You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Ben Reser <be...@reser.org> on 2013/10/22 21:36:50 UTC
1.8.4 up for testing/signing
The 1.8.4 release artifacts are now available for testing/signing.
Please get the tarballs from
https://dist.apache.org/repos/dist/dev/subversion
and add your signatures there. I plan to try and release on October
29th so please try and get your votes/signatures in place by October 28th.
Thanks!
Re: 1.8.4 up for testing/signing
Posted by Philip Martin <ph...@wandisco.com>.
Summary:
+1 to release
Platform:
Linux (Debian/wheezy) 64-bit
Tested:
(local, svn, svn/sasl, serf, serf/v1) x (fsfs, fsfs/pack/shard, bdb)
swig-pl, swig-py, swig-rb, ctypes-python
javahl x (fsfs, bdb)
Results:
All tests successful except for expected failures of checkout_tests.py 12.
Local dependencies:
apache2-threaded-dev : 2.2.22-13
libapr1-dev : 1.4.6-3+deb7u1
libaprutil1-dev : 1.4.1-3
libdb5.1-dev : 5.1.29-5
libsasl2-dev : 2.1.25.dfsg1-6+deb7u1
libsqlite3-dev : 3.7.16.2-1~bpo70+1
perl : 5.14.2-21+deb7u1
python2.7-dev : 2.7.3-6
ruby1.8-dev : 1.8.7.358-7
openjdk-7-jdk : 7u25-2.3.10-1~deb7u1
serf : 1.3.x@2142
subversion-1.8.4.tar.bz2
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iQEcBAABCAAGBQJSZ7VlAAoJEHbXiOHtGlmcp1YH/jJ1/y+9Qjt4Hjblo//r4Nk2
f2muNvozK0s6GLcTgv01e84Zx28tS/fcjyewdDdzkFlfVDyKnIRyRcFcEcNtwLek
LVTlOw4G+PbCXqd7e2T7DKx6rxjEc7swdkMA0HOQld4LLhUVY/YFVUHcWHteALBE
LMVSzSTHLkCyUhPwK1KIhq2GTyf1PaQJot/qp8Rt0rhto2hNFbLxy3GQNeC/1+yz
EiqcN2Ugiien92Ql7alMvunRmwPVcaw1Di02FXzFE30JPy4GMAoc59uSFoGbcV4k
PAJYl5J5QWJTPL9MkQk4r6yVct/3P6/Jrfc1U3ENX0RFPBDaGsvQhEE24g0yZtE=
=J8Mf
-----END PGP SIGNATURE-----
subversion-1.8.4.tar.gz
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iQEcBAABCAAGBQJSZ7VlAAoJEHbXiOHtGlmc+dIH/1a8KlhhYgqv/IRaOatfLB+D
5N2BhzySPWzYg0Nf/p2y/5pKtzWiQxiR9QixCcByp5PO2+LUPuStUiVA3BEJUCMK
DJBXvk40YSN1jvYo148yGDssUlKUuJmPbGFMLeJQ0zwrXQQJs3bDdFPqq6FB+HoD
Gsh6S0pFeRr2qWtoPQwqocSV8A2X2yxbKx8YqWQdQrx460ZfOP3gSQ6YUPXqNMvy
50SQa2fQO3gQnWepQlH1SCqCtQH0p8k48t2+YNFwVRWPjsGrf2RqVJGZMXui88kX
0C1NVaGo4y2Yv/lqAUoIVSZNoDEwv8YxetDeLyZZEcfe82u1dTRrbtosC/Le9tM=
=ofhH
-----END PGP SIGNATURE-----
--
Philip Martin | Subversion Committer
WANdisco // *Non-Stop Data*
Re: 1.8.4 up for testing/signing
Posted by Branko Čibej <br...@wandisco.com>.
On 23.10.2013 14:04, Ivan Zhakov wrote:
> Please note that I'm using scripts that was working fine with previous
> versions. I think we should not release Subversion 1.8.4 until this
> situation will be resolved.
Noted, but regressions scripts that someone uses to build Subversion,
that are not in our repository, are not part of the release criteria.
It's unfortunate that we don't have a reliable way to build Subversion
on Windows; but two others have already committed Windows signatures, so
clearly they /could/ build the binaries.
-- Brane
--
Branko Čibej | Director of Subversion
WANdisco // Non-Stop Data
e. brane@wandisco.com
RE: 1.8.4 up for testing/signing
Posted by Bert Huijben <be...@qqmail.nl>.
> -----Original Message-----
> From: Ivan Zhakov [mailto:ivan@visualsvn.com]
> Sent: woensdag 23 oktober 2013 14:04
> To: Ben Reser
> Cc: Subversion Development
> Subject: Re: 1.8.4 up for testing/signing
>
> On 22 October 2013 23:36, Ben Reser <be...@reser.org> wrote:
> > The 1.8.4 release artifacts are now available for testing/signing.
