All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Andreas Färber" <afaerber@suse.de>
To: Jens Freimann <jfrei@linux.vnet.ibm.com>
Cc: Heinz Graalfs <graalfs@linux.vnet.ibm.com>,
	Alexander Graf <agraf@suse.de>,
	qemu-devel <qemu-devel@nongnu.org>,
	Christian Borntraeger <borntraeger@de.ibm.com>,
	Cornelia Huck <cornelia.huck@de.ibm.com>,
	Einar Lueck <elelueck@linux.vnet.ibm.com>
Subject: Re: [Qemu-devel] [PATCH 2/3] s390: Add CPU reset handler
Date: Sun, 16 Dec 2012 16:30:21 +0100	[thread overview]
Message-ID: <50CDE90D.9000603@suse.de> (raw)
In-Reply-To: <1355503606-54131-3-git-send-email-jfrei@linux.vnet.ibm.com>

Am 14.12.2012 17:46, schrieb Jens Freimann:
> Add a CPU reset handler to have all CPUs in a PoP compliant
> state.
> 
> Signed-off-by: Jens Freimann <jfrei@linux.vnet.ibm.com>

The logic looks okay now. Some comments inline.

> ---
> v2 -> v3: 
> * explain in comment which code sets cpu 0 to running during IPL
> 
> v1 -> v2:
> * move setting of control registers and psa to s390_cpu_reset
>   and call it from the new s390_machine_cpu_reset_cb()
>   This makes it more similar to how it is done on x86
> * in s390_cpu_reset() set env->halted state of cpu after
>   the memset. This is needed to keep our s390_cpu_running
>   counter in sync when s390_cpu_reset is called via the
>   qemu_devices_reset path
> * set env->halted state in s390_cpu_initfn to 1 to avoid
>   decrementing the cpu counter during first reset
> ---
>  target-s390x/cpu.c | 32 ++++++++++++++++++++++++++++++--
>  target-s390x/kvm.c |  9 ++++++++-
>  2 files changed, 38 insertions(+), 3 deletions(-)
> 
> diff --git a/target-s390x/cpu.c b/target-s390x/cpu.c
> index 619b202..75d4036 100644
> --- a/target-s390x/cpu.c
> +++ b/target-s390x/cpu.c
> @@ -4,6 +4,7 @@
>   * Copyright (c) 2009 Ulrich Hecht
>   * Copyright (c) 2011 Alexander Graf
>   * Copyright (c) 2012 SUSE LINUX Products GmbH
> + * Copyright (c) 2012 IBM Corp.
>   *
>   * This library is free software; you can redistribute it and/or
>   * modify it under the terms of the GNU Lesser General Public
> @@ -18,9 +19,13 @@
>   * You should have received a copy of the GNU Lesser General Public
>   * License along with this library; if not, see
>   * <http://www.gnu.org/licenses/lgpl-2.1.html>
> + * Contributions after 2012-12-11 are licensed under the terms of the
> + * GNU GPL, version 2 or (at your option) any later version.
> + *
>   */
>  
>  #include "cpu.h"
> +#include "hw/hw.h"
>  #include "qemu-common.h"
>  #include "qemu-timer.h"
>  
> @@ -37,12 +42,30 @@ static void s390_cpu_reset(CPUState *s)
>          log_cpu_state(env, 0);
>      }
>  
> -    scc->parent_reset(s);
> +    s390_del_running_cpu(env);
>  
> +    scc->parent_reset(s);

If this gets respun, a white line separating the parent reset from the
local reset would be nice. :)

>      memset(env, 0, offsetof(CPUS390XState, breakpoints));
> +
> +    /* architectured initial values for CR 0 and 14 */
> +    env->cregs[0] = 0xE0UL;
> +    env->cregs[14] = 0xC2000000UL;
> +    /* set to z/Architecture mode */
> +    env->psw.mask = 0x0000000180000000ULL;
> +    env->psa = 0;
> +    /* set halted to 1 to make sure we can add the cpu in
> +     * s390_ipl_cpu code, where env->halted is set back to 0
> +     * after incrementing the cpu counter */
> +    env->halted = 1;
>      /* FIXME: reset vector? */

Do the above added cregs/psw/psa reset values resolve this FIXME? Or
does that refer to something different?

>      tlb_flush(env, 1);
> -    s390_add_running_cpu(env);
> +}
> +
> +static void s390_cpu_machine_reset_cb(void *opaque)
> +{
> +    S390CPU *cpu = opaque;
> +
> +    cpu_reset(CPU(cpu));
>  }
>  
>  static void s390_cpu_initfn(Object *obj)
> @@ -66,7 +89,12 @@ static void s390_cpu_initfn(Object *obj)
>      env->cpu_num = cpu_num++;
>      env->ext_index = -1;
>  
> +    /* set env->halted state to 1 to avoid decrementing the running
> +     * cpu counter in s390_cpu_reset to a negative number at 
> +     * initial ipl */
> +    env->halted = 1;
>      cpu_reset(CPU(cpu));
> +    qemu_register_reset(s390_cpu_machine_reset_cb, cpu);

Since we register the reset handler in instance_init, we should
unregister it in a instance_finalize callback (uninitfn?). Since we do
not hot-unplug s390 CPUs yet to my knowledge, that could be done in a
follow-up.

