This is the mail archive of the gdb@sourceware.org mailing list for the GDB project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: status of Darwin support


On Tue, Aug 11, 2009 at 1:27 AM, Christian Thalinger
<Christian.Thalinger@sun.com> wrote:

> The assert you removed was at line 801, but my assert is at 792. ?And,
> obviously, it still asserts. ?These are the values:

Indeed, this is a different assert from the one that fired on Linux,
sorry I missed that.

> (gdb) r
> Starting program:
> /Users/twisti/bsd-port/hotspot/build/bsd/bsd_i486_compiler2/jvmg/gamma

I assume these are from debug prints of (section addr? or endaddr?) you've
added? They look like possibly non-relocated section addresses.

> 0x4fd3 <= 0x7040
> 0x70d1 <= 0xc9b7
> 0x5ffd <= 0x7040
> 0x6015 <= 0x7040
> 0x6034 <= 0x7040
> 0x6074 <= 0x7040
> 0x7014 <= 0x7040
> 0x4fd3 <= 0x7040
> 0x4fd3 <= 0x7040
> 0x6074 <= 0x7040
> 0x6015 <= 0x7040
> 0x6eee <= 0x7040
> 0x5ffd <= 0x6000
> 0x6015 <= 0x6018
> 0x6034 <= 0x6034
> 0x6074 <= 0x7000
> 0x4fd3 <= 0x1bc2
> objfiles.c:793: internal-error: qsort_cmp: Assertion
> `obj_section_endaddr (sect1) <= sect2_addr' failed.
>
> I'm not sure that helps...

Could you run gdb under /usr/bin/gdb, like this:

/usr/bin/gdb --args ./gdb -nx
/Users/twisti/bsd-port/hotspot/build/bsd/bsd_i486_compiler2/jvmg/gamma
(top-gdb) break error
(top-gdb) run
(gdb) run

### You may hit error several times before the 'qsort_cmp' fires.
### Once it does, go up to qsort_cmp level and (at the (top-gdb) prompt):

  print *sect1->the_bfd_section
  print *sect1->objfile
  print *sect2->the_bfd_section
  print *sect2->objfile

Thanks,
--
Paul Pluzhnikov


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]