Xen devel - May 2006 - Gdbserver and --attach (stepping disabled since 8ed131452f27)

Hi,

Changeset 8ed131452f27 (March 6th, patch submitted by Horms) disabled
single-stepping for attached debugging:

     case PTRACE_SETREGS:
+        if (!current_isfile)
+                goto out_unspported; /* XXX not yet supported */
         SET_XC_REGS(((struct gdb_regs *)data), ctxt[cpu].user_regs);
         if ((retval = xc_vcpu_setcontext(xc_handle, current_domid, cpu, 
                                 &ctxt[cpu])))
@@ -407,6 +390,8 @@
         break;
 
     case PTRACE_SINGLESTEP:
+        if (!current_isfile)
+              goto out_unspported; /* XXX not yet supported */
         /*  XXX we can still have problems if the user switches threads
          *  during single-stepping - but that just seems retarded
          */

Is there any special reason while this was done? The description does
not comment this. Reverting this part of the patch allows me to
single-step the attached domain again (but see below).


There appears to be some kind of heap corruption as the gdb server
dies with

*** glibc detected *** malloc(): memory corruption: 0x08094fc8 ***

Program received signal SIGABRT, Aborted.
0xbfffe402 in __kernel_vsyscall ()
(gdb) bt
#0  0xbfffe402 in __kernel_vsyscall ()
#1  0xb7dc1991 in raise () from /lib/tls/i686/cmov/libc.so.6
#2  0xb7dc333b in abort () from /lib/tls/i686/cmov/libc.so.6
#3  0xb7df8af7 in __fsetlocking () from /lib/tls/i686/cmov/libc.so.6
#4  0xb7e000d0 in free () from /lib/tls/i686/cmov/libc.so.6
#5  0xb7e013f6 in malloc () from /lib/tls/i686/cmov/libc.so.6
#6  0x08049f23 in putpkt (buf=0x8054048
"T0505:344a2ec0;04:2c4a2ec0;08:a90900c0;thread:0;") at
../../../gdb-6.2.1/gdb/gdbserver/remote-utils.c:273
#7  0x0804bf8a in main (argc=4, argv=0xbfe0b904) at
../../../gdb-6.2.1/gdb/gdbserver/server.c:630

// Simon

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

On 11 May 2006, at 11:16, Simon Kagstrom wrote:
> Is there any special reason while this was done? The description does
> not comment this. Reverting this part of the patch allows me to
> single-step the attached domain again (but see below).
He quite obviously got the condition round the wrong way. You can''t 
single-step a core dump.

  -- Keir


_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel

At Thu, 11 May 2006 11:29:10 +0100,
Keir Fraser wrote:
> 
> 
> On 11 May 2006, at 11:16, Simon Kagstrom wrote:
> 
> > Is there any special reason while this was done? The description does
> > not comment this. Reverting this part of the patch allows me to
> > single-step the attached domain again (but see below).
> 
> He quite obviously got the condition round the wrong way. You
can''t
> single-step a core dump.
Sorry, I should have seen that...

// Simon

_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xensource.com
http://lists.xensource.com/xen-devel