> > Please get the tarballs from
> > https://dist.apache.org/repos/dist/dev/subversion
> > and add your signatures there. I plan to try and release on October
> > 29th so please try and get your votes/signatures in place by October 28th.
> >
> I cannot build Subversion 1.8.4 on Windows. First problem that I have
> to update scripts that was working fine for Subversion 1.8.3 to build
> zlib and serf externally, while it was compiled by Subversion before.
> Even after fixing this I get the following error:
> [[[
> [exec] ------ Build started: Project: libsvn_ra_dll,
> Configuration: Release Win32 ------
> [exec] Linking...
> [exec] LINK : fatal error LNK1181: cannot open input file 'ssleay32.lib'
> ]]]
>
>
> From what I see problem in gen_win.py:1095:
> [[[
> if self.serf_lib and dep.external_lib == '$(SVN_SERF_LIBS)':
> nondeplibs.append(serflib)
> if (self.serf_ver_maj, self.serf_ver_min) >= (1, 3):
> nondeplibs.append('ssleay32.lib')
> nondeplibs.append('libeay32.lib')
> ]]]
>
> So dependency to ssleay32.lib added, but OPENSS\out32dll is not added
> to library search path (fakelibdirs).
>
> This code introduced on branch in r1533874:
> [[[
> rhuijben 20.10.2013 14:41:47
> Merge ^/subversion/branches/1.8.x-serf-1.3+-windows from trunk
> manually,
> because our automatic merge script doesn't like the branch name.
>
> * ^/subversion/branches/1.8.x-serf-1.3+-windows
> Allow compiling against serf 1.3 and later on Windows
> Justification:
> Serf 1.3 brings a lot of fixes that we need for 1.8.x stability.
> Notes:
> The dependency handling on Windows was updated for 1.9+, so this
> needs a specific backport patch (r1517123)
> Votes:
> +1: rhuijben, brane, stefan2
>
> ../trunk/tools/dist/backport.pl: sh(256): (in 'the
> 1.8.x-serf-1.3+-windows branch') at
> ../trunk/tools/dist/backport.pl line 258, <STATUS> chunk 21.
> Warning summary
> ===============
>
> 1.8.x-serf-1.3+-windows
> (^/subversion/branches/1.8.x-serf-1.3+-windows[...]): sh(256):
> ]]]
>
> Please note that I'm using scripts that was working fine with previous
> versions. I think we should not release Subversion 1.8.4 until this
> situation will be resolved.
Then we have a conflict between the requirements of the serf and Subversion projects.
Before this patch Subversion requires a serf makefile and wants to build zlib itself.
While serf wants a pre-build zlib, and to build itself with scons, preferable before Subversion.
The out32dll directory points to a OpenSSL source location (And only to that of a DLL build, not a static build), while both Serf and Subversion document for many releases that the argument should point to an install location where the build type (Static vs Shared) doesn't matter.
Yes, your old situation may have worked, but as far as I can see not in a way that was promised to work by either recent Serf or recent Subversion.
If you put everything of openssl in the apr directories, things build ok... But we never promised that it works (and we never will).
For 1.9/trunk the dependency system is optimized to be in some ways stricter (Version checks, etc.), while in more ways relaxed: you can use output directories for all dependencies; no more source requirements.
Bert
Re: 1.8.4 up for testing/signing
Posted by Ivan Zhakov <iv...@visualsvn.com>.
On 22 October 2013 23:36, Ben Reser <be...@reser.org> wrote:
> The 1.8.4 release artifacts are now available for testing/signing.
> Please get the tarballs from
> https://dist.apache.org/repos/dist/dev/subversion
> and add your signatures there. I plan to try and release on October
> 29th so please try and get your votes/signatures in place by October 28th.
>
I cannot build Subversion 1.8.4 on Windows. First problem that I have
to update scripts that was working fine for Subversion 1.8.3 to build
zlib and serf externally, while it was compiled by Subversion before.
Even after fixing this I get the following error:
[[[
[exec] ------ Build started: Project: libsvn_ra_dll,
Configuration: Release Win32 ------
[exec] Linking...
[exec] LINK : fatal error LNK1181: cannot open input file 'ssleay32.lib'
]]]
>From what I see problem in gen_win.py:1095:
[[[
if self.serf_lib and dep.external_lib == '$(SVN_SERF_LIBS)':
nondeplibs.append(serflib)
if (self.serf_ver_maj, self.serf_ver_min) >= (1, 3):
nondeplibs.append('ssleay32.lib')
nondeplibs.append('libeay32.lib')
]]]
So dependency to ssleay32.lib added, but OPENSS\out32dll is not added
to library search path (fakelibdirs).
