On Tue, Jan 06, 2004 at 05:52:05PM -0500, Nathanael Nerode wrote: > >kdeutils > >-------- > >mips - failed - needs the SYS_sysinfo fix [1] > ... > >[1] - Was broken after supposed toolchain freeze by f*cking wonderful > >toolchain changes. > > Yeah, that one sucks a lot; I've noted some other packages it bites in the > appropriate bug trail. That one showed up in a supposed bug-fix-only release > of glibc, which is especially galling. glibc has been the most troublesome > package for a very long time now. > > I've been meaning to try to work on that bug myself,and not having a MIPS, it > seemed it might not be the best use of my time, so I've been fighting with > (upstream) GCC 3.4 configury issues instead.
Its trivial to fix just need to substitute in sysinfo() instead. The same thing had to be done for kdebase. I didn't notice kdeutils had that problem until I was writing the status update last night. > >[0] - > > > >The failure above seem to be caused by: > > > >mpeglib/lib/input/cdromAccess_Linux.cpp:13 > > > >typedef unsigned long long __u64; > > > >Anyone happen to know if its safe to remove that or some other easy way to > >work around it? > > The 'conflicting' definition is "typedef long unsigned int __u64" from /usr/ > include/asm/types.h (included via linux/cdrom.h). > > Obviously __u64 is meant to be an unsigned 64-bit integer in both cases. > > It looks like the 'other' (asm/types.h) definition is included in the same way > from the same place on all arches. If this is in fact the case, then it > should be safe to remove the cdromAccess_Linux.cpp definition. Sounds good then, I wasn't sure if simply removing that line would be enough or not. Hopefully there aren't any other alpha/ia64 bugs lurking behind it. Chris
signature.asc
Description: Digital signature
