| Message ID | 20221209021823.1232874-1-konstantin.meskhidze@huawei.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | fs: NFSv4.1: Fix memory leakage | expand |
On 8 Dec 2022, at 21:18, Konstantin Meskhidze wrote: > This commit fixes potential memory leakage of 'calldata' memory chunk > in _nfs41_proc_sequence() function. > > Signed-off-by: Konstantin Meskhidze <konstantin.meskhidze@huawei.com> > --- > fs/nfs/nfs4proc.c | 4 +++- > 1 file changed, 3 insertions(+), 1 deletion(-) > > diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c > index 86ed5c0142c3..b7aa66167341 100644 > --- a/fs/nfs/nfs4proc.c > +++ b/fs/nfs/nfs4proc.c > @@ -9416,8 +9416,10 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, > task_setup_data.callback_data = calldata; > > ret = rpc_run_task(&task_setup_data); > - if (IS_ERR(ret)) > + if (IS_ERR(ret)) { > + kfree(calldata); > goto out_err; > + } > return ret; > out_put_clp: > nfs_put_client(clp); > -- > 2.25.1 Did you observe this leak, or find it by code inspecton? I don't think there's a leak here because there's no way rpc_run_task() can return an error withouth also doing rpc_call_ops->rpc_release, which should free the calldata. Ben
> On Dec 9, 2022, at 07:29, Benjamin Coddington <bcodding@redhat.com> wrote: > > On 8 Dec 2022, at 21:18, Konstantin Meskhidze wrote: > >> This commit fixes potential memory leakage of 'calldata' memory chunk >> in _nfs41_proc_sequence() function. >> >> Signed-off-by: Konstantin Meskhidze <konstantin.meskhidze@huawei.com> >> --- >> fs/nfs/nfs4proc.c | 4 +++- >> 1 file changed, 3 insertions(+), 1 deletion(-) >> >> diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c >> index 86ed5c0142c3..b7aa66167341 100644 >> --- a/fs/nfs/nfs4proc.c >> +++ b/fs/nfs/nfs4proc.c >> @@ -9416,8 +9416,10 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, >> task_setup_data.callback_data = calldata; >> >> ret = rpc_run_task(&task_setup_data); >> - if (IS_ERR(ret)) >> + if (IS_ERR(ret)) { >> + kfree(calldata); >> goto out_err; >> + } >> return ret; >> out_put_clp: >> nfs_put_client(clp); >> -- >> 2.25.1 > > Did you observe this leak, or find it by code inspecton? > > I don't think there's a leak here because there's no way rpc_run_task() can > return an error withouth also doing rpc_call_ops->rpc_release, which should > free the calldata. Correct. We will not be applying this patch.
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 86ed5c0142c3..b7aa66167341 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -9416,8 +9416,10 @@ static struct rpc_task *_nfs41_proc_sequence(struct nfs_client *clp, task_setup_data.callback_data = calldata; ret = rpc_run_task(&task_setup_data); - if (IS_ERR(ret)) + if (IS_ERR(ret)) { + kfree(calldata); goto out_err; + } return ret; out_put_clp: nfs_put_client(clp);
This commit fixes potential memory leakage of 'calldata' memory chunk in _nfs41_proc_sequence() function. Signed-off-by: Konstantin Meskhidze <konstantin.meskhidze@huawei.com> --- fs/nfs/nfs4proc.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)