Re: [PATCH v3 00/35] Qualcomm Camera Subsystem driver - 8x96 support
From: Todor Tomov
Date: Tue Jul 24 2018 - 08:47:41 EST
On 24.07.2018 14:41, Hans Verkuil wrote:
> On 23/07/18 13:02, Todor Tomov wrote:
>> Changelog v3:
>> - split patch 08 to device tree binding patch and driver patch and
>> improve commit message.
>>
>> --------------------------------------------------------------------------------
>>
>> This patchset adds support for the Qualcomm Camera Subsystem found
>> on Qualcomm MSM8996 and APQ8096 SoC to the existing driver which
>> used to support MSM8916 and APQ8016.
>>
>> The camera subsystem hardware on 8x96 is similar to 8x16 but
>> supports more cameras and features. More details are added in the
>> driver document by the last patch.
>>
>> The first 3 patches are dependencies which have already been on
>> the mainling list but I'm adding them here for completeness.
>>
>> The following 12 patches add general updates and fixes to the driver.
>> Then the rest add the support for the new hardware.
>>
>> The driver is tested on Dragonboard 410c (APQ8016) and Dragonboard 820c
>> (APQ8096) with OV5645 camera sensors. media-ctl [1], yavta [2] and
>> GStreamer were used for testing.
>>
>> [1] https://git.linuxtv.org//v4l-utils.git
>> [2] http://git.ideasonboard.org/yavta.git
>>
>> --------------------------------------------------------------------------------
>
> When running sparse I get these warnings:
>
> camss/camss-csiphy-2ph-1-0.c:170:28: warning: symbol 'csiphy_ops_2ph_1_0' was not declared. Should it be static?
> camss/camss-csiphy-3ph-1-0.c:250:28: warning: symbol 'csiphy_ops_3ph_1_0' was not declared. Should it be static?
> camss/camss-vfe-4-1.c:976:25: warning: symbol 'vfe_ops_4_1' was not declared. Should it be static?
> camss/camss-vfe-4-7.c:1098:25: warning: symbol 'vfe_ops_4_7' was not declared. Should it be static?
>
> checkpatch.pl complains about this as well.
>
> There are externs in vfe.c, but those should probably be moved to camss-vfe.h.
Yes, this is correct. I can fix this in an additional small patch.
>
> It should solves both checkpatch and sparse.
>
> Regards,
>
> Hans
>
--
Best regards,
Todor Tomov