[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <1263340563.2854.1011.camel@sbs-t61.sc.intel.com>
Date: Tue, 12 Jan 2010 15:56:02 -0800
From: Suresh Siddha <suresh.b.siddha@...el.com>
To: Yinghai Lu <yinghai@...nel.org>
Cc: Linus Torvalds <torvalds@...ux-foundation.org>,
"ananth@...ibm.com" <ananth@...ibm.com>,
Ingo Molnar <mingo@...e.hu>,
Thomas Gleixner <tglx@...utronix.de>,
"H. Peter Anvin" <hpa@...or.com>,
Andrew Morton <akpm@...ux-foundation.org>,
"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH -v2 5/5] x86: use dmi check to treat disabled cpus as
hotplug cpus.
On Tue, 2010-01-12 at 15:17 -0800, Yinghai Lu wrote:
> some systems that have disable cpus entries because same
> BIOS will support 2 sockets and 4 sockets and more at
> same time, BIOS just leave some disable entries, but
> those system do not support cpu hotplug. we don't need
> treat disabled_cpus as hotplug cpus.
> so we can make nr_cpu_ids smaller and save more space
> (pcpu data allocations), and could make some systems run
> with logical flat instead of physical flat apic mode
>
> -v2: change to black list instead
I don't think this list will work because every HT system with HT
disabled in the bios shows that logical cpu as disabled.
> Signed-off-by: Yinghai Lu <yinghai@...nel.org>
>
> ---
> arch/x86/kernel/smpboot.c | 76 ++++++++++++++++++++++++++++++++++++++++++++--
> 1 file changed, 74 insertions(+), 2 deletions(-)
>
> Index: linux-2.6/arch/x86/kernel/smpboot.c
> ===================================================================
> --- linux-2.6.orig/arch/x86/kernel/smpboot.c
> +++ linux-2.6/arch/x86/kernel/smpboot.c
> @@ -47,6 +47,7 @@
> #include <linux/bootmem.h>
> #include <linux/err.h>
> #include <linux/nmi.h>
> +#include <linux/dmi.h>
> #include <linux/tboot.h>
>
> #include <asm/acpi.h>
> @@ -1178,6 +1179,59 @@ static int __init _setup_possible_cpus(c
> }
> early_param("possible_cpus", _setup_possible_cpus);
>
> +static __initdata int treat_disabled_cpus_as_hotplug = 1;
> +static __init int hotplug_cpus_check(const struct dmi_system_id *d)
> +{
> + printk(KERN_NOTICE "%s detected: treat disabled cpus as hotplug ones\n", d->ident);
> + treat_disabled_cpus_as_hotplug = 0;
> +
> + return 0;
> +}
> +
> +static struct dmi_system_id hotplug_cpus_dmi_table[] __initdata = {
> + {
> + .callback = hotplug_cpus_check,
> + .ident = "Sun Microsystems Sun Fire X4440",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "Sun Microsystems"),
> + DMI_MATCH(DMI_PRODUCT_NAME, "Sun Fire X4440"),
> + },
> + },
> + {
> + .callback = hotplug_cpus_check,
> + .ident = "Sun Microsystems Sun Fire X4240",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "Sun Microsystems"),
> + DMI_MATCH(DMI_PRODUCT_NAME, "Sun Fire X4240"),
> + },
> + },
> + {
> + .callback = hotplug_cpus_check,
> + .ident = "Sun Microsystems Sun Fire X4140",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "Sun Microsystems"),
> + DMI_MATCH(DMI_PRODUCT_NAME, "Sun Fire X4140"),
> + },
> + },
> + {
> + .callback = hotplug_cpus_check,
> + .ident = "Sun Microsystems Sun Fire X4600",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "Sun Microsystems"),
> + DMI_MATCH(DMI_PRODUCT_NAME, "Sun Fire X4600"),
> + },
> + },
> + {
> + .callback = hotplug_cpus_check,
> + .ident = "Sun Microsystems Sun Fire X4640",
> + .matches = {
> + DMI_MATCH(DMI_SYS_VENDOR, "Sun Microsystems"),
> + DMI_MATCH(DMI_PRODUCT_NAME, "Sun Fire X4640"),
> + },
> + },
> + { } /* NULL entry stops DMI scanning */
> +};
> +
>
> /*
> * cpu_possible_mask should be static, it cannot change as cpu's
> @@ -1204,8 +1258,26 @@ __init void prefill_possible_map(void)
> if (!num_processors)
> num_processors = 1;
>
> - if (setup_possible_cpus == -1)
> - possible = num_processors + disabled_cpus;
> + if (setup_possible_cpus == -1) {
> + possible = num_processors;
> + /*
> + * do we have better way to detect hotplug cpus?
> + *
> + * some systems that have disable cpus entries because same
> + * BIOS will support 2 sockets and 4 sockets and more at
> + * same time, BIOS just leave some disabled entries with wild
> + * apicid, but those system do not support cpu hotplug.
> + * we don't need treat disabled_cpus as hotplug cpus.
> + * so we can make nr_cpu_ids smaller and save more space
> + * (pcpu data allocations), and could make some systems run
> + * with logical flat instead of physical flat apic mode
> + */
> + if (disabled_cpus) {
> + dmi_check_system(hotplug_cpus_dmi_table);
> + if (treat_disabled_cpus_as_hotplug)
> + possible += disabled_cpus;
> + }
> + }
> else
> possible = setup_possible_cpus;
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists