From freedesktop-dbus Wed Nov 05 15:41:31 2008 From: "Jerome Philbert" Date: Wed, 05 Nov 2008 15:41:31 +0000 To: freedesktop-dbus Subject: Re: About performance of D-Bus Message-Id: <9c0d19af0811050741y4fd7377ci43daed0f3d0319a8 () mail ! gmail ! com> X-MARC-Message: https://marc.info/?l=freedesktop-dbus&m=122589969825150 MIME-Version: 1 Content-Type: multipart/mixed; boundary="--===============0076201741==" --===============0076201741== Content-Type: multipart/alternative; boundary="----=_Part_7444_5236675.1225899691838" ------=_Part_7444_5236675.1225899691838 Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline In your previous email, you also suggest : -did not try to "multiplex" code modules with "path registration" or "handler registration" What do you mean exactly ? Jerome 2008/11/5 Havoc Pennington > Hi, > > On Wed, Nov 5, 2008 at 8:41 AM, Jerome Philbert > wrote: > > I think I have also tried your suggestions number 3 and 4. > > Is not it equivalent to: > > - remove paranoid validation ? (by using > > DBUS_VALIDATION_MODE_WE_TRUST_THIS_DATA_ABSOLUTELY in the code) > > - build D-Bus with disable-checks option ? > > In this way, I have measured that D-Bus was 7% faster. > > > > I think if you wrote a library from scratch, you could disable even > more validation, basically - write code that would just crash and burn > horribly on invalid data, that never checked anything at all, but was > pretty fast. > > If you profile just plain raw sockets with no protocol parsing, > exchanging messages in a blocking way, that will tell you what the > absolute theoretical fastest library you could write would be like. > > Havoc > ------=_Part_7444_5236675.1225899691838 Content-Type: text/html; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Content-Disposition: inline In your previous email, you also suggest :
-did not try to "multiplex" code modules with "path registration" or
"handler registration"
What do you mean exactly ?

Jerome



2008/11/5 Havoc Pennington <hp@pobox.com>
Hi,

On Wed, Nov 5, 2008 at 8:41 AM, Jerome Philbert
<jerome.philbert@gmail.com> wrote:
> I think I have also tried your suggestions number 3 and 4.
> Is not it equivalent to:
> - remove paranoid validation ? (by using
> DBUS_VALIDATION_MODE_WE_TRUST_THIS_DATA_ABSOLUTELY in the code)
> - build D-Bus with disable-checks option ?
> In this way, I have measured that D-Bus was 7% faster.
>

I think if you wrote a library from scratch, you could disable even
more validation, basically - write code that would just crash and burn
horribly on invalid data, that never checked anything at all, but was
pretty fast.

If you profile just plain raw sockets with no protocol parsing,
exchanging messages in a blocking way, that will tell you what the
absolute theoretical fastest library you could write would be like.

Havoc

------=_Part_7444_5236675.1225899691838-- --===============0076201741== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ dbus mailing list dbus@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/dbus --===============0076201741==--