March 2012 - Arch-general - lists.archlinux.org

[arch-general] ../libtool: line 2083: ark_part.cpp: command not found?
by David C. Rankin 19 Mar '12

19 Mar '12

Guys, I need help, or a link, to find out why libtool isn't working. libtool is attempting to use an undeclared variable as a command. The libtool file generated includes (beginning at line 2071): # func_to_tool_file ARG LAZY # converts the file name ARG from $build format to toolchain format. Return # result in func_to_tool_file_result. If the conversion in use is listed # in (the comma separated) LAZY, no conversion takes place. func_to_tool_file () { $opt_debug case ,$2, in *,"$to_tool_file_cmd",*) func_to_tool_file_result=$1 ;; *) $to_tool_file_cmd "$1" func_to_tool_file_result=$func_to_host_file_result ;; esac } # end func_to_tool_file The tdeutils build fails at make on the very first file (ark_part.cpp). The line that fails is line 2083: 2083| $to_tool_file_cmd "$1" The problem is '$to_tool_file_cmd' is never declared in 'libtool' resulting in the error: ../libtool: line 2083: ark_part.cpp: command not found Basically, libtool is trying to execute 'ark_part.cpp' as the command since '$to_tool_file_cmd' is never declared or initialized and is just blank in the command '$to_tool_file_cmd "$1"'. What I need to learn is whether this is a libtool problem or whether it is a tdeutils problem. tdeutils from the same source builds fine on slack, debian, etc.., so I wonder if this is a libtool version problem or something similar. I am building in an arch chroot and he PKGBUILD file used does: cd ${srcdir}/${pkgname#*-} cp /usr/share/aclocal/libtool.m4 ./admin cp /usr/share/libtool/config/ltmain.sh ./admin make -f admin/Makefile.common ./configure \ --prefix=${TDEDIR} \ --with-qt-dir=${QTDIR} \ --sysconfdir=/etc \ --localstatedir=/var \ --enable-debug=full \ --enable-closure make The complete error, which I don't fully understand is: /bin/sh ../libtool --tag=CXX --mode=compile g++ -DHAVE_CONFIG_H -I. -I.. -D_LARGEFILE64_SOURCE -I/opt/trinity/include -I/opt/tqt3/include -I. -include tqt.h -DQT_THREAD_SUPPORT -D_REENTRANT -Wno-long-long -Wundef -ansi -D_XOPEN_SOURCE=500 -D_BSD_SOURCE -Wcast-align -Wchar-subscripts -Wall -W -Wpointer-arith -fno-builtin -g3 -fno-inline -march=x86-64 -mtune=generic -O2 -pipe -fstack-protector --param=ssp-buffer-size=4 -D_FORTIFY_SOURCE=2 -Wformat-security -Wmissing-format-attribute -Wno-non-virtual-dtor -fno-exceptions -fno-check-new -fno-common -I/usr/include/tqt -DQT_CLEAN_NAMESPACE -DQT_NO_ASCII_CAST -DQT_NO_STL -DQT_NO_COMPAT -DQT_NO_TRANSLATION -MT ark_part.lo -MD -MP -MF .deps/ark_part.Tpo -c -o ark_part.lo ark_part.cpp ../libtool: line 2083: ark_part.cpp: command not found I have also tried 'make SHELL=/bin/bash' after googling that the environment could cause shell extension issues if part was done with bash and make was using sh, but that made no difference. Anybody familiar with libtool know if this is a libtool problem or if this is a problem with the tdeutil code? If it is libtool, any work-arounds? -- David C. Rankin, J.D.,P.E.

2 3

[arch-general] An empty /etc/crontab file maybe needed bo be add in to package filesystem.
by 郑文辉(Techlive Zheng) 19 Mar '12

19 Mar '12

Due to cronie doesn't contain an /etc/crontab anymore, there are losts of msgs of missing crontab as shown below in errors.log. > /usr/sbin/crond[1236]: (root) CAN'T OPEN (/etc/crontab): No such file or directory

6 17

Re: [arch-general] mach64 "undefined symbol: xf86MapDomainMemory" fix
by Ionut Biru 19 Mar '12

19 Mar '12

On 03/19/2012 05:41 AM, Daniel F. Savarese wrote: > > Anyone using the xf86-video-mach64 package who has had their X server > fail after updating to xorg-server 1.12, reporting > "undefined symbol: xf86MapDomainMemory", you can get it working again > by applying the following patch (I've inlined it because I don't know > if the mailing list accepts attachments). This is an upstream issue > fixed in revision control, so there's no point opening a bug report. > It will be resolved when a new release of xf86-video-mach64 becomes available. > report a bug report. > --- src/atividmem.c.orig 2012-03-18 23:01:56.666029002 -0400 > +++ src/atividmem.c 2012-03-18 23:03:28.809154407 -0400 > @@ -84,7 +84,11 @@ > if (!pATI->pBank) > return; > > +#ifndef XSERVER_LIBPCIACCESS > xf86UnMapVidMem(iScreen, pATI->pBank, 0x00010000U); > +#else > + (void) pci_device_unmap_legacy(pATI->PCIInfo, pATI->pBank, 0x00010000U); > +#endif > > pATI->pBank = NULL; > } > @@ -219,9 +223,14 @@ > * No relocation, resizing, caching or write-combining of this > * aperture is supported. Hence, the hard-coded values here... > */ > - pATI->pBank = xf86MapDomainMemory(iScreen, VIDMEM_MMIO_32BIT, > - Tag, 0x000A0000U, 0x00010000U); > - > +#ifndef XSERVER_LIBPCIACCESS > + pATI->pBank = xf86MapDomainMemory(iScreen, VIDMEM_MMIO_32BIT, > + Tag, 0x000A0000U, 0x00010000U); > +#else > + (void) pci_device_map_legacy(Tag, 0x000A0000U, 0x00010000U, > + PCI_DEV_MAP_FLAG_WRITABLE, > + &pATI->pBank); > +#endif > if (!pATI->pBank) > return FALSE; > > > -- Ionuț

1 0

[arch-general] package information error
by Auguste Pop 19 Mar '12

19 Mar '12

hi, somtimes the installed size information is simply wrong.e $ pacman -Si thunderbird Repository : extra Name : thunderbird Version : 11.0-1 ...... Download Size : 15409.89 KiB Installed Size : 7968.00 KiB Architecture : x86_64 $ xz -l thunderbird-11.0-1-x86_64.pkg.tar.xz Strms Blocks Compressed Uncompressed Ratio Check Filename 1 1 15.0 MiB 40.0 MiB 0.376 CRC64 thunderbird-11.0-1-x86_64.pkg.tar.xz $ mkdir thunderbird && tar xf thunderbird-11.0-1-x86_64.pkg.tar.xz -C thunderbird && du -sh thunderbird tar: Ignoring unknown extended header keyword `SCHILY.fflags' 41M thunderbird the installed size should be around 40M while pacman and the package website show only about 8M. best regards,

2 2

[arch-general] Erlang R15B
by Kevin Arthur 18 Mar '12

18 Mar '12

Hello! I noticed that Erlang recently got moved from extra to community, and was flagged as out-of-date. I think someone mentioned R15B, the latest release, on this list about a month ago, but it doesn't look like it went anywhere. I've attached a patch that will update the PKGBUILD to the latest version. The ejabberd package will need to be rebuilt, but doesn't require any other changes (I've tested it, and it's working fine). The couchdb and yaws packages will probably have to be rebuilt too, although they've both been marked out-of-date as well. Could one of the developers update the package? Thanks! Kevin Arthur

2 2

[arch-general] spamd failure
by pete 17 Mar '12

17 Mar '12

Hi just noticed that that spamd has failed reporting /usr/bin/perl: error while loading shared libraries: libperl.so: cannot open shared object file: No such file or directory but locate libperl.so gives /usr/lib/perl5/core_perl/CORE/libperl.so but there is no libperl.so in /usr/lib/perl5/core_perl/CORE/ this was working fine untill yesterdays pacman -Suy which completed without a reported problem Pete . -- Linux 7-of-9 3.2.9-1-ARCH #1 SMP PREEMPT Thu Mar 1 09:31:13 CET 2012 x86_64 AMD Phenom(tm) 9600B Quad-Core Processor AuthenticAMD GNU/Linux

