From: Keir Fraser <keir.xen@gmail.com>
To: Jan Beulich <JBeulich@novell.com>,
"xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
Cc: Jinsong Liu <jinsong.liu@intel.com>
Subject: Re: [PATCH] x86/mce: CPU notifiers must not be registered a second time during resume
Date: Fri, 18 Mar 2011 15:43:52 +0000 [thread overview]
Message-ID: <C9A92E38.150A6%keir.xen@gmail.com> (raw)
In-Reply-To: <4D8383340200007800037412@vpn.id2.novell.com>
On 18/03/2011 15:07, "Jan Beulich" <JBeulich@novell.com> wrote:
> While c/s 22964:f71212f712fd and 23051:93c864c16ab1 fixed issues with
> CPU onlining, they introduced a problem with resume: mcheck_init() is
> also being called on that path, and hence checking whether it's running
> on CPU 0, which is generally not a really good thing, is particularly
> inappropriate here.
Just have a 'static bool_t early_init_done' or similar in
intel_mcheck_init().
if ( !early_init_done ) {
BUG_ON(smp_processor_id() != 0);
...
early_init_done = 1;
}
It's clearer anyway -- we're simply protecting one-time-only early-boot-time
initialisation stuff.
-- Keir
> Signed-off-by: Jan Beulich <jbeulich@novell.com>
>
> --- a/xen/arch/x86/acpi/power.c
> +++ b/xen/arch/x86/acpi/power.c
> @@ -187,7 +187,7 @@ static int enter_state(u32 state)
>
> device_power_up();
>
> - mcheck_init(&boot_cpu_data);
> + mcheck_init(&boot_cpu_data, 0);
> write_cr4(cr4);
>
> printk(XENLOG_INFO "Finishing wakeup from ACPI S%d state.\n", state);
> --- a/xen/arch/x86/cpu/common.c
> +++ b/xen/arch/x86/cpu/common.c
> @@ -431,19 +431,13 @@ void __cpuinit identify_cpu(struct cpuin
> /* AND the already accumulated flags with these */
> for ( i = 0 ; i < NCAPINTS ; i++ )
> boot_cpu_data.x86_capability[i] &= c->x86_capability[i];
> - }
> -
> - /* Init Machine Check Exception if available. */
> - mcheck_init(c);
>
> -#if 0
> - if (c == &boot_cpu_data)
> - sysenter_setup();
> - enable_sep_cpu();
> -#endif
> + mcheck_init(c, 0);
> + } else {
> + mcheck_init(c, 1);
>
> - if (c == &boot_cpu_data)
> mtrr_bp_init();
> + }
> }
>
> /* cpuid returns the value latched in the HW at reset, not the APIC ID
> --- a/xen/arch/x86/cpu/mcheck/mce.c
> +++ b/xen/arch/x86/cpu/mcheck/mce.c
> @@ -793,7 +793,7 @@ static struct notifier_block cpu_nfb = {
> };
>
> /* This has to be run for each processor */
> -void mcheck_init(struct cpuinfo_x86 *c)
> +void mcheck_init(struct cpuinfo_x86 *c, bool_t bsp)
> {
> enum mcheck_type inited = mcheck_none;
>
> @@ -822,7 +822,7 @@ void mcheck_init(struct cpuinfo_x86 *c)
> switch (c->x86) {
> case 6:
> case 15:
> - inited = intel_mcheck_init(c);
> + inited = intel_mcheck_init(c, bsp);
> break;
> }
> break;
> @@ -844,7 +844,7 @@ void mcheck_init(struct cpuinfo_x86 *c)
> /* Turn on MCE now */
> set_in_cr4(X86_CR4_MCE);
>
> - if ( smp_processor_id() == 0 )
> + if ( bsp )
> {
> /* Early MCE initialisation for BSP. */
> if ( cpu_poll_bankmask_alloc(0) )
> --- a/xen/arch/x86/cpu/mcheck/mce.h
> +++ b/xen/arch/x86/cpu/mcheck/mce.h
> @@ -42,7 +42,7 @@ enum mcheck_type amd_k7_mcheck_init(stru
> enum mcheck_type amd_k8_mcheck_init(struct cpuinfo_x86 *c);
> enum mcheck_type amd_f10_mcheck_init(struct cpuinfo_x86 *c);
>
> -enum mcheck_type intel_mcheck_init(struct cpuinfo_x86 *c);
> +enum mcheck_type intel_mcheck_init(struct cpuinfo_x86 *c, bool_t bsp);
>
> void intel_mcheck_timer(struct cpuinfo_x86 *c);
> void mce_intel_feature_init(struct cpuinfo_x86 *c);
> --- a/xen/arch/x86/cpu/mcheck/mce_intel.c
> +++ b/xen/arch/x86/cpu/mcheck/mce_intel.c
> @@ -1297,9 +1297,9 @@ static struct notifier_block cpu_nfb = {
> };
>
> /* p4/p6 family have similar MCA initialization process */
> -enum mcheck_type intel_mcheck_init(struct cpuinfo_x86 *c)
> +enum mcheck_type intel_mcheck_init(struct cpuinfo_x86 *c, bool_t bsp)
> {
> - if ( smp_processor_id() == 0 )
> + if ( bsp )
> {
> /* Early MCE initialisation for BSP. */
> if ( cpu_mcabank_alloc(0) )
> --- a/xen/include/asm-x86/processor.h
> +++ b/xen/include/asm-x86/processor.h
> @@ -555,7 +555,7 @@ void compat_show_guest_stack(struct vcpu
> extern void mtrr_ap_init(void);
> extern void mtrr_bp_init(void);
>
> -void mcheck_init(struct cpuinfo_x86 *c);
> +void mcheck_init(struct cpuinfo_x86 *c, bool_t bsp);
>
> #define DECLARE_TRAP_HANDLER(_name) \
> asmlinkage void _name(void); \
>
>
>
> _______________________________________________
> Xen-devel mailing list
> Xen-devel@lists.xensource.com
> http://lists.xensource.com/xen-devel
next prev parent reply other threads:[~2011-03-18 15:43 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-03-18 15:07 [PATCH] x86/mce: CPU notifiers must not be registered a second time during resume Jan Beulich
2011-03-18 15:43 ` Keir Fraser [this message]
2011-03-18 16:35 ` Jan Beulich
2011-03-18 17:11 ` Keir Fraser
2011-03-19 15:53 ` Liu, Jinsong
2011-03-19 22:20 ` Keir Fraser
2011-03-21 8:22 ` Jan Beulich
2011-03-21 13:18 ` Liu, Jinsong
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=C9A92E38.150A6%keir.xen@gmail.com \
--to=keir.xen@gmail.com \
--cc=JBeulich@novell.com \
--cc=jinsong.liu@intel.com \
--cc=xen-devel@lists.xensource.com \
/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).