2024-01-15 14:34:47

by Pierre Gondois

[permalink] [raw]
Subject: [PATCH v2 0/3] Small cleanup around sparse report

v2:
- Added __rcu annotations to variabbles in sched_reset_numa(),
and casts for calls to kfree().
- Renamed idle_cpu_without() to idle_rq_without()
- Added sched/fair: Use rq in update_blocked_averages()

While checking spare's tool report, some RCU pointers appeared
to be not annotated. In the same effort, idle_cpu_without() and
update_blocked_averages() seemed to be subject to a small
optimizations.

Pierre Gondois (3):
sched/topology: Annotate RCU pointers properly
sched/fair: Use rq in idle_cpu_without()
sched/fair: Use rq in update_blocked_averages()

kernel/sched/fair.c | 25 +++++++++++--------------
kernel/sched/topology.c | 12 ++++++------
2 files changed, 17 insertions(+), 20 deletions(-)

--
2.25.1



2024-01-15 14:35:01

by Pierre Gondois

[permalink] [raw]
Subject: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly

Cleanup RCU-related spare errors by annotating RCU pointers.

sched_domains_numa_distance:
error: incompatible types in comparison expression
(different address spaces):
int [noderef] __rcu *
int *

sched_domains_numa_masks:
error: incompatible types in comparison expression
(different address spaces):
struct cpumask **[noderef] __rcu *
struct cpumask ***

The cast to (void *) adds the following sparse warning:
warning: cast removes address space '__rcu' of expression
but this should be normal.

Fixes: 0fb3978b0aac ("sched/numa: Fix NUMA topology for systems with CPU-less nodes")
Signed-off-by: Pierre Gondois <[email protected]>
---
kernel/sched/topology.c | 12 ++++++------
1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
index 10d1391e7416..2a2da9b33e31 100644
--- a/kernel/sched/topology.c
+++ b/kernel/sched/topology.c
@@ -1542,8 +1542,8 @@ static int sched_domains_numa_levels;
static int sched_domains_curr_level;

int sched_max_numa_distance;
-static int *sched_domains_numa_distance;
-static struct cpumask ***sched_domains_numa_masks;
+static int __rcu *sched_domains_numa_distance;
+static struct cpumask ** __rcu *sched_domains_numa_masks;
#endif

/*
@@ -1988,8 +1988,8 @@ void sched_init_numa(int offline_node)

static void sched_reset_numa(void)
{
- int nr_levels, *distances;
- struct cpumask ***masks;
+ int nr_levels, __rcu *distances;
+ struct cpumask ** __rcu *masks;

nr_levels = sched_domains_numa_levels;
sched_domains_numa_levels = 0;
@@ -2003,7 +2003,7 @@ static void sched_reset_numa(void)
int i, j;

synchronize_rcu();
- kfree(distances);
+ kfree((void *)distances);
for (i = 0; i < nr_levels && masks; i++) {
if (!masks[i])
continue;
@@ -2011,7 +2011,7 @@ static void sched_reset_numa(void)
kfree(masks[i][j]);
kfree(masks[i]);
}
- kfree(masks);
+ kfree((void *)masks);
}
if (sched_domain_topology_saved) {
kfree(sched_domain_topology);
--
2.25.1


2024-01-15 14:35:13

by Pierre Gondois

[permalink] [raw]
Subject: [PATCH v2 2/3] sched/fair: Use rq in idle_cpu_without()

idle_cpu_without() could receive a 'struct rq' instead of a
CPU number to avoid converting the CPU number to a 'struct rq'
two times. Indeed update_sg_wakeup_stats() already makes the
conversion.
idle_cpu_without() is also renamed to idle_rq_without()
to match the input parameter.

Reviewed-by: Shrikanth Hegde <[email protected]>
Signed-off-by: Pierre Gondois <[email protected]>
---
kernel/sched/fair.c | 16 +++++++---------
1 file changed, 7 insertions(+), 9 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index d7a3c63a2171..97d9e0d32337 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -10105,23 +10105,21 @@ static unsigned int task_running_on_cpu(int cpu, struct task_struct *p)
}

/**
- * idle_cpu_without - would a given CPU be idle without p ?
- * @cpu: the processor on which idleness is tested.
+ * idle_rq_without - would a given rq be idle without p ?
+ * @rq: the rq on which idleness is tested.
* @p: task which should be ignored.
*
- * Return: 1 if the CPU would be idle. 0 otherwise.
+ * Return: 1 if the rq would be idle. 0 otherwise.
*/
-static int idle_cpu_without(int cpu, struct task_struct *p)
+static int idle_rq_without(struct rq *rq, struct task_struct *p)
{
- struct rq *rq = cpu_rq(cpu);
-
if (rq->curr != rq->idle && rq->curr != p)
return 0;

/*
* rq->nr_running can't be used but an updated version without the
* impact of p on cpu must be used instead. The updated nr_running
- * be computed and tested before calling idle_cpu_without().
+ * be computed and tested before calling idle_rq_without().
*/

#ifdef CONFIG_SMP
@@ -10166,9 +10164,9 @@ static inline void update_sg_wakeup_stats(struct sched_domain *sd,
sgs->sum_nr_running += nr_running;

/*
- * No need to call idle_cpu_without() if nr_running is not 0
+ * No need to call idle_rq_without() if nr_running is not 0
*/
- if (!nr_running && idle_cpu_without(i, p))
+ if (!nr_running && idle_rq_without(rq, p))
sgs->idle_cpus++;

/* Check if task fits in the CPU */
--
2.25.1


2024-01-15 14:35:26

by Pierre Gondois

[permalink] [raw]
Subject: [PATCH v2 3/3] sched/fair: Use rq in update_blocked_averages()

update_blocked_averages() could receive a 'struct rq' instead of a
CPU number to avoid converting the CPU number to a 'struct rq'
multiple times.

Suggested-by: Shrikanth Hegde <[email protected]>
Signed-off-by: Pierre Gondois <[email protected]>
---
kernel/sched/fair.c | 9 ++++-----
1 file changed, 4 insertions(+), 5 deletions(-)

diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c
index 97d9e0d32337..c70bb993f473 100644
--- a/kernel/sched/fair.c
+++ b/kernel/sched/fair.c
@@ -9349,10 +9349,9 @@ static unsigned long task_h_load(struct task_struct *p)
}
#endif

-static void update_blocked_averages(int cpu)
+static void update_blocked_averages(struct rq *rq)
{
bool decayed = false, done = true;
- struct rq *rq = cpu_rq(cpu);
struct rq_flags rf;

rq_lock_irqsave(rq, &rf);
@@ -12026,7 +12025,7 @@ static bool update_nohz_stats(struct rq *rq)
if (!time_after(jiffies, READ_ONCE(rq->last_blocked_load_update_tick)))
return true;

- update_blocked_averages(cpu);
+ update_blocked_averages(rq);

return rq->has_blocked_load;
}
@@ -12286,7 +12285,7 @@ static int newidle_balance(struct rq *this_rq, struct rq_flags *rf)
raw_spin_rq_unlock(this_rq);

t0 = sched_clock_cpu(this_cpu);
- update_blocked_averages(this_cpu);
+ update_blocked_averages(this_rq);

rcu_read_lock();
for_each_domain(this_cpu, sd) {
@@ -12376,7 +12375,7 @@ static __latent_entropy void run_rebalance_domains(struct softirq_action *h)
return;

/* normal load balance */
- update_blocked_averages(this_rq->cpu);
+ update_blocked_averages(this_rq);
rebalance_domains(this_rq, idle);
}

--
2.25.1


2024-01-16 01:54:34

by Huang, Ying

[permalink] [raw]
Subject: Re: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly

Pierre Gondois <[email protected]> writes:

> Cleanup RCU-related spare errors by annotating RCU pointers.
>
> sched_domains_numa_distance:
> error: incompatible types in comparison expression
> (different address spaces):
> int [noderef] __rcu *
> int *
>
> sched_domains_numa_masks:
> error: incompatible types in comparison expression
> (different address spaces):
> struct cpumask **[noderef] __rcu *
> struct cpumask ***
>
> The cast to (void *) adds the following sparse warning:
> warning: cast removes address space '__rcu' of expression
> but this should be normal.
>
> Fixes: 0fb3978b0aac ("sched/numa: Fix NUMA topology for systems with CPU-less nodes")
> Signed-off-by: Pierre Gondois <[email protected]>
> ---
> kernel/sched/topology.c | 12 ++++++------
> 1 file changed, 6 insertions(+), 6 deletions(-)
>
> diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
> index 10d1391e7416..2a2da9b33e31 100644
> --- a/kernel/sched/topology.c
> +++ b/kernel/sched/topology.c
> @@ -1542,8 +1542,8 @@ static int sched_domains_numa_levels;
> static int sched_domains_curr_level;
>
> int sched_max_numa_distance;
> -static int *sched_domains_numa_distance;
> -static struct cpumask ***sched_domains_numa_masks;
> +static int __rcu *sched_domains_numa_distance;
> +static struct cpumask ** __rcu *sched_domains_numa_masks;
> #endif
>
> /*
> @@ -1988,8 +1988,8 @@ void sched_init_numa(int offline_node)
>
> static void sched_reset_numa(void)
> {
> - int nr_levels, *distances;
> - struct cpumask ***masks;
> + int nr_levels, __rcu *distances;
> + struct cpumask ** __rcu *masks;

No. distances and masks are not accessed via RCU in the function.
Instead, they should be assigned like below,

distances = rcu_dereference_raw(sched_domains_numa_distance);

Because sched_domains_numa_distance is protected by cpu_hotplug_lock,
but the lock is static. Some comments are deserved.

Anyway, please read RCU document before making the change.

https://docs.kernel.org/RCU/index.html

> nr_levels = sched_domains_numa_levels;
> sched_domains_numa_levels = 0;
> @@ -2003,7 +2003,7 @@ static void sched_reset_numa(void)
> int i, j;
>
> synchronize_rcu();
> - kfree(distances);
> + kfree((void *)distances);
> for (i = 0; i < nr_levels && masks; i++) {
> if (!masks[i])
> continue;
> @@ -2011,7 +2011,7 @@ static void sched_reset_numa(void)
> kfree(masks[i][j]);
> kfree(masks[i]);
> }
> - kfree(masks);
> + kfree((void *)masks);
> }
> if (sched_domain_topology_saved) {
> kfree(sched_domain_topology);

--
Best Regards,
Huang, Ying

2024-01-16 04:03:50

by kernel test robot

[permalink] [raw]
Subject: Re: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly

Hi Pierre,

kernel test robot noticed the following build warnings:

[auto build test WARNING on tip/sched/core]
[also build test WARNING on linus/master v6.7 next-20240112]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url: https://github.com/intel-lab-lkp/linux/commits/Pierre-Gondois/sched-topology-Annotate-RCU-pointers-properly/20240115-223658
base: tip/sched/core
patch link: https://lore.kernel.org/r/20240115143427.1820628-2-pierre.gondois%40arm.com
patch subject: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly
config: arm64-randconfig-r133-20240116 (https://download.01.org/0day-ci/archive/20240116/[email protected]/config)
compiler: aarch64-linux-gcc (GCC) 13.2.0
reproduce: (https://download.01.org/0day-ci/archive/20240116/[email protected]/reproduce)

If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <[email protected]>
| Closes: https://lore.kernel.org/oe-kbuild-all/[email protected]/

sparse warnings: (new ones prefixed by >>)
kernel/sched/build_utility.c: note: in included file:
kernel/sched/debug.c:468:17: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/debug.c:468:17: sparse: expected struct sched_domain *[assigned] sd
kernel/sched/debug.c:468:17: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/debug.c:793:9: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct task_struct *tsk @@ got struct task_struct [noderef] __rcu *curr @@
kernel/sched/debug.c:793:9: sparse: expected struct task_struct *tsk
kernel/sched/debug.c:793:9: sparse: got struct task_struct [noderef] __rcu *curr
kernel/sched/debug.c:793:9: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct task_struct *tsk @@ got struct task_struct [noderef] __rcu *curr @@
kernel/sched/debug.c:793:9: sparse: expected struct task_struct *tsk
kernel/sched/debug.c:793:9: sparse: got struct task_struct [noderef] __rcu *curr
kernel/sched/build_utility.c: note: in included file:
kernel/sched/stats.c:148:17: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/stats.c:148:17: sparse: expected struct sched_domain *[assigned] sd
kernel/sched/stats.c:148:17: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/build_utility.c: note: in included file:
kernel/sched/topology.c:107:56: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:107:56: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:107:56: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:126:60: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:126:60: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:126:60: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:149:20: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:149:20: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:149:20: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:485:19: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct perf_domain *pd @@ got struct perf_domain [noderef] __rcu *pd @@
kernel/sched/topology.c:485:19: sparse: expected struct perf_domain *pd
kernel/sched/topology.c:485:19: sparse: got struct perf_domain [noderef] __rcu *pd
kernel/sched/topology.c:647:49: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected struct sched_domain *parent @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:647:49: sparse: expected struct sched_domain *parent
kernel/sched/topology.c:647:49: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:732:50: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected struct sched_domain *parent @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:732:50: sparse: expected struct sched_domain *parent
kernel/sched/topology.c:732:50: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:740:55: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain [noderef] __rcu *[noderef] __rcu child @@ got struct sched_domain *[assigned] tmp @@
kernel/sched/topology.c:740:55: sparse: expected struct sched_domain [noderef] __rcu *[noderef] __rcu child
kernel/sched/topology.c:740:55: sparse: got struct sched_domain *[assigned] tmp
kernel/sched/topology.c:753:29: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] tmp @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:753:29: sparse: expected struct sched_domain *[assigned] tmp
kernel/sched/topology.c:753:29: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:758:20: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:758:20: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:758:20: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:779:13: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] tmp @@ got struct sched_domain [noderef] __rcu *sd @@
kernel/sched/topology.c:779:13: sparse: expected struct sched_domain *[assigned] tmp
kernel/sched/topology.c:779:13: sparse: got struct sched_domain [noderef] __rcu *sd
kernel/sched/topology.c:941:70: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:941:70: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:941:70: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:970:59: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:970:59: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:970:59: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1016:57: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1016:57: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:1016:57: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1018:25: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *sibling @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1018:25: sparse: expected struct sched_domain *sibling
kernel/sched/topology.c:1018:25: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1026:55: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1026:55: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:1026:55: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1028:25: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *sibling @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1028:25: sparse: expected struct sched_domain *sibling
kernel/sched/topology.c:1028:25: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1098:62: sparse: sparse: incorrect type in argument 1 (different address spaces) @@ expected struct sched_domain *sd @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1098:62: sparse: expected struct sched_domain *sd
kernel/sched/topology.c:1098:62: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1202:40: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected struct sched_domain *child @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:1202:40: sparse: expected struct sched_domain *child
kernel/sched/topology.c:1202:40: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1622:43: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected struct sched_domain [noderef] __rcu *child @@ got struct sched_domain *child @@
kernel/sched/topology.c:1622:43: sparse: expected struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:1622:43: sparse: got struct sched_domain *child
>> kernel/sched/topology.c:2006:24: sparse: sparse: cast removes address space '__rcu' of expression
kernel/sched/topology.c:2014:24: sparse: sparse: cast removes address space '__rcu' of expression
kernel/sched/topology.c:2321:31: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain [noderef] __rcu *parent @@ got struct sched_domain *sd @@
kernel/sched/topology.c:2321:31: sparse: expected struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:2321:31: sparse: got struct sched_domain *sd
kernel/sched/topology.c:2425:57: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:2425:57: sparse: expected struct sched_domain *[assigned] sd
kernel/sched/topology.c:2425:57: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:2446:56: sparse: sparse: incorrect type in initializer (different address spaces) @@ expected struct sched_domain *child @@ got struct sched_domain [noderef] __rcu *child @@
kernel/sched/topology.c:2446:56: sparse: expected struct sched_domain *child
kernel/sched/topology.c:2446:56: sparse: got struct sched_domain [noderef] __rcu *child
kernel/sched/topology.c:2445:57: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:2445:57: sparse: expected struct sched_domain *[assigned] sd
kernel/sched/topology.c:2445:57: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/topology.c:2500:57: sparse: sparse: incorrect type in assignment (different address spaces) @@ expected struct sched_domain *[assigned] sd @@ got struct sched_domain [noderef] __rcu *parent @@
kernel/sched/topology.c:2500:57: sparse: expected struct sched_domain *[assigned] sd
kernel/sched/topology.c:2500:57: sparse: got struct sched_domain [noderef] __rcu *parent
kernel/sched/build_utility.c: note: in included file:
kernel/sched/core_sched.c:276:37: sparse: sparse: incompatible types in conditional expression (different address spaces):
kernel/sched/core_sched.c:276:37: sparse: struct task_struct *
kernel/sched/core_sched.c:276:37: sparse: struct task_struct [noderef] __rcu *
kernel/sched/build_utility.c: note: in included file:
kernel/sched/sched.h:2142:25: sparse: sparse: incompatible types in comparison expression (different address spaces):
kernel/sched/sched.h:2142:25: sparse: struct task_struct [noderef] __rcu *
kernel/sched/sched.h:2142:25: sparse: struct task_struct *

vim +/__rcu +2006 kernel/sched/topology.c

1987
1988
1989 static void sched_reset_numa(void)
1990 {
1991 int nr_levels, __rcu *distances;
1992 struct cpumask ** __rcu *masks;
1993
1994 nr_levels = sched_domains_numa_levels;
1995 sched_domains_numa_levels = 0;
1996 sched_max_numa_distance = 0;
1997 sched_numa_topology_type = NUMA_DIRECT;
1998 distances = sched_domains_numa_distance;
1999 rcu_assign_pointer(sched_domains_numa_distance, NULL);
2000 masks = sched_domains_numa_masks;
2001 rcu_assign_pointer(sched_domains_numa_masks, NULL);
2002 if (distances || masks) {
2003 int i, j;
2004
2005 synchronize_rcu();
> 2006 kfree((void *)distances);
2007 for (i = 0; i < nr_levels && masks; i++) {
2008 if (!masks[i])
2009 continue;
2010 for_each_node(j)
2011 kfree(masks[i][j]);
2012 kfree(masks[i]);
2013 }
2014 kfree((void *)masks);
2015 }
2016 if (sched_domain_topology_saved) {
2017 kfree(sched_domain_topology);
2018 sched_domain_topology = sched_domain_topology_saved;
2019 sched_domain_topology_saved = NULL;
2020 }
2021 }
2022

--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki

2024-02-02 17:36:03

by Valentin Schneider

[permalink] [raw]
Subject: Re: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly

On 16/01/24 09:23, Huang, Ying wrote:
> Pierre Gondois <[email protected]> writes:
>
>> Cleanup RCU-related spare errors by annotating RCU pointers.
>>
>> sched_domains_numa_distance:
>> error: incompatible types in comparison expression
>> (different address spaces):
>> int [noderef] __rcu *
>> int *
>>
>> sched_domains_numa_masks:
>> error: incompatible types in comparison expression
>> (different address spaces):
>> struct cpumask **[noderef] __rcu *
>> struct cpumask ***
>>
>> The cast to (void *) adds the following sparse warning:
>> warning: cast removes address space '__rcu' of expression
>> but this should be normal.
>>
>> Fixes: 0fb3978b0aac ("sched/numa: Fix NUMA topology for systems with CPU-less nodes")
>> Signed-off-by: Pierre Gondois <[email protected]>
>> ---
>> kernel/sched/topology.c | 12 ++++++------
>> 1 file changed, 6 insertions(+), 6 deletions(-)
>>
>> diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
>> index 10d1391e7416..2a2da9b33e31 100644
>> --- a/kernel/sched/topology.c
>> +++ b/kernel/sched/topology.c
>> @@ -1542,8 +1542,8 @@ static int sched_domains_numa_levels;
>> static int sched_domains_curr_level;
>>
>> int sched_max_numa_distance;
>> -static int *sched_domains_numa_distance;
>> -static struct cpumask ***sched_domains_numa_masks;
>> +static int __rcu *sched_domains_numa_distance;
>> +static struct cpumask ** __rcu *sched_domains_numa_masks;
>> #endif
>>
>> /*
>> @@ -1988,8 +1988,8 @@ void sched_init_numa(int offline_node)
>>
>> static void sched_reset_numa(void)
>> {
>> - int nr_levels, *distances;
>> - struct cpumask ***masks;
>> + int nr_levels, __rcu *distances;
>> + struct cpumask ** __rcu *masks;
>
> No. distances and masks are not accessed via RCU in the function.
> Instead, they should be assigned like below,
>
> distances = rcu_dereference_raw(sched_domains_numa_distance);
>
> Because sched_domains_numa_distance is protected by cpu_hotplug_lock,
> but the lock is static. Some comments are deserved.
>
> Anyway, please read RCU document before making the change.
>

