- 03 Feb, 2017 34 commits
-
-
Mats Randgaard authored
The number of CSI lanes that should be used is set to the CSI_CONTROL register by indirectly writing to the CSI_CONFW register. When the number of lanes is read back from the CSI_CONTROL register the value is usually correct, but we have seen that it suddenly is 1 for a short moment before the correct value is restored again. Toshiba have not figured out why that happen, but we have found it safer to store the value in the driver. Signed-off-by:
Mats Randgaard <matrandg@cisco.com> Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@s-opensource.com>
-
Bhumika Goyal authored
Check for v4l2_subdev_ops structures that are only passed as an argument to the function v4l2_subdev_init. This argument is of type const, so v4l2_subdev_ops structures having this property can also be declared const. Done using Coccinelle: @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_ops i@p = {...}; @ok1@ identifier r1.i; position p; @@ v4l2_subdev_init(...,&i@p) @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_ops i; File size before: text data bss dec hex filename 17171 1912 20 19103 4a9f platform/s3c-camif/camif-capture.o File size after: text data bss dec hex filename 17235 1848 20 19103 4a9f platform/s3c-camif/camif-capture.o Signed-off-by:Bhumika Goyal <bhumirks@gmail.com> Acked-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
-
Martin Wache authored
dib7000p_read_word() may return zero on i2c errors, resulting in dib7000p_get_internal_freq() returning zero. So don't divide by the result of dib7000p_get_internal_freq() without checking it for zero in dib7000p_set_dds(). On one of my machines the device ID 2304:0229 Pinnacle Systems, Inc. PCTV Dual DVB-T 2001e about once a day/every two days gets into a state, where most (all?) I2C reads return with an error. Tuning during this state will result in a divide by zero without this patch. This patch doesn't fix the root cause for the device getting into a bad state, but it allows me to unload/reload the drivers, bringing it back into a usable state. Signed-off-by:
Martin Wache <M.Wache@gmx.net> Signed-off-by:
-
Bhumika Goyal authored
v4l2_subdev_{core/video}_ops structures are stored in the fields of the v4l2_subdev_ops structure which are of type const. Also, v4l2_subdev_ops structure is passed to a function having its argument of type const. As these structures are never modified, so declare them as const. Done using Coccinelle: (One of the scripts used) @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_ops i@p = {...}; @ok1@ identifier r1.i; position p; @@ v4l2_i2c_subdev_init(...,&i@p) @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_ops i; File size before: text data bss dec hex filename 2732 608 8 3348 d14 drivers/media/i2c/ml86v7667.o File size after: text data bss dec hex filename 3100 232 8 3340 d0c drivers/media/i2c/ml86v7667.o Signed-off-by: -
Bhumika Goyal authored
v4l2_subdev_{core/video}_ops structures are stored in the fields of the v4l2_subdev_ops structure which are of type const. Also, v4l2_subdev_ops structure is passed to a function having its argument of type const. As these structures are never modified, so declare them as const. Done using Coccinelle: (One of the scripts used) @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_video_ops i@p = {...}; @ok1@ identifier r1.i; position p; struct v4l2_subdev_ops obj; @@ obj.video=&i@p; @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_video_ops i; File size before: text data bss dec hex filename 1809 872 0 2681 a79 drivers/media/i2c/ak881x.o File size after: text data bss dec hex filename 2185 496 0 2681 a79 drivers/media/i2c/ak881x.o Signed-off-by: -
Bhumika Goyal authored
v4l2_subdev_pad_ops structures are stored in the pad field of the v4l2_subdev_ops structure and this field is of type const. As the v4l2_subdev_pad_ops structures are never modified, they can be declared as const. Done using Coccinelle: @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_pad_ops i@p = {...}; @ok1@ identifier r1.i; position p; struct v4l2_subdev_ops obj; @@ obj.pad=&i@p; @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_pad_ops i; File size before: text data bss dec hex filename 7633 512 20 8165 1fe5 media/i2c/m5mols/m5mols_core.o File size after: text data bss dec hex filename 7761 384 20 8165 1fe5 media/i2c/m5mols/m5mols_core.o Signed-off-by: -
Nicholas Mc Guire authored
This delay is in non-atomic context and it does not seem to be time-critical so relax it to allow the timer subsystem to optimize hrtimers. Signed-off-by:
Nicholas Mc Guire <hofrat@osadl.org> Signed-off-by:
-
Laurent Pinchart authored
Use the vb2 ioctl handler directly. Signed-off-by:
Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com> Tested-by:
Jacopo Mondi <jacopo@jmondi.org> Signed-off-by:
-
Bhumika Goyal authored
v4l2_subdev_{core/pad/video}_ops structures are stored in the fields of the v4l2_subdev_ops structure which are of type const. Also, v4l2_subdev_ops structure is passed to a function having its argument of type const. As these structures are never modified, so declare them as const. Done using Coccinelle: (One of the scripts used) @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_video_ops i@p = {...}; @ok1@ identifier r1.i; position p; struct v4l2_subdev_ops obj; @@ obj.video=&i@p; @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_video_ops i; File sizes before: text data bss dec hex filename 7810 736 16 8562 2172 drivers/media/i2c/mt9p031.o 9652 736 24 10412 28ac drivers/media/i2c/mt9v032.o 4613 552 20 5185 1441 drivers/media/i2c/noon010pc30.o 2615 552 8 3175 c67 drivers/media/i2c/s5k6a3.o File sizes after: text data bss dec hex filename 8322 232 16 8570 217a drivers/media/i2c/mt9p031.o 10164 232 24 10420 28b4 drivers/media/i2c/mt9v032.o 4933 232 20 5185 1441 drivers/media/i2c/noon010pc30.o 2935 232 8 3175 c67 drivers/media/i2c/s5k6a3.o Signed-off-by:Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
-
Bhumika Goyal authored
v4l2_subdev_{core/pad/video}_ops structures are stored in the fields of the v4l2_subdev_ops structure which are of type const. Also, v4l2_subdev_ops structure is passed to a function having its argument of type const. As these structures are never modified, so declare them as const. Done using Coccinelle: (One of the scripts used) @r1 disable optional_qualifier @ identifier i; position p; @@ static struct v4l2_subdev_video_ops i@p = {...}; @ok1@ identifier r1.i; position p; struct v4l2_subdev_ops obj; @@ obj.video=&i@p; @bad@ position p!={r1.p,ok1.p}; identifier r1.i; @@ i@p @depends on !bad disable optional_qualifier@ identifier r1.i; @@ +const struct v4l2_subdev_video_ops i; File size before: text data bss dec hex filename 6170 2752 144 9066 236a media/platform/xilinx/xilinx-tpg.o File size after: text data bss dec hex filename 6666 2384 8 9058 2362 media/platform/xilinx/xilinx-tpg.o Signed-off-by: -
Markus Elfring authored
Two local variables will be set to an appropriate value a bit later. Thus omit the explicit initialisation at the beginning. Signed-off-by:
Markus Elfring <elfring@users.sourceforge.net> Signed-off-by:
-
Markus Elfring authored
Omit extra messages for a memory allocation failure in two functions. Link: http://events.linuxfoundation.org/sites/events/files/slides/LCJ16-Refactor_Strings-WSang_0.pdfSigned-off-by:
-
Javi Merino authored
In asds configured with V4L2_ASYNC_MATCH_OF, the v4l2 subdev can be part of a devicetree overlay, for example: &media_bridge { ... my_port: port@0 { #address-cells = <1>; #size-cells = <0>; reg = <0>; ep: endpoint@0 { remote-endpoint = <&camera0>; }; }; }; / { fragment@0 { target = <&i2c0>; __overlay__ { my_cam { compatible = "foo,bar"; port { camera0: endpoint { remote-endpoint = <&my_port>; ... }; }; }; }; }; }; Each time the overlay is applied, its of_node pointer will be different. We are not interested in matching the pointer, what we want to match is that the path is the one we are expecting. Change to use of_node_cmp() so that we continue matching after the overlay has been removed and reapplied. Cc: Mauro Carvalho Chehab <mchehab@kernel.org> Cc: Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by:Javi Merino <javi.merino@kernel.org> Reviewed-by:
Javier Martinez Canillas <javier@osg.samsung.com> Reviewed-by:
-
Pan Bian authored
The value of return variable ret is 0 on some error paths, for example, when pci_resource_start() returns a NULL pointer. 0 means no error in this context, which is contrary to the fact. This patch fixes the bug. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=189011Signed-off-by:
Pan Bian <bianpan2016@163.com> Signed-off-by:
-
Pan Bian authored
Function bdisp_debugfs_create() returns 0 even on errors. So its caller cannot detect the errors. It may be better to return "-ENOMEM" on the exception paths. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=188801Signed-off-by:
-
Tiffany Lin authored
Also remove the code using size-0 OUTPUT buffer to flush. Signed-off-by:
Tiffany Lin <tiffany.lin@mediatek.com> Signed-off-by:
Wu-Cheng Li <wuchengli@chromium.org> Reviewed-by:
Kuang-che Wu <kcwu@chromium.org> Signed-off-by:
-
Rasmus Villemoes authored
gcc isn't smart enough to realize it can share most of the argument buildup and the actual function call between the two branches, so help it a little. Signed-off-by:
Rasmus Villemoes <linux@rasmusvillemoes.dk> Signed-off-by:
-
Rasmus Villemoes authored
It takes more .text to initialize these on the stack than they occupy in .rodata, so just make them static const. Signed-off-by:
-
Rasmus Villemoes authored
ibuf and rbuf in lme2510_int_response are always assigned to before they are read, and their addresses do not escape the function, so they have no reason to be static. Signed-off-by:
-
Rasmus Villemoes authored
Using the %ph printf extension for hex dumps like this makes the generated code quite a bit smaller. Signed-off-by:
-
Shuah Khan authored
Drivers might try to access and run enable_source and disable_source handlers when the driver that implements these handlers is clearing the handlers during its unregister. Fix the following race condition: process 1 process 2 request video streaming unbind au0828 v4l2 checks if tuner is free ... ... au0828_unregister_media_device() ... ... (doesn't hold graph_mutex) mdev->enable_source = NULL; if (mdev && mdev->enable_source) mdev->disable_source = NULL; mdev->enable_source() (enable_source holds graph_mutex) As shown above enable_source check is done without holding the graph_mutex. If unbind happens to be in progress, au0828 could clear enable_source and disable_source handlers leading to null pointer de-reference. Fix it by protecting enable_source and disable_source set and clear and protecting enable_source and disable_source handler access and the call itself. process 1 process 2 request video streaming unbind au0828 v4l2 checks if tuner is free ... ... au0828_unregister_media_device() ... ... (hold graph_mutex while clearing) mdev->enable_source = NULL; if (mdev) mdev->disable_source = NULL; (hold graph_mutex to check and call enable_source) if (mdev->enable_source) mdev->enable_source() If graph_mutex is held to just heck for handler being null and needs to be released before calling the handler, there will be another window for the handlers to be cleared. Hence, enable_source and disable_source handlers no longer hold the graph_mutex and expect callers to hold it to avoid forcing them release the graph_mutex before calling the handlers. Signed-off-by:Shuah Khan <shuahkh@osg.samsung.com> Signed-off-by:
-
Nicolas Iooss authored
Using sprintf() with a non-literal string makes some compiler complain when building with -Wformat-security (eg. clang reports "format string is not a string literal (potentially insecure)"). Here sprintf() format parameter is indirectly a literal string so there is no security issue. Nevertheless adding a "%s" format string to silent the warning helps to detect real bugs in the kernel. Signed-off-by:
Nicolas Iooss <nicolas.iooss_linux@m4x.org> Signed-off-by:
-
Dan Carpenter authored
The check assumes that we end on zero but actually we end on -1. Change the post-op to a pre-op so that we do end on zero. Techinically now we only loop 499 times instead of 500 but that's fine. Fixes: dc12b124 ("[media] media: ti-vpe: vpdma: Add abort channel desc and cleanup APIs") Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
Benoit Parrot <bparrot@ti.com> Signed-off-by:
-
Shailendra Verma authored
Fix to avoid possible memory leak if the decoder initialization got failed. Free the allocated memory for file handle object before return in case decoder initialization fails. Signed-off-by:
Shailendra Verma <shailendra.v@samsung.com> Signed-off-by:
-
Arnd Bergmann authored
The FE_SUPPORTED() macro is basically the same as IS_REACHABLE, except that it causes a warning with gcc-7: common/b2c2/flexcop-fe-tuner.c:30:1: error: this use of "defined" may not be portable [-Werror=expansion-to-defined] common/b2c2/flexcop-fe-tuner.c:30:1: error: this use of "defined" may not be portable [-Werror=expansion-to-defined] common/b2c2/flexcop-fe-tuner.c:30:1: error: this use of "defined" may not be portable [-Werror=expansion-to-defined] Using IS_REACHABLE() to define it avoids the warning. Fixes: 3785bc17 ("[media] b2c2: break it into common/pci/usb directories") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Mauro Carvalho Chehab authored
