Bill Mark wrote:
>
> To the maintainer of Mesa's "span.c" file,
>
> While making some changes to Mesa for my own use, I stumbled across what
> I believe is a bug in span.c (version 3.1 of Mesa):
>
> The following code snippet appears twice:
>
> if (primitive==GL_BITMAP || (ctx->RasterMask & MULTI_DRAW_BIT)) {
> /* must make a copy of the colors since they may be modified */
> MEMCPY(rgbaBackup, rgbaIn, 4 * sizeof(GLubyte));
> rgba = rgbaBackup;
> }
>
> I believe it should be:
>
> if (primitive==GL_BITMAP || (ctx->RasterMask & MULTI_DRAW_BIT)) {
> /* must make a copy of the colors since they may be modified */
> MEMCPY(rgbaBackup, rgbaIn, 4 * n * sizeof(GLubyte));
> rgba = rgbaBackup;
> }
>
> The difference is the multiply by 'n'. Without this change, the
> MULTI_DRAW_BIT case won't work as intended. The BITMAP case might be
> OK without the change; I'm not as familiar with the BITMAP code paths
> in Mesa.
Thanks. I'm checking in the fix now.
-Brian
_______________________________________________
Mesa-bug maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-bug
_______________________________________________
Mesa-dev maillist - [EMAIL PROTECTED]
http://lists.mesa3d.org/mailman/listinfo/mesa-dev
