Introduce v4l2-get-device tool
PATCH 1/1 Introduce v4l2-get-device tool
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Usage:
v4l2-get-device -d uvcvideo -c V4L2_CAP_VIDEO_CAPTURE /dev/video0/dev/video2
--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.c
configure.ac +1 -0
@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefile
utils/Makefile.am +1 -0
@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \
utils/v4l2-get-device/.gitignore new file +1 -0
@@ -0,0 +1 @@ +v4l2-get-device
utils/v4l2-get-device/Makefile.am new file +4 -0
@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)
utils/v4l2-get-device/v4l2-get-device.c new file +147 -0
147 lines not on this page · raw message
v4l-utils.spec.in +1 -0
@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*
Signature · 1 line
2.19.1
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
Quoted 111 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +44
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ;
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 10 lines
utils/v4l2-get-device/v4l2-get-device.c new file +10
-0+const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver; +unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },
I'd like a bus-info option as well, if possible.
Quoted 10 lines
utils/v4l2-get-device/v4l2-get-device.c new file +10
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) {
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 10 lines
utils/v4l2-get-device/v4l2-get-device.c new file +10
-0+if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" )) + caps |=V4L2_CAP_VIDEO_OVERLAY ; +else if (!strcmp (s,"V4L2_CAP_VBI_CAPTURE" )) + caps |=V4L2_CAP_VBI_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VBI_OUTPUT" )) + caps |=V4L2_CAP_VBI_OUTPUT ;
Let's support all CAPS here. I'd also drop the V4L2_CAP_ prefix (or make it optional) and make the strcmp case-insensitive to ease typing.
Quoted 29 lines
utils/v4l2-get-device/v4l2-get-device.c new file +29
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[])
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 10 lines
utils/v4l2-get-device/v4l2-get-device.c new file +10
-0+{ +const char *vid; +struct args args; +void *md; + + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();
I never really liked this. My main problem is that it doesn't know about media devices.
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0
v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*
Regards,
Hans
Quoted 8 lines · Hans Verkuil
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
Quoted 114 lines
Quoted 114 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;
Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))
Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.
OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;
Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Sorry, not sure what you don't really like?
The discover_media_devices() is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).
Perhaps we can use this for now and pospone improving the discovery library?
Quoted 7 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.
Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 51 lines
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*Regards,
Hans
Quoted 5 lines · Ezequiel Garcia
Quoted 4 lines · Hans Verkuil
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Quoted 8 lines
Quoted 4 lines
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
I agree with that.
Quoted 197 lines
Quoted 115 lines
Quoted 115 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Quoted 6 lines
Sorry, not sure what you don't really like?
The
discover_media_devices()is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).Perhaps we can use this for now and pospone improving the discovery library?
Yes, that can be postponed (I should have been clear about that, sorry).
Quoted 9 lines
Quoted 8 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
Quoted 5 lines
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
Not for codecs, but this should become a generic utility that you can also use to find e.g. v4l-subdev device nodes from a complex camera driver.
Regards,
Hans
Quoted 56 lines
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 51 lines
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*Regards,
Hans
Quoted 11 lines · Hans Verkuil
Quoted 10 lines · Ezequiel Garcia
Quoted 8 lines · Hans Verkuil
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
Quoted 6 lines
Quoted 3 lines
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
I agree with that.
Great!
About adding it to v4l2-sysfs-path, is there any scenario where we'll want to use udev-based scanning instead of sysfs-based? Maybe we can rename the tool, and still install a v4l2-sysfs-path alias?
Quoted 213 lines
Quoted 196 lines
Quoted 114 lines
Quoted 114 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Quoted 6 lines
Sorry, not sure what you don't really like?
The
discover_media_devices()is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).Perhaps we can use this for now and pospone improving the discovery library?
Yes, that can be postponed (I should have been clear about that, sorry).
Quoted 8 lines
Quoted 7 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
Quoted 9 lines
Quoted 5 lines
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
Not for codecs, but this should become a generic utility that you can also use to find e.g. v4l-subdev device nodes from a complex camera driver.
Ah, that makes sense.
Thanks for the quick review!
Quoted 58 lines
Regards,
Hans
Quoted 54 lines
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 51 lines
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*Regards,
Hans
Resurrecting this thread, since this functionality would be really useful :-)
Quoted 16 lines · Ezequiel Garcia
Quoted 15 lines · Hans Verkuil
Quoted 13 lines · Ezequiel Garcia
Quoted 8 lines · Hans Verkuil
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
Quoted 8 lines
Quoted 2 lines
I agree with that.
Great!
About adding it to v4l2-sysfs-path, is there any scenario where we'll want to use udev-based scanning instead of sysfs-based? Maybe we can rename the tool, and still install a v4l2-sysfs-path alias?
