On Fri, Dec 10, 2010 at 10:34:06PM +, Chris Clayton wrote:
I'm not subscribed, so please cc me on any reply.
With rc kernels from 2.6.37, X frequently (approx 3 boots out of every 4)
fails
to start. dmesg shows the oops below. I can bisect over the weekend -
probably
Sunday - if no
oops happen in bttv_open while locking uninitialized mutex fh-cap.vb_lock
add mutex_init before usage
Signed-off-by: Dave Young hidave.darks...@gmail.com
Tested-by: Chris Clayton chris2...@googlemail.com
---
drivers/media/video/bt8xx/bttv-driver.c |2 ++
1 file changed, 2 insertions(+)
---
On Sun, Dec 12, 2010 at 2:15 PM, Dave Young hidave.darks...@gmail.com wrote:
oops happen in bttv_open while locking uninitialized mutex fh-cap.vb_lock
add mutex_init before usage
I have seen the same problem twice since I switched of the BKL in
2.6.37-rc2, but only had the time today to search
We're trying to make sure that no one is writing to the btv-init struct
while we copy it over to the newly allocated fh struct. The original
code doesn't make sense because fh-cap.vb_lock hasn't been
initialized and no one else can be writing to it anyway.
Addresses:
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/cs5345.c | 87 -
1 files changed, 59 insertions(+), 28 deletions(-)
diff --git a/drivers/media/video/cs5345.c b/drivers/media/video/cs5345.c
index 9358fe7..5909f25 100644
---
This patch series converts a number of subdev drivers to use the control
framework. In addition, it also converts the cx18 and vivi drivers to the
control
framework.
This has been tested with the vivi, cx18 and em28xx drivers.
As a result of these tests bugs relating to ENUM_INPUT/OUTPUT and
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/saa7110.c | 115
1 files changed, 46 insertions(+), 69 deletions(-)
diff --git a/drivers/media/video/saa7110.c b/drivers/media/video/saa7110.c
index 7913f93..9966420 100644
---
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/bt819.c | 129 --
1 files changed, 49 insertions(+), 80 deletions(-)
diff --git a/drivers/media/video/bt819.c b/drivers/media/video/bt819.c
index c38300f..f872044 100644
---
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/ov7670.c | 296 -
1 files changed, 116 insertions(+), 180 deletions(-)
diff --git a/drivers/media/video/ov7670.c b/drivers/media/video/ov7670.c
index c881a64..9d6458f 100644
---
Implement the control framework in tvaudio.
While doing that I noticed that the handling of balance was completely
broken. It's now fixed.
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/tvaudio.c | 221 ++---
1 files changed, 75
Also fixed a memory leak in the probe function if an error occurred.
The gain control range was also fixed (a proper range from -64 to 64).
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/adv7343.c | 167 +---
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/tlv320aic23b.c | 74 +++-
1 files changed, 47 insertions(+), 27 deletions(-)
diff --git a/drivers/media/video/tlv320aic23b.c
b/drivers/media/video/tlv320aic23b.c
index dfc4dd7..286ec7e
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/tda9875.c | 192 -
1 files changed, 57 insertions(+), 135 deletions(-)
diff --git a/drivers/media/video/tda9875.c b/drivers/media/video/tda9875.c
index 35b6ff5..6d8ff21 100644
---
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/tvp514x.c | 236 ++---
1 files changed, 57 insertions(+), 179 deletions(-)
diff --git a/drivers/media/video/tvp514x.c b/drivers/media/video/tvp514x.c
index 45bcf03..9b3e828 100644
---
The 'def' field is only valid for menus, integers and booleans.
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/v4l2-ctrls.c |9 -
1 files changed, 8 insertions(+), 1 deletions(-)
diff --git a/drivers/media/video/v4l2-ctrls.c b/drivers/media/video/v4l2-ctrls.c
The v4l core sets the v4l2_input.capabilities field based on the supplied
v4l2_ioctl_ops. However, several drivers do a memset or memcpy of the v4l2_input
struct, thus overwriting that field incorrectly.
