From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:48570) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WLugo-0008Oq-6O for qemu-devel@nongnu.org; Fri, 07 Mar 2014 08:19:56 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WLugf-0005IE-JN for qemu-devel@nongnu.org; Fri, 07 Mar 2014 08:19:50 -0500 Message-ID: <5319C768.2000903@weilnetz.de> Date: Fri, 07 Mar 2014 14:19:36 +0100 From: Stefan Weil MIME-Version: 1.0 References: <1394187082-31986-1-git-send-email-sw@weilnetz.de> <871tyecixy.fsf@linaro.org> <5319BE2C.7040801@weilnetz.de> <5319C21A.6010705@suse.de> In-Reply-To: <5319C21A.6010705@suse.de> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH] tests: Fix 'make test' for i686 hosts (buildregression)) List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?UTF-8?B?QW5kcmVhcyBGw6RyYmVy?= Cc: Peter Maydell , qemu-trivial@nongnu.org, qemu-devel@nongnu.org, qemu-stable@nongnu.org, =?UTF-8?B?QWxleCBCZW5uw6ll?= , Richard Henderson Am 07.03.2014 13:56, schrieb Andreas F=C3=A4rber: > Am 07.03.2014 13:40, schrieb Stefan Weil: >> Am 07.03.2014 13:06, schrieb Alex Benn=C3=A9e: >>> >>> Stefan Weil writes: >>> >>>> 'make test' is broken at least since commit >>>> baacf04799ace72a9c735dd9306a1ceaf305e7cf. Several source files were = moved >>>> to util/, and some of them there split, so add the missing prefix an= d new >>>> files to fix the compiler and linker errors. > [...] >> In the meantime I tried older versions. Even version 1.1 already shows >> unwanted differences between native and QEMU emulated FPU operations. >> >> The test binary is now available from this URL: >> http://qemu.weilnetz.de/test/user/i386/test-i386 >> >> Run it native and with qemu-i386 and compare the output of both runs >> to see the problems. >=20 > So... is QEMU or the test mistaken? :) >=20 > Regards, > Andreas Hi Andreas, test-i386 does some calculations and prints the results (see source code tests/tcg/test-i386.c). If the user mode emulation of QEMU works, it should not matter whether that executable runs native or emulated and both outputs be identical. They aren't - that's why I think we have a TCG problem to solve. Regards Stefan