This code introduced on branch in r1533874:
[[[
rhuijben 20.10.2013 14:41:47
Merge ^/subversion/branches/1.8.x-serf-1.3+-windows from trunk manually,
because our automatic merge script doesn't like the branch name.
* ^/subversion/branches/1.8.x-serf-1.3+-windows
Allow compiling against serf 1.3 and later on Windows
Justification:
Serf 1.3 brings a lot of fixes that we need for 1.8.x stability.
Notes:
The dependency handling on Windows was updated for 1.9+, so this
needs a specific backport patch (r1517123)
Votes:
+1: rhuijben, brane, stefan2
../trunk/tools/dist/backport.pl: sh(256): (in 'the
1.8.x-serf-1.3+-windows branch') at
../trunk/tools/dist/backport.pl line 258, <STATUS> chunk 21.
Warning summary
===============
1.8.x-serf-1.3+-windows
(^/subversion/branches/1.8.x-serf-1.3+-windows[...]): sh(256):
]]]
Please note that I'm using scripts that was working fine with previous
versions. I think we should not release Subversion 1.8.4 until this
situation will be resolved.
--
Ivan Zhakov
CTO | VisualSVN | http://www.visualsvn.com
Re: 1.8.4 up for testing/signing
Posted by Paul Burba <pt...@gmail.com>.
On Tue, Oct 22, 2013 at 3:36 PM, Ben Reser <be...@reser.org> wrote:
> The 1.8.4 release artifacts are now available for testing/signing.
> Please get the tarballs from
> https://dist.apache.org/repos/dist/dev/subversion
> and add your signatures there. I plan to try and release on October
> 29th so please try and get your votes/signatures in place by October 28th.
>
> Thanks!
SUMMARY:
---------
+1 to release
VERIFIED:
---------
Other than the expected differences in
subversion/include/svn_version.h,
tools/server-side/svnpubsub/rc.d/svnpubsub and
tools/server-side/svnpubsub/rc.d/svnwcsub[1],
https://dist.apache.org/repos/dist/dev/subversion/subversion-1.8.4.zip
is identical to
https://svn.apache.org/repos/asf/subversion/branches/1.8.x@1534716.
[1] See http://svn.haxx.se/dev/archive-2013-05/0079.shtml) for details.
SHA1 checksum of
https://dist.apache.org/repos/dist/dev/subversion/subversion-1.8.4.zip
is 6753989f5c6909dac74e3109547bf4794d3002d4 and agrees with the
published value at
https://dist.apache.org/repos/dist/dev/subversion/subversion-1.8.4.zip.sha1
C:\SVN\Google Drive\Subversion\NOTES\Release Testing\1.8.4>gpg2
--verify subversion-1.8.4.zip.asc
gpg: Signature made 10/22/13 15:30:20 Eastern Daylight Time using RSA
key ID 16A0DE01
gpg: Good signature from "Ben Reser <be...@reser.org>"
gpg: aka "Ben Reser <br...@apache.org>"
gpg: aka "Ben Reser <be...@wandisco.com>"
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the owner.
Primary key fingerprint: 19BB CAEF 7B19 B280 A0E2 175E 62D4 8FAD 16A0 DE01
TESTED:
-------
[Release-Build] x[ fsfs | bdb ] x [ ra_file | ra_svn | ra_serf) ]
Ruby bindings
JavaHL Bindings
RESULTS:
--------
All Tests Pass
PLATFORM:
---------
MS Windows 7 Home Premium Service Pack 1
Microsoft Visual Studio 2008 Version 9.0.30729.1 SP
DEPENDENCIES:
-------------
APR: 1.4.6
APR-UTIL: 1.4.1
Neon: 0.29.5
zlib: 1.2.4
OpenSSL: 0.9.8q
Apache: 2.2.22
BDB: 4.8.30
sqlite: 3.7.7.1
Python: 2.7.2 (ActivePython 2.7.2.5)
Perl: 5.14.2
Ruby: ruby 1.8.7
java: 1.7.0_10
junit: 4.8.2
swig: 1.3.40
serf: 1.2.1
SIGNATURE:
----------
user: "Paul T. Burba <pburba_at_collab.net>"
4096-bit RSA key, ID 56F3D7BC, created 2012-12-11
For https://dist.apache.org/repos/dist/dev/subversion/subversion-1.8.4.zip:
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.17 (MingW32)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=N3z3
-----END PGP SIGNATURE-----
--
Paul T. Burba
CollabNet, Inc. -- www.collab.net -- Enterprise Cloud Development
Skype: ptburba
Re: 1.8.4 up for testing/signing
Posted by Julian Foad <ju...@btopenworld.com>.
Ben Reser wrote:
> The 1.8.4 release artifacts are now available for testing/signing.
I've built this but am running into OS or hardware problems when trying to run the test suite, so don't know if I'll be able to sign it in time.
(Linux is throwing SSD I/O errors and remounting the FS read-only when I run the test suite. Repeatably. It's the first time I've tried to run a full multi-way svn test suite on this new machine, although I have successfully run single "make check"s. After re-booting, the file system (ext4) checks OK, the drive's SMART data shows no errors except "runtime-bad-blk-total = 2" which I think is insignificant, a "disk surface self-test" checks out OK, and everything works just fine until I try to run the tests again. So it looks like a hardware/firmware/OS-driver interaction bug. I have found a number of similar reports on the internet, but no specific diagnosis or solution for my case.)
- Julian
Re: 1.8.4 up for testing/signing
Posted by Ivan Zhakov <iv...@visualsvn.com>.
On 22 October 2013 23:36, Ben Reser <be...@reser.org> wrote:
> The 1.8.4 release artifacts are now available for testing/signing.
> Please get the tarballs from
> https://dist.apache.org/repos/dist/dev/subversion
> and add your signatures there. I plan to try and release on October
> 29th so please try and get your votes/signatures in place by October 28th.
>
Summary:
--------
+1 to release.
TESTED:
-------
[ fsfs ] x [ file | svn | http v1 | http v2]
RESULTS:
--------
All Tests Pass
PLATFORM:
---------
MS Windows 7 Ultimate (x64)
Microsoft Visual Studio 2008 Version 9.0.30729.1 SP
DEPENDENCIES:
-------------
APR: 1.4.8
APR-Util: 1.5.2
Apache HTTPD: 2.2.25 with r1497121 and r1497441 reverted.
zlib: 1.2.8
OpenSSL: 0.9.8y
sqlite: 3.7.12.1
Python: 2.6.6
serf: 1.3.2
Signatures:
-----------
(See the appropriate distribution files.)
Issues:
---------
- I had to apply r1535139 from 1.8.x-openssl-dirs branch to enable
linking OpenSSL with serf 1.3.x. But it should not be big problem,
since serf 1.3.x builds was not supported before.
--
Ivan Zhakov
CTO | VisualSVN | http://www.visualsvn.com
RE: 1.8.4 up for testing/signing
Posted by Bert Huijben <be...@qqmail.nl>.
> -----Original Message-----
> From: Ben Reser [mailto:ben@reser.org]
> Sent: dinsdag 22 oktober 2013 21:37
> To: Subversion Development
> Subject: 1.8.4 up for testing/signing
>
> The 1.8.4 release artifacts are now available for testing/signing.
> Please get the tarballs from
> https://dist.apache.org/repos/dist/dev/subversion
> and add your signatures there. I plan to try and release on October
> 29th so please try and get your votes/signatures in place by October 28th.
+1 for release.
Completed a full [bdb | fsfs] x [local | svn | http] testrun on Windows 7
x64.
Using the Visual C++ 2012 (Update 3) compiler, using the x64 cross compile
chain.
(So actually testing the far less common 64 bit binaries)
Dependencies:
apr 1.4.8
apr-util 1.5.2
bdb 4.4.20
expat 1.95.7 (from apr)
openssl 1.0.1e
python 2.7.2 (32 bit)
sasl 2.1.23
serf 1.3.2
sqlite 3.7.17
zlib 1.2.8 (without assembly optimizations)
Without r1499470, merge_tests.py 135 sometimes doesn't print its final lines
of output right before failing with an error, which makes the test fail as
the expected conflict summary is not available.
(Revision nominated for 1.8.5)
This is not a regression, but the result of running the testsuite faster
every time is that these timing issues become more visible.
Signatures are already commited.
Bert
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/27/13 10:03 AM, Stefan Fuhrmann wrote:
> Summary:
>
> +1 to release
>
Thanks Stefan. I'll stage this first thing Monday morning in case anyone else
wants to sign the release.
Re: 1.8.4 up for testing/signing
Posted by Stefan Fuhrmann <st...@wandisco.com>.
Summary:
+1 to release
Platform
Ubuntu 12.04.3 64 bit
Standard dependencies:
GCC 4.6.3-1ubuntu5
APR 1.4.6
APR-Util 1.3.12
zlib 1.2.3
httpd 2.2.22
OpenSSL 1.0.1
Python 2.7.3
Perl 5.14.2
Dependencies from homebrew:
SQLite 3.7.15.1
Serf 1.3.2
BDB 5.1.4
libmagic 5.9
Verified:
Tarball contents and signatures
(fsfs, bdb) x (local, svnserve, dav)
GPG Signatures committed to the dist/dev/subversion repository.
