Hello community, here is the log from the commit of package camsource for openSUSE:Factory checked in at 2020-02-09 20:48:43 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Comparing /work/SRC/openSUSE:Factory/camsource (Old) and /work/SRC/openSUSE:Factory/.camsource.new.26092 (New) ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Package is "camsource" Sun Feb 9 20:48:43 2020 rev:24 rq:772316 version:0.7.1 Changes: -------- --- /work/SRC/openSUSE:Factory/camsource/camsource.changes 2015-07-14 17:44:49.000000000 +0200 +++ /work/SRC/openSUSE:Factory/.camsource.new.26092/camsource.changes 2020-02-09 20:48:47.854888044 +0100 @@ -1,0 +2,15 @@ +Fri Feb 7 21:07:50 UTC 2020 - David Mair <[email protected]> + +- The camsource v4l1 module has logic errors opening the video + device that incorrectly perform a channel set when a grab window + set (frame size, fps, etc) is intended. The error handling from + the invalid channel set causes the configuration frame size and + fps settings to be discarded and replaced with the current + settings active on the video device. The result is that running a + v4l(1/2) application on a given camera before using camsource on + the same camera means camsource has the grab window attributes + from the other application not the one specified in camsource + configuration +- camsource-v4l-conf-video_window-overwrite.patch + +------------------------------------------------------------------- New: ---- camsource-v4l-conf-video_window-overwrite.patch ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Other differences: ------------------ ++++++ camsource.spec ++++++ --- /var/tmp/diff_new_pack.B80dh1/_old 2020-02-09 20:48:48.742888551 +0100 +++ /var/tmp/diff_new_pack.B80dh1/_new 2020-02-09 20:48:48.746888554 +0100 @@ -1,7 +1,7 @@ # # spec file for package camsource # -# Copyright (c) 2015 SUSE LINUX GmbH, Nuernberg, Germany. +# Copyright (c) 2020 SUSE LLC # # All modifications and additions to the file contributed by third parties # remain the property of their copyright owners, unless otherwise agreed @@ -12,7 +12,7 @@ # license that conforms to the Open Source Definition (Version 1.9) # published by the Open Source Initiative. -# Please submit bugfixes or comments via http://bugs.opensuse.org/ +# Please submit bugfixes or comments via https://bugs.opensuse.org/ # @@ -26,12 +26,13 @@ Version: 0.7.1 Release: 0 Summary: Camsource Grabs Images from a Video4Linux Device -License: GPL-2.0+ +License: GPL-2.0-or-later Group: Amusements/Toys/Graphics -Url: http://camsource.sourceforge.net +URL: http://camsource.sourceforge.net Source: camsource-%{version}.tar.gz Source1: camsource-rpmlintrc Patch0: camsource-no_implicit_decls.diff +Patch1: camsource-v4l-conf-video_window-overwrite.patch BuildRoot: %{_tmppath}/%{name}-%{version}-build %description @@ -53,6 +54,7 @@ %prep %setup -q %patch0 -p1 +%patch1 -p1 chmod ugo+x configure %build ++++++ camsource-v4l-conf-video_window-overwrite.patch ++++++ >From 03bfd892ee6a6454d92b10d1afd1f86fa0910fba Mon Sep 17 00:00:00 2001 From: David Mair <[email protected]> Date: Fri, 7 Feb 2020 13:39:12 -0700 Subject: [PATCH] opendev() in input_v4l.c functional execution is not implemented as described by the output messages it can display. One result is that it reads the configuration width and height settings and prepares them as the video source settings. But, it makes two unrelated attempts to set the video device source channel and if the second one succeeds it reads the current device grab window settings into the same structure it used to load the configuration, overwriting and losing the configuration. It then sets the grab window using the value it just got from the same device, i.e. redundant set device grab window to current grab window. It has the effect that if: * Another video application is used before running camsource - for the device camsource will use - Uses frame resolution different from that in the camsource configuration * Then the other application exited * camsource started The frame resolution used by camsource will be that set by the other application and not that in the camsource configuration. When reading what can be output to the console from the second set channel it says the "set grab window failed: Trying _again_ without the fps option" and performs a "set grab window" operation as the supposed "again" of a set video channel operation (when a full set video channel was already performed earlier in the code anyway). I modified the logic to operate as the messaging describes: * Perform a "get grab window" before loading the configuration * Load the configuration over the get grab window result * Perform a "set grab window" with the fps option * If that "set grab window" fails: - Perform a "set grab window" _again_ but without the fps This matches the behavior described in the messaging and camsource no longer loses the settings loaded from configuration. --- src/input_v4l.c | 11 ++++------- 1 file changed, 4 insertions(+), 7 deletions(-) diff --git a/src/input_v4l.c b/src/input_v4l.c index efbd7b3..f247d9b 100644 --- a/src/input_v4l.c +++ b/src/input_v4l.c @@ -145,7 +145,9 @@ closenerr: printf("ioctl \"set input channel\" failed, continuing anyway: %s\n", strerror(errno)); } - memset(&vidwin, 0, sizeof(vidwin)); + ret = v4l1_ioctl(newcamdev.fd, VIDIOCGWIN, &vidwin); + if (ret != 0) + memset(&vidwin, 0, sizeof(vidwin)); newcamdev.autobrightness = autobrightness; @@ -158,7 +160,7 @@ closenerr: vidwin.flags |= (fps & 0x3f) << 16; - ret = v4l1_ioctl(newcamdev.fd, VIDIOCSCHAN, &vidchan); + ret = v4l1_ioctl(newcamdev.fd, VIDIOCSWIN, &vidwin); if (ret != 0) { printf("ioctl \"set grab window\" failed: %s\n", strerror(errno)); @@ -180,11 +182,6 @@ closenerr: printf("we will continue anyway and hope that everything goes ok.\n"); } } - else { - ret = v4l1_ioctl(newcamdev.fd, VIDIOCGWIN, &vidwin); - if (!ret) - v4l1_ioctl(newcamdev.fd, VIDIOCSWIN, &vidwin); - } ret = v4l1_ioctl(newcamdev.fd, VIDIOCGPICT, &newcamdev.vidpic); if (ret != 0) -- 2.25.0
