Re: [PATCH/i386newframe] amd64newframe

[Mark Kettenis <kettenis@chello.nl>]

   From: Andreas Jaeger <aj@suse.de>
   Date: Sat, 17 May 2003 20:33:56 +0200

Andreas, thanks for testing this.

   building the kettenis-newframe-branch on x86-64 I get:

   make[1]: *** No rule to make target `inux-proc.o', needed by `libgdb.a'.  Stop.
   make[1]: Leaving directory `/usr/src/aj/gdb-test/build/gdb'
   make: *** [all-gdb] Error 2

   Note the following:
   ./config/i386/x86-64linux.mh:   proc-service.o thread-db.o lin-lwp.o inux-proc.o gcore.o 

Oops!  Thanks, fixed now.

   After fixing this problem, gdb build fine but running it I got:

   byrd:/usr/src/aj/gdb-test/build/gdb:[0]$ ./gdb ~/tmp/a.out 
   A handler for the OS ABI "GNU/Linux" is not built into this configuration of GDB.  Attempting to continue with the default i386 settings

Ah, GDB is trying to build a default architecture vector, for which it
happens to choose Linux/i386.  This can be solved by linking in
i386-linux-tdep.o.  I'll fix this too.

   GNU gdb 2003-05-04-cvs
   Copyright 2003 Free Software Foundation, Inc.
   GDB is free software, covered by the GNU General Public License, and you are
   welcome to change it and/or distribute copies of it under certain conditions.
   Type "show copying" to see the conditions.
   There is absolutely no warranty for GDB.  Type "show warranty" for details.
   This GDB was configured as "x86_64-unknown-linux-gnu"...
   Setting up the environment for debugging gdb.
   .gdbinit:5: Error in sourced command file:
   Function "internal_error" not defined.
   (gdb) b main
   Breakpoint 1 at 0x40040c: file b.c, line 4.
   (gdb) r
   Starting program: /suse/aj/tmp/a.out 

   Breakpoint 1, main () at b.c:4
   4       lg1[1] = 1;
   (gdb) n
   warning: Unmapped DWARF Register #16 encountered


   warning: Unmapped DWARF Register #16 encountered

   5       lg2[1] = 1;
   (gdb) 
   warning: Unmapped DWARF Register #16 encountered

   6       return 0;
   (gdb) 
   warning: Unmapped DWARF Register #16 encountered

   7       }

Ah, that must be the new DWARF CFI frame unwinder.  The return address
RA isn't mapped.  I could change the DWARF register mapping in
x86-64-tdep.c, but this really is a problem with the DWARF CFI frame
unwinder since the compiler is free to choose any number it wants for
the return address column.  Should be fixed now.

   And this also:

   $ ./gdb /bin/ls 
   A handler for the OS ABI "GNU/Linux" is not built into this configuration of GDB.  Attempting to continue with the default i386 settingsGNU gdb 2003-05-04-cvs
   Copyright 2003 Free Software Foundation, Inc.
   GDB is free software, covered by the GNU General Public License, and you are
   welcome to change it and/or distribute copies of it under certain conditions.
   Type "show copying" to see the conditions.
   There is absolutely no warranty for GDB.  Type "show warranty" for details.
   This GDB was configured as "x86_64-unknown-linux-gnu"...
   (no debugging symbols found)...
   Setting up the environment for debugging gdb.
   .gdbinit:5: Error in sourced command file:
   Function "internal_error" not defined.
   (gdb) b main
   Function "main" not defined.
   (gdb) r
   Starting program: /bin/ls 
   (no debugging symbols found)...[New Thread 16384 (LWP 4561)]

   ../../src/gdb/arch-utils.c:443: internal-error: generic_register_size: Assertion `regnum >= 0 && regnum < NUM_REGS + NUM_PSEUDO_REGS' failed.
   A problem internal to GDB has been detected.  Further
   debugging may prove unreliable.

Hmm.  Something is defenitely wrong here.  Hard to tell without a
backtrace though what's really wrong.

   I don't have time to go into details here but hope that Michal can.

Hope he can.

Thanks again,

Mark