| Message ID | 20240308121021.1732-1-yongsuyoo0215@gmail.com |
|---|---|
| State | New |
| Headers | show |
| Series | media: dvb_ca_en50221: Fix a bug for detecting CI MODULE | expand |
Dear All Can you review this patch ? Can you share how this modification is going ? 2024년 3월 8일 (금) 오후 9:10, <yongsuyoo0215@gmail.com>님이 작성: > > From: Yongsu yoo <yongsuyoo0215@gmail.com> > > Signed-off-by:Yongsu Yoo <yongsuyoo0215@gmail.com> > > In source/drivers/media/dvb-core/dvb_ca_en50221.c, if the CA_RESET > ioctl is called, the dvb_ca_en50221_slot_shutdown will also be called. > Inside of the dvb_ca_en50221_slot_shutdown, > the ca->slot_info[slot].slot_state will become DVB_CA_SLOTSTATE_NONE. > In the most of cases, the ca->slot_info[slot].slot_state will quickly > becomes restored to other states by the subsequent operations of the > thread dvb_ca_en50221_thread_state_machine. > But in some rare cases, when the CA_GET_SLOT_INFO ioctl is immediately > called after the CA_RESET ioctl is called, the > the ca->slot_info[slot].slot_state can still remains at > DVB_CA_SLOTSTATE_NONE, and this causes CA_GET_SLOT_INFO ioctl not to > return CA_CI_MODULE_PRESENT as info->flags even if CA_CI_MODULE is > really connected on TV. This means that the CA_GET_SLOT_INFO ioctl > does not return right informtion. This is a Bug. We fix this bug. > --- > drivers/media/dvb-core/dvb_ca_en50221.c | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/drivers/media/dvb-core/dvb_ca_en50221.c b/drivers/media/dvb-core/dvb_ca_en50221.c > index baf64540dc00..8d37c3c13227 100644 > --- a/drivers/media/dvb-core/dvb_ca_en50221.c > +++ b/drivers/media/dvb-core/dvb_ca_en50221.c > @@ -1403,6 +1403,10 @@ static int dvb_ca_en50221_io_do_ioctl(struct file *file, > (sl->slot_state != DVB_CA_SLOTSTATE_INVALID)) { > info->flags = CA_CI_MODULE_PRESENT; > } > + if ((sl->slot_state == DVB_CA_SLOTSTATE_NONE) && > + (sl->camchange_type == DVB_CA_EN50221_CAMCHANGE_INSERTED)) { > + info->flags = CA_CI_MODULE_PRESENT; > + } > if (sl->slot_state == DVB_CA_SLOTSTATE_RUNNING) > info->flags |= CA_CI_MODULE_READY; > break; > -- > 2.17.1 >
Dear All Can you review this patch ? Can you share how this is going ? 2024년 4월 11일 (목) 오후 9:03, YongSu Yoo <yongsuyoo0215@gmail.com>님이 작성: > > Dear All > > Can you review this patch ? > Can you share how this modification is going ? > > 2024년 3월 8일 (금) 오후 9:10, <yongsuyoo0215@gmail.com>님이 작성: > > > > From: Yongsu yoo <yongsuyoo0215@gmail.com> > > > > Signed-off-by:Yongsu Yoo <yongsuyoo0215@gmail.com> > > > > In source/drivers/media/dvb-core/dvb_ca_en50221.c, if the CA_RESET > > ioctl is called, the dvb_ca_en50221_slot_shutdown will also be called. > > Inside of the dvb_ca_en50221_slot_shutdown, > > the ca->slot_info[slot].slot_state will become DVB_CA_SLOTSTATE_NONE. > > In the most of cases, the ca->slot_info[slot].slot_state will quickly > > becomes restored to other states by the subsequent operations of the > > thread dvb_ca_en50221_thread_state_machine. > > But in some rare cases, when the CA_GET_SLOT_INFO ioctl is immediately > > called after the CA_RESET ioctl is called, the > > the ca->slot_info[slot].slot_state can still remains at > > DVB_CA_SLOTSTATE_NONE, and this causes CA_GET_SLOT_INFO ioctl not to > > return CA_CI_MODULE_PRESENT as info->flags even if CA_CI_MODULE is > > really connected on TV. This means that the CA_GET_SLOT_INFO ioctl > > does not return right informtion. This is a Bug. We fix this bug. > > --- > > drivers/media/dvb-core/dvb_ca_en50221.c | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/drivers/media/dvb-core/dvb_ca_en50221.c b/drivers/media/dvb-core/dvb_ca_en50221.c > > index baf64540dc00..8d37c3c13227 100644 > > --- a/drivers/media/dvb-core/dvb_ca_en50221.c > > +++ b/drivers/media/dvb-core/dvb_ca_en50221.c > > @@ -1403,6 +1403,10 @@ static int dvb_ca_en50221_io_do_ioctl(struct file *file, > > (sl->slot_state != DVB_CA_SLOTSTATE_INVALID)) { > > info->flags = CA_CI_MODULE_PRESENT; > > } > > + if ((sl->slot_state == DVB_CA_SLOTSTATE_NONE) && > > + (sl->camchange_type == DVB_CA_EN50221_CAMCHANGE_INSERTED)) { > > + info->flags = CA_CI_MODULE_PRESENT; > > + } > > if (sl->slot_state == DVB_CA_SLOTSTATE_RUNNING) > > info->flags |= CA_CI_MODULE_READY; > > break; > > -- > > 2.17.1 > >
