{"thread":{"id":"42730","subject":"git svn clone segmentation faul issue","startedAt":"2016-06-28T08:50:48Z","lastAt":"2016-07-01T13:41:31Z","messageCount":12,"participants":["ioannis.kappas@rbs.com","Johannes Schindelin","Ioannis Kappas","Eric Wong"],"isPatch":false,"patchVersion":null,"patchTotal":null},"messages":[{"id":"290351","messageId":"0BCA1E695085C645B9CD4A27DD59F6FA39AAD5CF@GBWGCEUHUBD0101.rbsres07.net","threadId":"42730","inReplyTo":null,"subject":"git svn clone segmentation faul issue","fromName":"","fromEmail":"ioannis.kappas@rbs.com","sentAt":"2016-06-28T08:50:34Z","receivedAt":"2016-06-28T08:50:48Z","isPatch":false,"sender":{"key":"ioannis.kappas@rbs.com","avatar":null},"body":"Hello,\n\nGit can fail with a \"malformed index nnn\" error or cause a segmentation fault when executing the \"git svn clone\" command. \n\nThis is because the git-svn perl script calling into the external subversion library during fetch, wherein a bug exists that can corrupt the Perl Interpreter's stack maintained on the heap. The bug is in a custom swig typemap around the \"result_digest\" argument which is not handled in a safe manner: An absolute address on the perl stack is calculated beforehand to store the result of svn_swig_pl_from_md5(), but the function can trigger the stack to be reallocated elsewhere in memory (i.e. when trying to grow it but there is not enough space left) thus invaliding the address calculated earlier, causing memory corruption or segmentation fault.\n\nThis call is made once for each file retrieved from subversion. The fault is caused when the free available perl stack size is very small at the call, triggering a realloc() moving it elsewhere in memory. In my test environment --MSYS2 on windows 7--, this was caused when the available stack size left was 24 SV* slots. This means that a subversion repository with many file entries has a higher probability to fail while cloning than one with only a few file entries big.\n\nPerl stack trace to the entry point of the offending code is:\n\n\n        SVN::TxDelta::apply(GLOB(0x60107ece8), GLOB(0x6011ee3a0), undef, SVN::Pool=REF(0x6011ee268)) called at /usr/share/perl5/site_perl/Git/SVN/Fetcher.pm line 368\n        Git::SVN::Fetcher::apply_textdelta(Git::SVN::Fetcher=HASH(0x6010236a0), HASH(0x6011ee2e0), undef, _p_apr_pool_t=SCALAR(0x60113e510)) called at /usr/lib/perl5/vendor_perl/SVN/Ra.pm line 623\n        SVN::Ra::Reporter::AUTOLOAD(SVN::Ra::Reporter=ARRAY(0x6010e9ee0), SVN::Pool=REF(0x6010ea228)) called at /usr/share/perl5/site_perl/Git/SVN/Ra.pm line 312\n        Git::SVN::Ra::gs_do_update(Git::SVN::Ra=HASH(0x60101f4f0), 384103, 384103, Git::SVN=HASH(0x60101fc58), Git::SVN::Fetcher=HASH(0x6010236a0)) called at /usr/share/perl5/site_perl/Git/SVN.pm line 1205\n        Git::SVN::do_fetch(Git::SVN=HASH(0x60101fc58), HASH(0x6010e9b20), 384103) called at /usr/share/perl5/site_perl/Git/SVN/Ra.pm line 475\n        Git::SVN::Ra::gs_fetch_loop_common(Git::SVN::Ra=HASH(0x60101f4f0), 384103, 384103, ARRAY(0x600788cc8), ARRAY(0x600788ce0)) called at /usr/share/perl5/site_perl/Git/SVN.pm line 179\n        Git::SVN::fetch_all(\"svn\") called at /usr/lib/git-core/git-svn line 522\n        main::cmd_clone(\"http://...\", \"debug\") called at /usr/lib/git-core/git-svn line 386\n        eval {...