Aren
/
NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

ftruncate(2) cause a SETATTR with only va_size set, and some filesystems 

(e.g.: glusterfs) will do a custom handling in such a situation. This
breaks because libpuffs folds a metadata (va_atime and va_mtime) update
in each SETATTR. We try to identify SETATTR caused by ftruncate(2) and
remove va_atime and va_mtime in such situation.

This fixes a bug with glusterfs, where parts of a file downloaded by
FTP was filled with zeros because of a ftruncate(2) sent out of order
with write(2) requests. glusterfs behavior depends on the undocumented
FUSE rule that ftruncate(2) will only set va_size in SETATTR.
This commit is contained in:
manu 2011-07-18 02:14:01 +00:00
parent c751831386
commit 7541af315c
1 changed files with 22 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
lib/libperfuse/ops.c
View File

@ -1,4 +1,4 @@
/* $NetBSD: ops.c,v 1.33 2011/07/14 15:37:32 manu Exp $ */
/* $NetBSD: ops.c,v 1.34 2011/07/18 02:14:01 manu Exp $ */
/*-
* Copyright (c) 2010-2011 Emmanuel Dreyfus. All rights reserved.
@ -1651,6 +1651,27 @@ perfuse_node_setattr(pu, opc, vap, pcr)
fsi->valid |= FUSE_FATTR_LOCKOWNER;
}
/*
* ftruncate() sends only va_size, and metadata cache
* flush adds va_atime and va_mtime. Some FUSE
* filesystems will attempt to detect ftruncate by
* checking for FATTR_SIZE being set without
* FATTR_UID|FATTR_GID|FATTR_ATIME|FATTR_MTIME|FATTR_MODE
*
* Try to adapt and remove FATTR_ATIME|FATTR_MTIME
* if we suspect a ftruncate().
*/
if ((va_size != (u_quad_t)PUFFS_VNOVAL) &&
((vap->va_mode == (mode_t)PUFFS_VNOVAL) &&
(vap->va_uid == (uid_t)PUFFS_VNOVAL) &&
(vap->va_gid == (gid_t)PUFFS_VNOVAL))) {
fsi->atime = 0;
fsi->atimensec = 0;
fsi->mtime = 0;
fsi->mtimensec = 0;
fsi->valid &= ~(FUSE_FATTR_ATIME|FUSE_FATTR_MTIME);
}
/*
* If nothing remain, discard the operation.
*/