Hi Wendy, Thanks for the fast answer, i have applied your patch, recompiled and its working fine now. The patch was already included on the 2.6.9-22 source, but it isnt applied by default. Regards, Jaime. 2006/2/9, Wendy Cheng <wcheng@xxxxxxxxxx>: > On Thu, 2006-02-09 at 00:53 +0100, Jaime Peñalba wrote: > > Hi, > > > > Im getting this error when trying to load gfs module: > > > > [root@apache2 ~]# modprobe gfs > > FATAL: Error inserting gfs > > (/lib/modules/2.6.9-5.ELsmp/kernel/fs/gfs/gfs.ko): Unknown symbol in > > module, or unknown parameter (see dmesg) > > gfs: Unknown symbol blockdev_direct_IO_cluster_locking > > This is a patch we added to fix a direct IO deadlock issue. We've tried > hard not to have dependency on base kernel but simply couldn't do it in > this case. > > The change has to pair with Red Hat RHEL 4 U3 kernel (>=2.6.9-25) that > is currently in Beta testing. > > If you can build your own base kernel, the base kernel patch is placed > in Red Hat bugzilla 173912 comment 4. If you can't read that bugzilla, > let me know. > > > Can i safely change this? > > > > return blockdev_direct_IO_cluster_locking(rw, iocb, inode, > > inode->i_sb->s_bdev, iov, > > offset, nr_segs, gb, NULL); > > > > with "return 1" or something like? Im going to use always lock_dlm. > > If you don't use Direct IO at all, this should be ok. > > > > > Anyway there is any other workaround? > > > > > > Let us know if above can't work out for you. > > -- Wendy > > -- > > Linux-cluster@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/linux-cluster > -- Linux-cluster@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/linux-cluster
