| Message ID | 20200605162518.28099-1-florian.fainelli@broadcom.com (mailing list archive) |
|---|---|
| Headers | show |
| Series | Unbreak 32-bit DVB applications on 64-bit kernels | expand |
On 6/5/2020 9:24 AM, Florian Fainelli wrote: > Hi all, > > This long patch series was motivated by backporting Jaedon's changes > which add a proper ioctl compatibility layer for 32-bit applications > running on 64-bit kernels. We have a number of Android TV-based products > currently running on the 4.9 kernel and this was broken for them. > > Thanks to Robert McConnell for identifying and providing the patches in > their initial format. > > In order for Jaedon's patches to apply cleanly a number of changes were > applied to support those changes. If you deem the patch series too big > please let me know. Mauro, can you review this? I would prefer not to maintain those patches in our downstream 4.9 kernel as there are quite a few of them, and this is likely beneficial to other people. Thank you! > > Thanks > > Colin Ian King (2): > media: dvb_frontend: ensure that inital front end status initialized > media: dvb_frontend: initialize variable s with FE_NONE instead of 0 > > Jaedon Shin (3): > media: dvb_frontend: Add unlocked_ioctl in dvb_frontend.c > media: dvb_frontend: Add compat_ioctl callback > media: dvb_frontend: Add commands implementation for compat ioct > > Katsuhiro Suzuki (1): > media: dvb_frontend: fix wrong cast in compat_ioctl > > Mauro Carvalho Chehab (14): > media: dvb/frontend.h: move out a private internal structure > media: dvb/frontend.h: document the uAPI file > media: dvb_frontend: get rid of get_property() callback > media: stv0288: get rid of set_property boilerplate > media: stv6110: get rid of a srate dead code > media: friio-fe: get rid of set_property() > media: dvb_frontend: get rid of set_property() callback > media: dvb_frontend: cleanup dvb_frontend_ioctl_properties() > media: dvb_frontend: cleanup ioctl handling logic > media: dvb_frontend: get rid of property cache's state > media: dvb_frontend: better document the -EPERM condition > media: dvb_frontend: fix return values for FE_SET_PROPERTY > media: dvb_frontend: be sure to init dvb_frontend_handle_ioctl() > return code > media: dvb_frontend: fix return error code > > Satendra Singh Thakur (1): > media: dvb_frontend: dtv_property_process_set() cleanups > > .../media/uapi/dvb/fe-get-property.rst | 7 +- > drivers/media/dvb-core/dvb_frontend.c | 571 +++++++++++------ > drivers/media/dvb-core/dvb_frontend.h | 13 - > drivers/media/dvb-frontends/lg2160.c | 14 - > drivers/media/dvb-frontends/stv0288.c | 7 - > drivers/media/dvb-frontends/stv6110.c | 9 - > drivers/media/usb/dvb-usb/friio-fe.c | 24 - > fs/compat_ioctl.c | 17 - > include/uapi/linux/dvb/frontend.h | 592 +++++++++++++++--- > 9 files changed, 881 insertions(+), 373 deletions(-) >
On 6/11/2020 9:45 PM, Florian Fainelli wrote: > > > On 6/5/2020 9:24 AM, Florian Fainelli wrote: >> Hi all, >> >> This long patch series was motivated by backporting Jaedon's changes >> which add a proper ioctl compatibility layer for 32-bit applications >> running on 64-bit kernels. We have a number of Android TV-based products >> currently running on the 4.9 kernel and this was broken for them. >> >> Thanks to Robert McConnell for identifying and providing the patches in >> their initial format. >> >> In order for Jaedon's patches to apply cleanly a number of changes were >> applied to support those changes. If you deem the patch series too big >> please let me know. > > Mauro, can you review this? I would prefer not to maintain those patches > in our downstream 4.9 kernel as there are quite a few of them, and this > is likely beneficial to other people. Hello? Anybody here?