(For x86 it it registered in the provisional realize function and not
unregistered lacking a matching unrealization mechanism today; elsewhere
reset registration is done in the machine.)

>  }
>  
>  static void s390_cpu_class_init(ObjectClass *oc, void *data)
> diff --git a/target-s390x/kvm.c b/target-s390x/kvm.c
> index 94de764..fda9f1f 100644
> --- a/target-s390x/kvm.c
> +++ b/target-s390x/kvm.c
> @@ -85,7 +85,14 @@ int kvm_arch_init_vcpu(CPUS390XState *env)
>  
>  void kvm_arch_reset_vcpu(CPUS390XState *env)
>  {

Note to Alex: In my upcoming KVM CPUState series, this argument type
changes to CPUState ...

> -    /* FIXME: add code to reset vcpu. */
> +   /* The initial reset call is needed here to reset in-kernel
> +    * vcpu data that we can't access directly from QEMU
> +    * (i.e. with older kernels which don't support sync_regs/ONE_REG).
> +    * Before this ioctl cpu_synchronize_state() is called in common kvm
> +    * code (kvm-all) */
> +    if (kvm_vcpu_ioctl(env, KVM_S390_INITIAL_RESET, NULL)) {

... but so does the argument for kvm_vcpu_ioctl(), so merging becomes a
trivial env -> cpu change.

> +        perror("Can't reset vcpu\n");
> +    }
>  }
>  
>  int kvm_arch_put_registers(CPUS390XState *env, int level)

Andreas

-- 
SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg, Germany
GF: Jeff Hawn, Jennifer Guild, Felix Imendörffer; HRB 16746 AG Nürnberg

  reply	other threads:[~2012-12-16 15:30 UTC|newest]

Thread overview: 24+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-14 16:46 [Qemu-devel] [PATCH 0/3] s390: ipl device, cpu reset handler and cpu model support Jens Freimann
2012-12-14 16:46 ` [Qemu-devel] [PATCH 1/3] s390: Move IPL code into a separate device Jens Freimann
2012-12-16 16:26   ` Andreas Färber
2012-12-16 21:15     ` Christian Borntraeger
2012-12-14 16:46 ` [Qemu-devel] [PATCH 2/3] s390: Add CPU reset handler Jens Freimann
2012-12-16 15:30   ` Andreas Färber [this message]
2012-12-17  8:49     ` Jens Freimann
2012-12-17 14:49   ` Alexander Graf
2012-12-17 15:41     ` Jens Freimann
2012-12-17 17:21     ` Andreas Färber
2012-12-17 17:27       ` Alexander Graf
2012-12-14 16:46 ` [Qemu-devel] [PATCH 3/3] S390: Enable -cpu help and QMP query-cpu-definitions Jens Freimann
2012-12-16 15:42   ` Andreas Färber
2012-12-17 14:47   ` Alexander Graf
2012-12-17 17:32     ` Andreas Färber
  -- strict thread matches above, loose matches on Subject: below --
2012-12-18 17:50 [Qemu-devel] [PATCH 0/3] s390: ipl device, cpu reset handler and cpu model support Jens Freimann
2012-12-18 17:50 ` [Qemu-devel] [PATCH 2/3] s390: Add CPU reset handler Jens Freimann
2013-01-03 12:50   ` Alexander Graf
2013-01-04 13:55     ` Jens Freimann
2013-01-03 12:55   ` Alexander Graf
2013-01-04 14:09     ` Jens Freimann
2013-01-04 14:11       ` Alexander Graf
2012-12-12 13:08 [Qemu-devel] [PATCH 0/3] s390: ipl device, cpu reset handler and cpu model support Jens Freimann
2012-12-12 13:08 ` [Qemu-devel] [PATCH 2/3] s390: Add CPU reset handler Jens Freimann
2012-12-12 13:38   ` Alexander Graf
2012-12-12 15:04     ` Jens Freimann

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=50CDE90D.9000603@suse.de \
    --to=afaerber@suse.de \
    --cc=agraf@suse.de \
    --cc=borntraeger@de.ibm.com \
    --cc=cornelia.huck@de.ibm.com \
    --cc=elelueck@linux.vnet.ibm.com \
    --cc=graalfs@linux.vnet.ibm.com \
    --cc=jfrei@linux.vnet.ibm.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.