On Sat, 03 Mar 2007 01:29:14 +0900 (JST), Atsushi Nemoto <anemo@xxxxxxxxxxxxx> wrote: > > For N32 and O32, kernel should be fixed anyway, but which syscall > > should be supported? And whether kernel or libc should take care of > > 'long long' issue? > > Then how about this absolutely untested patch? I had forgotten changing __NR_Linux_syscalls, etc. And I found readahead() has same problem, and sys32_readahead() and sys32_sync_file_range() does not match with glibc. I'll send a new patch, fixing them all in kernel side. --- Atsushi Nemoto