} called at /usr/lib/git-core/git-svn line 384\n\nThe solution is to change the swig typemap in subversion to delay calculating the stack address where the result is stored after the call to the md5 function:\n\nIn subversion's subversion/subversion/bindings/swig/include/svn_types.swg, one can change\n\nFrom \n\n%typemap(argout) unsigned char *result_digest {\n    %append_output($1);\n}\n\nTo\n\n%typemap(argout) unsigned char *result_digest {\n    SV* ucrdTemp = svn_swig_pl_from_md5($1);             \n    %append_output(ucrdTemp);\n}\n\n\nSwig will translate the above typemap in subversion/subversion/bindings/swig/perl/native/svn_delta.c\n\nFrom\n\n      if (argvi >= items) EXTEND(sp,1);  ST(argvi) = svn_swig_pl_from_md5(arg3); argvi++  ;\n\n\nto\n\n      SV* ucrdTemp = svn_swig_pl_from_md5(arg3);             \n      if (argvi >= items) EXTEND(sp,1);  ST(argvi) = ucrdTemp; argvi++  ;\n\n(The stack address calculation is the results of calling the macro ST(argvi) ).\n\nI am not a Perl expert, but ST(n) = any_function() seems to be a dangerous idiom because of the potential of any_function() to cause a Perl stack reallocation invaliding the address calculated beforehand by ST(n).\n\n\nThe above patch has been tested to work in my codebase both on windows 7 (with MSYS2) and on Arch Linux, where it would have caused a segmentation fault otherwise.\n\n\nFortunately, a patch has already been submitted to subversion with (github) revision a074af86c8764404b28ce99d0bedcb668a321408 (at https://github.com/apache/subversion/commit/a074af86c8764404b28ce99d0bedcb668a321408 ) on the trunk to handle this and a couple of other similar cases. But by the looks of it has not been picked up yet in the latest subversion 1.9.4 release or the 1.9.x branch, perhaps because this patch was identified in sanity checks rather than coming out from a perceivable production issue?\n\nAlthough this issue is not in the Git source code, I thought reporting this error here first since it is Git that stresses the subversion perl bindings in such a level to cause the failure. Perhaps the patch can be applied first in those git binary releases that include subversion (I believe Git for windows is such a case) while in the mean time I submit this case for consideration in the subversion bug tracking system.\n\nYannis\n\n*********************************************************************************** \nThe Royal Bank of Scotland plc. Registered in Scotland No 90312. \nRegistered Office: 36 St Andrew Square, Edinburgh EH2 2YB. \nAuthorised by the Prudential Regulation Authority and regulated \nby the Financial Conduct Authority and Prudential Regulation Authority. \nThe Royal Bank of Scotland N.V. is authorised and regulated by the \nDe Nederlandsche Bank and has its seat at Amsterdam, the \nNetherlands, and is registered in the Commercial Register under \nnumber 33002587. Registered Office: Gustav Mahlerlaan 350, \nAmsterdam, The Netherlands. The Royal Bank of Scotland N.V. and \nThe Royal Bank of Scotland plc are authorised to act as agent for each \nother in certain jurisdictions. \n  \nThis e-mail message is confidential and for use by the addressee only. \nIf the message is received by anyone other than the addressee, please \nreturn the message to the sender by replying to it and then delete the \nmessage from your computer. Internet e-mails are not necessarily \nsecure. The Royal Bank of Scotland plc and The Royal Bank of Scotland \nN.V. including its affiliates (\"RBS group\") does not accept responsibility \nfor changes made to this message after it was sent. For the protection\nof RBS group and its clients and customers, and in compliance with\nregulatory requirements, the contents of both incoming and outgoing\ne-mail communications, which could include proprietary information and\nNon-Public Personal Information, may be read by authorised persons\nwithin RBS group other than the intended recipient(s). \n\nWhilst all reasonable care has been taken to avoid the transmission of \nviruses, it is the responsibility of the recipient to ensure that the onward \ntransmission, opening or use of this message and any attachments will \nnot adversely affect its systems or data. No responsibility is accepted \nby the RBS group in this regard and the recipient should carry out such \nvirus and other checks as it considers appropriate. \n\nVisit our website at www.rbs.com \n***********************************************************************************  \n\n"},{"id":"290373","messageId":"alpine.DEB.2.20.1606281334450.12947@virtualbox","threadId":"42730","inReplyTo":"0BCA1E695085C645B9CD4A27DD59F6FA39AAD5CF@GBWGCEUHUBD0101.rbsres07.net","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-06-28T11:47:06Z","receivedAt":"2016-06-28T11:54:22Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Ioannis,\n\nOn Tue, 28 Jun 2016, Ioannis.Kappas@rbs.com wrote:\n\n> Git can fail with a \"malformed index nnn\" error or cause a segmentation\n> fault when executing the \"git svn clone\" command. \n>\n> [...]\n>\n> Fortunately, a patch has already been submitted to subversion with\n> (github) revision a074af86c8764404b28ce99d0bedcb668a321408 (at\n> https://github.com/apache/subversion/commit/a074af86c8) on the trunk to\n> handle this and a couple of other similar cases. But by the looks of it\n> has not been picked up yet in the latest subversion 1.9.4 release or the\n> 1.9.x branch, perhaps because this patch was identified in sanity checks\n> rather than coming out from a perceivable production issue?\n\nThis is an excellent analysis and a silver lining on the horizon to\nresolve those vexing git svn issues we keep having in Git for Windows.\n\nDo you have a test case that is reliably reproducing the issue?\n\n> Although this issue is not in the Git source code, I thought reporting\n> this error here first since it is Git that stresses the subversion perl\n> bindings in such a level to cause the failure. Perhaps the patch can be\n> applied first in those git binary releases that include subversion (I\n> believe Git for windows is such a case) while in the mean time I submit\n> this case for consideration in the subversion bug tracking system.\n\nYes, absolutely. Thank you for reporting this!\n\n> The Royal Bank of Scotland plc. Registered in Scotland No 90312. \n\nOh. I am sorry!\n\n> This e-mail message is confidential and for use by the addressee only. \n\nThis footer does not make much sense when sending a mail to a public\nmailing list ;-)\n\nCiao,\nJohannes\n"},{"id":"290378","messageId":"alpine.DEB.2.20.1606281530420.12947@virtualbox","threadId":"42730","inReplyTo":"alpine.DEB.2.20.1606281334450.12947@virtualbox","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-06-28T13:32:09Z","receivedAt":"2016-06-28T13:32:20Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Ioannis,\n\nOn Tue, 28 Jun 2016, Johannes Schindelin wrote:\n\n> On Tue, 28 Jun 2016, Ioannis.Kappas@rbs.com wrote:\n> \n> > Git can fail with a \"malformed index nnn\" error or cause a segmentation\n> > fault when executing the \"git svn clone\" command. \n> >\n> > [...]\n> >\n> > Fortunately, a patch has already been submitted to subversion with\n> > (github) revision a074af86c8764404b28ce99d0bedcb668a321408 (at\n> > https://github.com/apache/subversion/commit/a074af86c8) on the trunk to\n> > handle this and a couple of other similar cases. But by the looks of it\n> > has not been picked up yet in the latest subversion 1.9.4 release or the\n> > 1.9.x branch, perhaps because this patch was identified in sanity checks\n> > rather than coming out from a perceivable production issue?\n> \n> This is an excellent analysis and a silver lining on the horizon to\n> resolve those vexing git svn issues we keep having in Git for Windows.\n> \n> Do you have a test case that is reliably reproducing the issue?\n\nI hope you do! I patched the MSYS2 build script to apply a074af86c8 before\ncompiling, and uploaded the resulting packages for i686 and x86_64\narchitectures to\n\n\thttps://github.com/dscho/MSYS2-packages/releases/tag/subversion-1.9.4-2\n\nWould you mind giving them a whirl?\n\nThanks,\nJohannes\n"},{"id":"290388","messageId":"loom.20160628T175016-898@post.gmane.org","threadId":"42730","inReplyTo":"alpine.DEB.2.20.1606281530420.12947@virtualbox","subject":"Re: git svn clone segmentation faul issue","fromName":"Ioannis Kappas","fromEmail":"ioannis.kappas@rbs.com","sentAt":"2016-06-28T16:13:37Z","receivedAt":"2016-06-28T16:15:14Z","isPatch":false,"sender":{"key":"ioannis.kappas@rbs.com","avatar":null},"body":"Hi Johanes,\n\nJohannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n\n> \n> Hi Ioannis,\n> \n> On Tue, 28 Jun 2016, Johannes Schindelin wrote:\n> \n> > On Tue, 28 Jun 2016, Ioannis.Kappas <at> rbs.com wrote:\n> > \n> > > Git can fail with a \"malformed index nnn\" error or cause a \nsegmentation\n> > > fault when executing the \"git svn clone\" command. \n> > >\n> > > [...]\n> > >\n> > > Fortunately, a patch has already been submitted to subversion with\n> > > (github) revision a074af86c8764404b28ce99d0bedcb668a321408 (at\n> > > https://github.com/apache/subversion/commit/a074af86c8) on the trunk \nto\n> > > handle this and a couple of other similar cases. But by the looks of \nit\n> > > has not been picked up yet in the latest subversion 1.9.4 release or \nthe\n> > > 1.9.x branch, perhaps because this patch was identified in sanity \nchecks\n> > > rather than coming out from a perceivable production issue?\n> > \n> > This is an excellent analysis and a silver lining on the horizon to\n> > resolve those vexing git svn issues we keep having in Git for Windows.\n> > \n> > Do you have a test case that is reliably reproducing the issue?\n> \n> I hope you do! \n\nYes, my codebase always fails on git svn clone with something like:\n\nfatal: malformed index info 100644 a6d6543f625fb4a7200e\nerror: git-svn died of signal 11\n\n\n> I patched the MSYS2 build script to apply a074af86c8 before\n> compiling, and uploaded the resulting packages for i686 and x86_64\n> architectures to\n> \n> \thttps://github.com/dscho/MSYS2-\npackages/releases/tag/subversion-1.9.4-2\n> \n> Would you mind giving them a whirl?\n\nThe patch in \"source code (zip)\" seems to be missing the line in the prepare\n() section of PKGBUILD to actually apply the fix:\n\ndiff --git a/subversion/PKGBUILD b/subversion/PKGBUILD\n--- a/subversion/PKGBUILD\n+++ b/subversion/PKGBUILD\n\n@@ -101,6 +103,7 @@ prepare() {\n   patch -p1 -i ${srcdir}/17-fix-test-link.patch\n   patch -p1 -i ${srcdir}/18-fix-serf-config.patch\n   patch -p1 -i ${srcdir}/19-remove-contrib-from-configure.patch\n+  patch -p1 -i ${srcdir}/20-fix-stack-corruption-in-swig-wrappers.patch\n   patch -p1 -i ${srcdir}/subversion-1.9.1-msys2.patch\n   patch -p1 -i ${srcdir}/remove-checking-symlink.patch\n   patch -p1 -i ${srcdir}/90-use-copy-instead-symlink.patch\n\n\nWould you be so kind to add the above and rebuild. \n\nI have just tested it locally to work successfully, but I do not mind \nretesting if you wish to provide another build :)\n\nThanks,\nYannis\n\n> \n> Thanks,\n> Johannes\n> \n\n\n\n\n"},{"id":"290391","messageId":"alpine.DEB.2.20.1606281825190.12947@virtualbox","threadId":"42730","inReplyTo":"loom.20160628T175016-898@post.gmane.org","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-06-28T16:27:37Z","receivedAt":"2016-06-28T16:27:46Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Yannis,\n\nOn Tue, 28 Jun 2016, Ioannis Kappas wrote:\n\n> Hi Johanes,\n\nAlmost ;-)\n\n> Johannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n> \n> > Would you mind giving them a whirl?\n> \n> The patch in \"source code (zip)\" seems to be missing the line in the prepare\n> () section of PKGBUILD to actually apply the fix:\n> \n> diff --git a/subversion/PKGBUILD b/subversion/PKGBUILD\n> --- a/subversion/PKGBUILD\n> +++ b/subversion/PKGBUILD\n> \n> @@ -101,6 +103,7 @@ prepare() {\n>    patch -p1 -i ${srcdir}/17-fix-test-link.patch\n>    patch -p1 -i ${srcdir}/18-fix-serf-config.patch\n>    patch -p1 -i ${srcdir}/19-remove-contrib-from-configure.patch\n> +  patch -p1 -i ${srcdir}/20-fix-stack-corruption-in-swig-wrappers.patch\n>    patch -p1 -i ${srcdir}/subversion-1.9.1-msys2.patch\n>    patch -p1 -i ${srcdir}/remove-checking-symlink.patch\n>    patch -p1 -i ${srcdir}/90-use-copy-instead-symlink.patch\n> \n> \n> Would you be so kind to add the above and rebuild. \n\nOh bummer. Sorry for that! I will fix it and rebuild tomorrow.\n\n> I have just tested it locally to work successfully, but I do not mind \n> retesting if you wish to provide another build :)\n\nThanks, that would be much appreciated! I'll send a reply when I rebuilt\nthe packages.\n\nCiao,\nJohannes\n"},{"id":"290427","messageId":"20160628194804.GA3653@dcvr.yhbt.net","threadId":"42730","inReplyTo":"0BCA1E695085C645B9CD4A27DD59F6FA39AAD5CF@GBWGCEUHUBD0101.rbsres07.net","subject":"Re: git svn clone segmentation faul issue","fromName":"Eric Wong","fromEmail":"e@80x24.org","sentAt":"2016-06-28T19:48:04Z","receivedAt":"2016-06-28T19:48:13Z","isPatch":false,"sender":{"key":"e@80x24.org","avatar":null},"body":"Ioannis.Kappas@rbs.com wrote:\n> Fortunately, a patch has already been submitted to subversion\n> with (github) revision\n> a074af86c8764404b28ce99d0bedcb668a321408 (at\n> https://github.com/apache/subversion/commit/a074af86c8764404b28ce99d0bedcb668a321408\n> ) on the trunk to handle this and a couple of other similar\n> cases. But by the looks of it has not been picked up yet in\n> the latest subversion 1.9.4 release or the 1.9.x branch,\n> perhaps because this patch was identified in sanity checks\n> rather than coming out from a perceivable production issue?\n\nThank you for documenting this.  Curious, does this affect older\nSVN versions or only 1.9.x?\n\nI don't know Perl internals well or SWIG at all; so reports\nlike these are very much appreciated.\n"},{"id":"290460","messageId":"alpine.DEB.2.20.1606290911040.12947@virtualbox","threadId":"42730","inReplyTo":"alpine.DEB.2.20.1606281825190.12947@virtualbox","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-06-29T07:13:39Z","receivedAt":"2016-06-29T07:14:00Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Yannis,\n\nOn Tue, 28 Jun 2016, Johannes Schindelin wrote:\n\n> On Tue, 28 Jun 2016, Ioannis Kappas wrote:\n> \n> > Johannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n> > \n> > > Would you mind giving them a whirl?\n> > \n> > The patch in \"source code (zip)\" seems to be missing the line in the prepare\n> > () section of PKGBUILD to actually apply the fix:\n> > \n> > diff --git a/subversion/PKGBUILD b/subversion/PKGBUILD\n> > --- a/subversion/PKGBUILD\n> > +++ b/subversion/PKGBUILD\n> > \n> > @@ -101,6 +103,7 @@ prepare() {\n> >    patch -p1 -i ${srcdir}/17-fix-test-link.patch\n> >    patch -p1 -i ${srcdir}/18-fix-serf-config.patch\n> >    patch -p1 -i ${srcdir}/19-remove-contrib-from-configure.patch\n> > +  patch -p1 -i ${srcdir}/20-fix-stack-corruption-in-swig-wrappers.patch\n> >    patch -p1 -i ${srcdir}/subversion-1.9.1-msys2.patch\n> >    patch -p1 -i ${srcdir}/remove-checking-symlink.patch\n> >    patch -p1 -i ${srcdir}/90-use-copy-instead-symlink.patch\n> > \n> > \n> > Would you be so kind to add the above and rebuild. \n> \n> Oh bummer. Sorry for that! I will fix it and rebuild tomorrow.