-- Stefan^2.
On Sat, Oct 26, 2013 at 10:23 PM, Ben Reser <be...@reser.org> wrote:
> On 10/22/13 12:36 PM, Ben Reser wrote:
> > The 1.8.4 release artifacts are now available for testing/signing.
> > Please get the tarballs from
> > https://dist.apache.org/repos/dist/dev/subversion
> > and add your signatures there. I plan to try and release on October
> > 29th so please try and get your votes/signatures in place by October
> 28th.
>
> We have 3/3 votes for Windows and 2/3 for Unix. I know Julian was having
> some
> issues with his laptop trying to do his testing. If someone else has some
> time
> to do some testing over the weekend it would be great.
>
>
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/22/13 12:36 PM, Ben Reser wrote:
> The 1.8.4 release artifacts are now available for testing/signing.
> Please get the tarballs from
> https://dist.apache.org/repos/dist/dev/subversion
> and add your signatures there. I plan to try and release on October
> 29th so please try and get your votes/signatures in place by October 28th.
We have 3/3 votes for Windows and 2/3 for Unix. I know Julian was having some
issues with his laptop trying to do his testing. If someone else has some time
to do some testing over the weekend it would be great.
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/23/13 1:47 PM, Ben Reser wrote:
> Yeah that's definitely the problem. When Apple built APR they did the
> following to set their C compiler: CC=`xcrun --find cc` which finds the C
> compiler in the XCode install instead of using the command line tool mapping
> that is now optional. The problem with doing that is that the compiler path is
> entirely dependent on the specific version of XCode and of course you don't
> have a compiler at the path they used.
>
> Using apr-1-config to determine the compiler is just silly but I'm guessing
> that's legacy from ctypes-python existing outside of our project before it was
> moved in.
I believe this is fixed in r1536537. Appreciate if you could test it out.
The commit doesn't do anything other than fix the one problem. There's
probably a lot more work that could go into the ctypes build setup.
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/23/13 12:27 PM, Branko Čibej wrote:
> This looks familiar:
>
> $ apr-1-config --cc
> /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc
>
>
> This is the bit from the (generated) makefile:
>
> # ctypes-python make targets
> ctypes-python: local-all
> $(SHELL) $(abs_srcdir)/build/run_ctypesgen.sh "$(LT_EXECUTE)" "$(CPPFLAGS)" "$(EXTRA_CTYPES_LDFLAGS)" "$(PYTHON)" "$(CTYPESGEN)" "$(abs_srcdir)" "$(abs_builddir)" "$(libdir)" "$(SVN_APR_CONFIG)" "$(SVN_APRUTIL_CONFIG)"
>
>
> and run_ctypesgen.sh has:
>
> apr_config="$9"
> ...
> cpp="`$apr_config --cpp`"
> ...
> $LT_EXECUTE $PYTHON $CTYPESGEN --cpp "$cpp $CPPFLAGS $cppflags" $ldflags $includes -o $output --no-macro-warnings --strip-build-path=$abs_srcdir
>
>
> Certainly looks like that could be the culprit. Not sure why we do that; it
> should be OK to use the same preprocessor and compiler as Subversion, I expect.
Yeah that's definitely the problem. When Apple built APR they did the
following to set their C compiler: CC=`xcrun --find cc` which finds the C
compiler in the XCode install instead of using the command line tool mapping
that is now optional. The problem with doing that is that the compiler path is
entirely dependent on the specific version of XCode and of course you don't
have a compiler at the path they used.
Using apr-1-config to determine the compiler is just silly but I'm guessing
that's legacy from ctypes-python existing outside of our project before it was
moved in.
> If that were the case, the plain vanilla tests would fail in the same way, as I
> suspect would the Python and Perl tests. We may have missed a path for ruby,
> though.
Found the Ruby problem. Apparently the Ruby bindings have been complete enough
that a mismatched libsvn_swig_ruby library wasn't blowing up on me. When I
fixed it getting the other SVN libraries loading out of the build was enough to
fix things for me so I missed that I completely screwed up getting the swig-rb
libraries properly setup. Fixed in r1535161.
Re: 1.8.4 up for testing/signing
Posted by Branko Čibej <br...@wandisco.com>.
On 23.10.2013 20:38, Ben Reser wrote:
> On 10/23/13 11:16 AM, Branko Čibej wrote:
>> $ ./configure ......
>> $ make
>> $ make ctypes-python
>> ...
>> ERROR: /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc -E -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK -I/usr/include/apr-1
> I'm assuming you're not setting your C compiler to that /Applications path. So
> I'd guess that's an upstream bug in ctypesgen with OS X. I'm still on OS X
> 10.7 and an older version of Xcode as a result so that's why I may not be
> seeing it.
