Re: ghostscript issues

[Lemures Lemniscati <lemures.lemniscati@gmail.com>]

Date: Wed, 3 Jun 2020 07:56:25 -0400
From: Ken Brown via Cygwin
> On 6/2/2020 2:14 PM, Ken Brown via Cygwin wrote:
> > On 6/2/2020 8:12 AM, Ken Brown via Cygwin wrote:
> >> On 6/2/2020 7:58 AM, Lemures Lemniscati via Cygwin wrote:
> >>> Date: Mon, 1 Jun 2020 12:43:44 -0400
> >>> From: Ken Brown via Cygwin
> >>>
> >>>> On 6/1/2020 11:26 AM, Thomas Bodine via Cygwin wrote:
> >>>>> My favorite use of ghost script is to combine PDF's. I upgraded Cygwin this
> >>>>> weekend, and ghost script fails to perform this task
> >>>>>
> >>>>> I am running this:
> >>>>> Cygwin Setup version 2.904 (64 bit)
> >>>>> base-cygwin 3.8-1
> >>>>> ghostscript 9.52-1
> >>>>> ghostscript-fonts-other 6.0-1
> >>>>> Windows 10 Pro version 1909 build 18363.836
> >>>>>
> >>>>> When I run the script:
> >>>>>    $ gs -dNOPAUSE -dBATCH -sDEVICE=pdfwrite  -sOutputFile=combine.pdf
> >>>>> able.pdf baker.pdf charly.pdf
> >>>>>>
> >>>>> GPL Ghostscript 9.52 (2020-03-19)
> >>>>> Copyright (C) 2020 Artifex Software, Inc.  All rights reserved.
> >>>>> This software is supplied under the GNU AGPLv3 and comes with NO WARRANTY:
> >>>>> see the file COPYING for details.
> >>>>> Processing pages 1 through 1.
> >>>>> Page 1
> >>>>> Error: /rangecheck in --restore--
> >>>>> Operand stack:
> >>>>>      --dict:8/16(L)--   --dict:38/38(ro)(L)--
> >>>>> Execution stack:
> >>>>>      %interp_exit   .runexec2   --nostringval--   restore  >>>>> --nostringval--   2
> >>>>> %stopped_push   --nostringval--   restore   restore   false   1  >>>>> %stopped_push
> >>>>>     1990   1   3   %oparray_pop   1989   1   3   %oparray_pop   1977   1
> >>>>> 3   %op
> >>>>> array_pop   1978   1   3   %oparray_pop   restore   restore   2   1   1 restor
> >>>>> e   %for_pos_int_continue   1981   1   7   %oparray_pop   restore restore   18
> >>>>> 94   2   9   %oparray_pop
> >>>>> Dictionary stack:
> >>>>>      --dict:737/1123(ro)(G)--   --dict:1/20(G)--   --dict:80/200(L)--
> >>>>> --dict:80/
> >>>>> 200(L)--   --dict:135/256(ro)(G)--   --dict:318/325(ro)(G)-- --dict:33/64(L)--
> >>>>>      --dict:6/9(L)--   --dict:6/20(L)--
> >>>>> Current allocation mode is local
> >>>>> GPL Ghostscript 9.52: Unrecoverable error, exit code 1
> >>>>
> >>>> I can confirm this.  As a workaround, downgrade ghostscript and libgs9 to
> >>>> the previous version, 9.50-1, while I investigate.
> >>>>
> >>>> Thanks for the report.
> >>>>
> >>>> Ken
> >>>
> >>> Hi!
> >>>
> >>> The first bad commit causing the issue seems to be 327dc89:
> >>>
> >>> http://git.ghostscript.com/?p=ghostpdl.git;a=commitdiff;h=327dc8943c0aac1c7be7a1fb6e93346b61cd0900
> >>>
> >>
> >> Thanks!  I think the problem is here:
> >>
> >> @@ -234,18 +242,14 @@ CONTRIBINCLUDE="include $srcdir/contrib/contrib.mak"
> >>   INSTALL_CONTRIB="install-contrib-extras"
> >>
> >>   if test x"$enable_contrib" = x; then
> >> -  if test x"$cross_compiling" = x"yes"; then
> >> -    enable_contrib=
> >> -  else
> >> -    case `uname` in
> >> -      MINGW*|MSYS*)
> >> +    case $host in
> >> +      *-mingw*|*-msys*|*-cygwin*)
> >>           AC_MSG_WARN([disabling contrib devices])
> >>           enable_contrib=no
> >>         ;;
> >>         *)
> >>         ;;
> >>       esac
> >> -  fi
> >>   fi
> >>
> >>   if test x"$enable_contrib" != x"no"; then
> >>
> >> Do you want to send the obvious fix upstream or should I do it?
> >
> > Let's hold off on this.  Just making this one change doesn't solve the problem.
> > I've found several other commits that have mistakenly lumped cygwin with mingw
> > and msys.  I'm still working on it.
> 
> It turned out that fixing that one commit was all that was required, but the hunk I quoted above wasn't the culprit.
> 
> I've just uploaded ghostscript 9.52-2, which should fix the problem
> 
> Ken

Now, it works.
You are great.
Thank you!

Lem