Re: Partial fix for 687029, Different results on Alpha, part 4

This patch is out my area. Dan, please review it.

Igor.

----- Original Message ----- 
From: "Alex Cherepanov" <[email protected]>
To: "gs-code-review" <[email protected]>
Sent: Tuesday, November 25, 2003 4:20 AM
Subject: [gs-code-review] Partial fix for 687029, Different results on
Alpha, part 4


> Avoid division by 0. Don't interpolate constant functions.
>
>


----------------------------------------------------------------------------
----


> Index: gs/src/gscie.c
> ===================================================================
> RCS file: /cvs/ghostscript/gs/src/gscie.c,v
> retrieving revision 1.13
> diff -b -u -r1.13 gscie.c
> --- gs/src/gscie.c 16 Jun 2002 05:48:55 -0000 1.13
> +++ gs/src/gscie.c 24 Nov 2003 23:36:45 -0000
> @@ -589,6 +589,15 @@
>      cie_cache_range_temp_t temp[3];
>      int i, j;
>
> +    if (factor == 0) {
> +        /* Disable interpolation by setting an empty range. */
> +        for (j = 0; j < 3; ++j) {
> +      pcache->vecs.params.interpolation_ranges[j].rmin = 0;
> +     pcache->vecs.params.interpolation_ranges[j].rmax = -1;
> +        }
> +        return;
> +    }
> +
>      for (j = 0; j < 3; ++j)
>   temp[j].imin = gx_cie_cache_size, temp[j].imax = -1;
>      temp[0].prev = pcache->vecs.values[0].u;
> @@ -630,9 +639,8 @@
>
>      pcache->vecs.params.base = float2cie_cached(pcf->params.base);
>      pcache->vecs.params.factor = float2cie_cached(pcf->params.factor);
> -    pcache->vecs.params.limit =
> - float2cie_cached((gx_cie_cache_size - 1) / pcf->params.factor +
> - pcf->params.base);
> +    pcache->vecs.params.limit = float2cie_cached( (pcf->params.factor ==
0 ? 0 :
> +       (gx_cie_cache_size - 1) / pcf->params.factor) + pcf->params.base);
>      for (i = 0; i < gx_cie_cache_size; ++i) {
>   float f = pcf->values[i];
>
>


----------------------------------------------------------------------------
----


> _______________________________________________
> gs-code-review mailing list
> [email protected]
> http://www.ghostscript.com/mailman/listinfo/gs-code-review
>