{"thread":{"id":"11128","subject":"Building git-1.5.3.7 on HP-UX 11.00","startedAt":"2007-12-04T13:09:22Z","lastAt":"2007-12-11T13:53:03Z","messageCount":27,"participants":["H.Merijn Brand","Johannes Schindelin","Andreas Ericsson","Junio C Hamano","Shawn O. Pearce","Johannes Sixt"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"61904","messageId":"20071204130922.731c407a@pc09.procura.nl","threadId":"11128","inReplyTo":null,"subject":"Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T13:09:22Z","receivedAt":"2007-12-04T13:09:22Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"1. It won't compile at all with HP C-ANSI-C, so I have to use gcc-3.4.6\n\nI added this to Makefile to get it compiled\n\nifeq ($(uname_S),HP-UX)\n\t# HP-UX\n\tBASIC_LDFLAGS += -L/usr/local/ssl/lib -L/usr/local/lib\n\tEXTLIBS += -ldce\n\tNO_HSTRERROR = YesPlease\n\tNO_ICONV = YesPlease\n\tNO_INET_NTOP = YesPlease\n\tNO_INET_PTON = YesPlease\nendif\n\nmake test then fails:\n\n/pro/3gl/LINUX/git-1.5.3.7 178 > make test\n LINK git-convert-objects\n LINK git\n BUILTIN git-format-patch\n BUILTIN git-show\n BUILTIN git-whatchanged\n BUILTIN git-cherry\n BUILTIN git-get-tar-commit-id\n BUILTIN git-init\n BUILTIN git-repo-config\n BUILTIN git-fsck-objects\n BUILTIN git-cherry-pick\n BUILTIN git-add\n BUILTIN git-annotate\n BUILTIN git-apply\n BUILTIN git-archive\n BUILTIN git-blame\n BUILTIN git-branch\n BUILTIN git-bundle\n BUILTIN git-cat-file\n BUILTIN git-check-attr\n BUILTIN git-checkout-index\n BUILTIN git-check-ref-format\n BUILTIN git-commit-tree\n BUILTIN git-count-objects\n BUILTIN git-describe\n BUILTIN git-diff\n BUILTIN git-diff-files\n BUILTIN git-diff-index\n BUILTIN git-diff-tree\n BUILTIN git-fetch--tool\n BUILTIN git-fmt-merge-msg\n BUILTIN git-for-each-ref\n BUILTIN git-fsck\n BUILTIN git-gc\n BUILTIN git-grep\n BUILTIN git-init-db\n BUILTIN git-log\n BUILTIN git-ls-files\n BUILTIN git-ls-tree\n BUILTIN git-mailinfo\n BUILTIN git-mailsplit\n BUILTIN git-merge-base\n BUILTIN git-merge-file\n BUILTIN git-mv\n BUILTIN git-name-rev\n BUILTIN git-pack-objects\n BUILTIN git-prune\n BUILTIN git-prune-packed\n BUILTIN git-push\n BUILTIN git-read-tree\n BUILTIN git-reflog\n BUILTIN git-config\n BUILTIN git-rerere\n BUILTIN git-rev-list\n BUILTIN git-rev-parse\n BUILTIN git-revert\n BUILTIN git-rm\n BUILTIN git-runstatus\n BUILTIN git-shortlog\n BUILTIN git-show-branch\n BUILTIN git-stripspace\n BUILTIN git-symbolic-ref\n BUILTIN git-tag\n BUILTIN git-tar-tree\n BUILTIN git-unpack-objects\n BUILTIN git-update-index\n BUILTIN git-update-ref\n BUILTIN git-upload-archive\n BUILTIN git-verify-pack\n BUILTIN git-verify-tag\n BUILTIN git-write-tree\n BUILTIN git-show-ref\n BUILTIN git-pack-refs\n SUBDIR git-gui\n INDEX lib/\n SUBDIR perl\n SUBDIR templates\nmake -C t/ all\nmake[1]: Entering directory `/pro/3gl/LINUX/git-1.5.3.7/t'\n*** t0000-basic.sh ***\n* ok 1: .git/objects should be empty after git init in an empty repo.\n* ok 2: .git/objects should have 3 subdirectories.\n* ok 3: git update-index without --add should fail adding.\n* ok 4: git update-index with --add should succeed.\n* ok 5: writing tree out with git write-tree\n* ok 6: validate object ID of a known tree.\n* ok 7: git update-index without --remove should fail removing.\n* ok 8: git update-index with --remove should be able to remove.\n* ok 9: git write-tree should be able to write an empty tree.\n* ok 10: validate object ID of a known tree.\n* ok 11: adding various types of objects with git update-index --add.\n* ok 12: showing stage with git ls-files --stage\n* ok 13: validate git ls-files output for a known tree.\n* ok 14: writing tree out with git write-tree.\n* ok 15: validate object ID for a known tree.\n* ok 16: showing tree with git ls-tree\n* ok 17: git ls-tree output for a known tree.\n* ok 18: showing tree with git ls-tree -r\n* ok 19: git ls-tree -r output for a known tree.\n* ok 20: showing tree with git ls-tree -r -t\n* ok 21: git ls-tree -r output for a known tree.\n* ok 22: writing partial tree out with git write-tree --prefix.\n* ok 23: validate object ID for a known tree.\n* ok 24: writing partial tree out with git write-tree --prefix.\n* ok 25: validate object ID for a known tree.\n* ok 26: put invalid objects into the index.\n* ok 27: writing this tree without --missing-ok.\n* ok 28: writing this tree with --missing-ok.\n* ok 29: git read-tree followed by write-tree should be idempotent.\n* ok 30: validate git diff-files output for a know cache/work tree state.\n* ok 31: git update-index --refresh should succeed.\n* ok 32: no diff after checkout and git update-index --refresh.\n* ok 33: git commit-tree records the correct tree in a commit.\n* ok 34: git commit-tree records the correct parent in a commit.\n* ok 35: git commit-tree omits duplicated parent in a commit.\n* ok 36: update-index D/F conflict\n* ok 37: absolute path works as expected\n* passed all 37 test(s)\n*** t0001-init.sh ***\n* FAIL 1: plain\n\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n* ok 2: plain with GIT_WORK_TREE\n* FAIL 3: plain bare\n\n (\n unset GIT_DIR GIT_WORK_TREE GIT_CONFIG &&\n mkdir plain-bare-1 &&\n cd plain-bare-1 &&\n git --bare init\n ) &&\n check_config plain-bare-1 true unset\n\n* ok 4: plain bare with GIT_WORK_TREE\n* ok 5: GIT_DIR bare\n* ok 6: GIT_DIR non-bare\n* ok 7: GIT_DIR & GIT_WORK_TREE (1)\n* ok 8: GIT_DIR & GIT_WORK_TREE (2)\n* failed 2 among 8 test(s)\nmake[1]: *** [t0001-init.sh] Error 1\nmake[1]: Leaving directory `/pro/3gl/LINUX/git-1.5.3.7/t'\nmake: *** [test] Error 2\nExit 2\n\n\nand make install probably expects a broken^Wdifferent install than HP-UX has:\n\n/pro/3gl/LINUX/git-1.5.3.7 181 > make install\n SUBDIR git-gui\n INDEX lib/\n SUBDIR perl\n SUBDIR templates\ninstall -d -m755 '/pro/local/bin'\nrm: /pro/local/bin/ directory\nUsage: mv [-f] [-i] [-e warn|force|ignore] f1 f2\n mv [-f] [-i] [-e warn|force|ignore] f1 ... fn d1\n mv [-f] [-i] [-e warn|force|ignore] d1 d2\ninstall -d -m755 '/pro/local/bin'\nrm: /pro/local/bin/ directory\nUsage: mv [-f] [-i] [-e warn|force|ignore] f1 f2\n mv [-f] [-i] [-e warn|force|ignore] f1 ... fn d1\n mv [-f] [-i] [-e warn|force|ignore] d1 d2\ninstall git-convert-objects git-fetch-pack git-hash-object git-index-pack git-local-fetch git-fast-import git-daemon git-merge-index git-mktag git-mktree git-patch-id git-peek-remote git-receive-pack git-send-pack git-shell git-show-index git-ssh-fetch git-ssh-upload git-unpack-file git-update-server-info git-upload-pack git-pack-redundant git-var git-merge-tree git-imap-send git-merge-recursive git-ssh-pull git-ssh-push git-bisect git-checkout git-clean git-clone git-commit git-fetch git-ls-remote git-merge-one-file git-mergetool git-parse-remote git-pull git-rebase git-rebase--interactive git-repack git-request-pull git-reset git-sh-setup git-am git-merge git-merge-stupid git-merge-octopus git-merge-resolve git-merge-ours git-lost-found git-quiltimport git-submodule git-filter-branch gi\n t-stash git-add--interactive git-archimport git-cvsimport git-relink git-cvsserver git-remote git-svnimport git-cvsexportcommit git-send-email git-svn git-status git-instaweb git-merge-subt!\n ree '/pro/local/bin'\ninstall git '/pro/local/bin'\nmake -C templates DESTDIR='' install\nmake[1]: Entering directory `/pro/3gl/LINUX/git-1.5.3.7/templates'\ninstall -d -m755 '/pro/local/share/git-core/templates/'\nUsage: mv [-f] [-i] [-e warn|force|ignore] f1 f2\n mv [-f] [-i] [-e warn|force|ignore] f1 ... fn d1\n mv [-f] [-i] [-e warn|force|ignore] d1 d2\n(cd blt && tar cf - .) | \\\n(cd '/pro/local/share/git-core/templates/' && tar xf -)\n/bin/sh: /pro/local/share/git-core/templates/: not found.\nmake[1]: *** [install] Error 1\nmake[1]: Leaving directory `/pro/3gl/LINUX/git-1.5.3.7/templates'\nmake: *** [install] Error 2\nExit 2\n\nI have two:\n\n 14241 100555 -r-x 1 bin 1888 12 Jun 2002 10:58 /opt/imake/bin/install\n 634 100555 -r-x 1 bin 5147 7 Nov 1997 09:00 /usr/sbin/install\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61907","messageId":"Pine.LNX.4.64.0712041343040.27959@racer.site","threadId":"11128","inReplyTo":"20071204130922.731c407a@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T13:44:02Z","receivedAt":"2007-12-04T13:44:02Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 4 Dec 2007, H.Merijn Brand wrote:\n\n> make test then fails:\n\nCould you run the failing test manually with -i -v, like so:\n\n\t$ cd t/\n\t$ sh t0001*.sh -i -v\n\nThanks,\nDscho\n"},{"id":"61909","messageId":"20071204140326.14d9e7a0@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712041343040.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T14:03:26Z","receivedAt":"2007-12-04T14:03:26Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 13:44:02 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > make test then fails:\n> \n> Could you run the failing test manually with -i -v, like so:\n> \n> \t$ cd t/\n> \t$ sh t0001*.sh -i -v\n\n/pro/3gl/LINUX/git-1.5.3.7 183 > cd t\n/pro/3gl/LINUX/git-1.5.3.7/t 184 > sh t0001*.sh -i -v\n* error: cannot run git init -- have you built things yet?\nExit 1\n/pro/3gl/LINUX/git-1.5.3.7/t 185 >\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61911","messageId":"Pine.LNX.4.64.0712041439590.27959@racer.site","threadId":"11128","inReplyTo":"20071204140326.14d9e7a0@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T14:40:47Z","receivedAt":"2007-12-04T14:40:47Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 4 Dec 2007, H.Merijn Brand wrote:\n\n> On Tue, 4 Dec 2007 13:44:02 +0000 (GMT), Johannes Schindelin\n> wrote:\n> \n> > Hi,\n> > \n> > On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> > \n> > > make test then fails:\n> > \n> > Could you run the failing test manually with -i -v, like so:\n> > \n> > \t$ cd t/\n> > \t$ sh t0001*.sh -i -v\n> \n> /pro/3gl/LINUX/git-1.5.3.7 183 > cd t\n> /pro/3gl/LINUX/git-1.5.3.7/t 184 > sh t0001*.sh -i -v\n> * error: cannot run git init -- have you built things yet?\n> Exit 1\n> /pro/3gl/LINUX/git-1.5.3.7/t 185 >\n\nNext step: add the sh option -x:\n\n\t$ sh -x t0001*.sh -i -v\n\nto find out what made it fail.