[padb-devel] [padb] r205 committed - Effectively disable the code to re-try getting stack traces with gdb b...

[codesite-noreply at google.com]

Revision: 205
Author: apittman
Date: Mon Sep  7 14:11:38 2009
Log: Effectively disable the code to re-try getting stack traces with gdb  
by setting
gdb-retry-count to zero.  This used to be helpful on ia64 but just costs  
time
on machines where gdb is more capable.

http://code.google.com/p/padb/source/detail?r=205

Modified:
  /branches/cleanup/src/padb

=======================================
--- /branches/cleanup/src/padb	Mon Sep  7 13:54:28 2009
+++ /branches/cleanup/src/padb	Mon Sep  7 14:11:38 2009
@@ -6174,6 +6174,15 @@
  # not spinning whilst gdb is doing it's thing which will mean a quicker  
runtime
  # but also that the resulting stack traces will have less artifacts  
because running
  # processes bunch up behind the non-running ones.
+
+# We used to reguarly get garbage from gdb so there is a test here for  
finding
+# main, with code to detach and try again if we don't.  This served us  
well on
+# ia64 where gdb isn't very good however it as on most machines gdb gives
+# you results below main (__libc_start_main()) this test fails which causes
+# padb to loop a number of times for each procees on a node.  For now I've  
left
+# the code to loop here but have set the default gdb_retry_count to zero to
+# disable the code in the simple case.  This option will probably be  
removed
+# completely soon.
  sub stack_trace_from_pids {
      my ( $carg, $list ) = @_;

@@ -7434,7 +7443,7 @@
          options_i   => {
              stack_shows_params => 0,
              stack_shows_locals => 0,
-            gdb_retry_count    => 3,
+            gdb_retry_count    => 0,
              stack_strip_above =>
  'elan_waitWord,elan_pollWord,elan_deviceCheck,opal_condition_wait,opal_progress',
              stack_strip_below => 'main',