Nir Radian wrote:
> Hello,
>
>
>
> I came across this part of the code in s_texfilter.c when trying to
> debug a glitch in the texture application -
>
>
>
> /* DEBUG CODE: Verify that lambda[] is monotonic.
>
> * We can't really use this because the inaccuracy in the LOG2 function
>
> * causes this test to fail, yet the resulting texturing is correct.
>
> */
>
> if (n > 1) {
>
> GLuint i;
>
> printf("lambda delta = %g\n", lambda[0] - lambda[n-1]);
>
> if (lambda[0] >= lambda[n-1]) { /* decreasing */
>
> for (i = 0; i < n - 1; i++) {
>
> ASSERT((GLint) (lambda[i] * 10) >= (GLint) (lambda[i+1] * 10));
>
> }
>
> }
>
> else { /* increasing */
>
> for (i = 0; i < n - 1; i++) {
>
> ASSERT((GLint) (lambda[i] * 10) <= (GLint) (lambda[i+1] * 10));
>
> }
>
> }
>
> }
>
>
>
> I’ve encountered a situation where lambda drops below zero during a
> non-monotonic span and the results are disastrous and not “yet the
> resulting texturing is correct“ !
lambda < 0 is nothing unusual. It indicates texture magnification.
> Has anyone seen something like this?
>
> Thank you…
>
>
>
> Btw – My name is nir, and I’m working on linux-opengl-HW compatibility
> tests, and am using MESA as reference code – I came upon this when
> trying to mipmap between bitmaps of solid color and see blue where red
> should be J If u need code examples call me up.
If you can provide a small test case I'll take a look.
-Brian
------------------------------------------------------------------------------
Open Source Business Conference (OSBC), March 24-25, 2009, San Francisco, CA
-OSBC tackles the biggest issue in open source: Open Sourcing the Enterprise
-Strategies to boost innovation and cut costs with open source participation
-Receive a $600 discount off the registration fee with the source code: SFAD
http://p.sf.net/sfu/XcvMzF8H
_______________________________________________
Mesa3d-dev mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mesa3d-dev
