Re: [PATCH] drivers: use __free attribute instead of of_node_put()
From: Vincenzo Mezzela
Date: Mon Apr 22 2024 - 04:27:59 EST
On 19/04/24 16:01, Sudeep Holla wrote:
On Fri, Apr 19, 2024 at 03:19:56PM +0200, Vincenzo Mezzela wrote:
Introduce the __free attribute for scope-based resource management.
Resources allocated with __free are automatically released at the end of
the scope. This enhancement aims to mitigate memory management issues
associated with forgetting to release resources by utilizing __free
instead of of_node_put().
The declaration of the device_node used within the do-while loops is
moved directly within the loop so that the resource is automatically
freed at the end of each iteration.
Suggested-by: Julia Lawall <julia.lawall@xxxxxxxx>
Signed-off-by: Vincenzo Mezzela <vincenzo.mezzela@xxxxxxxxx>
---
drivers/base/arch_topology.c | 41 ++++++++++++++----------------------
1 file changed, 16 insertions(+), 25 deletions(-)
diff --git a/drivers/base/arch_topology.c b/drivers/base/arch_topology.c
index 024b78a0cfc1..58eeb8183747 100644
--- a/drivers/base/arch_topology.c
+++ b/drivers/base/arch_topology.c
@@ -513,10 +513,10 @@ core_initcall(free_raw_capacity);
*/
static int __init get_cpu_for_node(struct device_node *node)
{
- struct device_node *cpu_node;
int cpu;
- cpu_node = of_parse_phandle(node, "cpu", 0);
+ struct device_node *cpu_node __free(device_node) =
Missing include <linux/cleanup.h> for this ?
+ of_parse_phandle(node, "cpu", 0);
if (!cpu_node)
return -1;
@@ -527,7 +527,6 @@ static int __init get_cpu_for_node(struct device_node *node)
pr_info("CPU node for %pOF exist but the possible cpu range is :%*pbl\n",
cpu_node, cpumask_pr_args(cpu_possible_mask));
- of_node_put(cpu_node);
return cpu;
}
@@ -538,11 +537,11 @@ static int __init parse_core(struct device_node *core, int package_id,
bool leaf = true;
int i = 0;
int cpu;
- struct device_node *t;
do {
snprintf(name, sizeof(name), "thread%d", i);
- t = of_get_child_by_name(core, name);
+ struct device_node *t __free(device_node) =
+ of_get_child_by_name(core, name);
if (t) {
leaf = false;
cpu = get_cpu_for_node(t);
@@ -553,10 +552,8 @@ static int __init parse_core(struct device_node *core, int package_id,
cpu_topology[cpu].thread_id = i;
} else if (cpu != -ENODEV) {
pr_err("%pOF: Can't get CPU for thread\n", t);
- of_node_put(t);
return -EINVAL;
}
- of_node_put(t);
OK you moved 't' inside the loop and this must be taken care, but...
}
i++;
} while (t);
....now, will it even compile if 't' is not in scope ? I think you might get
compilation here. If not, I still don't understand what is the value of
't' being checked there.
@@ -586,7 +583,6 @@ static int __init parse_cluster(struct device_node *cluster, int package_id,
char name[20];
bool leaf = true;
bool has_cores = false;
- struct device_node *c;
int core_id = 0;
int i, ret;
@@ -598,13 +594,13 @@ static int __init parse_cluster(struct device_node *cluster, int package_id,
i = 0;
do {
snprintf(name, sizeof(name), "cluster%d", i);
- c = of_get_child_by_name(cluster, name);
+ struct device_node *c __free(device_node) =
+ of_get_child_by_name(cluster, name);
if (c) {
leaf = false;
ret = parse_cluster(c, package_id, i, depth + 1);
if (depth > 0)
pr_warn("Topology for clusters of clusters not yet supported\n");
- of_node_put(c);
if (ret != 0)
return ret;
}
@@ -615,14 +611,14 @@ static int __init parse_cluster(struct device_node *cluster, int package_id,
i = 0;
do {
snprintf(name, sizeof(name), "core%d", i);
- c = of_get_child_by_name(cluster, name);
+ struct device_node *c __free(device_node) =
+ of_get_child_by_name(cluster, name);
if (c) {
has_cores = true;
if (depth == 0) {
pr_err("%pOF: cpu-map children should be clusters\n",
c);
- of_node_put(c);
return -EINVAL;
}
@@ -635,7 +631,6 @@ static int __init parse_cluster(struct device_node *cluster, int package_id,
ret = -EINVAL;
}
- of_node_put(c);
if (ret != 0)
return ret;
}
@@ -651,17 +646,16 @@ static int __init parse_cluster(struct device_node *cluster, int package_id,
static int __init parse_socket(struct device_node *socket)
{
char name[20];
- struct device_node *c;
bool has_socket = false;
int package_id = 0, ret;
do {
snprintf(name, sizeof(name), "socket%d", package_id);
- c = of_get_child_by_name(socket, name);
+ struct device_node *c __free(device_node) =
+ of_get_child_by_name(socket, name);
if (c) {
has_socket = true;
ret = parse_cluster(c, package_id, -1, 0);
- of_node_put(c);
if (ret != 0)
return ret;
}
Same thing applies to these while(c) loop. I don't understand how this
could work even if it is compiling fine which I doubt.
@@ -676,11 +670,11 @@ static int __init parse_socket(struct device_node *socket)
static int __init parse_dt_topology(void)
{
- struct device_node *cn, *map;
int ret = 0;
int cpu;
- cn = of_find_node_by_path("/cpus");
+ struct device_node *cn __free(device_node) =
+ of_find_node_by_path("/cpus");
if (!cn) {
pr_err("No CPU information found in DT\n");
return 0;
@@ -690,13 +684,14 @@ static int __init parse_dt_topology(void)
* When topology is provided cpu-map is essentially a root
* cluster with restricted subnodes.
*/
- map = of_get_child_by_name(cn, "cpu-map");
+ struct device_node *map __free(devide_node) =
If not above ones, this must fail to compile. Perhaps s/devide_node/device_node/ ?
I now doubt if this patch is compile tested ?
--
Regards,
Sudeep
Hi,
As you rightly pointed out, I inadvertently omitted to compile this file
during the kernel build process. Consequently, certain errors remained
undetected. I apologize for the oversight.
I'll send an updated version of this patch soon.
Regards,
Vincenzo