On Tue, 20 Feb 2024, Radhakrishna Sripada <radhakrishna.sripada@xxxxxxxxx> wrote: > This series is originally based out of [1], and built on top of [2]. > > The primary departure from [1] was that vbt is no longer cached. During vbt > show, based on the source of vbt, it would simply be re-read reducing the > read/cleanup complexity. With this series debugfs dump of vbt should work on > all the platforms that support display. I think I'd structure this series differently. Please consider something like this: 1) Add intel_opregion_vbt_present() to check if the VBT is there. Use it in intel_bios_is_lvds_present(). 2) Always kmemdup and return an allocated buffer from intel_opregion_get_vbt(). Always kfree() it. See how that plays with the above, and helps clean up the paths and remove the oprom_vbt variable in intel_bios_init(). The extra allocation is a bummer, but otherwise the cleanup paths will be messy. 3) Add a static function in intel_bios.c to abstract the ways to get the VBT. Use it in both intel_bios_init() and intel_bios_vbt_show(). 4) Add the size return parameters. This could perhaps be done earlier. *shrug* 5) Move firmware VBT to the above function. BR, Jani. > > 1. https://patchwork.freedesktop.org/series/128341/ > 2. https://patchwork.freedesktop.org/series/128683/ > > Radhakrishna Sripada (4): > drm/i915: Pass size to oprom_get_vbt > drm/i915: Pass size to spi_oprom_get_vbt > drm/i915: Move vbt read from firmware to intel_bios.c > drm/i915: Show bios vbt when read from firmware/spi/oprom > > drivers/gpu/drm/i915/display/intel_bios.c | 104 +++++++++++++++--- > drivers/gpu/drm/i915/display/intel_opregion.c | 46 -------- > 2 files changed, 86 insertions(+), 64 deletions(-) -- Jani Nikula, Intel
