[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20191231094706.GD24469@gofer.mess.org>
Date: Tue, 31 Dec 2019 09:47:06 +0000
From: Sean Young <sean@...s.org>
To: Mauro Carvalho Chehab <mchehab@...nel.org>
Cc: "Daniel W. S. Almeida" <dwlsalmeida@...il.com>,
gregkh@...uxfoundation.org, rfontana@...hat.com,
kstewart@...uxfoundation.org, tglx@...utronix.de,
skhan@...uxfoundation.org,
linux-kernel-mentees@...ts.linuxfoundation.org,
linux-media@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] media: dvb_dummy_fe: Add error messages in case of
attach failure
Hi Daniel, Mauro,
On Sat, Nov 30, 2019 at 07:54:15AM +0100, Mauro Carvalho Chehab wrote:
> Em Wed, 6 Nov 2019 22:37:45 -0300
> "Daniel W. S. Almeida" <dwlsalmeida@...il.com> escreveu:
>
> > From: "Daniel W. S. Almeida" <dwlsalmeida@...il.com>
> >
> > Complain if the attach functions fail, for any reason. This is helpful
> > when debugging.
> >
> > Suggested-by: Shuah Khan <skhan@...uxfoundation.org>
> > Signed-off-by: Daniel W. S. Almeida <dwlsalmeida@...il.com>
>
> Looks OK to me.
>
> > ---
> > drivers/media/dvb-frontends/dvb_dummy_fe.c | 18 +++++++++++++++---
> > 1 file changed, 15 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/media/dvb-frontends/dvb_dummy_fe.c b/drivers/media/dvb-frontends/dvb_dummy_fe.c
> > index 4db679cb70ad..ca86857c3667 100644
> > --- a/drivers/media/dvb-frontends/dvb_dummy_fe.c
> > +++ b/drivers/media/dvb-frontends/dvb_dummy_fe.c
> > @@ -114,12 +114,16 @@ struct dvb_frontend* dvb_dummy_fe_ofdm_attach(void)
> > /* allocate memory for the internal state */
> > state = kzalloc(sizeof(struct dvb_dummy_fe_state), GFP_KERNEL);
> > if (!state)
> > - return NULL;
> > + goto err;
> >
> > /* create dvb_frontend */
> > memcpy(&state->frontend.ops, &dvb_dummy_fe_ofdm_ops, sizeof(struct dvb_frontend_ops));
> > state->frontend.demodulator_priv = state;
> > return &state->frontend;
> > +
> > +err:
> > + pr_err("%s: DVB Dummy frontend driver attach failed\n", __func__);
> > + return NULL;
> > }
If kzalloc() fails, it will already complain before returning. Lately we've
been removing error logs after kmalloc() failures. I'm sure someone will send
in patches reverting this if it gets merged.
Is this a good idea?
Sean
> >
> > static const struct dvb_frontend_ops dvb_dummy_fe_qpsk_ops;
> > @@ -131,12 +135,16 @@ struct dvb_frontend *dvb_dummy_fe_qpsk_attach(void)
> > /* allocate memory for the internal state */
> > state = kzalloc(sizeof(struct dvb_dummy_fe_state), GFP_KERNEL);
> > if (!state)
> > - return NULL;
> > + goto err;
> >
> > /* create dvb_frontend */
> > memcpy(&state->frontend.ops, &dvb_dummy_fe_qpsk_ops, sizeof(struct dvb_frontend_ops));
> > state->frontend.demodulator_priv = state;
> > return &state->frontend;
> > +
> > +err:
> > + pr_err("%s: DVB Dummy frontend driver attach failed\n", __func__);
> > + return NULL;
> > }
> >
> > static const struct dvb_frontend_ops dvb_dummy_fe_qam_ops;
> > @@ -148,12 +156,16 @@ struct dvb_frontend *dvb_dummy_fe_qam_attach(void)
> > /* allocate memory for the internal state */
> > state = kzalloc(sizeof(struct dvb_dummy_fe_state), GFP_KERNEL);
> > if (!state)
> > - return NULL;
> > + goto err;
> >
> > /* create dvb_frontend */
> > memcpy(&state->frontend.ops, &dvb_dummy_fe_qam_ops, sizeof(struct dvb_frontend_ops));
> > state->frontend.demodulator_priv = state;
> > return &state->frontend;
> > +
> > +err:
> > + pr_err("%s: DVB Dummy frontend driver attach failed\n", __func__);
> > + return NULL;
> > }
> >
> > static const struct dvb_frontend_ops dvb_dummy_fe_ofdm_ops = {
>
>
>
> Cheers,
> Mauro
Powered by blists - more mailing lists