[Devel] [PATCH vz10] sched: enforce the 0..1 bound on sched_cpulimit_scale_cpufreq

[Devel] [PATCH vz10] sched: enforce the 0..1 bound on sched_cpulimit_scale_cpufreq

[Konstantin Khorenko]

The sched_cpulimit_scale_cpufreq sysctl was registered with
  .extra1 = SYSCTL_ZERO
  .extra2 = SYSCTL_ONE
but with .proc_handler = proc_dointvec.

proc_dointvec() ignores extra1/extra2, so the intended [0, 1] clamp was
never applied. sysctl_sched_cpulimit_scale_cpufreq is also unsigned int,
so proc_dointvec() is the wrong signedness for it in the first place.

Use proc_douintvec_minmax() so the already-declared bounds are honoured
and an out-of-range write is rejected.

Fixes: 4e99efb641da0 ("sched: Port CONFIG_CFS_CPULIMIT feature")
Feature: sched: ability to limit number of CPUs available to a CT
https://virtuozzo.atlassian.net/browse/VSTOR-137234
Signed-off-by: Konstantin Khorenko <khorenko@virtuozzo.com>
---
 kernel/sched/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index d5b4d8c97a0c7..7580067e7f44a 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -4753,7 +4753,7 @@ static struct ctl_table sched_core_sysctls[] = {
 		.data		= &sysctl_sched_cpulimit_scale_cpufreq,
 		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= proc_dointvec,
+		.proc_handler	= proc_douintvec_minmax,
 		.extra1		= SYSCTL_ZERO,
 		.extra2		= SYSCTL_ONE,
 	},
-- 
2.43.0

Re: [Devel] [PATCH vz10] sched: enforce the 0..1 bound on sched_cpulimit_scale_cpufreq

[Konstantin Khorenko]

a kind ping

--
Best regards,

Konstantin Khorenko,
Virtuozzo Linux Kernel Team

On 8/5/26 15:45, Konstantin Khorenko wrote:
> The sched_cpulimit_scale_cpufreq sysctl was registered with
>   .extra1 = SYSCTL_ZERO
>   .extra2 = SYSCTL_ONE
> but with .proc_handler = proc_dointvec.
> 
> proc_dointvec() ignores extra1/extra2, so the intended [0, 1] clamp was
> never applied. sysctl_sched_cpulimit_scale_cpufreq is also unsigned int,
> so proc_dointvec() is the wrong signedness for it in the first place.
> 
> Use proc_douintvec_minmax() so the already-declared bounds are honoured
> and an out-of-range write is rejected.
> 
> Fixes: 4e99efb641da0 ("sched: Port CONFIG_CFS_CPULIMIT feature")
> Feature: sched: ability to limit number of CPUs available to a CT
> https://virtuozzo.atlassian.net/browse/VSTOR-137234
> Signed-off-by: Konstantin Khorenko <khorenko@virtuozzo.com>
> ---
>  kernel/sched/core.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index d5b4d8c97a0c7..7580067e7f44a 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -4753,7 +4753,7 @@ static struct ctl_table sched_core_sysctls[] = {
>  		.data		= &sysctl_sched_cpulimit_scale_cpufreq,
>  		.maxlen		= sizeof(unsigned int),
>  		.mode		= 0644,
> -		.proc_handler	= proc_dointvec,
> +		.proc_handler	= proc_douintvec_minmax,
>  		.extra1		= SYSCTL_ZERO,
>  		.extra2		= SYSCTL_ONE,
>  	},

_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel

Re: [Devel] [PATCH RHEL10 COMMIT] sched: enforce the 0..1 bound on sched_cpulimit_scale_cpufreq

[Konstantin Khorenko]

The commit is pushed to "branch-rh10-6.12.0-211.39.1.16.x.vz10-ovz" and will appear at git@bitbucket.org:openvz/vzkernel.git
after rh10-6.12.0-211.39.1.16.11.vz10
------>
commit 58d8489a47affd692c11d97592c092a3c377e047
Author: Konstantin Khorenko <khorenko@virtuozzo.com>
Date:   Wed Aug 5 15:45:32 2026 +0200

    sched: enforce the 0..1 bound on sched_cpulimit_scale_cpufreq
    
    The sched_cpulimit_scale_cpufreq sysctl was registered with
      .extra1 = SYSCTL_ZERO
      .extra2 = SYSCTL_ONE
    but with .proc_handler = proc_dointvec.
    
    proc_dointvec() ignores extra1/extra2, so the intended [0, 1] clamp was
    never applied. sysctl_sched_cpulimit_scale_cpufreq is also unsigned int,
    so proc_dointvec() is the wrong signedness for it in the first place.
    
    Use proc_douintvec_minmax() so the already-declared bounds are honoured
    and an out-of-range write is rejected.
    
    Fixes: 4e99efb641da0 ("sched: Port CONFIG_CFS_CPULIMIT feature")
    Feature: sched: ability to limit number of CPUs available to a CT
    https://virtuozzo.atlassian.net/browse/VSTOR-137234
    Signed-off-by: Konstantin Khorenko <khorenko@virtuozzo.com>
    Reviewed-by: Vasileios Almpanis <vasileios.almpanis@virtuozzo.com>
---
 kernel/sched/core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index fb2477cb1d076..9275501bc3ad0 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -4755,7 +4755,7 @@ static struct ctl_table sched_core_sysctls[] = {
 		.data		= &sysctl_sched_cpulimit_scale_cpufreq,
 		.maxlen		= sizeof(unsigned int),
 		.mode		= 0644,
-		.proc_handler	= proc_dointvec,
+		.proc_handler	= proc_douintvec_minmax,
 		.extra1		= SYSCTL_ZERO,
 		.extra2		= SYSCTL_ONE,
 	},
_______________________________________________
Devel mailing list
Devel@openvz.org
https://lists.openvz.org/mailman/listinfo/devel