[Piglit] [PATCH 4/6] Get rid of reshape funcs

[Chad Versace]

On 08/26/2012 12:43 PM, Paul Berry wrote:
> On 25 August 2012 01:12, Chris Forbes <chrisf at ijw.co.nz
> <mailto:chrisf at ijw.co.nz>> wrote:
> 
>     Converts most of the custom reshape funcs to set the projection
>     in either piglit_init or piglit_display, depending on whether
>     it depends on the size.
> 
>     Signed-off-by: Chris Forbes <chrisf at ijw.co.nz <mailto:chrisf at ijw.co.nz>>
>     ---
>      tests/bugs/tri-tex-crash.c            | 21 ++-------------------
>      tests/general/linestipple.c           | 13 +------------
>      tests/general/occlusion_query.c       | 15 +--------------
>      tests/general/point-line-no-cull.c    | 15 ++-------------
>      tests/general/sync_api.c              | 14 +-------------
>      tests/general/texgen.c                | 17 +----------------
>      tests/general/varray-disabled.c       | 16 ++--------------
>      tests/shaders/fp-fog.c                | 21 ++-------------------
>      tests/shaders/fp-incomplete-tex.c     | 18 +-----------------
>      tests/shaders/fp-kil.c                | 20 ++------------------
>      tests/shaders/fp-lit-mask.c           | 17 +----------------
>      tests/shaders/fp-lit-src-equals-dst.c | 17 +----------------
>      tests/shaders/vpfp-generic.cpp        | 17 +----------------
>      tests/texturing/crossbar.c            | 21 +--------------------
>      tests/texturing/getteximage-formats.c | 24 +++---------------------
>      tests/texturing/getteximage-simple.c  | 23 ++++-------------------
>      tests/texturing/tex3d-maxsize.c       | 12 ++----------
>      tests/texturing/tex3d-npot.c          | 13 ++-----------
>      tests/texturing/tex3d.c               | 13 ++-----------
>      tests/texturing/texdepth.c            | 25 ++++++-------------------
>      tests/texturing/texrect-many.c        | 17 +----------------
>      tests/texturing/texredefine.c         | 13 ++-----------
>      22 files changed, 41 insertions(+), 341 deletions(-)
> 
> 
> (snip)
>  
> 
> 
>     diff --git a/tests/shaders/fp-fog.c b/tests/shaders/fp-fog.c
>     index 2351544..e977a74 100644
>     --- a/tests/shaders/fp-fog.c
>     +++ b/tests/shaders/fp-fog.c
>     @@ -60,6 +60,8 @@ piglit_display(void)
>             int pass = 1;
>             unsigned i;
> 
>     +       piglit_ortho_projection(2.0, 2.0, GL_FALSE);
>     +
> 
> 
> This will set up a projection with near=-1 and far=1, but the code you're
> deleting below uses near=-2 and far=6.  I assume you've verified that this
> change is inconsequential?  If so, it would be nice to have a short explanation
> in the commit message as to why it's inconsequential.  A similar comment applies
> to fp-incomplete-tex.c, fp-kil.c, and getteximage-simple.c.

Even if Chris has verified that changing znear and zfar is inconsequential, I
would prefer if this patch did not change their values. The patch's commit
message claims to replace glutReshape with piglit_*_projection, and the patch
should restrict itself to exactly that; it should not additionally change test
parameters.

In each of fp-incomplete-tex.c, fp-kil.c, and getteximage-simple.c, I think the
correct thing to do is to use `piglit_gen_orthoprojection(..., near=-2, far=6)`
rather than piglit_ortho_projection().