Tags:
create new tag
view all tags

GPDL Issues

This is the list of issues that need to be solved before we can fold GPDL down onto the master branch and have it tested regularly by the cluster.

I've removed cases where the same file fails in both banded and non-banded (or pdfwritten) cases, assuming that it's the same issue, and solving the non-banded one will solve the others.

SEGVs

tests_private/comparefiles/Bug695090.ps.ppmraw.300.0

./gs/bin/gpdl -sOutputFile='| md5sum >>./temp/tests_private__comparefiles__Bug695090.ps.ppmraw.300.0..gpdl.md5' -dMaxBitmap=400000000 -sDEVICE=ppmraw -r300 -Z: -dNOPAUSE -dBATCH -K2000000 -dClusterJob ./tests_private/comparefiles/Bug695090.ps 
PDL detected as POSTSCRIPT
GPL Ghostscript 9.27 (2018-09-13)
Copyright (C) 2018 Artifex Software, Inc. All rights reserved.
This software comes with NO WARRANTY: see the file PUBLIC for details.
% Start time = 0.074818, memory allocated = 5428305, used = 3291503, max_used = 4624193
% Init (Part 2/2) done, instance 0x0x3e35390.% Outputpage start time = 0.764548, memory allocated = 31736612, used = 4148955, max_used = 30732828
% Outputpage end time = 0.856342, memory allocated = 31736612, used = 4148979, max_used = 30732828
%%[Page: 1]%%
Unrecoverable error: stackunderflow in dup
Command terminated by signal 11
0.74 0.03 0:01.14 67%
return: 139
8d8454da16ca56131d4c857d9f41b74a -

The file itself exits from the PS interpreter having hit zforceinterp_exit (because there is a UEL in the file at the end).

The PS interpreter's handling of this causes us to leave in a state that upsets the setting of the device to NULL in the dnit_job. This is shown by the 'stackunderflow in dup'.

The actual SEGV is caused when the cursor is freed, and cursor->strm is not valid any more.

Handing this one off to Chris.

Errors

tests_private/customer_tests/ps-pcl-marigoldeuro.prn.ppmraw.75.0

 ./gs/bin/gpdl -sOutputFile='| md5sum >>./temp/tests_private__customer_tests__ps-pcl-marigoldeuro.prn.ppmraw.75.0..gpdl.md5' -dMaxBitmap=400000000 -sDEVICE=ppmraw -r75 -Z: -dNOPAUSE -dBATCH -K2000000 -dClusterJob ./tests_private/customer_tests/ps-pcl-marigoldeuro.prn 
PDL detected as PJL
PDL detected as POSTSCRIPT
GPL Ghostscript 9.27 (2018-09-13)
Copyright (C) 2018 Artifex Software, Inc. All rights reserved.
This software comes with NO WARRANTY: see the file PUBLIC for details.
% Start time = 0.112399, memory allocated = 6164729, used = 4016131, max_used = 5339969
% Init (Part 2/2) done, instance 0x0x7fc40a504e80.Loading NimbusMonoPS-Regular font from %rom%Resource/Font/NimbusMonoPS-Regular... 4531328 2944380 4693272 3406717 1 done.
%%[ ProductName: GPL Ghostscript ]%%
Loading NimbusSans-Bold font from %rom%Resource/Font/NimbusSans-Bold... 4597440 3123340 7554416 6218318 1 done.
Querying operating system for font files...
Can't find (or can't open) font file %rom%Resource/Font/Marigold.
Can't find (or can't open) font file Marigold.
Didn't find this font on the system!
Substituting font ZapfChancery-MediumItalic for Marigold.
Loading Z003-MediumItalic font from %rom%Resource/Font/Z003-MediumItalic... 4683752 3360466 7837216 6481487 2 done.
% Outputpage start time = 0.221023, memory allocated = 13872736, used = 10186490, max_used = 14321655
% Outputpage end time = 0.240159, memory allocated = 13872736, used = 10186490, max_used = 14321655
Unrecoverable error: stackunderflow in --%s_proc_read_continue--
...

Device Frobbing

These ones seem to be fiddling with the device within PS:

  • tests_private/ps/ps3cet/09-59.PS.ppmraw.300.0

Postscript Problems

/rangecheck in cvs:

  • tests_private/ps/ps3cet/25-13.PS.ppmraw.300.0
  • tests_private/ps/ps3cet/26-02.PS.ppmraw.300.0

/undefined in ......:

  • tests_private/ps/ps3cet/18-02F.PS.ppmraw.300.0
  • tests_private/ps/ps3cet/23-12B.PS.ppmraw.300.0
  • tests_private/ps/ps3cet/23-28.PS.ppmraw.300.0
  • tests_private/ps/ps3fts/119-23.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/119-28.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/119-35.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/119-41.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/119-47.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/206-01.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/246-01.ps.ppmraw.300.0
  • tests_private/ps/ps3fts/272-01.ps.ppmraw.300.0

typecheck in --show--:

  • tests_private/ps/ps3fts/299-01.ps.ppmraw.300.0

Known DIffs

tests_private/customer_tests/ip2000.pcl

tests_private/customer_tests/testip.pcl

Hits a UEL after 6 pages. PCL and GPDL differ after that. GPDL looks better, I think. PCL exhibits blocks of crap at various points. Possibly to do with fonts not being reset or something? I'm going to count this as a progression.

Umknown Diffs

And there are loads of files that give different results to gs. I will investigate why.

-- Robin Watts - 2018-11-13

Comments

Edit | Attach | Watch | Print version | History: r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r2 - 2018-11-19 - RobinWatts
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright 2014 Artifex Software Inc