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

Fix md(4) panic which occurs when I/O request different than

Browse Source 
BIO_READ/BIO_WRITE is sent to vnode-backed provider (BIO_DELETE or
BIO_FLUSH).

Reported by:	ceri

Add support for BIO_FLUSH to vnode-backed md(4) devices based on
VOP_FSYNC().
This commit is contained in:
pjd 2006-11-01 18:56:18 +00:00
parent f66a72bc4c
commit 6058d96715
1 changed files with 53 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

63
sys/dev/md/md.c
View File

@ -406,6 +406,15 @@ mdstart_malloc(struct md_s *sc, struct bio *bp)
off_t secno, nsec, uc;
uintptr_t sp, osp;
switch (bp->bio_cmd) {
case BIO_READ:
case BIO_WRITE:
case BIO_DELETE:
break;
default:
return (EOPNOTSUPP);
}
nsec = bp->bio_length / sc->sectorsize;
secno = bp->bio_offset / sc->sectorsize;
dst = bp->bio_data;
@ -491,6 +500,20 @@ mdstart_vnode(struct md_s *sc, struct bio *bp)
struct uio auio;
struct iovec aiov;
struct mount *mp;
struct vnode *vp;
struct thread *td;
switch (bp->bio_cmd) {
case BIO_READ:
case BIO_WRITE:
case BIO_FLUSH:
break;
default:
return (EOPNOTSUPP);
}
td = curthread;
vp = sc->vnode;
/*
* VNODE I/O
@ -500,6 +523,17 @@ mdstart_vnode(struct md_s *sc, struct bio *bp)
* still valid.
*/
if (bp->bio_cmd == BIO_FLUSH) {
vfslocked = VFS_LOCK_GIANT(vp->v_mount);
(void) vn_start_write(vp, &mp, V_WAIT);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
error = VOP_FSYNC(vp, MNT_WAIT, td);
VOP_UNLOCK(vp, 0, td);
vn_finished_write(mp);
VFS_UNLOCK_GIANT(vfslocked);
return (error);
}
bzero(&auio, sizeof(auio));
aiov.iov_base = bp->bio_data;
@ -515,22 +549,22 @@ mdstart_vnode(struct md_s *sc, struct bio *bp)
else
panic("wrong BIO_OP in mdstart_vnode");
auio.uio_resid = bp->bio_length;
auio.uio_td = curthread;
auio.uio_td = td;
/*
* When reading set IO_DIRECT to try to avoid double-caching
* the data. When writing IO_DIRECT is not optimal.
*/
vfslocked = VFS_LOCK_GIANT(sc->vnode->v_mount);
vfslocked = VFS_LOCK_GIANT(vp->v_mount);
if (bp->bio_cmd == BIO_READ) {
vn_lock(sc->vnode, LK_EXCLUSIVE | LK_RETRY, curthread);
error = VOP_READ(sc->vnode, &auio, IO_DIRECT, sc->cred);
VOP_UNLOCK(sc->vnode, 0, curthread);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
error = VOP_READ(vp, &auio, IO_DIRECT, sc->cred);
VOP_UNLOCK(vp, 0, td);
} else {
(void) vn_start_write(sc->vnode, &mp, V_WAIT);
vn_lock(sc->vnode, LK_EXCLUSIVE | LK_RETRY, curthread);
error = VOP_WRITE(sc->vnode, &auio,
sc->flags & MD_ASYNC ? 0 : IO_SYNC, sc->cred);
VOP_UNLOCK(sc->vnode, 0, curthread);
(void) vn_start_write(vp, &mp, V_WAIT);
vn_lock(vp, LK_EXCLUSIVE | LK_RETRY, td);
error = VOP_WRITE(vp, &auio, sc->flags & MD_ASYNC ? 0 : IO_SYNC,
sc->cred);
VOP_UNLOCK(vp, 0, td);
vn_finished_write(mp);
}
VFS_UNLOCK_GIANT(vfslocked);
@ -547,6 +581,15 @@ mdstart_swap(struct md_s *sc, struct bio *bp)
vm_page_t m;
u_char *p;
switch (bp->bio_cmd) {
case BIO_READ:
case BIO_WRITE:
case BIO_DELETE:
break;
default:
return (EOPNOTSUPP);
}
p = bp->bio_data;
/*