openafs-1.6-sl-module-tools-0:1.6.10-0.152.pre1.el7> H HtxHFT ?*}}&982` nE l7f#p:52ba41090023dee115dea0b386328a75c8916d934&s=E¯ ?FT ?*}}xFB a-'LJ$,5Wm9!ω%l>9H?8d%4 8 Y     A DHMRx|=(Z8d9:%FGHIXY\]^b#defltuvwxy z$4Copenafs-1.6-sl-module-tools1.6.100.152.pre1.el7OpenAFS analogon to weak-modulesThe AFS distributed filesystem. AFS is a distributed filesystem allowing cross-platform sharing of files among multiple computers. Facilities are provided for access control, authentication, backup and administrative management. This package provides scripts for making the kernel module availabe to compatible kernels.Tcompose-7.fnal.gov$dScientific LinuxGPLv2+Networking/Filesystemslinuxx86_64$dT20e91b69ef6f9f037cd5f3de5ebed674de8497ce53e1d3fcdfe58358b50afac70rootrootopenafs.SLx-1.6.10-0.152.pre1.src.rpmopenafs-1.6-sl-module-toolsopenafs-1.6-sl-module-tools(x86-64)@    /bin/bashrpmlib(CompressedFileNames)rpmlib(FileDigests)rpmlib(PayloadFilesHavePrefix)rpmlib(PayloadIsXz)3.0.4-14.6.0-14.0-15.2-14.11.1S<@S@S@SG@S=M@RR;QQ,Q,Q+R@Q*Q']Q@Q @Q @QsPPP@P@P@PCPO@PP @Pf@P@P @PPPP@PPPPBOxC@OTOKp@O!@@N@NJ[N?N*M@M@MM[@M6@M$]@M hM@ME@M\@L,@L|LmK@K@K#@K\6@K\6@KZKRK@J@JrI@@IR$@I&I[HH @HB@H,%GRG^@G^@G G.@GGnMG<4G,bG@GFF@F@FF} @FZ@F8}@FFoE5@E@E@E@Ez@E(ETE%E^E[@EW E~E@E 8EDr@D7DDDm@DLCiCBnB=BwgBv@Bss@B@B$Y@B!@B dBAAKAA@*@@@@@r@8?W)@?U?O@@?F?Cb?<@?;y?7@?3Stephan Wiesand 1.6.10-152.SLxStephan Wiesand 1.6.9-151.SLxStephan Wiesand 1.6.9-150.SLxStephan Wiesand 1.6.5.1-149.SLxStephan Wiesand 1.6.5.1-148.SLxStephan Wiesand 1.6.5.1-147.SLxStephan Wiesand 1.6.5.1-146.SLxStephan Wiesand 1.6.2-145.SLxStephan Wiesand 1.6.2-0.144.SLxStephan Wiesand 1.6.2-0.143.SLxStephan Wiesand 1.6.2-0.142.SLxStephan Wiesand 1.6.2-0.141.SLxStephan Wiesand 1.6.2-0.140.SLxStephan Wiesand 1.6.2-0.139.SLxStephan Wiesand 1.6.2-0.138.1.pre4.SLxStephan Wiesand 1.6.2-0.137.1.pre4.SLxStephan Wiesand 1.6.2-0.136.1.pre4.SLxStephan Wiesand 1.6.2-0.135.1.pre3.SLxStephan Wiesand 1.6.2-0.134.1.pre3.SLxStephan Wiesand 1.6.2-0.134.pre3.SLxStephan Wiesand 1.6.2-0.133.pre3.SLxStephan Wiesand 1.6.2-0.132.pre2.SLxStephan Wiesand 1.6.2-0.131.pre2.SLxStephan Wiesand 1.6.2-0.129.pre2.SLxStephan Wiesand 1.6.2-0.128.pre2.SLxStephan Wiesand 1.6.2-0.127.pre2.SLxStephan Wiesand 1.6.2-0.126.pre1.SLxStephan Wiesand 1.6.2-0.124.pre1.SLxStephan Wiesand 1.6.2-0.123.pre1.SLxStephan Wiesand 1.6.2-0.122.pre0.SLxStephan Wiesand 1.6.2-0.120.pre0.SLxStephan Wiesand 1.6.2-0.119.pre0.SLxStephan Wiesand 1.6.2-0.118.pre0.SLxStephan Wiesand 1.6.2-0.117.pre0.SLxStephan Wiesand 1.6.2-0.116.pre0.SLxStephan Wiesand 1.6.2-0.114.pre0.SLxStephan Wiesand 1.6.1-112.SLxStephan Wiesand 1.6.1-110.pre4.SLxStephan Wiesand 1.6.1-109.pre3.SLxStephan Wiesand 1.6.1-99.pre2.SLxStephan Wiesand 1.6.1-98.pre1.SLxStephan Wiesand 1.6.0-97.SLxStephan Wiesand 1.6.0-96.pre7.SLxStephan Wiesand 1.6.0-95.pre6.SLxStephan Wiesand 1.6.0-94.pre6.SLxStephan Wiesand 1.6.0-93.pre4.SLxStephan Wiesand 1.6.0-92.pre3.SLxStephan Wiesand 1.6.0-91.pre2.SLxStephan Wiesand 1.6.0-90.pre1.SLxStephan Wiesand 1.6.0-89.pre1.SLxStephan Wiesand 1.6.0pre1-88.SLxTroy Dawson 1.5.78-87.SLxStephan Wiesand 1.5.78-86.SLxStephan Wiesand 1.5.78-84.SLxStephan Wiesand 1.5.78-83.SLxStephan Wiesand 1.5.77-82.SLxStephan Wiesand 1.5.76-81.SLxStephan Wiesand 1.5.74-80.z2.SLxStephan Wiesand 1.5.74-80.z1.SLxStephan Wiesand 1.5.74-79.SLxStephan Wiesand 1.5.69-78.z0.SLxStephan Wiesand 1.4.12-77.pre1.z2.SLxStephan Wiesand 1.4.12-77.pre1.z1.SLxStephan Wiesand 1.4.12-77.pre1.SLxStephan Wiesand 1.4.11-76.1.SLxStephan Wiesand 1.4.11-76.SLxStephan Wiesand 1.4.11-75.SLxStephan Wiesand 1.4.10-74.SLxStephan Wiesand 1.4.8-73.SLxStephan Wiesand 1.4.8-71.SLxStephan Wiesand 1.4.8-0.70.pre3.SLxStephan Wiesand 1.4.8-0.69.pre2.SLxStephan Wiesand 1.4.8-0.69.pre1.SLxStephan Wiesand 1.4.7-68.SLxStephan Wiesand 1.4.7-67.SLxStephan Wiesand 1.4.7-0.65.pre3.SLxStephan Wiesand 1.4.7-0.63.pre3.SLxStephan Wiesand 1.4.7-0.62.pre2.SLxStephan Wiesand 1.4.7-0.61.pre2.SLxStephan Wiesand 1.4.6-60.SLxStephan Wiesand 1.4.6-59.SLxStephan Wiesand 1.4.6-58.SLxStephan Wiesand 1.4.5-56.SLxStephan Wiesand 1.4.5-55.SLxStephan Wiesand 1.4.5-0.54.pre4.SLxStephan Wiesand 1.4.5-0.53.pre3.SLxStephan Wiesand 1.4.4-52.SLxStephan Wiesand 1.4.4-51.SLxStephan Wiesand 1.4.4-50.SLxStephan