\n\nThanks,\nDscho\n"},{"id":"61912","messageId":"20071204150102.7f3ec3e9@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712041439590.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T15:01:02Z","receivedAt":"2007-12-04T15:01:02Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 14:40:47 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > On Tue, 4 Dec 2007 13:44:02 +0000 (GMT), Johannes Schindelin\n> > wrote:\n> > \n> > > Hi,\n> > > \n> > > On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> > > \n> > > > make test then fails:\n> > > \n> > > Could you run the failing test manually with -i -v, like so:\n> > > \n> > > \t$ cd t/\n> > > \t$ sh t0001*.sh -i -v\n> > \n> > /pro/3gl/LINUX/git-1.5.3.7 183 > cd t\n> > /pro/3gl/LINUX/git-1.5.3.7/t 184 > sh t0001*.sh -i -v\n> > * error: cannot run git init -- have you built things yet?\n> > Exit 1\n> > /pro/3gl/LINUX/git-1.5.3.7/t 185 >\n> \n> Next step: add the sh option -x:\n\nI also meanwhile forced copied the git binaries to my bin dir\n\n/pro/3gl/LINUX/git-1.5.3.7/t 106 > git --version\ngit version 1.5.3.7\n\n\n> \t$ sh -x t0001*.sh -i -v\n\n/pro/3gl/LINUX/git-1.5.3.7/t 105 > sh -x t0001*.sh -i -v\n+ test_description=git init\n+ . ./test-lib.sh\n+ LANG=C\n+ LC_ALL=C\n+ PAGER=cat\n+ TZ=UTC\n+ export LANG LC_ALL PAGER TZ\n+ EDITOR=:\n+ VISUAL=:\n+ unset GIT_EDITOR\n+ unset AUTHOR_DATE\n+ unset AUTHOR_EMAIL\n+ unset AUTHOR_NAME\n+ unset COMMIT_AUTHOR_EMAIL\n+ unset COMMIT_AUTHOR_NAME\n+ unset EMAIL\n+ unset GIT_ALTERNATE_OBJECT_DIRECTORIES\n+ unset GIT_AUTHOR_DATE\n+ GIT_AUTHOR_EMAIL=author@example.com\n+ GIT_AUTHOR_NAME=A U Thor\n+ unset GIT_COMMITTER_DATE\n+ GIT_COMMITTER_EMAIL=committer@example.com\n+ GIT_COMMITTER_NAME=C O Mitter\n+ unset GIT_DIFF_OPTS\n+ unset GIT_DIR\n+ unset GIT_WORK_TREE\n+ unset GIT_EXTERNAL_DIFF\n+ unset GIT_INDEX_FILE\n+ unset GIT_OBJECT_DIRECTORY\n+ unset SHA1_FILE_DIRECTORIES\n+ unset SHA1_FILE_DIRECTORY\n+ GIT_MERGE_VERBOSITY=5\n+ export GIT_MERGE_VERBOSITY\n+ export GIT_AUTHOR_EMAIL GIT_AUTHOR_NAME\n+ export GIT_COMMITTER_EMAIL GIT_COMMITTER_NAME\n+ export EDITOR VISUAL\n+ unset CDPATH\n+ tr [A-Z] [a-z]\n+ echo\n+ test git init !=\n+ test 2 -ne 0\n+ immediate=t\n+ shift\n+ test 1 -ne 0\n+ verbose=t\n+ shift\n+ test 0 -ne 0\n+ exec\n+ 5>& 1\n+ test t = t\n+ exec\n+ 4>& 2 3>& 1\n+ test_failure=0\n+ test_count=0\n+ trap echo >&5 \"FATAL: Unexpected exit with code $?\"; exit 1 exit\n+ + pwd\nPATH=/pro/3gl/LINUX/git-1.5.3.7/t/..:.:/u/usr/merijn/bin/private:/u/usr/merijn/bin:/pro/local/bin:/pro/bin:/usr/hosts:/usr/bin:/opt/ansic/bin:/usr/ccs/bin:/usr/contrib/bin:/opt/nettladm/bin:/opt/fc/bin:/opt/fcms/bin:/opt/upgrade/bin:/opt/pd/bin:/usr/bin/X11:/usr/contrib/bin/X11:/opt/resmon/bin:/usr/sbin/diag/contrib:/opt/pred/bin:/opt/hparray/bin:/opt/perf/bin:/opt/langtools/bin:/opt/imake/bin:/opt/ignite/bin:/opt/sec_mgmt/spc/bin:/usr/local/bin:/usr/local/sbin:/opt/java1.5/jre/bin:/opt/graphics/common/bin:/pro/local/libexec:/pro/local/sbin:/pro/local/samba/bin:/pro/local/samba/sbin:/opt/hpnpl/bin:/pro/tu/bin:/usr/sbin:/sbin:/usr/local/pa20_64/bin:/pro/3gl/CPAN/bin\n+ + pwd\nGIT_EXEC_PATH=/pro/3gl/LINUX/git-1.5.3.7/t/..\n+ + pwd\nGIT_TEMPLATE_DIR=/pro/3gl/LINUX/git-1.5.3.7/t/../templates/blt\n+ GIT_CONFIG=.git/config\n+ export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n+ + pwd\n+ pwd\nGITPERLLIB=/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/lib:/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/arch/auto/Git\n+ export GITPERLLIB\n+ test -d ../templates/blt\n+ test -x ../test-chmtime\n+ test=trash\n+ rm -fr trash\n+ test_create_repo trash\n* error: cannot run git init -- have you built things yet?\nExit 1\n\npro/3gl/LINUX/git-1.5.3.7/t 108 > ll /pro/local/bin/git*\n127035 -rwxrwxrwx 1 merijn softwr 1081204 Dec 4 15:19 /pro/local/bin/git\n 86787 -rwxrwxrwx 1 merijn softwr 16467 Dec 4 15:22 /pro/local/bin/git-add--interactive\n 86627 -rwxrwxrwx 1 merijn softwr 11285 Dec 4 15:22 /pro/local/bin/git-am\n 86909 -rwxrwxrwx 1 merijn softwr 36968 Dec 4 15:22 /pro/local/bin/git-archimport\n 86420 -rwxrwxrwx 1 merijn softwr 8508 Dec 4 15:22 /pro/local/bin/git-bisect\n 86423 -rwxrwxrwx 1 merijn softwr 6905 Dec 4 15:22 /pro/local/bin/git-checkout\n 86424 -rwxrwxrwx 1 merijn softwr 1972 Dec 4 15:22 /pro/local/bin/git-clean\n 86427 -rwxrwxrwx 1 merijn softwr 11560 Dec 4 15:22 /pro/local/bin/git-clone\n 86428 -rwxrwxrwx 1 merijn softwr 14693 Dec 4 15:22 /pro/local/bin/git-commit\n127036 -rwxrwxrwx 1 merijn softwr 327540 Dec 4 15:19 /pro/local/bin/git-convert-objects\n 87086 -rwxrwxrwx 1 merijn softwr 9394 Dec 4 15:22 /pro/local/bin/git-cvsexportcommit\n 86968 -rwxrwxrwx 1 merijn softwr 26590 Dec 4 15:22 /pro/local/bin/git-cvsimport\n 86997 -rwxrwxrwx 1 merijn softwr 101558 Dec 4 15:22 /pro/local/bin/git-cvsserver\n 86356 -rwxrwxrwx 1 merijn softwr 343924 Dec 4 15:22 /pro/local/bin/git-daemon\n 86355 -rwxrwxrwx 1 merijn softwr 420048 Dec 4 15:22 /pro/local/bin/git-fast-import\n 86440 -rwxrwxrwx 1 merijn softwr 9083 Dec 4 15:22 /pro/local/bin/git-fetch\n 87107 -rwxrwxrwx 1 merijn softwr 360308 Dec 4 15:22 /pro/local/bin/git-fetch-pack\n 86724 -rwxrwxrwx 1 merijn softwr 10240 Dec 4 15:22 /pro/local/bin/git-filter-branch\n 86351 -rwxrwxrwx 1 merijn softwr 323444 Dec 4 15:22 /pro/local/bin/git-hash-object\n 86390 -rwxrwxrwx 1 merijn softwr 343924 Dec 4 15:22 /pro/local/bin/git-imap-send\n 87114 -rwxrwxrwx 1 merijn softwr 348020 Dec 4 15:22 /pro/local/bin/git-index-pack\n 87100 -rwxrwxrwx 1 merijn softwr 177519 Dec 4 15:22 /pro/local/bin/git-instaweb\n 86342 -rwxrwxrwx 1 merijn softwr 339828 Dec 4 15:22 /pro/local/bin/git-local-fetch\n 86680 -rwxrwxrwx 1 merijn softwr 473 Dec 4 15:22 /pro/local/bin/git-lost-found\n 86444 -rwxrwxrwx 1 merijn softwr 2544 Dec 4 15:22 /pro/local/bin/git-ls-remote\n 86633 -rwxrwxrwx 1 merijn softwr 11464 Dec 4 15:22 /pro/local/bin/git-merge\n 86359 -rwxrwxrwx 1 merijn softwr 323444 Dec 4 15:22 /pro/local/bin/git-merge-index\n 86636 -rwxrwxrwx 1 merijn softwr 2415 Dec 4 15:22 /pro/local/bin/git-merge-octopus\n 86445 -rwxrwxrwx 1 merijn softwr 3364 Dec 4 15:22 /pro/local/bin/git-merge-one-file\n 86656 -rwxrwxrwx 1 merijn softwr 338 Dec 4 15:22 /pro/local/bin/git-merge-ours\n 86391 -rwxrwxrwx 1 merijn softwr 491380 Dec 4 15:22 /pro/local/bin/git-merge-recursive\n 86643 -rwxrwxrwx 1 merijn softwr 954 Dec 4 15:22 /pro/local/bin/git-merge-resolve\n 86634 -rwxrwxrwx 1 merijn softwr 1456 Dec 4 15:22 /pro/local/bin/git-merge-stupid\n 86399 -rwxrwxrwx 1 merijn softwr 491380 Dec 4 15:21 /pro/local/bin/git-merge-subtree\n 86389 -rwxrwxrwx 1 merijn softwr 348020 Dec 4 15:22 /pro/local/bin/git-merge-tree\n 86446 -rwxrwxrwx 1 merijn softwr 8649 Dec 4 15:22 /pro/local/bin/git-mergetool\n 86361 -rwxrwxrwx 1 merijn softwr 323444 Dec 4 15:22 /pro/local/bin/git-mktag\n 86363 -rwxrwxrwx 1 merijn softwr 323444 Dec 4 15:22 /pro/local/bin/git-mktree\n 86387 -rwxrwxrwx 1 merijn softwr 331636 Dec 4 15:22 /pro/local/bin/git-pack-redundant\n 86447 -rwxrwxrwx 1 merijn softwr 5892 Dec 4 15:22 /pro/local/bin/git-parse-remote\n 86365 -rwxrwxrwx 1 merijn softwr 319348 Dec 4 15:22 /pro/local/bin/git-patch-id\n 86367 -rwxrwxrwx 1 merijn softwr 348020 Dec 4 15:22 /pro/local/bin/git-peek-remote\n 86450 -rwxrwxrwx 1 merijn softwr 3868 Dec 4 15:22 /pro/local/bin/git-pull\n 86692 -rwxrwxrwx 1 merijn softwr 3203 Dec 4 15:22 /pro/local/bin/git-quiltimport\n 86454 -rwxrwxrwx 1 merijn softwr 8737 Dec 4 15:22 /pro/local/bin/git-rebase\n 86456 -rwxrwxrwx 1 merijn softwr 11642 Dec 4 15:22 /pro/local/bin/git-rebase--interactive\n 86369 -rwxrwxrwx 1 merijn softwr 335732 Dec 4 15:22 /pro/local/bin/git-receive-pack\n 86988 -rwxrwxrwx 1 merijn softwr 4189 Dec 4 15:22 /pro/local/bin/git-relink\n 87012 -rwxrwxrwx 1 merijn softwr 9524 Dec 4 15:22 /pro/local/bin/git-remote\n 86457 -rwxrwxrwx 1 merijn softwr 2886 Dec 4 15:22 /pro/local/bin/git-repack\n 86458 -rwxrwxrwx 1 merijn softwr 1364 Dec 4 15:22 /pro/local/bin/git-request-pull\n 86591 -rwxrwxrwx 1 merijn softwr 2304 Dec 4 15:22 /pro/local/bin/git-reset\n 87087 -rwxrwxrwx 1 merijn softwr 20052 Dec 4 15:22 /pro/local/bin/git-send-email\n 86371 -rwxrwxrwx 1 merijn softwr 356212 Dec 4 15:22 /pro/local/bin/git-send-pack\n 86616 -rwxrwxrwx 1 merijn softwr 2960 Dec 4 15:22 /pro/local/bin/git-sh-setup\n 86372 -rwxrwxrwx 1 merijn softwr 319348 Dec 4 15:22 /pro/local/bin/git-shell\n 86375 -rwxrwxrwx 1 merijn softwr 323444 Dec 4 15:22 /pro/local/bin/git-show-index\n 86376 -rwxrwxrwx 1 merijn softwr 335732 Dec 4 15:22 /pro/local/bin/git-ssh-fetch\n 86392 -rwxrwxrwx 1 merijn softwr 335732 Dec 4 15:22 /pro/local/bin/git-ssh-pull\n 86414 -rwxrwxrwx 1 merijn softwr 327540 Dec 4 15:22 /pro/local/bin/git-ssh-push\n 86379 -rwxrwxrwx 1 merijn softwr 327540 Dec 4 15:22 /pro/local/bin/git-ssh-upload\n 86749 -rwxrwxrwx 1 merijn softwr 4613 Dec 4 15:22 /pro/local/bin/git-stash\n 87098 -rwxrwxrwx 1 merijn softwr 14693 Dec 4 15:22 /pro/local/bin/git-status\n 86697 -rwxrwxrwx 1 merijn softwr 6388 Dec 4 15:22 /pro/local/bin/git-submodule\n 87097 -rwxrwxrwx 1 merijn softwr 119623 Dec 4 15:22 /pro/local/bin/git-svn\n 87085 -rwxrwxrwx 1 merijn softwr 24869 Dec 4 15:22 /pro/local/bin/git-svnimport\n 86381 -rwxrwxrwx 1 merijn softwr 319348 Dec 4 15:22 /pro/local/bin/git-unpack-file\n 86382 -rwxrwxrwx 1 merijn softwr 327540 Dec 4 15:22 /pro/local/bin/git-update-server-info\n 86386 -rwxrwxrwx 1 merijn softwr 487284 Dec 4 15:22 /pro/local/bin/git-upload-pack\n 86388 -rwxrwxrwx 1 merijn softwr 319348 Dec 4 15:22 /pro/local/bin/git-var\n\n\n> to find out what made it \n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61913","messageId":"20071204151147.3bda2292@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712041439590.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T15:11:47Z","receivedAt":"2007-12-04T15:11:47Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 14:40:47 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > On Tue, 4 Dec 2007 13:44:02 +0000 (GMT), Johannes Schindelin\n> > wrote:\n> > \n> > > Hi,\n> > > \n> > > On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> > > \n> > > > make test then fails:\n> > > \n> > > Could you run the failing test manually with -i -v, like so:\n> > > \n> > > \t$ cd t/\n> > > \t$ sh t0001*.sh -i -v\n> > \n> > /pro/3gl/LINUX/git-1.5.3.7 183 > cd t\n> > /pro/3gl/LINUX/git-1.5.3.7/t 184 > sh t0001*.sh -i -v\n> > * error: cannot run git init -- have you built things yet?