From doug.chapman at hp.com Mon Jun 2 00:00:06 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Sun, 1 Jun 2008 18:00:06 -0600 (MDT) Subject: [Fedora-ia64-list] Fedora ia64 Buglist 2008-6-1 Message-ID: <20080602000006.237BC39C084@ldl.fc.hp.com> Bugs filed against fedora-ia64. If a you have filed a bug and it is not listed below, then it is likely you did not file it under the appropriate Product (Fedora) or Hardware Category (ia64). Total bugs in each state: NEEDINFO: 2 MODIFIED: 8 ASSIGNED: 5 CLOSED : 311 NEW : 24 ON_QA : 1 RECENT MOVERS (changed in past 7 days): 7 448327: [NEW] shutdown or restart 448323: [CLOSED] Anaconda bug report 446951: [NEW] Crash system-config-network under XEN 445642: [NEEDINFO] add ia64 support for livecd-tools 435742: [ON_QA] libelf-devel conflicts with itself 433670: [NEW] bootstrap ghc on ia64 392741: [CLOSED] Crash during software update (using gnome and not yum through terminal) NEW state: 24 448327: shutdown or restart 447975: 2.5GHz Q9300 CPU frequency defaults to 2.0 446992: libnss_wins.so segfaults when invoked 446951: Crash system-config-network under XEN 444867: Add Remove Programs has no categories 444821: unaligned access messages during kickstart install 441547: system-config-network ignores DNS settings of wlan interface (iwl3945) 433921: xen ia64 build fails with array subscript is above array bounds 433913: need to bootstrap plplot on ia64 433911: google-perftools needs stacktrace code for ia64 433910: eclipse-setools missing setools.linux.ia64/build.properties 433844: libEMF fails to build on ia64 433670: bootstrap ghc on ia64 426333: [kernel-headers] Compiles using the macro 'NBPG' fail as PAGE_SIZE is not defined in /usr/include/asm/page.h 388361: eog unable to display files having non-US-ASCII names 386151: Cannot create bridged network for paravirtualized guest. 360411: Install failed on 256p system 350971: Limit disk request ignored. 350941: Wrong entry in fstab 313441: system-config-network-tui crash 275131: Nvidia MCP55 eth0,eth1 not working after updating from initial install 251696: Atlas config bounds error leads to build failure 249254: USB hard drive not shown as hotpluggable 243185: Error in execution of RPM clean-up for librsvg2 From yi.zhan at intel.com Mon Jun 2 06:06:24 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Mon, 2 Jun 2008 14:06:24 +0800 Subject: [Fedora-ia64-list] Fedora ia64 packages status 2008-6-2 Message-ID: Fedora IA64 packages status, compared with f9 final on x86. Packages available: 5392 / 5551 (97.1%) Packages updated: 5334 / 5551 (96.1%) All failed builds: 217 (5 newly fixed) Arch excluded: 61 Build errors: 74 Dependency issues: 71 Others: 11 =============== Packages with build errors =============== jgroups crystal-stacker Glide3 python-tag njam xen straw rafkill monotone oggconvert zasx ocaml-cil callweaver scorched3d bmpx planets vdrift boswars ccrtp crystalspace libzzub GtkAda libEMF postgresql-plruby csound newsx Pixie upx tachyon cpdup warzone2100 gpa ocaml-camomile atlascpp ikarus vbetool kgtk google-perftools seamonkey tbb sipp gnuradio R-Matrix rekall QuantLib eclipse-setools gambas gridengine linkage guile frysk esc ekiga grub qca-gnupg ice sunbird gwget oorexx mono-debugger OpenSceneGraph clisp nx polyml gcl thunderbird ladspa-swh-plugins plt-scheme uim plotmm mosml openoffice.org pingus cln From doug.chapman at hp.com Mon Jun 9 00:00:04 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Sun, 8 Jun 2008 18:00:04 -0600 (MDT) Subject: [Fedora-ia64-list] Fedora ia64 Buglist 2008-6-8 Message-ID: <20080609000004.DA39239C00B@ldl.fc.hp.com> Bugs filed against fedora-ia64. If a you have filed a bug and it is not listed below, then it is likely you did not file it under the appropriate Product (Fedora) or Hardware Category (ia64). Total bugs in each state: NEEDINFO: 4 MODIFIED: 8 ASSIGNED: 5 CLOSED : 311 NEW : 22 ON_QA : 1 RECENT MOVERS (changed in past 7 days): 3 444867: [NEW] Add Remove Programs has no categories 350971: [NEEDINFO] Limit disk request ignored. 350941: [NEEDINFO] Wrong entry in fstab NEW state: 22 448327: shutdown or restart 447975: 2.5GHz Q9300 CPU frequency defaults to 2.0 446992: libnss_wins.so segfaults when invoked 446951: Crash system-config-network under XEN 444867: Add Remove Programs has no categories 444821: unaligned access messages during kickstart install 441547: system-config-network ignores DNS settings of wlan interface (iwl3945) 433921: xen ia64 build fails with array subscript is above array bounds 433913: need to bootstrap plplot on ia64 433911: google-perftools needs stacktrace code for ia64 433910: eclipse-setools missing setools.linux.ia64/build.properties 433844: libEMF fails to build on ia64 433670: bootstrap ghc on ia64 426333: [kernel-headers] Compiles using the macro 'NBPG' fail as PAGE_SIZE is not defined in /usr/include/asm/page.h 388361: eog unable to display files having non-US-ASCII names 386151: Cannot create bridged network for paravirtualized guest. 360411: Install failed on 256p system 313441: system-config-network-tui crash 275131: Nvidia MCP55 eth0,eth1 not working after updating from initial install 251696: Atlas config bounds error leads to build failure 249254: USB hard drive not shown as hotpluggable 243185: Error in execution of RPM clean-up for librsvg2 From yi.zhan at intel.com Mon Jun 9 14:45:28 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Mon, 9 Jun 2008 22:45:28 +0800 Subject: [Fedora-ia64-list] Too many subdirectories under /mnt/koji/work/tasks/ Message-ID: Hi Doug, I hit the error "exceptions.OSError: [Errno 31] Too many links: '/mnt/koji/work/tasks/75012'" hundreds times today when catching up with x86's dist-f10. See http://koji.fedoraproject.org/koji/taskinfo?taskID=418705, http://ia64.koji.fedoraproject.org/koji/taskinfo?taskID=75003. It seems the cause of this is that the number of subdirectories under /mnt/koji/work/tasks is approaching the limit (~32k). So the file system refused to create new directory under /mnt/koji/work/tasks and then we failed to launch any task on koji. I have created a new directory /mnt/koji/work/tasks.old and move the oldest ~2.4k directories there to work around this issue. The side effect is that logs of these old tasks are no longer available on the koji's web interface (e.g. http://ia64.koji.fedoraproject.org/koji/taskinfo?taskID=102) but seems it's same on x86 (e.g. http://koji.fedoraproject.org/koji/taskinfo?taskID=365824). Please let me know if there are better solutions for this. Thanks, Yi From yi.zhan at intel.com Mon Jun 9 18:03:03 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Tue, 10 Jun 2008 02:03:03 +0800 Subject: [Fedora-ia64-list] Fedora ia64 packages status 2008-6-9 Message-ID: Fedora IA64 packages status, compared with f9 final on x86. Packages available: 5394 / 5551 (97.2%) Packages updated: 5337 / 5551 (96.1%) All failed builds: 214 (3 newly fixed) Arch excluded: 61 Build errors: 74 Dependency issues: 68 Others: 11 =============== Packages with build errors =============== jgroups crystal-stacker Glide3 python-tag njam xen straw rafkill monotone oggconvert zasx ocaml-cil callweaver scorched3d bmpx planets vdrift boswars ccrtp crystalspace libzzub GtkAda libEMF postgresql-plruby csound newsx Pixie upx tachyon cpdup warzone2100 gpa ocaml-camomile atlascpp ikarus vbetool kgtk google-perftools seamonkey tbb sipp gnuradio R-Matrix rekall QuantLib eclipse-setools gambas gridengine linkage guile frysk esc ekiga grub qca-gnupg ice sunbird gwget oorexx mono-debugger OpenSceneGraph clisp nx polyml gcl thunderbird ladspa-swh-plugins plt-scheme uim plotmm mosml openoffice.org pingus cln From doug.chapman at hp.com Tue Jun 10 15:04:50 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Tue, 10 Jun 2008 11:04:50 -0400 Subject: [Fedora-ia64-list] Re: Too many subdirectories under /mnt/koji/work/tasks/ In-Reply-To: References: Message-ID: <1213110290.32662.7.camel@deimos.americas.hpqcorp.net> On Mon, 2008-06-09 at 22:45 +0800, Zhan, Yi wrote: > Hi Doug, > > I