- 15 Apr, 2024 25 commits
-
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings: WARNING:PRINTK_WITHOUT_KERN_LEVEL: printk() should include KERN_<LEVEL> facility level WARNING:PREFER_PR_LEVEL: Prefer [subsystem eg: netdev]_dbg([subsystem]dev, ... then dev_dbg(dev, ... then pr_debug(... to printk(KERN_DEBUG ... WARNING:EMBEDDED_FUNCTION_NAME: Prefer using '"%s...", __func__' to using 'ciintf_slot_ts_enable', this function's name, in a string CHECK:MACRO_ARG_PRECEDENCE: Macro argument 'level' may be better as '(level)' to avoid precedence issues There was no consistent style of logging functions used so far. Convert all logging to the appropriate pr_* macros. Add a global pr_fmt macro for the whole driver. Signed-off-by:
Stefan Herdler <herdler@nurfuerspam.de> Signed-off-by:
Hans Verkuil <hverkuil-cisco@xs4all.nl>
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings: WARNING:RETURN_VOID: void function return statements are not generally useful WARNING:TRAILING_SEMICOLON: macros should not use a trailing semicolon Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings: WARNING:CONSTANT_COMPARISON: Comparisons should place the constant on the right side of the test Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch errors: ERROR:TRAILING_STATEMENTS: trailing statements should be on next line Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch errors: ERROR:ASSIGN_IN_IF: do not use assignment in if condition Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings: WARNING:EXPORT_SYMBOL: EXPORT_SYMBOL(foo); should immediately follow its function/variable Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings and errors: WARNING:BRACES: braces {} are not necessary for any arm of this statement WARNING:BRACES: braces {} are not necessary for single statement blocks ERROR:OPEN_BRACE: that open brace { should be on the previous line CHECK:BRACES: braces {} should be used on all arms of this statement CHECK:BRACES: Unbalanced braces around else statement Signed-off-by: -
Stefan Herdler authored
This patch fixes the following checkpatch warnings: WARNING:BLOCK_COMMENT_STYLE: Block comments use * on subsequent lines WARNING:BLOCK_COMMENT_STYLE: Block comments use a trailing */ on a separate line WARNING: It's generally not useful to have the filename in the file Signed-off-by:
-
Stefan Herdler authored
It fixes the following checkpatch errors: ERROR:SPACING: space prohibited before that ',' (ctx:WxW) ERROR:SPACING: space required after that ',' (ctx:VxV) ERROR:SWITCH_CASE_INDENT_LEVEL: switch and case should be at the same indent This patch contains the large multi line blocks. Signed-off-by:
-
Stefan Herdler authored
This patch fixes the following checkpatch warnings and errors: ERROR:POINTER_LOCATION: "foo * bar" should be "foo *bar" ERROR:POINTER_LOCATION: "foo* bar" should be "foo *bar" ERROR:SPACING: space prohibited after that open parenthesis '(' ERROR:SPACING: space prohibited before that close parenthesis ')' ERROR:SPACING: space prohibited before that ',' (ctx:WxW) ERROR:SPACING: space required after that ',' (ctx:VxV) ERROR:SPACING: space required before that '-' (ctx:OxV) ERROR:SPACING: space required before the open parenthesis '(' ERROR:SPACING: spaces required around that '!=' (ctx:VxO) ERROR:SPACING: spaces required around that '<' (ctx:VxV) ERROR:SPACING: spaces required around that '=' (ctx:VxV) ERROR:SPACING: spaces required around that '>=' (ctx:VxV) WARNING:SPACING: space prohibited between function name and open parenthesis '(' WARNING:TABSTOP: Statements should start on a tabstop WARNING:LEADING_SPACE: please, no spaces at the start of a line WARNING:SUSPECT_CODE_INDENT: suspect code indent for conditional statements WARNING:LINE_SPACING: Missing a blank line after declarations Signed-off-by: -
Ricardo Ribalda authored
i2c_device_probe(), seems to assume that irq = 0 means that there is no irq to request. The driver also believes that on the clean path. So lets be consistent here. Also make smatch happy. Fix: drivers/media/i2c/adv7180.c:1526 adv7180_probe() warn: 'client->irq' from request_threaded_irq() not released on lines: 1526 Signed-off-by:
Ricardo Ribalda <ribalda@chromium.org> Signed-off-by:
-
Ricardo Ribalda authored
Simplify a bit the code introducing a new variable for iterating through the control list. It also makes smatch happy: drivers/media/v4l2-core/v4l2-ctrls-api.c:1091 v4l2_query_ext_ctrl() warn: iterator used outside loop: 'ref' Signed-off-by:
-
Ricardo Ribalda authored
