Re: Request assistance with monotone package on alpha, hppa, mipsel, powerpc, s390
Hello Zack,
On Mon, Sep 03, 2007 at 06:59:28PM -0700, Zack Weinberg wrote:
in unstable, we decided to run the program's testsuite during the
build. This found failures on many architectures - a mixed blessing,
Well, in the end this is positive, of course.
This message is going out to all the porter lists for which the
monotone package presently FTBFS because of testsuite failures. I am
not a DD (the package is being sponsored by Ludovic Brenta) so I do
not have access to any of the Debian porting hosts, nor do I have
other access to any of your architectures.
Sorry, I cannot grant to access to an alpha, but maybe someone on this
list can.
I would therefore greatly appreciate assistance with these bugs. If
you have time to debug the thing yourself and send patches to
[EMAIL PROTECTED], that is of course the most helpful thing
you could do. Failing that, the next most helpful possibility is if
you are willing to grant me access (unprivileged only is fine) to a
machine under your control, and install monotone's build-dependencies
on that machine. If you can't do that either, at the least we might
have a chance of guessing what the problem is if you built the package
to the point where it fails (using a method that does not destroy the
build tree when the build fails), then tar up the tester_dir/
directory and the mtn binary and send them to me.
First, I see several cast from pointer to integer of different size,
I think these are the low hanging portability fixes to catch:
sqlite/func.c: In function 'trimFunc':
sqlite/func.c:891: warning: cast from pointer to integer of different size
sqlite/func.c: In function 'sqlite3RegisterBuiltinFunctions':
sqlite/func.c:1382: warning: cast to pointer from integer of different size
sqlite/func.c:1401: warning: cast to pointer from integer of different size
sqlite/table.c: In function 'sqlite3_get_table':
sqlite/table.c:148: warning: cast to pointer from integer of different size
sqlite/table.c: In function 'sqlite3_free_table':
sqlite/table.c:193: warning: cast from pointer to integer of different size
sqlite/vdbemem.c: In function 'sqlite3ValueText':
sqlite/vdbemem.c:857: warning: cast from pointer to integer of different size
sqlite/vtab.c: In function 'sqlite3VtabRollback':
sqlite/vtab.c:587: warning: cast from pointer to integer of different size
sqlite/vtab.c: In function 'sqlite3VtabCommit':
sqlite/vtab.c:596: warning: cast from pointer to integer of different size
For the tests:
echo '#!/bin/sh' run_tester_tests ; \
echo 'PATH=.:$PATH' run_tester_tests ; \
echo 'exec ./tester ./tester-testsuite.lua $@' run_tester_tests ;
\chmod 755 run_tester_tests
echo '#!/bin/sh' run_lua_tests ; \
echo 'PATH=.:$PATH' run_lua_tests ; \
echo 'exec ./tester ./testsuite.lua $@' run_lua_tests ; \
chmod 755 run_lua_tests
Running unit tests...
basic_io:binary_transparency ok
charset:idna_encoding ok
charset:utf8_validation ok
commands:command_complete_command ok
commands:command_find_command ok
commands:complete_command ok
commands:make_command_id ok
crypto:calculate_identok
cset:basic_csets ok
cset:cset_written /bin/sh: line 4: 15213
Segmentation fault AUTOTEST_PATH=. ${dir}$tst
FAIL: unit_tests
Running tests...
1 isolated-1ok
2 isolated-2ok
3 cleanup-1 ok
4 cleanup-2 ok
5 remove-unwriteableok
Of 5 tests run:
5 succeeded
0 failed
0 had expected failures
0 succeeded unexpectedly
0 were skipped
PASS: run_tester_tests
/bin/sh: line 4: 15224 Segmentation fault AUTOTEST_PATH=. ${dir}$tst
FAIL: run_lua_tests
==
2 of 3 tests failed
Please report to [EMAIL PROTECTED]
==
make[4]: *** [check-TESTS] Error 1
If I run the tests manually, I get the following results (abbreviated,
I cannot get the output captured (the ordinary foo 21 does not
work):
run_tester_tests:
Running tests...
1 isolated-1ok
2 isolated-2ok
3 cleanup-1 ok
4 cleanup-2 ok
5 remove-unwriteableok
Of 5 tests run:
5 succeeded
0 failed
0 had expected failures
0 succeeded unexpectedly
0 were skipped
./tester ./tester-testsuite.lua
Running tests...