hit the error "exceptions.OSError: [Errno 31] Too many links: > '/mnt/koji/work/tasks/75012'" hundreds times today when catching up with > x86's dist-f10. See > http://koji.fedoraproject.org/koji/taskinfo?taskID=418705, > http://ia64.koji.fedoraproject.org/koji/taskinfo?taskID=75003. > > It seems the cause of this is that the number of subdirectories under > /mnt/koji/work/tasks is approaching the limit (~32k). So the file system > refused to create new directory under /mnt/koji/work/tasks and then we > failed to launch any task on koji. I have created a new directory > /mnt/koji/work/tasks.old and move the oldest ~2.4k directories there to > work around this issue. The side effect is that logs of these old tasks > are no longer available on the koji's web interface (e.g. > http://ia64.koji.fedoraproject.org/koji/taskinfo?taskID=102) but seems > it's same on x86 (e.g. > http://koji.fedoraproject.org/koji/taskinfo?taskID=365824). Please let > me know if there are better solutions for this. > > Thanks, > Yi > Yi, Yes, I think this is the right thing to do. On the primary arch's koji server they have a script that tries to clean up old tasks. I had not been concerned about this since we still have plenty of disk space. I did not think about the problem of having too many directories. - Doug From yanmin_zhang at linux.intel.com Wed Jun 11 00:58:03 2008 From: yanmin_zhang at linux.intel.com (Zhang, Yanmin) Date: Wed, 11 Jun 2008 08:58:03 +0800 Subject: [Fedora-ia64-list] Kernel directory change Message-ID: <1213145883.3054.8.camel@ymzhang> Yi, I installed FC9 Beta ia64 on my Montvale machine and found the kernel directory is changed. The old directory is /boot/efi/efi/redhat, but the new one is /boot/efi/EFI/redhat. Is there any special reason to do so? A couple of test suites assumes ?/boot/efi/efi/redhat. -yanmin From dchapman at redhat.com Wed Jun 11 13:03:42 2008 From: dchapman at redhat.com (Doug Chapman) Date: Wed, 11 Jun 2008 09:03:42 -0400 Subject: [Fedora-ia64-list] Kernel directory change In-Reply-To: <1213145883.3054.8.camel@ymzhang> References: <1213145883.3054.8.camel@ymzhang> Message-ID: <1213189422.3673.2.camel@centrino> On Wed, 2008-06-11 at 08:58 +0800, Zhang, Yanmin wrote: > Yi, > > I installed FC9 Beta ia64 on my Montvale machine and found the kernel > directory is changed. The old directory is /boot/efi/efi/redhat, but the > new one is /boot/efi/EFI/redhat. Is there any special reason to do so? > > A couple of test suites assumes ?/boot/efi/efi/redhat. > > -yanmin > This changed a while back. What it appears happened was that the vfat filesystem for some reason now understands upper/lower case. In actuality it was always /boot/efi/EFI but the linux implementation of vfat shifted the directory name to lower case. Note that under the EFI shell it was always an upper case EFI as that is what is actually on the filesystem. I would suggest modifying your test to handle either case. - Doug From yi.zhan at intel.com Wed Jun 11 14:08:17 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Wed, 11 Jun 2008 22:08:17 +0800 Subject: [Fedora-ia64-list] Kernel directory change In-Reply-To: <1213189422.3673.2.camel@centrino> References: <1213145883.3054.8.camel@ymzhang> <1213189422.3673.2.camel@centrino> Message-ID: > -----Original Message----- > From: Doug Chapman [mailto:dchapman at redhat.com] > Sent: Wednesday, June 11, 2008 9:04 PM > To: ia64 Fedora Core Development > Cc: Zhan at redhat.com; Zhan, Yi; Development discussions related to Fedora > Subject: Re: [Fedora-ia64-list] Kernel directory change > > On Wed, 2008-06-11 at 08:58 +0800, Zhang, Yanmin wrote: > > Yi, > > > > I installed FC9 Beta ia64 on my Montvale machine and found the kernel > > directory is changed. The old directory is /boot/efi/efi/redhat, but the > > new one is /boot/efi/EFI/redhat. Is there any special reason to do so? > > > > A couple of test suites assumes ?