Re: 00_rwsem-11, 2.4.4-ac11 and gcc-3(2001-05-14)

2001-05-21 Thread David Howells
The compiler should be now fixed in this respect, for both my stuff that's in the kernel and Andrea's desired replacement. The problem appears to have been triggered by having two "input+output" constraints (eg: "+r", "+m"). However, I can't test this because the head of the CVS trunk doesn't

Re: 00_rwsem-11, 2.4.4-ac11 and gcc-3(2001-05-14)

2001-05-21 Thread David Howells
The compiler should be now fixed in this respect, for both my stuff that's in the kernel and Andrea's desired replacement. The problem appears to have been triggered by having two input+output constraints (eg: +r, +m). However, I can't test this because the head of the CVS trunk doesn't seem to

00_rwsem-11, 2.4.4-ac11 and gcc-3(2001-05-14)

2001-05-18 Thread mirabilos
Andrea, I applied rwsem-11 (a bit by hand) to -ac11 and tried to compile. By changing CFLAGS_sys.o to -O (instead of -O2) as I read earlier I nearly could compile, it only barfed when it came to assemble the xaddl procedure by itself: static inline long rwsem_xchgadd(long value, long * count) {

00_rwsem-11, 2.4.4-ac11 and gcc-3(2001-05-14)

2001-05-18 Thread mirabilos
Andrea, I applied rwsem-11 (a bit by hand) to -ac11 and tried to compile. By changing CFLAGS_sys.o to -O (instead of -O2) as I read earlier I nearly could compile, it only barfed when it came to assemble the xaddl procedure by itself: static inline long rwsem_xchgadd(long value, long * count) {