Wiesand 1.4.4-49.SLxStephan Wiesand 1.4.4-47.SLxStephan Wiesand 1.4.4-46.SLxStephan Wiesand 1.4.4-44.SLxStephan Wiesand 1.4.4-43.SLxStephan Wiesand 1.4.4-42.SLxStephan Wiesand 1.4.3-0.rc3.41.SLxStephan Wiesand 1.4.3-0.rc3.39.SLxStephan Wiesand 1.4.3-0.rc3.38.SLxStephan Wiesand 1.4.3-0.rc2.37.SLxStephan Wiesand 1.4.3-0.rc2.36.SLxStephan Wiesand 1.4.2-35.SLxStephan Wiesand 1.4.2-33.SLxStephan Wiesand 1.4.2-32.SLxStephan Wiesand 1.4.2-31.SLxStephan Wiesand 1.4.2-30.SLxStephan Wiesand 1.4.2-29.SLxStephan Wiesand 1.4.2-0.fc4.18.SL5Stephan Wiesand 1.4.2-0.fc3.17.SL5Stephan Wiesand 1.4.2-0.fc3.16.SL5Stephan Wiesand 1.4.2-0.fc3.15.SL5Stephan Wiesand 1.4.2-0.fc2.15.SL5Stephan Wiesand 1.4.2-0.rc1.14.SL5Stephan Wiesand 1.4.2-0.rc1.12.SL5Stephan Wiesand 1.4.2-0.beta3.11.SL5Stephan Wiesand 1.4.0-11.SLStephan Wiesand 1.4.0-10.SLStephan Wiesand 1.4.0-8.SLStephan Wiesand 1.4.0-0rc6.8.SLStephan Wiesand 1.3.82-5.SLStephan Wiesand 1.3.82-3.SLStephan Wiesand 1.3.82-2.SLStephan Wiesand 1.3.82-1.SLStephan Wiesand 1.3.80-1.SLStephan Wiesand 1.3.79-4.SLStephan Wiesand 1.3.79-3.SLJaroslaw Polok Jaroslaw Polok Troy Dawson 15.12.SLTroy Dawson 15.11.SLEnrico M.V. Fasanelli 15.10.SLTroy Dawson 15.7.SLTroy Dawson 15.6.SLTroy Dawson 15.4.SLTroy Dawson 15.3.SLJaroslaw.Polok@cern.chJaroslaw.Polok@cern.chJaroslaw.Polok@cern.chJaroslaw.Polok@cern.chDavid Howells > 1.2.10-4David Howells 1.2.10-4> 1.2.10-3David Howells 1.2.10-3Nalin Dahyabhai > 1.2.10-3Nalin Dahyabhai > 1.2.10-2> 1.2.10-1- updated to 1.6.10-pre1 - package the new volscan executable + manpage, in -server - don't package the xfs_size_check manpage - buildrequire systemd on SL7 (replaces systemd-units) - client/server: require systemd in post, preun and postun on SL7 - client/server: run systemd_ macros in post, preun and postun on SL7 - client/server: prepended a . to the generated sysconfig file names on SL7 - client/server: ship preset files on SL7 rather than disabling in %post - simply install the init scripts as usual on SL7, no need to move/rename them - support "--with modsign" to sign the kernel module (works on SL7+ only)- allow building with a "name suffix" - defaults to "" on SL5/6 and "-1.6-sl" on SL7 - can be overridden on the command line (nsfx macro)- update to release 1.6.9 - 1st stab at support for SL7 - ship systemd unit files on this platform, rather than init scripts - the init scripts are then installed as /sbin/openafs-{client,server}-init.sh - they are called in "prepare" mode from the unit files - in this mode, they do the usual checks/autoconfig, ... - and then they write the environment files used in the unit files - added BOSSERVER_OPTIONS variable to sysconfig (and server startup files) - added ENABLE_DYNROOT_SPARSE to sysconfig (and client init script) - defaults to on on SL7, off on SL5/6 - patches 1080..83 are already in - use the right path to depmod, and generally the right sbin- modify kmods to clean up links when updating to build against other kernel- patches 1081/2/3 with security content from openafs-1.6.7- patch1080 with http://gerrit.openafs.org/10578 to fix RHBZ #1038315 on EL 4.6 - fix build on EL 5.10- 1.6.5.1 bugfix release - modify buildroot to use mktemp, allowing simultaneous builds (Pat) - fix the AFS mountpoint in the cacheinfo file generated in the SPEC- 1.6.5 security release addressing openafs-sa-2013-0003 and -0004 - see SECURITY.txt for important info on how to actually secure your site - just updating to this build is *not* sufficient- do some more work in modules script - print debug output from kmod scripts if variable "verbose" is set- several fixes in modules script; does less unnecessary work- fixed location of modules script- (untested, unfinished) - use our own script for dealing with "weak updates", don't patch weak-modules - use different paths, don't interfere with weak-modules - new subpackage for the script: module-tools (license), require from kmods - script is run from scripts and triggers in the kmods - no longer provide kabi-mods - added a purgecache action to the client init script (Pat Riehecky)- the corrected actual, still "inofficial" 1.6.2 release- the actual, still "inofficial" 1.6.2 release- the actual pre4, downloaded from grand.central.org, + bos.patch - adapted client init script to changed ifconfig output on F18 (->EL7)- update to current head of 1.6.x (a15c406) + gerrit 9895, 8750 + bos patch - updated CellServDB to latest from grand.central.org (Jan 28, 2013)- pseudo pre4 - has gerrit 8912, 8937, 8940, 8939, 8945, + bos.patch - this could be 1.6.2 final - require /sbin/modprobe, not modutils (SL7/FC18)- made it build and install on SL7 (and recognize F18 as that)- added bos.patch- this is the actual pre3 uploaded to grand.central.org - and downloaded from there to another system- current head of 1_6_x (fa0c841d) + changes in the pipeline: - 8866..7, 8869, 8872, 8889..90, 8896..97 - + "make pre3" not yet on gerrit- this is the actual pre2 as uploaded to grand.central.org- current head of 1_6_x (5cfb720) + changes in the pipeline: - 8775..6, 8791, 8795..6, 8798, 8804..5, 8799..800, 8808..11 - + "make pre2" not yet on gerrit - renamed kmod packages from openafs_... to openafs-... - removed unused patches - client init script changes to make updates from 1.4 smoother: - on start, check for the libafs module as well, bail out if loaded - on stop, check for a loaded libafs module, remove it if loaded- pre1 + changes in the pipeline: - 8775/6, 8785, 8779/80/81, 8786..91, 8795, 8796- pre1 + EL5 build fixes (f87d49c/8021, 74c1881/8019, 5daa08e/8006)- pre1 + changes foreseen for pre2 (8753, 8757, 8758, 8761)- build kmod packages for each minor SL release kernel series - patch weak-modules so it won't link our modules across minor release kernels - use the doc macro to package documentation files - added several files from the doc tarball to documentation files - mark manpages as documentation files- these are the tarballs uploaded to grand.central.org, still w/o tag - add a sanity check for the module to load in the client init script- first, still "inofficial" pre1- build and package rxperf and th_rxperf- latest pre1 candidate: current head (fb95823) + gerrit 8421, 8423 (rxperf stuff)- latest pre1 candidate: current head (75af57b) + gerrit 8609, 8563, 8564, 8715- includes the latest patch sets for gerrit 8548, 8604, 8605, 8696, 8608 - also includes cherry-picked 8617 from master, supposed to fix cache-bypass- this one is just the current HEAD of openafs-stable_1_6_x (e1baf1af4)- next 1.6.2 prerelease lookalike - needs patch1078 to fix gerrit 8608/1 which is currently broken- 1st attempt to build a 1.6.2 prerelease lookalike - needs patch1077 beacuse erroneously version one of gerrit 8549 was applied- update to 1.6.1 stable release, no patches- update to prerelease 4- update to prerelease 3 (-100..-108 were builds with debug patches)- update to prerelease 2 - no changelog, no release notes- update to prerelease 1 for 1.6.1 release - there are no release notes yet - package new library libafscp.a in -devel - the old patch 302 (afsdump_scan) is finally upstream- update to final 1.6 release - updated CellServDB to latest from grand.central.org (Aug 14, 2011)- update to next prerelease, last patch is already in- patch1076: attempt an RX fix (backport of http://gerrit.openafs.org/4837)- update to next prerelease (pre5 had a serious bug and was not released)- update to next prerelease- update to next prerelease - init scripts: fixed condrestart- update to next prerelease - package /afs as ghost, and create it in post again - and fix the label- let openafs-client own the /afs mountpoint to avoid selinux troubles - change initdir macro to /etc/rc.d/init.d (like upstream), and use it everywhere- fix version/release for prerelease to allow updating to final later- update to first 1.6 prerelease- added i386 case version for sysname incase arch shows i386 despite a target=i686- don't strip any binaries, let rpm do the job - new in openafs-debug: cacheout, afscp, fsx - new in openafs-client: afsio - moved into main package: restorevol - moved into openafs-server: ka-forwarder + manpage - server package should own /usr/afs/etc - moved into openafs-authlibs: libkopenafs.so.1[.1] - moved into openafs-authlibs-devel: libkopenafs.