/boot/efi/efi/redhat. > > > > -yanmin > > > > This changed a while back. What it appears happened was that the vfat > filesystem for some reason now understands upper/lower case. In > actuality it was always /boot/efi/EFI but the linux implementation of > vfat shifted the directory name to lower case. > Yes the file system changing might be the point. There is a "%define image_install_path boot/efi/EFI/redhat" for ia64 in kernel.spec since fc4. So the kernel dir was always /boot/efi/EFI/redhat. Yi From yanmin_zhang at linux.intel.com Thu Jun 12 01:04:06 2008 From: yanmin_zhang at linux.intel.com (Zhang, Yanmin) Date: Thu, 12 Jun 2008 09:04:06 +0800 Subject: [Fedora-ia64-list] Kernel directory change In-Reply-To: <1213189422.3673.2.camel@centrino> References: <1213145883.3054.8.camel@ymzhang> <1213189422.3673.2.camel@centrino> Message-ID: <1213232646.3054.16.camel@ymzhang> On Wed, 2008-06-11 at 09:03 -0400, Doug Chapman wrote: > On Wed, 2008-06-11 at 08:58 +0800, Zhang, Yanmin wrote: > > Yi, > > > > I installed FC9 Beta ia64 on my Montvale machine and found the kernel > > directory is changed. The old directory is /boot/efi/efi/redhat, but the > > new one is /boot/efi/EFI/redhat. Is there any special reason to do so? > > > > A couple of test suites assumes ?/boot/efi/efi/redhat. > > > > -yanmin > > > > This changed a while back. What it appears happened was that the vfat > filesystem for some reason now understands upper/lower case. In > actuality it was always /boot/efi/EFI but the linux implementation of > vfat shifted the directory name to lower case. > > Note that under the EFI shell it was always an upper case EFI as that is > what is actually on the filesystem. I would suggest modifying your test > to handle either case. That's not a good idea. Although it's caused by vfat improvements, I would suggest to keep the old interface. Long long ago, there was a long discussion about interface change on LKML. It's a bad idea to change interfaces or API when they become de facto. From doug.chapman at hp.com Mon Jun 16 00:00:05 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Sun, 15 Jun 2008 18:00:05 -0600 (MDT) Subject: [Fedora-ia64-list] Fedora ia64 Buglist 2008-6-15 Message-ID: <20080616000005.EF3F739D019@ldl.fc.hp.com> Bugs filed against fedora-ia64. If a you have filed a bug and it is not listed below, then it is likely you did not file it under the appropriate Product (Fedora) or Hardware Category (ia64). Total bugs in each state: NEEDINFO: 4 MODIFIED: 8 ASSIGNED: 5 CLOSED : 312 NEW : 21 ON_QA : 1 RECENT MOVERS (changed in past 7 days): 2 446992: [CLOSED] libnss_wins.so segfaults when invoked 433921: [NEW] xen ia64 build fails with array subscript is above array bounds NEW state: 21 448327: shutdown or restart 447975: 2.5GHz Q9300 CPU frequency defaults to 2.0 446951: Crash system-config-network under XEN 444867: Add Remove Programs has no categories 444821: unaligned access messages during kickstart install 441547: system-config-network ignores DNS settings of wlan interface (iwl3945) 433921: xen ia64 build fails with array subscript is above array bounds 433913: need to bootstrap plplot on ia64 433911: google-perftools needs stacktrace code for ia64 433910: eclipse-setools missing setools.linux.ia64/build.properties 433844: libEMF fails to build on ia64 433670: bootstrap ghc on ia64 426333: [kernel-headers] Compiles using the macro 'NBPG' fail as PAGE_SIZE is not defined in /usr/include/asm/page.h 388361: eog unable to display files having non-US-ASCII names 386151: Cannot create bridged network for paravirtualized guest. 