- 18 Nov, 2015 2 commits
-
-
Javier Martinez Canillas authored
The SPI core always reports the MODALIAS uevent as "spi:<modalias>" regardless of the mechanism that was used to register the device (i.e: OF or board code) and the table that is used later to match the driver with the device (i.e: SPI id table or OF match table). So drivers needs to export the SPI id table and this be built into the module or udev won't have the necessary information to autoload the needed driver module when the device is added. But this means that OF-only drivers needs to have both OF and SPI id tables that have to be kept in sync and also the dev node compatible manufacturer prefix is stripped when reporting the MODALIAS. Which can lead to issues if two vendors use the same SPI device name for example. To avoid the above, the SPI core behavior may be changed in the future to not require an SPI device table for OF-only drivers and report the OF module alias. So, it's better to also export the OF table even when is unused now to prevent breaking module loading when the core changes. Signed-off-by:
Javier Martinez Canillas <javier@osg.samsung.com> Reviewed-by:
Andrzej Hajda <a.hajda@samsung.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@osg.samsung.com>
-
Nicholas Mc Guire authored
API conformance testing for completions with coccinelle spatches are being used to locate API usage inconsistencies: ./drivers/media/pci/ddbridge/ddbridge-core.c:89 incorrect check for negative return Return type of wait_event_timeout is signed long not int and the return type is >=0 always thus the negative check is unnecessary.. As stat is used here exclusively its type is simply changed and the negative return check dropped. Signed-off-by:Nicholas Mc Guire <hofrat@osadl.org> Signed-off-by:
-
- 17 Nov, 2015 38 commits
-
-
Julia Lawall authored
These v4l2_ctrl_ops structures are never modified, like all the other v4l2_ctrl_ops structures, so declare them as const. Done with the help of Coccinelle. Signed-off-by:
Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by:
Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by:
-
Josh Wu authored
This patch enable Atmel ISI preview path to convert the YUV to RGB format. Signed-off-by:
Josh Wu <josh.wu@atmel.com> [g.liakhovetski@gmx.de: removed superfluous parentheses] Signed-off-by:
Guennadi Liakhovetski <g.liakhovetski@gmx.de> Signed-off-by:
-
Josh Wu authored
The preview path only can convert UYVY format to RGB data. To make preview path work correctly, we need to set up YCC_SWAP according to sensor output and convert them to UYVY. Signed-off-by:
-
Josh Wu authored
Not like codec path, preview path can do downsampling, so we should setup a extra preview width, height for it. This patch add preview resolution setup without down sampling. So currently preview path will output same size as sensor output size. Signed-off-by:
-
Josh Wu authored
Atmel ISI support a preview path which can output RGB data. So this patch introduces a bool variable to choose which path is enabled currently. And also we need setup corresponding path registers. By default the preview path is disabled. We only use Codec path. Signed-off-by:
-
Josh Wu authored
we need to configure the YCC_SWAP bits in ISI_CFG2 according to current sensor output and Atmel ISI output format. Current there are two cases Atmel ISI supported: 1. Atmel ISI outputs YUYV format. This case we need to setup YCC_SWAP according to sensor output format. 2. Atmel ISI output a pass-through formats, which means no swap. Just setup YCC_SWAP as default with no swap. Signed-off-by: -
Josh Wu authored
The soc-camera host will be probed and register a v4l2_clk, but if at that moment, the i2c device is not available, then the registered v4l2_clk name is an OF string not a I2C string. So when i2c sensor probes and calls v4l2_clk_get(), it only searches a clock with I2C string, like "1-0030". This patch will search the clock with OF string name if no clock with I2C string name could be found. Signed-off-by:
-
Josh Wu authored
Make all v4l2-clk's clock name use V4L2_CLK_NAME_SIZE definition. In future, if the string is increased we just need to change the V4L2_CLK_NAME_SIZE once. Signed-off-by:
-
Josh Wu authored
This macro is used to generate an OF string for a v4l2 clock. Signed-off-by:
-
Josh Wu authored
Since v4l2_clk_name_i2c() is defined, reuse it. Signed-off-by:
-
Prashant Laddha authored
With this patch, vivid capture thread can now generate a video with frame rate reduced by a factor of 1000 / 1001. This option can be selected using a control Reduced Framerate from gui. Signed-off-by:
Prashant Laddha <prladdha@cisco.com> Signed-off-by:
-
Prashant Laddha authored
A boolean control Reduced Framerate is added to vivid controls for controlling the reduced fps option for vivid capture from gui. Signed-off-by:
-
Prashant Laddha authored
If reduced fps flag is set then check if other necessary conditions are true for the given bt timing. If yes, then reduce the frame rate. For vivid transmitter, timeperframe_vid_out controls the frame rate. Adjusting the timeperframe_vid_out by scaling down pixel clock by factor of 1000 / 1001. Signed-off-by:
-
Prashant Laddha authored
Added a helper function to check necessary conditions required for reduced fps. The reduced fps is supported for CVT and CEA861 timings. CVT supports reduced fps only if reduced blanking v2 (indicated by vsync = 8) is true. Whereas CEA861 supports reduced fps if V4L2_DV_FL_CAN_REDUCE_FPS flag is true. Signed-off-by:
-
Tommi Franttila authored
When a V4L2 sub-device backed by a DT or ACPI based device was removed, the device was unregistered as well which certainly was not intentional, as the client device would not be re-created by simply reinstating the V4L2 sub-device (indeed the device would have to be there first!). Skip unregistering the device in case it has non-NULL of_node or fwnode. Signed-off-by:
Tommi Franttila <tommi.franttila@intel.com> Signed-off-by:
Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by:
-
Ulrich Hecht authored
Initialization of the ADV7180 chip fails on the Renesas R8A7790-based Lager board about 50% of the time. This patch resolves the issue by increasing the minimum delay after reset from 2 ms to 5 ms, following the recommendation in the ADV7180 datasheet: "Executing a software reset takes approximately 2 ms. However, it is recommended to wait 5 ms before any further I2C writes are performed." Signed-off-by:
Ulrich Hecht <ulrich.hecht+renesas@gmail.com> Acked-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Acked-by:
Lars-Peter Clausen <lars@metafoo.de> Signed-off-by:
-
Oliver Neukum authored
This fixes the crash reported in: http://seclists.org/bugtraq/2015/Oct/35 The interface number needs a sanity check. Signed-off-by:
Oliver Neukum <oneukum@suse.com> Cc: Vladis Dronov <vdronov@redhat.com> Signed-off-by:
-
Arnd Bergmann authored
The sh-vou driver has been converted from videobuf to videobuf2, but the Kconfig file still lists VIDEOBUF_DMA_CONTIG as a dependency. Consequently we can build the driver without VIDEOBUF2_DMA_CONTIG and get a link error: drivers/built-in.o: In function `sh_vou_probe': vf610-ocotp.c:(.text+0x2dbf5c): undefined reference to `vb2_dma_contig_init_ctx' vf610-ocotp.c:(.text+0x2dc0b4): undefined reference to `vb2_dma_contig_cleanup_ctx' vf610-ocotp.c:(.text+0x2dc144): undefined reference to `vb2_dma_contig_memops' drivers/built-in.o: In function `sh_vou_remove': vf610-ocotp.c:(.text+0x2dc190): undefined reference to `vb2_dma_contig_cleanup_ctx' This changes the dependency to VIDEOBUF2_DMA_CONTIG instead. Fixes: 57af3ad5 ("[media] sh-vou: convert to vb2") Signed-off-by:
Arnd Bergmann <arnd@arndb.de> Signed-off-by:
-
Kosuke Tatsukawa authored
