[PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_relase does not held the lock although it is modifying the queue-owner field. This could lead to race conditions on the vb2_perform_io function when multiple applications are accessing the video device via read/write API: [ 308.297741]

Re: [PATCH v4] videobuf2: Add missing lock held on vb2_fop_relase

Hello Hans and Sywester I have just posted a new patch. I think it fits the suggestions from both of you, please take a look to it and please post any comment. I will also send a patch about the em28xx, to swap the lock order. Thanks for your comments On Mon, Nov 4, 2013 at 4:19 PM, Hans

[PATCH] em28xx-video: Swap release order to avoid lock nesting

vb2_fop_release might take the video queue mutex lock. In order to avoid nesting mutexes the private mutex is taken after the fop_release has finished. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- drivers/media/usb/em28xx/em28xx-video.c | 2 +- 1 file changed, 1

Re: [PATCH v4] videobuf2: Add missing lock held on vb2_fop_relase

Here are the patches https://patchwork.linuxtv.org/patch/20668/ https://patchwork.linuxtv.org/patch/20669/ Thanks! On Wed, Nov 6, 2013 at 9:26 AM, Ricardo Ribalda Delgado ricardo.riba...@gmail.com wrote: Hello Hans and Sywester I have just posted a new patch. I think it fits the

Re: [PATCH] em28xx-video: Swap release order to avoid lock nesting

On 11/06/13 09:39, Ricardo Ribalda Delgado wrote: vb2_fop_release might take the video queue mutex lock. In order to avoid nesting mutexes the private mutex is taken after the fop_release has finished. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com ---

Re: [PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

On 11/06/13 09:24, Ricardo Ribalda Delgado wrote: From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_relase does not held the lock although it is modifying the Small typo: _relase - _release queue-owner field. This could lead to race conditions on the vb2_perform_io function when

Re: [PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

Hello Hans I will fix both things, but I will wait a couple of hours in case there is more changes, to avoid spamming the list. Thanks On Wed, Nov 6, 2013 at 10:07 AM, Hans Verkuil hverk...@xs4all.nl wrote: On 11/06/13 09:24, Ricardo Ribalda Delgado wrote: From: Ricardo Ribalda

Re: [PATCH 2/6] [media] mt9p031: Include linux/of.h header

On 6 November 2013 06:58, Laurent Pinchart laurent.pinch...@ideasonboard.com wrote: Hi Sachin, Thank you for the patch, and sorry for the late reply. On Friday 18 October 2013 08:37:11 Sachin Kamat wrote: 'of_match_ptr' is defined in linux/of.h. Include it explicitly to avoid build breakage

Re: [PATCH v11 03/12] [media] exynos5-fimc-is: Add common driver header files

Hi, On 05/11/13 14:16, Arun Kumar K wrote: +struct is_common_reg { + u32 hicmd; + u32 hic_sensorid; + u32 hic_param[4]; + + u32 reserved1[3]; [...] + u32 meta_iflag; + u32 meta_sensor_id; + u32 meta_param1; + + u32 reserved9[1]; + + u32 fcount;

Re: [PATCH v11 03/12] [media] exynos5-fimc-is: Add common driver header files

Hi Sylwester, On Wed, Nov 6, 2013 at 4:53 PM, Sylwester Nawrocki s.nawro...@samsung.com wrote: Hi, On 05/11/13 14:16, Arun Kumar K wrote: +struct is_common_reg { + u32 hicmd; + u32 hic_sensorid; + u32 hic_param[4]; + + u32 reserved1[3]; [...] + u32 meta_iflag; +

[PATCH] videodev2: Set vb2_rect's width and height as unsigned

As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- drivers/media/i2c/mt9m032.c| 16

Re: [PATCH v11 03/12] [media] exynos5-fimc-is: Add common driver header files

Hi Sylwester and Arun, On Wed, Nov 06, 2013 at 12:23:07PM +0100, Sylwester Nawrocki wrote: Hi, On 05/11/13 14:16, Arun Kumar K wrote: +struct is_common_reg { + u32 hicmd; + u32 hic_sensorid; + u32 hic_param[4]; + + u32 reserved1[3]; [...] + u32 meta_iflag;

Re: [PATCH] Full DVB driver package for Earthsoft PT3 (ISDB-S/T) cards

On Tue, Nov 5, 2013 at 5:30 PM, ほち knightri...@are.ma wrote: Michael Krufky mkrufky at linuxtv.org writes: As the DVB maintainer, I am telling you that I won't merge this as a monolithic driver. The standard is to separate the driver into modules where possible, unless there is a valid