hva->clk cannot be NULL at this point. Simplify the code and make smatch happy: drivers/media/platform/st/sti/hva/hva-hw.c:412 hva_hw_probe() warn: 'hva->clk' from clk_prepare() not released on lines: 412 Signed-off-by:
-
Ricardo Ribalda authored
Fix smatch error: drivers/media/spi/cxd2880-spi.c:391 cxd2880_start_feed() warn: Please consider using kzalloc instead of kmalloc Signed-off-by:
-
Ricardo Ribalda authored
Add missing release_firmware on the error paths. drivers/media/usb/dvb-usb/dib0700_devices.c:2415 stk9090m_frontend_attach() warn: 'state->frontend_firmware' from request_firmware() not released on lines: 2415. drivers/media/usb/dvb-usb/dib0700_devices.c:2497 nim9090md_frontend_attach() warn: 'state->frontend_firmware' from request_firmware() not released on lines: 2489,2497. Signed-off-by:
-
Ricardo Ribalda authored
Do not cache the file, instead load it on demand. This makes smatch a happy parser: drivers/media/dvb-frontends/drx39xyj/drxj.c:11908 drx_ctrl_u_code() warn: 'fw' from request_firmware() not released on lines: 11877,11886,11896. Signed-off-by:
-
Ricardo Ribalda authored
Make sure the firmware is released when we leave xc_load_fw_and_init_tuner() This change makes smatch happy: drivers/media/tuners/xc5000.c:1213 xc_load_fw_and_init_tuner() warn: 'fw' from request_firmware() not released on lines: 1213. Cc: Shuah Khan <shuah.kh@samsung.com> Signed-off-by:
-
Ricardo Ribalda authored
release_firmware() in the same function that it was requested. It is more clear and makes smatch happy. drivers/media/platform/st/sti/c8sectpfe/c8sectpfe-core.c:1146 load_c8sectpfe_fw() warn: 'fw' from request_firmware() not released on lines: 1125,1132. Signed-off-by:
-
Ricardo Ribalda authored
The handled version of request_irq let us remove the free_irq and makes smatch happier: drivers/media/pci/saa7134/saa7134-alsa.c:1186 alsa_card_saa7134_create() warn: 'dev->pci->irq' from request_irq() not released on lines: 1186 Signed-off-by:
-
Ricardo Ribalda authored
Let devm handle the life cycle of the irq request. Makes smatch happier: drivers/media/pci/solo6x10/solo6x10-core.c:631 solo_pci_probe() warn: 'pdev' from pci_request_regions() not released on lines: 631. Signed-off-by:
-
Ricardo Ribalda authored
Instead of handling manually the release of the memory regions let devm do that for us. Makes smatch happy: drivers/media/pci/solo6x10/solo6x10-core.c:631 solo_pci_probe() warn: 'pdev' from pci_request_regions() not released on lines: 631. Signed-off-by:
-
Ricardo Ribalda authored
The structure is packed, which requires that all its fields need to be also packed. ./include/uapi/linux/videodev2.h:1810:2: warning: field within 'struct v4l2_ext_control' is less aligned than 'union v4l2_ext_control::(anonymous at ./include/uapi/linux/videodev2.h:1810:2)' and is usually due to 'struct v4l2_ext_control' being packed, which can lead to unaligned accesses [-Wunaligned-access] Explicitly set the inner union as packed. Marking the inner union as 'packed' does not change the layout, since the whole struct is already packed, it just silences the clang warning. See also this llvm discussion: https://github.com/llvm/llvm-project/issues/55520Signed-off-by:
-
Ricardo Ribalda authored
The structure is packed, which requires that all its fields need to be also packed. ./include/uapi/linux/dvb/frontend.h:854:2: warning: field within 'struct dtv_stats' is less aligned than 'union dtv_stats::(anonymous at ./include/uapi/linux/dvb/frontend.h:854:2)' and is usually due to 'struct dtv_stats' being packed, which can lead to unaligned accesses [-Wunaligned-access] Explicitly set the inner union as packed. Marking the inner union as 'packed' does not change the layout, since the whole struct is already packed, it just silences the clang warning. See also this llvm discussion: https://github.com/llvm/llvm-project/issues/55520Signed-off-by:
-
Ricardo Ribalda authored
This structure is embedded in multiple other structures that are packed, which conflicts with it being aligned. drivers/media/usb/as102/as10x_cmd.h:379:30: warning: field reg_addr within 'struct as10x_dump_memory::(unnamed at drivers/media/usb/as102/as10x_cmd.h:373:2)' is less aligned than 'struct as10x_register_addr' and is usually due to 'struct as10x_dump_memory::(unnamed at drivers/media/usb/as102/as10x_cmd.h:373:2)' being packed, which can lead to unaligned accesses [-Wunaligned-access] Mark it as being packed. Marking the inner struct as 'packed' does not change the layout, since the whole struct is already packed, it just silences the clang warning. See also this llvm discussion: https://github.com/llvm/llvm-project/issues/55520Signed-off-by:
