- 15 Aug, 2012 40 commits
-
-
Albert Wang authored
In JPEG mode, the size of image is variable due to different JPEG compression rate. We only can get the pix->sizeimage from the user. If we clear pix->sizeimage in soc_camera_try_fmt() then we will get it from: ret = soc_mbus_image_size(xlate->host_fmt, pix->bytesperline, pix->height); if (ret < 0) return ret; pix->sizeimage = max_t(u32, pix->sizeimage, ret); In general, this sizeimage will be larger than the actul JPEG image size. But vb2 will check the buffer and size of image in __qbuf_userptr(): /* Check if the provided plane buffer is large enough */ if (planes[plane].length < q->plane_sizes[plane]) So we shouldn't clear the pix->sizeimage and also shouldn't re-calculate the pix->sizeimage in soc_mbus_image_size() in JPEG mode We also shouldn't re-calculate pix->bytesperline: ret = soc_mbus_bytes_per_line(pix->width, xlate->host_fmt); if (ret < 0) return ret; pix->bytesperline = max_t(u32, pix->bytesperline, ret); pix->bytesperline also should be set by the user or by the driver's try_fmt() implementation. Change-Id: I700690a2287346127a624b5260922eaa5427a596 Signed-off-by:
Albert Wang <twang13@marvell.com> Signed-off-by:
Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by:
Mauro Carvalho Chehab <mchehab@redhat.com>
-
Liu Ying authored
This patch contains code change only to use the present macro- MX3_CAMERA_DATAWIDTH_MASK to check valid camera platform data bus width instead of enumerating every possible data bus width. Signed-off-by:
Liu Ying <Ying.Liu@freescale.com> Signed-off-by:
-
Alex Gershgorin authored
Driver for mt9v022 camera sensor is fully compatible for mt9v024 camera sensor with the exception of several registers which have been changed addresses. mt9v024 also has improved and additional features, but they are currently not in use. Signed-off-by:
Alex Gershgorin <alexg@meprolight.com> Signed-off-by:
-
Javier Martin authored
On i.MX27 two clocks are required: emma-ipg and emma-ahb. The ahb clock has to be requested using both a device and a connection ID. Signed-off-by:
Javier Martin <javier.martin@vista-silicon.com> [g.liakhovetski@gmx.de: rebase to the current media tree] Signed-off-by:
-
Alex Gershgorin authored
This patch checks the state of the buffer when calling .buf_init() method. This is needed for the USERPTR buffer type, because in that case .buf_init() is called every time a buffer is queued, and not only once during the preparation stage, like in the MMAP case. Without this check buffers get initialised repeatedly, which also leads to the allocation of new DMA descriptors, of which there is only a final relatively small number available. Both MMAP and USERPTR methods were successfully tested. Signed-off-by:
-
Antti Palosaari authored
All these seems to survive .reset_resume. Signed-off-by:
Antti Palosaari <crope@iki.fi> Signed-off-by:
-
Antti Palosaari authored
All these seems to survive .reset_resume. Signed-off-by:
-
Antti Palosaari authored
Add .reset_resume() support. Also some other small changes for suspend / resume. Signed-off-by:
-
Antti Palosaari authored
Move initial suspend / resume support from dvb_usb_v2 to dvb_frontend as it is dvb general feature that could be used all dvb devices. Signed-off-by:
-
Antti Palosaari authored
Put device full sleep on suspend, wake-up it on resume and acquire retune in order to return same television channel. Signed-off-by:
-
Antti Palosaari authored
We need to retune when resume from suspend. Signed-off-by:
-
Emil Goode authored
The commit 98d7bbb9 changed *of_device_id.data to const which introduced warnings in various places that have mostly been fixed. This patch fixes one such warning by introducing two const qualifiers. GCC warning: drivers/media/platform/coda.c:1785:16: warning: assignment discards ‘const’ qualifier from pointer target type [enabled by default] Signed-off-by:Emil Goode <emilgoode@gmail.com> Signed-off-by:
-
Antti Palosaari authored
Signed-off-by:
-
Antti Palosaari authored
debug_dump macro was defined twice when CONFIG_DVB_USB_DEBUG was not set. Move debug_dump macro to correct place. Reported-by:
Randy Dunlap <rdunlap@xenotime.net> Signed-off-by:
-
Julia Lawall authored
[media] drivers/media/usb/{s2255drv.c, tm6000/tm6000-alsa.c, tm6000/tm6000-input.c}: Remove potential NULL dereferences If the NULL test is necessary, the initialization involving a dereference of the tested value should be moved after the NULL test. The sematic patch that fixes this problem is as follows: (http://coccinelle.lip6.fr/) // <smpl> @@ type T; expression E; identifier i,fld; statement S; @@ - T i = E->fld; + T i; ... when != E when != i if (E == NULL) S + i = E->fld; // </smpl> Signed-off-by:Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by:
-
Sachin Kamat authored
devm_* functions are device managed functions and make error handling and cleanup cleaner and simpler. Signed-off-by:
