bug-hurd
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Core dump (semi) update


From: Alfred M. Szmidt
Subject: Re: Core dump (semi) update
Date: Mon, 9 Feb 2004 19:48:17 +0100 (MET)

   All the pr_*id values look sane to me.  pr_[ug]id and pr_egid are
   (unsigned)-1.  If you look at elfcore.c, this is what to expect.

Ahhh... Why didn't I think of that.

   > If you want to check it yourself, I can always send you the core
   > dump, it is only 30k compressed.

   Sure.

You can snatch it from: http://user.it.uu.se/~alsz8747/core.23268

   > |       pr_reg = {31, 31, 31, 31, 0, 3, 16873244, 2, 18340376,
   > |                 18579400, 0, -25, 0, 0, 17004764, 23, 17004764, 23, 
17004764}, 

   Are these sane register values for your program?

No, not really.  It was from gencat waiting for input from stdin.
eflags seems completely bogus.  And so do the other registers...

   Ok.  The first thing to check here is that the pseudo-sections are
   showing the right locations.  That is, use objdump -h and see that
   the file offsets given for .reg and the like match what they
   should.

,----
| Idx Name          Size      VMA       LMA       File off  Algn
|   1 .reg/23268    0000004c  00000000  00000000  00002554  2**2
|                   CONTENTS
|   2 .reg          0000004c  00000000  00000000  00002554  2**2
|                   CONTENTS
|   3 .reg2/23268   00000000  00000000  00000000  00000000  2**2
|                   CONTENTS
|   4 .reg2         00000000  00000000  00000000  00000000  2**2
|                   CONTENTS
|   5 .reg/88804    0000004c  00000000  00000000  00002744  2**2
|                   CONTENTS
|   6 .reg2/88804   00000000  00000000  00000000  00000000  2**2
|                   CONTENTS
`----

Offsets look more or less decent I think, though I only did a quick
look.

   > ,----
   > | hurd:/ams# ./sources.redhat.com/sources.redhat.com/obj/gdb/gdb 
/sub-hurd/bin/gencat core.23268 
[...snip...]
   > | #0  0x00000000 in ?? ()
   > | (gdb) 
   > `----

   Show `info regs' here.  That will show what info gdb got, and if
   it's all zeros that's suspicious in one way, while if it has some
   values, exactly what those are might tell us something else.

Seems to be the same as in pr_reg.

,----
| (gdb) info registers 
| eax            0xffffffe7       -25
| ecx            0x0      0
| edx            0x11b7fc8        18579400
| ebx            0x117da18        18340376
| esp            0x17     0x17
| ebp            0x101771c        0x101771c
| esi            0x3      3
| edi            0x0      0
| eip            0x0      0x0
| eflags         0x10378dc        17004764
| cs             0x0      0
| ss             0x10378dc        17004764
| ds             0x1f     31
| es             0x1f     31
| fs             0x1f     31
| gs             0x1f     31
`----

Sorry for the short info. here, in a hurry....




reply via email to

[Prev in Thread] Current Thread [Next in Thread]