-
Ricardo Ribalda authored
Make dependency on DEBUG_FS conditional, that way we are not forced to enable DEBUG_FS if we can to use this driver. Signed-off-by:
-
- 11 Apr, 2024 11 commits
-
-
Hans de Goede authored
The atomisp driver's Android heritage makes it weird in that even though it uses MC + subdev-s it is designed to primarily be controlled through its /dev/video# node. It implements s_input() on /dev/video# to select which sensor to use, while ignoring link_setup() calls to enable a link to another sensor. Add support for selecting the active sensor the MC way by adding link_setup() support. The implementation is a bit convoluted due to the atomisp driver's heritage. Reviewed-by:
Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by:
Hans de Goede <hdegoede@redhat.com> Signed-off-by:
Mauro Carvalho Chehab <mchehab@kernel.org>
-
Hans de Goede authored
The link from the ISP's source pad to the /dev/video# output v4l2-dev is always enabled and immutable, mark it as such. Reviewed-by:
Kieran Bingham <kieran.bingham@ideasonboard.com> Reviewed-by:
-
Hans de Goede authored
Change the generic "ATOMISP_SUBDEV" name to "ATOM ISP" to make clear that this is the subdev for the ISP itself. Reviewed-by:
-
Hans de Goede authored
Add DMABUF support and while at it drop userptr support. Now that atomisp has been ported to videobuf2 this is trivial. Reviewed-by:
Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Reviewed-by:
-
Hans de Goede authored
Stop setting sd->devnode for the ATOMISP_SUBDEV v4l2-subdev, so that a proper /dev/v4l-subdev# gets created for it. Reviewed-by:
-
Hans de Goede authored
Drop all ISP controls and ioctls which just get passed through to the camera subdev. Instead these calls should be done directly at the sensor subdev. Reviewed-by:
-
Hans de Goede authored
Remove the custom VCM handling, instead the VCM should be controlled through its own v4l2-subdev and the new ipu-bridge.c code already supports instantiating an i2c_client for this and setting up the necessary endpoints in the fwnode graph. Reviewed-by:
-
Hans de Goede authored
Rename atomisp_set_crop_and_fmt() to atomisp_set_sensor_crop_and_fmt() to make clear that it operates on the sensor subdev. Reviewed-by:
-
Hans de Goede authored
isp_subdev_propagate() is a wrapper around atomisp_subdev_set_selection() which gets only used in a single place. Call atomisp_subdev_set_selection() directly in that single place instead. Reviewed-by:
-
Zhipeng Lu authored
The allocation failure of mycs->yuv_scaler_binary in load_video_binaries() is followed with a dereference of mycs->yuv_scaler_binary after the following call chain: sh_css_pipe_load_binaries() |-> load_video_binaries(mycs->yuv_scaler_binary == NULL) | |-> sh_css_pipe_unload_binaries() |-> unload_video_binaries() In unload_video_binaries(), it calls to ia_css_binary_unload with argument &pipe->pipe_settings.video.yuv_scaler_binary[i], which refers to the same memory slot as mycs->yuv_scaler_binary. Thus, a null-pointer dereference is triggered. Link: https://lore.kernel.org/r/20240118151303.3828292-1-alexious@zju.edu.cn Fixes: a49d2536 ("staging/atomisp: Add support for the Intel IPU v2") Signed-off-by:Zhipeng Lu <alexious@zju.edu.cn> Reviewed-by:
-
Hans de Goede authored
s/is not unsupported/is not supported/ in the unsupported revision error. Suggested-by:
-
- 10 Apr, 2024 4 commits
-
-
Laurent Pinchart authored
When support for streams was added to the V4L2 subdev API, the v4l2_subdev_crop structure was extended with a stream field, but the field was not handled in the core code that translates the VIDIOC_SUBDEV_[GS]_CROP ioctls to the selection API. Fix it. Fixes: 2f91e10e ("media: subdev: add stream based configuration") Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Reviewed-by:
Tomi Valkeinen <tomi.valkeinen@ideasonboard.com> Signed-off-by:
Sakari Ailus <sakari.ailus@linux.intel.com> Signed-off-by:
-
Tomi Valkeinen authored
The graph walk tries to follow all links, even if they are not between pads. This causes a crash with, e.g. a MEDIA_LNK_FL_ANCILLARY_LINK link. Fix this by allowing the walk to proceed only for MEDIA_LNK_FL_DATA_LINK links. Signed-off-by:
-
Sakari Ailus authored
Remove e-mail addresses no longer in use. Signed-off-by:
-
Sakari Ailus authored
Remove inactive e-mail addresses. Signed-off-by:
-
