Can we remove the eloop exception listener?
Mon Jul 2 21:24:11 PDT 2012
On 06/30/2012 03:24 AM, Jouni Malinen wrote:
> On Mon, Jun 25, 2012 at 12:23:08PM -0700, Ben Greear wrote:
>> I wanted to remove it mainly because it cannot be properly
>> implemented when using poll() because you can't poll just
>> for exceptions.
> By the way, I just noticed that the poll() version of eloop seemed to
> break WPS external registrar functionality (TCP socket used for HTTP;
> i.e., something that needs both read and write events). I haven't had a
> chance to look at closer details on what was happening, but the HTTP
> connections were timing out in my tests. Going back to select() removed
> the issue.
I don't think I've ever used that. Do you have any details on
how to reproduce the problem and/or setup WPS external registrar?
>> My plan would be to just fail the debus request to add
>> an exception handler and hope that their code can deal
>> with that properly (if indeed they use that interface
>> at all).
>> But, I don't have any way of testing this, so perhaps
>> it is not worth the effort?
> Probably not.. I don't have much time or interest to work on this and I
> cannot accept a patch to remove the exception case unless it has been
> fully tested to work with D-Bus.
I agree. Select() is good enough for most folks anyway. We don't use
DBUS either, but if we ever start and I can do some decent regression
testing, I'll work on a patch.
Ben Greear <greearb at candelatech.com>
Candela Technologies Inc http://www.candelatech.com
More information about the Hostap