Re: [PATCH] videodev2: Set vb2_rect's width and height as unsigned

Hi Ricardo, Thanks for the patch. On Wed, Nov 06, 2013 at 01:49:25PM +0100, Ricardo Ribalda Delgado wrote: ... diff --git a/drivers/media/i2c/smiapp/smiapp-core.c b/drivers/media/i2c/smiapp/smiapp-core.c index ae66d91..9036e64 100644 --- a/drivers/media/i2c/smiapp/smiapp-core.c +++

[PATCH] smiapp: Fix BUG_ON() on an impossible condition

internal_csi_format_idx and csi_format_idx are unsigned integers, therefore they can never be nevative. CC drivers/media/i2c/smiapp/smiapp-core.o In file included from include/linux/err.h:4:0, from include/linux/clk.h:15, from

[PATCH] ths7303: Declare as static a private function

git grep shows that the function is only called from ths7303.c Fix this build warning: CC drivers/media/i2c/ths7303.o drivers/media/i2c/ths7303.c:86:5: warning: no previous prototype for ‘ths7303_setval’ [-Wmissing-prototypes] int ths7303_setval(struct v4l2_subdev *sd, enum

Re: [PATCH] ths7303: Declare as static a private function

Hi Ricardo, Thank you for the patch. On Wednesday 06 November 2013 15:27:48 Ricardo Ribalda Delgado wrote: git grep shows that the function is only called from ths7303.c Fix this build warning: CC drivers/media/i2c/ths7303.o drivers/media/i2c/ths7303.c:86:5: warning: no previous

[PATCH v6] videobuf2: Add missing lock held on vb2_fop_relase

From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_release does not held the lock although it is modifying the queue-owner field. This could lead to race conditions on the vb2_perform_io function when multiple applications are accessing the video device via read/write API: [ 308.297741]

[PATCH v2] ths7303: Declare as static a private function

git grep shows that the function is only called from ths7303.c Fix this build warning: CC drivers/media/i2c/ths7303.o drivers/media/i2c/ths7303.c:86:5: warning: no previous prototype for ‘ths7303_setval’ [-Wmissing-prototypes] int ths7303_setval(struct v4l2_subdev *sd, enum

Re: [PATCH] ths7303: Declare as static a private function

Hello Laurent I have just uploaded a patch, please take a look if I align it properly. Thanks! On Wed, Nov 6, 2013 at 3:33 PM, Laurent Pinchart laurent.pinch...@ideasonboard.com wrote: Hi Ricardo, Thank you for the patch. On Wednesday 06 November 2013 15:27:48 Ricardo Ribalda Delgado

Re: [PATCH] videodev2: Set vb2_rect's width and height as unsigned

Hello Sakari. Thanks for your comments. I have fixed them, but I will wait some time to get more comments, to avoid spamming the list. Thanks! On Wed, Nov 6, 2013 at 3:19 PM, Sakari Ailus sakari.ai...@iki.fi wrote: Hi Ricardo, Thanks for the patch. On Wed, Nov 06, 2013 at 01:49:25PM +0100,

Re: [PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

Hello, (dropping some unrelated e-mail addresses from Cc) On 06/11/13 10:07, Hans Verkuil wrote: On 11/06/13 09:24, Ricardo Ribalda Delgado wrote: From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_relase does not held the lock although it is modifying the Small typo: _relase -

[GIT PULL v2] Exynos5 SoC FIMC-IS imaging subsystem driver

Hi Mauro, This change set adds a V4L2 driver for the Exynos5 SoC series camera subsystem. It also includes a minimal v4l2 subdev driver for s5k4e5 raw image sensor. Comparing to the original pull request, the DT binding documentation is on separate patches and some DT binding maintainer acks were

Re: [PATCH] videodev2: Set vb2_rect's width and height as unsigned

Hi Ricardo, Thank you for the patch. On Wednesday 06 November 2013 13:49:25 Ricardo Ribalda Delgado wrote: As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned.