Sachin Kamat <sachin.kamat@linaro.org> Signed-off-by:
Sylwester Nawrocki <s.nawrocki@samsung.com> Signed-off-by:
-
Sachin Kamat authored
Replace printk with pr_* functions to silence checkpatch warnings. Signed-off-by:
-
Sachin Kamat authored
devm_regulator_get() is a device managed function and makes the exit code a bit simpler and cleaner. Signed-off-by:
-
Julia Lawall authored
Using devm_kzalloc simplifies the code and ensures that the use of devm_request_irq is safe. When kzalloc and kfree were used, the interrupt could be triggered after the handler's data argument had been freed. This also introduces some missing initializations of the return variable ret, and uses devm_request_and_ioremap instead of the combination of devm_request_mem_region and devm_ioremap. The problem of a free after a devm_request_irq was found using the following semantic match (http://coccinelle.lip6.fr/) // <smpl> @r exists@ expression e1,e2,x,a,b,c,d; identifier free; position p1,p2; @@ devm_request_irq@p1(e1,e2,...,x) ... when any when != e2 = a when != x = b if (...) { ... when != e2 = c when != x = d free@p2(...,x,...); ... return ...; } // </smpl> Signed-off-by:
-
Hans Verkuil authored
More validation fixes as reported by xmllint. There are still three xmllint errors after this remaining regarding SVG file support. Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
-
Hans Verkuil authored
Signed-off-by:
-
Dan Carpenter authored
"card" is a valid pointer here because we checked snd_card_create() for error returns. Checking after a dereference makes the static checkers complain. Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Acked-by:
Ezequiel Garcia <elezegarcia@gmail.com> Signed-off-by:
-
Dan Carpenter authored
There are some unlocks missing on error. Signed-off-by:
-
Dan Carpenter authored
We recently added locking to this function, but there was an error path which accidentally returned holding a lock. Also we returned zero on failure on some paths instead of the error code. Signed-off-by:
-
Dan Carpenter authored
We need to move the unlock before the kfree(dev); Signed-off-by:
-
Dan Carpenter authored
This code looks suspicious, but it turns out that "nv" is an array of u8 so sizeof() is the same as ARRAY_SIZE(). Using ARRAY_SIZE() is more readable though. Signed-off-by:
-
Guennadi Liakhovetski authored
Add .get_selection() and .set_selection() soc-camera host driver operations. Additionally check, that the user is not trying to change the output sizes during a running capture. Signed-off-by:
-
Javier Martin authored
Add explicit conversions from UYVY and YUYV to YUYV so that csicr1 configuration can be set properly for each format. Signed-off-by:
-
Fabio Estevam authored
Prepare the clock before enabling it. Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de> Cc: <linux-media@vger.kernel.org> Signed-off-by:
Fabio Estevam <fabio.estevam@freescale.com> Signed-off-by:
-
Fabio Estevam authored
Prepare the clock before enabling it. Cc: Guennadi Liakhovetski <g.liakhovetski@gmx.de> Cc: <linux-media@vger.kernel.org> Signed-off-by:
-
Laurent Pinchart authored
The sensor starts streaming video as soon as it gets powered or is reset. Disable the output in the reset function. Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
-
Laurent Pinchart authored
Instead of hardcoding register arrays, compute the values at runtime. Signed-off-by:
-
Laurent Pinchart authored
Signed-off-by:
-
Laurent Pinchart authored
And use them instead of calling SMBus access functions directly. Signed-off-by:
-
Laurent Pinchart authored
Conversion from i2c_client to ov772x_priv is only needed in a single location, while conversion from v4l2_subdev to ov772x_priv is needed in several locations. Perform the former manually, and use to_ov772x for the later. Signed-off-by:
-
Laurent Pinchart authored
If the requested format isn't supported, return a fixed default format instead of the current format. Signed-off-by:
-
Laurent Pinchart authored
Select a default format instead. Signed-off-by:
-
Laurent Pinchart authored
The format and window size are only initialized during the first g_fmt call. This leaves the device in an inconsistent state after initialization, which will cause problems when implementing pad operations. Move the format and window size initialization to probe time. Signed-off-by:
-
Laurent Pinchart authored
The control handler isn't freed if its initialization fails. Fix it. Signed-off-by:
-
Laurent Pinchart authored
Several client drivers access the hardware at probe time, for instance to read the probe chip ID. Such chips need to be powered up when being probed. soc-camera handles this by powering chips up in the soc-camera probe implementation. However, this will break with non soc-camera hosts that don't perform the same operations. Fix the problem by pushing the power up/down from the soc-camera core down to individual drivers on a needs basis. Signed-off-by:
-
