From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:50805) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1fQDNQ-0006kl-3o for qemu-devel@nongnu.org; Tue, 05 Jun 2018 10:56:01 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1fQDNN-00054q-0l for qemu-devel@nongnu.org; Tue, 05 Jun 2018 10:56:00 -0400 Received: from mail-wr0-x22a.google.com ([2a00:1450:400c:c0c::22a]:34337) by eggs.gnu.org with esmtps (TLS1.0:RSA_AES_128_CBC_SHA1:16) (Exim 4.71) (envelope-from ) id 1fQDNM-000529-FD for qemu-devel@nongnu.org; Tue, 05 Jun 2018 10:55:56 -0400 Received: by mail-wr0-x22a.google.com with SMTP id a12-v6so2789162wro.1 for ; Tue, 05 Jun 2018 07:55:56 -0700 (PDT) References: <20180528071332.9424.27343.stgit@pasha-VirtualBox> <87in6yd146.fsf@linaro.org> <001201d3fc8d$e8a06390$b9e12ab0$@ru> <87fu21d3gh.fsf@linaro.org> <001c01d3fcae$f0443fb0$d0ccbf10$@ru> From: Alex =?utf-8?Q?Benn=C3=A9e?= In-reply-to: <001c01d3fcae$f0443fb0$d0ccbf10$@ru> Date: Tue, 05 Jun 2018 15:55:53 +0100 Message-ID: <87bmcpcmfa.fsf@linaro.org> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [PATCH v4 00/19] reverse debugging List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Pavel Dovgalyuk 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 Pavel Dovgalyuk writes: >> From: Alex Benn=C3=A9e [mailto:alex.bennee@linaro.org] >> Pavel Dovgalyuk writes: >> >> >> 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,g= ic- >> >> version=3D3,virtualization=3Don -cpu cortex-a53 --serial mon:stdio -d= isplay 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,g= ic- >> >> version=3D3,virtualization=3Don -cpu cortex-a53 --serial mon:stdio -d= isplay 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. >> >> I suspect a recent patch has broken locking again: >> >> Starting program: /home/alex/lsrc/qemu/qemu.git/aarch64-softmmu/qemu-sys= tem-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: (repl= ay_file && >> replay_mutex_locked()) > > Have you recorded it with the attached disk before replaying? Yes. I assume the -drive doesn't actually have to be visible to the guest, it's just the mechanism rr needs for saving snapshots? > Are you using the latest version? > If the both answers are 'yes', then can you share the kernel? My i386 > runs work normally. I'll have a go with x86 first as aarch64 hasn't been proven yet. > >> 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/q= emu-test/blob/master/arm/rr > > Pavel Dovgalyuk -- Alex Benn=C3=A9e