- 19 Jul, 2020 40 commits
-
-
Alexander A. Klimov authored
Rationale: Reduces attack surface on kernel devs opening the links for MITM as HTTPS traffic is much harder to manipulate. Deterministic algorithm: For each file: If not .svg: For each line: If doesn't contain `\bxmlns\b`: If neither `\bgnu\.org/license`, nor `\bmozilla\.org/MPL\b`: If both the HTTP and HTTPS versions return 200 OK and serve the same content: Replace HTTP with HTTPS. Signed-off-by:Alexander A. Klimov <grandmaster@al2klimov.de> Signed-off-by:
Hans Verkuil <hverkuil-cisco@xs4all.nl> Signed-off-by:
Mauro Carvalho Chehab <mchehab+huawei@kernel.org>
-
Michael Tretter authored
Encode messages as necessary for the firmware 2019.2 and, thus, support the more recent firmware version in the driver, too. Signed-off-by:
Michael Tretter <m.tretter@pengutronix.de> Signed-off-by:
-
Michael Tretter authored
In order to distinguish the message format that is expected by the firmware, add a version field to the message header. This allows to encode and decode the messages for the version of the firmware that was loaded by the driver. Signed-off-by:
-
Michael Tretter authored
The length of the message will be determined when the message is encoded. Writing the size of the struct into the message in the driver won't be the actual length of the message that is send to the firmware. Therefore, drop the length field from the message. Since the header is the same for all response messages, it does not make sense to read the header in each decoding function, but we can simply decode it once before dispatching to the respective functions. Signed-off-by:
-
Michael Tretter authored
The firmware decides how many reference frames shall be used und returns this information via the config blob. In order to set the number of reference frames in the PPS, the driver has to read these values from the config blob after the channel has been created in the firmware. Signed-off-by:
-
Michael Tretter authored
Firmware versions >= 2019.2 do not configure the channel via the mailbox interface anymore, but use a separate chunk of memory that is only referenced by the message. As the configuration must be in a format that is understood by the firmware and this format can change between firmware versions, handle it similar to the message and treat is as a blob. In order to support both methods in the driver, always use a separate blob for the channel configuration and copy that blob into the mailbox if the firmware requires it and otherwise reference it. Signed-off-by:
-
Michael Tretter authored
Even bits in bitfields do not keep their position, but move around or move to other bitfields. Therefore, the driver has to handle bitfields differently depending on the firmware version. Create separate fields for the options that have been in bitfields and handle the bitfields when encoding the message. As a side effect, this makes the messages a bit more readable. Signed-off-by:
-
Michael Tretter authored
Like the message format, also the identifiers in the messages differ between firmware versions. This especially affects the identifier for the codec that shall be used. As the messages used by the driver are now independent from the firmware, we can use the values defined by V4L2 as identifiers in the messages. Convert the V4L2 codec format to the respective firmware value when encoding the messages to binary format instead beforehand. Signed-off-by:
-
Michael Tretter authored
The messages are now explicitly converted from the struct to the binary representation used by the firmware. Therefore, there is no need to keep the structs packed anymore. Drop the attribute and avoid confusion why the mails are packed. Signed-off-by:
-
Michael Tretter authored
When pushing the buffers for the intermediate and reference frames to the MCU, the driver relied on the message size to calculate the number of buffers. As it is not necessary anymore to keep the messages binary compatible to the firmware, we can just explicitly write the number of buffers into the message. Signed-off-by:
-
Michael Tretter authored
The message format in the mailboxes differ between firmware versions. Therefore, it is necessary to decouple the mailbox format of the driver from the message format of the firmware. This allows to keep a consistent message format in the driver while still supporting various firmware versions. Add an intermediate step to encode and decode message before writing the mails to the mailboxes. On the other hand, this allows to handle optional fields in the messages, which is required for advanced features of the encoder and was not possible until now. Signed-off-by:
-
Michael Tretter authored
Rework the functions that read and write the SRAM that is used to communicate with the MCU. As the functions will not operate on structs but on prepared binary buffers, make the buffer stride more explicit. Also, avoid any uses of struct mcu_msg_header to analyze messages in memory, because the header will be made independent of the binary representation in the mailbox. Instead explicitly access the mail size field in the mailbox. As at it, further reduce the dependency between the mailboxes and struct allegro_dev. Signed-off-by:
-
Michael Tretter authored
Add a send/notify abstraction for the mailbox and separate the message handling in the driver from the code to read and write message to the mailbox. This untangles how mails are written into the MCU's SRAM and signaled to the MCU from the protocol between the driver and the firmware. Signed-off-by:
-
Randy Dunlap authored
Delete the doubled words "the" in comments. Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Signed-off-by:
-
Randy Dunlap authored
Delete the doubled word "flag" in a comment. Signed-off-by:
-
Randy Dunlap authored
Change the doubled word "the" in a comment to "to the". Signed-off-by:
-
Randy Dunlap authored
Delete the doubled word "the" in a comment. Signed-off-by:
-
Randy Dunlap authored
Delete the doubled word "the" in a comment. Signed-off-by:
-
Krzysztof Hałasa authored
I don't know what minimal image dimensions are, but 32x32 appears to be ok according to the docs. This is needed for small sensors like 80x80 thermal imagers. Signed-off-by:
Krzysztof Halasa <khalasa@piap.pl> Reviewed-by:
Philipp Zabel <p.zabel@pengutronix.de> Tested-by:
Steve Longerbeam <slongerbeam@gmail.com> Signed-off-by:
-
Kaaira Gupta authored
Add a control in VIMC to display information such as the correct order of colors for a given test pattern, counter, brightness, hue, saturation, contrast, width and height at sensor over test image. Signed-off-by:
Kaaira Gupta <kgupta@es.iitr.ac.in> Acked-by:
Helen Koike <helen.koike@collabora.com> Reviewed-by:
Kieran Bingham <kieran.bingham@ideasonboard.com> Signed-off-by:
-
Kaaira Gupta authored
Currently there is no method to know the correct order of the colors for a test image generated by tpg. Write a function that returns a string of colors' order given a tpg. It returns a NULL pointer in case of test patterns which do not have a well defined colors' order. Hence add a NULL check for text in tpg_gen_text(). [hverkuil: white -> White (for consistency)] Signed-off-by:
-
Kaaira Gupta authored
Change the argument of type char * to const char * for function tpg_gen_text(). This function should take in a const char * as opposed to char * as it does not make changes to the text. This issue was found while passing the order of colors of tpg generated test image (which is a const char *) to this function. Signed-off-by:
-
Evgeny Novikov authored
If platform_driver_register() fails within vpss_init() resources are not cleaned up. The patch fixes this issue by introducing the corresponding error handling. Found by Linux Driver Verification project (linuxtesting.org). Signed-off-by:
Evgeny Novikov <novikov@ispras.ru> Signed-off-by:
-
Ezequiel Garcia authored
The hardware only supports 4:2:0 or 4:0:0 (monochrome), 8-bit depth content. Verify that the SPS refers to a supported bitstream, and refuse unsupported bitstreams by failing at TRY_EXT_CTRLS time. Given the JPEG compression level control is the only one that needs setting, a specific ops is provided. Signed-off-by:
Ezequiel Garcia <ezequiel@collabora.com> Reviewed-by:
Jonas Karlman <jonas@kwiboo.se> Signed-off-by:
-
Ezequiel Garcia authored
The hardware only supports 4:2:2, 4:2:0 or 4:0:0 (monochrome), 8-bit or 10-bit depth content. Verify that the SPS refers to a supported bitstream, and refuse unsupported bitstreams by failing at TRY_EXT_CTRLS time. The driver is currently broken on 10-bit and 4:2:2 so disallow those as well. Signed-off-by:
-
Ezequiel Garcia authored