\n> > Exit 1\n> > /pro/3gl/LINUX/git-1.5.3.7/t 185 >\n> \n> Next step: add the sh option -x:\n> \n> \t$ sh -x t0001*.sh -i -v\n\nAfter I did # make git\nand # make test\n\n# cd t\n# h -x t0001*.sh -i -v\n:\n:\n+ + pwd\nGIT_EXEC_PATH=/pro/3gl/LINUX/git-1.5.3.7/t/..\n+ + pwd\nGIT_TEMPLATE_DIR=/pro/3gl/LINUX/git-1.5.3.7/t/../templates/blt\n+ GIT_CONFIG=.git/config\n+ export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n+ + pwd\n+ pwd\nGITPERLLIB=/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/lib:/pro/3gl/LINUX/git-1.5.\n3.7/t/../perl/blib/arch/auto/Git\n+ export GITPERLLIB\n+ test -d ../templates/blt\n+ test -x ../test-chmtime\n+ test=trash\n+ rm -fr trash\n+ test_create_repo trash\n+ cd trash\n+ + expr ./t0001-init.sh : .*/\\(t[0-9]*\\)-[^/]*$\nthis_test=t0001\n+ test_expect_success plain\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n* expecting success:\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n* FAIL 1: plain\n\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n\n> to find out what made it fail.\n> \n> Thanks,\n> Dscho\n> \n> \n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61914","messageId":"47556EE2.6040105@op5.se","threadId":"11128","inReplyTo":"20071204150102.7f3ec3e9@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Andreas Ericsson","fromEmail":"ae@op5.se","sentAt":"2007-12-04T15:14:42Z","receivedAt":"2007-12-04T15:14:42Z","isPatch":false,"sender":{"key":"ae@op5.se","avatar":"https://gravatar.com/avatar/426e89595c75a8f5252dd0c989e5fabe5bcac616e68557427ad9aef6b0ca342a?d=mp&s=160"},"body":"H.Merijn Brand wrote:\n> \n> I also meanwhile forced copied the git binaries to my bin dir\n> \n\nThat's no good. \"make test\" will test binaries residing in ..\nrelative to where the tests are executed.\n\n-- \nAndreas Ericsson andreas.ericsson@op5.se\nOP5 AB www.op5.se\nTel: +46 8-230225 Fax: +46 8-230231\n"},{"id":"61915","messageId":"20071204152240.6cb6018e@pc09.procura.nl","threadId":"11128","inReplyTo":"47556EE2.6040105@op5.se","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T15:22:40Z","receivedAt":"2007-12-04T15:22:40Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 04 Dec 2007 16:14:42 +0100, Andreas Ericsson wrote:\n\n> H.Merijn Brand wrote:\n> > \n> > I also meanwhile forced copied the git binaries to my bin dir\n> > \n> \n> That's no good. \"make test\" will test binaries residing in ..\n> relative to where the tests are executed.\n\nOK, I removed those again :)\n\nI also removed the -ldce in the HP-UX piece, which was there to\nsupport unsetenv, but using NO_UNSETENV seems to be safer\n\nifeq ($(uname_S),HP-UX)\n\t# HP-UX\n\tBASIC_LDFLAGS += -L/usr/local/ssl/lib -L/usr/local/lib\n\t#EXTLIBS += -ldce\n\tNO_HSTRERROR = YesPlease\n\tNO_ICONV = YesPlease\n\tNO_INET_NTOP = YesPlease\n\tNO_INET_PTON = YesPlease\n\tNO_UNSETENV = YesPlease\nendif\n\ncauses a clean build\n\n+ + pwd\nGIT_EXEC_PATH=/pro/3gl/LINUX/git-1.5.3.7/t/..\n+ + pwd\nGIT_TEMPLATE_DIR=/pro/3gl/LINUX/git-1.5.3.7/t/../templates/blt\n+ GIT_CONFIG=.git/config\n+ export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n+ + pwd\n+ pwd\nGITPERLLIB=/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/lib:/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/arch/auto/Git\n+ export GITPERLLIB\n+ test -d ../templates/blt\n+ test -x ../test-chmtime\n+ test=trash\n+ rm -fr trash\n+ test_create_repo trash\n+ cd trash\n+ + expr ./t0001-init.sh : .*/\\(t[0-9]*\\)-[^/]*$\nthis_test=t0001\n+ test_expect_success plain\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n* expecting success:\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n* FAIL 1: plain\n\n (\n unset GIT_DIR GIT_WORK_TREE &&\n mkdir plain &&\n cd plain &&\n git init\n ) &&\n check_config plain/.git false unset\n\n\n\n/pro/3gl/LINUX/git-1.5.3.7/t 130 > ldd ../git\n /usr/lib/libc.2 => /usr/lib/libc.2\n /usr/lib/libdld.2 => /usr/lib/libdld.2\n /usr/lib/libc.2 => /usr/lib/libc.2\n /usr/local/ssl/lib/libcrypto.sl => /usr/local/ssl/lib/libcrypto.sl\n /usr/local/lib/libz.sl => /usr/local/lib/libz.sl\n\n\nI will leave for a meeting in 30 minutes, so don't think I don't want\nto communicate any further :)\n\nSam is pretty enthusiastic on IRC, he's to blame!\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61918","messageId":"Pine.LNX.4.64.0712041536180.27959@racer.site","threadId":"11128","inReplyTo":"20071204152240.6cb6018e@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T15:39:47Z","receivedAt":"2007-12-04T15:39:47Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 4 Dec 2007, H.Merijn Brand wrote:\n\n> + + pwd\n> GIT_EXEC_PATH=/pro/3gl/LINUX/git-1.5.3.7/t/..\n> + + pwd\n> GIT_TEMPLATE_DIR=/pro/3gl/LINUX/git-1.5.3.7/t/../templates/blt\n> + GIT_CONFIG=.git/config\n> + export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n> + + pwd\n> + pwd\n> GITPERLLIB=/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/lib:/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/arch/auto/Git\n> + export GITPERLLIB\n> + test -d ../templates/blt\n> + test -x ../test-chmtime\n> + test=trash\n> + rm -fr trash\n> + test_create_repo trash\n> + cd trash\n> + + expr ./t0001-init.sh : .*/\\(t[0-9]*\\)-[^/]*$\n> this_test=t0001\n> + test_expect_success plain\n> (\n> unset GIT_DIR GIT_WORK_TREE &&\n> mkdir plain &&\n> cd plain &&\n> git init\n> ) &&\n> check_config plain/.git false unset\n> \n> * expecting success:\n> (\n> unset GIT_DIR GIT_WORK_TREE &&\n> mkdir plain &&\n> cd plain &&\n> git init\n> ) &&\n> check_config plain/.git false unset\n> \n> * FAIL 1: plain\n> \n> (\n> unset GIT_DIR GIT_WORK_TREE &&\n> mkdir plain &&\n> cd plain &&\n> git init\n> ) &&\n> check_config plain/.git false unset\n\nThat's not good. The relevant part here reads:\n\n-- snip --\n+ test_description='git init'\n+ . ./test-lib.sh\n++ LANG=C\n++ LC_ALL=C\n++ PAGER=cat\n++ TZ=UTC\n++ export LANG LC_ALL PAGER TZ\n++ EDITOR=:\n++ VISUAL=:\n++ unset GIT_EDITOR\n++ unset AUTHOR_DATE\n++ unset AUTHOR_EMAIL\n++ unset AUTHOR_NAME\n++ unset COMMIT_AUTHOR_EMAIL\n++ unset COMMIT_AUTHOR_NAME\n++ unset EMAIL\n++ unset GIT_ALTERNATE_OBJECT_DIRECTORIES\n++ unset GIT_AUTHOR_DATE\n++ GIT_AUTHOR_EMAIL=author@example.com\n++ GIT_AUTHOR_NAME='A U Thor'\n++ unset GIT_COMMITTER_DATE\n++ GIT_COMMITTER_EMAIL=committer@example.com\n++ GIT_COMMITTER_NAME='C O Mitter'\n++ unset GIT_DIFF_OPTS\n++ unset GIT_DIR\n++ unset GIT_WORK_TREE\n++ unset GIT_EXTERNAL_DIFF\n++ unset GIT_INDEX_FILE\n++ unset GIT_OBJECT_DIRECTORY\n++ unset SHA1_FILE_DIRECTORIES\n++ unset SHA1_FILE_DIRECTORY\n++ GIT_MERGE_VERBOSITY=5\n++ export GIT_MERGE_VERBOSITY\n++ export GIT_AUTHOR_EMAIL GIT_AUTHOR_NAME\n++ export GIT_COMMITTER_EMAIL GIT_COMMITTER_NAME\n++ export EDITOR VISUAL\n++ unset CDPATH\n++ case $(echo $GIT_TRACE |tr \"[A-Z]\" \"[a-z]\") in\n+++ echo\n+++ tr '[A-Z]' '[a-z]'\n++ '[' xxterm '!=' xdumb ']'\n++ '[' -t 1 ']'\n++ test 2 -ne 0\n++ case \"$1\" in\n++ immediate=t\n++ shift\n++ test 1 -ne 0\n++ case \"$1\" in\n++ verbose=t\n++ shift\n++ test 0 -ne 0\n++ test -n ''\n++ test 'git init' '!=' ''\n++ test '' = t\n++ exec\n++ test t = t\n++ exec\n++ test_failure=0\n++ test_count=0\n++ trap 'echo >&5 \"FATAL: Unexpected exit with code $?\"; exit 1' exit\n+++ pwd\n++ PATH=/home/gitte/my/git/t/..:/home/gene099/bin:/usr/local/bin:/usr/bin:/usr/X11R6/bin:/bin:/usr/games:/opt/gnome/bin:/opt/kde3/bin:/usr/lib/mit/bin:/usr/lib/mit/sbin:/usr/lib/qt3/bin\n+++ pwd\n++ GIT_EXEC_PATH=/home/gitte/my/git/t/..\n+++ pwd\n++ GIT_TEMPLATE_DIR=/home/gitte/my/git/t/../templates/blt\n++ GIT_CONFIG=.git/config\n++ export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n+++ pwd\n+++ pwd\n++ GITPERLLIB=/home/gitte/my/git/t/../perl/blib/lib:/home/gene099/my/git/t/../perl/blib/arch/auto/Git\n++ export GITPERLLIB\n++ test -d ../templates/blt\n++ test -x ../test-chmtime\n++ test=trash\n++ rm -fr trash\n++ test_create_repo trash\n++ test 1 = 1\n+++ pwd\n++ owd=/home/gitte/my/git/t\n++ repo=trash\n++ mkdir trash\n++ cd trash\n++ /home/gitte/my/git/t/../git init --template=/home/gene099/my/git/t/../templates/blt/\n++ mv .git/hooks .git/hooks-disabled\n++ cd /home/gitte/my/git/t\n++ cd trash\n+++ expr ./t0001-init.sh : '.*/\\(t[0-9]*\\)-[^/]*$'\n++ this_test=t0001\n+ test_expect_success plain '\n\t(\n\t\tunset GIT_DIR GIT_WORK_TREE &&\n\t\tmkdir plain &&\n\t\tcd plain &&\n\t\tgit init\n\t) &&\n\tcheck_config plain/.git false unset\n'\n+ test 2 = 2\n+ test_skip plain '\n\t(\n\t\tunset GIT_DIR GIT_WORK_TREE &&\n\t\tmkdir plain &&\n\t\tcd plain &&\n\t\tgit init\n\t) &&\n\tcheck_config plain/.git false unset\n'\n-- snap --\n\nSo I get way more information. Which let's me suspect that your shell \nmight be the culprit? Can you try with bash?\n\nCiao,\nDscho\n"},{"id":"61921","messageId":"20071204155655.053f4fb4@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712041536180.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T15:56:55Z","receivedAt":"2007-12-04T15:56:55Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > + + pwd\n> > GIT_EXEC_PATH=/pro/3gl/LINUX/git-1.5.3.7/t/..\n> > + + pwd\n> > GIT_TEMPLATE_DIR=/pro/3gl/LINUX/git-1.5.3.7/t/../templates/blt\n> > + GIT_CONFIG=.git/config\n> > + export PATH GIT_EXEC_PATH GIT_TEMPLATE_DIR GIT_CONFIG\n> > + + pwd\n> > + pwd\n> > GITPERLLIB=/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/lib:/pro/3gl/LINUX/git-1.5.3.7/t/../perl/blib/arch/auto/Git\n> > + export GITPERLLIB\n> > + test -d ../templates/blt\n> > + test -x ../test-chmtime\n> > + test=trash\n> > + rm -fr trash\n> > + test_create_repo trash\n> > + cd trash\n> > + + expr ./t0001-init.sh : .