[PATCH ath-next v3] wifi: ath12k: avoid dynamic alloc when parsing wmi tb

[Baochen Qiang]

On 3/16/2026 6:32 PM, Nicolas Escande wrote:
> On each WMI message received from the hardware, we alloc a temporary array
> of WMI_TAG_MAX entries of type void *. This array is then populated with
> pointers of parsed structs depending on the WMI type, and then freed. This
> alloc can fail when memory pressure in the system is high enough.
> 
> Given the fact that it is scheduled in softirq with the system_bh_wq, we
> should not be able to parse more than one WMI message per CPU at any time.
> 
> So instead lets move to a per cpu allocated array, that is reused across
> calls: ath12K_wmi_tb that lives in wmi.c of the ath12K module. To alloc &
> free we added two new module_init/exit functions for the module and two
> new wmi functions to alloc/free this memory.
> 
> ath12k_wmi_tlv_parse_alloc() and ath12k_wmi_tlv_parse() are merged
> together as it no longer allocs mem but returns the existing per-cpu one.
> 
> Signed-off-by: Nicolas Escande <nico.escande at gmail.com>
> ---
> changes from v2:
>   - removed now superfluous return in ath12k_wmi_event_teardown_complete()
>   - moved ath12k_wmi_tb declaration to wmi.c & added two functions to
>     alloc / free it
>   - removed useless error message on memory allocation failure
> 
> changes from v1:
>   - rebased on ath-next 27401c9b1432
>   - changed wording according to Jeff's comment
>   - moved alloc/cleanup to new module_init/exit functions in the
>     ath12k module as per Baochen's comment
> 
> changes from RFC:
>   - rebased on ath-next 8e0ab5b9adb7
>   - converted missing call sites ath12k_wmi_obss_color_collision_event()
>     & ath12k_wmi_pdev_temperature_event()
>   - changed alloc order & cleanup path in ath12k_core_alloc() as it seems
>     it confused people
>   - used sizeof(*tb) in ath12k_wmi_tlv_parse()
> ---
>  drivers/net/wireless/ath/ath12k/core.c |  19 +++
>  drivers/net/wireless/ath/ath12k/core.h |   1 +
>  drivers/net/wireless/ath/ath12k/wmi.c  | 200 ++++++++-----------------
>  drivers/net/wireless/ath/ath12k/wmi.h  |   3 +
>  4 files changed, 84 insertions(+), 139 deletions(-)
> 
> diff --git a/drivers/net/wireless/ath/ath12k/core.c b/drivers/net/wireless/ath/ath12k/core.c
> index c31c47fb5a73..16815a27f8f2 100644
> --- a/drivers/net/wireless/ath/ath12k/core.c
> +++ b/drivers/net/wireless/ath/ath12k/core.c
> @@ -2321,5 +2321,24 @@ struct ath12k_base *ath12k_core_alloc(struct device *dev, size_t priv_size,
>  	return NULL;
>  }
>  
> +static int ath12k_init(void)
> +{
> +	int ret;
> +
> +	ret = ath12k_wmi_alloc();
> +	if (ret)
> +		return ret;
> +
> +	return 0;

nit: you can simply

	return ath12k_wmi_alloc();

> +}
> +
> +static void ath12k_exit(void)
> +{
> +	ath12k_wmi_free();
> +}
> +
> +module_init(ath12k_init);
> +module_exit(ath12k_exit);
> +
>  MODULE_DESCRIPTION("Driver support for Qualcomm Technologies WLAN devices");
>  MODULE_LICENSE("Dual BSD/GPL");
> diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
> index 59c193b24764..02ee6c718621 100644
> --- a/drivers/net/wireless/ath/ath12k/core.h
> +++ b/drivers/net/wireless/ath/ath12k/core.h
> @@ -19,6 +19,7 @@
>  #include <linux/average.h>
>  #include <linux/of.h>
>  #include <linux/rhashtable.h>
> +#include <linux/percpu.h>

does it make more sense to add this to wmi.c instead?