2018-05-31 11:47:41

by Wei Yongjun

Subject: [PATCH -next] dm writecache: fix return value check in writecache_ctr()

Function dm_io_client_create() and dm_kcopyd_client_create() return
ERR_PTR() not NULL in case of error. The NULL test in the return value
check should be replaced with IS_ERR()

Fixes: 2105231db61b ("dm: add writecache target")
Signed-off-by: Wei Yongjun <[email protected]>
---
drivers/md/dm-writecache.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/md/dm-writecache.c b/drivers/md/dm-writecache.c
index 844c4fb..0b2b436 100644
--- a/drivers/md/dm-writecache.c
+++ b/drivers/md/dm-writecache.c
@@ -1872,9 +1872,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
}

wc->dm_io = dm_io_client_create();
- if (!wc->dm_io) {
- r = -ENOMEM;
+ if (IS_ERR(wc->dm_io)) {
+ r = PTR_ERR(wc->dm_io);
ti->error = "Unable to allocate dm-io client";
+ wc->dm_io = NULL;
goto bad;
}

@@ -2096,9 +2097,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
}

wc->dm_kcopyd = dm_kcopyd_client_create(&dm_kcopyd_throttle);
- if (!wc->dm_kcopyd) {
- r = -ENOMEM;
+ if (IS_ERR(wc->dm_kcopyd)) {
+ r = PTR_ERR(wc->dm_kcopyd);
ti->error = "Unable to allocate dm-kcopyd client";
+ wc->dm_kcopyd = NULL;
goto bad;
}

2018-05-31 16:02:57

by Mike Snitzer

Subject: Re: [PATCH -next] dm writecache: fix return value check in writecache_ctr()

On Thu, May 31 2018 at 7:53am -0400,
Wei Yongjun <[email protected]> wrote:

> Function dm_io_client_create() and dm_kcopyd_client_create() return
> ERR_PTR() not NULL in case of error. The NULL test in the return value
> check should be replaced with IS_ERR()
>
> Fixes: 2105231db61b ("dm: add writecache target")
> Signed-off-by: Wei Yongjun <[email protected]>
> ---
> drivers/md/dm-writecache.c | 10 ++++++----
> 1 file changed, 6 insertions(+), 4 deletions(-)
>
> diff --git a/drivers/md/dm-writecache.c b/drivers/md/dm-writecache.c
> index 844c4fb..0b2b436 100644
> --- a/drivers/md/dm-writecache.c
> +++ b/drivers/md/dm-writecache.c
> @@ -1872,9 +1872,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
> }
>
> wc->dm_io = dm_io_client_create();
> - if (!wc->dm_io) {
> - r = -ENOMEM;
> + if (IS_ERR(wc->dm_io)) {
> + r = PTR_ERR(wc->dm_io);
> ti->error = "Unable to allocate dm-io client";
> + wc->dm_io = NULL;
> goto bad;
> }
>
> @@ -2096,9 +2097,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
> }
>
> wc->dm_kcopyd = dm_kcopyd_client_create(&dm_kcopyd_throttle);
> - if (!wc->dm_kcopyd) {
> - r = -ENOMEM;
> + if (IS_ERR(wc->dm_kcopyd)) {
> + r = PTR_ERR(wc->dm_kcopyd);
> ti->error = "Unable to allocate dm-kcopyd client";
> + wc->dm_kcopyd = NULL;
> goto bad;
> }
>


Yes, I forgot to pull this in.. it got lost in the shuffle. Mikulas had
provided this same change (folded into a larger patch) yesterday, see:
https://patchwork.kernel.org/patch/10439345/

I'll get this folded in ASAP.

Thanks,
Mike

2018-05-31 16:07:53

by Mike Snitzer

Subject: Re: [PATCH -next] dm writecache: fix return value check in writecache_ctr()

On Thu, May 31 2018 at 12:02pm -0400,
Mike Snitzer <[email protected]> wrote:

> On Thu, May 31 2018 at 7:53am -0400,
> Wei Yongjun <[email protected]> wrote:
>
> > Function dm_io_client_create() and dm_kcopyd_client_create() return
> > ERR_PTR() not NULL in case of error. The NULL test in the return value
> > check should be replaced with IS_ERR()
> >
> > Fixes: 2105231db61b ("dm: add writecache target")
> > Signed-off-by: Wei Yongjun <[email protected]>
> > ---
> > drivers/md/dm-writecache.c | 10 ++++++----
> > 1 file changed, 6 insertions(+), 4 deletions(-)
> >
> > diff --git a/drivers/md/dm-writecache.c b/drivers/md/dm-writecache.c
> > index 844c4fb..0b2b436 100644
> > --- a/drivers/md/dm-writecache.c
> > +++ b/drivers/md/dm-writecache.c
> > @@ -1872,9 +1872,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
> > }
> >
> > wc->dm_io = dm_io_client_create();
> > - if (!wc->dm_io) {
> > - r = -ENOMEM;
> > + if (IS_ERR(wc->dm_io)) {
> > + r = PTR_ERR(wc->dm_io);
> > ti->error = "Unable to allocate dm-io client";
> > + wc->dm_io = NULL;
> > goto bad;
> > }
> >
> > @@ -2096,9 +2097,10 @@ static int writecache_ctr(struct dm_target *ti, unsigned argc, char **argv)
> > }
> >
> > wc->dm_kcopyd = dm_kcopyd_client_create(&dm_kcopyd_throttle);
> > - if (!wc->dm_kcopyd) {
> > - r = -ENOMEM;
> > + if (IS_ERR(wc->dm_kcopyd)) {
> > + r = PTR_ERR(wc->dm_kcopyd);
> > ti->error = "Unable to allocate dm-kcopyd client";
> > + wc->dm_kcopyd = NULL;
> > goto bad;
> > }
> >
>
>
> Yes, I forgot to pull this in.. it got lost in the shuffle. Mikulas had
> provided this same change (folded into a larger patch) yesterday, see:
> https://patchwork.kernel.org/patch/10439345/
>
> I'll get this folded in ASAP.

Turns out I already did fold it in. My push to linux-next just missed
today's tree.

So already taken care of.

Thanks for sending the patch though.

Mike