[SOLVED] feedback request on possible upgrade of ghostscript

[aus9]

solved as submitted

I have permisson to try to upgrade, rather than pester Juanito lets share the load?

For each list I ran diff with the following results

ghostscript
no changes

ghostscript_common loses dumphint and pdfopt
IMHO they don't appear to be essential

libgs
so bumped to 9.07
packages that depend on libgs are evince.tcz.dep graphviz.tcz.dep libgs-dev.tcz.dep libspectre.tcz.dep

libgs-dev
new ------gxfapi.h icc34.h wrfont.h write_t1.h write_t2.h
missing-- iapi.h  ierrors.h sjpx.h

these missing headers can be found in one or the other of libgs-dev or libgs_ps
but I am building libgs-dev

and libgs_ps uses so.8 and this stage I prefer not to build that as its not a frequent used package IMHO


trivia here is current build, sorry if I am simple coder but I am trying to be smarter

Code: [Select]

build removed as it had wrong owner and permissions as per wiki

thanks for reading

I also have permission to supply a link thru pm for those who opt in to offer to test any new packages as my printer is not working
any one interested?

I have not uploaded these tczs yet to my cloud storage in case feedback suggests I still need to rebuild.

cheers

[Juanito]

You can use:

Code: [Select]

export CXXFLAGS="-march=i486 -mtune=i686 -Os -pipe -fno-exceptions -fno-rtti"
..and it will usually produce smaller output unless exceptions are specifically required in which case "make" will halt with errors.

[aus9]

 may have mislead

the packages build, otherwise I would not be able to compare the lists.
I grab the lists from APPS and compare them to the new tcz lists.
--I used leafpad and its replace all function to adjust for the first /  and 9.05 vs 9.07

I will try to look at what difference -fno-exceptions -fno-rtti makes to my build depends later thanks

[aus9]

ok I have submitted foomatic-db and its engine so can now start to test this script and prepare tczs for submission

If no-one has time to test tczs before submission thats ok, I will mark info files as TESTING and ask for a backup of existing tczs when I submit. Here is the output for what devices gs may be good for...in a code box

Code: [Select]

Input formats: PostScript PostScriptLevel1 PostScriptLevel2 PostScriptLevel3 PDF
Default output device: x11alpha
Available devices:
   alc1900 alc2000 alc4000 alc4100 alc8500 alc8600 alc9100 ap3250 appledmp
   atx23 atx24 atx38 bbox bit bitcmyk bitrgb bitrgbtags bj10e bj10v bj10vh
   bj200 bjc600 bjc800 bjc880j bjccmyk bjccolor bjcgray bjcmono bmp16 bmp16m
   bmp256 bmp32b bmpgray bmpmono bmpsep1 bmpsep8 ccr cdeskjet cdj1600 cdj500
   cdj550 cdj670 cdj850 cdj880 cdj890 cdj970 cdjcolor cdjmono cdnj500 cfax
   chp2200 cif cljet5 cljet5c cljet5pr coslw2p coslwxl cp50 cups declj250
   deskjet devicen dfaxhigh dfaxlow dj505j djet500 djet500c dl2100 dnj650c
   epl2050 epl2050p epl2120 epl2500 epl2750 epl5800 epl5900 epl6100 epl6200
   eplcolor eplmono eps9high eps9mid epson epsonc epswrite escp escpage
   faxg3 faxg32d faxg4 fmlbp fmpr fs600 gdi hl1240 hl1250 hl7x0 hpdj1120c
   hpdj310 hpdj320 hpdj340 hpdj400 hpdj500 hpdj500c hpdj510 hpdj520 hpdj540
   hpdj550c hpdj560c hpdj600 hpdj660c hpdj670c hpdj680c hpdj690c hpdj850c
   hpdj855c hpdj870c hpdj890c hpdjplus hpdjportable ibmpro ijs imagen
   inferno inkcov iwhi iwlo iwlq jetp3852 jj100 jpeg jpegcmyk jpeggray la50
   la70 la75 la75plus laserjet lbp310 lbp320 lbp8 lex2050 lex3200 lex5700
   lex7000 lips2p lips3 lips4 lips4v lj250 lj3100sw lj4dith lj4dithp lj5gray
   lj5mono ljet2p ljet3 ljet3d ljet4 ljet4d ljet4pjl ljetplus ln03 lp1800
   lp1900 lp2000 lp2200 lp2400 lp2500 lp2563 lp3000c lp7500 lp7700 lp7900
   lp8000 lp8000c lp8100 lp8200c lp8300c lp8300f lp8400f lp8500c lp8600
   lp8600f lp8700 lp8800c lp8900 lp9000b lp9000c lp9100 lp9200b lp9200c
   lp9300 lp9400 lp9500c lp9600 lp9600s lp9800c lps4500 lps6500 lq850
   lxm3200 lxm5700m m8510 mag16 mag256 md1xMono md2k md50Eco md50Mono md5k
   mgr4 mgr8 mgrgray2 mgrgray4 mgrgray8 mgrmono miff24 mj500c mj6000c
   mj700v2c mj8000c ml600 necp6 npdl nullpage oce9050 oki182 oki4w okiibm
   oprp opvp paintjet pam pamcmyk32 pamcmyk4 pbm pbmraw pcl3 pcx16 pcx24b
   pcx256 pcx256 pcx2up pcxcmyk pcxgray pcxmono pdfwrite pdfwrite pgm pgmraw
   pgnm pgnmraw photoex picty180 pj pjetxl pjxl pjxl300 pkm pkmraw pksm
   pksmraw plan plan9bm planc plang plank planm png16 png16m png256 png48
   pngalpha pnggray pngmono pnm pnmraw ppm ppmraw pr1000 pr1000_4 pr150
   pr201 ps2write psdcmyk psdrgb psgray psmono psrgb pswrite pxlcolor
   pxlmono r4081 rinkj rpdl samsunggdi sgirgb sj48 spotcmyk st800 stcolor
   sunhmono t4693d2 t4693d4 t4693d8 tek4696 tiff12nc tiff24nc tiff32nc
   tiff48nc tiff64nc tiffcrle tiffg3 tiffg32d tiffg4 tiffgray tifflzw
   tiffpack tiffscaled tiffscaled24 tiffscaled8 tiffsep tiffsep1 txtwrite
   uniprint x11 x11alpha x11cmyk x11cmyk2 x11cmyk4 x11cmyk8 x11gray2
   x11gray4 x11mono xcf xes