… > really connected on TV. This means that the CA_GET_SLOT_INFO ioctl > does not return right informtion. This is a Bug. We fix this bug. 1. I hope that typos will be avoided for subsequent change descriptions. 2. How do you think about to improve the patch further with a corresponding imperative wording? https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.9-rc6#n94 3. Would you like to add the tag “Fixes” accordingly? Regards, Markus
Dear Markus Elfring Can you explain what I shall do more ? 1. What is the "typo" ? 2. Can you take some examples which show the "Imperative wording" ? 3. Can you give me some examples for adding "Fixes" ? Do I have to use "Fixes" instead of "Fix" ? 2024년 5월 2일 (목) 오전 1:25, Markus Elfring <Markus.Elfring@web.de>님이 작성: > > … > > really connected on TV. This means that the CA_GET_SLOT_INFO ioctl > > does not return right informtion. This is a Bug. We fix this bug. > > 1. I hope that typos will be avoided for subsequent change descriptions. > > 2. How do you think about to improve the patch further with a corresponding imperative wording? > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.9-rc6#n94 > > 3. Would you like to add the tag “Fixes” accordingly? > > Regards, > Markus
> 1. What is the "typo" ? Please take another look at opportunities for spell checking also in commit messages. >>>> … right informtion. … a Bug. … > 2. Can you take some examples which show the "Imperative wording" ? I hope that you can take published guidance better into account (instead of the wording “We fix this bug.” in the discussed changelog). https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.9#n94 > 3. Can you give me some examples for adding "Fixes" ? An other tag specification might be more appropriate than the following suggestion. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?h=v6.9#n145 Fixes: a75aa90c7c2f ("media: dvb_ca_en50221: Used a helper variable") Regards, Markus
diff --git a/drivers/media/dvb-core/dvb_ca_en50221.c b/drivers/media/dvb-core/dvb_ca_en50221.c index baf64540dc00..8d37c3c13227 100644 --- a/drivers/media/dvb-core/dvb_ca_en50221.c +++ b/drivers/media/dvb-core/dvb_ca_en50221.c @@ -1403,6 +1403,10 @@ static int dvb_ca_en50221_io_do_ioctl(struct file *file, (sl->slot_state != DVB_CA_SLOTSTATE_INVALID)) { info->flags = CA_CI_MODULE_PRESENT; } + if ((sl->slot_state == DVB_CA_SLOTSTATE_NONE) && + (sl->camchange_type == DVB_CA_EN50221_CAMCHANGE_INSERTED)) { + info->flags = CA_CI_MODULE_PRESENT; + } if (sl->slot_state == DVB_CA_SLOTSTATE_RUNNING) info->flags |= CA_CI_MODULE_READY; break;
From: Yongsu yoo <yongsuyoo0215@gmail.com> Signed-off-by:Yongsu Yoo <yongsuyoo0215@gmail.com> In source/drivers/media/dvb-core/dvb_ca_en50221.c, if the CA_RESET ioctl is called, the dvb_ca_en50221_slot_shutdown will also be called. Inside of the dvb_ca_en50221_slot_shutdown, the ca->slot_info[slot].slot_state will become DVB_CA_SLOTSTATE_NONE. In the most of cases, the ca->slot_info[slot].slot_state will quickly becomes restored to other states by the subsequent operations of the thread dvb_ca_en50221_thread_state_machine. But in some rare cases, when the CA_GET_SLOT_INFO ioctl is immediately called after the CA_RESET ioctl is called, the the ca->slot_info[slot].slot_state can still remains at DVB_CA_SLOTSTATE_NONE, and this causes CA_GET_SLOT_INFO ioctl not to return CA_CI_MODULE_PRESENT as info->flags even if CA_CI_MODULE is really connected on TV. This means that the CA_GET_SLOT_INFO ioctl does not return right informtion. This is a Bug. We fix this bug. --- drivers/media/dvb-core/dvb_ca_en50221.c | 4 ++++ 1 file changed, 4 insertions(+)