From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:52788) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1c8E98-0004HL-U5 for qemu-devel@nongnu.org; Sat, 19 Nov 2016 17:30:07 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1c8E98-0004a1-0A for qemu-devel@nongnu.org; Sat, 19 Nov 2016 17:30:06 -0500 Received: from hera.aquilenet.fr ([2a01:474::1]:54063) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1c8E97-0004YR-Pn for qemu-devel@nongnu.org; Sat, 19 Nov 2016 17:30:05 -0500 Date: Sat, 19 Nov 2016 23:30:03 +0100 From: Samuel Thibault Message-ID: <20161119223003.GD10378@var.home> References: <20161118175128.17192-1-samuel.thibault@ens-lyon.org> <20161118175128.17192-2-samuel.thibault@ens-lyon.org> <6e88c2ca-d43b-27ab-6acb-fa4c958ca976@redhat.com> <87twb3dhhf.fsf@luffy.cx> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline In-Reply-To: <87twb3dhhf.fsf@luffy.cx> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PULL] tftp: fake support for netascii protocol List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Vincent Bernat Cc: Thomas Huth , qemu-devel@nongnu.org, Jan Kiszka , Stefan Hajnoczi Vincent Bernat, on Sat 19 Nov 2016 09:03:08 +0100, wrote: > =E2=9D=A6 19 novembre 2016 08:32 +0100, Thomas Huth = =C2=A0: >=20 > >> Some network equipments are requesting a file using the netascii > >> protocol and this is not configurable. Currently, qemu's tftpd only > >> supports the octet protocol. This commit makes it accept the netasci= i > >> protocol as well but do not perform the requested transformation (LF= -> > >> CR,LF) as it would be far more complex. > > > > That sounds somewhat wrong to me. QEMU now seems to support a transfe= r > > mode that is not really implemented. >=20 > On the other hand, the current implementation may not RFC compliant as > the three modes (netascii, octet, mail) are not supported (but the RFC > predates the usage of MAY/SHOULD/MUST, so it's unclear for me). >=20 > I have tried to add proper support, but this is more invasive as we > cannot just seek in the file to get each block. However, this is > something that I can do if compliance is important for QEMU. >=20 > > I think you should at least issue a qemu_log_mask(LOG_UNIMP, "...") > > call in that case. >=20 > I can do that if needed. That'd be better indeed. Otherwise people might wonder why things are not working. Warning that they have to do the LF -> CR,LF conversion by hand is important here. Samuel