From bug 189672: "I got an ACCESS VIOLATION while I emerged gnuplot 4.2.2-r1: --------------------------- ACCESS VIOLATION SUMMARY --------------------------- LOG FILE = "/var/log/sandbox/sandbox-sci-visualization_-_gnuplot-4.2.2-r1-14203.log" open_wr: /var/cache/fonts/ls-R"
Please elaborate: Output of emerge --info and emerge -pv gnuplot plus attach the build.log file.
This is absolutely standard situation, I guess a very large fraction of all ebuilds in the tree have this problem. Some try to work around it, some don't. Emerge calls latex to compile documentation (usually with the USE flag doc). Some font which has not yet been used on this particular computer is encountered. latex calls metafont to generate it. metafont wants to write it to /var/cache/fonts. Access violation. It is difficult to catch this situation, because it depends on the history of prior font usage on a particular computer (for example, I use LaTeX often and for many documents, and, probably, all popular fonts are already in my /var/cache/fonts). Writing tp /var/cache/fonts has *absolutely* nothing to do with files the package wants to install. Some ebuilds use the following ugly hack: for dir in `find /var/cache/fonts -type d`; do addwrite ${dir}; done or some variation of it. Not nice. In the thread http://thread.gmane.org/gmane.linux.gentoo.devel/52950/ it was suggested to VARTEXFONTS="${T}/fonts" in the ebuild. Then *all* fonts encountered in the package LaTeX documentation will be missing; they will be generated by metafont, used to produce dvi, and then thrown away. Perhaps better, though inefficient. It would be better to solve this problem once and for all hundreds of packages containing LaTeX documentation. But I don't know how.
Yes, I suspected something like that, but wanted to get sure with the build.log... Anyway, I added VARTEXFONTS="${T}/fonts" to the ebuild. René, would you please try again (in about 30 minutes, when the rsync servers have spread the word).
(In reply to comment #3) > Anyway, I added VARTEXFONTS="${T}/fonts" to the ebuild. René, would you please > try again (in about 30 minutes, when the rsync servers have spread the word). I take this as a yes. Closing.
I'm currently as of 8-23-08 haveing this problem emerge -av gives sci-visualization/gnuplot-4.2.3-r2 USE="X emacs gd ggi latex pdf plotutils readline wxwindows xemacs -doc (-svga)" 0 kB and it errors kdir -p -- /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot /usr/bin/install -c -m 644 gnuplot-gui.el /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/gnuplot-gui.el /usr/bin/install -c -m 644 gnuplot.el /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/gnuplot.el /usr/bin/install -c -m 644 info-look.20.2.el /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/info-look.20.2.el /usr/bin/install -c -m 644 info-look.20.3.el /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/info-look.20.3.el Compiling /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/gnuplot.el...Compiling /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/gnuplot-gui.el...Compiling /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/info-look.el... Wrote /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/gnuplot.elc Done Wrote /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/info-look.elc Done Wrote /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/work/gnuplot-4.2.3/lisp/gnuplot-gui.elc Done /usr/bin/install -c -m 644 gnuplot.elc /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/gnuplot.elc /usr/bin/install -c -m 644 gnuplot-gui.elc /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/gnuplot-gui.elc /usr/bin/install -c -m 644 info-look.elc /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image//usr/lib/xemacs/site-packages/gnuplot/info-look.elc >>> Completed installing gnuplot-4.2.3-r2 into /var/tmp/portage/sci-visualization/gnuplot-4.2.3-r2/image/ emerge --info reports Portage 2.1.4.4 (default-linux/amd64/2007.0, gcc-4.2.4, glibc-2.6.1-r0, 2.6.26-sabayon x86_64) ================================================================= System uname: 2.6.26-sabayon x86_64 AMD Athlon(tm) 64 X2 Dual Core Processor 4600+ Timestamp of tree: Sat, 23 Aug 2008 13:00:01 +0000 distcc 2.18.3 x86_64-pc-linux-gnu (protocols 1 and 2) (default port 3632) [disabled] app-shells/bash: 3.2_p33 dev-java/java-config: 1.3.7, 2.1.6 dev-lang/python: 2.5.2-r7 dev-python/pycrypto: 2.0.1-r6 sys-apps/baselayout: 2.0.0 sys-apps/openrc: 0.2.5-r1 sys-apps/sandbox: 1.2.18.1-r2 sys-devel/autoconf: 2.13, 2.62-r1 sys-devel/automake: 1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r2, 1.10.1 sys-devel/binutils: 2.18-r1 sys-devel/gcc-config: 1.4.0-r4 sys-devel/libtool: 1.5.26 virtual/os-headers: 2.6.24 ACCEPT_KEYWORDS="amd64 ~amd64" CBUILD="x86_64-pc-linux-gnu" CFLAGS="-Os -march=x86-64 -pipe" CHOST="x86_64-pc-linux-gnu" CONFIG_PROTECT="/etc /usr/kde/3.5/env /usr/kde/3.5/share/config /usr/kde/3.5/share/config/kdm /usr/kde/3.5/shutdown /usr/kde/4.0/env /usr/kde/4.0/share/config /usr/kde/4.0/shutdown /usr/share/config" CONFIG_PROTECT_MASK="/etc/env.d /etc/env.d/java/ /etc/fonts/fonts.conf /etc/gconf /etc/gentoo-release /etc/php/apache2-php5/ext-active/ /etc/php/cgi-php5/ext-active/ /etc/php/cli-php5/ext-active/ /etc/revdep-rebuild /etc/splash /etc/terminfo /etc/texmf/web2c /etc/udev/rules.d" CXXFLAGS="-Os -march=x86-64 -pipe" DISTDIR="/usr/portage/distfiles" FEATURES="collision-protect distlocks metadata-transfer sandbox sfperms strict unmerge-orphans userfetch" GENTOO_MIRRORS="http://distfiles.gentoo.org http://distro.ibiblio.org/pub/linux/distributions/gentoo" LANG="en_US.UTF-8" LC_ALL="en_US.UTF-8" LINGUAS="en" MAKEOPTS="-s -j3" PKGDIR="/usr/portage/packages" PORTAGE_RSYNC_OPTS="--recursive --links --safe-links --perms --times --compress --force --whole-file --delete --stats --timeout=180 --exclude=/distfiles --exclude=/local --exclude=/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR="/usr/portage" PORTDIR_OVERLAY="/usr/portage/local/layman/sabayon /usr/portage/local/Joe" SYNC="rsync://rsync.gentoo.org/gentoo-portage" USE="7Zip X a52 aac aalib accessibility acl acpi aiglx aim alsa amd64 artswrappersuid asf audiofile avahi berkdb bidi bluetooth bzip2 cairo cdda cddb cdr chm cjk cli commercial config_wizard cpudetection cracklib crypt css cups dbox2 dbus device-mapper dga dri dts dv dvb dvd dvdr dvdread dvi dxr3 encode esd exscalibar extramodules fam fame fat ffmpeg fftw firefox flac flash foomatic-db fortran freetype gcj gdbm gif gimpprint gnutls gphoto2 gpm gs gsm gstreamer gtk hal hfs iconv icq ieee1394 imap inotify ipod ipv6 irc irda isdnlog jabber jack jfs jingle joystick jpeg kde kdeenablefinal kdehiddenvisibility kerberos kickoff kipi lame lcd libao libnotify lirc live lj lm_sensors logitech-mouse lzo mad mail midi mikmod mjpeg mmx mng mozdevelop mp3 mp3rtp mpeg msn mudflap musepack musicbrainz ncurses network new-login nls nptl nptlonly nsplugin ntfs ogg openal openexr opengl openmp pam pcmcia pcre pda perl png povray ppds pppd pulseaudio pwdb python qt3 qt3support qt4 quotas rar rdesktop readline reflection reiserfs rss scanner sdl session slp sms speex spell spl spreadsheet sse sse2 ssl startup-notification stream svg switch_all_desktops sysfs tcpd theora tiff tracker truetype udev unicode usb v4l v4l2 visualization voice vorbis weather wifi wmf x264 xcb xfs xine xinerama xml xorg xpm xprint xulrunner xvid xvmc yahoo zeroconf zlib" ALSA_CARDS="hda-intel bt87x" ALSA_PCM_PLUGINS="adpcm alaw asym copy dmix dshare dsnoop empty extplug file hooks iec958 ioplug ladspa lfloat linear meter mmap_emul mulaw multi null plug rate route share shm softvol" APACHE2_MODULES="actions alias auth_basic authn_alias authn_anon authn_dbm authn_default authn_file authz_dbm authz_default authz_groupfile authz_host authz_owner authz_user autoindex cache dav dav_fs dav_lock deflate dir disk_cache env expires ext_filter file_cache filter headers include info log_config logio mem_cache mime mime_magic negotiation rewrite setenvif speling status unique_id userdir usertrack vhost_alias" APACHE2_MPMS="prefork" CAMERAS="agfa_cl20 casio_qv dimagev dimera3500 kodak_dc120 kodak_dc210 kodak_dc240 kodak_dc3200 kodak_ez200 konica_qm150 panasonic_coolshot panasonic_dc1000 panasonic_dc1580 panasonic_l859 polaroid_pdc320 polaroid_pdc640 polaroid_pdc700 ricoh_g3 sipix_blink sipix_blink2 sipix_web2 sony_dscf1 sony_dscf55 toshiba_pdrm11" ELIBC="glibc" INPUT_DEVICES="keyboard mouse evdev" KERNEL="linux" LCD_DEVICES="bayrad cfontz cfontz633 glk hd44780 lb216 lcdm001 mtxorb ncurses text" LINGUAS="en" LIRC_DEVICES="all" USERLAND="GNU" VIDEO_CARDS="nvidia" --------------------------- ACCESS VIOLATION SUMMARY --------------------------- LOG FILE = "/var/log/sandbox/sandbox-8876.log" open_wr: /usr/share/texmf/ls-R open_wr: /usr/share/texmf-site/ls-R --------------------------------------------------------------------------------
Reopening and adding ulm... Joe, do you use Sabayon or Gentoo?