cpia2_usb_disconnect() seems to be missing a memory barrier which might cause the waker to not notice the waiter and miss sending a wake_up as in the following figure. cpia2_usb_disconnect sync ------------------------------------------------------------------------ mutex_unlock(&cam->v4l2_lock); if (waitqueue_active(&cam->wq_stream)) /* The CPU might reorder the test for the waitqueue up here, before prior writes complete */ /* wait_event_interruptible */ /* __wait_event_interruptible */ /* ___wait_event */ long __int = prepare_to_wait_event( &wq, &__wait, state); if (!cam->streaming || frame->status == FRAME_READY) cam->curbuff->status = FRAME_READY; cam->curbuff->length = 0; schedule() ------------------------------------------------------------------------ The attached patch removes the call to waitqueue_active() leaving just wake_up() behind. This fixes the problem because the call to spin_lock_irqsave() in wake_up() will be an ACQUIRE operation. I found this issue when I was looking through the linux source code for places calling waitqueue_active() before wake_up*(), but without preceding memory barriers, after sending a patch to fix a similar issue in drivers/tty/n_tty.c (Details about the original issue can be found here: https://lkml.org/lkml/2015/9/28/849). Signed-off-by:
Kosuke Tatsukawa <tatsu@ab.jp.nec.com> Signed-off-by:
-
Antonio Ospite authored
v4l2-compliance sends a zeroed struct v4l2_streamparm in v4l2-test-formats.cpp::testParmType(), and this results in a division by 0 in some gspca subdrivers: divide error: 0000 [#1] SMP Modules linked in: gspca_ov534 gspca_main ... CPU: 0 PID: 17201 Comm: v4l2-compliance Not tainted 4.3.0-rc2-ao2 #1 Hardware name: System manufacturer System Product Name/M2N-E SLI, BIOS ASUS M2N-E SLI ACPI BIOS Revision 1301 09/16/2010 task: ffff8800818306c0 ti: ffff880095c4c000 task.ti: ffff880095c4c000 RIP: 0010:[<ffffffffa079bd62>] [<ffffffffa079bd62>] sd_set_streamparm+0x12/0x60 [gspca_ov534] RSP: 0018:ffff880095c4fce8 EFLAGS: 00010296 RAX: 0000000000000000 RBX: ffff8800c9522000 RCX: ffffffffa077a140 RDX: 0000000000000000 RSI: ffff880095e0c100 RDI: ffff8800c9522000 RBP: ffff880095e0c100 R08: ffffffffa077a100 R09: 00000000000000cc R10: ffff880067ec7740 R11: 0000000000000016 R12: ffffffffa07bb400 R13: 0000000000000000 R14: ffff880081b6a800 R15: 0000000000000000 FS: 00007fda0de78740(0000) GS:ffff88012fc00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 00000000014630f8 CR3: 00000000cf349000 CR4: 00000000000006f0 Stack: ffffffffa07a6431 ffff8800c9522000 ffffffffa077656e 00000000c0cc5616 ffff8800c9522000 ffffffffa07a5e20 ffff880095e0c100 0000000000000000 ffff880067ec7740 ffffffffa077a140 ffff880067ec7740 0000000000000016 Call Trace: [<ffffffffa07a6431>] ? v4l_s_parm+0x21/0x50 [videodev] [<ffffffffa077656e>] ? vidioc_s_parm+0x4e/0x60 [gspca_main] [<ffffffffa07a5e20>] ? __video_do_ioctl+0x280/0x2f0 [videodev] [<ffffffffa07a5ba0>] ? video_ioctl2+0x20/0x20 [videodev] [<ffffffffa07a59b9>] ? video_usercopy+0x319/0x4e0 [videodev] [<ffffffff81182dc1>] ? page_add_new_anon_rmap+0x71/0xa0 [<ffffffff811afb92>] ? mem_cgroup_commit_charge+0x52/0x90 [<ffffffff81179b18>] ? handle_mm_fault+0xc18/0x1680 [<ffffffffa07a15cc>] ? v4l2_ioctl+0xac/0xd0 [videodev] [<ffffffff811c846f>] ? do_vfs_ioctl+0x28f/0x480 [<ffffffff811c86d4>] ? SyS_ioctl+0x74/0x80 [<ffffffff8154a8b6>] ? entry_SYSCALL_64_fastpath+0x16/0x75 Code: c7 93 d9 79 a0 5b 5d e9 f1 f3 9a e0 0f 1f 00 66 2e 0f 1f 84 00 00 00 00 00 66 66 66 66 90 53 31 d2 48 89 fb 48 83 ec 08 8b 46 10 <f7> 76 0c 80 bf ac 0c 00 00 00 88 87 4e 0e 00 00 74 09 80 bf 4f RIP [<ffffffffa079bd62>] sd_set_streamparm+0x12/0x60 [gspca_ov534] RSP <ffff880095c4fce8> ---[ end trace 279710c2c6c72080 ]--- Following what the doc says about a zeroed timeperframe (see http://www.linuxtv.org/downloads/v4l-dvb-apis/vidioc-g-parm.html): ... To reset manually applications can just set this field to zero. fix the issue by resetting the frame rate to a default value in case of an unusable timeperframe. The fix is done in the subdrivers instead of gspca.c because only the subdrivers have notion of a default frame rate to reset the camera to. Signed-off-by:Antonio Ospite <ao2@ao2.it> Cc: stable@vger.kernel.org Reviewed-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
