[Piglit] [PATCH 3/6] glsl-es-3.00: add uniform block type mismatch link test

[Eric Anholt]

Jordan Justen <jljusten at gmail.com> writes:

> On Thu, Feb 28, 2013 at 4:41 PM, Eric Anholt <eric at anholt.net> wrote:
>> Jordan Justen <jordan.l.justen at intel.com> writes:
>>
>>> Signed-off-by: Jordan Justen <jordan.l.justen at intel.com>
>>> ---
>>>  ...terface-blocks-member-type-mismatch.shader_test |   39 ++++++++++++++++++++
>>>  1 file changed, 39 insertions(+)
>>>  create mode 100644 tests/spec/glsl-es-3.00/linker/interface-blocks-member-type-mismatch.shader_test
>>>
>>> diff --git a/tests/spec/glsl-es-3.00/linker/interface-blocks-member-type-mismatch.shader_test b/tests/spec/glsl-es-3.00/linker/interface-blocks-member-type-mismatch.shader_test
>>> new file mode 100644
>>> index 0000000..d586867
>>> --- /dev/null
>>> +++ b/tests/spec/glsl-es-3.00/linker/interface-blocks-member-type-mismatch.shader_test
>>> @@ -0,0 +1,39 @@
>>> +# Tests that a link error occurs when an interface block member's type
>>> +# differs between the vertex and fragment shaders.
>>> +#
>>> +# GLSL_ES_Specification_3.00.3, 4.3.7 Interface Blocks:
>>> +# "Matched block names within an interface (as defined above) must match
>>> +#  in terms of having the same number of declarations with the same
>>> +#  sequence of types, precisions and the same sequence of member names,
>>> +#  as well as having the same member-wise layout qualification (see next
>>> +#  section)."
>>> +[require]
>>> +GL ES >= 3.0
>>> +GLSL ES >= 3.00
>>> +
>>> +[vertex shader]
>>> +#version 300 es
>>
>> This looks like a single case of what's tested in link-mismatch-blocks.
>> I don't know whether it's important to have ES-specific version of that
>> test, but we might want to test the same things on both sides if we
>> decide to.
>
> This test fails for us currently. The differing instance name seems to
> be significant in masking the linker failure.
>
> I guess ARB_uniform_buffer_object doesn't support instance names, so
> this test can't be added to link-mismatch-blocks.

Ah, I see.  I'm still confused by the spec about what should happen with
the same block name and uniforms within differently-named interfaces,
and I'll leave that interpretation up to others :)
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 197 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/piglit/attachments/20130301/c62f27bb/attachment.pgp>