On Jul 19, 2022, at 11:49 AM, Jeff Layton [off-list ref] wrote:

On Tue, 2022-07-19 at 09:18 -0400, Chuck Lever wrote:

quoted

I discovered that xdr_encode_bool() was returning the same address
that was passed in the @p parameter. The documenting comment states
that the intent is to return the address of the next buffer
location, just like the other "xdr_encode_*" helpers.

The result was the encoded results of NFSv3 PATHCONF operations were
not formed correctly.

Fixes: ded04a587f6c ("NFSD: Update the NFSv3 PATHCONF3res encoder to use struct xdr_stream")
Signed-off-by: Chuck Lever <redacted>
---
include/linux/sunrpc/xdr.h |    4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/include/linux/sunrpc/xdr.h b/include/linux/sunrpc/xdr.h
index 5860f32e3958..986c8a17ca5e 100644
--- a/include/linux/sunrpc/xdr.h
+++ b/include/linux/sunrpc/xdr.h

@@ -419,8 +419,8 @@ static inline int xdr_stream_encode_item_absent(struct xdr_stream *xdr)
 */

static inline __be32 *xdr_encode_bool(__be32 *p, u32 n)
{
-	*p = n ? xdr_one : xdr_zero;
-	return p++;
+	*p++ = n ? xdr_one : xdr_zero;
+	return p;
}

/**

Nice catch. Postincrement operators strike again!

The original patch description read "D'oh!".

Reviewed-by: Jeff Layton <jlayton@kernel.org>

--
Chuck Lever