*/\\(t[0-9]*\\)-[^/]*$\n> > this_test=t0001\n> > + test_expect_success plain\n> > (\n> > unset GIT_DIR GIT_WORK_TREE &&\n> > mkdir plain &&\n> > cd plain &&\n> > git init\n> > ) &&\n> > check_config plain/.git false unset\n> > \n> > * expecting success:\n> > (\n> > unset GIT_DIR GIT_WORK_TREE &&\n> > mkdir plain &&\n> > cd plain &&\n> > git init\n> > ) &&\n> > check_config plain/.git false unset\n> > \n> > * FAIL 1: plain\n> > \n> > (\n> > unset GIT_DIR GIT_WORK_TREE &&\n> > mkdir plain &&\n> > cd plain &&\n> > git init\n> > ) &&\n> > check_config plain/.git false unset\n> \n> That's not good. The relevant part here reads:\n\nI found it! unset returns false\n\n/pro/3gl/LINUX/git-1.5.3.7/t 152 > sh t0001*.sh\n* ok 1: plain\n* ok 2: plain with GIT_WORK_TREE\n* ok 3: plain bare\n* ok 4: plain bare with GIT_WORK_TREE\n* ok 5: GIT_DIR bare\n* ok 6: GIT_DIR non-bare\n* ok 7: GIT_DIR & GIT_WORK_TREE (1)\n* ok 8: GIT_DIR & GIT_WORK_TREE (2)\n* passed all 8 test(s)\n\n--8<---\n--- t0001-init.sh.org 2007-12-04 16:55:25 +0100\n+++ t0001-init.sh 2007-12-04 16:52:37 +0100\n@@ -25,7 +25,7 @@ check_config () {\n\n test_expect_success 'plain' '\n (\n- unset GIT_DIR GIT_WORK_TREE &&\n+ unset GIT_DIR GIT_WORK_TREE ;\n mkdir plain &&\n cd plain &&\n git init\n@@ -35,7 +35,7 @@ test_expect_success 'plain' '\n\n test_expect_success 'plain with GIT_WORK_TREE' '\n if (\n- unset GIT_DIR &&\n+ unset GIT_DIR ;\n mkdir plain-wt &&\n cd plain-wt &&\n GIT_WORK_TREE=$(pwd) git init\n@@ -48,7 +48,7 @@ test_expect_success 'plain with GIT_WORK\n\n test_expect_success 'plain bare' '\n (\n- unset GIT_DIR GIT_WORK_TREE GIT_CONFIG &&\n+ unset GIT_DIR GIT_WORK_TREE GIT_CONFIG ;\n mkdir plain-bare-1 &&\n cd plain-bare-1 &&\n git --bare init\n@@ -58,7 +58,7 @@ test_expect_success 'plain bare' '\n\n test_expect_success 'plain bare with GIT_WORK_TREE' '\n if (\n- unset GIT_DIR GIT_CONFIG &&\n+ unset GIT_DIR GIT_CONFIG ;\n mkdir plain-bare-2 &&\n cd plain-bare-2 &&\n GIT_WORK_TREE=$(pwd) git --bare init\n@@ -72,7 +72,7 @@ test_expect_success 'plain bare with GIT\n test_expect_success 'GIT_DIR bare' '\n\n (\n- unset GIT_CONFIG &&\n+ unset GIT_CONFIG ;\n mkdir git-dir-bare.git &&\n GIT_DIR=git-dir-bare.git git init\n ) &&\n@@ -82,7 +82,7 @@ test_expect_success 'GIT_DIR bare' '\n test_expect_success 'GIT_DIR non-bare' '\n\n (\n- unset GIT_CONFIG &&\n+ unset GIT_CONFIG ;\n mkdir non-bare &&\n cd non-bare &&\n GIT_DIR=.git git init\n@@ -93,7 +93,7 @@ test_expect_success 'GIT_DIR non-bare' '\n test_expect_success 'GIT_DIR & GIT_WORK_TREE (1)' '\n\n (\n- unset GIT_CONFIG &&\n+ unset GIT_CONFIG ;\n mkdir git-dir-wt-1.git &&\n GIT_WORK_TREE=$(pwd) GIT_DIR=git-dir-wt-1.git git init\n ) &&\n@@ -103,7 +103,7 @@ test_expect_success 'GIT_DIR & GIT_WORK_\n test_expect_success 'GIT_DIR & GIT_WORK_TREE (2)' '\n\n if (\n- unset GIT_CONFIG &&\n+ unset GIT_CONFIG ;\n mkdir git-dir-wt-2.git &&\n GIT_WORK_TREE=$(pwd) GIT_DIR=git-dir-wt-2.git git --bare init\n )\n-->8---\n\nI must leave now.\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61925","messageId":"Pine.LNX.4.64.0712041625530.27959@racer.site","threadId":"11128","inReplyTo":"20071204155655.053f4fb4@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T16:28:40Z","receivedAt":"2007-12-04T16:28:40Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 4 Dec 2007, H.Merijn Brand wrote:\n\n> I found it! unset returns false\n\nOh, wow! The question is now: how to deal with it (there are quite a few \nunsets in the test scripts).\n\nIs there a way to override \"unset\" with something like this:\n\n\tunset_real () {\n\t\tunset \"$@\"\n\t}\n\n\tunset () {\n\t\tunset_real \"$@\"\n\t\ttrue\n\t}\n\n?\n\nCiao,\nDscho\n"},{"id":"61933","messageId":"7vve7e49or.fsf@gitster.siamese.dyndns.org","threadId":"11128","inReplyTo":"20071204155655.053f4fb4@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2007-12-04T18:05:40Z","receivedAt":"2007-12-04T18:05:40Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"\"H.Merijn Brand\" writes:\n\n> On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n> ...\n> I found it! unset returns false\n> ...\n> I must leave now.\n\nThanks, you two.\n\nI do not see \"unset VAR... &&\" outside t0001 test, but there are\ninstances of \"unset VAR... &&\" in git-submodule implementations as well.\n\nIn short, not too many places to fix.\n"},{"id":"61955","messageId":"20071204232542.425d2ead@pc09.procura.nl","threadId":"11128","inReplyTo":"7vve7e49or.fsf@gitster.siamese.dyndns.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-04T22:25:42Z","receivedAt":"2007-12-04T22:25:42Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 04 Dec 2007 10:05:40 -0800, Junio C Hamano wrote:\n\n> \"H.Merijn Brand\" writes:\n> \n> > On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n> > ...\n> > I found it! unset returns false\n> > ...\n> > I must leave now.\n> \n> Thanks, you two.\n> \n> I do not see \"unset VAR... &&\" outside t0001 test, but there are\n> instances of \"unset VAR... &&\" in git-submodule implementations as well.\n> \n> In short, not too many places to fix.\n\nGoing from there ....\n\n# make test\n:\n:\n:\n*** t1300-repo-config.sh ***\n* ok 1: initial\n* ok 2: mixed case\n* ok 3: similar section\n* ok 4: similar section\n* ok 5: replace with non-match\n* ok 6: replace with non-match (actually matching)\n* ok 7: non-match result\n* ok 8: multiple unset\n* ok 9: multiple unset is correct\n* ok 10: --replace-all\n* ok 11: all replaced\n* ok 12: really mean test\n* ok 13: really really mean test\n* ok 14: get value\n* ok 15: unset\n* ok 16: multivar\n* FAIL 17: non-match\n git config --get nextsection.nonewline !for\n* FAIL 18: non-match value\n test wow = $(git config --get nextsection.nonewline !for)\n* ok 19: ambiguous get\n* ok 20: get multivar\n* ok 21: multivar replace\n* ok 22: ambiguous value\n* ok 23: ambiguous unset\n* ok 24: invalid unset\n* ok 25: multivar unset\n* ok 26: invalid key\n* ok 27: correct key\n* ok 28: hierarchical section\n* ok 29: hierarchical section value\n* ok 30: working --list\n* ok 31: --get-regexp\n* ok 32: --add\n* ok 33: get variable with no value\n* ok 34: get-regexp variable with no value\n* ok 35: no arguments, but no crash\n* ok 36: new section is partial match of another\n* ok 37: new variable inserts into proper section\n* ok 38: alternative GIT_CONFIG (non-existing file should fail)\n* ok 39: alternative GIT_CONFIG\n* ok 40: alternative GIT_CONFIG (--file)\n* ok 41: --set in alternative GIT_CONFIG\n* ok 42: rename section\n* ok 43: rename succeeded\n* ok 44: rename non-existing section\n* ok 45: rename succeeded\n* ok 46: rename another section\n* ok 47: rename succeeded\n* ok 48: remove section\n* ok 49: section was removed properly\n* ok 50: section ending\n* ok 51: numbers\n* ok 52: bool\n* ok 53: invalid bool (--get)\n* ok 54: invalid bool (set)\n* ok 55: set --bool\n* ok 56: set --int\n* ok 57: quoting\n* ok 58: key with newline\n* ok 59: value with newline\n* ok 60: value continued on next line\n* FAIL 61: --null --list\n cmp result expect\n* FAIL 62: --null --get-regexp\n cmp result expect\n* ok 63: symlinked configuration\n* failed 4 among 63 test(s)\n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61956","messageId":"Pine.LNX.4.64.0712042242310.27959@racer.site","threadId":"11128","inReplyTo":"7vve7e49or.fsf@gitster.siamese.dyndns.org","subject":"[PATCH] Do not rely on the exit status of \"unset\" for unset variables","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T22:45:16Z","receivedAt":"2007-12-04T22:45:16Z","isPatch":true,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"\nFrom: H.Merijn Brand \n\nPOSIX says that exit status \"0\" means that \"unset\" successfully unset\nthe variable. However, it is kind of ambiguous if an environment\nvariable which was not set could be successfully unset.\n\nAt least the default shell on HP-UX insists on reporting an error in\nsuch a case, so just ignore the exit status of \"unset\".\n\n[Dscho: extended the patch to git-submodule.sh, as Junio realized that\n this is the only other place where we check the exit status of \"unset\".]\n\nSigned-off-by: Johannes Schindelin \n---\n\n\tOn Tue, 4 Dec 2007, Junio C Hamano wrote:\n\n\t> \"H.Merijn Brand\" writes:\n\t> \n\t> > On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n\t> > ...\n\t> > I found it! unset returns false\n\t> > ...\n\t> > I must leave now.\n\t> \n\t> Thanks, you two.\n\t> \n\t> I do not see \"unset VAR... &&\" outside t0001 test, but there are\n\t> instances of \"unset VAR... &&\" in git-submodule implementations \n\t> as well.\n\t> \n\t> In short, not too many places to fix.\n\n\tYou're right. I grepped for \"unset\" in t/*.sh, but that catches \n\tonly false positives other than t0001.\n\n\tMerijn, maybe you want to have your sign-off in the commit \n\tmessage?\n\n git-submodule.sh | 10 +++++-----\n t/t0001-init.sh | 16 ++++++++--------\n 2 files changed, 13 insertions(+), 13 deletions(-)\n\ndiff --git a/git-submodule.sh b/git-submodule.sh\nindex 82ac28f..ad9fe62 100755\n--- a/git-submodule.sh\n+++ b/git-submodule.sh\n@@ -158,7 +158,7 @@ module_add()\n \tdie \"'$path' already exists in the index\"\n \n \tmodule_clone \"$path\" \"$realrepo\" || exit\n-\t(unset GIT_DIR && cd \"$path\" && git checkout -q ${branch:+-b \"$branch\" \"origin/$branch\"}) ||\n+\t(unset GIT_DIR; cd \"$path\" && git checkout -q ${branch:+-b \"$branch\" \"origin/$branch\"}) ||\n \tdie \"Unable to checkout submodule '$path'\"\n \tgit add \"$path\" ||\n \tdie \"Failed to add submodule '$path'\"\n@@ -228,14 +228,14 @@ modules_update()\n \t\t\tmodule_clone \"$path\" \"$url\" || exit\n \t\t\tsubsha1=\n \t\telse\n-\t\t\tsubsha1=$(unset GIT_DIR && cd \"$path\" &&\n+\t\t\tsubsha1=$(unset GIT_DIR; cd \"$path\" &&\n \t\t\t\tgit rev-parse --verify HEAD) ||\n \t\t\tdie \"Unable to find current revision in submodule path '$path'\"\n \t\tfi\n \n \t\tif test \"$subsha1\" != \"$sha1\"\n \t\tthen\n-\t\t\t(unset GIT_DIR && cd \"$path\" && git-fetch &&\n+\t\t\t(unset GIT_DIR; cd \"$path\" && git-fetch &&\n \t\t\t\tgit-checkout -q \"$sha1\") ||\n \t\t\tdie \"Unable to checkout '$sha1' in submodule path '$path'\"\n \n@@ -246,7 +246,7 @@ modules_update()\n \n set_name_rev () {\n \trevname=$( (\n-\t\tunset GIT_DIR &&\n+\t\tunset GIT_DIR\n \t\tcd \"$1\" && {\n \t\t\tgit describe \"$2\" 2>/dev/null ||\n \t\t\tgit describe --tags \"$2\" 2>/dev/null ||\n@@ -285,7 +285,7 @@ modules_list()\n \t\telse\n \t\t\tif test -z \"$cached\"\n \t\t\tthen\n-\t\t\t\tsha1=$(unset GIT_DIR && cd \"$path\" && git rev-parse --verify HEAD)\n+\t\t\t\tsha1=$(unset GIT_DIR; cd \"$path\" && git rev-parse --verify HEAD)\n \t\t\t\tset_name_rev \"$path\" \"$sha1\"\n \t\t\tfi\n \t\t\tsay \"+$sha1 $path$revname\"\ndiff --git a/t/t0001-init.sh b/t/t0001-init.sh\nindex b14b3ec..c015405 100755\n--- a/t/t0001-init.sh\n+++ b/t/t0001-init.sh\n@@ -25,7 +25,7 @@ check_config () {\n \n test_expect_success 'plain' '\n \t(\n-\t\tunset GIT_DIR GIT_WORK_TREE &&\n+\t\tunset GIT_DIR GIT_WORK_TREE\n \t\tmkdir plain &&\n \t\tcd plain &&\n \t\tgit init\n@@ -35,7 +35,7 @@ test_expect_success 'plain' '\n \n test_expect_success 'plain with GIT_WORK_TREE' '\n \tif (\n-\t\tunset GIT_DIR &&\n+\t\tunset GIT_DIR\n \t\tmkdir plain-wt &&\n \t\tcd plain-wt &&\n \t\tGIT_WORK_TREE=$(pwd) git init\n@@ -48,7 +48,7 @@ test_expect_success 'plain with GIT_WORK_TREE' '\n \n test_expect_success 'plain bare' '\n \t(\n-\t\tunset GIT_DIR GIT_WORK_TREE GIT_CONFIG &&\n+\t\tunset GIT_DIR GIT_WORK_TREE GIT_CONFIG\n \t\tmkdir plain-bare-1 &&\n \t\tcd plain-bare-1 &&\n \t\tgit --bare init\n@@ -58,7 +58,7 @@ test_expect_success 'plain bare' '\n \n test_expect_success 'plain bare with GIT_WORK_TREE' '\n \tif (\n-\t\tunset GIT_DIR GIT_CONFIG &&\n+\t\tunset GIT_DIR GIT_CONFIG\n \t\tmkdir plain-bare-2 &&\n \t\tcd plain-bare-2 &&\n \t\tGIT_WORK_TREE=$(pwd) git --bare init\n@@ -72,7 +72,7 @@ test_expect_success 'plain bare with GIT_WORK_TREE' '\n test_expect_success 'GIT_DIR bare' '\n \n \t(\n-\t\tunset GIT_CONFIG &&\n+\t\tunset GIT_CONFIG\n \t\tmkdir git-dir-bare.git &&\n \t\tGIT_DIR=git-dir-bare.git git init\n \t) &&\n@@ -82,7 +82,7 @@ test_expect_success 'GIT_DIR bare' '\n test_expect_success 'GIT_DIR non-bare' '\n \n \t(\n-\t\tunset GIT_CONFIG &&\n+\t\tunset GIT_CONFIG\n \t\tmkdir non-bare &&\n \t\tcd non-bare &&\n \t\tGIT_DIR=.git git init\n@@ -93,7 +93,7 @@ test_expect_success 'GIT_DIR non-bare' '\n test_expect_success 'GIT_DIR & GIT_WORK_TREE (1)' '\n \n \t(\n-\t\tunset GIT_CONFIG &&\n+\t\tunset GIT_CONFIG\n \t\tmkdir git-dir-wt-1.git &&\n \t\tGIT_WORK_TREE=$(pwd) GIT_DIR=git-dir-wt-1.git git init\n \t) &&\n@@ -103,7 +103,7 @@ test_expect_success 'GIT_DIR & GIT_WORK_TREE (1)' '\n test_expect_success 'GIT_DIR & GIT_WORK_TREE (2)' '\n \n \tif (\n-\t\tunset GIT_CONFIG &&\n+\t\tunset GIT_CONFIG\n \t\tmkdir git-dir-wt-2.git &&\n \t\tGIT_WORK_TREE=$(pwd) GIT_DIR=git-dir-wt-2.git git --bare init\n \t)\n-- \n1.5.3.7.2139.g2a5a3\n"},{"id":"61957","messageId":"Pine.LNX.4.64.0712042245440.27959@racer.site","threadId":"11128","inReplyTo":"20071204232542.425d2ead@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2007-12-04T22:46:30Z","receivedAt":"2007-12-04T22:46:30Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi,\n\nOn Tue, 4 Dec 2007, H.Merijn Brand wrote:\n\n> * FAIL 17: non-match\n> git config --get nextsection.nonewline !for\n> * FAIL 18: non-match value\n> test wow = $(git config --get nextsection.nonewline !for)\n\nCan you investigate further with \"-i -v\" and/or \"-x\"? Could be the \nexclamation mark...\n\nCiao,\nDscho\n"},{"id":"61981","messageId":"20071205080157.5121bfee@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712042242310.27959@racer.site","subject":"Re: [PATCH] Do not rely on the exit status of \"unset\" for unset variables","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-05T08:01:57Z","receivedAt":"2007-12-05T08:01:57Z","isPatch":true,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 22:45:16 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> From: H.Merijn Brand \n> \n> POSIX says that exit status \"0\" means that \"unset\" successfully unset\n> the variable. However, it is kind of ambiguous if an environment\n> variable which was not set could be successfully unset.\n> \n> At least the default shell on HP-UX insists on reporting an error in\n\nplease, for now make that HP-UX 11.11 and older. I'll check if it also\nfails in 11.23/IPF.\n\nOn 11.11 HP C-ANSI-C cannot be used either.\n\nAnd I have to remove \"#include \" from pager.c on HP-UX, I\nforgot to tell. With the Makefile change in place, building with 64bit\ngcc, \n\n--8<--- skip this part if you're not interested in 64bit builds on HP-UX\nOn 64bit gcc on HP-UX, there is no strtoull (). Nowhere! strtoul () is\nthe same there. But this is only in the 64bit world, so NO_STRTOULL can\nnot be set to YesPlease unconditionally. When I also set NO_STRTOUMAX,\nI get shiploads of warnings like:\n\n CC commit.o\nIn file included from cache.h:4,\n from commit.c:1:\ngit-compat-util.h:166:1: warning: \"strtoumax\" redefined\nIn file included from git-compat-util.h:62,\n from cache.h:4,\n from commit.c:1:\n/usr/include/inttypes.h:527:1: warning: this is the location of the previous\ndefinition\n\nAnd that is NOT your fault, as this include file has defines like\n\n** When compiling in ILP32 mode long long will be used for the 64-bit data\n** types. This will cause compilation errors if 64-bit data types are\n** requested and the compiler in use does not support them.\n\n#define strtoimax(__a, __b, __c) __strtoll(__a, __b, __c)\n#define strtoumax(__a, __b, __c) __strtoull(__a, __b, __c)\n\nand that is not guarded with something like\n\n/* LP64 takes precedence */\n#if (defined(__STDC_EXT__) || defined(_INCLUDE_LONGLONG))\n&& !defined(__LP64__)\n\nso I ended up replacing all strtoumax () to strtoul () in git-fast-import.c\n\nThen I end up with the same error as on 11.00.\n-->8---\n\n\n> such a case, so just ignore the exit status of \"unset\".\n> \n> [Dscho: extended the patch to git-submodule.sh, as Junio realized that\n> this is the only other place where we check the exit status of \"unset\".]\n> \n> Signed-off-by: Johannes Schindelin \n> ---\n> \n> \tOn Tue, 4 Dec 2007, Junio C Hamano wrote:\n> \n> \t> \"H.Merijn Brand\" writes:\n> \t> \n> \t> > On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n> \t> > ...\n> \t> > I found it! unset returns false\n> \t> > ...\n> \t> > I must leave now.\n> \t> \n> \t> Thanks, you two.\n> \t> \n> \t> I do not see \"unset VAR... &&\" outside t0001 test, but there are\n> \t> instances of \"unset VAR... &&\" in git-submodule implementations \n> \t> as well.\n> \t> \n> \t> In short, not too many places to fix.\n> \n> \tYou're right. I grepped for \"unset\" in t/*.sh, but that catches \n> \tonly false positives other than t0001.\n> \n> \tMerijn, maybe you want to have your sign-off in the commit \n> \tmessage?\n\nFeel free to do so, I don't really care.\n\nWill you also be looking into the install issue?\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61982","messageId":"20071205080822.0f41379b@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712042245440.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-05T08:08:22Z","receivedAt":"2007-12-05T08:08:22Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 22:46:30 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > * FAIL 17: non-match\n> > git config --get nextsection.nonewline !for\n> > * FAIL 18: non-match value\n> > test wow = $(git config --get nextsection.nonewline !for)\n> \n> Can you investigate further with \"-i -v\" and/or \"-x\"? Could be the \n> exclamation mark...\n\n* expecting success: git config --get nextsection.nonewline !for\n* FAIL 17: non-match\n git config --get nextsection.nonewline !for\n\n+ git config nextsection.NoNewLine wow2 for me for me$\n+ cat\n+ 1> expect 0< /var/tmp/sh4066.13\n+ test_expect_success multivar cmp .git/config expect\n* expecting success: cmp .git/config expect\n* ok 16: multivar\n\n+ test_expect_success non-match git config --get nextsection.nonewline !for\n* expecting success: git config --get nextsection.nonewline !for\n* FAIL 17: non-match\n git config --get nextsection.nonewline !for\n\n\n> Ciao,\n> Dscho\n> \n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"61998","messageId":"20071205104911.06a77de7@pc09.procura.nl","threadId":"11128","inReplyTo":"Pine.LNX.4.64.0712041625530.27959@racer.site","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-05T10:49:11Z","receivedAt":"2007-12-05T10:49:11Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 4 Dec 2007 16:28:40 +0000 (GMT), Johannes Schindelin\n wrote:\n\n> Hi,\n> \n> On Tue, 4 Dec 2007, H.Merijn Brand wrote:\n> \n> > I found it! unset returns false\n> \n> Oh, wow! The question is now: how to deal with it (there are quite a few \n> unsets in the test scripts).\n\nHP-UX 11.23/IPF 64bit gcc-4.2.1\n\nenv CC=gcc configure --disable-nls --prefix=/pro/local --without-iconv --with-perl=/pro/bin/perl\n\nifeq ($(uname_S),HP-UX)\n\tBASIC_CFLAGS += -mlp64\n\tNO_UNSETENV = YesPlease\nendif\n\nneeds the unset patch\n\nfails at the same point.\n\nSo at least all of HP-UX is consistent, which means that any of the HP\ntestdrive systems might be worth looking at for you lot.\n\nhttp://www.testdrive.hp.com/current.shtml\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"62579","messageId":"20071210145123.7c34af6d@pc09.procura.nl","threadId":"11128","inReplyTo":"7vve7e49or.fsf@gitster.siamese.dyndns.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-10T14:51:23Z","receivedAt":"2007-12-10T14:51:23Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 04 Dec 2007 10:05:40 -0800, Junio C Hamano wrote:\n\n> \"H.Merijn Brand\" writes:\n> \n> > On Tue, 4 Dec 2007 15:39:47 +0000 (GMT), Johannes Schindelin\n> > ...\n> > I found it! unset returns false\n> > ...\n> > I must leave now.\n> \n> Thanks, you two.\n> \n> I do not see \"unset VAR... &&\" outside t0001 test, but there are\n> instances of \"unset VAR... &&\" in git-submodule implementations as well.\n> \n> In short, not too many places to fix.