This is the mail archive of the
gdb@sourceware.cygnus.com
mailing list for the GDB project.
Re: remote protocol: error codes
- To: Quality Quorum <qqi at world dot std dot com>
- Subject: Re: remote protocol: error codes
- From: jtc at redback dot com (J.T. Conklin)
- Date: 18 Oct 1999 10:58:53 -0700
- Cc: gdb at sourceware dot cygnus dot com
- References: <Pine.SGI.3.95.991017223617.4444A-100000@world.std.com>
- Reply-To: jtc at redback dot com
>>>>> "qqi" == Quality Quorum <qqi@world.std.com> writes:
qqi> Are there any consistent description of error codes for remote
qqi> protocol ?
No. One of the longstanding deficiencies of the remote protocol.
It's probably much too late to change now.
qqi> It seems to me that i386-stub.c has something like
qqi> E01 - malformed read request
qqi> E02 - malformed write request
qqi> E03 - access error
True, many stubs derived from the sample stubs distributed with GDB
return the same values. But there are stubs that exist that aren't
related to the samples, so even those error values are unreliable.
GDB itself does not check beyond the 'E' (in many cases, any response
at all is considered an error).
qqi> at the same time sparcl-stub.c has something like P01 ??????
I think this is simply a bug, errors should always be reported in the
E?? form. Fortunately, in this case GDB is only checking whether the
target returns anything.
--jtc
--
J.T. Conklin
RedBack Networks