(In reply to comment #6) > Reopening and adding ulm... > > Joe, do you use Sabayon or Gentoo? > I using sabayon just woundering what dose that have to do with it? It use a gentoo ebuild useing the latest none mask portage and sandbox. And sabyon dose not have an overlay for gnuplot.(I checked)
(In reply to comment #7) > I using sabayon [...] Please reproduce the bug on a Gentoo system and reopen.
(In reply to comment #8) > (In reply to comment #7) > > I using sabayon [...] > > Please reproduce the bug on a Gentoo system and reopen. > Ulrich Müller wrote Please reproduce the bug on a Gentoo system and reopen. Please don't take this the wrong way but How i'm I suposed to do that It like tell you to use sabayon and reproduce it wich in my oppion far more doable. I Have a litmit on how much i can download thou my isp which makes this imposible to do. The only reason i post this here is I have see an ebuild do dumb stuff already they put chmod with the wrong periminters wich could been easly fixed if the devloper check his ebuild(even thou i admit it was in testing) I put this here to let Devloper know there a problem(please note i did not open a bug). If the maintaner of an ebuild don't what to fix there problem why become a maintaner? I know you guyes are busy and all I respect that you don't what to spend all day and work on a sabayon system wich has far to many mask packages but could you at least point me in the right direction like tell me if there some packages that likely to cause problems and (could you tell me it workes with gentoo) if you tested it on gentoo. As stated above i see ebuild that where not test buy devloper (even thou it was mask.) I don't think it is fair to totaly thou out helping out pepole with sabayon because they use mask packages(It encuarge pepole to lie aboout the os there using, which i am not going to do)
(In reply to comment #9) Hi Joe, The latest ebuilds in Gentoo's portage should have a fix for these access violations by latex. Not knowing much about Sabayon can you make sure that these changes have propagated to Sabayon already? You'd be looking for the command VARTEXFONTS="${T}/fonts" right at the beginning of src_compile(). If this is missing you should contact the sabayon folks to propagate this fix. More generally, it happens quite frequently that small fixes (like the above) are applied to ebuilds without a revision bump. Hence, the fact that you are emerging gnuplot-4.2.2-r1 on Sabayon does not mean that you are using the most recent ebuild present in Gentoo's portage depending on how often the Sabayon folks sync with our tree which I don't know. If they sync at least once a day that's fine as long as you use the most recent tree as well. Otherwise, as a Sabayon user it will be crucial that you check Gentoo's portage first and make sure you're using the latest ebuild version and if not, try Gentoo's more recent one first. Then, if that fails as well, I'd personally be more than happy to look at the problem. > The only reason i post this here is I have see an ebuild do dumb stuff already > they put chmod with the wrong periminters wich could been easly fixed if the > devloper check his ebuild(even thou i admit it was in testing) > Could you please elaborate! It is most easy for us to fix things if you are as specific as possible. Thank you very much! Best, Markus
(In reply to comment #10) > VARTEXFONTS="${T}/fonts" Hm, but the issue from comment #5 was a different one: open_wr: /usr/share/texmf/ls-R open_wr: /usr/share/texmf-site/ls-R In fact, the build system of the package calls texhash during the install phase (which shouldn't harm since it will detect that the cache is not writable and do nothing in this case). Can you please post the output of the following two commands? $ texhash --version $ sandbox texhash
(In reply to comment #11) > (In reply to comment #10) > > VARTEXFONTS="${T}/fonts" > > Hm, but the issue from comment #5 was a different one: > open_wr: /usr/share/texmf/ls-R > open_wr: /usr/share/texmf-site/ls-R > > In fact, the build system of the package calls texhash during the install phase > (which shouldn't harm since it will detect that the cache is not writable and > do nothing in this case). > > Can you please post the output of the following two commands? > $ texhash --version > $ sandbox texhash > Sure $ texhash --version reports texhash $Id: mktexlsr,v 1.46 2005/06/21 14:32:26 olaf Exp $ kpathsea version 3.5.5 Copyright 2005 Karl Berry & Olaf Weber. There is NO warranty. You may redistribute this software under the terms of the GNU General Public License. For more information about these matters, see the files named GPL and LGPL. sandbox texhash reports: ACCESS DENIED open_wr: /usr/share/texmf/ls-R texhash: /usr/share/texmf/ls-R: no write permission. Skipping... ACCESS DENIED open_wr: /usr/share/texmf-site/ls-R touch: cannot touch `/usr/share/texmf-site/ls-R': Permission denied chmod: cannot access `/usr/share/texmf-site/ls-R': No such file or directory texhash: /usr/share/texmf-site/ls-R: no write permission. Skipping... texhash: /var/cache/fonts/ls-R: no write permission. Skipping... texhash: /var/lib/texmf/ls-R: no write permission. Skipping... texhash: Done. --------------------------- ACCESS VIOLATION SUMMARY --------------------------- LOG FILE = "/var/log/sandbox/sandbox-28543.log" open_wr: /usr/share/texmf/ls-R open_wr: /usr/share/texmf-site/ls-R after i remebered to run it as root Please note: i do not know what the last statment dose.