\n\nI took the snapshot from 10-12-2007 and applied the patch below\n\nSummary of the changes:\n\n1 Added a section for HP-UX in the Makefile. Note that this will\n cover most of HP-UX, but might need several changes for both\n newer HP-UX versions and 64bit environments. Will come to that\n once I've got it all running\n\n2 HP-UX does not have . I mentioned this before\n\n3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n at least does not conform to the expectations in the GNU world,\n but chickening out like the way strbuf_addf () does is maybe a\n bit too rude, so I forced a minimum of 64bytes available. That\n fixes a lot!\n\n but it still breaks t4013 :(\n\n4 'tr' doesn't like '\\0', but wants '\\000' instead.\n\n5 'tr' cannot deal with character classes\n\n6 I don't know how to attack another HP specific problem: HP has\n a system command called 'patch' and it is *very* incompatible\n with GNU patch. Very. On all my systems I have GNU patch as\n 'npatch', but I also know of people that have it as 'gpatch'.\n In whatever case, HP's patch will fail, so you'd have to ask\n if the poor user does have a GNU patch lying around. In t4109\n I changed the three calls to patch to calls to npatch, but did\n not include this in the patch below. Same for the single call\n in t4110.\n\n7 What do you expect for LOCALE's? below is the first failure for\n locale testing, but there are more, like\n\n *** t4201-shortlog.sh ***\n t4201-shortlog.sh[22]: Invalid multibyte character.\n\n and somewhere inside t5100-mailinfo.sh:\n fatal: cannot convert from iso-2022-jp to utf-8\n\nAfter working my way down, I stopped after hitting this:\n\n+ git-index-pack -o tmp.idx test-2-7f8ead892057e78576c0329a70cc83afb113f117.pack\nfatal: serious inflate inconsistency\n\ninside t5300-pack-object.sh\n\n# make configure\n:\n# env CC=gcc configure --prefix=/pro/local --disable-nls --without-iconv \\\n --with-perl=/pro/bin/perl\n:\n# make\n:\n# make test\n:\n:\n*** t3900-i18n-commit.sh ***\n* ok 1: setup\n* ok 2: no encoding header for base case\n* ok 3: ISO-8859-1 setup\n* ok 4: EUCJP setup\n* ok 5: ISO-2022-JP setup\n* ok 6: check encoding header for ISO-8859-1\n* ok 7: check encoding header for EUCJP\n* ok 8: check encoding header for ISO-2022-JP\n* ok 9: config to remove customization\n* FAIL 10: ISO-8859-1 should be shown in UTF-8 now\n\n compare_with ISO-8859-1 ../t3900/1-UTF-8.txt\n\n* FAIL 11: EUCJP should be shown in UTF-8 now\n\n compare_with EUCJP ../t3900/2-UTF-8.txt\n\n* FAIL 12: ISO-2022-JP should be shown in UTF-8 now\n\n compare_with ISO-2022-JP ../t3900/2-UTF-8.txt\n\n* ok 13: config to add customization\n* ok 14: ISO-8859-1 should be shown in itself now\n* ok 15: EUCJP should be shown in itself now\n* ok 16: ISO-2022-JP should be shown in itself now\n* ok 17: config to tweak customization\n* FAIL 18: ISO-8859-1 should be shown in UTF-8 now\n\n compare_with ISO-8859-1 ../t3900/1-UTF-8.txt\n\n* FAIL 19: EUCJP should be shown in UTF-8 now\n\n compare_with EUCJP ../t3900/2-UTF-8.txt\n\n* FAIL 20: ISO-2022-JP should be shown in UTF-8 now\n\n compare_with ISO-2022-JP ../t3900/2-UTF-8.txt\n\n* ok 21: EUCJP should be shown in EUCJP now\n* FAIL 22: ISO-2022-JP should be shown in EUCJP now\n\n compare_with ISO-2022-JP ../t3900/EUCJP.txt\n\n* FAIL 23: EUCJP should be shown in ISO-2022-JP now\n\n compare_with EUCJP ../t3900/ISO-2022-JP.txt\n\n* ok 24: ISO-2022-JP should be shown in ISO-2022-JP now\n* ok 25: No conversion with ISO-8859-1\n* ok 26: No conversion with EUCJP\n* ok 27: No conversion with ISO-2022-JP\n* failed 8 among 27 test(s)\nmake[1]: *** [t3900-i18n-commit.sh] Error 1\n\n# locale -a\nC\nPOSIX\nC.iso88591\nC.utf8\nuniv.utf8\nnl_NL.iso88591\nnl_NL.roman8\nen_GB.iso88591\nen_GB.roman8\nen_US.iso88591\nen_US.roman8\nC.iso885915\nnl_NL.iso885915@euro\nen_GB.iso885915@euro\n\n--8<--- git-hpux.diff\ndiff -pur git-2007-12-10_01/Makefile git-2007-12-10/Makefile\n--- git-2007-12-10_01/Makefile 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/Makefile 2007-12-10 13:41:39 +0100\n@@ -398,6 +398,18 @@ EXTLIBS =\n # because maintaining the nesting to match is a pain. If\n # we had \"elif\" things would have been much nicer...\n\n+ifeq ($(uname_S),HP-UX)\n+ # HP-UX\n+ BASIC_LDFLAGS += -L/usr/local/ssl/lib -L/usr/local/lib\n+ #EXTLIBS += -lc_r -lgcc\n+ NO_HSTRERROR = YesPlease\n+ NO_ICONV = YesPlease\n+ NO_INET_NTOP = YesPlease\n+ NO_INET_PTON = YesPlease\n+ NO_UNSETENV = YesPlease\n+ NO_STRTOULL = YesPlease\n+ NO_STRTOUMAX = YesPlease\n+endif\n ifeq ($(uname_S),Linux)\n NO_STRLCPY = YesPlease\n endif\ndiff -pur git-2007-12-10_01/git-compat-util.h git-2007-12-10/git-compat-util.h\n--- git-2007-12-10_01/git-compat-util.h 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/git-compat-util.h 2007-12-10 14:25:26 +0100\n@@ -68,7 +68,9 @@\n #include \n #include \n #include \n+#ifndef _HPUX_SOURCE\n #include \n+#endif\n #include \n #include \n #include \ndiff -pur git-2007-12-10_01/strbuf.c git-2007-12-10/strbuf.c\n--- git-2007-12-10_01/strbuf.c 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/strbuf.c 2007-12-10 15:19:46 +0100\n@@ -12,7 +12,7 @@ void strbuf_init(struct strbuf *sb, size\n sb->alloc = sb->len = 0;\n sb->buf = strbuf_slopbuf;\n if (hint)\n- strbuf_grow(sb, hint);\n+ strbuf_grow(sb, hint);\n }\n\n void strbuf_release(struct strbuf *sb)\n@@ -118,7 +118,7 @@ void strbuf_addf(struct strbuf *sb, cons\n int len;\n va_list ap;\n\n- if (!strbuf_avail(sb))\n+ if (strbuf_avail(sb) < 64)\n strbuf_grow(sb, 64);\n va_start(ap, fmt);\n len = vsnprintf(sb->buf + sb->len, sb->alloc - sb->len, fmt, ap);\ndiff -pur git-2007-12-10_01/t/t0020-crlf.sh git-2007-12-10/t/t0020-crlf.sh\n--- git-2007-12-10_01/t/t0020-crlf.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t0020-crlf.sh 2007-12-10 14:25:58 +0100\n@@ -5,7 +5,7 @@ test_description='CRLF conversion'\n . ./test-lib.sh\n\n q_to_nul () {\n- tr Q '\\0'\n+ tr Q '\\000'\n }\n\n append_cr () {\ndiff -pur git-2007-12-10_01/t/t1300-repo-config.sh git-2007-12-10/t/t1300-repo-config.sh\n--- git-2007-12-10_01/t/t1300-repo-config.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t1300-repo-config.sh 2007-12-10 09:49:48 +0100\n@@ -591,12 +591,12 @@ Qsection.sub=section.val4\n Qsection.sub=section.val5Q\n EOF\n\n-git config --null --list | tr '[\\000]' 'Q' > result\n+git config --null --list | tr '\\000' 'Q' > result\n echo >>result\n\n test_expect_success '--null --list' 'cmp result expect'\n\n-git config --null --get-regexp 'val[0-9]' | tr '[\\000]' 'Q' > result\n+git config --null --get-regexp 'val[0-9]' | tr '\\000' 'Q' > result\n echo >>result\n\n test_expect_success '--null --get-regexp' 'cmp result expect'\n-->8--- git-hpux.diff\n\nCan I put the ball back in your court now?\n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"62685","messageId":"7vfxy9fxh3.fsf@gitster.siamese.dyndns.org","threadId":"11128","inReplyTo":"20071210145123.7c34af6d@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Junio C Hamano","fromEmail":"gitster@pobox.com","sentAt":"2007-12-11T08:26:48Z","receivedAt":"2007-12-11T08:26:48Z","isPatch":false,"sender":{"key":"gitster@pobox.com","avatar":"https://avatars.githubusercontent.com/u/54884?v=4"},"body":"\"H.Merijn Brand\" writes:\n\n> Summary of the changes:\n>\n> 1 Added a section for HP-UX in the Makefile. Note that this will\n> cover most of HP-UX, but might need several changes for both\n> newer HP-UX versions and 64bit environments. Will come to that\n> once I've got it all running\n>\n> 2 HP-UX does not have . I mentioned this before\n\nUnfortunate. It is even in POSIX.\nBut nothing \"make -DNO_SYS_SELECT=Unfortunately\" cannot fix.\n\n> 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n> at least does not conform to the expectations in the GNU world,\n> but chickening out like the way strbuf_addf () does is maybe a\n> bit too rude, so I forced a minimum of 64bytes available. That\n> fixes a lot!\n>\n> but it still breaks t4013 :(\n\nI think Shawn had something similar for Solaris, and if it is a small\ncost to help portability, it does not feel so bad.\n\n> 4 'tr' doesn't like '\\0', but wants '\\000' instead.\n> 5 'tr' cannot deal with character classes\n\nOk, that's odd (and not old fashioned, which makes even odder), but\n\"\\000\" is not too bad (unless we have to do that million places).\n\n> 6 I don't know how to attack another HP specific problem: HP has\n> a system command called 'patch' and it is *very* incompatible\n> with GNU patch. Very.\n\nI do not think there is any reason for us to rely on \"GNU patch\"\nanymore. 4109 can have precomputed test vector instead of comparing our\noutput with what GNU patch of the day happens to do.\n\n> 7 What do you expect for LOCALE's?\n\nI am not opposed to a change that makes these tests to conditionally\nskip on systems (not necessarily \"platforms\" but individual\n\"installations\") that do not have locales necessary to run tests\ninstalled, just like cvs or svn tests are skipped when they are not\ninstalled. Especially you are not using iconv...\n\n> + git-index-pack -o tmp.idx test-2-7f8ead892057e78576c0329a70cc83afb113f117.pack\n> fatal: serious inflate inconsistency\n\nThat sounds like a broken zlib X-<.\n"},{"id":"62687","messageId":"20071211085440.GK14735@spearce.org","threadId":"11128","inReplyTo":"7vfxy9fxh3.fsf@gitster.siamese.dyndns.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Shawn O. Pearce","fromEmail":"spearce@spearce.org","sentAt":"2007-12-11T08:54:40Z","receivedAt":"2007-12-11T08:54:40Z","isPatch":false,"sender":{"key":"spearce@spearce.org","avatar":"https://avatars.githubusercontent.com/u/34844?v=4"},"body":"Junio C Hamano wrote:\n> \"H.Merijn Brand\" writes:\n> \n> > 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n> > at least does not conform to the expectations in the GNU world,\n> > but chickening out like the way strbuf_addf () does is maybe a\n> > bit too rude, so I forced a minimum of 64bytes available. That\n> > fixes a lot!\n> >\n> > but it still breaks t4013 :(\n> \n> I think Shawn had something similar for Solaris, and if it is a small\n> cost to help portability, it does not feel so bad.\n\nMy Solaris change is already in your tree; on Solaris vnsprintf\nwill return -1 if the input buffer we gave it had a length of 0.