This looks familiar:
$ apr-1-config --cc
/Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc
This is the bit from the (generated) makefile:
# ctypes-python make targets
ctypes-python: local-all
$(SHELL) $(abs_srcdir)/build/run_ctypesgen.sh "$(LT_EXECUTE)" "$(CPPFLAGS)" "$(EXTRA_CTYPES_LDFLAGS)" "$(PYTHON)" "$(CTYPESGEN)" "$(abs_srcdir)" "$(abs_builddir)" "$(libdir)" "$(SVN_APR_CONFIG)" "$(SVN_APRUTIL_CONFIG)"
and run_ctypesgen.sh has:
apr_config="$9"
...
cpp="`$apr_config --cpp`"
...
$LT_EXECUTE $PYTHON $CTYPESGEN --cpp "$cpp $CPPFLAGS $cppflags" $ldflags $includes -o $output --no-macro-warnings --strip-build-path=$abs_srcdir
Certainly looks like that could be the culprit. Not sure why we do that;
it should be OK to use the same preprocessor and compiler as Subversion,
I expect.
>> Probably. We should be able to make it cross-platform if -Djava.library.path
>> works. It's an inconvenience, not a blocker.
> I took a look. -Djava.library.path is setup fine. The problem is that the
> dynamic loader can't find the libraries that the JNI library needs. OS X
> rpath's can't be as easily adjusted as other platforms so the final rpath is in
> the library when we build it in tree. Other platforms the final rpath is
> applied during the install process. So we need to setup DYLD_LIBRARY_PATH
> here. Trunk is fixed with r1535115.
Oh, that's nice.
>>>> - 'make check-swig-rb' requires 'make install-swig-rb'
>>> Can't replicate this problem and in fact I remember specifically fixing this
>>> before 1.8.x branched. What error did you get?
>> Continuing from above:
>>
>> $ make swig-rb
>> $ make check-swig-rb
>> ...
>> /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle: dlopen(/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle, 9): Library not loaded: /opt/mine/svn-rel-1.8.x/lib/libsvn_swig_ruby-1.0.dylib (LoadError)
>> Referenced from: /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
>> Reason: image not found - /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
>> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/error.rb:21
>> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25:in `require'
>> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25
>> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47:in `require'
>> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47
>> make: *** [check-swig-rb] Error 1
>>
>>
>> I should point out that I give configure a custom prefix, in this case
>> /opt/mine/svn-rel-1.8.x, where it's looking for the native library. Possibly
>> another missing DYLD_LIBRARY_PATH. Again, an inconvenience rather than a blocker.
> Yeah that's rather peculiar. Since we do jump through hoops to setup the
> DYLD_LIBRARY_PATH. Perhaps there's some different behavior on 10.8 with
> respect to the the dynamic loader that's causing this problem.
If that were the case, the plain vanilla tests would fail in the same
way, as I suspect would the Python and Perl tests. We may have missed a
path for ruby, though.
-- Brane
--
Branko Čibej | Director of Subversion
WANdisco // Non-Stop Data
e. brane@wandisco.com
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/23/13 11:16 AM, Branko Čibej wrote:
> $ ./configure ......
> $ make
> $ make ctypes-python
> ...
> ERROR: /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc -E -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK -I/usr/include/apr-1 -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK -I/usr/include/apr-1 -Isubversion/include: /bin/sh: /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc: No such file or directory
> INFO: Status: Parsing /var/folders/zv/qq43zzc547zcbm_h9k900hnh0000gn/T/tmpnDfWGq.h
> INFO: Status: Processing description list.
> INFO: Status: Writing to subversion/bindings/ctypes-python/svn_all.py.
> INFO: Status: Wrapping complete.
> $ make check-ctypes-python
> ...
> cd /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python; \
> /bin/sh /Users/brane/src/svn/rel/subversion-1.8.4/libtool --tag=CC --silent --mode=execute `for f in /Users/brane/src/svn/rel/subversion-1.8.4/subversion/*/*.la; do echo -dlopen $f; done` /usr/bin/python test/run_all.py
> Traceback (most recent call last):
> File "test/run_all.py", line 22, in <module>
> import setup_path
> File "/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python/test/setup_path.py", line 31, in <module>
> import csvn.core
> File "/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python/csvn/core/__init__.py", line 134, in <module>
> f.argtypes = [POINTER(POINTER(svn_auth_provider_object_t)), POINTER(apr_pool_t)]
> NameError: name 'svn_auth_provider_object_t' is not defined
> make: *** [check-ctypes-python] Error 1
>
>
> I didn't investigate further, since apparently no-one actually uses the
> ctypes-python bindings.
I'm assuming you're not setting your C compiler to that /Applications path. So
I'd guess that's an upstream bug in ctypesgen with OS X. I'm still on OS X
10.7 and an older version of Xcode as a result so that's why I may not be
seeing it.