-
Terry Heo authored
The current cx231xx driver doesn't work with bulk transfer mode. This patch makes it possible to use bulk transfer mode. Signed-off-by:
Terry Heo <terryheo@google.com> Signed-off-by:
-
Julia Lawall authored
The radio_tea5777_ops structure is never modified, so declare it as const. Done with the help of Coccinelle. Signed-off-by:
-
Julia Lawall authored
Move constants to the right of binary operators. Generated by: scripts/coccinelle/misc/compare_const_fl.cocci CC: Junghak Sung <jh1009.sung@samsung.com> Signed-off-by:
Fengguang Wu <fengguang.wu@intel.com> Signed-off-by:
-
Ricardo Ribalda authored
Vidioc-g-ext-ctrls can now be used to get the default value of the controls. Signed-off-by:
Ricardo Ribalda Delgado <ricardo.ribalda@gmail.com> Signed-off-by:
-
Ricardo Ribalda authored
This driver does not use the control infrastructure. Add support for the new field which on structure v4l2_ext_controls Acked-by:
Mike Isely <isely@pobox.com> Signed-off-by:
-
Ricardo Ribalda authored
This driver does not use the control infrastructure. Add support for the new field which on structure v4l2_ext_controls Signed-off-by:
-
Ricardo Ribalda authored
Support for new field which on v4l2_ext_controls, used to get the default value of one or more controls. Signed-off-by:
-
Ricardo Ribalda authored
Replace the obsolete field ctrl_class with "which". Make sure it not used in future modules by commenting out the field with ifndef __KERNEL_ . The field cannot be simply removed because that would be change on the kenel API to the userspace (and we don't like that). Signed-off-by:
-
Ricardo Ribalda authored
So it can be used to get the default value of a control. Without this change it is not possible to get the default value of array controls. Signed-off-by:
-
Hans Verkuil authored
The quantization bits are in bits 7-6, not 7-4, so shift by 6 instead of 4. This bug is caused by a typo in the adv7511 datasheet. Signed-off-by:
-
Mats Randgaard authored
hsync and hbackporch must also be compared Signed-off-by:
Mats Randgaard <matrandg@cisco.com> Signed-off-by:
-
Ricardo Ribalda authored
If modify_range is called but no range is changed, do not send the CH_RANGE event. Reported-by:
Dimitrios Katsaros <patcherwork@gmail.com> Signed-off-by:
-
Dan Carpenter authored
"len" comes from dvb_video_ioctl() and there is a possibility that it is zero. We do a divide by len later in the function so that's not ok. Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
Dan Carpenter authored
The issue here is that we there is a static checker warning because we have a user controlled volume setting and we cap the upper bound but we allow negative numbers. Negative volumes don't make sense, so let's make these variables unsigned. Signed-off-by:
-
Hans Verkuil authored
In the 3.17 kernel the poll() behavior changed for output streams: as long as not all buffers were queued up poll() would return that userspace can write. This is fine for the write() call, but when using stream I/O this changed the behavior since the expectation was that it would wait for buffers to become available for dequeuing. This patch only enables the check whether you can queue buffers for file I/O only, and skips it for stream I/O. Signed-off-by:
-
Hans Verkuil authored
If vivid is loaded with the no_error_inj=1 option, then v4l2-compliance will fail for the video and vbi output nodes because the vivid control class has no controls. Don't add the control class for video and vbi output if no_error_inj is true. Signed-off-by:
-
Hans Verkuil authored
The wrong flags field was tested for the GO7007_BOARD_HAS_AUDIO flag: that flag is in board->main_info.flags, not in board->flags. Reported-by:
-
Hans Verkuil authored
The term "input stream" is confusing since it is not clear whether this is an input stream from the point of view of the application or from the point of view of the hardware. So replace it with the more standard term "capture stream". Signed-off-by:
-