Re: [PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

On 11/06/13 15:46, Sylwester Nawrocki wrote: Hello, (dropping some unrelated e-mail addresses from Cc) On 06/11/13 10:07, Hans Verkuil wrote: On 11/06/13 09:24, Ricardo Ribalda Delgado wrote: From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_relase does not held the lock although

Re: [PATCH v2] ths7303: Declare as static a private function

Hi Ricardo, Thank you for the patch. On Wednesday 06 November 2013 15:40:18 Ricardo Ribalda Delgado wrote: git grep shows that the function is only called from ths7303.c Fix this build warning: CC drivers/media/i2c/ths7303.o drivers/media/i2c/ths7303.c:86:5: warning: no previous

Re: [PATCH] smiapp: Fix BUG_ON() on an impossible condition

Hi Ricardo, Thanks for the patch. I've removed LKML from cc since I don't think this is anything but noise there. On Wed, Nov 06, 2013 at 03:21:30PM +0100, Ricardo Ribalda Delgado wrote: internal_csi_format_idx and csi_format_idx are unsigned integers, therefore they can never be nevative.

Re: [PATCH] smiapp: Fix BUG_ON() on an impossible condition

Hello Sakari I always try to send the patches to the mails found by get-maintainer.pl, I thought it was unpolite not doint so :). In the future I will try to send the mails only to the media list. I don't have a tree, so please do as you wish. Thanks! On Wed, Nov 6, 2013 at 3:58 PM, Sakari

Re: [PATCH v5] videobuf2: Add missing lock held on vb2_fop_relase

On 06/11/13 15:51, Hans Verkuil wrote: Rename __vb2_fop_release to vb2_fop_release_unlock and rearrange arguments I know I suggested the vb2_fop_release_unlock name, but on second thoughts that's not a good name. I suggest vb2_fop_release_no_lock instead. '_unlock' suggests that

[PATCH v2] videodev2: Set vb2_rect's width and height as unsigned

As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- v2: Comments by Sakari Ailus and Laurent Pinchart -Fix

[PATCH v7] videobuf2: Add missing lock held on vb2_fop_relase

From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_release does not held the lock although it is modifying the queue-owner field. This could lead to race conditions on the vb2_perform_io function when multiple applications are accessing the video device via read/write API: [ 308.297741]

[PATCH v2] videodev2: Set vb2_rect's width and height as unsigned

As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- v2: Comments by Sakari Ailus and Laurent Pinchart -Fix

Re: [PATCH] smiapp: Fix BUG_ON() on an impossible condition

Hi Ricardo, On Wed, Nov 06, 2013 at 04:03:37PM +0100, Ricardo Ribalda Delgado wrote: Hello Sakari I always try to send the patches to the mails found by get-maintainer.pl, I thought it was unpolite not doint so :). In the future I will try to send the mails only to the media list. Yeah, it

Re: [PATCH v2] videodev2: Set vb2_rect's width and height as unsigned

On Wed, Nov 06, 2013 at 04:43:04PM +0100, Ricardo Ribalda Delgado wrote: As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. Signed-off-by: Ricardo Ribalda Delgado

Re: [PATCH v2] videodev2: Set vb2_rect's width and height as unsigned

Hello Sakai It has to be done in the same patch? or on a separated patch just changing the xml file? Thanks! On Wed, Nov 6, 2013 at 4:53 PM, Sakari Ailus sakari.ai...@iki.fi wrote: On Wed, Nov 06, 2013 at 04:43:04PM +0100, Ricardo Ribalda Delgado wrote: As addressed on the media summit 2013,

Re: [PATCH v7] videobuf2: Add missing lock held on vb2_fop_relase

Hi, On 06/11/13 16:41, Ricardo Ribalda Delgado wrote: From: Ricardo Ribalda ricardo.riba...@gmail.com vb2_fop_release does not held the lock although it is modifying the queue-owner field. This could lead to race conditions on the vb2_perform_io function when multiple applications are

Re: [PATCH v2] videodev2: Set vb2_rect's width and height as unsigned

Hi Ricardo, On Wed, Nov 06, 2013 at 04:56:12PM +0100, Ricardo Ribalda Delgado wrote: It has to be done in the same patch? or on a separated patch just changing the xml file? Good question. Now that you ask, I also realise the documentation must also be changed --- struct v4l2_rect is

Re: [PATCH] Full DVB driver package for Earthsoft PT3 (ISDB-S/T) cards

On 06.11.2013 15:14, Michael Krufky wrote: On Tue, Nov 5, 2013 at 5:30 PM, ほち knightri...@are.ma wrote: Michael Krufky mkrufky at linuxtv.org writes: As the DVB maintainer, I am telling you that I won't merge this as a monolithic driver. The standard is to separate the driver into modules

re: [media] coda: update CODA7541 to firmware 1.4.50

Hello Philipp Zabel, This is a semi-automatic email about new static checker warnings. The patch 5677e3b04d3b: [media] coda: update CODA7541 to firmware 1.4.50 from Jun 21, 2013, leads to the following Smatch complaint: drivers/media/platform/coda.c:1530 coda_alloc_framebuffers()