Hey Florian, Thank you for the time and effort that you put into this patch series. I was excited to see this, when I first saw it posted a few weeks ago. I have every intention of giving it a review, but just haven't found the time yet. I'm sure that Mauro would say the same. I'm sure that he and I both will find some time, hopefully over the next few weeks or sooner, to give this a thorough review and provide some feedback. Hopefully we can put this on its way for merge soon. Please bear with us.. Thanks again for your contribution. -Mike Krufky On Wed, Jun 17, 2020 at 12:39 AM Florian Fainelli <f.fainelli@gmail.com> wrote: > > > > On 6/11/2020 9:45 PM, Florian Fainelli wrote: > > > > > > On 6/5/2020 9:24 AM, Florian Fainelli wrote: > >> Hi all, > >> > >> This long patch series was motivated by backporting Jaedon's changes > >> which add a proper ioctl compatibility layer for 32-bit applications > >> running on 64-bit kernels. We have a number of Android TV-based products > >> currently running on the 4.9 kernel and this was broken for them. > >> > >> Thanks to Robert McConnell for identifying and providing the patches in > >> their initial format. > >> > >> In order for Jaedon's patches to apply cleanly a number of changes were > >> applied to support those changes. If you deem the patch series too big > >> please let me know. > > > > Mauro, can you review this? I would prefer not to maintain those patches > > in our downstream 4.9 kernel as there are quite a few of them, and this > > is likely beneficial to other people. > > Hello? Anybody here? > -- > Florian
On Wed, Jun 17, 2020 at 12:39 AM Florian Fainelli <f.fainelli@gmail.com> wrote: > > > > On 6/11/2020 9:45 PM, Florian Fainelli wrote: > > > > > > On 6/5/2020 9:24 AM, Florian Fainelli wrote: > >> Hi all, > >> > >> This long patch series was motivated by backporting Jaedon's changes > >> which add a proper ioctl compatibility layer for 32-bit applications > >> running on 64-bit kernels. We have a number of Android TV-based products > >> currently running on the 4.9 kernel and this was broken for them. > >> > >> Thanks to Robert McConnell for identifying and providing the patches in > >> their initial format. > >> > >> In order for Jaedon's patches to apply cleanly a number of changes were > >> applied to support those changes. If you deem the patch series too big > >> please let me know. > > > > Mauro, can you review this? I would prefer not to maintain those patches > > in our downstream 4.9 kernel as there are quite a few of them, and this > > is likely beneficial to other people. > > Hello? Anybody here? > -- > Florian Ouch. I top-posted - oops! Please reply on this email rather than the previous. Hey Florian, Thank you for the time and effort that you put into this patch series. I was excited to see this, when I first saw it posted a few weeks ago. I have every intention of giving it a review, but just haven't found the time yet. I'm sure that Mauro would say the same. I'm sure that he and I both will find some time, hopefully over the next few weeks or sooner, to give this a thorough review and provide some feedback. Hopefully we can put this on its way for merge soon. Please bear with us.. Thanks again for your contribution. -Mike Krufky
On Fri, Jun 05, 2020 at 09:24:57AM -0700, Florian Fainelli wrote: > Hi all, > > This long patch series was motivated by backporting Jaedon's changes > which add a proper ioctl compatibility layer for 32-bit applications > running on 64-bit kernels. We have a number of Android TV-based products > currently running on the 4.9 kernel and this was broken for them. > > Thanks to Robert McConnell for identifying and providing the patches in > their initial format. > > In order for Jaedon's patches to apply cleanly a number of changes were > applied to support those changes. If you deem the patch series too big > please let me know. Now queued up,t hanks. greg k-h
On 6/23/2020 12:13 PM, Greg KH wrote: > On Fri, Jun 05, 2020 at 09:24:57AM -0700, Florian Fainelli wrote: >> Hi all, >> >> This long patch series was motivated by backporting Jaedon's changes >> which add a proper ioctl compatibility layer for 32-bit applications >> running on 64-bit kernels. We have a number of Android TV-based products >> currently running on the 4.9 kernel and this was broken for them. >> >> Thanks to Robert McConnell for identifying and providing the patches in >> their initial format. >> >> In order for Jaedon's patches to apply cleanly a number of changes were >> applied to support those changes. If you deem the patch series too big >> please let me know. > > Now queued up,t hanks. Thanks a lot, I did not get an email about "[PATCH stable 4.9 02/21] media: dvb_frontend: initialize variable s with FE_NONE instead of 0" being applied, not that it is a very important change, > > greg k-h >
