ACK/cmnt: [SRU][Xenial][PATCH] x86/speculation: Remove redundant arch_smt_update() invocation

Kleber Souza kleber.souza at canonical.com
Wed Jun 26 16:44:19 UTC 2019


On 6/13/19 12:56 PM, Juerg Haefliger wrote:
> From: Zhenzhong Duan <zhenzhong.duan at oracle.com>
> 
> With commit a74cfffb03b7 ("x86/speculation: Rework SMT state change"),
> arch_smt_update() is invoked from each individual CPU hotplug function.
> 
> Therefore the extra arch_smt_update() call in the sysfs SMT control is
> redundant.
> 
> Fixes: a74cfffb03b7 ("x86/speculation: Rework SMT state change")
> Signed-off-by: Zhenzhong Duan <zhenzhong.duan at oracle.com>
> Signed-off-by: Thomas Gleixner <tglx at linutronix.de>
> Cc: <konrad.wilk at oracle.com>
> Cc: <dwmw at amazon.co.uk>
> Cc: <bp at suse.de>
> Cc: <srinivas.eeda at oracle.com>
> Cc: <peterz at infradead.org>
> Cc: <hpa at zytor.com>
> Link: https://lkml.kernel.org/r/e2e064f2-e8ef-42ca-bf4f-76b612964752@default
> 
> CVE-2018-12126
> CVE-2018-12127
> CVE-2018-12130
> 
> (cherry picked from commit 34d66caf251df91ff27b24a3a786810d29989eca)
> Signed-off-by: Juerg Haefliger <juergh at canonical.com>

Juerg,

is this needed for newer series as well?


Acked-by: Kleber Sacilotto de Souza <kleber.souza at canonical.com>

Thank you,
Kleber

> ---
>  kernel/cpu.c | 5 +----
>  1 file changed, 1 insertion(+), 4 deletions(-)
> 
> diff --git a/kernel/cpu.c b/kernel/cpu.c
> index 5d48161476a1..d9465123f7ec 100644
> --- a/kernel/cpu.c
> +++ b/kernel/cpu.c
> @@ -896,10 +896,8 @@ static int cpuhp_smt_disable(enum cpuhp_smt_control ctrlval)
>  		 */
>  		cpuhp_offline_cpu_device(cpu);
>  	}
> -	if (!ret) {
> +	if (!ret)
>  		cpu_smt_control = ctrlval;
> -		arch_smt_update();
> -	}
>  	cpu_maps_update_done();
>  	return ret;
>  }
> @@ -910,7 +908,6 @@ static int cpuhp_smt_enable(void)
>  
>  	cpu_maps_update_begin();
>  	cpu_smt_control = CPU_SMT_ENABLED;
> -	arch_smt_update();
>  	for_each_present_cpu(cpu) {
>  		/* Skip online CPUs and CPUs on offline nodes */
>  		if (cpu_online(cpu) || !node_online(cpu_to_node(cpu)))
> 




More information about the kernel-team mailing list