On 30/08/11 at 21:08 +0200, Petr Salinger wrote: > >Could you try make test-all ? While some failures and errors are > >expected, it stresses the interpreter a bit more, so it's a good way to > >check that it doesn't block. > > Under sid, it fails to start, probably due to > multiarch changes of libc location: > > /build/manual/ruby1.9.1-1.9.3~preview1+svn33077/.ext/common/dl/import.rb:227:in > `import_function': cannot find the function: strcpy() (DL::DLError) > from > /build/manual/ruby1.9.1-1.9.3~preview1+svn33077/.ext/common/dl/import.rb:133:in > `extern' > from > /build/manual/ruby1.9.1-1.9.3~preview1+svn33077/test/dl/test_import.rb:12:in > `<module:LIBC>' > from > /build/manual/ruby1.9.1-1.9.3~preview1+svn33077/test/dl/test_import.rb:5:in > `<module:DL>'
Oh, ./test/dl/test_base.rb and ./test/fiddle/helper.rb probably need some changes. That code is really stupid. I've added to my todo list to fix it, but if you want to take a look so that you can continue testing, it should be trivial. > Under squeeze runs up to > > TestDRbCore#test_03 = 0.06 s = . > TestDRbCore#test_04 = > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/lib/timeout.rb:61: > SEGV received in SEGV handler > [BUG] Segmentation fault > ruby 1.9.3dev (2011-08-26) [x86_64-kfreebsd-gnu] > > -- Control frame information > ----------------------------------------------- > c:0004 p:---- s:0011 b:0011 l:000010 d:000010 CFUNC :sleep > c:0003 p:0014 s:0007 b:0007 l:001c10 d:000006 BLOCK > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/lib/timeout.rb:61 > c:0002 p:---- s:0004 b:0004 l:000003 d:000003 FINISH > c:0001 p:---- s:0002 b:0002 l:000001 d:000001 TOP > > -- Ruby level backtrace information > ---------------------------------------- > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/lib/timeout.rb:61:in > `block in timeout' > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/lib/timeout.rb:61:in > `sleep' > > -- C level backtrace information > ------------------------------------------- > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/libruby-1.9.1.so.1.9(+0x15256f) > [0x80097256f] vm_dump.c:796 > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/libruby-1.9.1.so.1.9(+0x59754) > [0x800879754] error.c:258 > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/libruby-1.9.1.so.1.9(rb_bug+0xb8) > [0x8008798c8] error.c:277 > /opt/sid/build/manual/ruby1.9.1-1.9.3~preview1+svn33077/libruby-1.9.1.so.1.9(+0xf0cd5) > [0x800910cd5] signal.c:609 > /lib/libpthread.so.0(+0xc244) [0x800c31244] > [0x7fffffffffc3] This is related to timer thread (which is used to handle timeouts). It might be worth debugging, as it could be an indication that your patch doesn't handle a corner case. I've added info on how to run specific tests to git: http://anonscm.debian.org/gitweb/?p=collab-maint/ruby1.9.1.git;a=blob;f=debian/README.porters;h=94aff7dd3ecd9f748498f2e285b4a4313b4b8f36;hb=f3ac1c548cfc23a32ec54df480547c6634381f53 Due to the differences of kfreebsd, I think that we should aim for: - interpreter never crashes or hangs - it's "OK" if some identified tests fail due to identified problems, because it would be very hard to avoid. What worries me is that some applications, such as puppet (used by DSA), might break with 1.9. That would be a huge problem if we decide to switch to 1.9 by default and drop 1.8 in wheezy. Lucas -- To UNSUBSCRIBE, email to debian-bugs-rc-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
