On Fri, Dec 12, 2003 at 05:21:53PM +0000, Matthew Byng-Maddick wrote:
> On Fri, Dec 12, 2003 at 05:01:54PM +0000, Philip Hazel wrote:
>> On Fri, 12 Dec 2003, Matthew Byng-Maddick wrote:
>>> If so, should you not return an error to DATA, but not log a protocol error,
>>> and should this check (or similar) not also be implemented for "MAIL FROM",
>>> as that is valid in the middle of a RFC2920 group.
>> Probably. :-)
>> I will check this all out in due course. Thanks for the additional
>> information and diagnosis.
> Depending on whether or not I have time, I think I may be able to produce
> a patch in the near future against smtp_in.c. It will involve changing
> smtp_read_command (which is static to that file), but I'll give it a go.
I have produced a patch, available at
http://colon.colondot.net/~mbm/exim_pipeline.patch
Which I believe produces the right results for pipelining, of penalising
clients that have been told the results already, but where they have done
everything in one go, it doesn't call synprot_error(), so the error count
doesn't get increased.
Phil, if it's possible for you to regression test this patch with your
test-suite, I'd really appreciate it, but if you don't have time, don't
worry. Feel free to re-write it as necessary. :-)
David, I believe that this fixes your problem, too.
Cheers
MBM
--
Matthew Byng-Maddick <mbm@???> http://colondot.net/
