9p: add refcount to p9_fid struct

Fix race issue in fid contention.

Eric's and Greg's patch offer a mechanism to fix open-unlink-f*syscall
bug in 9p. But there is race issue in fid parallel accesses.
As Greg's patch stores all of fids from opened files into according inode,
so all the lookup fid ops can retrieve fid from inode preferentially. But
there is no mechanism to handle the fid contention issue. For example,
there are two threads get the same fid in the same time and one of them
clunk the fid before the other thread ready to discard the fid. In this
scenario, it will lead to some fatal problems, even kernel core dump.

I introduce a mechanism to fix this race issue. A counter field introduced
into p9_fid struct to store the reference counter to the fid. When a fid
is allocated from the inode or dentry, the counter will increase, and
will decrease at the end of its occupation. It is guaranteed that the
fid won't be clunked before the reference counter go down to 0, then
we can avoid the clunked fid to be used.

tests:
race issue test from the old test case:
for file in {01..50}; do touch f.${file}; done
seq 1 1000 | xargs -n 1 -P 50 -I{} cat f.* > /dev/null

open-unlink-f*syscall test:
I have tested for f*syscall include: ftruncate fstat fchown fchmod faccessat.

Link: http://lkml.kernel.org/r/20200923141146.90046-5-jianyong.wu@arm.com
Fixes: 478ba09edc1f ("fs/9p: search open fids first")
Signed-off-by: Jianyong Wu <jianyong.wu@arm.com>
Signed-off-by: Dominique Martinet <asmadeus@codewreck.org>

1 files changed, 5 insertions, 4 deletions

diff --git a/net/9p/client.c b/net/9p/client.c
index 1a3f72bf45fc..a6c8a915e0d8 100644
--- a/net/9p/client.c
+++ b/net/9p/client.c
@@ -901,6 +901,7 @@ static struct p9_fid *p9_fid_create(struct p9_client *clnt)
 	fid->clnt = clnt;
 	fid->rdir = NULL;
 	fid->fid = 0;
+	atomic_set(&fid->count, 1);
 
 	idr_preload(GFP_KERNEL);
 	spin_lock_irq(&clnt->lock);
@@ -908,7 +909,6 @@ static struct p9_fid *p9_fid_create(struct p9_client *clnt)
 			    GFP_NOWAIT);
 	spin_unlock_irq(&clnt->lock);
 	idr_preload_end();
-
 	if (!ret)
 		return fid;
 
@@ -1187,7 +1187,6 @@ struct p9_fid *p9_client_walk(struct p9_fid *oldfid, uint16_t nwname,
 
 	p9_debug(P9_DEBUG_9P, ">>> TWALK fids %d,%d nwname %ud wname[0] %s\n",
 		 oldfid->fid, fid->fid, nwname, wnames ? wnames[0] : NULL);
-
 	req = p9_client_rpc(clnt, P9_TWALK, "ddT", oldfid->fid, fid->fid,
 								nwname, wnames);
 	if (IS_ERR(req)) {
@@ -1461,12 +1460,14 @@ int p9_client_clunk(struct p9_fid *fid)
 	struct p9_req_t *req;
 	int retries = 0;
 
-	if (!fid) {
-		pr_warn("%s (%d): Trying to clunk with NULL fid\n",
+	if (!fid || IS_ERR(fid)) {
+		pr_warn("%s (%d): Trying to clunk with invalid fid\n",
 			__func__, task_pid_nr(current));
 		dump_stack();
 		return 0;
 	}
+	if (!atomic_dec_and_test(&fid->count))
+		return 0;
 
 again:
 	p9_debug(P9_DEBUG_9P, ">>> TCLUNK fid %d (try %d)\n", fid->fid,