[Rspamd-Users] rspamd: *** stack smashing detected ***: rspamd: controller process (localhost:11334) terminated

[Vsevolod Stakhov]

On 25/04/2019 12:54, Vsevolod Stakhov wrote:
> On 24/04/2019 11:49, Thomas Plant via Users wrote:
>> Hello,
>>
>> found this message in /var/log/messages:
>>
>> Apr 24 09:24:05 rspamd rspamd: *** stack smashing detected ***: rspamd:
>> controller process (localhost:11334) terminated
>>
>> Anyone seen such a message in the logs? In rspamd.log I see at the same
>> time the following:
>>
>> 2019-04-24 09:24:05 #13379(controller) rspamd_crash_sig_handler: caught
>> fatal signal 11(Segmentation fault), pid: 13379, trace:
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 0:
>> 00007F3709D500B8: _Unwind_GetTextRelBase()+0x1bf8
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 1:
>> 00007F3709D50FB9: _Unwind_Backtrace()+0x69
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 2:
>> 00007F3709A87F16: backtrace()+0x56
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 3:
>> 00007F37099ECD44: __libc_message()+0x3f4
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 4:
>> 00007F3709A8B9E7: __fortify_fail()+0x37
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 5:
>> 00007F3709A8B9A2: __stack_chk_fail()+0x12
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 6:
>> 00007F370E996F96: rspamd_vprintf_common()+0x1066
>> 2019-04-24 09:24:05 #13379(controller) rspamd_print_crash: 7:
>> 3030303030303030: <unknown>
>> 2019-04-24 09:24:05 #12306(main) <6281a5>; main; rspamd_cld_handler:
>> catch SIGCHLD signal, finding terminated workers
>> 2019-04-24 09:24:05 #12306(main) <6281a5>; main; rspamd_cld_handler:
>> controller process 13379 terminated abnormally by signal: Segmentation
>> fault but NOT created core file (throttled=no); core file limits: 0
>> current, -1 max
>> 2019-04-24 09:24:07 #26574(controller) <6281a5>; main;
>> rspamd_worker_set_limits: use system max file descriptors limit: 1024KiB
>> cur and 1024KiB max
>> 2019-04-24 09:24:07 #26574(controller) <6281a5>; main;
>> rspamd_worker_set_limits: use system max core size limit: 0B cur and -1B max
>>
> 
> 
> I will ignore this issue unless you all stop ignoring the FAQ, in
> particular the following entry:
> 
> https://rspamd.com/doc/faq.html#how-to-figure-out-why-rspamd-process-crashed
> 
> 

It seems it was caused by printing of some small negative floating point
values. I have found something similar in one of my artificial tests.