[Mesa-dev] Implementing an HW accelerated GL_SELECT

[Ian Romanick]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 07/29/2011 06:19 AM, Micael wrote:
> Hi
> 
> There has been some work towards having GL_SELECT hardware accelerated
> accelerated on [1] and it's possible I may have some time to rewrite
> this code in order to be able to switch between the current behavior and
> the new HW accelerated behavior using a runtime command line option.
> 
> However, before starting this work, I would like to know if there is a
> chance that this will be included into master, and which eventual
> problems I may hit that will prevent it from going into master.

Selection is one of those topics that comes up every now and then, but
nobody ever quite gets excited enough to tackle it.  Thanks for jumping
on this.

I'd say fix the code formatting and post the patch to the mailing list.
 The formatting in your patch doesn't match the rest of the modified
files at all, and there's a bunch of spurious whitespace changes.

If possible, you should also break it into a series of logical patches.
 That will make it a lot easier for people to review.  It's always
easier to digest a series 5 line changes each with a two sentence
rationale than a 50 line patch with a twenty sentence rationale.

A good first message (e.g., a "[PATCH 0/n]" post) leading the series
that explains the overall approach would also be good.  As it is, the
code is fairly impenetrable and difficult to review.  That will be a
significant barrier.

We also don't have any test cases for GL_SELECT that I know of.  Before
a significant change is made to the way a large, mostly untested piece
of code works, we'll need a bunch of tests.

> [1] https://bugs.freedesktop.org/show_bug.cgi?id=34495

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAk4y6XcACgkQX1gOwKyEAw/6+gCePclQMfX8ag5Y0/3j5FPJH47C
cSEAnitT2H396OXHTwectXrABrC6qUeA
=jskE
-----END PGP SIGNATURE-----