From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-lf0-f65.google.com ([209.85.215.65]:36759 "EHLO mail-lf0-f65.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750719AbdBOUnr (ORCPT ); Wed, 15 Feb 2017 15:43:47 -0500 Date: Wed, 15 Feb 2017 23:43:44 +0300 From: Cyrill Gorcunov To: Andrew Morton Cc: LINUXFS-ML , LKML , Al Viro , Andrew Morton , Andrey Vagin , Pavel Emelyanov Subject: Re: [RFC] fs,eventpoll: Add ability to install target file by its number Message-ID: <20170215204344.GA14561@uranus> References: <20170215161453.GA13021@uranus> <20170215122904.33499c5b0af3de12d351e7a1@linux-foundation.org> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20170215122904.33499c5b0af3de12d351e7a1@linux-foundation.org> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: On Wed, Feb 15, 2017 at 12:29:04PM -0800, Andrew Morton wrote: > On Wed, 15 Feb 2017 19:14:54 +0300 Cyrill Gorcunov wrote: > > > When we checkpoint a process we look into /proc//fdinfo/ of eventpoll > > file and parse target files list from there. In most situations this is fine > > because target file is present in the /proc//fd/ list. But in case if file > > descriptor was dup'ed or transferred via unix socket and closed after, > > it might not be in the list and we can't figure out which file descriptor > > to pass into epoll_ctl call. > > > > To resolve this tie lets add EPOLL_CTL_ITF ("itf" stands for install > > target file) operation which simply takes target file descriptor number > > and installs it into a caller's file table, thus we can use kcmp() > > syscall and figure out which exactly file to be added into > > eventpoll on restore procedure. > > Can we please see the proposed manpage update. And Cc linux-api and > Michael Kerrisk. Sure! Will do (I didn't write it immediately 'cause wanted to see if the idea of new operation won't be rejected immediately, that is why it is rfc). > > #define EPOLL_CTL_ADD 1 > > #define EPOLL_CTL_DEL 2 > > #define EPOLL_CTL_MOD 3 > > +#define EPOLL_CTL_ITF 4 > > Somewhere we should tell the poor reader what "itf" means. A comment > here would suit. I thought maybe some better name come to mind... Say EPOLL_CTL_INS, except INS usually associated with "insert"?