I keep on thinking epson xp-100 is there, but I am the local village idiot

[aus9]

error fixed due to build error

[Juanito]

I'm not sure if you did it already, but it's worth reading the ghostscript entry in blfs - I seem to recall you need to delete several subdirectories in the source and hack the Makefile to force it to use system zlib, png, etc.

[aus9]

Yes I have see something similar which is LFS based on 9.06 and earlier docs want you/me to sym link an unix-gcc.mak to Makefile/makefile
but thats in the past
http://ghostscript.com/doc/9.07/Install.htm#Install_Unix

2) make takes a long time, but I mananged to see a build error scroll by but not the actual message so I have gone backwards and moving forward from here---without showing full build

Code: [Select]

export CFLAGS="-march=i486 -mtune=i686 -Os -pipe"
export CXXFLAGS="-march=i486 -mtune=i686 -Os -pipe -fno-exceptions -fno-rtti"
export LDFLAGS="-Wl,-O1"
/usr/bin/tce-load -i compiletc submitqc4 wget coreutils
/usr/bin/tce-load -i zlib_base-dev jpeg_base-dev libpaper-dev gtk2-dev cups-dev texlive
/usr/bin/tce-load -i dbus-dev tiff3-dev libxml2-dev lcms2-dev libidn-dev

./configure --prefix=/usr/local --with-drivers=ALL --with-install-cups  --sysconfdir=/usr/local \
--enable-dynamic --enable-fontconfig

vfork.h had 2 candidates and neither looked the right deal I did try cdrtools-dev but any how here is current = a shorter error list, best yet

Code: [Select]

tc@box:/tmp/ghostscript-9.07$ cat config.log | grep error
gcc: error: unrecognized option '-V'
gcc: fatal error: no input files
gcc: error: unrecognized option '-qversion'
gcc: fatal error: no input files
conftest.c:9:28: fatal error: ac_nonexistent.h: No such file or directory
conftest.c:9:28: fatal error: ac_nonexistent.h: No such file or directory
conftest.c:68:18: fatal error: ndir.h: No such file or directory
conftest.c:35:18: fatal error: ndir.h: No such file or directory
conftest.c:69:22: fatal error: sys/ndir.h: No such file or directory
conftest.c:36:22: fatal error: sys/ndir.h: No such file or directory
conftest.c:70:24: fatal error: sys/window.h: No such file or directory
conftest.c:37:24: fatal error: sys/window.h: No such file or directory
conftest.c:73:21: error: expected expression before ')' token
conftest.c:73:20: error: expected expression before ')' token
conftest.c:73:21: error: expected expression before ')' token
configure:8528: checking for strerror
conftest.c:91:20: error: expected expression before ')' token
conftest.c:88:19: fatal error: vfork.h: No such file or directory
conftest.c:55:19: fatal error: vfork.h: No such file or directory
conftest.c:67:1: error: void value not ignored as it ought to be
| /* Override any GCC internal prototype to avoid an error.
/tmp/tcloop/gcc/usr/local/bin/../lib/gcc/i686-pc-linux-gnu/4.6.1/include/emmintrin.h:32:3: error: #error "SSE2 instruction set not enabled"
conftest.c:88:3: error: unknown type name '__m128i'
conftest.c:90:3: error: unknown type name '__m128i'
ac_cv_func_strerror=yes

no need to reply, but if anyone can tell me how to resolve those kernel type messages that might help greatly eh?

make takes a while, and I now know I have to watch it to see anything naughty scroll by. In the past I was dozing off in front of tv

I also changed my depends to add libpaper and force ghostscript as a dep to _common even after reading your info file
By doing it, I actually had less execuables complaining.

cheers

[aus9]

looks like I might revert to my smaller flag style mentioned in build script first post

after running various configs options and not yet make stuff I get

Code: [Select]

checking if gcc supports -fno-rtti -fno-exceptions... no

to stop sse2 I will change to

Code: [Select]

