Sven, As I'm stuck home today (unusual snow at Bordeaux, France), I made some tests:
1. time ls -l > /dev/null > /dev/null (with a 10 thousands files in current directory over NFS) a) With coreutils 8.5-1 real 0m4.242s user 0m0.132s sys 0m0.504s b) With coreutils devel 8.15 real 0m4.162s user 0m0.116s sys 0m0.392s 2. strace -c ls -l > /dev/null (on the same directory) a) With coreutils 8.5-1 % time seconds usecs/call calls errors syscall ------ ----------- ----------- --------- --------- ---------------- 60.27 0.044688 4 10000 lstat 37.77 0.028002 2546 11 getdents 1.18 0.000876 0 20000 20000 getxattr 0.65 0.000480 0 10000 10000 lgetxattr 0.13 0.000095 0 9999 stat 0.00 0.000000 0 19 read 0.00 0.000000 0 122 write [...] b) With coreutils devel 8.15 % time seconds usecs/call calls errors syscall ------ ----------- ----------- --------- --------- ---------------- 65.22 0.037658 4 10000 lstat 34.64 0.020002 1818 11 getdents 0.13 0.000076 0 9999 stat 0.00 0.000000 0 13 read 0.00 0.000000 0 122 write [...) => First, it looks like that the devel version does not call getxattr() and lgetxattr() when unneeded. Second, the getxattr and lgetxattr syscall are not the cause of the speed problem you talk about. Cheers, Jérémy
