Bug 2936 - [tracker] Backtraces
Summary: [tracker] Backtraces
Status: ASSIGNED
Alias: None
Product: frysk
Classification: Unclassified
Component: general (show other bugs)
Version: unspecified
: P2 normal
Target Milestone: ---
Assignee: Unassigned
URL:
Keywords:
Depends on: 3076 3259 3590 3601 3649 3728 3068 3070 3071 3258 3529 3723 3814
Blocks: 2935 3177 3192 3198 3250
  Show dependency tree / graph
 
Reported: 2006-07-17 18:51 UTC by Adam Jocksch
Modified: 2006-12-15 17:25 UTC (History)
6 users (show)

See Also:
Host:
Target:
Build:
Last reconfirmed:

Attachments
Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Adam Jocksch 2006-07-17 18:51:04 UTC 
We are currently unable to get a stackbacktrace using libunwind
Comment 1 Mark Wielaard 2006-09-11 09:25:02 UTC 
When running the frysk.rt.tests.TestStackBacktrace test on something that is not
x86 you get:

1) testBacktrace(frysk.rt.tests.TestStackBacktrace)java.lang.NullPointerException
   at frysk.rt.StackCallbacks.accessReg(TestRunner)
   at lib.unwind.StackTraceCreator.unwind_setup(TestRunner)
   at lib.unwind.StackTraceCreator.createStackTrace(TestRunner)
   at frysk.rt.StackFactory.createStackFrame(TestRunner)
   at frysk.rt.tests.TestStackBacktrace.testBacktrace(TestRunner)
   at frysk.junit.Runner.<init>(TestRunner)
   at TestRunner.main(TestRunner)

Currently the callbacks are setup through the RegisterX86 class which only
hardcodes x86 registers.
Comment 2 Mike Cvet 2006-09-12 15:27:25 UTC 
Currently libunwind is not using dwarf2 - based remote unwinding, and is instead
falling back to using frame pointers - which will not work on anything non-x86.
Since binaries are often compiled with -fomit-frame-pointers anyway, this needs
to get fixed.