On Wed, Jun 24, 2020 at 08:41:06AM -0700, Florian Fainelli wrote: > > > On 6/23/2020 12:13 PM, Greg KH wrote: > > On Fri, Jun 05, 2020 at 09:24:57AM -0700, Florian Fainelli wrote: > >> Hi all, > >> > >> This long patch series was motivated by backporting Jaedon's changes > >> which add a proper ioctl compatibility layer for 32-bit applications > >> running on 64-bit kernels. We have a number of Android TV-based products > >> currently running on the 4.9 kernel and this was broken for them. > >> > >> Thanks to Robert McConnell for identifying and providing the patches in > >> their initial format. > >> > >> In order for Jaedon's patches to apply cleanly a number of changes were > >> applied to support those changes. If you deem the patch series too big > >> please let me know. > > > > Now queued up,t hanks. > > Thanks a lot, I did not get an email about "[PATCH stable 4.9 02/21] > media: dvb_frontend: initialize variable s with FE_NONE instead of 0" > being applied, not that it is a very important change, That should be there, I merged it before I did the whole long series and realised I should just automate it, sorry about that. thanks, greg k-h
On Tue, Jun 23, 2020 at 09:13:34PM +0200, Greg KH wrote: > On Fri, Jun 05, 2020 at 09:24:57AM -0700, Florian Fainelli wrote: > > Hi all, > > > > This long patch series was motivated by backporting Jaedon's changes > > which add a proper ioctl compatibility layer for 32-bit applications > > running on 64-bit kernels. We have a number of Android TV-based products > > currently running on the 4.9 kernel and this was broken for them. > > > > Thanks to Robert McConnell for identifying and providing the patches in > > their initial format. > > > > In order for Jaedon's patches to apply cleanly a number of changes were > > applied to support those changes. If you deem the patch series too big > > please let me know. > > Now queued up,t hanks. Sorry about the delay getting this reviewed. I've spent the morning going through them and it looks good. Of the all the dvb ioctl, only the FE_SET_PROPERTY and FE_GET_PROPERTY ioctls need special handling and this series fixes that. Belated, Reviewed-by: Sean Young <sean@mess.org> Sean
Hi all, This long patch series was motivated by backporting Jaedon's changes which add a proper ioctl compatibility layer for 32-bit applications running on 64-bit kernels. We have a number of Android TV-based products currently running on the 4.9 kernel and this was broken for them. Thanks to Robert McConnell for identifying and providing the patches in their initial format. In order for Jaedon's patches to apply cleanly a number of changes were applied to support those changes. If you deem the patch series too big please let me know. Thanks Colin Ian King (2): media: dvb_frontend: ensure that inital front end status initialized media: dvb_frontend: initialize variable s with FE_NONE instead of 0 Jaedon Shin (3): media: dvb_frontend: Add unlocked_ioctl in dvb_frontend.c media: dvb_frontend: Add compat_ioctl callback media: dvb_frontend: Add commands implementation for compat ioct Katsuhiro Suzuki (1): media: dvb_frontend: fix wrong cast in compat_ioctl Mauro Carvalho Chehab (14): media: dvb/frontend.h: move out a private internal structure media: dvb/frontend.h: document the uAPI file media: dvb_frontend: get rid of get_property() callback media: stv0288: get rid of set_property boilerplate media: stv6110: get rid of a srate dead code media: friio-fe: get rid of set_property() media: dvb_frontend: get rid of set_property() callback media: dvb_frontend: cleanup dvb_frontend_ioctl_properties() media: dvb_frontend: cleanup ioctl handling logic media: dvb_frontend: get rid of property cache's state media: dvb_frontend: better document the -EPERM condition media: dvb_frontend: fix return values for FE_SET_PROPERTY media: dvb_frontend: be sure to init dvb_frontend_handle_ioctl() return code media: dvb_frontend: fix return error code Satendra Singh Thakur (1): media: dvb_frontend: dtv_property_process_set() cleanups .../media/uapi/dvb/fe-get-property.rst | 7 +- drivers/media/dvb-core/dvb_frontend.c | 571 +++++++++++------ drivers/media/dvb-core/dvb_frontend.h | 13 - drivers/media/dvb-frontends/lg2160.c | 14 - drivers/media/dvb-frontends/stv0288.c | 7 - drivers/media/dvb-frontends/stv6110.c | 9 - drivers/media/usb/dvb-usb/friio-fe.c | 24 - fs/compat_ioctl.c | 17 - include/uapi/linux/dvb/frontend.h | 592 +++++++++++++++--- 9 files changed, 881 insertions(+), 373 deletions(-)