> On Apr 15, 2024, at 1:21 PM, Petr Vorel <pvorel@xxxxxxx> wrote:
>
> /proc/fs/nfsd/nfsv4recoverydir started from kernel 6.8 report EINVAL.
>
> Signed-off-by: Petr Vorel <pvorel@xxxxxxx>
> ---
> Hi,
>
> @ Jeff, Chuck, Neil, NFS devs: The patch itself whitelist reading
> /proc/fs/nfsd/nfsv4recoverydir in LTP test. I suspect reading failed
> with EINVAL in 6.8 was a deliberate change and expected behavior when
> CONFIG_NFSD_LEGACY_CLIENT_TRACKING is not set:
I'm not sure it was deliberate. This seems like a behavior
regression. Jeff?
> $ sudo cat /proc/fs/nfsd/nfsv4recoverydir
> cat: /proc/fs/nfsd/nfsv4recoverydir: Invalid argument
>
> I'm asking because It worked fine in kernel 6.7:
>
> $ sudo cat /proc/fs/nfsd/nfsv4recoverydir
> /var/lib/nfs/v4recovery
>
> I did not bisect but I suspect suspect 74fd48739d04 ("nfsd: new Kconfig
> option for legacy client tracking") from v6.8-rc1. The system I test
> (openSUSE Tumbleweed) has not CONFIG_NFSD_LEGACY_CLIENT_TRACKING set and
> 74fd48739d04 wraps write_recoverydir setup, thus it's not set.
>
> +#ifdef CONFIG_NFSD_LEGACY_CLIENT_TRACKING
> [NFSD_RecoveryDir] = write_recoverydir,
> +#endif
>
> Kind regards,
> Petr
>
> testcases/kernel/fs/proc/proc01.c | 1 +
> 1 file changed, 1 insertion(+)
>
> diff --git a/testcases/kernel/fs/proc/proc01.c b/testcases/kernel/fs/proc/proc01.c
> index c90e509a3..08b9bbc75 100644
> --- a/testcases/kernel/fs/proc/proc01.c
> +++ b/testcases/kernel/fs/proc/proc01.c
> @@ -113,6 +113,7 @@ static const struct mapping known_issues[] = {
> {"read", "/proc/fs/nfsd/filehandle", EINVAL},
> {"read", "/proc/fs/nfsd/.getfs", EINVAL},
> {"read", "/proc/fs/nfsd/.getfd", EINVAL},
> + {"read", "/proc/fs/nfsd/nfsv4recoverydir", EINVAL},
> {"read", "/proc/self/net/rpc/use-gss-proxy", EAGAIN},
> {"read", "/proc/sys/net/ipv6/conf/*/stable_secret", EIO},
> {"read", "/proc/sys/vm/nr_hugepages", EOPNOTSUPP},
> --
> 2.43.0
>
>
--
Chuck Lever
