* [PATCH] ACPI: scan: fix bus ID cleanup on device_add() failure
@ 2026-08-04 15:48 Hongyan Xu
2026-08-07 17:38 ` Rafael J. Wysocki (Intel)
2026-08-08 8:59 ` [PATCH v2] " Hongyan Xu
0 siblings, 2 replies; 3+ messages in thread
From: Hongyan Xu @ 2026-08-04 15:48 UTC (permalink / raw)
To: Rafael J . Wysocki; +Cc: Len Brown, linux-acpi, jianhao.xu, Hongyan Xu
When device_add() fails after acpi_device_set_name() has allocated an
instance ID and linked a new acpi_device_bus_id into
acpi_bus_id_list, the rollback path only removes wakeup_list and
detaches the ACPI handle data.
That leaves the bus-ID bookkeeping behind and keeps the allocated
instance number consumed.
Factor the bus-ID removal into a helper, use it from both the normal
device teardown path and the device_add() rollback path, and only drop
wakeup_list when it was actually linked.
Found by manual review of reports from the
kernel70rc2-fail11-retry-20260801 run.
Signed-off-by: Hongyan Xu <getshell@seu.edu.cn>
---
drivers/acpi/scan.c | 18 +++++++++++++-----
1 file changed, 13 insertions(+), 5 deletions(-)
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index ff7000b71..852682436 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -520,12 +520,10 @@ static void acpi_device_release(struct device *dev)
kfree(acpi_dev);
}
-static void acpi_device_del(struct acpi_device *device)
+static void acpi_device_remove_bus_id(struct acpi_device *device)
{
struct acpi_device_bus_id *acpi_device_bus_id;
- mutex_lock(&acpi_device_lock);
-
list_for_each_entry(acpi_device_bus_id, &acpi_bus_id_list, node)
if (!strcmp(acpi_device_bus_id->bus_id,
acpi_device_hid(device))) {
@@ -538,8 +536,16 @@ static void acpi_device_del(struct acpi_device *device)
}
break;
}
+}
- list_del(&device->wakeup_list);
+static void acpi_device_del(struct acpi_device *device)
+{
+ mutex_lock(&acpi_device_lock);
+
+ acpi_device_remove_bus_id(device);
+
+ if (device->wakeup.flags.valid)
+ list_del(&device->wakeup_list);
mutex_unlock(&acpi_device_lock);
@@ -800,7 +806,9 @@ int acpi_device_add(struct acpi_device *device)
err:
mutex_lock(&acpi_device_lock);
- list_del(&device->wakeup_list);
+ acpi_device_remove_bus_id(device);
+ if (device->wakeup.flags.valid)
+ list_del(&device->wakeup_list);
err_unlock:
mutex_unlock(&acpi_device_lock);
--
2.50.1.windows.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH] ACPI: scan: fix bus ID cleanup on device_add() failure
2026-08-04 15:48 [PATCH] ACPI: scan: fix bus ID cleanup on device_add() failure Hongyan Xu
@ 2026-08-07 17:38 ` Rafael J. Wysocki (Intel)
2026-08-08 8:59 ` [PATCH v2] " Hongyan Xu
1 sibling, 0 replies; 3+ messages in thread
From: Rafael J. Wysocki (Intel) @ 2026-08-07 17:38 UTC (permalink / raw)
To: Hongyan Xu; +Cc: Rafael J . Wysocki, Len Brown, linux-acpi, jianhao.xu
On Tue, Aug 4, 2026 at 5:49 PM Hongyan Xu <getshell@seu.edu.cn> wrote:
>
> When device_add() fails after acpi_device_set_name() has allocated an
> instance ID and linked a new acpi_device_bus_id into
> acpi_bus_id_list, the rollback path only removes wakeup_list and
> detaches the ACPI handle data.
>
> That leaves the bus-ID bookkeeping behind and keeps the allocated
> instance number consumed.
>
> Factor the bus-ID removal into a helper, use it from both the normal
> device teardown path and the device_add() rollback path, and only drop
> wakeup_list when it was actually linked.
>
> Found by manual review of reports from the
> kernel70rc2-fail11-retry-20260801 run.
