| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <942d2b65-3c28-4f6d-aa6b-8365fa5cea2c@redhat.com>
Date: Tue, 18 Nov 2025 14:53:27 -0500
From: Waiman Long <llong@...hat.com>
To: Sun Shaojie <sunshaojie@...inos.cn>, llong@...hat.com
Cc: chenridong@...weicloud.com, mkoutny@...e.com, cgroups@...r.kernel.org,
hannes@...xchg.org, linux-kernel@...r.kernel.org,
linux-kselftest@...r.kernel.org, shuah@...nel.org, tj@...nel.org
Subject: Re: [PATCH v4 1/1] cpuset: relax the overlap check for cgroup-v2
On 11/16/25 8:57 PM, Sun Shaojie wrote:
> In cgroup v2, a mutual overlap check is required when at least one of two
> cpusets is exclusive. However, this check should be relaxed and limited to
> cases where both cpusets are exclusive.
>
> This patch ensures that for sibling cpusets A1 (exclusive) and B1
> (non-exclusive), change B1 cannot affect A1's exclusivity.
>
> for example. Assume a machine has 4 CPUs (0-3).
>
> root cgroup
> / \
> A1 B1
>
> Case 1:
> Table 1.1: Before applying the patch
> Step | A1's prstate | B1'sprstate |
> #1> echo "0-1" > A1/cpuset.cpus | member | member |
> #2> echo "root" > A1/cpuset.cpus.partition | root | member |
> #3> echo "0" > B1/cpuset.cpus | root invalid | member |
>
> After step #3, A1 changes from "root" to "root invalid" because its CPUs
> (0-1) overlap with those requested by B1 (0-3). However, B1 can actually
> use CPUs 2-3(from B1's parent), so it would be more reasonable for A1 to
> remain as "root."
>
> Table 1.2: After applying the patch
> Step | A1's prstate | B1'sprstate |
> #1> echo "0-1" > A1/cpuset.cpus | member | member |
> #2> echo "root" > A1/cpuset.cpus.partition | root | member |
> #3> echo "0" > B1/cpuset.cpus | root | member |
>
> All other cases remain unaffected. For example, cgroup-v1, both A1 and B1
> are exclusive or non-exlusive.
>
> Signed-off-by: Sun Shaojie <sunshaojie@...inos.cn>
> ---
> kernel/cgroup/cpuset-internal.h | 3 ++
> kernel/cgroup/cpuset-v1.c | 20 +++++++++
> kernel/cgroup/cpuset.c | 43 ++++++++++++++-----
> .../selftests/cgroup/test_cpuset_prs.sh | 5 ++-
> 4 files changed, 58 insertions(+), 13 deletions(-)
>
> diff --git a/kernel/cgroup/cpuset-internal.h b/kernel/cgroup/cpuset-internal.h
> index 337608f408ce..c53111998432 100644
> --- a/kernel/cgroup/cpuset-internal.h
> +++ b/kernel/cgroup/cpuset-internal.h
> @@ -292,6 +292,7 @@ void cpuset1_hotplug_update_tasks(struct cpuset *cs,
> struct cpumask *new_cpus, nodemask_t *new_mems,
> bool cpus_updated, bool mems_updated);
> int cpuset1_validate_change(struct cpuset *cur, struct cpuset *trial);
> +bool cpuset1_cpus_excl_conflict(struct cpuset *cs1, struct cpuset *cs2);
> #else
> static inline void fmeter_init(struct fmeter *fmp) {}
> static inline void cpuset1_update_task_spread_flags(struct cpuset *cs,
> @@ -302,6 +303,8 @@ static inline void cpuset1_hotplug_update_tasks(struct cpuset *cs,
> bool cpus_updated, bool mems_updated) {}
> static inline int cpuset1_validate_change(struct cpuset *cur,
> struct cpuset *trial) { return 0; }
> +static inline bool cpuset1_cpus_excl_conflict(struct cpuset *cs1,
> + struct cpuset *cs2) {return false; }
> #endif /* CONFIG_CPUSETS_V1 */
>
> #endif /* __CPUSET_INTERNAL_H */
> diff --git a/kernel/cgroup/cpuset-v1.c b/kernel/cgroup/cpuset-v1.c
> index 12e76774c75b..5c1296bf6a34 100644
> --- a/kernel/cgroup/cpuset-v1.c
> +++ b/kernel/cgroup/cpuset-v1.c
> @@ -373,6 +373,26 @@ int cpuset1_validate_change(struct cpuset *cur, struct cpuset *trial)
> return ret;
> }
>
> +/*
> + * cpuset1_cpus_excl_conflict() - Check if two cpusets have exclusive CPU conflicts
> + * to legacy (v1)
> + * @cs1: first cpuset to check
> + * @cs2: second cpuset to check
> + *
> + * Returns: true if CPU exclusivity conflict exists, false otherwise
> + *
> + * If either cpuset is CPU exclusive, their allowed CPUs cannot intersect.
> + */
> +bool cpuset1_cpus_excl_conflict(struct cpuset *cs1, struct cpuset *cs2)
> +{
> + if (is_cpu_exclusive(cs1) || is_cpu_exclusive(cs2))
> + if (cpumask_intersects(cs1->cpus_allowed,
> + cs2->cpus_allowed))
> + return true;
> +
> + return false;
> +}
> +
> #ifdef CONFIG_PROC_PID_CPUSET
> /*
> * proc_cpuset_show()
> diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
> index 52468d2c178a..0fd803612513 100644
> --- a/kernel/cgroup/cpuset.c
> +++ b/kernel/cgroup/cpuset.c
> @@ -580,35 +580,56 @@ static inline bool cpusets_are_exclusive(struct cpuset *cs1, struct cpuset *cs2)
>
> /**
> * cpus_excl_conflict - Check if two cpusets have exclusive CPU conflicts
> - * @cs1: first cpuset to check
> - * @cs2: second cpuset to check
> + * @cs1: current cpuset to check
> + * @cs2: cpuset involved in the check
> *
> * Returns: true if CPU exclusivity conflict exists, false otherwise
> *
> * Conflict detection rules:
> - * 1. If either cpuset is CPU exclusive, they must be mutually exclusive
> + * For cgroup-v1:
> + * see cpuset1_cpus_excl_conflict()
> + * For cgroup-v2:
> + * 1. If cs1 is exclusive, cs1 and cs2 must be mutually exclusive
> * 2. exclusive_cpus masks cannot intersect between cpusets
> - * 3. The allowed CPUs of one cpuset cannot be a subset of another's exclusive CPUs
> + * 3. If cs2 is exclusive, cs2's allowed CPUs cannot be a subset of cs1's exclusive CPUs
> + * 4. if cs1 and cs2 are not exclusive, the allowed CPUs of one cpuset cannot be a subset
> + * of another's exclusive CPUs
> */
> static inline bool cpus_excl_conflict(struct cpuset *cs1, struct cpuset *cs2)
As cs1 and cs2 is going to be handled differently, their current naming
will make it hard to understand why they are treated differently. I will
recommended changing the parameter name to "trial, sibling" as the
caller call it with "cpus_excl_conflict(trial, c)" where trial is the
new cpuset data to be tested and sibling is one of its sibling cpusets.
It has to be clearly document what each parameter is for and the fact
that swapping the parameters will cause it to return incorrect result.
> {
> - /* If either cpuset is exclusive, check if they are mutually exclusive */
> - if (is_cpu_exclusive(cs1) || is_cpu_exclusive(cs2))
> + /* For cgroup-v1 */
> + if (!cpuset_v2())
> + return cpuset1_cpus_excl_conflict(cs1, cs2);
> +
> + /* If cs1 are exclusive, check if they are mutually exclusive */
> + if (is_cpu_exclusive(cs1))
> return !cpusets_are_exclusive(cs1, cs2);
Code change like the following can eliminate the need to introduce a new
cpuset1_cpus_excl_conflict() helper.
diff --git a/kernel/cgroup/cpuset.c b/kernel/cgroup/cpuset.c
index ec8bebc66469..201c70fb7401 100644
--- a/kernel/cgroup/cpuset.c
+++ b/kernel/cgroup/cpuset.c
@@ -599,9 +599,15 @@ static inline bool cpusets_are_exclusive(struct
cpuset *cs1, struct cpuset *cs2)
*/
static inline bool cpus_excl_conflict(struct cpuset *cs1, struct
cpuset *cs2)
{
- /* If either cpuset is exclusive, check if they are mutually
exclusive */
- if (is_cpu_exclusive(cs1) || is_cpu_exclusive(cs2))
- return !cpusets_are_exclusive(cs1, cs2);
+ /*
+ * If trial is exclusive or sibling is exclusive & in v1,
+ * check if they are mutually exclusive
+ */
+ if (is_cpu_exclusive(trial) || (!cpuset_v2() &&
is_cpu_exclusive(sibling)))
+ return !cpusets_are_exclusive(trial, sibling);
+
+ if (!cpuset_v2())
+ return false; /* The checking below is irrelevant to
cpuset v1 */
/* Exclusive_cpus cannot intersect */
if (cpumask_intersects(cs1->exclusive_cpus, cs2->exclusive_cpus))
>
> + /* The following check applies when either
> + * both cs1 and cs2 are non-exclusive，or
> + * only cs2 is exclusive.
> + */
> +
> /* Exclusive_cpus cannot intersect */
> if (cpumask_intersects(cs1->exclusive_cpus, cs2->exclusive_cpus))
> return true;
>
> - /* The cpus_allowed of one cpuset cannot be a subset of another cpuset's exclusive_cpus */
> - if (!cpumask_empty(cs1->cpus_allowed) &&
> - cpumask_subset(cs1->cpus_allowed, cs2->exclusive_cpus))
> - return true;
> -
> + /* cs2's allowed CPUs cannot be a subset of cs1's exclusive CPUs */
> if (!cpumask_empty(cs2->cpus_allowed) &&
> cpumask_subset(cs2->cpus_allowed, cs1->exclusive_cpus))
> return true;
>
> + /* If cs2 is exclusive, check finished here */
> + if (is_cpu_exclusive(cs2))
> + return false;
> +
> + /* The following check applies only if both cs1 and cs2 are non-exclusive. */
> +
> + /* cs1's allowed CPUs cannot be a subset of cs1's exclusive CPUs */
"sibling's exclusive CPUs"
> + if (!cpumask_empty(cs1->cpus_allowed) &&
> + cpumask_subset(cs1->cpus_allowed, cs2->exclusive_cpus))
> + return true;
> +
As said before, we can't fail change to cpuset.cpus by default, but we
can fail change to cpuset.cpus.exclusive. So this additional check isn't
OK unless this check is under a special mode that is opted in via other
means like an additional cgroup control file or a boot command line
option and so on.
Cheers,
Longman
Powered by blists - more mailing lists