> Probably. We should be able to make it cross-platform if -Djava.library.path
> works. It's an inconvenience, not a blocker.
I took a look. -Djava.library.path is setup fine. The problem is that the
dynamic loader can't find the libraries that the JNI library needs. OS X
rpath's can't be as easily adjusted as other platforms so the final rpath is in
the library when we build it in tree. Other platforms the final rpath is
applied during the install process. So we need to setup DYLD_LIBRARY_PATH
here. Trunk is fixed with r1535115.
>>> - 'make check-swig-rb' requires 'make install-swig-rb'
>> Can't replicate this problem and in fact I remember specifically fixing this
>> before 1.8.x branched. What error did you get?
>
> Continuing from above:
>
> $ make swig-rb
> $ make check-swig-rb
> ...
> /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle: dlopen(/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle, 9): Library not loaded: /opt/mine/svn-rel-1.8.x/lib/libsvn_swig_ruby-1.0.dylib (LoadError)
> Referenced from: /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
> Reason: image not found - /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/error.rb:21
> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25:in `require'
> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25
> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47:in `require'
> from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47
> make: *** [check-swig-rb] Error 1
>
>
> I should point out that I give configure a custom prefix, in this case
> /opt/mine/svn-rel-1.8.x, where it's looking for the native library. Possibly
> another missing DYLD_LIBRARY_PATH. Again, an inconvenience rather than a blocker.
Yeah that's rather peculiar. Since we do jump through hoops to setup the
DYLD_LIBRARY_PATH. Perhaps there's some different behavior on 10.8 with
respect to the the dynamic loader that's causing this problem.
Re: 1.8.4 up for testing/signing
Posted by Branko Čibej <br...@wandisco.com>.
On 23.10.2013 18:49, Ben Reser wrote:
> On 10/23/13 1:07 AM, Branko Čibej wrote:
>> Issues:
>>
>> - 'make ctypes-python' fails
> Works for me. I had ctypesgen's at r147. I notice yours is at r151. I
> updated to r151 and rebuilt and it still worked just fine.
>
> Care to elaborate on what the problem or error message was?
$ ./configure ......
$ make
$ make ctypes-python
...
ERROR: /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc -E -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK -I/usr/include/apr-1 -DDARWIN -DSIGPROCMASK_SETS_THREAD_MASK -I/usr/include/apr-1 -Isubversion/include: /bin/sh: /Applications/Xcode.app/Contents/Developer/Toolchains/OSX10.8.xctoolchain/usr/bin/cc: No such file or directory
INFO: Status: Parsing /var/folders/zv/qq43zzc547zcbm_h9k900hnh0000gn/T/tmpnDfWGq.h
INFO: Status: Processing description list.
INFO: Status: Writing to subversion/bindings/ctypes-python/svn_all.py.
INFO: Status: Wrapping complete.
$ make check-ctypes-python
...
cd /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python; \
/bin/sh /Users/brane/src/svn/rel/subversion-1.8.4/libtool --tag=CC --silent --mode=execute `for f in /Users/brane/src/svn/rel/subversion-1.8.4/subversion/*/*.la; do echo -dlopen $f; done` /usr/bin/python test/run_all.py
Traceback (most recent call last):
File "test/run_all.py", line 22, in <module>
import setup_path
File "/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python/test/setup_path.py", line 31, in <module>
import csvn.core
File "/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/ctypes-python/csvn/core/__init__.py", line 134, in <module>
f.argtypes = [POINTER(POINTER(svn_auth_provider_object_t)), POINTER(apr_pool_t)]
NameError: name 'svn_auth_provider_object_t' is not defined
make: *** [check-ctypes-python] Error 1
I didn't investigate further, since apparently no-one actually uses the
ctypes-python bindings.
>> - Swig bindings do not build with the generated wrappers included in
>> the tarball. Running autogen.sh regenerates them and the bindings
>> then build succesfully.
> Known issue.
Ack.
>> - C tests do not work with the --enable-optimize configure option
> Should be fixed with the removal of the -fwhole-program option in the next release.
Ack. I expect it'll be gone by 1.8.5.
>> - wc-queries-test #3 fails with SQLite 3.8.0, but passes with
>> the default system version, 3.7.12.
>> -- There is a bug in the test: we should not fail a regression
>> test because we do not like how particular version of SQLite
>> does queries, as long as the query results are correct.
> I'd suggest that we start an independent thread on this. Because I suspect
> it's going to be a much longer discussion.
Ack.
>> - 'make check-javahl' requires 'make install'
> Looks like we need to setup the DYLD_LIBRARY_PATH or add more paths to
> -Djava.library.path for this to work properly on the Mac. Other platforms get
> away with rpaths for this.