IIUC, something like so could also do?

distances = rcu_dereference_check(sched_domains_numa_distance, lockdep_is_cpus_held());


2024-02-04 01:30:22

by Huang, Ying

[permalink] [raw]
Subject: Re: [PATCH v2 1/3] sched/topology: Annotate RCU pointers properly

Valentin Schneider <[email protected]> writes:

> On 16/01/24 09:23, Huang, Ying wrote:
>> Pierre Gondois <[email protected]> writes:
>>
>>> Cleanup RCU-related spare errors by annotating RCU pointers.
>>>
>>> sched_domains_numa_distance:
>>> error: incompatible types in comparison expression
>>> (different address spaces):
>>> int [noderef] __rcu *
>>> int *
>>>
>>> sched_domains_numa_masks:
>>> error: incompatible types in comparison expression
>>> (different address spaces):
>>> struct cpumask **[noderef] __rcu *
>>> struct cpumask ***
>>>
>>> The cast to (void *) adds the following sparse warning:
>>> warning: cast removes address space '__rcu' of expression
>>> but this should be normal.
>>>
>>> Fixes: 0fb3978b0aac ("sched/numa: Fix NUMA topology for systems with CPU-less nodes")
>>> Signed-off-by: Pierre Gondois <[email protected]>
>>> ---
>>> kernel/sched/topology.c | 12 ++++++------
>>> 1 file changed, 6 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/kernel/sched/topology.c b/kernel/sched/topology.c
>>> index 10d1391e7416..2a2da9b33e31 100644
>>> --- a/kernel/sched/topology.c
>>> +++ b/kernel/sched/topology.c
>>> @@ -1542,8 +1542,8 @@ static int sched_domains_numa_levels;
>>> static int sched_domains_curr_level;
>>>
>>> int sched_max_numa_distance;
>>> -static int *sched_domains_numa_distance;
>>> -static struct cpumask ***sched_domains_numa_masks;
>>> +static int __rcu *sched_domains_numa_distance;
>>> +static struct cpumask ** __rcu *sched_domains_numa_masks;
>>> #endif
>>>
>>> /*
>>> @@ -1988,8 +1988,8 @@ void sched_init_numa(int offline_node)
>>>
>>> static void sched_reset_numa(void)
>>> {
>>> - int nr_levels, *distances;
>>> - struct cpumask ***masks;
>>> + int nr_levels, __rcu *distances;
>>> + struct cpumask ** __rcu *masks;
>>
>> No. distances and masks are not accessed via RCU in the function.
>> Instead, they should be assigned like below,
>>
>> distances = rcu_dereference_raw(sched_domains_numa_distance);
>>
>> Because sched_domains_numa_distance is protected by cpu_hotplug_lock,
>> but the lock is static. Some comments are deserved.
>>
>> Anyway, please read RCU document before making the change.
>>
>
> IIUC, something like so could also do?
>
> distances = rcu_dereference_check(sched_domains_numa_distance, lockdep_is_cpus_held());

Yes. You are right. We should do that.

--
Best Regards,
Huang, Ying