Re: remote packet size for memory read

On Fri, 2013-07-05 at 15:25 +0100, Pedro Alves wrote:
> On 07/02/2013 09:44 AM, Christian Groessler wrote:
> 
> I haven't analyzed the issue thoroughly, but it's quite plausible
> you're right.
Problem of GDB asking to read packetsize/2 has also be detected  with
the Valgrind gdbserver, which expects that
   nr_bytes_requested_in_m_packet*2 + 4 <= packetsize
but GDB only guarantees
    nr_bytes_requested_in_m_packet*2 <= packetsize


> If you don't plan on contributing fixes, could you file (separate) bug
> reports for these, so they don't get lost?
Have these been filed ? Not seen ...
> 
> Thanks,