[Intel-gfx] [PATCH i-g-t v5 0/7] CRC testing with Chamelium improvements

[Lyude Paul]

Thank you for all of the great work you're doing! This looks perfect,
so for the whole series

Reviewed-by: Lyude <lyude at redhat.com>

I've pushed everything upstream, congrats!

On Wed, 2017-07-19 at 16:46 +0300, Paul Kocialkowski wrote:
> Changes since v4:
> * Moved igt_get_cairo_surface out of the thread function to properly
>   handle assert failure
> * Rebased on top of current master
> 
> Changes since v3:
> * Renamed structure used by async crc calculation for more clarity
> * Used const qualifier for untouched buffer when hashing
> * Split actual calculation to a dedicated function
> * Reworked async functions names for more clarity
> * Reworked descriptions for better accuracy
> * Exported framebuffer cairo surface and use it directly instead of
>   (unused) png dumping
> * Fix how the framebuffer cairo surface is obtained to avoid
> destroying
>   it too early
> 
> * Made CRC checking logic common
> * Added a check for the same number of words
> * Made frame dumping configuration and helpers common
> * Added an extended crc to string helper
> * Added a chamelium helper for crc checking and frame dumping
> * Split the merging of crc functions to a separate patch
> * Added support for frame dump path global variable
> * Added listing the dumped images in a file, possibly identified with
>   an id global variable
> 
> The latter allows having one "dump report" file per run, possibly
> identified with the id global variable, that indicates which files
> are the output, while keeping the possibility to have the same files
> for different runs. This allows saving significant disk space when
> the images are identified with e.g. their crc, so that duplicate
> results
> only lead to duplicate dump reports and not duplicate images.
>