[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

Tapani Pälli  changed:

   What|Removed |Added

 Status|ASSIGNED|RESOLVED
 Resolution|--- |NOTOURBUG

--- Comment #8 from Tapani Pälli  ---
This is now fixed in opengl-es-cts-3.2.4 by following commit, resolving this
one as NOTOURBUG.

--- 8< ---
commit c693c85a5f9983caab94c3973e6dc0efaae57b0c
Author: Tapani Pälli 
Date:   Thu Apr 5 08:13:46 2018 +0300

Prefer KHR entrypoints instead of EXT for robustness tests

When resolving function entrypoints, framework resolves EXT
entrypoints after KHR to the same pointers. There are drivers that
implement only KHR entrypoints, prefer KHR over EXT so that KHR
entrypoints will be the ones used if both extensions are supported
by the driver.

Components: OpenGL ES
VK-GL-CTS issue: 1107

Affects:
KHR-NoContext.es32.robustness.getnuniform
KHR-NoContext.es32.robustness.readnpixels

Change-Id: Iec5f7cbdd53061e105b3445f7613ee41fccc4553
Signed-off-by: Tapani Pälli 

-- 
You are receiving this mail because:
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #7 from Tapani Pälli  ---
OK, sent a fix to vk-gl-cts suite to reorder EXT and KHR resolving so that KHR
is preferred over EXT. It might be this is not the only case where such
reordering is required and maybe more elegant fix to generator scripts is
required.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #6 from Tapani Pälli  ---
OK .. after yet another gdb session I think now I got it. What is happening is
that when the test suite generates functions from khronos xml, it first sets
gl.readnPixels to glReadnPixelsKHR but later on replaces this same pointer with
glReadnPixelsEXT (which Mesa does not support). This is a bit unfortunate,
maybe it should check if the function was already succesfully resolved. Will
attempt this.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #5 from Tapani Pälli  ---
Updating status ... my fix to dEQP went in and now we pass 'getnuniform' test,
but readnpixels still fails.

I did a separate GLES 3.2 app that utilizes these entrypoints, they are being
exposed and work correctly. Also dEQP has tests that call these entrypoints and
those tests are passing. So, as said in comment #4 there seems to be something
going wrong with how the api functions are resolved by the test suite.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #4 from Tapani Pälli  ---
(In reply to Kenneth Graunke from comment #0)
> The readnpixels test does:
> 
>PFNGLREADNPIXELS pReadnPixels =
> (PFNGLREADNPIXELS)context->getRenderContext().
> getProcAddress("glReadnPixels");
> 
> and then calls that function pointer...but ends up in generic_nop().

It does not actually call it, instead it calls api via "gl.readnPixels", gl
contains the API functions that have been resolved already earlier. The
pReadnPixels would actually work if it would be utilized, I changed test to use
that and it passes.

I'm wanted to write this down here so that I won't forget this. Will need to
figure out how context functions in 'gl.*' get resolved.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

Tapani Pälli  changed:

   What|Removed |Added

 Status|NEW |ASSIGNED
   Assignee|mesa-dev@lists.freedesktop. |lem...@gmail.com
   |org |

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #3 from Tapani Pälli  ---
(In reply to Tapani Pälli from comment #2)
> (In reply to Tapani Pälli from comment #1)
> > It seems compiler/linker should realize that they are written to shared
> > values so we should not optimize those away.
> 
> and alternative theory: since we never write these out to any buffer maybe
> it is just ok to optimize these away and test should be changed instead :)

I've created issue about this and sent a fix proposal for getnuniform test:
https://gitlab.khronos.org/Tracker/vk-gl-cts/issues/1036

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #2 from Tapani Pälli  ---
(In reply to Tapani Pälli from comment #1)
> It seems compiler/linker should realize that they are written to shared
> values so we should not optimize those away.

and alternative theory: since we never write these out to any buffer maybe it
is just ok to optimize these away and test should be changed instead :)

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

--- Comment #1 from Tapani Pälli  ---
It looks like these 2 tests fail for different reasons:

- 'readnpix' hits generic_nop (glapi issue)

- 'getnuniform', problem seems to be that according to Mesa these uniforms are
inactive and therefore no uniformstorage is available. I can see that linker
reserves explicit locations correctly (inputf : 11, inputi : 12, inputu : 13)
but opt_dead_code removes the variables:

--- 8< ---
Removed declaration of inputf@0x34f1f80
Removed declaration of inputu@0x34f26b0
Removed declaration of inputi@0x34f2540
--- 8< ---

If I modify opt_dead_code to not remove those variables the test passes. I
hacked it like this:

--- 8< ---
+if (entry->var->data.explicit_location)
+   continue;
--- 8< ---

It seems compiler/linker should realize that they are written to shared values
so we should not optimize those away.

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev

[Mesa-dev] [Bug 105105] Suffixless KHR_robustness functions aren't exposed in ES 3.2

[bugzilla-daemon]

https://bugs.freedesktop.org/show_bug.cgi?id=105105

Bug ID: 105105
   Summary: Suffixless KHR_robustness functions aren't exposed in
ES 3.2
   Product: Mesa
   Version: git
  Hardware: Other
OS: All
Status: NEW
  Severity: major
  Priority: medium
 Component: Mesa core
  Assignee: mesa-dev@lists.freedesktop.org
  Reporter: kenn...@whitecape.org
QA Contact: mesa-dev@lists.freedesktop.org
CC: i...@freedesktop.org

These two tests fail on i965:

   KHR-NoContext.es32.robustness.readnpixels
   KHR-NoContext.es32.robustness.getnuniform

The readnpixels test does:

   PFNGLREADNPIXELS pReadnPixels =
(PFNGLREADNPIXELS)context->getRenderContext().getProcAddress("glReadnPixels");

and then calls that function pointer...but ends up in generic_nop().

It looks like the suffixless versions aren't getting exposed correctly, despite
being part of ES 3.2.

There's a fair bit of complex aliasing going on for ARB/KHR/suffixless, and I'm
getting a bit lost trying to figure out the right mapi/glapi/gen XML snippets
to do the right thing.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are the assignee for the bug.___
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev