From mboxrd@z Thu Jan 1 00:00:00 1970 From: Davide Libenzi Subject: Re: [take23 0/5] kevent: Generic event handling mechanism. Date: Wed, 8 Nov 2006 15:56:28 -0800 (PST) Message-ID: References: <1154985aa0591036@2ka.mipt.ru> <20061107141718.f7414b31.akpm@osdl.org> <20061108082147.GA2447@2ka.mipt.ru> <200611081551.14671.dada1@cosmosbay.com> <20061108140307.da7d815f.akpm@osdl.org> <45526339.3040506@cosmosbay.com> Mime-Version: 1.0 Content-Type: MULTIPART/MIXED; BOUNDARY="1795850513-710282051-1163029622=:23154" Cc: Andrew Morton , Evgeniy Polyakov , David Miller , Ulrich Drepper , netdev , Zach Brown , Christoph Hellwig , Chase Venters , Johann Borck , Linux Kernel Mailing List , Jeff Garzik Return-path: Received: from x35.xmailserver.org ([69.30.125.51]:47077 "EHLO x35.xmailserver.org") by vger.kernel.org with ESMTP id S1161759AbWKHX4b (ORCPT ); Wed, 8 Nov 2006 18:56:31 -0500 Received: from alien.or.mcafeemobile.com by x35.xmailserver.org with [XMail 1.23 ESMTP Server] id for from ; Wed, 8 Nov 2006 15:56:29 -0800 To: Eric Dumazet In-Reply-To: <45526339.3040506@cosmosbay.com> Sender: netdev-owner@vger.kernel.org List-Id: netdev.vger.kernel.org This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. --1795850513-710282051-1163029622=:23154 Content-Type: TEXT/PLAIN; CHARSET=X-UNKNOWN Content-Transfer-Encoding: QUOTED-PRINTABLE On Thu, 9 Nov 2006, Eric Dumazet wrote: > Davide Libenzi a =E9crit : > >=20 > > I don't care about both ways, but sys_poll() does the same thing epoll = does > > right now, so I would not change epoll behaviour. > >=20 >=20 > Sure poll() cannot return a partial count, since its return value is : >=20 > On success, a positive number is returned, where the number returned is > the number of structures which have non-zero revents fields (in = other > words, those descriptors with events or errors reported). >=20 > poll() is non destructive (it doesnt change any state into kernel). Retur= ning > EFAULT in case of an error in the very last bit of user area is mandatory= =2E >=20 > On the contrary : >=20 > epoll_wait() does return a count of transfered events, and update some st= ate > in kernel (it consume Edge Trigered events : They can be lost forever if = not > reported to user) >=20 > So epoll_wait() is much more like read(), that also updates file state in > kernel (current file position) Lost forever means? If there are more processes watching some fd=20 (external events), they all get their own copy of the events in their own= =20 private epoll fd. It's not that we "steal" things out of the kernel, is=20 not a 1:1 producer/consumer thing (one producer, 1 queue). It's one=20 producer, broadcast to all listeners (consumers) thing. The only case=20 where it'd matter is in the case of multiple threads sharing the same=20 epoll fd. In general, I'd be more for having the userspace get his own SEGFAULT=20 instead of letting it go with broken parameters. If I'm coding userspace,= =20 and I'm doing something wrong, I like the kernel to let me know, instead=20 of trying to fix things for me. Also, epoll can easily be fixed (add a param to ep_reinject_items() to=20 re-inject items in case of error/EFAULT) to leave events in the ready-list= =20 and let the EFAULT emerge.=20 Anyone else has opinions about this? PS: Next time it'd be great if you Cc: me when posting epoll patches, so=20 you avoid Andrew the job of doing it. - Davide --1795850513-710282051-1163029622=:23154--