[PATCH v3] videodev2: Set vb2_rect's width and height as unsigned

As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. v3: Comments by Sakari -Update also doc v2: Comments by Sakari Ailus and Laurent Pinchart -Fix alignment on all drivers

[PATCH v4] videodev2: Set vb2_rect's width and height as unsigned

As addressed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields into unsigned. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- v3: Comments by Sakari -Update also doc v2: Comments by

Re: [PATCH v4] videodev2: Set vb2_rect's width and height as unsigned

Hi Ricardo, Thanks for the update. (Dropping LKML again.) On Wed, Nov 06, 2013 at 05:44:38PM +0100, Ricardo Ribalda Delgado wrote: As addressed on the media summit 2013, there is no reason for the width s/addressed/discussed/ ? and height to be signed. Therefore this patch is an attempt

Re: [PATCH] v4l2-fh: Include linux/videodev2.h for enum v4l2_priority definition

On Thu, Oct 03, 2013 at 11:59:30PM +0200, Laurent Pinchart wrote: struct v4l2_fh has an enum v4l2_priority field. Make sure the enum definition is available by including linux/videodev2.h. Signed-off-by: Laurent Pinchart laurent.pinch...@ideasonboard.com Acked-by: Sakari Ailus

Re: [PATCH] v4l2-fh: Include linux/fs.h for struct file definition

On Thu, Oct 03, 2013 at 11:55:52PM +0200, Laurent Pinchart wrote: v4l2-fh.h dereferences struct file, the structure must thus be defined. Pull in its definition by including linux/fs.h. Signed-off-by: Laurent Pinchart laurent.pinch...@ideasonboard.com Acked-by: Sakari Ailus

[PATCH 7/8] MAINTAINERS: add M88DS3103

It is Montage M88DS3103 DVB-S/S2 demodulator driver. Signed-off-by: Antti Palosaari cr...@iki.fi --- MAINTAINERS | 10 ++ 1 file changed, 10 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index e61c2e8..85cfc6f 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5244,6 +5244,16 @@ W:

[PATCH 8/8] MAINTAINERS: add M88TS2022

It is Montage M88TS2022 DVB-S/S2 silicon tuner driver. Signed-off-by: Antti Palosaari cr...@iki.fi --- MAINTAINERS | 10 ++ 1 file changed, 10 insertions(+) diff --git a/MAINTAINERS b/MAINTAINERS index 85cfc6f..314b485 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -5262,6 +5262,16 @@ Q:

[PATCH 4/8] Montage M88TS2022 silicon tuner driver

--- drivers/media/tuners/Kconfig | 7 + drivers/media/tuners/Makefile | 1 + drivers/media/tuners/m88ts2022.c | 664 ++ drivers/media/tuners/m88ts2022.h | 72 drivers/media/tuners/m88ts2022_priv.h | 38 ++ 5 files changed, 782

[PATCH 6/8] m88ds3103: add parent for I2C adapter

Every I2C adapter should have a parent. Signed-off-by: Antti Palosaari cr...@iki.fi --- drivers/media/dvb-frontends/m88ds3103.c | 1 + 1 file changed, 1 insertion(+) diff --git a/drivers/media/dvb-frontends/m88ds3103.c b/drivers/media/dvb-frontends/m88ds3103.c index 91b3729..fe4a67e 100644 ---

[PATCH 1/8] em28xx: add support for Empia EM28178

Signed-off-by: Antti Palosaari cr...@iki.fi --- drivers/media/usb/em28xx/em28xx-cards.c | 5 + drivers/media/usb/em28xx/em28xx-core.c | 9 ++--- drivers/media/usb/em28xx/em28xx-input.c | 2 ++ drivers/media/usb/em28xx/em28xx-reg.h | 1 + 4 files changed, 14 insertions(+), 3

[PATCH 3/8] Montage M88DS3103 DVB-S/S2 demodulator driver

--- drivers/media/dvb-frontends/Kconfig |7 + drivers/media/dvb-frontends/Makefile |1 + drivers/media/dvb-frontends/m88ds3103.c | 1293 ++ drivers/media/dvb-frontends/m88ds3103.h | 108 +++ drivers/media/dvb-frontends/m88ds3103_priv.h |