360411: Install failed on 256p system 313441: system-config-network-tui crash 275131: Nvidia MCP55 eth0,eth1 not working after updating from initial install 251696: Atlas config bounds error leads to build failure 249254: USB hard drive not shown as hotpluggable 243185: Error in execution of RPM clean-up for librsvg2 From yi.zhan at intel.com Sun Jun 22 16:34:34 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Mon, 23 Jun 2008 00:34:34 +0800 Subject: [Fedora-ia64-list] Fedora ia64 packages status 2008-6-22 Message-ID: The package building process has moved on to catch up with dist-f10 of primary arches. The f9 tree would not continue to take new changes other than critical fixes. Compared to f9 tree, we have some regression but not big. Many broken dependencies in f9 tree have been fixed and the rate of "Packages updated" would grow higher soon when the packages failed with dependencies issues are limit to just the ones who depend on arch excluded packages and build error packages. Packages available: 5569 / 5814 (97.1%) Packages updated: 5337 / 5551 (94.6%) All failed builds: 312 Arch excluded: 67 Build errors: 80 Dependency issues: 165 =============== Packages with build errors =============== jgroups newsx mesa gtk-sharp tachyon guile gstm google-perftools sudo airsnort ice snort ruby-bdb clisp OpenSceneGraph gdmap oorexx dietlibc pcmanx-gtk2 gcl sunbird straw mosml gnome-power-manager oggconvert plt-scheme libx86 sos Glide3 chess callweaver monotone crystalspace GtkAda libzzub gkrellm-wifi planets postgresql-plruby ikarus ccrtp Pixie gstreamer-plugins-good csound grub xen atlascpp gnuradio java-1.6.0-openjdk R-Matrix mono-debugger perl-Convert-Binary-C linkage ocaml-cil perl-Net-Netmask scalapack nx xbase esc ocaml-camomile uim gwget sipp hdf5 lush cln eclipse polyml seamonkey bmpx ladspa-swh-plugins thunderbird cluster plotmm atlas frysk libEMF gridengine pidgin-otr vdrift upx From doug.chapman at hp.com Mon Jun 23 00:00:06 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Sun, 22 Jun 2008 18:00:06 -0600 (MDT) Subject: [Fedora-ia64-list] Fedora ia64 Buglist 2008-6-22 Message-ID: <20080623000006.257DD39C043@ldl.fc.hp.com> Bugs filed against fedora-ia64. If a you have filed a bug and it is not listed below, then it is likely you did not file it under the appropriate Product (Fedora) or Hardware Category (ia64). Total bugs in each state: NEEDINFO: 4 MODIFIED: 4 ASSIGNED: 5 CLOSED : 324 NEW : 16 ON_QA : 1 RECENT MOVERS (changed in past 7 days): 17 452373: [CLOSED] kernels after 2.6.24 broke intel 82801h sound 451770: [NEW] Cannot create bridged network for paravirtualized guest. 451740: [NEW] system-config-services seems to fail to start all services in correct runlevel 442751: [CLOSED] anaconda does not create boot.iso on ia64 441846: [CLOSED] Unable to determine kernel type/version 441708: [CLOSED] mk-images sets KERNELDIR wrong on ia64 441470: [CLOSED] mk-images failed on ia64 439876: [MODIFIED] No kernel packages were installed on your system 435750: [MODIFIED] [PATCH] fix booty default file params on ia64 elilo 433913: [ASSIGNED] need to bootstrap plplot on ia64 388361: [CLOSED] eog unable to display files having non-US-ASCII names 386151: [CLOSED] Cannot create bridged network for paravirtualized guest. 313441: [CLOSED] system-config-network-tui crash 294381: [CLOSED] XGL(i810) + Xinerama = Crash 275131: [CLOSED] Nvidia MCP55 eth0,eth1 not working after updating from initial install 249254: [CLOSED] USB hard drive not shown as hotpluggable 243185: [CLOSED] Error in execution of RPM clean-up for librsvg2 NEW state: 16 451770: Cannot create bridged network for paravirtualized guest. 451740: system-config-services seems to fail to start all services in correct runlevel 448327: shutdown or restart 447975: 2.5GHz Q9300 CPU frequency defaults to 2.0 446951: Crash system-config-network under XEN 444867: Add Remove Programs has no categories 444821: unaligned access messages during kickstart install 441547: system-config-network ignores DNS settings of wlan interface (iwl3945) 433921: xen ia64 build fails with array subscript is above array bounds 433911: google-perftools needs stacktrace code for ia64 433910: eclipse-setools missing setools.linux.ia64/build.properties 433844: libEMF fails to build on ia64 433670: bootstrap ghc on ia64 426333: [kernel-headers] Compiles using the macro 'NBPG' fail as PAGE_SIZE is not defined in /usr/include/asm/page.h 360411: Install failed on 256p system 251696: Atlas config bounds error leads to build failure