Probably. We should be able to make it cross-platform if
-Djava.library.path works. It's an inconvenience, not a blocker.
>> - 'make check-swig-rb' requires 'make install-swig-rb'
> Can't replicate this problem and in fact I remember specifically fixing this
> before 1.8.x branched. What error did you get?
Continuing from above:
$ make swig-rb
$ make check-swig-rb
...
/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle: dlopen(/Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle, 9): Library not loaded: /opt/mine/svn-rel-1.8.x/lib/libsvn_swig_ruby-1.0.dylib (LoadError)
Referenced from: /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
Reason: image not found - /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/.ext/svn/ext/core.bundle
from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/error.rb:21
from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25:in `require'
from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/svn/core.rb:25
from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47:in `require'
from /Users/brane/src/svn/rel/subversion-1.8.4/subversion/bindings/swig/ruby/test/run-test.rb:47
make: *** [check-swig-rb] Error 1
I should point out that I give configure a custom prefix, in this case
/opt/mine/svn-rel-1.8.x, where it's looking for the native library.
Possibly another missing DYLD_LIBRARY_PATH. Again, an inconvenience
rather than a blocker.
-- Brane
--
Branko Čibej | Director of Subversion
WANdisco // Non-Stop Data
e. brane@wandisco.com
Re: 1.8.4 up for testing/signing
Posted by Ben Reser <be...@reser.org>.
On 10/23/13 1:07 AM, Branko Čibej wrote:
> Issues:
>
> - 'make ctypes-python' fails
Works for me. I had ctypesgen's at r147. I notice yours is at r151. I
updated to r151 and rebuilt and it still worked just fine.
Care to elaborate on what the problem or error message was?
> - Swig bindings do not build with the generated wrappers included in
> the tarball. Running autogen.sh regenerates them and the bindings
> then build succesfully.
Known issue.
> - C tests do not work with the --enable-optimize configure option
Should be fixed with the removal of the -fwhole-program option in the next release.
> - wc-queries-test #3 fails with SQLite 3.8.0, but passes with
> the default system version, 3.7.12.
> -- There is a bug in the test: we should not fail a regression
> test because we do not like how particular version of SQLite
> does queries, as long as the query results are correct.
I'd suggest that we start an independent thread on this. Because I suspect
it's going to be a much longer discussion.
> - 'make check-javahl' requires 'make install'
Looks like we need to setup the DYLD_LIBRARY_PATH or add more paths to
-Djava.library.path for this to work properly on the Mac. Other platforms get
away with rpaths for this.
> - 'make check-swig-rb' requires 'make install-swig-rb'
Can't replicate this problem and in fact I remember specifically fixing this
before 1.8.x branched. What error did you get?
> -- 'make install-swig-rb' tries to install "something" into the
> Ruby site dir, which requires root privilege
Confirmed.
Re: 1.8.4 up for testing/signing
Posted by Branko Čibej <br...@wandisco.com>.
Summary:
+1 to release
Platform
Mac OS X 10.8.5 Mountain Lion, build 12F45
Standard dependencies:
Apple clang(++) 5.0 (clang-500.2.75)/LLVM 3.3svn
APR 1.4.5
APR-Util 1.3.12
zlib 1.2.5
httpd 2.2.24
OpenSSL 0.9.8y
Java 1.6.0_51
Python 2.7.2
Perl 5.12.4
Ruby 1.8.7
(SQLite 3.7.12 ... see comments)
Dependencies from homebrew:
SQLite 3.8.0.2
Serf 1.3.2
BDB 5.3.21
Swig 2.0.11
libmagic 5.15
Other dependencies:
JUnit 4.11
ctypesgen svn-r151
Verified:
Tarball contents and signatures
(fsfs, bdb) x (local, svnserve, dav)
check-all-javahl
check-swig-py
check-swig-rb
check-swig-pl
Issues:
- 'make ctypes-python' fails
- Swig bindings do not build with the generated wrappers included in
the tarball. Running autogen.sh regenerates them and the bindings
then build succesfully.
- C tests do not work with the --enable-optimize configure option
- wc-queries-test #3 fails with SQLite 3.8.0, but passes with
the default system version, 3.7.12.
-- There is a bug in the test: we should not fail a regression
test because we do not like how particular version of SQLite
does queries, as long as the query results are correct.
- 'make check-javahl' requires 'make install'
- 'make check-swig-rb' requires 'make install-swig-rb'
-- 'make install-swig-rb' tries to install "something" into the
Ruby site dir, which requires root privilege.
GPG Signatures committed to the dist/dev/subversion repository.
--
Branko Čibej | Director of Subversion
WANdisco // Non-Stop Data
e. brane@wandisco.com