[ 
https://issues.apache.org/jira/browse/TS-272?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12849028#action_12849028
 ] 

Leif Hedstrom commented on TS-272:
----------------------------------

The problem is a "new" requirement on finding Records stuff in signals.cc. 
Disabling this fixes the problem:

Index: proxy/signals.cc
===================================================================
--- proxy/signals.cc    (revision 926615)
+++ proxy/signals.cc    (working copy)
@@ -465,13 +465,13 @@ check_signal_thread(void *)
 void
 init_signals()
 {
-  RecInt stackDump;
-  bool found = (RecGetRecordInt("proxy.config.stack_dump_enabled", &stackDump) 
== REC_ERR_OKAY);
+  //RecInt stackDump;
+  //  bool found = (RecGetRecordInt("proxy.config.stack_dump_enabled", 
&stackDump) == REC_ERR_OKAY);
 
-  if(found == false) {
-    Warning("Unable to determine stack_dump_enabled , assuming enabled");
-    stackDump = 1;
-  }
+  //  if(found == false) {
+  //    Warning("Unable to determine stack_dump_enabled , assuming enabled");
+  //    stackDump = 1;
+  //  }
 
   sigset_t sigsToBlock;
   sigemptyset(&sigsToBlock);
@@ -482,10 +482,10 @@ init_signals()
   set_signal(SIGTERM, (SigActionFunc_t) signal_handler);
   set_signal(SIGHUP, (SigActionFunc_t) interrupt_handler);
   set_signal(SIGILL, (SigActionFunc_t) signal_handler);
-  if(stackDump == 1) {
+  //  if(stackDump == 1) {
     set_signal(SIGBUS, (SigActionFunc_t) signal_handler);
     set_signal(SIGSEGV, (SigActionFunc_t) signal_handler);
-  }
+    //  }
 
 //
 //    Presviously the following lines were #if 0


We'll have to check with VJ if this is something that needs to be in there or 
not.


> logcat and logstats segfault
> ----------------------------
>
>                 Key: TS-272
>                 URL: https://issues.apache.org/jira/browse/TS-272
>             Project: Traffic Server
>          Issue Type: Bug
>          Components: Logging
>    Affects Versions: 2.1.0, 2.0.0a
>            Reporter: Leif Hedstrom
>             Fix For: 2.0.0
>
>
> Both trunk and 2.0.x branch segfaults when trying to run traffic_logcat or 
> traffic_logstats:
> (gdb) bt
> #0  0x00000000006ef566 in ink_hash_table_lookup_entry (ht_ptr=0x0, 
> key=0x709b58 "proxy.config.stack_dump_enabled") at ink_hash_table.cc:249
> #1  0x00000000006ef4aa in ink_hash_table_lookup (ht_ptr=0x0, key=0x709b58 
> "proxy.config.stack_dump_enabled", value_ptr=0x7fffffffe160) at 
> ink_hash_table.cc:190
> #2  0x00000000006db02a in RecGetRecord_Xmalloc (name=0x709b58 
> "proxy.config.stack_dump_enabled", data_type=RECD_INT, data=0x7fffffffe1a0, 
> lock=true) at RecCore.cc:915
> #3  0x00000000006da333 in RecGetRecordInt (name=0x709b58 
> "proxy.config.stack_dump_enabled", rec_int=0x7fffffffe248, lock=true) at 
> RecCore.cc:454
> #4  0x00000000004f95ad in init_signals () at signals.cc:469
> #5  0x00000000004cdb5a in init_system () at ./logging/LogStandalone.cc:125
> #6  0x00000000004ce2d2 in init_log_standalone_basic (pgm_name=0x7034d8 
> "traffic_logcat") at ./logging/LogStandalone.cc:329
> #7  0x00000000004cebf6 in main (argc=1, argv=0x7fffffffe468) at logcat.cc:271

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to