From yi.zhan at intel.com Tue Jun 24 16:08:02 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Wed, 25 Jun 2008 00:08:02 +0800 Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 Message-ID: All, David Woodhouse suggested that we should have a bug filed for each use of ExcludeArch and marked as blocking the corresponding FE-ExcludeArch-ia64 tracker bug. I think it's nice to have this for ia64. And we have filed a FE-ExcludeArch-ia64 for this: https://bugzilla.redhat.com/show_bug.cgi?id=FE-ExcludeArch-IA64. With this we can track why a package is arch excluded on ia64 and see if we can get it build or if it worth the effort. Also we could get a package arch excluded in the spec if needed, e.g. the ones do not build on ia64 and currently not arch excluded. Below is the description of FE-ExcludeArch-ia64. ======================================================================== ======= >From the package review guidelines: --- MUST: If the package does not successfully compile, build or work on an architecture, then those architectures should be listed in the spec in ExcludeArch. Each architecture listed in ExcludeArch needs to have a bug filed in bugzilla, describing the reason that the package does not compile/build/work on that architecture. The bug number should then be placed in a comment, next to the corresponding ExcludeArch line. New packages will not have bugzilla entries during the review process, so they should put this description in the comment until the package is approved, then file the bugzilla entry, and replace the long explanation with the bug number. --- This Bug is used as a tracker for such issues. If you added a bug for a package that uses ExcludeArch mark is as blocking this bug. If it's unlikely that the issue is ever fixed feel free to close the bug directy after reporting. Otherwise leave it open. ======================================================================== ======= Thanks, Yi From yi.zhan at intel.com Tue Jun 24 16:44:31 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Wed, 25 Jun 2008 00:44:31 +0800 Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 In-Reply-To: References: Message-ID: > From: fedora-ia64-list-bounces at redhat.com [mailto:fedora-ia64-list- > bounces at redhat.com] On Behalf Of Zhan, Yi > Sent: Wednesday, June 25, 2008 12:08 AM > To: ia64 Fedora Core Development > Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 > > All, > > David Woodhouse suggested that we should have a bug filed for each use > of ExcludeArch and marked as blocking the corresponding > FE-ExcludeArch-ia64 tracker bug. I think it's nice to have this for > ia64. And we have filed a FE-ExcludeArch-ia64 for this: > https://bugzilla.redhat.com/show_bug.cgi?id=FE-ExcludeArch-IA64. > I think it's useful to list ia64's currently arch excluded packages here. I will take a look at each of them and file BZs for them if applicable. For guys who want to help with this please pick the ones you interested in, check if there was a corresponding bug filed for it and then look into it. Any contribution is appreciated. Note, most packages below have explicit ExcludeArch & ExclusiveArch statement in their spec file but not all. We use "FAILED: BuildError: No matching arches were found" in the koji build output message to judge if a package currently does not build on ia64. So some of below packages may have explicit declaration saying do not support ia64 in the code but have no corresponding statement in the spec file. ==================== 67 arch excluded packages ==================== apmd openhpi apmud powerpc-utils athcool powerpc-utils-papr atitvout ppc64-utils camstream prelink cmucl ps3pf-utils cpufrequtils python-psyco cpufreq-utils qemu dev86 qpidc edac-utils rhm f-spot s390utils Glide3-libGL s3switch gnome-applet-vm sbcl gpart spicctrl gprolog svgalib gsynaptics synaptics HelixPlayer syslinux i8kutils system-config-boot ibmasm tpb iprutils tuxcmd kvm valgrind libhugetlbfs viaideinfo librtas wine lightning wxMaxima