Gentoo Websites Logo
Go to: Gentoo Home Documentation Forums Lists Bugs Planet Store Wiki Get Gentoo!
Bug 672762 - media-libs/netpbm-10.76.00: pbmtog3.test, pbm-misc-converters.test, g3-roundtrip.test fail on ppc, hppa, sparc
Summary: media-libs/netpbm-10.76.00: pbmtog3.test, pbm-misc-converters.test, g3-roundt...
Status: RESOLVED FIXED
Alias: None
Product: Gentoo Linux
Classification: Unclassified
Component: Current packages (show other bugs)
Hardware: PPC Linux
: Normal normal (vote)
Assignee: Viorel Munteanu
URL:
Whiteboard:
Keywords: REGRESSION, TESTFAILURE
Depends on:
Blocks: 671068
  Show dependency tree
 
Reported: 2018-12-08 18:27 UTC by ernsteiswuerfel
Modified: 2021-11-10 07:40 UTC (History)
2 users (show)

See Also:
Package list:
Runtime testing required: ---


Attachments
build.log (netpbm-10.76.00:20181208-180937.log,703.21 KB, text/plain)
2018-12-08 18:27 UTC, ernsteiswuerfel
Details
emerge --info (file_672762.txt,6.16 KB, text/plain)
2018-12-08 18:28 UTC, ernsteiswuerfel
Details
build.log (10.70.00) (netpbm-10.70.00:20181208-184332.log,668.72 KB, text/plain)
2018-12-08 18:50 UTC, ernsteiswuerfel
Details

Note You need to log in before you can comment on or make changes to this bug.
Description ernsteiswuerfel archtester 2018-12-08 18:27:11 UTC
Created attachment 557336 [details]
build.log

== pbmtog3.test ==
pbmtog3.test: FAILURE
== pbm-misc-converters.test ==
pbmtoibm23xx: opening 'testgrid.pbm'
pbm-misc-converters.test: FAILURE
== g3-roundtrip.test ==
g3topbm: Problem reading Row 0.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Warning: Image contains a line of 33 pixels.  You specified lines should be 14 pixels..
g3topbm: Problem reading Row 1.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 2.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 3.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 4.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 5.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 6.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 7.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 8.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 9.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 10.  Skipping rest of row.  bad code word at Column 33.  No prefix of the 14 bits 0x0 matches any recognized code word and no code words longer than 13 bits are defined.  
g3topbm: Problem reading Row 11.  Skipping rest of row.  EOF or error reading file
g3topbm: Unable to read Row 12.  Skipping rest of image.  EOF or error reading file
[...]
g3-roundtrip.test: FAILURE
Comment 1 ernsteiswuerfel archtester 2018-12-08 18:28:02 UTC
Created attachment 557338 [details]
emerge --info
Comment 2 ernsteiswuerfel archtester 2018-12-08 18:50:22 UTC
Created attachment 557340 [details]
build.log (10.70.00)

Regression over current stable 10.70.00.

pbmtog3.test and g3-roundtrip.test pass in 10.70.00. pbm-misc-converters.test is not included in 10.70.00.
Comment 3 Rolf Eike Beer archtester 2018-12-15 11:13:07 UTC
Same tests fail on sparc.
Comment 4 ernsteiswuerfel archtester 2019-05-04 23:47:54 UTC
Re-tested from latest svn stable 10.73.27 and sent a mail to bryanh@giraffe-data.com for upstream bug report.
Comment 5 Rolf Eike Beer archtester 2020-05-18 16:58:58 UTC
Fails on hppa, too. Endianess issue? What happened to your upstream report?
Comment 6 ernsteiswuerfel archtester 2020-05-18 20:01:20 UTC
(In reply to Rolf Eike Beer from comment #5)
> Fails on hppa, too. Endianess issue? What happened to your upstream report?
Last email I got as an answer 26.05.2019:

"With the /tmp/netpbm-test stuff you sent, I was able to find the
pbm-misc-converters problem.  It's an endianness bug introduced 4 years ago in
Pbmtozinc, a program that is unlikely ever to be used again.  I have fixed it
in the current Development release and will fix it in the next Stable and
Advanced releases in about two weeks.

The palm-roundtrip and palm-roundtrip2 failures are with new pnmtopalm
-packbits_compression code, and I think I see endian errors in there too; I am
asking the author of that code to look at it.

For the ps-alt-roundtrip problem, I need you to do an experiment and send
me the results:

  CD to the 'test' directory of the Netpbm source tree

  $ pbmtolps -dpi 72 testgrid.pbm >/tmp/testgrid.ps

Send me that /tmp/testgrid.ps that creates, and any error or warning messages
you see.

-- 
Bryan Henderson                                   San Jose, California"

Mailed the requested /tmp/testgrid.ps as an answer.
Comment 7 ernsteiswuerfel archtester 2020-05-18 20:16:24 UTC
According to the changelog https://sourceforge.net/p/netpbm/code/HEAD/tree/advanced/doc/HISTORY one endianness and several pnmtopalm (see email) bugs were fixed in more recent versions.

19.06.28 BJH  Release 10.87.00
[...]
pbmtozinc: fix wrong output on big-endian machines.  Broken in
Netpbm 10.71 (June 2015).

17.09.30 BJH  Release 10.80.00
[...]
pnmtopalm: Refuse to create a compressed image with more than 8
bits per pixel.

pnmtopalm: fix incorrect output with certain input files and
-packbits_compression.  Always broken.  -packbits_compression
was new in Netpbm 10.27 (March 2005).

pnmtopalm: Correct error message recommending running
pnmquant when the real solution is to run pnmremap to modify the
image to Palm standard colors.

pnmtopalm: Fix bug causing "color not in colormap" failure with
-colormap and maxval other than 255.
Comment 8 Viorel Munteanu gentoo-dev 2021-11-10 07:40:46 UTC
netpbm-10.86.24 has been tested on sparc and looks ok.

Closing the bug as resolved.