Re: [ECOS] More question on GDB Stub for NECVR4373 board.

["Ling Su" <lingsu at palmmicro dot com>]

Thanks a lot! It is the Endian problem. The PMON is little endian and the
compiler produces big endian code.

Best Regards,
-Ling

> > Hello, everyone,
> >
> > Since I met troubles to burn the gdbstub ROM for NECvr4373 board, I'd
like
> > to looking some help from the list. More specifically, I hope someone
can
> > help answer following questions, Thanks in advance!
> >
> > <1>. Has anyone successfully run eCos via the gdbstub on NECvr4373
board? I
> > knew Jifl did it, I'd like to query if anyone else succeed in doing
this?
> We did and we're doing.
>
> >
> > <2>. I observed the PCI bus signal after I mounted  the gdbstub EPROM
that I
> > burned, there is no any active signal showing out there. Since the UART
is
> > on South bridge chip, it must be accessed via PCI bus, this mean
something
> > wrong with my gdbstub or my board. I doubt if my board doesn't work, or
at
> > least not fully functional. (But I am pretty sure the PMON comes along
with
> > the board works very well.) This is the reason I'd like to check if
anyone
> > else successd in burning the GDBSTUB except Jifl.
> Please notice the endianess problem. The PMON runs in little endian while
the
> default comipler settings of ecos is big endian (if you don't know what
> endianess is, read any MIPS books). You have to disconnect the jumper JP66
on
> the board to set the endianess to be big.  After you have done this, I
guess
> you can connect to the gdb by your stub