lrmi x86info maxima xenner mcelog xeuphoric memtest86+ xfce4-gsynaptics-mcs-plugin microcode_ctl xorg-x11-drv-geode mkbootdisk xorg-x11-drv-neomagic mknbi xorg-x11-drv-openchrome msr-tools xorg-x11-drv-vmmouse muine-scrobbler yaboot nspluginwrapper Thanks, Yi From yi.zhan at intel.com Wed Jun 25 12:06:00 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Wed, 25 Jun 2008 20:06:00 +0800 Subject: [Fedora-ia64-list] Fedora ia64 packages status 2008-6-22 In-Reply-To: References: Message-ID: > From: fedora-ia64-list-bounces at redhat.com [mailto:fedora-ia64-list- > bounces at redhat.com] On Behalf Of Zhan, Yi > Sent: Monday, June 23, 2008 12:35 AM > To: ia64 Fedora Core Development > Subject: [Fedora-ia64-list] Fedora ia64 packages status 2008-6-22 > > > The package building process has moved on to catch up with dist-f10 of > primary arches. The f9 tree would not continue to take new changes other > than critical fixes. Compared to f9 tree, we have some regression but > not big. Many broken dependencies in f9 tree have been fixed and the > rate of "Packages updated" would grow higher soon when the packages > failed with dependencies issues are limit to just the ones who depend on > arch excluded packages and build error packages. > > Packages available: 5569 / 5814 (97.1%) > Packages updated: 5337 / 5551 (94.6%) > Oops, sent wrong numbers. It should be: Packages available: 5643 / 5814 (97.1%) Packages updated: 5502 / 5814 (94.6%) From dchapman at redhat.com Wed Jun 25 15:10:48 2008 From: dchapman at redhat.com (Doug Chapman) Date: Wed, 25 Jun 2008 11:10:48 -0400 Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 In-Reply-To: References: Message-ID: <1214406648.25291.6.camel@deimos.americas.hpqcorp.net> On Wed, 2008-06-25 at 00:44 +0800, Zhan, Yi wrote: > > From: fedora-ia64-list-bounces at redhat.com [mailto:fedora-ia64-list- > > bounces at redhat.com] On Behalf Of Zhan, Yi > > Sent: Wednesday, June 25, 2008 12:08 AM > > To: ia64 Fedora Core Development > > Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 > > > > All, > > > > David Woodhouse suggested that we should have a bug filed for each use > > of ExcludeArch and marked as blocking the corresponding > > FE-ExcludeArch-ia64 tracker bug. I think it's nice to have this for > > ia64. And we have filed a FE-ExcludeArch-ia64 for this: > > https://bugzilla.redhat.com/show_bug.cgi?id=FE-ExcludeArch-IA64. > > > > I think it's useful to list ia64's currently arch excluded packages > here. I will take a look at each of them and file BZs for them if > applicable. For guys who want to help with this please pick the ones you > interested in, check if there was a corresponding bug filed for it and > then look into it. Any contribution is appreciated. > > Note, most packages below have explicit ExcludeArch & ExclusiveArch > statement in their spec file but not all. We use "FAILED: BuildError: No > matching arches were found" in the koji build output message to judge if > a package currently does not build on ia64. So some of below packages > may have explicit declaration saying do not support ia64 in the code but > have no corresponding statement in the spec file. > > ==================== 67 arch excluded packages ==================== > apmd openhpi > apmud powerpc-utils > athcool powerpc-utils-papr > atitvout ppc64-utils > camstream prelink > cmucl ps3pf-utils > cpufrequtils python-psyco > cpufreq-utils qemu > dev86 qpidc > edac-utils rhm > f-spot s390utils > Glide3-libGL s3switch > gnome-applet-vm sbcl > gpart spicctrl > gprolog svgalib > gsynaptics synaptics > HelixPlayer syslinux > i8kutils system-config-boot > ibmasm tpb > iprutils tuxcmd > kvm valgrind > libhugetlbfs viaideinfo > librtas wine > lightning wxMaxima > lrmi x86info > maxima xenner > mcelog xeuphoric > memtest86+ xfce4-gsynaptics-mcs-plugin > microcode_ctl xorg-x11-drv-geode > mkbootdisk xorg-x11-drv-neomagic > mknbi xorg-x11-drv-openchrome > msr-tools xorg-x11-drv-vmmouse > muine-scrobbler yaboot > nspluginwrapper > > Thanks, > Yi > I just talked with Alex Williamson about KVM. According to him the ia64 bits should be upstream and working so this one should be able to be fixed. I will file this one and investigate it. The system-config-boot package probably should be addressed. The only reason it doesn't get built on ia64 is because we use elilo instead of grub however that will change eventually since grub should be able to boot from efi soon. A few others on the list seem strange. For example s390utils shouldn't ExcludeArch anything, it should be ExlusiveArch (I think that is the keyword at least) since it ONLY makes sense on s390. I don't think it makes sense to file BZ's on things like that. However if it excludes arches when it should just include s390 then I agree that is a bug. Same goes for the powerpc* and ppc* packages. - Doug From yi.zhan at intel.com Thu Jun 26 01:43:59 2008 From: yi.zhan at intel.com (Zhan, Yi) Date: Thu, 26 Jun 2008 09:43:59 +0800 Subject: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 In-Reply-To: <1214406648.25291.6.camel@deimos.americas.hpqcorp.net> References: <1214406648.25291.6.camel@deimos.americas.hpqcorp.net> Message-ID: > From: fedora-ia64-list-bounces at redhat.com [mailto:fedora-ia64-list- > bounces at redhat.com] On Behalf Of Doug Chapman > Sent: Wednesday, June 25, 2008 11:11 PM > To: ia64 Fedora Core Development > Subject: RE: [Fedora-ia64-list] ExcludeArch tracker bug for IA64 > > I just talked with Alex Williamson about KVM. According to him the ia64 > bits should be upstream and working so this one should be able to be > fixed. I will file this one and investigate it. > > The system-config-boot package probably should be addressed. The only > reason it doesn't get built on ia64 is because we use elilo instead of > grub however that will change eventually since grub should be able to > boot from efi soon. > > A few others on the list seem strange. For example s390utils shouldn't > ExcludeArch anything, it should be ExlusiveArch (I think that is the > keyword at least) since it ONLY makes sense on s390. I don't think it > makes sense to file BZ's on things like that. However if it excludes > arches when it should just include s390 then I agree that is a bug. > > Same goes for the powerpc* and ppc* packages. > Yes the packages which have "ExclusiveArch 'some arches other than ia64'" are in above list. Originally the list is just used to indicate our catch up script that these packages do not build on ia64 currently in order to not launch koji build for them and fails over and over again. So the ExcludeArch & ExclusiveArch cases are both went into the same list. I agree that we should not file BZs for the ExclusiveArch ones. Yi From doug.chapman at hp.com Mon Jun 30 00:00:06 2008 From: doug.chapman at hp.com (Doug Chapman) Date: Sun, 29 Jun 2008 18:00:06 -0600 (MDT) Subject: [Fedora-ia64-list] Fedora ia64 Buglist 2008-6-29 Message-ID: <20080630000006.47EED39C003@ldl.fc.hp.com> Bugs filed against fedora-ia64. If a you have filed a bug and it is not listed below, then it is likely you did not file it under the appropriate Product (Fedora) or Hardware Category (ia64). Total bugs in each state: NEEDINFO: 4 MODIFIED: 4 ASSIGNED: 7 CLOSED : 323 NEW : 9 ON_QA : 1 RECENT MOVERS (changed in past 7 days): 3 452637: [ASSIGNED] ExcludeArch tracker for IA64 452373: [ASSIGNED] kernels after 2.6.24 broke intel 82801h sound 163350: [ASSIGNED] Tracking bug for Fedora ia64/Itanium support NEW state: 9 444821: unaligned access messages during kickstart install 433921: xen ia64 build fails with array subscript is above array bounds 433911: google-perftools needs stacktrace code for ia64 433910: eclipse-setools missing setools.linux.ia64/build.properties 433844: libEMF fails to build on ia64 433670: bootstrap ghc on ia64 426333: [kernel-headers] Compiles using the macro 'NBPG' fail as PAGE_SIZE is not defined in /usr/include/asm/page.h 360411: Install failed on 256p system 251696: Atlas config bounds error leads to build failure