{a,so}, kopenafs.h, libafs{authent,rpc}.a - fixed chcon in server's post script for SL6 - introduced the ability to build a kabi-tracking kmod (define build_kmod 1) - introduced a switch to disable building the fuse client - and unless that switch is used, buildrequire fuse-devel and package afsd.fuse - other minor changes to make rpmlint less unhappy - renamed openafs-debug to openafs-plumbing-tools - moved fssync-debug and manpages into -plumbing-tools - enabled debugsyms by default again, no effect on kernel module- removed support for SL3/4, greatly simplifying the spec - removed support for ia32e, ia64, athlon, ... - removed all old patches and obsolete sources (like openafs-krb5.tar.gz) - relaxed requirements between userland and kernel-module package - updated Changelog & Relnotes, and made sure we do so in the future - don't package the afs_compile_et manpage in the main packages as well - don't add -fPIC to CFLAGS (be compatible with openafs.org spec) - configure: no --enable-debug[-kernel] by default (be compatible with openafs.org spec) - configure: added --disable-strip-binaries (be compatible with openafs.org spec) - configure: added --with-linux-kernel-packaging (be compatible with openafs.org spec) - this also renames the module to openafs.ko, so change in rc script - configure: added --disable-kernel-module to non-module builds - and moved the build requirement on kernel-devel to the module package - the server package shouldn't require the kernel module - debugspec can now be defined on the command line, and defaults to 0 - moved much of the kernel/kvers/****dir voodoo into openafs-sl-defs.sh - which also replaces openas-lslrelease-helper.sh (and no longer stes SL5 on SL6) - updated CellServDB to latest from grand.central.org (Dec 13, 2010)- update to latest feature release - enable dugging symbols by default - there is no more /usr/include/potpourri.h - module is called ...mp.ko again?- update to latest feature release- update to latest feature release - patch1075 should no longer be required - libkopenafs is now 1.1 - DAFS is built by default - added files to openafs-server - added the new krb.excl(5) manpage to openafs-server- make this work with kernel releases prefixed with ".z[0-9]"- more SL6 changes: - use afsvers macro instead of PACKAGE_VERSION which is no longer available - patch1075 to fix kernel panic on module load - and: buildrequire make...- 1.5.74 release - F10 -> SL6 - changed how to determine the domain (hostname -d/-f may not work) - restorevol moved from sbin to bin- first stab at 1.5- patch1074 test fix, backed out 1073 for now- patch1073 test fix- prerelease 1 for next upstream release - patches 1065..1069 are already in - added patches since prerelease from stable-1_4_x git branch - updated CellServDB from grand.central.org (dementia.org entry changed) - changed CellservDB entry for phy.bris.ac.uk (mail from Winnie Lasseco) - package afs_compile_et + manpage (in -devel)- handle current lustre kernels (packaged a bit more like EL5)- added more post-1.4.11 fixes from stable-1_4_x git branch: 1068: fix-linux24-builds-20090718 1069: kernel-init-vrequest-structure-20090914- upstream release 1.4.11 - updated CellServDB from grand.central.org - install release notes and changelog into /usr/share/doc/openafs- - explicitly build src/auth/setkey (for -debug package) - do not install the new compile_et manpage, it clashes with e2fsprogs-devel - do not install the new rmtsysd manpage, this doesn't exist for linux yet - copyauth is no longer built; delete the manpage as well - package new manpages for restorevol, vsys - added post-1.4.11 fixes from stable-1_4_x git branch: 1066: export-ktc-curpag-everywhere-20090713 1067: fix-getvolume-for-nonroot-dynroot-fids-20090722- upstream release 1.4.10 - removed patches 1062..65 - updated CellServDB from grand.central.org; remove wallace from RAL entry- make this fit for F10- upstream release 1.4.8 - post release patches: 1062: viced-helper-thread-count-20081111 (1-liner to stay within 128 threads) 1063..65: fileserver man page updates- prerelease 3 - package a link for k5log.1.gz, and package both in openafs-krb5 not openafs- simply prerelease 2- patches 1058..1061 are in pre1 - package new manpages for krb.conf and vldb_convert (in -server) - upstream now installs the K5 klog as klog.krb5; package, + a symlink to k5log - should no longer need 1057- replace 1056 by 1057: linux-hlist-unhashed-opencoding-20080520 - more post-1.4.7 patches (either trivial, or serious and simple): 1058: butc-xbsa-lwp-protoize-damage-20080501 1059: uuid-corrected-duplicate-check-20080501 1060: viced-large-more-threads-20080506 1061: vos-sync-flag-voltype-properly-20080521- 1.4.7 release - should not require patch 1056 - but still does :-(- 1056: kernel 2.4 has no hlist_unhashed/insert_inode_hash- next prerelease - patches 1052..1054 are in pre3- actually apply 1052..1054, stupid- first stab at 1.4.7 prerelease - had to change twiddle patch (301 -> 303) (one of two typos is now fixed) - 1051 is already in now - parallel make is hopelessly broken now; remove retry kludges and don't try - remove useless symlink manpages - package the new read_tape manpage - temp. patches from cvs (should be in pre3): 1052: linux-flush-unlock-20080402 1053: linux-flush-compare-20080402 1054: vlserver-checksignal-returns-voidstar-20080401- when building for lustre kernels, fetch Module.symvers- fix build requirement for lustre case- new upstream release; fixes fileserver bugs (OPENAFS-SA-2007-003, DOS) - this obsoletes patches 1052,54,55; - also removed linux-nsec-timestamp-zero (1053) for the time being - renamed klog.krb5 to k5log - => this is a build of pristine openafs-1.4.6, except for the extra k5log - largefile-fileserver is the default since 1.4.5 -> adapt configure options - added "--with debugsyms" build option- first patches from cvs 1051: krb5-klog-20071101 1052: cbd-use-callback-size-for-callbacks-20071105 1053: linux-nsec-timestamp-zero-20071106 1054: viced-missing-lock-20071109 1055: viced-accurately-track-file-callbacks-20071112 - krb5 capable klog packaged as klog.krb5 in openafs-krb5- upstream release 1.4.5 - backout patch 1050, it's not working right => unmodified release- prerelease 4 - fix missing percent sign in is_lustre macro- prerelease 3 of next upstream version, obsoletes patches 1018..1049 - new patch 1050: oldfs-2gb (see RT #73720, but this one also works on amd64) - new man pages for CellAlias and pt_util - there is no more kseal executable - ka-forwarder now comes with upstream, and with a man page - => we no longer need afs-krb5 at all; this also obsoletes patches 103, 104, 204 - removed unused SuidCells - use man-pages from release, meanwhile they are better than those from #19268 - make sure man pages are packaged once only - support for building against lustre-kernel-source on SL4/5- more patches from CVS: 1044: viced-assert-less-20070719 1045: viced-multi-probe-addr-20070808 1046: find-dcache-just-hold-the-lock-slightly-longer-20070820 1047: linux-modparam-269-updates-20070821 1048: linux-module-error-handling-20070821-modified 1049: viced-remove-asserts-20070821- SL5: actually add the improved init script... - more patches from CVS: 1037: bucoord-dump-levels-20070517.patch 1038: budb-ol_verify-20070706.patch 1039: clone-dont-hold-vol-lock-20070516.patch 1040: des-p-temp-volatile-20070703.patch 1041: memcache-alloc-failures-20070623.patch 1042: pagsh-krb5-20070710.patch 1043: tvolser-locking-updates-20070322.patch- SL5: improved afs-server init script; also set it to unconfined_exec_t in postinstall if selinux active - updated CellServDB to current version from grand.central.org - more patches from CVS: 1033: volser-delete-clone-not-source-20070702 1034: avoid-empty-uuid-20070618 1035: forceallnewconns-prototype-20070627 1036: volser-earlier-logging-20070627- SL3: added corrected patch1032 from CVS to find 64bit syscall table on amd64 (openafs-1.4.4-amd64-linux-24-syscall-probe-symbol-change-20070623-fixed)- SL5: fix /etc/init.d/afs-server (missing fi in stop code) - more patches from CVS: - 1029: linux-kmem-destroy-fix-20070609 - 1030: vsprocs-avoid-bogus-error-20070606 - 1031: sl5-vfs-flush-3arg-20070612 (modified patch w/o ifdefs since that would require running regen.sh - which doesn't work on SL5)- added recent patches from CVS: - 1022: namei-dont-randomly-full-salvage-20070514 - 1023: vos-namelen-for-dump-restore-is-stupid-20070509 - 1024: supergroup-cleanup-20070516 - 1025: ptserver-fix-bitmap-20070516 - 1026: glibc24-jmp-buf-mangling-20070516 (SL5 only) - 1027: rx-call-abort-release-refcount-20070425 - 1028: dont-fclose-null-20070514- SL4/5 init script: fix cache partition detection on SL4/5 - SL4: replace patch1020 by slightly fixed version now in CVS - add accompanying patch1021 (vmalloc-no-glock-20070418)- SL4: added patch1020 by Chris Wing to turn off broken symlink caching, see https://lists.openafs.org/pipermail/openafs-devel/2007-March/015069.html - add buildhost and version-release to AFS version string- upstream release with new default of -nosuid for local cell, see OPENAFS-SA-2007-001 * this may break things for sites relying on setuid files in AFS * - added a word of warning to the SuidCells config file - updated CellServDB to version from grand.central.org as of 2007-03-21 - added linux-task-pointer-safety-20070320 from CVS (see RT #55973)- added support for xenU kernel from xensource - fixed release detection on plain RHEL5- added patch to make sure afs_pag_destroy() holds the glock, from https://lists.openafs.org/pipermail/openafs-devel/2007-March/014985.html - for SL5, changed comments on afsd options in the sysconfig file - for SL5, changed default afsd options to " " (as recommended by developers)- upstream rc3, could remove all recent patches - support new xenU kernel variant on SL4 - automated macro for rc sources - put back kernel module dependencies (except if built --without kreqs)- added forecenewconns patch (from CVS)- release candidate 2 of next upstream version - still needs patches for the tasklist_lock issue - added patch against leaks in the kernel module (from CVS) - fixed retry of parallel build of only_libafs_tree- put SLx macros into a helper script, recognize 4.9x as SL5 - put debuginfo override into a helper script, required by RPM on SL3/4 - fix debuginfo release - omit kernel dependency from kernel-module unless built --with kreqs (work around yum limitation to help the kernel module plugin)- override debuginfo package name for kernel module package - to make this work, never build the module and userland in same run- added three patches from CVS: Patch1011: openafs-1.4.2-linux-osi-cred-pool-byebye-20061218.patch will hopefully rid us of the panics due to "cannot malloc 4xxx bytes" observed on SL3 systems (2.4.21-47.0.1.ELsmp), even if unloaded Patch1012: openafs-1.4.2-linux26-disable-backing-readahead-20061109.patch is just there to make the noext one succeed, and should not do any harm Patch1013: openafs-1.4.2-linux-statfs-dentry-20061109.patch should provide a reasonable df output (not 0kB or 166YB) on recent 2.6 kernels (SL5)- el5 beta2: tasklist_lock is back but now GPL-only... do not use...- add patch on SL3/amd64 to get the syscall hooks (should be in 1.4.3)- should build on {SL3 SL4 SL5} x {i686 ia64 x86_64} x {any variant} - renamed base package to openafs.SLx to get a common src.rpm; binary packages retain their usual names, with SL3 etc. appended to release; added build instructions to base package description - krb5: build aklog and asetkey from source coming with the openafs tarball now; only ka-forwarder is still built from krb5-afs - run parallel makes (and try again if the first attempt fails) - make use of new (build) requirements on kernel[-devel] on SL4/5 - updated CellServDB to latest from grand.central.org (w/o openafs.org entry!) - allow changing configure options from the command line - record configure option in description of server package - init script: check more thoroughly for a defined cell; try to guess it if not - init script: subtract 32MB from cache partition size if it is ext3 when calculating the usable cache size, then use 70% (not 80) - init script: use 100MB (not just 8) if cache is not on dedicated filesystem - on SL5: separate init scripts for client and server - client: /etc/init.d/afs (as before), owned by openafs-client - server: /etc/init.d/afs-server (new), owned by openafs-server- replace last patch by openafs-1.4.2fc4-rcu_read_lock.patch - 32bit syscalls are in upstream now- add openafs-1.4.2fc3-no-tasklist_lock.patch for build on RHEL5beta1 x86_64- add openafs-1.4.2fc3-x64-32bitsyscalls.patch for build on RHEL5beta1 x86_64- upstream final candidate 3 - besides a real fix for amd64, some quite substantial changes in salvager and volserver- drop the remaining two patches that were not in 