From mboxrd@z Thu Jan 1 00:00:00 1970 From: Evgeniy Polyakov Subject: Re: [take33 10/10] kevent: Kevent based AIO (aio_sendfile()/aio_sendfile_path()). Date: Fri, 19 Jan 2007 12:23:43 +0300 Message-ID: <20070119092343.GA14605@2ka.mipt.ru> References: <11690154353959@2ka.mipt.ru> <11690154352501@2ka.mipt.ru> <20070117135142.GA24866@in.ibm.com> <20070117143950.GA19434@2ka.mipt.ru> <20070119062700.GA14705@in.ibm.com> Mime-Version: 1.0 Content-Type: text/plain; charset=koi8-r Cc: David Miller , Ulrich Drepper , Andrew Morton , netdev , Zach Brown , Christoph Hellwig , Chase Venters , Johann Borck , linux-kernel@vger.kernel.org, Jeff Garzik , Jamal Hadi Salim , Ingo Molnar , linux-fsdevel@vger.kernel.org Return-path: To: Suparna Bhattacharya Content-Disposition: inline In-Reply-To: <20070119062700.GA14705@in.ibm.com> Sender: linux-fsdevel-owner@vger.kernel.org List-Id: netdev.vger.kernel.org On Fri, Jan 19, 2007 at 11:57:00AM +0530, Suparna Bhattacharya (suparna@in.ibm.com) wrote: > > > Since you are implementing new APIs here, have you considered doing an > > > aio_sendfilev to be able to send a header with the data ? > > > > It is doable, but why people do not like corking? > > With Linux less than microsecond syscall overhead it is better and more > > flexible solution, doesn't it? > > That is what I used to think as well. However ... > > The problem as I understand it now is not about bunching data together, but > of ensuring some sort of atomicity between the header and the data, when > there can be multiple outstanding aio requests on the same socket - i.e > ensuring strict ordering without other data coming in between, when data > to be sent is not already in cache, and in the meantime another sendfile > or aio write requests comes in for the same socket. Without having to lock > the socket when reading data from disk. No, socket locking is not solution at all here. But the same applies to header - it will be copied into socket queue, then socket will be unlocked and populated VFS data will be put into that queue too, but there is a window between socket unlock after header copy and file data copy. If we will hold socket lock after header is copied, it is possible to lock it for too long - bad sectors on disk, and reading might take forever. > There are alternate ways to address this, aio_sendfilev is one of the options > I have heard people requesting. I bet those people worked with different Unix systems, which have much slower syscalls, so they combine several operations into one call. Only from this perspective I see any benefit from having header in the syscall related to file transfer. Since I already "optimized" open() syscall into file sending, things can not became worse if I will put there header pointer too. I will schedule new kevent release with this change somewhere after current work on M-on-N threading model. > Regards > Suparna -- Evgeniy Polyakov