(In reply to comment #10) > (In reply to comment #9) > > Hi Joe, > > The latest ebuilds in Gentoo's portage should have a fix for these > access violations by latex. Not knowing much about Sabayon can you make > sure that these changes have propagated to Sabayon already? You'd be > looking for the command > > VARTEXFONTS="${T}/fonts" > > right at the beginning of src_compile(). If this is missing you should > contact the sabayon folks to propagate this fix. > > More generally, it happens quite frequently that small fixes (like the > above) are applied to ebuilds without a revision bump. Hence, the fact > that you are emerging gnuplot-4.2.2-r1 on Sabayon does not mean that you > are using the most recent ebuild present in Gentoo's portage depending > on how often the Sabayon folks sync with our tree which I don't know. > If they sync at least once a day that's fine as long as you use > the most recent tree as well. Otherwise, as a Sabayon user it will be > crucial that you check Gentoo's portage first and make sure you're using > the latest ebuild version and if not, try Gentoo's more recent one first. > Then, if that fails as well, I'd personally be more than happy to look at > the problem. > First let me say Yes, it is in there and thankyou. Secoundly let me say that I am very confused about the above commit. What do you think sabayon dose? My only guess is you think it snyc with gentoo mirror and add patches. Witch would be a teribale way of doing things but I can't say for 100% sure it dose not do but guse strongly (99.9%) not. It looks to me sabayon put all there ebuild in an overlay and there is no gnuplot ebuild in the overlay looking in dir /usr/portage/local/layman/sabayon It is missing the sci-visualization dir. I have to use layman -S to update it. I also made sure I ran eix-sync in a fairly resent time frame of my first post. > > > > The only reason i post this here is I have see an ebuild do dumb stuff already > > they put chmod with the wrong periminters wich could been easly fixed if the > > devloper check his ebuild(even thou i admit it was in testing) > > > > Could you please elaborate! It is most easy for us to fix things if you are > as specific as possible. > > Thank you very much! > this was a very long time ago on another ebuild I think it was net-analyzer/jffnms I don't know if is still in there or is the correct ebuild.I think I was use a 32 bit system at the time before I switch to the 64-bit system. I did not feel like reporting it as a bug at the time because it seems like when every i Post something it turn out to be a nightmare on both side and far more simpler and probaly better to ingnore it. Please note i was not expecting you to fix it. > Best, > Markus > Thank again these type of comment I look forward too.
(In reply to comment #12) > sandbox texhash > > ACCESS DENIED open_wr: /usr/share/texmf/ls-R > texhash: /usr/share/texmf/ls-R: no write permission. Skipping... > ACCESS DENIED open_wr: /usr/share/texmf-site/ls-R > touch: cannot touch `/usr/share/texmf-site/ls-R': Permission denied > chmod: cannot access `/usr/share/texmf-site/ls-R': No such file or directory > texhash: /usr/share/texmf-site/ls-R: no write permission. Skipping... > [...] O.K., this information was helpful and I can reproduce the bug now. The following will trigger the sandbox violation: 1. rm /usr/share/texmf-site/ls-R 2. emerge =gnuplot-4.2.3-r2 Reopening.
I've disabled texhash in the Makefile of the package; in any case pkg_postinst() calls texmf-update which does the same thing (and more). No revbump because it was a build-time issue. Please re-sync and try if you can emerge gnuplot-4.2.3-r2 now.
(In reply to comment #15) > I've disabled texhash in the Makefile of the package; in any case > pkg_postinst() calls texmf-update which does the same thing (and more). No > revbump because it was a build-time issue. > > Please re-sync and try if you can emerge gnuplot-4.2.3-r2 now. > Thanks eix-sync and now it works Thank you for you help.
Thanks for reporting. Closing.
Finally fixed upstream, so the disable-texhash patch can be removed in >gnuplot-4.4.0-rc1.