1.4.1- find kernel headers in /usr/src/kernels - add rxubikdeclare patch (not sure it is important) - find kernel headers on FC6- drop patches no longer needed - add patch 1005 for last tiny configure problem, run regen, add buildreq on autoconf, automake- build on FC6T2 (ugly, ugly hacks) - update CellServDB - the last patch is now in upstream sources - tag "copyright" changed to "license" - aklog.1 is no longer in src/aklog- added openafs-1.4.1-rxkad-ticketsize.patch, see mail to openafs-info by H. Reuter 2006-05-17 (and the reply by R. Toebbicke)- new upstream release with critical bugs fixed in the fileserver - removed patch 1002 (rra-soname-bug18767.patch), it's in upstream now - new patch 1003 openafs-1.4.1-moduleparam.patch to get modules built on 2.6.9 (shamelessly stolen from Derek Atkins' "official" SRPM) - pthreaded butc is now the default; package butc.lwp in addition, instead of a second butc.pthread - adapted to upstream changes w.r.t. man pages (there are quite a few more now), again borrowing from Derek Atkins' spec, but still preferring the enhanced ones from #19268, and packaging many more of them - spec changes to allow module build for hugemem kernels (Fabien Wernli) - the same for the new largesmp variant - added build requirements on gcc, flex, bison, ncurses-devel (Chris Huebsch) - make sure /usr/kerberos/bin is leading in PATH when configuring krb5 stuff- final release of 1.4.0- latest release candidate, possibly the last one before 1.4.0 - added man pages (like "official" spec, but add those from RT #19268 and prefer them over the default ones where duplicate) - added openafs-LICENSE.Sun (from "official" package) - removed unused source6 openafs-modname helper - add x-bit to kernel module to get it stripped - really do not turn off __check_files - remove all unpackaged files - added xstat_*_test to debug package - added -authlibs and -authlibs-devel packages (like "official" package, but keep static libs in -devel to make updates painless) - added patch for those to correct sonames (RT #18767) - added copyauth, kseal, vsys to main package - added ka-forwarder to -krb5 package - added bos_util, voldump to -server package - added des_prototypes.h to -devel package - updated CellServDB to latest version from grand.central.org (June 1) and added hepix.org cell- removed LSB init info from start script (breaks recent chkconfig)- added a patch (1001) from Jason McCormick to fix cache coherence issues with this release- added missing ia64 changes by Jarek from 1.3.80-2.SL (kernel requirement, module built is .mp.ko, not .ko) - turned off leftover debugspec- new upstream release fixing linux 2.6 issues including the first write access hanging on x86/UP - hopefully guessed right about the fix for ia64 module build (by Jarek for 1.3.80-2.SL, but could not find the srpm) - made the init script a separate source file - in init script, do not run the on_network check if ENABLE_DYNROOT=1 - package the LWP builds of fileserver & volserver (which are built but not installed by default) in addition to the default pthreaded ones, with suffix ".lwp" - package the pthread build of butc in addition to the default LWP one, with suffix ".pthread" - disable patch101 (obsoleted by upstream hack) - removed the build hack from 1.3.80-1.SL, fixed upstream- new upstream version supposed to fix many problems on linux 2.6 - notice the default afsd behaviour finally has changed to -nosettime; add -settime to your afsd options to get back the old behaviour - added an ugly hack to get the kernel modules built - updated CellServDB to version from grand.central.org as of today- made default cache location and size macros, removed cacheinfo source5, generate cacheinfo in install instead - fixed default cache size to 100MB again, resurrected the message in -client's post, changed default location to /var/cache/openafs - removed unused source21 (kernel-version.sh) - changed last afs-krb5 build patch (patch configure, not configure.in, and do not wipe out 64bit patch to configure by rerunning autoconf)- this version is SL4/1.3 only - get afs-krb5 going - removed tons of old cruft & patches no longer used - adapt module build to kernel[-smp]-devel: we now require this to