Commit 986eee3a ("media: hantro: Prevent encoders from using post-processing") fixed hantro_needs_postproc condition, but missed one case. Encoders don't have any post-processor hardware block, so also can't be disabled. Fix it. Fixes: 986eee3a ("media: hantro: Prevent encoders from using post-processing") Signed-off-by:
-
Ezequiel Garcia authored
The driver doesn't need the bytesused argument. For decoders, the plane bytesused is known and therefore, buf_prepare is used to set it. For encoders, it's handled by the codec_ops.done hook. Signed-off-by:
-
Ezequiel Garcia authored
hantro_enc_buf_finish is used only for JPEG, and so should be moved to JPEG codec_ops.done. This cleanup is also taking care of addressing a subtle issue: checking the non-NULL bounce buffer using ctx->jpeg_enc, which is a member of a union is confusing and error-prone. Note that the issue is currently innocuous because an encoder context only supports JPEG. The codec_ops.done has an argument that codec-specific code shouldn't need, so drop that as well. Signed-off-by:
-
Ezequiel Garcia authored
So far we've been using the .buf_finish hook to distinguish decoder from encoder. This is unnecessarily obfuscated. Moreover, we want to move the buf_finish, so use a cleaner scheme to distinguish the driver decoder/encoder type. Signed-off-by:
-
Ezequiel Garcia authored
Commit e17f08e3 ("media: hantro: Do not reorder H264 scaling list") removed the scaling list reordering, which was wrong and not needed. However, the name of the function stayed, which is confusing for anyone reading the code. Rename from "reorder" to "assemble" which is cleaner. This is just a cosmetic cleanup. Signed-off-by:
-
Ezequiel Garcia authored
The generic H264 reference list builder moved all the users of this macro, but left the macro. Remove it. Signed-off-by:
-
Dafna Hirschfeld authored
Fix two spelling typos in file rkisp1-config.h. Signed-off-by:
Dafna Hirschfeld <dafna.hirschfeld@collabora.com> Acked-by:
-
Dafna Hirschfeld authored
The macro RKISP1_CIF_ISP_AWB_MODE_YCBCR_EN is defined twice. Remove the second define. Signed-off-by:
-
Dafna Hirschfeld authored
Currently the stats are read in a work queue. Defering the reading of the stats is not needed and it is fine to read them inside the irq handler. This patch fixes and remove the TODO item: 'Use threaded interrupt for rkisp1_stats_isr(), remove work queue.' Signed-off-by:
-
Dafna Hirschfeld authored
This patch removes two locks in the rkisp1_stats struct: - The mutex 'wq_lock' that is used to protect the buffers list 'stat' - The spin_lock 'irq_lock' that is used to protect the variable 'is_streaming' It replaces them with one spin_lock 'lock' that protects both the buffers list and the 'is_streaming' variable. In later patch the reading of the statistics will move to the isr so there will be no need for the mutex 'wq_lock' Signed-off-by:
-
Dafna Hirschfeld authored
The function 'rkisp1_stats_vb2_stop_streaming' runs in user context therefore it is enough to use spin_lock_irq Signed-off-by:
-
Dafna Hirschfeld authored
The mask of the ready statistics flags is used in several places in the code using bitwise-or. Use a macro for that to make the code more readable. Signed-off-by:
Laurent Pinchart <laurent.pinchart@ideasonboard.com> Signed-off-by:
-
Laurent Pinchart authored
Implementation of media controller centric device configuration will add code to the CAMERARX support section, which is already quite big. Move it to a separate file to make the code more manageable. No functional change is included. The cal_camerarx_init_regmap() function is kept in cal.c and renamed to cal_init_camerarx_regmap() as it's not specific to one CAMERARX instance, but related to the whole CAL device. Signed-off-by:
Benoit Parrot <bparrot@ti.com> Signed-off-by:
-
Laurent Pinchart authored
To prepare for the split of the camerarx code to a separate file, move the CAL I/O accessors to cal.h. This requires renaming the accessors with a cal_prefix, as the current names are too generic and prone to namespace clashes. The reg_read() and read_write() macros, that cover both CAL and CAMERARX register access, are split in two groups of inline functions, one for CAL access and one for CAMERARX access. Signed-off-by:
-
Laurent Pinchart authored
To prepare for implementation of media controller centric device configuration, move all the video node handling support to a separate file. Signed-off-by:
-
