Re: [PATCH] firmware: qcom: tzmem: use sleepable variant of DO_ONCE()

[Bartosz Golaszewski]

On Sat, 5 Sep 2026 05:04:41 +0200, Daniel J Blueman <daniel@xxxxxxxxx> said:
> qcom_tzmem_enable() runs qcom_tzmem_init() using DO_ONCE(), taking
> the once_lock spinlock. qcom_tzmem_init() subsequently takes the
> qcom_scm_lock mutex, potentially sleeping holding the spinlock,
> triggering:
>
>   BUG: sleeping function called from invalid context at drivers/firmware/qcom/qcom_scm.c:334
>   in_atomic(): 1, irqs_disabled(): 1, non_block: 0, pid: 112, name: kworker/u49:4
>   Call trace:
>    __might_resched+0x378/0x698
>    __might_sleep+0x9c/0x220
>    qcom_scm_call+0x3c/0xf0
>    __qcom_scm_is_call_available+0xec/0x198
>    qcom_scm_shm_bridge_enable+0xc4/0x158
>    qcom_tzmem_enable+0x178/0x218
>    qcom_scm_probe+0x4d4/0xcb8
>
> Fix this by using DO_ONCE_SLEEPABLE().
>
> Fixes: 9941fe8a04f3 ("firmware: qcom: scm: Fix tzmem state on probe retry")
> Signed-off-by: Daniel J Blueman <daniel@xxxxxxxxx>
> ---
>  drivers/firmware/qcom/qcom_tzmem.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/firmware/qcom/qcom_tzmem.c b/drivers/firmware/qcom/qcom_tzmem.c
> index 0fd9581275f1..ebcdee3c5975 100644
> --- a/drivers/firmware/qcom/qcom_tzmem.c
> +++ b/drivers/firmware/qcom/qcom_tzmem.c
> @@ -518,7 +518,7 @@ int qcom_tzmem_enable(struct device *dev)
>  	static int result;
>
>  	qcom_tzmem_dev = dev;
> -	DO_ONCE(qcom_tzmem_do_init, &result);
> +	DO_ONCE_SLEEPABLE(qcom_tzmem_do_init, &result);
>  	return result;
>  }
>  EXPORT_SYMBOL_GPL(qcom_tzmem_enable);
> --
> 2.53.0
>
>

Mukesh already sent the same change[1].

Bart

[1] https://lore.kernel.org/all/20260806112709.4018596-1-mukesh.ojha@xxxxxxxxxxxxxxxx/