kernel.git/drivers/media/cec/core, branch linux-6.16.y

Merge tag 'media/v6.15-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media

2025-03-26T04:00:31Z

Pull media updates from Mauro Carvalho Chehab: - platform: synopsys: hdmirx: Fix 64-bit division for 32-bit targets - vim2m: print device name after registering device - Synopsys DesignWare HDMI RX Driver and various fixes - cec/printk fixes and the removal of the vidioc_g/s_ctrl and vidioc_queryctrl callbacks - AVerMedia H789-C PCIe support and rc-core structs padding - Several camera sensor patches - uvcvideo improvements - visl: Fix ERANGE error when setting enum controls - codec fixes - V4L2 camera sensor patches mostly - chips-media: wave5: Fixes - Add SDM670 camera subsystem - Qualcomm iris video decoder driver - dt-bindings: update clocks for sc7280-camss - various fixes and enhancements * tag 'media/v6.15-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-media: (264 commits) media: pci: mgb4: include linux/errno.h media: synopsys: hdmirx: Fix signedness bug in hdmirx_parse_dt() media: platform: synopsys: hdmirx: Fix 64-bit division for 32-bit targets media: vim2m: print device name after registering device media: vivid: Introduce VIDEO_VIVID_OSD media: vivid: Move all fb_info references into vivid-osd media: platform: synopsys: hdmirx: Optimize struct snps_hdmirx_dev media: platform: synopsys: hdmirx: Remove unused HDMI audio CODEC relics media: platform: synopsys: hdmirx: Remove duplicated header inclusion media: qcom: Clean up Kconfig dependencies media: dvb-frontends: tda10048: Make the range of z explicit. media: platform: stm32: Add check for clk_enable() media: xilinx-tpg: fix double put in xtpg_parse_of() media: siano: Fix error handling in smsdvb_module_init() media: c8sectpfe: Call of_node_put(i2c_bus) only once in c8sectpfe_probe() media: i2c: tda1997x: Call of_node_put(ep) only once in tda1997x_parse_dt() dt-bindings: media: mediatek,vcodec: Revise description dt-bindings: media: mediatek,jpeg: Relax IOMMU max item count media: v4l2-dv-timings: prevent possible overflow in v4l2_detect_gtf() media: rockchip: rga: fix rga offset lookup ...

media: cec: use us_to_ktime() where appropriate

2025-03-06T16:28:25Z

[Why] There are several ns_to_ktime() calls that require using nanoseconds. It is better to replace them with us_to_ktime() to make code clear, getting rid of multiplication by 1000. Also the timer function code may have an integer wrap-around issue. Since both tx_custom_low_usecs and tx_custom_high_usecs can be set to up to 9999999 from the user space via cec_pin_error_inj_parse_line(), this may cause usecs to be overflowed when adap->monitor_pin_cnt is zero and usecs is multiplied by 1000. [How] Take advantage of using an appropriate helper func us_to_ktime() instead of ns_to_ktime() to improve readability and to make the code clearer. And this also mitigates possible integer wrap-arounds when usecs value is too large and it is multiplied by 1000. Found by Linux Verification Center (linuxtesting.org) with Svace. Signed-off-by: Vitaliy Shevtsov Signed-off-by: Hans Verkuil

media: cec: core: allow raw msg transmit while configuring

2025-03-05T09:43:23Z

While the CEC adapter is configuring, it is not possible to transmit a CEC message with the CEC_MSG_FL_RAW flag set as this is blocked at the ioctl level. Check if this flag is set, and if so, allow the message to be transmitted. This is useful for debugging if the display has no physical address (typically because the HPD is pulled low while in Standby). Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: Switch to use hrtimer_setup()

2025-02-18T10:19:04Z

hrtimer_setup() takes the callback function pointer as argument and initializes the timer completely. Replace hrtimer_init() and the open coded initialization of hrtimer::function with the new setup mechanism. Patch was created by using Coccinelle. Signed-off-by: Nam Cao Signed-off-by: Thomas Gleixner Acked-by: Hans Verkuil Acked-by: Zack Rusin Link: https://lore.kernel.org/all/c725ad8d0ecac3cf6bbc532af567e56d47a6b75c.1738746904.git.namcao@linutronix.de

media: cec: include linux/debugfs.h and linux/seq_file.h where needed

2025-01-07T11:21:38Z

Having cec.h include linux/debugfs.h leads to all users of all cec headers include and depend on debugfs.h and its dependencies for no reason. Drop the include from cec.h, and include debugfs.h and seq_file.h where needed. Sort all the modified include lists while at it. Signed-off-by: Jani Nikula Signed-off-by: Hans Verkuil

media: replace obsolete hans.verkuil@cisco.com alias

2024-11-08T12:38:09Z

The old hans.verkuil@cisco.com email address was discontinued years ago. Replace it with the correct hansverk@cisco.com email. Signed-off-by: Hans Verkuil

[tree-wide] finally take no_llseek out

2024-09-27T15:18:43Z

no_llseek had been defined to NULL two years ago, in commit 868941b14441 ("fs: remove no_llseek") To quote that commit, At -rc1 we'll need do a mechanical removal of no_llseek - git grep -l -w no_llseek | grep -v porting.rst | while read i; do sed -i '/\/d' $i done would do it. Unfortunately, that hadn't been done. Linus, could you do that now, so that we could finally put that thing to rest? All instances are of the form .llseek = no_llseek, so it's obviously safe. Signed-off-by: Al Viro Signed-off-by: Linus Torvalds

media: cec: move cec_get/put_device to header

2024-09-05T18:12:15Z

Move cec_get/put_device to the media/cec.h header. This allows CEC drivers to use this. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: cec: cec-adap.c: improve CEC_MSG_FL_REPLY_VENDOR_ID check

2024-08-08T13:24:03Z

The new CEC_MSG_FL_REPLY_VENDOR_ID flag only makes sense in combination with CEC_MSG_VENDOR_COMMAND_WITH_ID. So rather than reporting an error if that flag is set with another command, just clear the flag instead. Only keep the message length check, since otherwise the flag would not make sense. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab

media: cec: core: add new CEC_MSG_FL_REPLY_VENDOR_ID flag

2024-08-05T07:38:46Z

If this flag is set, then the reply is expected to consist of the CEC_MSG_VENDOR_COMMAND_WITH_ID opcode followed by the Vendor ID (as used in bytes 1-4 of the message), followed by the struct cec_msg reply field. Note that this assumes that the byte after the Vendor ID is a vendor-specific opcode. This flag makes it easier to wait for replies to vendor commands, using the same CEC framework support for waiting for regular replies. Support for this flag is indicated by setting the new CEC_CAP_REPLY_VENDOR_ID capability. Signed-off-by: Hans Verkuil Signed-off-by: Mauro Carvalho Chehab