./configure --prefix=/usr/local --with-drivers=ALL --with-install-cups  --sysconfdir=/usr/local \
--enable-dynamic --enable-fontconfig --disable-sse2

ok at config stage this is looking better

Libtiff is now configured for i686-pc-linux-gnu

  Installation directory:             /usr/local
  Documentation directory:            ${prefix}/share/doc/tiff-4.0.1
  C compiler:                         gcc -march=i486 -mtune=i686 -Os -pipe -DUSE_LIBPAPER -Wall -W
  C++ compiler:                       g++ -march=i486 -mtune=i686 -Os -pipe -fno-exceptions -fno-rtti
  Enable runtime linker paths:        no
  Enable linker symbol versioning:    no
  Support Microsoft Document Imaging: yes
  Use win32 IO:                       no

 Support for internal codecs:
  CCITT Group 3 & 4 algorithms:       yes
  Macintosh PackBits algorithm:       yes
  LZW algorithm:                      yes
  ThunderScan 4-bit RLE algorithm:    yes
  NeXT 2-bit RLE algorithm:           yes
  LogLuv high dynamic range encoding: yes

 Support for external codecs:
  ZLIB support:                       yes
  Pixar log-format algorithm:         yes
  JPEG support:                       yes
  Old JPEG support:                   yes
  JPEG 8/12 bit dual mode:            no
  ISO JBIG support:                   no
  LZMA2 support:                      no

  C++ support:                        yes

  OpenGL support:                     yes

that C++ compiler: still looks funny, I just went back to
export CFLAGS="-march=i486 -mtune=i686 -Os -pipe"
export CXXFLAGS="-march=i486 -mtune=i686 -Os -pipe"
export LDFLAGS="-Wl,-O1"

ahh well never mind time waits for no boy

[Juanito]

Code: [Select]

checking if gcc supports -fno-rtti -fno-exceptions... no


..but g++ (as opposed to gcc) does support this...

[aus9]

woohoo  except for the missing executables already mentioned above, they are now working.

some minor pests remain, with the new config settings...grep config.log for error still shows SSE2

I wonder if the earlier errors were because I did not force ghostscript as a dep of g*_common?

anyhow I am much happier

tce-load -i gs  test sym link goes to gsc-----------------------------OK
load -i ghostscript....sym link goes to tmp tcloop etc gs---------Great

edit see below

[aus9]

I got excited and submitted the new build and packages for approval and started tonite on gutenprint.

I was aware that hplip may no longer need ijs which can come from ghostscript so I did not enable it in build

silly me, gutenprint is wanting it, so the build may now look like

Code: [Select]

./configure --prefix=/usr/local --with-drivers=ALL --with-install-cups  --sysconfdir=/usr/local \
--enable-dynamic --enable-fontconfig --disable-sse2 --with-ijs

but not built and not tested. Naturally this means I can't submit any more inter-related packages until they are all built which being impatient means when it rains it will pour for my fav tcz checker.

cheers FYI

[aus9]

FYI
using above configure did not build any ijs, the subfolder ijs exists and has an autogen.sh
running autogen.sh gives some errors and I have decided after seeing the version has not changed from original source of ijs to leave ijs out of this build and will build ijs from source, as I can understand the config.log a bit easier

it builds so will test if the new ijs-config solves the gutenprint upgrade, and if no errors, will attempt submit ijs then gutenprint and come back to ghost later.

cheers

[aus9]

deleted offtopic discussion on gutenprint

[aus9]

ok so independent ijs is now submitted and when that lands I can submit gutenprint.

I am now thinking I can relook at submitting ghostscript unless I spot anything else,

[aus9]

back to testing and earlier I reported a strange box, silly me, its the preview to a pdf or other file

image shows I have selected only the top part of the pdf file on the right, thru mtpaint ,
but had no errors scrolling pages using the enter button


top of info file for feedback is now reading (some white space removed)

Title:          ghostscript.tcz
Description:    ghostscript
Version:        9.07
Author:         see AUTHORS
Original-site:  http://sourceforge.net/projects/ghostscript
Copying-policy: see LICENSE and COPYING
Size:      8.2M
Extension_by:   aus9
Tags:          ghostscript  printing PDF ps postscript
Comments:       Use input of PostScript(ps) or PDF and convert to something.  The main command is gs and for some info try $ gs -h
      eg $ gs -sDEVICE=jpeg -sOutputFile=foo.jpg foo.ps
      Can be used for pdf-> ps or the reverse.
      
      ** for Upgraders ** ghostscript-common is now a dep

      Be aware output to non-laser printers may be affected by inkjet dot size. You may wish to read up on dithering.

      Can be used to view various file formats including pdf eg if *-doc is loaded,
      $ gs /usr/local/share/doc/ghostscript/GS9_Color_Management.pdf
      running gs also causes it to change prompt to gs>..... like grub
      #### REMEMBER $ quit or < control and c keys > will exit the shell

      Possible encryption bypass http://www.cs.cmu.edu/~dst/Adobe/Gallery/Keating/