changeset da8ec560 ("[media] ivtv: implement new decoder command ioctls") implemented proper support for mpeg audio and video control at V4L2 API. Since then, the usage of the the DVB APIs is deprecated. However, we never actually marked it as deprecated nor provided any way to disable it. Let's do it now. This patch prepares for the removal of this bad usage on a couple of Kernel versions. Fixes: da8ec560 ("[media] ivtv: implement new decoder command ioctls") Acked-by:
-
Antti Palosaari authored
Remove unused demod driver. Device that used it never went public. If someone later decide to add support for a device using this chipset, it should be easy enough to retrieve it from git history. Signed-off-by:
Antti Palosaari <crope@iki.fi> Signed-off-by:
-
Antti Palosaari authored
I saw from ZyDAS ZD1301 sniffs it sets chip sleeping by using REG_MISC_CTRL. That has very huge effect for power management, around 0.9W. Sleep is still disabled for all the old hardware just to avoid possible regression as meaning of register bits are unknown. I tested it also with some other devices and it seems to be working, but I still consider it to be too risky to change it default. Signed-off-by:
-
Antti Palosaari authored
DVB USB interface driver for ZyDAS ZD1301 chip. Signed-off-by:
-
Antti Palosaari authored
ZyDAS ZD1301 is chip having USB interface and DVB-T demodulator integrated. This driver is for USB interface part. Device has USB ID 0ace:13a1. Used tuner is MT2060. Signed-off-by:
-
Antti Palosaari authored
DVB-T demodulator driver for ZyDAS ZD1301 chip. Signed-off-by:
-
Antti Palosaari authored
ZyDAS ZD1301 is chip having USB interface and DVB-T demodulator integrated. This driver is for demodulator part. Driver is very reduced, just basic demodulator functionality, no statistics at all. It registers as a platform driver to driver core. Signed-off-by:
-
Antti Palosaari authored
Add configuration parameter to split long i2c writes as some I2C adapters cannot write 10 bytes used as a one go. Signed-off-by:
-
Antti Palosaari authored
Add proper i2c driver model bindings. Signed-off-by:
-
- 31 Jan, 2017 6 commits
-
-
Mauro Carvalho Chehab authored
pr_foo() provides a convenient way for printk's, enforcing that they'll all prepend the error message with the driver's name. Use it inside ivtv. Signed-off-by:
-
Mauro Carvalho Chehab authored
Move the pr_fmt() macro to ivtv_driver.h and ensure that it will be the first file to be included on all ivtv files. While here, put the includes inside ivtv-driver.h on alphabetic order. Signed-off-by:
-
Mauro Carvalho Chehab authored
As reported by gcc: drivers/media/platform/sti/hva/hva-v4l2.c:227:6: warning: no previous prototype for 'hva_dbg_summary' [-Wmissing-prototypes] void hva_dbg_summary(struct hva_ctx *ctx) ^~~~~~~~~~~~~~~ This function is used only internally, so make it static. Signed-off-by: -
Jean-Christophe Trotin authored
This patch creates 4 static debugfs entries to dump: - the device-related information ("st-hva/device") - the list of registered encoders ("st-hva/encoders") - the current values of the hva registers ("st-hva/regs") - the information about the last closed instance ("st-hva/last") It also creates dynamically a debugfs entry for each opened instance, ("st-hva/<instance identifier>") to dump: - the information about the frame (format, resolution) - the information about the stream (format, profile, level, resolution) - the control parameters (bitrate mode, framerate, GOP size...) - the potential (system, encoding...) errors - the performance information about the encoding (HW processing duration, average bitrate, average framerate...) Each time a running instance is closed, its context (including the debug information) is saved to feed, on demand, the last closed instance debugfs entry. Signed-off-by:Yannick Fertre <yannick.fertre@st.com> Acked-by:
Jean-Christophe Trotin <jean-christophe.trotin@st.com> Signed-off-by:
-
Jean-Christophe Trotin authored
This patch adds a short summary about the encoding operation at each instance closing, for debug purpose (through dev_dbg()): - information about the frame (format, resolution) - information about the stream (format, profile, level, resolution) - number of encoded frames - potential (system, encoding...) errors Signed-off-by:
-
Nicholas Mc Guire authored
ulseep_range() uses hrtimers and provides no advantage over msleep() for larger delays. Fix up the 25ms delays here to use msleep() and reduce the load on the hrtimer subsystem. Link: http://lkml.org/lkml/2017/1/11/377Reviwed-by:
-
