On 6/29/07, Wojtek.Pilorz wpilorz@bdk.pl wrote:
On Fedora6, I have noted the folowing problems when using cifs to connect to NT shares:
- cp -vip does not copy mtime; tar x does copy mtime (it seems setting modtime by fname does work, setting by handle does not)
- when there is a share mounted from a NT/2000 server which is not used for long time mounting another share from the same servers reliably fails ; touching the already mounted share (ls is OK) allows the mount so succeed.
- when using vim to edit a file on the share and trying to write it, vim often says that the file has been modified in the meantime
- perhaps most annoying: somtimes modifying a file located on NT 4 Workstation share with vim, it becomes permanently unavailable until NT is rebooted. It happens rarely, I do not know how to trigger it on demand.
Those problems did not happen when using smbfs from CentOS 4.4.
I am wondering if CentOS 5 has these problems? Anyone using cifs from CentOS 5 to access NT-exported share, would care to comment?
Here is the updated info.
Steve French (samba team) made available a newer cifs version (1.48) backported to old kernels:
http://lists.samba.org/archive/linux-cifs-client/2007-April/001898.html
Using this cifs code Jeff Layton (RH) has built some test kernels after making appropriate adjustment for RHEL:
http://people.redhat.com/jlayton/
You can try the test kernel to see if the problems have been fixed, or if you would rather not use any test kernel, you can just get the cifs module compiled. I have compiled a x86_64 version of cifs for CentOS 5 for my own use and confirmed it has eliminated the bug that causes system crashes in kernel 2.6.18 (C5's default).
Akemi
On 7/8/07, Akemi Yagi amyagi@gmail.com wrote:
Here is the updated info.
Steve French (samba team) made available a newer cifs version (1.48) backported to old kernels:
http://lists.samba.org/archive/linux-cifs-client/2007-April/001898.html
Using this cifs code Jeff Layton (RH) has built some test kernels after making appropriate adjustment for RHEL:
FYI. A list of cifs bugs fixed in the above version (going from 1.45 to 1.48):
+Verison 1.48 +------------ +Fix mtime bouncing around from local idea of last write times to remote time. +Fix hang (in i_size_read) when simultaneous size update of same remote file +on smp system corrupts sequence number. Do not reread unnecessarily partial page +(which we are about to overwrite anyway) when writing out file opened rw. +When DOS attribute of file on non-Unix server's file changes on the server side +from read-only back to read-write, reflect this change in default file mode +(we had been leaving a file's mode read-only until the inode were reloaded). +Allow setting of attribute back to ATTR_NORMAL (removing readonly dos attribute +when archive dos attribute not set and we are changing mode back to writeable +on server which does not support the Unix Extensions). + +Version 1.47 +------------ +Fix oops in list_del during mount caused by unaligned string. +Fix file corruption which could occur on some large file +copies caused by writepages page i/o completion bug. +Seek to SEEK_END forces check for update of file size for non-cached +files. + +Version 1.46 +------------ +Support deep tree mounts. Better support OS/2, Win9x (DOS) time stamps. +Allow null user to be specified on mount ("username="). Do not return +EINVAL on readdir when filldir fails due to overwritten blocksize +(fixes FC problem). Return error in rename 2nd attempt retry (ie report +if rename by handle also fails, after rename by path fails, we were +not reporting whether the retry worked or not). Fix NTLMv2 to +work to Windows servers (mount with option "sec=ntlmv2").
-
Akemi Yagi