linux-kernel - Re: [PATCH] EPOLL_CTL_SET operatin for epoll

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Fri, 25 Sep 2009 00:22:54 +0000
From:	Carlo Marcelo Arenas Belon <carenas@...inet.com.pe>
To:	Davide Libenzi <davidel@...ilserver.org>
Cc:	Alexander Viro <viro@...iv.linux.org.uk>,
	Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] EPOLL_CTL_SET operatin for epoll_ctl when using
	EPOLLONESHOT

On Thu, Sep 24, 2009 at 04:46:20PM -0700, Davide Libenzi wrote:
> On Thu, 24 Sep 2009, Carlo Marcelo Arenas Belon wrote:
> 
> > Simplify userspace by eliminating the need to track whether a specific
> > filehandle was added to epoll already and only needs to be re-arm
> > after being internally disabled by EPOLLONESHOT.
> > 
> > EPOLL_CLT_SET can be used to add or modify a filehandle in epoll
> > using the kernel's internal tracking to apply the right equivalent
> > operation to EPOLL_CTL_ADD or EPOLL_CTL_MOD.
> 
> Applications (except poorly written ones) usually know what's in the 
> epoll set and what's not, in their bookkeeping.

agree, but since the kernel is also doing that bookkeeping using
ep_find it is possible by using this new operation to avoid doing
that in userspace, saving some cycles (and probably some contention in
multithreaded applications if the epoll set is global) in the hot path.

> I'm sorry, but I see no use of this code.

a very simple, probe of concept echoserver using this (once #defined)
is available from :

  http://sajino.sajinet.com.pe/echoserver.c

Carlo
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/