\n> \n> > I have just tested it locally to work successfully, but I do not mind \n> > retesting if you wish to provide another build :)\n> \n> Thanks, that would be much appreciated! I'll send a reply when I rebuilt\n> the packages.\n\nI just re-uploaded new packages, after rebuilding them and making sure\nthat the patch was actually applied.\n\nMay I ask you to re-test, just to make extra sure that the bug in question\nis fixed?\n\nCiao,\nJohannes\n"},{"id":"290463","messageId":"loom.20160629T095219-589@post.gmane.org","threadId":"42730","inReplyTo":"20160628194804.GA3653@dcvr.yhbt.net","subject":"Re: git svn clone segmentation faul issue","fromName":"Ioannis Kappas","fromEmail":"ioannis.kappas@rbs.com","sentAt":"2016-06-29T08:11:38Z","receivedAt":"2016-06-29T08:12:07Z","isPatch":false,"sender":{"key":"ioannis.kappas@rbs.com","avatar":null},"body":"Hi Eric,\n\nEric Wong <e <at> 80x24.org> writes:\n\n> \n> Ioannis.Kappas <at> rbs.com wrote:\n> > Fortunately, a patch has already been submitted to subversion\n> > with (github) revision\n> > a074af86c8764404b28ce99d0bedcb668a321408 (at\n> > \nhttps://github.com/apache/subversion/commit/a074af86c8764404b28ce99d0bedcb66\n8a321408\n> > ) on the trunk to handle this and a couple of other similar\n> > cases. But by the looks of it has not been picked up yet in\n> > the latest subversion 1.9.4 release or the 1.9.x branch,\n> > perhaps because this patch was identified in sanity checks\n> > rather than coming out from a perceivable production issue?\n> \n> Thank you for documenting this.  Curious, does this affect older\n> SVN versions or only 1.9.x?\n\nIt does not appear so, at least not in the degree of failures that I am \nseeing it now, i.e. failing consistently. It only came to my attention \nafter I upgraded to the latest git/perl/subversion release from a version \nthat was more than a year's old.\n\nGiven that the issue is with corruption in the perl interpreter stack, I \nsuspect is due to one, or combination, of the following:\n\n- git in latest releases makes additional perl subversion calls, thus \naffecting the perl stack usage.\n- perl in latest releases affected the stack allocation profile.\n- subversion perl module in latest releases refactored/added new calls that \nhas changed the perl stack usage.\n\n> \n> I don't know Perl internals well or SWIG at all; so reports\n> like these are very much appreciated.\n> \n\nThanks\n\n\n\n"},{"id":"290465","messageId":"loom.20160629T104814-936@post.gmane.org","threadId":"42730","inReplyTo":"alpine.DEB.2.20.1606290911040.12947@virtualbox","subject":"Re: git svn clone segmentation faul issue","fromName":"Ioannis Kappas","fromEmail":"ioannis.kappas@rbs.com","sentAt":"2016-06-29T09:00:35Z","receivedAt":"2016-06-29T09:01:24Z","isPatch":false,"sender":{"key":"ioannis.kappas@rbs.com","avatar":null},"body":"Hi Johannes,\n\nJohannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n\n> \n> Hi Yannis,\n> \n> [...]\n> \n> I just re-uploaded new packages, after rebuilding them and making sure\n> that the patch was actually applied.\n> \n> May I ask you to re-test, just to make extra sure that the bug in question\n> is fixed?\n\nI confirm the latest version, built from the 'source files (zip)' package, \nhas resolved the issue.\n\nThanks for the prompt response! Next task is to get the patch through to \nthe next subversion official release.