Either remove the memset (which is already done by the v4l core), or add the
proper
Calling subdevs to handle s_ctrl returned a non-zero return code even if
everything went fine.
Calling STREAMOFF if no STREAMON happened earlier would hit a BUG_ON
in res_free.
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/em28xx/em28xx-video.c | 14 +-
1
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/vpx3220.c | 137 +++--
1 files changed, 51 insertions(+), 86 deletions(-)
diff --git a/drivers/media/video/vpx3220.c b/drivers/media/video/vpx3220.c
index 91a01b3..75301d1 100644
---
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/cx18/cx18-av-audio.c | 92 ++-
drivers/media/video/cx18/cx18-av-core.c | 162 ++---
drivers/media/video/cx18/cx18-av-core.h | 12 +-
drivers/media/video/cx18/cx18-controls.c | 285
Signed-off-by: Hans Verkuil hverk...@xs4all.nl
---
drivers/media/video/tda7432.c | 277 -
1 files changed, 106 insertions(+), 171 deletions(-)
diff --git a/drivers/media/video/tda7432.c b/drivers/media/video/tda7432.c
index 3941f95..2e430a6 100644
---
* Andy Walls (awa...@md.metrocast.net) wrote:
On Sun, 2010-11-28 at 17:40 +, Dr. David Alan Gilbert wrote:
Hi,
Sparse pointed me at the following line in ivtv-fileops.c's
ivtv_v4l2_write:
ivtv_write_vbi(itv, (const struct v4l2_sliced_vbi_data
*)user_buf,
Hans,
This has at least the same two problems the change for cx25840 had:
1. Volume control init should use 65535 not 65335
2. You cannot trust reg 0x8d4 to have a value in it for the default volume that
won't give an ERANGE error when you go to init the volume control. Subdev
probe will
This message is generated daily by a cron job that builds v4l-dvb for
the kernels and architectures in the list below.
Results of the daily build of v4l-dvb:
date:Sun Dec 12 19:00:09 CET 2010
git master: 59365d136d205cc20fe666ca7f89b1c5001b0d5a
git media-master: gcc version:
On Sunday, December 12, 2010 19:09:36 Andy Walls wrote:
Hans,
This has at least the same two problems the change for cx25840 had:
1. Volume control init should use 65535 not 65335
2. You cannot trust reg 0x8d4 to have a value in it for the default volume
that won't give an ERANGE error
Hi Hans,
It looks like it should. I'm looking at things on my phone while out and
about - not the best environment for code review.
You just need to ensure default vol is = 0 and = 65535 which may not be the
case if reg 8d4 has a value near 0.
Two other topics while I'm here:
1. Why set
Hi Mauro et al,
Some bugfixes and support for a new sensor in the sonixb driver.
The following changes since commit dedb94adebe0fbdd9cafdbb170337810d8638bc9:
[media] timblogiw: Fix a merge conflict with v4l2_i2c_new_subdev_board
changes (2010-12-11 09:07:52 -0200)
are available in the git
On 12.12.2010 19:58, Dan Carpenter wrote:
We're trying to make sure that no one is writing to the btv-init struct
while we copy it over to the newly allocated fh struct. The original
code doesn't make sense because fh-cap.vb_lock hasn't been
initialized and no one else can be writing to it
Some color formats are mismatched in s5p-fimc driver.
CICICTRL[1:0], order422_out, should be set 2b'00 not 2b'11 to use
V4L2_PIX_FMT_YUYV.
Because in V4L2 standard V4L2_PIX_FMT_YUYV means start + 0: Y'00 Cb00 Y'01
Cr00 Y'02 Cb01 Y'03 Cr01.
According to datasheet 2b'00 is right value for
On Sunday, December 12, 2010 19:46:33 Andy Walls wrote:
Hi Hans,
It looks like it should. I'm looking at things on my phone while out and
about - not the best environment for code review.
You just need to ensure default vol is = 0 and = 65535 which may not be the
case if reg 8d4 has a
29 matches
Mail list logo