>
> Signed-off-by: Hongyan Xu <getshell@seu.edu.cn>
> ---
> drivers/acpi/scan.c | 18 +++++++++++++-----
> 1 file changed, 13 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
> index ff7000b71..852682436 100644
> --- a/drivers/acpi/scan.c
> +++ b/drivers/acpi/scan.c
> @@ -520,12 +520,10 @@ static void acpi_device_release(struct device *dev)
> kfree(acpi_dev);
> }
>
> -static void acpi_device_del(struct acpi_device *device)
> +static void acpi_device_remove_bus_id(struct acpi_device *device)
> {
> struct acpi_device_bus_id *acpi_device_bus_id;
>
> - mutex_lock(&acpi_device_lock);
> -
> list_for_each_entry(acpi_device_bus_id, &acpi_bus_id_list, node)
> if (!strcmp(acpi_device_bus_id->bus_id,
> acpi_device_hid(device))) {
> @@ -538,8 +536,16 @@ static void acpi_device_del(struct acpi_device *device)
> }
> break;
> }
> +}
>
> - list_del(&device->wakeup_list);
> +static void acpi_device_del(struct acpi_device *device)
> +{
> + mutex_lock(&acpi_device_lock);
> +
> + acpi_device_remove_bus_id(device);
> +
> + if (device->wakeup.flags.valid)
> + list_del(&device->wakeup_list);
>
> mutex_unlock(&acpi_device_lock);
>
> @@ -800,7 +806,9 @@ int acpi_device_add(struct acpi_device *device)
> err:
> mutex_lock(&acpi_device_lock);
>
> - list_del(&device->wakeup_list);
> + acpi_device_remove_bus_id(device);
> + if (device->wakeup.flags.valid)
> + list_del(&device->wakeup_list);
You have exactly the same code sequence in two places, so any chance
to avoid that duplication?
Also, it is not necessary to check device->wakeup.flags.valid before
doing list_del(&device->wakeup_list) if I'm not mistaken.
>
> err_unlock:
> mutex_unlock(&acpi_device_lock);
> --
^ permalink raw reply [flat|nested] 3+ messages in thread
* [PATCH v2] ACPI: scan: fix bus ID cleanup on device_add() failure
2026-08-04 15:48 [PATCH] ACPI: scan: fix bus ID cleanup on device_add() failure Hongyan Xu
2026-08-07 17:38 ` Rafael J. Wysocki (Intel)
@ 2026-08-08 8:59 ` Hongyan Xu
1 sibling, 0 replies; 3+ messages in thread
From: Hongyan Xu @ 2026-08-08 8:59 UTC (permalink / raw)
To: rafael; +Cc: Len Brown, linux-acpi, jianhao.xu, Hongyan Xu
When device_add() fails after acpi_device_set_name() has allocated an
instance ID and linked a new acpi_device_bus_id into acpi_bus_id_list, the
rollback path only removes wakeup_list and detaches the ACPI handle data.
That leaves the bus-ID bookkeeping behind and keeps the allocated instance
number consumed.
Move the bus-ID cleanup and wakeup-list removal into a single helper.
Use it from both the normal device teardown path and the device_add()
rollback path. The wakeup list node is initialized before registration, so
it can be deleted without checking whether the device is wakeup-capable, as
in the original teardown path.
Found by manual review of reports from the
kernel70rc2-fail11-retry-20260801 run.
Signed-off-by: Hongyan Xu <getshell@seu.edu.cn>
---
v2:
- Fold the bus-ID cleanup and wakeup-list removal into one helper to avoid
duplicating the same sequence in acpi_device_del() and the device_add()
rollback path.
- Drop the wakeup.flags.valid check before list_del(&device->wakeup_list).
v1: https://lore.kernel.org/linux-acpi/20260804154854.1555-1-getshell@seu.edu.cn/
drivers/acpi/scan.c | 13 +++++++++----
1 file changed, 9 insertions(+), 4 deletions(-)
diff --git a/drivers/acpi/scan.c b/drivers/acpi/scan.c
index ff7000b71fd1..a311af60d8d3 100644
--- a/drivers/acpi/scan.c
+++ b/drivers/acpi/scan.c
@@ -520,12 +520,10 @@ static void acpi_device_release(struct device *dev)
kfree(acpi_dev);
}
-static void acpi_device_del(struct acpi_device *device)
+static void acpi_device_del_list(struct acpi_device *device)
{
struct acpi_device_bus_id *acpi_device_bus_id;
- mutex_lock(&acpi_device_lock);
-
list_for_each_entry(acpi_device_bus_id, &acpi_bus_id_list, node)
if (!strcmp(acpi_device_bus_id->bus_id,
acpi_device_hid(device))) {
@@ -540,6 +538,13 @@ static void acpi_device_del(struct acpi_device *device)
}
list_del(&device->wakeup_list);
+}
+
+static void acpi_device_del(struct acpi_device *device)
+{
+ mutex_lock(&acpi_device_lock);
+
+ acpi_device_del_list(device);
mutex_unlock(&acpi_device_lock);
@@ -800,7 +805,7 @@ int acpi_device_add(struct acpi_device *device)
err:
mutex_lock(&acpi_device_lock);
- list_del(&device->wakeup_list);
+ acpi_device_del_list(device);
err_unlock:
mutex_unlock(&acpi_device_lock);
--
2.50.1.windows.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-08-08 8:59 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-08-04 15:48 [PATCH] ACPI: scan: fix bus ID cleanup on device_add() failure Hongyan Xu
2026-08-07 17:38 ` Rafael J. Wysocki (Intel)
2026-08-08 8:59 ` [PATCH v2] " Hongyan Xu
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox