From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:44800) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cHI5w-0004t1-Nr for qemu-devel@nongnu.org; Wed, 14 Dec 2016 17:32:17 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cHI5r-0005uG-Ok for qemu-devel@nongnu.org; Wed, 14 Dec 2016 17:32:16 -0500 Received: from mx5-phx2.redhat.com ([209.132.183.37]:51765) by eggs.gnu.org with esmtps (TLS1.0:DHE_RSA_AES_256_CBC_SHA1:32) (Exim 4.71) (envelope-from ) id 1cHI5r-0005tw-GV for qemu-devel@nongnu.org; Wed, 14 Dec 2016 17:32:11 -0500 Date: Wed, 14 Dec 2016 17:32:09 -0500 (EST) From: Paolo Bonzini Message-ID: <458039301.4070513.1481754729324.JavaMail.zimbra@redhat.com> In-Reply-To: <20161214222601.GP3808@thinpad.lan.raisama.net> References: <1481732391-882-1-git-send-email-vlad.lungu@windriver.com> <20161214170058.GJ3808@thinpad.lan.raisama.net> <20161214175144.GN3808@thinpad.lan.raisama.net> <1509132582.4065072.1481751487408.JavaMail.zimbra@redhat.com> <20161214222601.GP3808@thinpad.lan.raisama.net> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH] multiboot: copy the cmdline verbatim List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eduardo Habkost Cc: Vlad Lungu , qemu-devel@nongnu.org, rth@twiddle.net > > > In other words: this fixes the mb_add_cmdline(kcmdline) case, and > > > doesn't break comma escaping on the initrd case (because it was > > > already broken). I don't see a problem with this patch. > > > > ... there is one case of comma escaping that wasn't broken: > > > > $ qemu-system-x86_64 -kernel foo -initrd '/tmp/one > > arg,,with,,commas,/tmp/another arg,,with,,commas' > > > > Oh, I didn't notice the whitespace-based split for initrd > arguments. > > This is messier than I thought. Maybe the simplest solution is to > inline mb_add_cmdline() at both callers, and change the kcmdline > one to use memcpy(). Yes, I agree. Paolo