[PATCH 0/8] PCTV DVB-S2 Stick (461e) [2013:0258] driver

This patch serie adds support for PCTV Systems latest DVB-S/S2 USB stick, model numbered as 461e. I found only German version of product page [1], maybe English is coming soon as this is quite young device. Device looks 100% similar than older PCTV DVB-S2 Stick 460e, but internally demodulator

[PATCH 5/8] em28xx: add support for PCTV DVB-S2 Stick (461e) [2013:0258]

Device has following chips: Empia EM28178, Montage M88DS3103, Montage M88TS2022, Allegro A8293. --- drivers/media/usb/em28xx/Kconfig| 2 ++ drivers/media/usb/em28xx/em28xx-cards.c | 35 +++ drivers/media/usb/em28xx/em28xx-dvb.c | 49 +

[PATCH 2/8] a8293: add small sleep in order to settle LNB voltage

Signed-off-by: Antti Palosaari cr...@iki.fi --- drivers/media/dvb-frontends/a8293.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/media/dvb-frontends/a8293.c b/drivers/media/dvb-frontends/a8293.c index 74fbb5d..780da58 100644 --- a/drivers/media/dvb-frontends/a8293.c +++

Re: em28xx: new board id [0213:0258]

Sorry for top posting, but I just posted driver for that stick! Look latest LMML messages. regards Antti On 07.10.2013 23:34, Rodolphe M wrote: Hello Got this DVB-S2 USB stick from PCTV Model 461e. Firmware for 460e does not work. Chipset must be different. Here are data from Windows

[PATCH v5] videodev2: Set vb2_rect's width and height as unsigned

As discussed on the media summit 2013, there is no reason for the width and height to be signed. Therefore this patch is an attempt to convert those fields from __s32 to __u32. Acked-by: Sakari Ailus sakari.ai...@iki.fi (documentation and smiapp) Signed-off-by: Ricardo Ribalda Delgado

[PATCH] videobuf2-dma-sg: Fix typo on debug message

num_pages_from_user and buf-num_pages were swapped. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com --- drivers/media/v4l2-core/videobuf2-dma-sg.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/media/v4l2-core/videobuf2-dma-sg.c

[PATCH] af9035: add [0413:6a05] Leadtek WinFast DTV Dongle Dual

It is IT9135 dual design. Thanks to Michael Piko for reporting that! Reported-by: Michael Piko mich...@piko.com.au Signed-off-by: Antti Palosaari cr...@iki.fi --- drivers/media/usb/dvb-usb-v2/af9035.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/drivers/media/usb/dvb-usb-v2/af9035.c

[PATCH] Leadtek WinFast DTV Dongle Dual [0413:6a05]

Adds new USB ID for Leadtek WinFast DTV Dongle Dual [0413:6a05] and nothing more. I have not send that patch earlier as there was no test reports. Now there seems to be some reports on Wiki: http://linuxtv.org/wiki/index.php/LinuxTVWiki:Sandbox

[GIT PULL] af9035: add [0413:6a05] Leadtek WinFast DTV Dongle Dual

The following changes since commit 80f93c7b0f4599ffbdac8d964ecd1162b8b618b9: [media] media: st-rc: Add ST remote control driver (2013-10-31 08:20:08 -0200) are available in the git repository at: git://linuxtv.org/anttip/media_tree.git af9035_winfast_id for you to fetch changes up to

[PATCH] videobuf2-dma-sg: Support io userptr operations on io memory

Memory exported via remap_pfn_range cannot be remapped via get_user_pages. Other videobuf2 methods (like the dma-contig) supports io memory. This patch adds support for this kind of memory. Signed-off-by: Ricardo Ribalda Delgado ricardo.riba...@gmail.com ---

cron job: media_tree daily build: WARNINGS

This message is generated daily by a cron job that builds media_tree for the kernels and architectures in the list below. Results of the daily build of media_tree: date: Thu Nov 7 04:00:19 CET 2013 git branch: for-v3.13c git hash: 3adeac2c34cc28e05d0ec52f38f009dcce278555 gcc

Re: [REVIEW PATCH 6/9] si4713 : Added the USB driver for Si4713

On 11/05/2013 03:18 PM, edubez...@gmail.com wrote: Dinesh On Tue, Oct 15, 2013 at 11:24 AM, Dinesh Ram dinesh@cern.ch wrote: This is the USB driver for the Silicon Labs development board. It contains the Si4713 FM transmitter chip. I tried this driver again. The system attempts to