remote failure reply

[Daniel Jacobowitz]

On Thu, Nov 03, 2005 at 07:49:04PM +0100, Efim Monjak wrote:
> Hi all,
> 
> I try to debug a relatively big programm with arm-elf-gdb 6.3.50. After 
> a couple of steps I receive a failure message from GDB
> about register message. But later the GDB has receive an other register 
> message withother failure.
> Can you tell me what is hier a problem? As I used GDB 6.1 I have had a 
> problem with remote target
> depend on Walues in Registers. In this case GDB sent +---+--- .... In 
> snapshort 6.3.50 it is fixed.
> Possibly it is now a similar problem.

This is why it's traditional to use lowercase hex, not uppercase, in
remote replies.

> Packet received: 
> E03D004032000000D13C004002000000800000003CFEFF7F2C01004000000000DC813F28F26AC140F03B0040EC3D0040703C0040C83C0040F00F0080F40B0080000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000005F000020

This is interpreted as an error because it begins with 'E'.

Perhaps it shouldn't be, but for compatibility, I recommend fixing the
stub.

-- 
Daniel Jacobowitz
CodeSourcery, LLC