From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mailman by lists.gnu.org with tmda-scanned (Exim 4.43) id 1M8ufk-0003cE-SU for qemu-devel@nongnu.org; Tue, 26 May 2009 07:18:20 -0400 Received: from exim by lists.gnu.org with spam-scanned (Exim 4.43) id 1M8ufi-0003c1-M4 for qemu-devel@nongnu.org; Tue, 26 May 2009 07:18:19 -0400 Received: from [199.232.76.173] (port=43912 helo=monty-python.gnu.org) by lists.gnu.org with esmtp (Exim 4.43) id 1M8ufi-0003by-Id for qemu-devel@nongnu.org; Tue, 26 May 2009 07:18:18 -0400 Received: from flounder.pepperfish.net ([87.237.62.181]:48122) by monty-python.gnu.org with esmtps (TLS-1.0:RSA_AES_256_CBC_SHA1:32) (Exim 4.60) (envelope-from ) id 1M8ufi-0004lI-6o for qemu-devel@nongnu.org; Tue, 26 May 2009 07:18:18 -0400 Date: Tue, 26 May 2009 12:18:13 +0100 From: Vincent Sanders Subject: Re: [Qemu-devel] [PATCH 1/16] ARM Add ARM 920T identifiers Message-ID: <20090526111813.GE4784@derik> References: <20090423171503.GC4629@derik> <761ea48b0905260242v4290a835rd8b958c7544e667c@mail.gmail.com> <20090526095652.GA32352@shareable.org> <200905261116.19478.paul@codesourcery.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <200905261116.19478.paul@codesourcery.com> List-Id: qemu-devel.nongnu.org List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: qemu-devel@nongnu.org Cc: Paul Brook On Tue, May 26, 2009 at 11:16:18AM +0100, Paul Brook wrote: > > > It could be a major problem if you start playing with ARM/Thumb > > > interworking. > > > > I'm curious. How would code built for ARMv4T interworking fail on an > > ARMv5T emulator? > > Usually the reverse is true - code works on v5, but fails on v4t. > However the following will work on armv4t and fail on armv5: > > test: > adr r0, 1f > orr r0, r0, #1 > str r0, [sp, #-4]! > ldr pc, [sp], #4 > 1: > bx lr > > Admittedly this is a fairly contrived example, but the circumstances aren't > completely implausible. Imagine someone using the low bit of a pointer as an > additional flag, and relying on the hardware ignoring that bit. ok, aside from that (very very unlikey) case are you going to accept the new patch with the comment or do I need to alter translate.c to perform V4 operations and add the v5 feature flag. I am happy to do that if there is any chance of it being accepted. However I am now at the point I feel I am just wasting my time and you have no intention whatsoever of accepting anything we submit. If you have no desire whatsoever to accept our work, thats fine, and I shall desist from attempting to contribute any futher. Please communicate your intentions clearly so there is no futher ambiguity, thankyou. -- Regards Vincent