2014-07-30 13:26:30

by Kinglong Mee

Subject: [PATCH] NFSD: Decrease nfsd_users in nfsd_startup_generic fail

If nfsd_startup_generic fail, nfsd_users isn't decreased right now.

NFSD restarts will return 0 at the first nfsd_users checking,
after that, nfs4_state_start_net() will meet a bad laundry_wq
(I meet NULL), after nfsd4_grace timeout, the kernel will crash.

Signed-off-by: Kinglong Mee <[email protected]>
---
fs/nfsd/nfssvc.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 5d026dc..752d56b 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -221,7 +221,8 @@ static int nfsd_startup_generic(int nrservs)
*/
ret = nfsd_racache_init(2*nrservs);
if (ret)
- return ret;
+ goto dec_users;
+
ret = nfs4_state_start();
if (ret)
goto out_racache;
@@ -229,6 +230,8 @@ static int nfsd_startup_generic(int nrservs)

out_racache:
nfsd_racache_shutdown();
+dec_users:
+ nfsd_users--;
return ret;
}

--
1.9.3

2014-07-30 14:16:50

by J. Bruce Fields

Subject: Re: [PATCH] NFSD: Decrease nfsd_users in nfsd_startup_generic fail

Thanks!

On Wed, Jul 30, 2014 at 09:26:05PM +0800, Kinglong Mee wrote:
> If nfsd_startup_generic fail, nfsd_users isn't decreased right now.

Looks like that should only fail in out of memory cases?

> NFSD restarts will return 0 at the first nfsd_users checking,
> after that, nfs4_state_start_net() will meet a bad laundry_wq
> (I meet NULL), after nfsd4_grace timeout, the kernel will crash.

So this was introduced with 4539f14981ce02d48b212786a41c8bcfb62851b4
"nfsd: replace boolean nfsd_up flag by users counter".

--b.

>
> Signed-off-by: Kinglong Mee <[email protected]>
> ---
> fs/nfsd/nfssvc.c | 5 ++++-
> 1 file changed, 4 insertions(+), 1 deletion(-)
>
> diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
> index 5d026dc..752d56b 100644
> --- a/fs/nfsd/nfssvc.c
> +++ b/fs/nfsd/nfssvc.c
> @@ -221,7 +221,8 @@ static int nfsd_startup_generic(int nrservs)
> */
> ret = nfsd_racache_init(2*nrservs);
> if (ret)
> - return ret;
> + goto dec_users;
> +
> ret = nfs4_state_start();
> if (ret)
> goto out_racache;
> @@ -229,6 +230,8 @@ static int nfsd_startup_generic(int nrservs)
>
> out_racache:
> nfsd_racache_shutdown();
> +dec_users:
> + nfsd_users--;
> return ret;
> }
>
> --
> 1.9.3
>

2014-07-31 12:56:34

by Kinglong Mee

Subject: Re: [PATCH] NFSD: Decrease nfsd_users in nfsd_startup_generic fail

On 7/30/2014 22:16, J. Bruce Fields wrote:
> Thanks!
>
> On Wed, Jul 30, 2014 at 09:26:05PM +0800, Kinglong Mee wrote:
>> If nfsd_startup_generic fail, nfsd_users isn't decreased right now.
>
> Looks like that should only fail in out of memory cases?

Yes, it is.
I found this problem through reviewing code.

>
>> NFSD restarts will return 0 at the first nfsd_users checking,
>> after that, nfs4_state_start_net() will meet a bad laundry_wq
>> (I meet NULL), after nfsd4_grace timeout, the kernel will crash.
>
> So this was introduced with 4539f14981ce02d48b212786a41c8bcfb62851b4
> "nfsd: replace boolean nfsd_up flag by users counter".

Yes, I forget adding it.

thanks,
Kinglong Mee

>> Signed-off-by: Kinglong Mee <[email protected]>
>> ---
>> fs/nfsd/nfssvc.c | 5 ++++-
>> 1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
>> index 5d026dc..752d56b 100644
>> --- a/fs/nfsd/nfssvc.c
>> +++ b/fs/nfsd/nfssvc.c
>> @@ -221,7 +221,8 @@ static int nfsd_startup_generic(int nrservs)
>> */
>> ret = nfsd_racache_init(2*nrservs);
>> if (ret)
>> - return ret;
>> + goto dec_users;
>> +
>> ret = nfs4_state_start();
>> if (ret)
>> goto out_racache;
>> @@ -229,6 +230,8 @@ static int nfsd_startup_generic(int nrservs)
>>
>> out_racache:
>> nfsd_racache_shutdown();
>> +dec_users:
>> + nfsd_users--;
>> return ret;
>> }
>>
>> --
>> 1.9.3
>>
>

2014-08-01 20:29:32

by J. Bruce Fields

Subject: Re: [PATCH] NFSD: Decrease nfsd_users in nfsd_startup_generic fail

On Thu, Jul 31, 2014 at 08:56:13PM +0800, Kinglong Mee wrote:
> On 7/30/2014 22:16, J. Bruce Fields wrote:
> > Thanks!
> >
> > On Wed, Jul 30, 2014 at 09:26:05PM +0800, Kinglong Mee wrote:
> >> If nfsd_startup_generic fail, nfsd_users isn't decreased right now.
> >
> > Looks like that should only fail in out of memory cases?
>
> Yes, it is.
> I found this problem through reviewing code.

Got it, thanks! Applying for 3.17 and stable.

--b.

>
> >
> >> NFSD restarts will return 0 at the first nfsd_users checking,
> >> after that, nfs4_state_start_net() will meet a bad laundry_wq
> >> (I meet NULL), after nfsd4_grace timeout, the kernel will crash.
> >
> > So this was introduced with 4539f14981ce02d48b212786a41c8bcfb62851b4
> > "nfsd: replace boolean nfsd_up flag by users counter".
>
> Yes, I forget adding it.
>
> thanks,
> Kinglong Mee
>
> >> Signed-off-by: Kinglong Mee <[email protected]>
> >> ---
> >> fs/nfsd/nfssvc.c | 5 ++++-
> >> 1 file changed, 4 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
> >> index 5d026dc..752d56b 100644
> >> --- a/fs/nfsd/nfssvc.c
> >> +++ b/fs/nfsd/nfssvc.c
> >> @@ -221,7 +221,8 @@ static int nfsd_startup_generic(int nrservs)
> >> */
> >> ret = nfsd_racache_init(2*nrservs);
> >> if (ret)
> >> - return ret;
> >> + goto dec_users;
> >> +
> >> ret = nfs4_state_start();
> >> if (ret)
> >> goto out_racache;
> >> @@ -229,6 +230,8 @@ static int nfsd_startup_generic(int nrservs)
> >>
> >> out_racache:
> >> nfsd_racache_shutdown();
> >> +dec_users:
> >> + nfsd_users--;
> >> return ret;
> >> }
> >>
> >> --
> >> 1.9.3
> >>
> >