[PATCH] nfsd: check get_user() return when reading princhashlen

Dominik Woźniak posted 1 patch 3 days, 3 hours ago
fs/nfsd/nfs4recover.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
[PATCH] nfsd: check get_user() return when reading princhashlen
Posted by Dominik Woźniak 3 days, 3 hours ago
In __cld_pipe_inprogress_downcall(), the get_user() that reads
princhashlen from the userspace cld_msg_v2 buffer does not check its
return value. A failing copy leaves princhashlen with uninitialised
stack contents, which are then used to drive memdup_user() and stored
as princhash.len on the resulting reclaim record. The other get_user()
calls in this function all check the return; only this one is missed,
which is most likely a copy-paste oversight from when v2 upcalls were
introduced.

Mirror the existing pattern used a few lines above for namelen.
namecopy is declared with __free(kfree) so the early return cleans up
the already-allocated buffer automatically.

Fixes: 6ee95d1c8991 ("nfsd: add support for upcall version 2")
Signed-off-by: Dominik Woźniak <stalion@gmail.com>
---
 fs/nfsd/nfs4recover.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
index b338473d6e52..6ea25a52d2f4 100644
--- a/fs/nfsd/nfs4recover.c
+++ b/fs/nfsd/nfs4recover.c
@@ -718,7 +718,8 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
 				return PTR_ERR(namecopy);
 			name.data = namecopy;
 			name.len = namelen;
-			get_user(princhashlen, &ci->cc_princhash.cp_len);
+			if (get_user(princhashlen, &ci->cc_princhash.cp_len))
+				return -EFAULT;
 			if (princhashlen > 0) {
 				princhashcopy = memdup_user(
 					&ci->cc_princhash.cp_data,
-- 
2.50.1 (Apple Git-155)

Re: [PATCH] nfsd: check get_user() return when reading princhashlen
Posted by Chuck Lever 2 days, 23 hours ago
From: Chuck Lever <chuck.lever@oracle.com>

On Thu, 21 May 2026 17:46:56 +0200, Dominik Woźniak wrote:
> In __cld_pipe_inprogress_downcall(), the get_user() that reads
> princhashlen from the userspace cld_msg_v2 buffer does not check its
> return value. A failing copy leaves princhashlen with uninitialised
> stack contents, which are then used to drive memdup_user() and stored
> as princhash.len on the resulting reclaim record. The other get_user()
> calls in this function all check the return; only this one is missed,
> which is most likely a copy-paste oversight from when v2 upcalls were
> introduced.
> 
> [...]

Applied to nfsd-testing, thanks!

[1/1] nfsd: check get_user() return when reading princhashlen
      commit: aff199f9ef2d1870edbc27253798bd97ee9c65be

--
Chuck Lever <chuck.lever@oracle.com>
Re: [PATCH] nfsd: check get_user() return when reading princhashlen
Posted by Jeff Layton 3 days, 3 hours ago
On Thu, 2026-05-21 at 17:46 +0200, Dominik Woźniak wrote:
> In __cld_pipe_inprogress_downcall(), the get_user() that reads
> princhashlen from the userspace cld_msg_v2 buffer does not check its
> return value. A failing copy leaves princhashlen with uninitialised
> stack contents, which are then used to drive memdup_user() and stored
> as princhash.len on the resulting reclaim record. The other get_user()
> calls in this function all check the return; only this one is missed,
> which is most likely a copy-paste oversight from when v2 upcalls were
> introduced.
> 
> Mirror the existing pattern used a few lines above for namelen.
> namecopy is declared with __free(kfree) so the early return cleans up
> the already-allocated buffer automatically.
> 
> Fixes: 6ee95d1c8991 ("nfsd: add support for upcall version 2")
> Signed-off-by: Dominik Woźniak <stalion@gmail.com>
> ---
>  fs/nfsd/nfs4recover.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c
> index b338473d6e52..6ea25a52d2f4 100644
> --- a/fs/nfsd/nfs4recover.c
> +++ b/fs/nfsd/nfs4recover.c
> @@ -718,7 +718,8 @@ __cld_pipe_inprogress_downcall(const struct cld_msg_v2 __user *cmsg,
>  				return PTR_ERR(namecopy);
>  			name.data = namecopy;
>  			name.len = namelen;
> -			get_user(princhashlen, &ci->cc_princhash.cp_len);
> +			if (get_user(princhashlen, &ci->cc_princhash.cp_len))
> +				return -EFAULT;
>  			if (princhashlen > 0) {
>  				princhashcopy = memdup_user(
>  					&ci->cc_princhash.cp_data,

Nice catch.

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