2018-07-16 12:37:05

by Colin King

Subject: [PATCH] nfsd: fix memory leak of async_copy

From: Colin Ian King <[email protected]>

In the case where async_copy is successfully allocated but
the call to nfs4_init_cp_state fails, async_copy is not
currently freed and the memory is leaked. Fix this by kfree'ing
it before returning.

Detected by CoverityScan, CID#1471823 ("Resource leak")

Fixes: beb1814d5a8a ("NFSD create new stateid for async copy")
Signed-off-by: Colin Ian King <[email protected]>
---
fs/nfsd/nfs4proc.c | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
index 8f3368353aaf..3fb96a2708b9 100644
--- a/fs/nfsd/nfs4proc.c
+++ b/fs/nfsd/nfs4proc.c
@@ -1295,8 +1295,10 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
async_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
if (!async_copy)
goto out;
- if (!nfs4_init_cp_state(nn, copy))
+ if (!nfs4_init_cp_state(nn, copy)) {
+ kfree(async_copy);
goto out;
+ }
refcount_set(&async_copy->refcount, 1);
memcpy(&copy->cp_res.cb_stateid, &copy->cp_stateid,
sizeof(copy->cp_stateid));
--
2.17.1

2018-07-17 10:02:01

by Dan Carpenter

Subject: Re: [PATCH] nfsd: fix memory leak of async_copy

On Mon, Jul 16, 2018 at 01:09:54PM +0100, Colin King wrote:
> From: Colin Ian King <[email protected]>
>
> In the case where async_copy is successfully allocated but
> the call to nfs4_init_cp_state fails, async_copy is not
> currently freed and the memory is leaked. Fix this by kfree'ing
> it before returning.
>
> Detected by CoverityScan, CID#1471823 ("Resource leak")
>
> Fixes: beb1814d5a8a ("NFSD create new stateid for async copy")
> Signed-off-by: Colin Ian King <[email protected]>
> ---
> fs/nfsd/nfs4proc.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
> index 8f3368353aaf..3fb96a2708b9 100644
> --- a/fs/nfsd/nfs4proc.c
> +++ b/fs/nfsd/nfs4proc.c
> @@ -1295,8 +1295,10 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> async_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
> if (!async_copy)
> goto out;
> - if (!nfs4_init_cp_state(nn, copy))
> + if (!nfs4_init_cp_state(nn, copy)) {
> + kfree(async_copy);
> goto out;

It really feels like both this and the kzalloc() failure should be doing
an of fput() of copy->file_src and copy->file_dst. The goto out_err
does an list_del(&copy->copies); but it happens before the
"list_add(&async_copy->copies ..." so that's likely wrong as well.

regards,
dan carpenter

2018-07-17 10:05:46

by Colin King

Subject: Re: [PATCH] nfsd: fix memory leak of async_copy

On 17/07/18 10:30, Dan Carpenter wrote:
> On Mon, Jul 16, 2018 at 01:09:54PM +0100, Colin King wrote:
>> From: Colin Ian King <[email protected]>
>>
>> In the case where async_copy is successfully allocated but
>> the call to nfs4_init_cp_state fails, async_copy is not
>> currently freed and the memory is leaked. Fix this by kfree'ing
>> it before returning.
>>
>> Detected by CoverityScan, CID#1471823 ("Resource leak")
>>
>> Fixes: beb1814d5a8a ("NFSD create new stateid for async copy")
>> Signed-off-by: Colin Ian King <[email protected]>
>> ---
>> fs/nfsd/nfs4proc.c | 4 +++-
>> 1 file changed, 3 insertions(+), 1 deletion(-)
>>
>> diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
>> index 8f3368353aaf..3fb96a2708b9 100644
>> --- a/fs/nfsd/nfs4proc.c
>> +++ b/fs/nfsd/nfs4proc.c
>> @@ -1295,8 +1295,10 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
>> async_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
>> if (!async_copy)
>> goto out;
>> - if (!nfs4_init_cp_state(nn, copy))
>> + if (!nfs4_init_cp_state(nn, copy)) {
>> + kfree(async_copy);
>> goto out;
>
> It really feels like both this and the kzalloc() failure should be doing
> an of fput() of copy->file_src and copy->file_dst. The goto out_err
> does an list_del(&copy->copies); but it happens before the
> "list_add(&async_copy->copies ..." so that's likely wrong as well.

Good observation, thanks for spotting that. I suspect I'm a bit out of
my depth figuring out the exact error handling reaping steps here.
Perhaps this is one for the maintainers to figure out a safe cleanup on
these error paths.

>
> regards,
> dan carpenter
>
> --
> To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>

2018-07-17 10:45:13

by J. Bruce Fields

Subject: Re: [PATCH] nfsd: fix memory leak of async_copy

Cc'ing Olga.--b.

On Tue, Jul 17, 2018 at 10:33:59AM +0100, Colin Ian King wrote:
> On 17/07/18 10:30, Dan Carpenter wrote:
> > On Mon, Jul 16, 2018 at 01:09:54PM +0100, Colin King wrote:
> >> From: Colin Ian King <[email protected]>
> >>
> >> In the case where async_copy is successfully allocated but
> >> the call to nfs4_init_cp_state fails, async_copy is not
> >> currently freed and the memory is leaked. Fix this by kfree'ing
> >> it before returning.
> >>
> >> Detected by CoverityScan, CID#1471823 ("Resource leak")
> >>
> >> Fixes: beb1814d5a8a ("NFSD create new stateid for async copy")
> >> Signed-off-by: Colin Ian King <[email protected]>
> >> ---
> >> fs/nfsd/nfs4proc.c | 4 +++-
> >> 1 file changed, 3 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c
> >> index 8f3368353aaf..3fb96a2708b9 100644
> >> --- a/fs/nfsd/nfs4proc.c
> >> +++ b/fs/nfsd/nfs4proc.c
> >> @@ -1295,8 +1295,10 @@ nfsd4_copy(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate,
> >> async_copy = kzalloc(sizeof(struct nfsd4_copy), GFP_KERNEL);
> >> if (!async_copy)
> >> goto out;
> >> - if (!nfs4_init_cp_state(nn, copy))
> >> + if (!nfs4_init_cp_state(nn, copy)) {
> >> + kfree(async_copy);
> >> goto out;
> >
> > It really feels like both this and the kzalloc() failure should be doing
> > an of fput() of copy->file_src and copy->file_dst. The goto out_err
> > does an list_del(&copy->copies); but it happens before the
> > "list_add(&async_copy->copies ..." so that's likely wrong as well.
>
> Good observation, thanks for spotting that. I suspect I'm a bit out of
> my depth figuring out the exact error handling reaping steps here.
> Perhaps this is one for the maintainers to figure out a safe cleanup on
> these error paths.
>
> >
> > regards,
> > dan carpenter
> >
> > --
> > To unsubscribe from this list: send the line "unsubscribe kernel-janitors" in
> > the body of a message to [email protected]
> > More majordomo info at http://vger.kernel.org/majordomo-info.html
> >