4 6

[arch-general] About SHFS "Transport endpoint is not connected" error
by 郑文辉(Techlive Zheng) 17 Mar '12

17 Mar '12

I don't know if you guys ever encontered this problem while using sshfs. I set up a virtual machine to do some PKGBUILD makepking stuff, because some dependcies need bo be installed and I do not want messed up my host machine. So, I set up a sshfs filesystem connect from a virtualbox arch guest to my arch host. Every thing works fine, but each time I ran to following command, the sshfs just hangs up with the same error bellow. Then, I have to remount it again to get it working. command: find . -name Makefile -print | xargs sed -i "s:/usr/bin/install:/bin/install:g" error: sed: cannot remove ./maintscripts/sedP7sQnH: Transport endpoint is not connected Here is my sshfs mount command: sshfs -o sshfs_sync,no_readahead,sshfs_debug,port=64,idmap=user,reconnect,allow_other,workaround=all techlive@home:/home/abs /home/abs I have tested both following, they are all failed. sshfs -o port=64,idmap=user,reconnect,allow_other,workaround=all techlive@home:/home/abs /home/abs sshfs -o port=64,idmap=user,allow_other techlive@home:/home/abs /home/abs I have googled a lot, but none of them could help. So, What do you guys say, is there a known bug or something?

2 1

[arch-general] A Couple of Questions about files under etc which owned by no package
by 郑文辉(Techlive Zheng) 17 Mar '12

17 Mar '12

I ran the script below to see the files which owned by no package. export LC_ALL=C for dir in bin etc lib usr var sbin;do find /$dir -type f -exec pacman -Qo {} \; 2>$dir 1>/dev/null done I got the following. error: No package owns /etc/crontab error: No package owns /etc/localtime error: No package owns /etc/ssl/certs/ca-certificates.crt Are these files necessary now, could I delete them?

4 8

[arch-general] Firefox fonts are not anti-aliased after installing xorg-fonts-100dpi
by Ankur Sethi 16 Mar '12

16 Mar '12

Hi, I recently installed emacs-color-theme from AUR which, for some reason, required xorg-fonts-100dpi. After installing xorg-fonts-100dpi, some websites in Firefox no longer have antialiased fonts. For example, this is what the Mozilla Developer Center looks like on my computer: http://imagebin.org/203464 I see the same issue on Twitter, Hacker News and several blogs. Removing xorg-fonts-100dpi fixes the issue. I've checked my installed pacakges and I have all the web fonts that are recommended by the Arch Wiki. In any case, I don't think missing fonts are the issue here. The same websites that look wonky in Firefox work well on both Chrome and Opera. Secondly, the emacs-color-theme package from AUR works just fine without xorg-fonts-100dpi. I don't understand why that is listed as a dependency in the PKGBUILD. So can someone enlighten me? Why does emacs-color-theme require xorg-fonts-100dpi, and why does installing xorg-fonts-100dpi make Firefox sad? -- Ankur Sethi

2 2

[arch-general] X doesn't start for 3.0-rt, X 1.12, nv 2.1.18-5
by Ralf Mardorf 16 Mar '12

16 Mar '12

Perhaps the current X update has broken my install. At least in 2012-Feb-21 this kernel-rt worked with X and the nv driver. I tried to boot the kernel-rt to build a new module for VBox. # pacman -Qi linux-rt xorg-server xf86-video-nv | grep Version Version : 3.0.14_rt31-1 Today updated: Version : 1.12.0-1 Version : 2.1.18-5 This does work with X 1.12: # uname -r 3.2.9-1-ARCH # pacman -Qi nvidia | grep Version Version : 295.20-3 Anybody using current nv driver with kernel-rt 3.0 or current version 3.2.9_rt17-1 from AUR? -Ralf

3 5