\nThis can happen at odd times, depending on what is already in the\nstrbuf and how its buffer is. :)\n\nThe fix for Solaris turned out to be just make sure there is at\nleast 1 byte in the buffer before calling vsnprintf().\n \n> > + git-index-pack -o tmp.idx test-2-7f8ead892057e78576c0329a70cc83afb113f117.pack\n> > fatal: serious inflate inconsistency\n> \n> That sounds like a broken zlib X-<.\n\nI think this is actually a broken pread(). Its been reported\nmultiple times by different people on HP-UX. Compiling with\nNO_PREAD=1 resolves the issues for everyone who has tried it.\n\nJakub asked me for an autoconf test to look for this sort of\nbreakage in pread(), I have yet to come up with a theory for\nwhat sort of breakage HP-UX is showing us here, let alone write\na test for it.\n\n-- \nShawn.\n"},{"id":"62689","messageId":"475E5677.9090507@viscovery.net","threadId":"11128","inReplyTo":"7vfxy9fxh3.fsf@gitster.siamese.dyndns.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Sixt","fromEmail":"j.sixt@viscovery.net","sentAt":"2007-12-11T09:20:55Z","receivedAt":"2007-12-11T09:20:55Z","isPatch":false,"sender":{"key":"j6t@kdbg.org","avatar":"https://avatars.githubusercontent.com/u/14810926?v=4"},"body":"Junio C Hamano schrieb:\n> \"H.Merijn Brand\" writes:\n>> 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n>> at least does not conform to the expectations in the GNU world,\n>> but chickening out like the way strbuf_addf () does is maybe a\n>> bit too rude, so I forced a minimum of 64bytes available. That\n>> fixes a lot!\n>>\n>> but it still breaks t4013 :(\n> \n> I think Shawn had something similar for Solaris, and if it is a small\n> cost to help portability, it does not feel so bad.\n\nWindows's vsnprintf() is broken in two regards: (1) It returns -1 if the\nbuffer is too small; (2) the size parameter is not the size of the buffer,\nbut the max. number of chars to write (i.e. does not count the NUL). How\nbroken is HP-UX's? Maybe we can share the implementation? See:\n\nhttp://repo.or.cz/w/git/mingw.git?a=commitdiff;h=5fd41ffacdef5454acbe51f5e23a97eb5158226d\nhttp://repo.or.cz/w/git/mingw.git?a=commitdiff;h=e88aa8cbe816526bb0a7d37eaf2f5eb40ff36ae1\n\n-- Hannes\n"},{"id":"62697","messageId":"20071211114251.78ae357a@pc09.procura.nl","threadId":"11128","inReplyTo":"475E5677.9090507@viscovery.net","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-11T10:42:51Z","receivedAt":"2007-12-11T10:42:51Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 11 Dec 2007 10:20:55 +0100, Johannes Sixt \nwrote:\n\n> Junio C Hamano schrieb:\n> > \"H.Merijn Brand\" writes:\n> >> 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n> >> at least does not conform to the expectations in the GNU world,\n> >> but chickening out like the way strbuf_addf () does is maybe a\n> >> bit too rude, so I forced a minimum of 64bytes available. That\n> >> fixes a lot!\n> >>\n> >> but it still breaks t4013 :(\n> > \n> > I think Shawn had something similar for Solaris, and if it is a small\n> > cost to help portability, it does not feel so bad.\n> \n> Windows's vsnprintf() is broken in two regards: (1) It returns -1 if the\n> buffer is too small; (2) the size parameter is not the size of the buffer,\n> but the max. number of chars to write (i.e. does not count the NUL). How\n> broken is HP-UX's?\n\nDo you have a simple test program to show how `broken' it is?\nI have 10.20, 11.00, 11.11, 11.23/PA and 11.23/IPF and both\nHP C-ANSI-C and gcc\n\n> Maybe we can share the implementation? See:\n> \n> http://repo.or.cz/w/git/mingw.git?a=commitdiff;h=5fd41ffacdef5454acbe51f5e23a97eb5158226d\n> http://repo.or.cz/w/git/mingw.git?a=commitdiff;h=e88aa8cbe816526bb0a7d37eaf2f5eb40ff36ae1\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"62698","messageId":"475E6E07.60904@viscovery.net","threadId":"11128","inReplyTo":"20071211114251.78ae357a@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Sixt","fromEmail":"j.sixt@viscovery.net","sentAt":"2007-12-11T11:01:27Z","receivedAt":"2007-12-11T11:01:27Z","isPatch":false,"sender":{"key":"j6t@kdbg.org","avatar":"https://avatars.githubusercontent.com/u/14810926?v=4"},"body":"H.Merijn Brand schrieb:\n> On Tue, 11 Dec 2007 10:20:55 +0100, Johannes Sixt \n> wrote:\n> \n>> Junio C Hamano schrieb:\n>>> \"H.Merijn Brand\" writes:\n>>>> 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n>>>> at least does not conform to the expectations in the GNU world,\n>>>> but chickening out like the way strbuf_addf () does is maybe a\n>>>> bit too rude, so I forced a minimum of 64bytes available. That\n>>>> fixes a lot!\n>>>>\n>>>> but it still breaks t4013 :(\n>>> I think Shawn had something similar for Solaris, and if it is a small\n>>> cost to help portability, it does not feel so bad.\n>> Windows's vsnprintf() is broken in two regards: (1) It returns -1 if the\n>> buffer is too small; (2) the size parameter is not the size of the buffer,\n>> but the max. number of chars to write (i.e. does not count the NUL). How\n>> broken is HP-UX's?\n> \n> Do you have a simple test program to show how `broken' it is?\n> I have 10.20, 11.00, 11.11, 11.23/PA and 11.23/IPF and both\n> HP C-ANSI-C and gcc\n\nSorry, no. vsnprintf's behavior on Windows is actually documented, so to be\nfair it is not \"broken\", but \"unwanted\".\n\n-- Hannes\n"},{"id":"62710","messageId":"20071211135742.62d1d641@pc09.procura.nl","threadId":"11128","inReplyTo":"20071211085440.GK14735@spearce.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-11T12:57:42Z","receivedAt":"2007-12-11T12:57:42Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 11 Dec 2007 03:54:40 -0500, \"Shawn O. Pearce\" \nwrote:\n\n> Junio C Hamano wrote:\n> > \"H.Merijn Brand\" writes:\n> > \n> > > 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n> > > at least does not conform to the expectations in the GNU world,\n> > > but chickening out like the way strbuf_addf () does is maybe a\n> > > bit too rude, so I forced a minimum of 64bytes available. That\n> > > fixes a lot!\n> > >\n> > > but it still breaks t4013 :(\n> > \n> > I think Shawn had something similar for Solaris, and if it is a small\n> > cost to help portability, it does not feel so bad.\n> \n> My Solaris change is already in your tree; on Solaris vnsprintf\n> will return -1 if the input buffer we gave it had a length of 0.\n> This can happen at odd times, depending on what is already in the\n> strbuf and how its buffer is. :)\n> \n> The fix for Solaris turned out to be just make sure there is at\n> least 1 byte in the buffer before calling vsnprintf().\n> \n> > > + git-index-pack -o tmp.idx test-2-7f8ead892057e78576c0329a70cc83afb113f117.pack\n> > > fatal: serious inflate inconsistency\n> > \n> > That sounds like a broken zlib X-<.\n\nlibz is version 1.2.3\n\n> I think this is actually a broken pread(). Its been reported\n> multiple times by different people on HP-UX. Compiling with\n> NO_PREAD=1 resolves the issues for everyone who has tried it.\n\n:)\n\n* passed all 114 test(s)\n\n> Jakub asked me for an autoconf test to look for this sort of\n> breakage in pread(), I have yet to come up with a theory for\n> what sort of breakage HP-UX is showing us here, let alone write\n> a test for it.\n\nFor now, the HP section in Makefile for 11.00 now looks like\n\nifeq ($(uname_S),HP-UX)\n\t# HP-UX\n\tBASIC_LDFLAGS += -L/usr/local/ssl/lib -L/usr/local/lib\n#\tEXTLIBS += -lc_r -lgcc\n\tNO_HSTRERROR\t= YesPlease\n\tNO_ICONV\t= YesPlease\n\tNO_INET_NTOP\t= YesPlease\n\tNO_INET_PTON\t= YesPlease\n\tNO_UNSETENV\t= YesPlease\n\tNO_STRTOULL\t= YesPlease\n\tNO_STRTOUMAX\t= YesPlease\n\tNO_PREAD\t= YesPlease\nendif\n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"62714","messageId":"20071211143323.27c860db@pc09.procura.nl","threadId":"11128","inReplyTo":"7vfxy9fxh3.fsf@gitster.siamese.dyndns.org","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"H.Merijn Brand","fromEmail":"h.m.brand@xs4all.nl","sentAt":"2007-12-11T13:33:23Z","receivedAt":"2007-12-11T13:33:23Z","isPatch":false,"sender":{"key":"h.m.brand@xs4all.nl","avatar":"https://gravatar.com/avatar/5b8f83ee35c427a646cbea3b104346e00ab3663b99bbf435cddeb75cd4b3857b?d=mp&s=160"},"body":"On Tue, 11 Dec 2007 00:26:48 -0800, Junio C Hamano wrote:\n\n> \"H.Merijn Brand\" writes:\n> \n> > Summary of the changes:\n> >\n> > 1 Added a section for HP-UX in the Makefile. Note that this will\n> > cover most of HP-UX, but might need several changes for both\n> > newer HP-UX versions and 64bit environments. Will come to that\n> > once I've got it all running\n> >\n> > 2 HP-UX does not have . I mentioned this before\n> \n> Unfortunate. It is even in POSIX.\n\nThe *include* file, or the function? HP-UX has the function, but\nnot the include file.\n\n> But nothing \"make -DNO_SYS_SELECT=Unfortunately\" cannot fix.\n\nI didn't see any #ifdef/#ifndef around the include, otherwise I\nwould have found that already I think.\n\n> > 3 I am willing to believe that HP_UX' vsnprintf () is broken, or\n> > at least does not conform to the expectations in the GNU world,\n> > but chickening out like the way strbuf_addf () does is maybe a\n> > bit too rude, so I forced a minimum of 64bytes available. That\n> > fixes a lot!\n> >\n> > but it still breaks t4013 :(\n> \n> I think Shawn had something similar for Solaris, and if it is a small\n> cost to help portability, it does not feel so bad.\n\n:)\n\n> > 4 'tr' doesn't like '\\0', but wants '\\000' instead.\n> > 5 'tr' cannot deal with character classes\n> \n> Ok, that's odd (and not old fashioned, which makes even odder), but\n> \"\\000\" is not too bad (unless we have to do that million places).\n\n> perl -MFile::Find -e'find(sub{m/\\.sh$/ or return;local@ARGV=($_);while(<>){/\\btr\\b.*\\\\\\d(?\\!\\d\\d)/&&print\"$File::Find::name\\t$.\\t$_\"}},\".