From mboxrd@z Thu Jan 1 00:00:00 1970 From: Christoph Hellwig Subject: Re: io_uring vs in_compat_syscall() Date: Tue, 21 Jul 2020 09:07:09 +0200 Message-ID: <20200721070709.GB11432@lst.de> References: <8987E376-6B13-4798-BDBA-616A457447CF@amacapital.net> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit Return-path: Received: from verein.lst.de ([213.95.11.211]:50851 "EHLO verein.lst.de" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726474AbgGUHHM (ORCPT ); Tue, 21 Jul 2020 03:07:12 -0400 Content-Disposition: inline In-Reply-To: <8987E376-6B13-4798-BDBA-616A457447CF@amacapital.net> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Andy Lutomirski Cc: Jens Axboe , Christoph Hellwig , linux-arch@vger.kernel.org, linux-api@vger.kernel.org, linux-kernel@vger.kernel.org, io-uring@vger.kernel.org On Mon, Jul 20, 2020 at 10:28:55AM -0700, Andy Lutomirski wrote: > > Sure, I'd consider that implementation detail for the actual patch(es) > > for this issue. > > There’s a corner case, though: doesn’t io_uring submission frequently do the work synchronously in the context of the calling thread? Yes. > If so, can a thread do a 64-bit submit with 32-bit work or vice versa? In theory you could share an fd created in a 32-bit thread to a 64-bit thread or vice versa, but I think at that point you absolutely are in "you get to keep the pieces" land. > Sometimes I think that in_compat_syscall() should have a mode in which calling it warns (e.g. not actually in a syscall when doing things in io_uring). And the relevant operations should be properly wired up to avoid global state like this. What do you mean with "properly wired up". Do you really want to spread ->compat_foo methods everywhere, including read and write? I found in_compat_syscall() a lot small and easier to maintain than all the separate compat cruft.