\n\n> \n> Ciao,\n> Johannes\n> \n\n\n\n\n"},{"id":"290476","messageId":"alpine.DEB.2.20.1606291341530.12947@virtualbox","threadId":"42730","inReplyTo":"loom.20160629T104814-936@post.gmane.org","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-06-29T11:43:26Z","receivedAt":"2016-06-29T11:43:39Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Yannis,\n\nOn Wed, 29 Jun 2016, Ioannis Kappas wrote:\n\n> Johannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n> \n> > I just re-uploaded new packages, after rebuilding them and making sure\n> > that the patch was actually applied.\n> > \n> > May I ask you to re-test, just to make extra sure that the bug in\n> > question is fixed?\n> \n> I confirm the latest version, built from the 'source files (zip)'\n> package, has resolved the issue.\n\nThank you so much!\n\n> Thanks for the prompt response! Next task is to get the patch through to\n> the next subversion official release.\n\nMy more immediate concern is MSYS2 (the POSIX-emulating basis of Git for\nWindows), and I opened a Pull Request there:\n\n\thttps://github.com/Alexpux/MSYS2-packages/pull/647\n\nIf this Pull Request is not picked up in time for the next Git for Windows\nversion, I plan to use the packages that you tested.\n\nThanks,\nDscho\n"},{"id":"290649","messageId":"loom.20160701T120539-937@post.gmane.org","threadId":"42730","inReplyTo":"alpine.DEB.2.20.1606291341530.12947@virtualbox","subject":"Re: git svn clone segmentation faul issue","fromName":"Ioannis Kappas","fromEmail":"ioannis.kappas@rbs.com","sentAt":"2016-07-01T10:09:31Z","receivedAt":"2016-07-01T10:09:55Z","isPatch":false,"sender":{"key":"ioannis.kappas@rbs.com","avatar":null},"body":"Hi Johannes,\n\nJohannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n\n> [...]\n> \n> My more immediate concern is MSYS2 (the POSIX-emulating basis of Git \nfor\n> Windows), and I opened a Pull Request there:\n> \n> \thttps://github.com/Alexpux/MSYS2-packages/pull/647\n> \n> If this Pull Request is not picked up in time for the next Git for \nWindows\n> version, I plan to use the packages that you tested.\n\nThe patch has also been accepted for inclusion in the forthcoming \nsubversion 1.9.5 release (\"barring unexpectancies\"):\n\nhttp://thread.gmane.org/gmane.comp.version-\ncontrol.subversion.user/120144/focus=120146\n\nThanks,\nYannis\n\n> \n> Thanks,\n> Dscho\n> \n\n\n\n\n"},{"id":"290660","messageId":"alpine.DEB.2.20.1607011540470.12947@virtualbox","threadId":"42730","inReplyTo":"loom.20160701T120539-937@post.gmane.org","subject":"Re: git svn clone segmentation faul issue","fromName":"Johannes Schindelin","fromEmail":"johannes.schindelin@gmx.de","sentAt":"2016-07-01T13:41:21Z","receivedAt":"2016-07-01T13:41:31Z","isPatch":false,"sender":{"key":"johannes.schindelin@gmx.de","avatar":"https://avatars.githubusercontent.com/u/127790?v=4"},"body":"Hi Yannis,\n\nOn Fri, 1 Jul 2016, Ioannis Kappas wrote:\n\n> Johannes Schindelin <Johannes.Schindelin <at> gmx.de> writes:\n> \n> > [...]\n> > \n> > My more immediate concern is MSYS2 (the POSIX-emulating basis of Git\n> > for Windows), and I opened a Pull Request there:\n> > \n> > \thttps://github.com/Alexpux/MSYS2-packages/pull/647\n> > \n> > If this Pull Request is not picked up in time for the next Git for\n> > Windows version, I plan to use the packages that you tested.\n> \n> The patch has also been accepted for inclusion in the forthcoming \n> subversion 1.9.5 release (\"barring unexpectancies\"):\n> \n> http://thread.gmane.org/gmane.comp.version-\n> control.subversion.user/120144/focus=120146\n\nPerfect. Thanks for being diligent and keeping us informed!\n\nCiao,\nDscho\n"}]}