To be honest, I don't know. It appears that sysfs is the most generic approach and I have seen cases where udev isn't installed on an embedded system.
So I feel that sysfs is the best approach for now.
BTW, I've added support for swradio and v4l-touch devices to v4l2-sysfs-path, but I noticed that it didn't recognize media devices, I need to dig into it a bit to understand how to find those devices.
Regards,
Hans
Quoted 288 lines
Quoted 214 lines
Quoted 197 lines
Quoted 115 lines
Quoted 115 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Quoted 6 lines
Sorry, not sure what you don't really like?
The
discover_media_devices()is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).Perhaps we can use this for now and pospone improving the discovery library?
Yes, that can be postponed (I should have been clear about that, sorry).
Quoted 8 lines
Quoted 7 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
Quoted 9 lines
Quoted 5 lines
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
Not for codecs, but this should become a generic utility that you can also use to find e.g. v4l-subdev device nodes from a complex camera driver.
Ah, that makes sense.
Thanks for the quick review!
Quoted 58 lines
Regards,
Hans
Quoted 54 lines
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 51 lines
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*Regards,
Hans
Quoted 2 lines · Hans Verkuil
Resurrecting this thread, since this functionality would be really useful :-)
Thanks for the resurrection chant :-)
Quoted 26 lines · Ezequiel Garcia
Quoted 16 lines · Hans Verkuil
Quoted 15 lines · Ezequiel Garcia
Quoted 13 lines · Hans Verkuil
Quoted 8 lines · Ezequiel Garcia
Quoted 2 lines
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
Quoted 8 lines
Quoted 2 lines
I agree with that.
Great!
About adding it to v4l2-sysfs-path, is there any scenario where we'll want to use udev-based scanning instead of sysfs-based? Maybe we can rename the tool, and still install a v4l2-sysfs-path alias?
Quoted 9 lines
To be honest, I don't know. It appears that sysfs is the most generic approach and I have seen cases where udev isn't installed on an embedded system.
So I feel that sysfs is the best approach for now.
BTW, I've added support for swradio and v4l-touch devices to v4l2-sysfs-path, but I noticed that it didn't recognize media devices, I need to dig into it a bit to understand how to find those devices.
OK, I will try to post a new version shortly.
Thanks,
Eze
Quoted 289 lines
Regards,
Hans
Quoted 285 lines
Quoted 213 lines
Quoted 196 lines
Quoted 114 lines
Quoted 114 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Quoted 6 lines
Sorry, not sure what you don't really like?
The
discover_media_devices()is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).Perhaps we can use this for now and pospone improving the discovery library?
Yes, that can be postponed (I should have been clear about that, sorry).
Quoted 8 lines
Quoted 7 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
Quoted 9 lines
Quoted 5 lines
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
Not for codecs, but this should become a generic utility that you can also use to find e.g. v4l-subdev device nodes from a complex camera driver.
Ah, that makes sense.
Thanks for the quick review!
Quoted 58 lines
Regards,
Hans
Quoted 54 lines
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 51 lines
Quoted 40 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +0
-0v4l-utils.spec.in gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : · line 165
Quoted 7 lines
v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*Regards,
Hans
Hello,
Quoted 23 lines · Hans Verkuil
Resurrecting this thread, since this functionality would be really useful :-)
Quoted 20 lines · Ezequiel Garcia
Quoted 11 lines · Hans Verkuil
Quoted 10 lines · Ezequiel Garcia
Quoted 8 lines · Hans Verkuil
Quoted 2 lines · Ezequiel Garcia
This tool allows to find a device by driver name, this is useful for scripts to be written in a generic way.
Why not add support for this to v4l2-sysfs-path? And improve it at the same time (swradio devices do not show up when I run v4l2-sysfs-path, I also suspect v4l-touch devices aren't recognized. Ditto for media devices.
Quoted 5 lines
Quoted 3 lines
I can add the functionality to v4l2-sysfs-path, and we can document the rest in some TODO list, as I don't think we need to get everything solved right now :-)
I agree with that.
Great!
Quoted 8 lines
Quoted 3 lines
About adding it to v4l2-sysfs-path, is there any scenario where we'll want to use udev-based scanning instead of sysfs-based? Maybe we can rename the tool, and still install a v4l2-sysfs-path alias?
To be honest, I don't know. It appears that sysfs is the most generic approach and I have seen cases where udev isn't installed on an embedded system.
So I feel that sysfs is the best approach for now.
The problem with sysfs is that it won't give you the device node path. You can use it to get the device node name known to the kernel, but if the device node gets renamed by a udev policy, you won't know. That's why it's better to use udev if available, and fall back on sysfs otherwise.
Please see below for an additional comment.
Quoted 211 lines
BTW, I've added support for swradio and v4l-touch devices to v4l2-sysfs-path, but I noticed that it didn't recognize media devices, I need to dig into it a bit to understand how to find those devices.
