From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Wed, 25 Mar 2009 00:47:53 +0100 From: Renzo Davoli Subject: Re: [PATCH 0/2] ptrace_vm: ptrace for syscall emulation virtual machines Message-ID: <20090324234753.GH22695@cs.unibo.it> References: <20090204080236.GA17452@cs.unibo.it> <20090310214436.GC5213@cs.unibo.it> <20090316074520.GC3360@hack> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20090316074520.GC3360@hack> Sender: linux-kernel-owner@vger.kernel.org To: Am??rico Wang Cc: linux-kernel@vger.kernel.org, Jeff Dike , user-mode-linux-devel@lists.sourceforge.net, mtk.manpages@gmail.com, Roland McGrath List-ID: > I just finished reading all of them. Good work! :) > Thanks. Comments below. I am thanking you, not viceversa! > > Why not introduce a new request for PTRACE_VM but use *tags* in 'addr'? > We are taking risks of breaking the existing code. :) Yes, there is a minimal risk to break some code. This is a con. On the other side there are two main pros for this proposal: 1- the code is now extremely simple 2- if we define a different tag for syscall (e.g. PTRACE_VM), we need also different tags for PTRACE_VM_SINGLESTEP, PTRACE_VM_SINGLEBLOCK and maybe others in the future. Using the addr field we don't need this multiplication of tags (and we could soon delete PTRACE_SYSEMU and PTRACE_SYSEMU_SINGLESTEP). > My question is if there are any other usages of SYSEMU beyond UML? At this time I have not find any other usage. > > Well, since 'addr' is said to be unused, it can have any value beyond > 0 or 1, we are still having the risks of breaking existing code. :( True. > > >This change would eventually simplify both the kernel code > >(reducing tags and exceptions) and even user-mode linux and umview. > > I forgot to change this sentence > >The skip-exit feature can be implemented in a arch-independent > >manner, while for skip_call some simple changes are needed > >(the entry assembly code should process the return value of the syscall > >tracing function call, like in arch/x86/kernel/Entry_32.S). Both features are now arch independent after McGrath work on tracing hooks. > > BTW, please always update the corresponding man pages when you change > any syscall interface. So let's Cc Michael Kerrisk. You are right, I'll patch the man page as soon as possible. renzo