From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:37009) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cVZOV-0005MW-I1 for qemu-devel@nongnu.org; Mon, 23 Jan 2017 02:50:28 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1cVZOR-0003dX-ML for qemu-devel@nongnu.org; Mon, 23 Jan 2017 02:50:27 -0500 Received: from mail.ispras.ru ([83.149.199.45]:34596) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1cVZOR-0003dA-Fl for qemu-devel@nongnu.org; Mon, 23 Jan 2017 02:50:23 -0500 From: "Pavel Dovgalyuk" References: <000301d259dc$f9d097c0$ed71c740$@ru> <000601d25a95$12b1b9f0$38152dd0$@ru> <20161220102126.GE5602@stefanha-x1.localdomain> <002501d25ab1$af024b00$0d06e100$@ru> <000301d25b4f$20018440$60048cc0$@ru> <000801d26bd9$dca56db0$95f04910$@ru> <87o9zd3jta.fsf@linaro.org> <000e01d26caa$dfdb3150$9f9193f0$@ru> <87mveleiw8.fsf@linaro.org> In-Reply-To: <87mveleiw8.fsf@linaro.org> Date: Mon, 23 Jan 2017 10:50:22 +0300 Message-ID: <000c01d2754d$59a4cd70$0cee6850$@ru> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable Content-Language: ru Subject: Re: [Qemu-devel] qemu-2.8-rc4 is broken List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: =?utf-8?Q?'Alex_Benn=C3=A9e'?= Cc: 'Stefan Hajnoczi' , 'qemu-devel' , 'Paolo Bonzini' , 'Pavel Dovgalyuk' , 'Peter Maydell' > From: Alex Benn=C3=A9e [mailto:alex.bennee@linaro.org] > Pavel Dovgalyuk writes: >=20 > >> From: Alex Benn=C3=A9e [mailto:alex.bennee@linaro.org] > > > > Sorry, this is another problem which occurs only in icount replay = mode: > > 1. cpu_handle_exception tries to force exception when is cannot = occur due to > > running out all the planned instructions: > > } else if (replay_has_exception() > > && cpu->icount_decr.u16.low + cpu->icount_extra = =3D=3D 0) { > > /* try to cause an exception pending in the log */ > > cpu_exec_nocache(cpu, 1, tb_find(cpu, NULL, 0), true); > > *ret =3D -1; > > return true; > > > > 2. tb_find calls tb_gen_code, which cannot allocate new translation = block > > and calls tb_flush (which only queues the flushing) and = cpu_loop_exit > > 3. cpu_loop_exit returns to infinite loop of cpu_exec and the = condition > > if (cpu_handle_exception(cpu, &ret)) { > > break; > > } > > is checked again causing an infinite loop. > > > > TB cache is not flushed because we never execute that break and real = work of tb_flush > > is made outside this loop. >=20 > I think what we need is a: >=20 >=20 > if (cpu->exit_request) > break; Where this exit_request is supposed to be set? >=20 > before the cpu_handle_exception() call to ensure any queued work gets > processed first. Can you give me you current command line so I can > reproduce this and check the fix works? I solved the problem using following patch: --- a/cpu-exec.c +++ b/cpu-exec.c @@ -451,6 +451,10 @@ static inline bool cpu_handle_exception(CPUState = *cpu, int *ret) #ifndef CONFIG_USER_ONLY } else if (replay_has_exception() && cpu->icount_decr.u16.low + cpu->icount_extra =3D=3D = 0) { + /* Break the execution loop in case of running out of TB cache. + This is needed to make flushing of the TB cache, because + real flush is queued to be executed outside the cpu loop. */ + cpu->exception_index =3D EXCP_INTERRUPT; /* try to cause an exception pending in the log */ cpu_exec_nocache(cpu, 1, tb_find(cpu, NULL, 0), true); *ret =3D -1; Pavel Dovgalyuk