[Mesa-dev] XCOM: Enemy Unknown vs. NaN texture unit LOD bias

[Kenneth Graunke]

On Thursday, July 13, 2017 4:23:14 PM PDT Ian Romanick wrote:
> On 07/10/2017 11:25 PM, Kenneth Graunke wrote:
> > Hello,
> > 
> > Mesa master has been hitting assert failures when running "XCOM: Enemy
> > Unknown" since commit f8d69beed49c64f883bb8ffb28d4960306baf575, where we
> > started asserting that the SAMPLER_STATE LOD Bias value actually fits in
> > the correct number of bits.
> > 
> > Apparently, XCOM calls
> > 
> >    glTexEnv(GL_TEXTURE_FILTER_CONTROL_EXT, GL_TEXTURE_LOD_BIAS_EXT, val);
> > 
> > to set the texture unit LOD bias...but according to gdb, the value is:
> > 
> >    -nan(0x7ffff3)
> > 
> > In i965, we do CLAMP(lod bias, -16, 15)...but NaN ends up failing both
> > the < min and > max comparisons, so it slips through.  But, that raises
> > the question...what value *should* we be using?  0?  Min?  Max?
> > 
> > I couldn't find any immediately applicable GL spec text.  Anyone know of
> > any?  If not, does DirectX mandate something?
> 
> There is one place.  Section 2.3.4.1 (Floating-Point Computation) of the
> OpenGL 4.5 core profile spec says:
> 
>    "The result of providing a value that is not a floating-point number
>    to such a command is unspecified, but must not lead to GL
>    interruption or termination. In IEEE arithmetic, for example,
>    providing a negative zero or a denormalized number to a GL command
>    yields predictable results, while providing a NaN or an infinity
>    yields unspecified results."
> 
> Crashing is not allowed, but nearly any other behavior is.  Based on
> that, I like Roland's suggestion of changing the CLAMP() macro.  We
> should also report this to the developers.  I'd wager that some crazy
> NaN value is not what they intended.

Hi Marc,

I think I may have found a bug in "XCOM: Enemy Within" - it appears to be
setting the texture unit LOD bias to -nan(0x7ffff3) via glTexEnv.  Passing
in a NaN here can cause unspecified results...meaning you likely get an
arbitrary LOD bias.  I'm guessing this isn't what you intended.

See above for the details.  I just sent patches to work around the crash
(assertion failure) in i965 due to the NaN popping up in unexpected places,
but we'll end up setting the LOD bias to -16, which I imagine could have a
performance impact...

https://lists.freedesktop.org/archives/mesa-dev/2017-July/162921.html

The bogus glTexEnv call happens immediately after the intro videos end,
and you reach the main menu (new game, load, etc).

Any thoughts?

--Ken
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: This is a digitally signed message part.
URL: <https://lists.freedesktop.org/archives/mesa-dev/attachments/20170713/221d69f0/attachment.sig>