Re: [PATCH] wifi: ath10k: Stop unanswered temperature reads

[Loic Poulain]

On Mon, Oct 26, 2026 at 5:36 PM Filippo Falezza via B4 Relay
<devnull+effeffe99.gmail.com@xxxxxxxxxx> wrote:
>
> From: Filippo Falezza <effeffe99@xxxxxxxxx>
>
> Reading the ath10k hwmon temperature sends WMI_PDEV_GET_TEMPERATURE and
> waits up to 5 s (ATH10K_THERMAL_SYNC_TIMEOUT_HZ) for the reply event,
> holding conf_mutex. Not all firmware answers: the WCN3990 firmware
> WLAN.HL.3.3.1-00694-QCAHLSWMTPLZ-1 on a Motorola Moto G 5G Plus
> codename nairo (SM7250) never sends the event, so every read of
> temp1_input stalls for 5 s, logs
>
>   ath10k_snoc 18800000.wifi: failed to synchronize thermal read
>
> and returns -ETIMEDOUT. Tools that read all hwmon sensors, such as
> lm-sensors, hang for 5 s on every run, and configuration changes of the
> interface wait for conf_mutex in the meantime.
>
> Whether the firmware answers cannot be told from the chip: firmware that
> answers and firmware that does not both advertise
> WMI_SERVICE_THERM_THROT, and WCN3990 firmware differs between devices.
> Instead, remember that the running firmware did not answer, and return
> -EOPNOTSUPP right away on later reads. Forget it when the firmware
> starts, so that a restarted or updated firmware is asked again.
>
> The first read after each firmware start still takes 5 s when the
> firmware does not answer. Reads on firmware that answers are unchanged.
>
> This change was written with the help of an AI coding assistant; see
> the notes below.
>
> Fixes: e98199a8c27f ("ath10k: enable QCA6174/QCA9377 to read the chip temperature")
> Signed-off-by: Filippo Falezza <effeffe99@xxxxxxxxx>
> ---
> Notes, not for the changelog:
>
> - Tool: Claude Opus 5.5 in Claude Code wrote the code and this changelog
>   during an interactive bring-up session for the Moto G 5G Plus. The
>   request was to fix `sensors` hanging on every run because of ath10k.
>   A first version disabled the read for all WCN3990 through a hw_params
>   flag; it was replaced by this one so that it does not assume all
>   WCN3990 firmware behaves the same, as Loic Poulain asked of another
>   WCN3990 firmware workaround ("ath10k: WCN3990 firmware workarounds for
>   WLAN.HL.3.2", March 2026).
> - Tested on: Moto G 5G Plus (SM7250, WCN3990, firmware
>   WLAN.HL.3.3.1-00694-QCAHLSWMTPLZ-1), linux-next next-20261021 plus
>   device support patches, with this patch in place of a local WCN3990
>   quirk. First read of temp1_input: 5065 ms, -ETIMEDOUT, one warning;
>   next reads: 8 ms, -EOPNOTSUPP; a full `sensors` run: 25 ms (5 s or
>   more before). After switching Wi-Fi off and on, which restarts the
>   firmware session, the first read waits 5 s again, then -EOPNOTSUPP.
> - Not tested: firmware that does answer the request (QCA6174, QCA9377,
>   other WCN3990 firmware); for them only reads after a timeout change.
>   Not run: sparse.
> - Build: W=1 for arm64, no warnings; checkpatch --strict clean.
>   Applies to ath-next (cd026e68f).
> ---
>  drivers/net/wireless/ath/ath10k/core.c    | 1 +
>  drivers/net/wireless/ath/ath10k/thermal.c | 9 ++++++++-
>  drivers/net/wireless/ath/ath10k/thermal.h | 2 ++
>  3 files changed, 11 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/net/wireless/ath/ath10k/core.c b/drivers/net/wireless/ath/ath10k/core.c
> index 7c2939cbde5f..9467e4c3099a 100644
> --- a/drivers/net/wireless/ath/ath10k/core.c
> +++ b/drivers/net/wireless/ath/ath10k/core.c
> @@ -2985,6 +2985,7 @@ int ath10k_core_start(struct ath10k *ar, enum ath10k_firmware_mode mode,
>         clear_bit(ATH10K_FLAG_CRASH_FLUSH, &ar->dev_flags);
>
>         ar->running_fw = fw;
> +       ar->thermal.temperature_unanswered = false;
>
>         if (!test_bit(ATH10K_FW_FEATURE_NON_BMI,
>                       ar->running_fw->fw_file.fw_features)) {
> diff --git a/drivers/net/wireless/ath/ath10k/thermal.c b/drivers/net/wireless/ath/ath10k/thermal.c
> index 9bb486f8e920..1079d552cff4 100644
> --- a/drivers/net/wireless/ath/ath10k/thermal.c
> +++ b/drivers/net/wireless/ath/ath10k/thermal.c
> @@ -75,6 +75,12 @@ static ssize_t ath10k_thermal_show_temp(struct device *dev,
>                 goto out;
>         }
>
> +       /* Don't stall every read on firmware that never answers */
> +       if (ar->thermal.temperature_unanswered) {
> +               ret = -EOPNOTSUPP;
> +               goto out;
> +       }
> +
>         reinit_completion(&ar->thermal.wmi_sync);
>         ret = ath10k_wmi_pdev_get_temperature(ar);
>         if (ret) {
> @@ -90,7 +96,8 @@ static ssize_t ath10k_thermal_show_temp(struct device *dev,
>         time_left = wait_for_completion_timeout(&ar->thermal.wmi_sync,
>                                                 ATH10K_THERMAL_SYNC_TIMEOUT_HZ);
>         if (!time_left) {
> -               ath10k_warn(ar, "failed to synchronize thermal read\n");
> +               ath10k_warn(ar, "failed to synchronize thermal read, not asking again until the firmware restarts\n");
> +               ar->thermal.temperature_unanswered = true;
>                 ret = -ETIMEDOUT;
>                 goto out;
>         }
> diff --git a/drivers/net/wireless/ath/ath10k/thermal.h b/drivers/net/wireless/ath/ath10k/thermal.h
> index 1f4de9fbf2b3..cb930da5c04a 100644
> --- a/drivers/net/wireless/ath/ath10k/thermal.h
> +++ b/drivers/net/wireless/ath/ath10k/thermal.h
> @@ -19,6 +19,8 @@ struct ath10k_thermal {
>         /* protected by conf_mutex */
>         u32 throttle_state;
>         u32 quiet_period;
> +       /* the running firmware did not answer a temperature request */
> +       bool temperature_unanswered;

Maybe just "bool broken;" or alternatively a new
ATH10K_FLAG_TEMP_BROKEN dev_flag.

>         /* temperature value in Celsius degree
>          * protected by data_lock
>          */
>
> ---
> base-commit: cd026e68fdbf88b9cc522b76c7a9d92c9a41f709
> change-id: 20261022-ath10k-temperature-22b119fd615c
>
> Best regards,
> --
> Filippo Falezza <effeffe99@xxxxxxxxx>
>
>
>