From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:60461) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fQ8CD-0008Aw-1m for qemu-devel@nongnu.org; Tue, 05 Jun 2018 05:24:07 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fQ8C9-00081E-U2 for qemu-devel@nongnu.org; Tue, 05 Jun 2018 05:24:05 -0400 Received: from mail.ispras.ru ([83.149.199.45]:48362) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fQ8C9-00080a-Dm for qemu-devel@nongnu.org; Tue, 05 Jun 2018 05:24:01 -0400 From: "Pavel Dovgalyuk" References: <20180528071332.9424.27343.stgit@pasha-VirtualBox> <87in6yd146.fsf@linaro.org> <001201d3fc8d$e8a06390$b9e12ab0$@ru> <87fu21d3gh.fsf@linaro.org> In-Reply-To: <87fu21d3gh.fsf@linaro.org> Date: Tue, 5 Jun 2018 12:24:00 +0300 Message-ID: <001c01d3fcae$f0443fb0$d0ccbf10$@ru> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Language: ru Subject: Re: [Qemu-devel] [PATCH v4 00/19] reverse debugging List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?utf-8?Q?'Alex_Benn=C3=A9e'?= Cc: 'Pavel Dovgalyuk' , qemu-devel@nongnu.org, kwolf@redhat.com, peter.maydell@linaro.org, war2jordan@live.com, crosthwaite.peter@gmail.com, boost.lists@gmail.com, quintela@redhat.com, ciro.santilli@gmail.com, jasowang@redhat.com, mst@redhat.com, zuban32s@gmail.com, armbru@redhat.com, maria.klimushenkova@ispras.ru, kraxel@redhat.com, thomas.dullien@googlemail.com, pbonzini@redhat.com, mreitz@redhat.com, dgilbert@redhat.com, rth@twiddle.net > From: Alex Benn=C3=A9e [mailto:alex.bennee@linaro.org] > Pavel Dovgalyuk writes: >=20 > >> From: Alex Benn=C3=A9e [mailto:alex.bennee@linaro.org] > >> Pavel Dovgalyuk writes: > >> > >> > Ping? > >> > >> I started having a look but I ran into this straight away. First I > >> recorded a boot of the kernel: > >> > >> ./aarch64-softmmu/qemu-system-aarch64 -machine = virt,graphics=3Don,gic- > >> version=3D3,virtualization=3Don -cpu cortex-a53 --serial mon:stdio = -display none -kernel > >> ../images/aarch64-current-linux-initrd-guest.img -icount > shift=3D7,rr=3Drecord,rrfile=3Dreplay.bin > >> > >> Then played back: > >> > >> ./aarch64-softmmu/qemu-system-aarch64 -machine = virt,graphics=3Don,gic- > >> version=3D3,virtualization=3Don -cpu cortex-a53 --serial mon:stdio = -display none -kernel > >> ../images/aarch64-current-linux-initrd-guest.img -icount > shift=3D7,rr=3Dreplay,rrfile=3Dreplay.bin - > >> s -S > > > > This looks ok, but... > > > >> And did the following on gdb: > >> > >> (gdb) i > >> 0x0000000040000004 in ?? () > >> =3D> 0x40000004: mov x1, xzr > >> 0x40000008: mov x2, xzr > >> 0x4000000c: mov x3, xzr > >> (gdb) > >> 0x0000000040000008 in ?? () > >> =3D> 0x40000008: mov x2, xzr > >> 0x4000000c: mov x3, xzr > >> 0x40000010: ldr x4, 0x40000020 > >> (gdb) > >> 0x000000004000000c in ?? () > >> =3D> 0x4000000c: mov x3, xzr > >> 0x40000010: ldr x4, 0x40000020 > >> 0x40000014: br x4 > >> (gdb) > >> 0x0000000040000010 in ?? () > >> =3D> 0x40000010: ldr x4, 0x40000020 > >> 0x40000014: br x4 > >> 0x40000018: .inst 0x44000000 ; undefined > >> (gdb) > >> 0x0000000040000014 in ?? () > >> =3D> 0x40000014: br x4 > >> 0x40000018: .inst 0x44000000 ; undefined > >> 0x4000001c: .inst 0x00000000 ; undefined > >> (gdb) p/x $x4 > >> $1 =3D 0x40080000 > >> (gdb) reverse-stepi > >> warning: Remote failure reply: E14 > >> > >> Surely this is the simple case and doesn't require any snapshots = for > >> block devices as there are none. Am I missing something? > > > > Reverse debugging requires the snapshotting. QEMU can't revert the = VM state without the > snapshots. > > You can try adding an empty qcow2 image to allow snapshotting there. >=20 > I suspect a recent patch has broken locking again: >=20 > Starting program: = /home/alex/lsrc/qemu/qemu.git/aarch64-softmmu/qemu-system-aarch64 = -machine > virt,graphics=3Don,gic-version=3D3,virtualization=3Don -cpu cortex-a53 = --serial mon:stdio -display > none -kernel ../images/aarch64-current-linux-initrd-guest.img -icount > shift=3D7,rr=3Dreplay,rrfile=3Dreplay.bin,rrsnapshot=3Ddebug -drive > file=3Drr.qcow2,if=3Dnone,snapshot,id=3Drr -s -S > [Thread debugging using libthread_db enabled] > Using host libthread_db library = "/lib/x86_64-linux-gnu/libthread_db.so.1". > [New Thread 0x7fffd8536700 (LWP 32452)] > [New Thread 0x7fffd5bb3700 (LWP 32453)] > [New Thread 0x7fffd4eab700 (LWP 32454)] > ** > ERROR:replay/replay-time.c:49:replay_read_clock: assertion failed: = (replay_file && > replay_mutex_locked()) Have you recorded it with the attached disk before replaying? Are you using the latest version? If the both answers are 'yes', then can you share the kernel? My i386 = runs work normally. > Once I have the linux-user TCG tests merged I'm planning on focusing = on > the system emulation tests and we should be able to add some > record/replay tests to defend the behaviour. That will be great. There are some Ciro's attempts on that: = https://github.com/cirosantilli/qemu-test/blob/master/arm/rr Pavel Dovgalyuk