Quoted 207 lines
Quoted 207 lines
Quoted 195 lines
Quoted 114 lines
Quoted 114 lines
Usage:
v4l2-get-device -d uvcvideo -c
V4L2_CAP_VIDEO_CAPTURE/dev/video0/dev/video2Signed-off-by Ezequiel Garcia <[email protected]>--- configure.ac | 1 + utils/Makefile.am | 1 + utils/v4l2-get-device/.gitignore | 1 + utils/v4l2-get-device/Makefile.am | 4 + utils/v4l2-get-device/v4l2-get-device.c | 147 ++++++++++++++++++++++++ v4l-utils.spec.in | 1 + 6 files changed, 155 insertions(+) create mode 100644 utils/v4l2-get-device/.gitignore create mode 100644 utils/v4l2-get-device/Makefile.am create mode 100644 utils/v4l2-get-device/v4l2-get-device.cconfigure.ac +1
-0@@ -31,6 +31,7 @@ AC_CONFIG_FILES([Makefile utils/v4l2-compliance/Makefile utils/v4l2-ctl/Makefile utils/v4l2-dbg/Makefile + utils/v4l2-get-device/Makefile utils/v4l2-sysfs-path/Makefile utils/qv4l2/Makefile utils/cec-ctl/Makefileutils/Makefile.am +1
-0@@ -9,6 +9,7 @@ SUBDIRS = \ v4l2-compliance \ v4l2-ctl \ v4l2-dbg \ + v4l2-get-device \ v4l2-sysfs-path \ cec-ctl \ cec-compliance \utils/v4l2-get-device/.gitignore new file +1
-0@@ -0,0 +1 @@ +v4l2-get-deviceutils/v4l2-get-device/Makefile.am new file +4
-0@@ -0,0 +1,4 @@ +bin_PROGRAMS = v4l2-get-device +v4l2_get_device_SOURCES = v4l2-get-device.c +v4l2_get_device_LDADD = ../libmedia_dev/libmedia_dev.la +v4l2_get_device_LDFLAGS = $(ARGP_LIBS)utils/v4l2-get-device/v4l2-get-device.c new file +48
-0@@ -0,0 +1,147 @@ +/* +* Copyright © 2018 Collabora, Ltd. +* +* Based on v4l2-sysfs-path by Mauro Carvalho Chehab: +* +* Copyright © 2011 Red Hat, Inc. +* +* Permission to use, copy, modify, distribute, and sell this software +* and its documentation for any purpose is hereby granted without +* fee, provided that the above copyright notice appear in all copies +* and that both that copyright notice and this permission notice +* appear in supporting documentation, and that the name of Red Hat +* not be used in advertising or publicity pertaining to distribution +* of the software without specific, written prior permission. Red +* Hat makes no representations about the suitability of this software +* for any purpose. It is provided "as is" without express or implied +* warranty. +* +* THE AUTHORS DISCLAIM ALL WARRANTIES WITH REGARD TO THIS SOFTWARE, +* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS, IN +* NO EVENT SHALL THE AUTHORS BE LIABLE FOR ANY SPECIAL, INDIRECT OR +* CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS +* OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, +* NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN +* CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE. +* +*/ + +#include <argp.h> +#include <config.h> +#include <fcntl.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <sys/ioctl.h> +#include <sys/unistd.h> + +#include <linux/videodev2.h> + +#include "../libmedia_dev/get_media_devices.h" + +const char *argp_program_version ="v4l2-get-device " V4L_UTILS_VERSION ; +const char *argp_program_bug_address ="Ezequiel Garcia <[email protected]>" ; + +struct args { +const char *driver;Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 54
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+unsigned int device_caps; +}; + +static const struct argp_option options[] = { + {"driver" ,'d' ,"DRIVER" ,0 ,"device driver name" ,0 }, + {"v4l2-device-caps" ,'c' ,"CAPS" ,0 ,"v4l2 device capabilities" ,0 },I'd like a bus-info option as well, if possible.
Guess that works.
Quoted 15 lines
Quoted 15 lines
utils/v4l2-get-device/v4l2-get-device.c new file +15
-0+ { 0 } +}; + +static unsigned int parse_capabilities (const char *arg) +{ +char *s, *str =strdup (arg); +unsigned int caps =0 ; + + s =strtok (str,"," ); +while (s !=NULL ) { +if (!strcmp (s,"V4L2_CAP_VIDEO_CAPTURE" )) + caps |=V4L2_CAP_VIDEO_CAPTURE ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OUTPUT" )) + caps |=V4L2_CAP_VIDEO_OUTPUT ; +else if (!strcmp (s,"V4L2_CAP_VIDEO_OVERLAY" ))Quoted 9 lines
utils/v4l2-get-device/v4l2-get-device.c line 74
Quoted 5 lines
utils/v4l2-get-device/v4l2-get-device.c new file +5
-0+ caps |=V4L2_CAP_VIDEO_OVERLAY ; + else if (!strcmp(s, "V4L2_CAP_VBI_CAPTURE")) + caps |=V4L2_CAP_VBI_CAPTURE ; + else if (!strcmp(s, "V4L2_CAP_VBI_OUTPUT")) + caps |=V4L2_CAP_VBI_OUTPUT ;Let's support all CAPS here. I'd also drop the
V4L2_CAP_prefix (or make it optional) and make the strcmp case-insensitive to ease typing.OK.
