At Sun, 27 Nov 2011 16:56:56 +0800, Liu Yuan wrote: > > On 11/25/2011 08:59 PM, MORITA Kazutaka wrote: > > > Yuan, can you check whether you can start a sheep daemon with the > > patchset I sent just now? > > > > > Still got the following error: > > ... > sheep: worker_routine(206) started this thread 0 > *** longjmp causes uninitialized stack frame ***: sheep/sheep terminated > sheep: worker_routine(206) started this thread 0 > sheep: worker_routine(206) started this thread 0 > ======= Backtrace: ========= > /lib/libc.so.6(__fortify_fail+0x37)[0x7fa82e414537] > /lib/libc.so.6(+0xff4c9)[0x7fa82e4144c9] > /lib/libc.so.6(__longjmp_chk+0x33)[0x7fa82e414433] > sheep/sheep[0x417710] > sheep/sheep[0x4177ae] > sheep/sheep[0x415716] > sheep/sheep[0x403a2a] > /lib/libc.so.6(__libc_start_main+0xfe)[0x7fa82e333d8e] > sheep/sheep[0x403439] > ... Thanks for your report, it seems that glibc's stack check doesn't work with our coroutine codes. I sent an update patch to disable it. Thanks, Kazutaka |