all devices regression report - 2017-11-15-14:14:51 - a8b1611e7613781cd5c2c7defa2c5315254d9d1d

Differences in all devices errors using gs/examples files

Previous Revision: fac14b8c25422529acded1e9e920eb20190db340
 Current Revision: a8b1611e7613781cd5c2c7defa2c5315254d9d1d

commit a8b1611e7613781cd5c2c7defa2c5315254d9d1d
Author:     Robin Watts <[email protected]>
AuthorDate: Wed Nov 15 19:17:05 2017 +0000
CommitDate: Wed Nov 15 19:30:24 2017 +0000

    Fix operation of 1bpc, 3 component halftones.
    
    In the chunky world, it makes no sense to deal with 3 bit rgb as
    this doesn't pack neatly into a byte. You'd do it by handling
    4 bit chunks.
    
    Accordingly, parts of Ghostscript are written to assume that you'd
    not do anything this crazy.
    
    Unfortunately, in the planar world, a 1 bpc, 3 component device
    does make (some) sense. Sadly, stuff like the halftone generation
    gets deeply confused by this.
    
    Here we fix 2 specific instances of this problem. In the halftone
    generation, we generate 3bit chunky halftones as 4 bit ones, and in
    the planar_copy_color routines, we assume that 3 bit data is
    actually 4 bit (as a) it probably comes from these routines, and
    b) if it doesn't it'd be packed as 4 bit in any sane world).

base/gdevmpla.c
base/gxcht.c

new errors:

errors that went away:

errors that are not whitelisted:
=== ghostpdl/examples/transparency_example.ps === cp50 ===
=== ghostpdl/examples/transparency_example.ps === psdrgb ===