Quoted 33 lines
Quoted 33 lines
utils/v4l2-get-device/v4l2-get-device.c new file +33
-0+ s =strtok (NULL ,"," ); + } +free (str); +return caps; +} + +static error_t parse_opt (int k,char *arg,struct argp_state *state) +{ +struct args *args = state->input; + +switch (k) { +case 'd' : + args->driver = arg; +break ; +case 'c' : + args->device_caps =parse_capabilities (arg); +break ; +default : +return ARGP_ERR_UNKNOWN ; + } +return 0 ; +} + +static struct argp argp = { + .options = options, + .parser = parse_opt, +}; + +int main (int argc,char *argv[]) +{ +const char *vid; +struct args args; +void *md;Quoted 8 lines
utils/v4l2-get-device/v4l2-get-device.c line 113
Quoted 6 lines
utils/v4l2-get-device/v4l2-get-device.c new file +6
-0+ + args.driver =NULL ; + args.device_caps =0 ; +argp_parse (&argp, argc, argv,0 ,0 , &args); + + md =discover_media_devices ();I never really liked this. My main problem is that it doesn't know about media devices.
Quoted 6 lines
Sorry, not sure what you don't really like?
The
discover_media_devices()is maybe not the best API, but it's what v4l2-sysfs-path uses, and it seemed to fit what we need as a first step (to find v4l2 devices).Perhaps we can use this for now and pospone improving the discovery library?
Yes, that can be postponed (I should have been clear about that, sorry).
Quoted 3 lines
Quoted 3 lines
In my view it should look for media devices first, query them and get all the device nodes referenced in the topology, and then fall back to the old method to discover any remaining device nodes for drivers that do not create a media device.
I agree with Hans here. It's about time to make new development media device-centric, and only fall back to legacy methods for legacy drivers.
Quoted 64 lines
Quoted 64 lines
Quoted 11 lines
Quoted 8 lines
Quoted 2 lines
You need media device support anyway since you also want to be able to query the media device for a specific driver and find the device node for a specific entity.
I've been thinking about this (since I started with http://git.ideasonboard.org/media-enum.git),
but to be honest, I failed to see why I would want to query media devices.Let's say we want to find a H264 decoder, I don't think the topology is be of much use, is it?
Not for codecs, but this should become a generic utility that you can also use to find e.g. v4l-subdev device nodes from a complex camera driver.
Ah, that makes sense.
Thanks for the quick review!
Quoted 48 lines
Quoted 48 lines
In any case, I think this is part of a bigger discussion, would you consider merging v4l discovery for now?
Quoted 45 lines
Quoted 45 lines
utils/v4l2-get-device/v4l2-get-device.c new file +34
-0+ + vid =NULL ; +do { +struct v4l2_capability cap; +char devnode[64 ]; +int ret; +int fd; + + vid =get_associated_device (md, vid,MEDIA_V4L_VIDEO , +NULL ,NONE ); +if (!vid) +break ; +snprintf (devnode,64 ,"/dev/%s" , vid); + fd =open (devnode,O_RDWR ); +if (fd <0 ) +continue ; + +memset (&cap,0 ,sizeof cap); + ret =ioctl (fd,VIDIOC_QUERYCAP , &cap); +if (ret) { +close (fd); +continue ; + } +close (fd); + +if (strncmp (args.driver, (char *)cap.driver,sizeof (cap.driver))) +continue ; +if (args.device_caps && (args.device_caps & cap.device_caps) != args.device_caps) +continue ; +fprintf (stdout,"%s\n" , devnode); + }while (vid); +free_media_devices (md); +return 0 ; +}v4l-utils.spec.in +1
-0@@ -159,6 +159,7 @@ gtk-update-icon-cache %{_datadir}/icons/hicolor &>/dev/null || : %{_bindir}/ivtv-ctl %{_bindir}/v4l2-ctl %{_bindir}/v4l2-sysfs-path +%{_bindir}/v4l2-get-device %{_mandir}/man1/ir-keytable.1* %{_mandir}/man1/ir-ctl.1*
Signature · 3 lines
Regards, Laurent Pinchart