Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1360147
| From | Andy Lutomirski <luto@amacapital.net> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [PATCH 3/4] intel_idle: Fix MSRs after resume |
| Date | 2016-03-17 20:50 +0100 |
| Message-ID | <rdM5H-38A-7@gated-at.bofh.it> (permalink) |
| References | <rcssq-7f5-9@gated-at.bofh.it> <rcsC6-7iI-9@gated-at.bofh.it> <rcsC6-7iI-25@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
On Sun, Mar 13, 2016 at 9:48 PM, Andy Lutomirski <luto@kernel.org> wrote:
> Firmware that enables auto-promotion / auto-demotion flags we don't
> like will probably re-enable them after suspend/resume. Disable
> them again after resume so they stay fixed.
>
> I've seen this on my Dell XPS 13 9350.
This isn't right -- syscore's resume hook is called too early for
smp_call_function_many. I think the right way is through a device's
dev_pm_ops (SET_SYSTEM_SLEEP_PM_OPS), but there don't seem to be any
devices associated with intel_idle.
--Andy
>
> Signed-off-by: Andy Lutomirski <luto@kernel.org>
> ---
> drivers/idle/intel_idle.c | 12 ++++++++++++
> 1 file changed, 12 insertions(+)
>
> diff --git a/drivers/idle/intel_idle.c b/drivers/idle/intel_idle.c
> index 338df09ad60b..e3d7d8bbc843 100644
> --- a/drivers/idle/intel_idle.c
> +++ b/drivers/idle/intel_idle.c
> @@ -61,6 +61,7 @@
> #include <linux/notifier.h>
> #include <linux/cpu.h>
> #include <linux/module.h>
> +#include <linux/syscore_ops.h>
> #include <asm/cpu_device_id.h>
> #include <asm/mwait.h>
> #include <asm/msr.h>
> @@ -1026,6 +1027,15 @@ void intel_idle_state_table_update(void)
> return;
> }
>
> +static void intel_idle_resume(void)
> +{
> + on_each_cpu(fix_this_cpu, NULL, 1);
> +}
> +
> +static struct syscore_ops intel_idle_syscore_ops = {
> + .resume = intel_idle_resume,
> +};
> +
> /*
> * intel_idle_cpuidle_driver_init()
> * allocate, initialize cpuidle_states
> @@ -1119,6 +1129,7 @@ static int __init intel_idle_init(void)
> if (retval)
> return retval;
>
> + register_syscore_ops(&intel_idle_syscore_ops);
> intel_idle_cpuidle_driver_init();
> retval = cpuidle_register_driver(&intel_idle_driver);
> if (retval) {
> @@ -1153,6 +1164,7 @@ static void __exit intel_idle_exit(void)
> {
> intel_idle_cpuidle_devices_uninit();
> cpuidle_unregister_driver(&intel_idle_driver);
> + unregister_syscore_ops(&intel_idle_syscore_ops);
>
> cpu_notifier_register_begin();
>
> --
> 2.5.0
>
--
Andy Lutomirski
AMA Capital Management, LLC
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
C1E auto-promotion suspend/resume Andy Lutomirski <luto@amacapital.net> - 2016-03-14 02:40 +0100
RE: C1E auto-promotion suspend/resume "Brown, Len" <len.brown@intel.com> - 2016-03-14 05:40 +0100
[PATCH 0/4] intel_idle: Improve MSR fixup resume handling Andy Lutomirski <luto@kernel.org> - 2016-03-14 05:50 +0100
[PATCH 3/4] intel_idle: Fix MSRs after resume Andy Lutomirski <luto@kernel.org> - 2016-03-14 05:50 +0100
Re: [PATCH 3/4] intel_idle: Fix MSRs after resume Andy Lutomirski <luto@amacapital.net> - 2016-03-17 20:50 +0100
[PATCH 4/4] intel_idle: Move BYT/CHT auto-demotion fixup into fix_this_cpu Andy Lutomirski <luto@kernel.org> - 2016-03-14 05:50 +0100
[PATCH 1/4] intel_idle: Consolidate auto-promotion/auto-demotion fixups Andy Lutomirski <luto@kernel.org> - 2016-03-14 05:50 +0100
[PATCH 2/4] intel_idle: Remove a broadcast MSR fixup at boot Andy Lutomirski <luto@kernel.org> - 2016-03-14 05:50 +0100
Re: C1E auto-promotion suspend/resume Andy Lutomirski <luto@amacapital.net> - 2016-03-14 06:30 +0100
csiph-web