Thank you for your contribution! Sashiko AI review found 2 potential issues to consider: - [Low] media: tegra-video: dangling staging path in MAINTAINERS after directory move - [Low] media: tegra-video: incomplete destaging leaves TODO file and staging hacks --
commit a8916ec654eff39d7e3b979839383852a183e987 Author: Svyatoslav Ryhel <[email protected]> staging: media: tegra-video: Move driver to media/platform/nvidia This commit moves the Tegra Video Input driver from drivers/staging/media/tegra-video/ to drivers/media/platform/nvidia/tegra-video/. > diff --git a/drivers/media/platform/nvidia/Kconfig > b/drivers/media/platform/nvidia/Kconfig > index b211b46877f63..53f34308f5aff 100644 > --- a/drivers/media/platform/nvidia/Kconfig > +++ b/drivers/media/platform/nvidia/Kconfig > @@ -3,3 +3,5 @@ > comment "NVidia media platform drivers" > > source "drivers/media/platform/nvidia/tegra-vde/Kconfig" > + > +source "drivers/media/platform/nvidia/tegra-video/Kconfig" [Severity: Low] Should MAINTAINERS also be updated to point to drivers/media/platform/nvidia/tegra-video/? The TEGRA VIDEO DRIVER entry in MAINTAINERS still lists drivers/staging/media/tegra-video/, which causes scripts/get_maintainer.pl --self-test=patterns to report a broken file pattern and prevents get_maintainer.pl from matching the moved files. [ ... ] > diff --git a/drivers/staging/media/tegra-video/TODO > b/drivers/media/platform/nvidia/tegra-video/TODO > similarity index 100% > rename from drivers/staging/media/tegra-video/TODO > rename to drivers/media/platform/nvidia/tegra-video/TODO [Severity: Low] Should the staging TODO file and remaining staging workarounds be resolved or removed before moving this driver out of staging? Renaming drivers/staging/media/tegra-video/TODO leaves an open staging TODO list (ganged mode, Tegra210 RAW10 packed format support, suspend/resume, and v4l2-compliance verification) in drivers/media/platform/nvidia/tegra-video/TODO. In addition, tegra20_vi_enable() in tegra20.c still calls ioremap(TEGRA_APB_MISC_BASE, PAGE_SIZE) on a hardcoded physical address for an unsynchronized read-modify-write on a shared APB_MISC register, and __tegra_channel_try_format() in vi.c still calls the internal __v4l2_subdev_state_alloc() helper with a FIXME noting that drivers must not use it. -- Sashiko AI review ยท https://sashiko.dev/#/patchset/[email protected]?part=1