1 isolated-1ok
2 isolated-2ok
3 cleanup-1
Re: Request assistance with monotone package on alpha, hppa, mipsel, powerpc, s390
On Mon, Sep 03, 2007 at 06:59:28PM -0700, Zack Weinberg wrote: I'm one of the maintainers of the Debian package of monotone; I'm also one of its upstream developers. For the new 0.36-1 package presently in unstable, we decided to run the program's testsuite during the build. This found failures on many architectures - a mixed blessing, since on the one hand it appears that the program has catastrophic bugs that need fixing, but on the other hand, now the package has been kicked out of testing. :-( Zack, I kicked this off on my testing/unstable hybrid hppa box: dpkg-buildpackage -rfakeroot -uc -us Source/Build tree is on: iou.parisc-linux.org/~grundler/monotone-0.36/ hppa is a 32-bit user space and this system is running a 64-bit kernel. Please let me know off list if you would like access to that machine and/or ia64-linux machines. cheers, grant This message is going out to all the porter lists for which the monotone package presently FTBFS because of testsuite failures. I am not a DD (the package is being sponsored by Ludovic Brenta) so I do not have access to any of the Debian porting hosts, nor do I have other access to any of your architectures. I would therefore greatly appreciate assistance with these bugs. If you have time to debug the thing yourself and send patches to [EMAIL PROTECTED], that is of course the most helpful thing you could do. Failing that, the next most helpful possibility is if you are willing to grant me access (unprivileged only is fine) to a machine under your control, and install monotone's build-dependencies on that machine. If you can't do that either, at the least we might have a chance of guessing what the problem is if you built the package to the point where it fails (using a method that does not destroy the build tree when the build fails), then tar up the tester_dir/ directory and the mtn binary and send them to me. Thanks in advance, zw -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Request assistance with monotone package on alpha, hppa, mipsel, powerpc, s390
On Mon, Sep 03, 2007 at 06:59:28PM -0700, Zack Weinberg wrote: I'm one of the maintainers of the Debian package of monotone; I'm also one of its upstream developers. For the new 0.36-1 package presently in unstable, we decided to run the program's testsuite during the build. This found failures on many architectures - a mixed blessing, since on the one hand it appears that the program has catastrophic bugs that need fixing, but on the other hand, now the package has been kicked out of testing. :-( This message is going out to all the porter lists for which the monotone package presently FTBFS because of testsuite failures. I am not a DD (the package is being sponsored by Ludovic Brenta) so I do not have access to any of the Debian porting hosts, nor do I have other access to any of your architectures. I would therefore greatly appreciate assistance with these bugs. If you have time to debug the thing yourself and send patches to [EMAIL PROTECTED], that is of course the most helpful thing you could do. Failing that, the next most helpful possibility is if you are willing to grant me access (unprivileged only is fine) to a machine under your control, and install monotone's build-dependencies on that machine. If you can't do that either, at the least we might have a chance of guessing what the problem is if you built the package to the point where it fails (using a method that does not destroy the build tree when the build fails), then tar up the tester_dir/ directory and the mtn binary and send them to me. I am currently attempting a build in my unstable chroot on my alpha. It might take a while. My mipsel isn't hooked up at the moment so I can't try that one (and it would take more than a while on it). I don't own any hppa, ppc or s390 at this time. I only have m68k, mipsel, sparc, alpha, x86 and amd64, so far, and have access to arm. Well and a mips (not el) but it isn't supported by linux at this time. :) -- Len Sorensen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Request assistance with monotone package on alpha, hppa, mipsel, powerpc, s390
I'm one of the maintainers of the Debian package of monotone; I'm also one of its upstream developers. For the new 0.36-1 package presently in unstable, we decided to run the program's testsuite during the build. This found failures on many architectures - a mixed blessing, since on the one hand it appears that the program has catastrophic bugs that need fixing, but on the other hand, now the package has been kicked out of testing. :-( This message is going out to all the porter lists for which the monotone package presently FTBFS because of testsuite failures. I am not a DD (the package is being sponsored by Ludovic Brenta) so I do not have access to any of the Debian porting hosts, nor do I have other access to any of your architectures. I would therefore greatly appreciate assistance with these bugs. If you have time to debug the thing yourself and send patches to [EMAIL PROTECTED], that is of course the most helpful thing you could do. Failing that, the next most helpful possibility is if you are willing to grant me access (unprivileged only is fine) to a machine under your control, and install monotone's build-dependencies on that machine. If you can't do that either, at the least we might have a chance of guessing what the problem is if you built the package to the point where it fails (using a method that does not destroy the build tree when the build fails), then tar up the tester_dir/ directory and the mtn binary and send them to me. Thanks in advance, zw -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