be installed, and will only build the module package for a single kernel defined on the command line, or the running one in none is defined (this gets rid of all that redhat-buildsys voodoo which seems broken on SL4 anyway) - added livesys & kdump executables to main package - added optional openafs-debug package with additional tools - made module name in package & init script libafs instead of openafs, since that's what the module registers as anyway (modprobe -r openafs would fail) - added a patch (101) to fix the CACHESIZE=AUTOMATIC behaviour when the cache is mounted on some device with a long name- initial attempt at build for SL 4. (2.6 kernel)- added build of ia32e kernel module- All changes came from Stephan Wiesand. Many thanks for his pointers - afsd options chaned in spec file - /etc/sysconfig/afs set to be %config(noreplace) - requires wget - removed openafs-compat %pre scripts - Added CellAlias file, with commented out example- Put LC_ALL=C into the startup server test. This helps with internationalization. Submitted by Jaroslaw Polok- applied patch from Chris Wing for 64bit aklog & Co.- Put -fakestat in as a default option- Updated CellServDB - Put -dynroot in as a default option- Made more changes to startup script. Does sanity checks of ThisCell- Made change to /etc/init.d/afs startup script- rebuilt for Scientific Linux without site-specific setup - kerberos 5 enabled in build- changed packaging of modules again: to followup Fedora Core (draft) rules and allow yum/apt/up2date to handle these- changed packaging of modules- Rebuilt for CERN E. Linux (ix86/ia64/x86_64)- don't use rpm from within spec file as this can cause problems- added ia64 to the ExclusiveArch list- get rid of a %else that causes problems with rpm-4.0.4- added rpmbuild cmdline defines to control which bits to build- obey CFLAGS set at configure-time on i386/x86_64 linux - openafs-krb5: look in $krb5_prefix/lib64 for Kerberos libraries if they are not found in $krb5_prefix/lib - build the modules for PAM on x86_64 - when checking for res_search, try to link with it instead of checking for it in libraries, in case it's a macro or redefined by a macro - use %{ix86} instead of %{all_x86} where appropriate - install AFS libraries into %{_libdir} instead of %{_prefix}/lib- permit the kernel-to-build-against to be selected by cmdline argument to rpmbuild if preferred- fix compilation of openafs-krb5 bits against krb5 1.3- got rid of all rpmlint errors - delete certain tags (Packager/Vendor/Distribution)- Adapted OpenAFS version 1.2.10 to compile in beehivecompose-7.fnal.gov 14108007750:1.6.10-0.152.pre1.el70:1.6.10-0.152.pre1.el7openafs-modules/usr/sbin/-O2 -g -pipe -Wall -Wp,-D_FORTIFY_SOURCE=2 -fexceptions -fstack-protector-strong --param=ssp-buffer-size=4 -grecord-gcc-switches -m64 -mtune=genericcpioxz2x86_64-redhat-linux-gnuBourne-Again shell script, ASCII text executableRtx򛪮J?p7zXZ !#,%k ] b2u y-iSq߰3٤m{'xd1d@Ԅ6{ƒCTAJ.]dk$4txQ5zMu{3޽^\鷁x#݈Cς44ŽXwLbv6,SH7 lU) W3y`*=JkgyjY6v:Vm[XꏿRAvq{.Ԥ\q:9G2?T76Z K:3^,>9[Fр ?yTpÊ?]<v^+^0&1(ˏ*sڑOf`I⏺ڮ5-^.ҖAbVDIKi(ɑ@̸jq!_?M ئ\@ZVAî()HY&HZ6,/GyYa}$5*j,.tI D$vv.rlt"(K5sN[,V:': `(#hГŨy1H\rcS;H s754yl11uif*u~Սb?EI#;n}ff^Oį^ a6[ LQPgdɡ-bYNߛ*0!Hޏ㚼Qw:(ԟfDqT)IXQ(>l]M4#dM8$ 3,ɬ(442B'9\mV7X,KEl9!1im(,Jyi0y;РƤ3vKa4~bmCQ!% f@=6kҝYF ɓh"JvҸ֝Bٶ4r5`Z1?>bXف|`^@iigr #l3FW }FVQŒNteauٽ u}&ڂF wi}! ڱGav9*Q@œFW!Mf /Ɩgz F3t ^O x;8.L!ݞz"E/4 +3Urh\2 |^v @!? 㱳 0{b6F` Zi|@F8[hӰϒAH{+V =G&v,o"%"+$?dn10xDW*B̹e6!>HO=ҔIdG>&&DA  bǵ&`ĩqwAQUSŬOWӎo3tnS1oي=/cyEI$1*}&˨ <1WK"ٝ tЋ˜x{?"^#/Hy7i"FX  ځjUT:G?`%Zx|H6<)+n36'u:D'OHXqBGROHq,-+dAЌ 4vIЊB7/($C %DdYZ KHF7%Q,Х4khZ?@ȟQA>ٍel(Q#ĈV^Mq&O>yZ!HS; O4Ka]N"ۜ]+̟o.vZ;-ytN%0t]D?=!"|=ƥA0 A9/:c_f<ϰ5srchC2δ 1:@k[T Ob΃ʳ송hd֨аep9:8ECps| >ƣ'o 5%(Pc+Y2|~Wkh!= !m)a}asnWqIi5#iwCl`V{R+hJLACq h9;F]AƏEf쁕&Yh/x؂>̑ѷcdX^U{pU)0B egh4V(ϣKFfL9K n@XPǾ?t224t94!.")8 b9hA MA+pE?݇ⷰ]`IGF(]VnLQ Q+xG9:MR%]xa]>, 5~z;eaR)yk1eE_z^&(h& BK$dU r;+|T+B]I–Ri2F,5Sjkk}kɛ#4ɹ6;D5GY|RX dx'|aYHi1],]zS&>||W`ȕE\𐌥ZdQh,#00y#em3o7#iL߸:{1/'}i,}׸T {Pb*3ܧʠ7c! ~G8}j:mlڟu NJ ? [śT"RG@~~]\#*@(_ k Bs(8zgQ[=g"hADC6יDJpfkK18-Fk1"|RTY#9n*MLs"}8ySo\Ez1jjB{01ҞclPFH,Q%{_yK^!D|Cg`&{#1̓[!JTCk0"k:^_/|Yk]!)MYGv pjs:<$p#Z3Z{h{6D_li{Mus'D &ڷg;/H]X#=<è'P^0"G$+,(8Vƌ 9A%.P4 izM +" Qc0IHlwwܺ[@Rջo@|J lyCJ $ YZ