Coda: push BKL regions into coda_upcall()
Now that shared inode state is locked using the cii->c_lock, the BKL is
only used to protect the upcall queues used to communicate with the
userspace cache manager. The remaining state is all local and we can
push the lock further down into coda_upcall().
Signed-off-by: Yoshihisa Abe <yoshiabe@cs.cmu.edu>
Signed-off-by: Jan Harkes <jaharkes@cs.cmu.edu>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
diff --git a/fs/coda/psdev.c b/fs/coda/psdev.c
index fdc2f3e..9a9248e 100644
--- a/fs/coda/psdev.c
+++ b/fs/coda/psdev.c
@@ -108,16 +108,9 @@
return -EFAULT;
if (DOWNCALL(hdr.opcode)) {
- struct super_block *sb = NULL;
- union outputArgs *dcbuf;
+ union outputArgs *dcbuf;
int size = sizeof(*dcbuf);
- sb = vcp->vc_sb;
- if ( !sb ) {
- count = nbytes;
- goto out;
- }
-
if ( nbytes < sizeof(struct coda_out_hdr) ) {
printk("coda_downcall opc %d uniq %d, not enough!\n",
hdr.opcode, hdr.unique);
@@ -137,9 +130,7 @@
}
/* what downcall errors does Venus handle ? */
- lock_kernel();
- error = coda_downcall(hdr.opcode, dcbuf, sb);
- unlock_kernel();
+ error = coda_downcall(vcp, hdr.opcode, dcbuf);
CODA_FREE(dcbuf, nbytes);
if (error) {