\")'\n./git-filter-branch.sh 293 git diff-index -r $commit | cut -f 2- | tr '\\n' '\\0' | \\\n./test-sha1.sh 13 tr '[\\0]' '[g]'\n./test-sha1.sh 58 tr '[\\0]' '[g]'\n./t/diff-lib.sh 24 tr '\\0' '\\012' <\"$1\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-1\n./t/diff-lib.sh 25 tr '\\0' '\\012' <\"$2\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-2\n./t/t2001-checkout-cache-clash.sh 33 sed -e 's/^\\([0-9]*\\) [^ ]* [0-9a-f]* /tr: \\1 /'\n./t/t3300-funny-names.sh 57 'git ls-files -z | tr \\\\0 \\\\012 >current &&\n./t/t3300-funny-names.sh 86 'git diff-index -z --name-status $t0 | tr \\\\0 \\\\012 >current &&\n./t/t3300-funny-names.sh 90 'git diff-tree -z --name-status $t0 $t1 | tr \\\\0 \\\\012 >current &&\n./t/t4020-diff-external.sh 102 echo NULZbetweenZwords | tr Z '\\0' > file\n./t/t4103-apply-binary.sh 27 tr 'x' '\\0' file3\n./t/t4103-apply-binary.sh 30 tr '\\0' 'v' file1\n./t/t4116-apply-reverse.sh 15 tr \"[ijk]\" '\\''[\\0\\1\\2]'\\'' file2 &&\n./t/t4116-apply-reverse.sh 22 tr \"[mon]\" '\\''[\\0\\1\\2]'\\'' file2 &&\n./t/t4200-rerere.sh 132 echo \"$sha1 a1\" | tr '\\012' '\\0' > .git/rr-cache/MERGE_RR\n./t/t5300-pack-object.sh 18 dd if=/dev/zero bs=4k count=1 | tr \"\\\\0\" $i >$i &&\n\nSo with the recent history in the back of my mind, I'd expect all tr commands\nwith character class like notations to fail. Patch below.\n\n> > 6 I don't know how to attack another HP specific problem: HP has\n> > a system command called 'patch' and it is *very* incompatible\n> > with GNU patch. Very.\n> \n> I do not think there is any reason for us to rely on \"GNU patch\"\n> anymore. 4109 can have precomputed test vector instead of comparing our\n> output with what GNU patch of the day happens to do.\n\n\\o/\n\n> > 7 What do you expect for LOCALE's?\n> \n> I am not opposed to a change that makes these tests to conditionally\n> skip on systems (not necessarily \"platforms\" but individual\n> \"installations\") that do not have locales necessary to run tests\n> installed, just like cvs or svn tests are skipped when they are not\n> installed. Especially you are not using iconv...\n\n\\o/\n\nBTW it is rather unlikely to find Tcl and/or Tk on most HP-UX systems.\nHaving perl already required, would you cheer an attempt to rewrite\ngit-gui to Perl/Tk? (I didn't have a look at it yet, but that should\nmake git-gui an awful lot more portable and when used on windows with\nActivePerl, Tk is already available in all builds)\n\n\nI have put the output of\n\n% make -i test | & perl -ne'm/^\\*\\s+ok\\s+\\d+:/ or print' > git-make-test.log\n\non http://www.xs4all.nl/~procura/git-make-test.log\n\nwhich leaves only the following tests as failures:\n\nt3901-i18n-patch.sh t4201-shortlog.sh t5100-mailinfo.sh\nt5405-send-pack-rewind.sh t5500-fetch-pack.sh t5505-remote.sh\nt5510-fetch.sh t5515-fetch-merge-logic.sh t5516-fetch-push.sh\nt5520-pull.sh t5600-clone-fail-cleanup.sh t5700-clone-reference.sh\nt5701-clone-local.sh t6002-rev-list-bisect.sh t6003-rev-list-topo-order.sh\nt6022-merge-rename.sh t6026-merge-attr.sh t6200-fmt-merge-msg.sh\nt7003-filter-branch.sh t9300-fast-import.sh t9301-fast-export.sh\nt9500-gitweb-standalone-no-errors.sh\n\nHow close am I to success?\n\n\n\n--8<--- git-tr.diff\n--- git-2007-12-10_01/git-filter-branch.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/git-filter-branch.sh 2007-12-11 13:39:02 +0100\n@@ -290,7 +290,7 @@ while read commit parents; do\n eval \"$filter_tree\" < /dev/null ||\n die \"tree filter failed: $filter_tree\"\n\n- git diff-index -r $commit | cut -f 2- | tr '\\n' '\\0' | \\\n+ git diff-index -r $commit | cut -f 2- | tr '\\n' '\\000' | \\\n xargs -0 git update-index --add --replace --remove\n git ls-files -z --others | \\\n xargs -0 git update-index --add --replace --remove\n--- git-2007-12-10_01/test-sha1.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/test-sha1.sh 2007-12-11 13:39:29 +0100\n@@ -10,7 +10,7 @@ do\n {\n test -z \"$pfx\" || echo \"$pfx\"\n dd if=/dev/zero bs=1048576 count=$cnt 2>/dev/null |\n- tr '[\\0]' '[g]'\n+ tr '\\000' 'g'\n } | ./test-sha1 $cnt\n `\n if test \"$expect\" = \"$actual\"\n@@ -55,7 +55,7 @@ do\n {\n test -z \"$pfx\" || echo \"$pfx\"\n dd if=/dev/zero bs=1048576 count=$cnt 2>/dev/null |\n- tr '[\\0]' '[g]'\n+ tr '\\000' 'g'\n } | sha1sum |\n sed -e 's/ .*//'\n `\n--- git-2007-12-10_01/t/diff-lib.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/diff-lib.sh 2007-12-11 13:39:56 +0100\n@@ -21,8 +21,8 @@ compare_diff_raw_z () {\n # Also we do not check SHA1 hash generation in this test, which\n # is a job for t0000-basic.sh\n\n- tr '\\0' '\\012' <\"$1\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-1\n- tr '\\0' '\\012' <\"$2\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-2\n+ tr '\\000' '\\012' <\"$1\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-1\n+ tr '\\000' '\\012' <\"$2\" | sed -e \"$sanitize_diff_raw_z\" >.tmp-2\n git diff .tmp-1 .tmp-2 && rm -f .tmp-1 .tmp-2\n }\n\n--- git-2007-12-10_01/t/t3300-funny-names.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t3300-funny-names.sh 2007-12-11 13:40:32 +0100\n@@ -54,7 +54,7 @@ echo 'just space\n no-funny\n tabs ,\" (dq) and spaces' >expected\n test_expect_success 'git ls-files -z with-funny' \\\n- 'git ls-files -z | tr \\\\0 \\\\012 >current &&\n+ 'git ls-files -z | tr \\\\000 \\\\012 >current &&\n git diff expected current'\n\n t1=`git write-tree`\n@@ -83,11 +83,11 @@ test_expect_success 'git diff-tree with-\n echo 'A\n tabs ,\" (dq) and spaces' >expected\n test_expect_success 'git diff-index -z with-funny' \\\n- 'git diff-index -z --name-status $t0 | tr \\\\0 \\\\012 >current &&\n+ 'git diff-index -z --name-status $t0 | tr \\\\000 \\\\012 >current &&\n git diff expected current'\n\n test_expect_success 'git diff-tree -z with-funny' \\\n- 'git diff-tree -z --name-status $t0 $t1 | tr \\\\0 \\\\012 >current &&\n+ 'git diff-tree -z --name-status $t0 $t1 | tr \\\\000 \\\\012 >current &&\n git diff expected current'\n\n cat > expected <<\\EOF\n--- git-2007-12-10_01/t/t4020-diff-external.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t4020-diff-external.sh 2007-12-11 13:40:44 +0100\n@@ -99,7 +99,7 @@ test_expect_success 'no diff with -diff'\n git diff | grep Binary\n '\n\n-echo NULZbetweenZwords | tr Z '\\0' > file\n+echo NULZbetweenZwords | tr Z '\\000' > file\n\n test_expect_success 'force diff with \"diff\"' '\n echo >.gitattributes \"file diff\" &&\n--- git-2007-12-10_01/t/t4103-apply-binary.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t4103-apply-binary.sh 2007-12-11 13:40:57 +0100\n@@ -24,10 +24,10 @@ git update-index --add --remove file1 fi\n git-commit -m 'Initial Version' 2>/dev/null\n\n git-checkout -b binary\n-tr 'x' '\\0' file3\n+tr 'x' '\\000' file3\n cat file3 >file4\n git add file2\n-tr '\\0' 'v' file1\n+tr '\\000' 'v' file1\n rm -f file2\n git update-index --add --remove file1 file2 file3 file4\n git-commit -m 'Second Version'\n--- git-2007-12-10_01/t/t4116-apply-reverse.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t4116-apply-reverse.sh 2007-12-11 13:42:13 +0100\n@@ -12,14 +12,14 @@ test_description='git apply in reverse\n test_expect_success setup '\n\n for i in a b c d e f g h i j k l m n; do echo $i; done >file1 &&\n- tr \"[ijk]\" '\\''[\\0\\1\\2]'\\'' file2 &&\n+ tr \"ijk\" '\\''\\000\\001\\002'\\'' file2 &&\n\n git add file1 file2 &&\n git commit -m initial &&\n git tag initial &&\n\n for i in a b c g h i J K L m o n p q; do echo $i; done >file1 &&\n- tr \"[mon]\" '\\''[\\0\\1\\2]'\\'' file2 &&\n+ tr \"mon\" '\\''\\000\\001\\002'\\'' file2 &&\n\n git commit -a -m second &&\n git tag second &&\n--- git-2007-12-10_01/t/t4200-rerere.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t4200-rerere.sh 2007-12-11 13:42:28 +0100\n@@ -129,7 +129,7 @@ test_expect_success 'rerere kicked in' \"\n test_expect_success 'rerere prefers first change' 'git diff a1 expect'\n\n rm $rr/postimage\n-echo \"$sha1 a1\" | tr '\\012' '\\0' > .git/rr-cache/MERGE_RR\n+echo \"$sha1 a1\" | tr '\\012' '\\000' > .git/rr-cache/MERGE_RR\n\n test_expect_success 'rerere clear' 'git rerere clear'\n\n--- git-2007-12-10_01/t/t5300-pack-object.sh 2007-12-09 10:23:48 +0100\n+++ git-2007-12-10/t/t5300-pack-object.sh 2007-12-11 13:42:46 +0100\n@@ -15,7 +15,7 @@ test_expect_success \\\n 'rm -f .git/index*\n for i in a b c\n do\n- dd if=/dev/zero bs=4k count=1 | tr \"\\\\0\" $i >$i &&\n+ dd if=/dev/zero bs=4k count=1 | tr \"\\\\000\" $i >$i &&\n git update-index --add $i || return 1\n done &&\n cat c >d && echo foo >>d && git update-index --add d &&\n-->8--- git-tr.diff\n\n\n-- \nH.Merijn Brand Amsterdam Perl Mongers (http://amsterdam.pm.org/)\nusing & porting perl 5.6.2, 5.8.x, 5.10.x on HP-UX 10.20, 11.00, 11.11,\n& 11.23, SuSE 10.1 & 10.2, AIX 5.2, and Cygwin. http://qa.perl.org\nhttp://mirrors.develooper.com/hpux/ http://www.test-smoke.org\n http://www.goldmark.org/jeff/stupid-disclaimers/\n"},{"id":"62717","messageId":"475E963F.3000804@viscovery.net","threadId":"11128","inReplyTo":"20071211143323.27c860db@pc09.procura.nl","subject":"Re: Building git-1.5.3.7 on HP-UX 11.00","fromName":"Johannes Sixt","fromEmail":"j.sixt@viscovery.net","sentAt":"2007-12-11T13:53:03Z","receivedAt":"2007-12-11T13:53:03Z","isPatch":false,"sender":{"key":"j6t@kdbg.org","avatar":"https://avatars.githubusercontent.com/u/14810926?v=4"},"body":"H.Merijn Brand schrieb:\n> I have put the output of\n> \n> % make -i test | & perl -ne'm/^\\*\\s+ok\\s+\\d+:/ or print' > git-make-test.log\n> \n> on http://www.xs4all.nl/~procura/git-make-test.log\n> \n> which leaves only the following tests as failures:\n> \n> t3901-i18n-patch.sh t4201-shortlog.sh t5100-mailinfo.sh\n> t5405-send-pack-rewind.sh t5500-fetch-pack.sh t5505-remote.sh\n> t5510-fetch.sh t5515-fetch-merge-logic.sh t5516-fetch-push.sh\n> t5520-pull.sh t5600-clone-fail-cleanup.sh t5700-clone-reference.sh\n> t5701-clone-local.sh t6002-rev-list-bisect.sh t6003-rev-list-topo-order.sh\n> t6022-merge-rename.sh t6026-merge-attr.sh t6200-fmt-merge-msg.sh\n> t7003-filter-branch.sh t9300-fast-import.sh t9301-fast-export.sh\n> t9500-gitweb-standalone-no-errors.sh\n> \n> How close am I to success?\n\nThere are a number of tests that pass only if you have iconv, most notably\nt3901 and t5100. pull, clone, merge, filter-branch are shell scripts with\nplenty of opportunities to stumble over incompatible/unportable constructs.\n\n-- Hannes\n"}]}