qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Greg Kurz <gkurz@linux.vnet.ibm.com>
To: Alexey Kardashevskiy <aik@ozlabs.ru>
Cc: qemu-ppc@nongnu.org, qemu-devel@nongnu.org,
	"Andreas Färber" <afaerber@suse.de>
Subject: Re: [Qemu-devel] [Qemu-ppc] [PATCH] target-ppc: reset SPRs on CPU reset
Date: Wed, 19 Mar 2014 10:16:44 +0100	[thread overview]
Message-ID: <20140319101644.56ad5c36@bahia.local> (raw)
In-Reply-To: <1395199682-28772-1-git-send-email-aik@ozlabs.ru>

On Wed, 19 Mar 2014 14:28:02 +1100
Alexey Kardashevskiy <aik@ozlabs.ru> wrote:
> This resets SPR values to defaults on CPU reset. This should help
> with little-endian guests reboot issues.
> 
> Signed-off-by: Alexey Kardashevskiy <aik@ozlabs.ru>
> ---

Yeah ! Since we only set the endianness (LPCR_ILE) when the guest
kernel calls H_SET_MODE_RESOURCE_LE, we indeed have a window where
the guest keeps the endianness from the previous kernel. In case
the guest was running a LE kernel, when we reboot we end up running
SLOF in a LE environment... but SLOF assumes BE and breaks.

Reviewed-by: Greg Kurz <gkurz@linux.vnet.ibm.com>

Andreas,

This is really a bug fix to support LE->BE transition when we
reboot the guest. But first, we should support LE... thanks to:

https://lists.nongnu.org/archive/html/qemu-devel/2014-03/msg01316.html

(especially patch 2/3)

FWIW, there have been a lot of discussions about H_SET_MODE last autumn.
It was decided to kick the hcall out of KVM (now upstream in Linus's
tree with commit d682916a381)... We are halfway through :-\

Could you please consider applying this serie as well ?

Thanks.

--
Greg

>  target-ppc/cpu.h            |  1 +
>  target-ppc/translate_init.c | 12 +++++++++++-
>  2 files changed, 12 insertions(+), 1 deletion(-)
> 
> diff --git a/target-ppc/cpu.h b/target-ppc/cpu.h
> index 91b7ae5..8c181e7 100644
> --- a/target-ppc/cpu.h
> +++ b/target-ppc/cpu.h
> @@ -334,6 +334,7 @@ struct ppc_spr_t {
>      void (*hea_write)(void *opaque, int spr_num, int gpr_num);
>  #endif
>      const char *name;
> +    target_ulong default_value;
>  #ifdef CONFIG_KVM
>      /* We (ab)use the fact that all the SPRs will have ids for the
>       * ONE_REG interface will have KVM_REG_PPC to use 0 as meaning,
> diff --git a/target-ppc/translate_init.c b/target-ppc/translate_init.c
> index 6084f40..c63f4a1 100644
> --- a/target-ppc/translate_init.c
> +++ b/target-ppc/translate_init.c
> @@ -631,7 +631,7 @@ static inline void _spr_register(CPUPPCState *env, int num,
>  #if defined(CONFIG_KVM)
>      spr->one_reg_id = one_reg_id,
>  #endif
> -    env->spr[num] = initial_value;
> +    env->spr[num] = spr->default_value = initial_value;
>  }
> 
>  /* Generic PowerPC SPRs */
> @@ -8381,6 +8381,7 @@ static void ppc_cpu_reset(CPUState *s)
>      PowerPCCPUClass *pcc = POWERPC_CPU_GET_CLASS(cpu);
>      CPUPPCState *env = &cpu->env;
>      target_ulong msr;
> +    int i;
> 
>      pcc->parent_reset(s);
> 
> @@ -8434,6 +8435,15 @@ static void ppc_cpu_reset(CPUState *s)
>      env->dtl_size = 0;
>  #endif /* TARGET_PPC64 */
> 
> +    for (i = 0; i < sizeof(env->spr_cb)/sizeof(env->spr_cb[0]); i++) {
> +        ppc_spr_t *spr = &env->spr_cb[i];
> +
> +        if (!spr->name) {
> +            continue;
> +        }
> +        env->spr[i] = spr->default_value;
> +    }
> +
>      /* Flush all TLBs */
>      tlb_flush(s, 1);
>  }

  reply	other threads:[~2014-03-19  9:23 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-19  3:28 [Qemu-devel] [PATCH] target-ppc: reset SPRs on CPU reset Alexey Kardashevskiy
2014-03-19  9:16 ` Greg Kurz [this message]
2014-03-19  9:20 ` Andreas Färber
2014-03-19 13:03   ` [Qemu-devel] [PATCH v2] " Alexey Kardashevskiy

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=20140319101644.56ad5c36@bahia.local \
    --to=gkurz@linux.vnet.ibm.com \
    --cc=afaerber@suse.de \
    --cc=aik@ozlabs.ru \
    --cc=qemu-devel@nongnu.org \
    --cc=qemu-ppc@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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).