d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix two deadlocks in snapshots:

Browse Source 
1) Release the snapshot file lock while suspending the system. Otherwise
   a process trying to read the lock may block on its containing directory
   preventing the suspension from completing. Thanks to Sean Kelly
   <smkelly@zombie.org> for finding this deadlock.

2) Replace some bdwrite's with bawrite's so as not to fill all the
   buffers with dirty data. The buffers could not be cleaned as the
   snapshot vnode was locked hence the system could deadlock when
   making snapshots of really massive filesystems. Thanks to
   Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp> for figuring
   this out.

Sponsored by:   DARPA & NAI Labs.
This commit is contained in:
Kirk McKusick 2002-11-30 07:27:12 +00:00
parent a963b34a5c
commit 63cf5b0ee2
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
sys/ufs/ffs/ffs_snapshot.c
View File

@ -230,7 +230,7 @@ ffs_snapshot(mp, snapfile)
fs->fs_bsize, td->td_ucred, BA_METAONLY, &ibp);
if (error)
goto out;
bdwrite(ibp);
bawrite(ibp);
}
/*
* Allocate copies for the superblock and its summary information.
@ -257,7 +257,7 @@ ffs_snapshot(mp, snapfile)
fs->fs_bsize, KERNCRED, 0, &nbp);
if (error)
goto out;
bdwrite(nbp);
bawrite(nbp);
}
/*
* Copy all the cylinder group maps. Although the
@ -290,9 +290,12 @@ ffs_snapshot(mp, snapfile)
ip->i_flag |= IN_CHANGE | IN_UPDATE;
/*
* Ensure that the snapshot is completely on disk.
* Since we have marked it as a snapshot it is safe to
* unlock it as no process will be allowed to write to it.
*/
if ((error = VOP_FSYNC(vp, KERNCRED, MNT_WAIT, td)) != 0)
goto out;
VOP_UNLOCK(vp, 0, td);
/*
* All allocations are done, so we can now snapshot the system.
*
@ -309,12 +312,14 @@ ffs_snapshot(mp, snapfile)
vn_finished_write(wrtmp);
if ((error = vfs_write_suspend(vp->v_mount)) != 0) {
vn_start_write(NULL, &wrtmp, V_WAIT);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
goto out;
}
if (mp->mnt_kern_flag & MNTK_SUSPENDED)
break;
vn_start_write(NULL, &wrtmp, V_WAIT);
}
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
if (collectsnapstats)
nanotime(&starttime);
/*