[PATCH V2] lightnvm: if LUNs are already allocated fix return

[PATCH V2] lightnvm: if LUNs are already allocated fix return

[Rakesh Pandit]

While creating new device with NVM_DEV_CREATE if LUNs are already
allocated ioctl would return -ENOMEM which is wrong.  This patch
propagates -EBUSY from nvm_reserve_luns which is correct response.

Fixes: ade69e243 ("lightnvm: merge gennvm with core")
Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
---

V2: return error code directly instead of using ret variable (Frans)

 drivers/lightnvm/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
index b8f82f5..c5d71c6 100644
--- a/drivers/lightnvm/core.c
+++ b/drivers/lightnvm/core.c
@@ -253,7 +253,7 @@ static int nvm_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
 	mutex_unlock(&dev->mlock);
 
 	if (nvm_reserve_luns(dev, s->lun_begin, s->lun_end))
-		return -ENOMEM;
+		return -EBUSY;
 
 	t = kmalloc(sizeof(struct nvm_target), GFP_KERNEL);
 	if (!t) {
-- 
2.5.5

Re: [PATCH V2] lightnvm: if LUNs are already allocated fix return

[Frans Klaver]

On Tue, Jun 27, 2017 at 12:43 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
> While creating new device with NVM_DEV_CREATE if LUNs are already
> allocated ioctl would return -ENOMEM which is wrong.  This patch
> propagates -EBUSY from nvm_reserve_luns which is correct response.
>
> Fixes: ade69e243 ("lightnvm: merge gennvm with core")
> Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
> ---
>
> V2: return error code directly instead of using ret variable (Frans)
>
>  drivers/lightnvm/core.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
> index b8f82f5..c5d71c6 100644
> --- a/drivers/lightnvm/core.c
> +++ b/drivers/lightnvm/core.c
> @@ -253,7 +253,7 @@ static int nvm_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
>         mutex_unlock(&dev->mlock);
>
>         if (nvm_reserve_luns(dev, s->lun_begin, s->lun_end))
> -               return -ENOMEM;
> +               return -EBUSY;

Why aren't you propagating ret in this version?

Re: [PATCH V2] lightnvm: if LUNs are already allocated fix return

[Rakesh Pandit]

On Tue, Jun 27, 2017 at 01:01:22PM +0200, Frans Klaver wrote:
> On Tue, Jun 27, 2017 at 12:43 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
> > While creating new device with NVM_DEV_CREATE if LUNs are already
> > allocated ioctl would return -ENOMEM which is wrong.  This patch
> > propagates -EBUSY from nvm_reserve_luns which is correct response.
> >
> > Fixes: ade69e243 ("lightnvm: merge gennvm with core")
> > Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
> > ---
> >
> > V2: return error code directly instead of using ret variable (Frans)
> >
> >  drivers/lightnvm/core.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
> > index b8f82f5..c5d71c6 100644
> > --- a/drivers/lightnvm/core.c
> > +++ b/drivers/lightnvm/core.c
> > @@ -253,7 +253,7 @@ static int nvm_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
> >         mutex_unlock(&dev->mlock);
> >
> >         if (nvm_reserve_luns(dev, s->lun_begin, s->lun_end))
> > -               return -ENOMEM;
> > +               return -EBUSY;
> 
> Why aren't you propagating ret in this version?

Well nvm_reserve_luns either returns 0 or -EBUSY and it is unlikely
that return value would change and even if it does this can be
updated.

Re: [PATCH V2] lightnvm: if LUNs are already allocated fix return

[Frans Klaver]

On Tue, Jun 27, 2017 at 1:23 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
> On Tue, Jun 27, 2017 at 01:01:22PM +0200, Frans Klaver wrote:
>> On Tue, Jun 27, 2017 at 12:43 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
>> > While creating new device with NVM_DEV_CREATE if LUNs are already
>> > allocated ioctl would return -ENOMEM which is wrong.  This patch
>> > propagates -EBUSY from nvm_reserve_luns which is correct response.
>> >
>> > Fixes: ade69e243 ("lightnvm: merge gennvm with core")
>> > Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
>> > ---
>> >
>> > V2: return error code directly instead of using ret variable (Frans)
>> >
>> >  drivers/lightnvm/core.c | 2 +-
>> >  1 file changed, 1 insertion(+), 1 deletion(-)
>> >
>> > diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
>> > index b8f82f5..c5d71c6 100644
>> > --- a/drivers/lightnvm/core.c
>> > +++ b/drivers/lightnvm/core.c
>> > @@ -253,7 +253,7 @@ static int nvm_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
>> >         mutex_unlock(&dev->mlock);
>> >
>> >         if (nvm_reserve_luns(dev, s->lun_begin, s->lun_end))
>> > -               return -ENOMEM;
>> > +               return -EBUSY;
>>
>> Why aren't you propagating ret in this version?
>
> Well nvm_reserve_luns either returns 0 or -EBUSY and it is unlikely
> that return value would change and even if it does this can be
> updated.

If you propagate the result of nvm_reserve_luns(), the casual reader
will immediately understand that any possible faulty result is
returned. returning -EBUSY here might suggest you're overriding
whatever this function returns.

Re: [PATCH V2] lightnvm: if LUNs are already allocated fix return

[Rakesh Pandit]

On Tue, Jun 27, 2017 at 01:27:40PM +0200, Frans Klaver wrote:
> On Tue, Jun 27, 2017 at 1:23 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
> > On Tue, Jun 27, 2017 at 01:01:22PM +0200, Frans Klaver wrote:
> >> On Tue, Jun 27, 2017 at 12:43 PM, Rakesh Pandit <rakesh@tuxera.com> wrote:
> >> > While creating new device with NVM_DEV_CREATE if LUNs are already
> >> > allocated ioctl would return -ENOMEM which is wrong.  This patch
> >> > propagates -EBUSY from nvm_reserve_luns which is correct response.
> >> >
> >> > Fixes: ade69e243 ("lightnvm: merge gennvm with core")
> >> > Signed-off-by: Rakesh Pandit <rakesh@tuxera.com>
> >> > ---
> >> >
> >> > V2: return error code directly instead of using ret variable (Frans)
> >> >
> >> >  drivers/lightnvm/core.c | 2 +-
> >> >  1 file changed, 1 insertion(+), 1 deletion(-)
> >> >
> >> > diff --git a/drivers/lightnvm/core.c b/drivers/lightnvm/core.c
> >> > index b8f82f5..c5d71c6 100644
> >> > --- a/drivers/lightnvm/core.c
> >> > +++ b/drivers/lightnvm/core.c
> >> > @@ -253,7 +253,7 @@ static int nvm_create_tgt(struct nvm_dev *dev, struct nvm_ioctl_create *create)
> >> >         mutex_unlock(&dev->mlock);
> >> >
> >> >         if (nvm_reserve_luns(dev, s->lun_begin, s->lun_end))
> >> > -               return -ENOMEM;
> >> > +               return -EBUSY;
> >>
> >> Why aren't you propagating ret in this version?
> >
> > Well nvm_reserve_luns either returns 0 or -EBUSY and it is unlikely
> > that return value would change and even if it does this can be
> > updated.
> 
> If you propagate the result of nvm_reserve_luns(), the casual reader
> will immediately understand that any possible faulty result is
> returned. returning -EBUSY here might suggest you're overriding
> whatever this function returns.

I don't have a strong opinion for or against. That was being done
irrespective of this change already.  But you have a valid argument.
I would post V3.

Thanks,