Govind Singh <govinds@xxxxxxxxxxxxxx> writes:
> From: Rakesh Pillai <pillair@xxxxxxxxxxxxxx>
>
> HL2.0 firmware does not support setting quiet mode.
> If the host driver sends the quiet mode setting
> command to the HL2.0 firmware, it crashes with the
> below signature.
>
> fatal error received: err_qdi.c:456:EX:wlan_process:1:WLAN RT:207a:PC=b001b4f0
>
> The quiet mode command support is exposed by the firmware
> via thermal throttle wmi service. Enable ath10k thermal
> support if thermal throttle wmi service bit is set.
> 10X firmware version supports this feature by default, hence
> set this service bit by default.
>
> Tested HW: WCN3990
> Tested FW: WLAN.HL.2.0-01188-QCAHLSWMTPLZ-1
So this would also need testing with QCA988X or QCA9884 family of
devices, I don't think this patch would have worked as is.
> Co-developed-by: Govind Singh <govinds@xxxxxxxxxxxxxx>
> Signed-off-by: Rakesh Pillai <pillair@xxxxxxxxxxxxxx>
> Signed-off-by: Govind Singh <govinds@xxxxxxxxxxxxxx>
I had few comments below which I fixed myself. I'll send v2 soon.
> @@ -2621,6 +2621,10 @@ int ath10k_core_start(struct ath10k *ar, enum ath10k_firmware_mode mode,
> goto err_hif_stop;
> }
>
> + if (test_bit(ATH10K_FW_FEATURE_WMI_10X,
> + ar->normal_mode_fw.fw_file.fw_features)) {
> + set_bit(WMI_SERVICE_THERM_THROT, ar->wmi.svc_map);
> + }
I don't think testing for ATH10K_FW_FEATURE_WMI_10X works, that's a
really old and deprecated flag. wmi_op_version should be more reliable.
Also I refactored this to it's own function.
> /* Some firmware revisions do not properly set up hardware rx filter
> * registers.
> *
> @@ -2919,11 +2923,13 @@ static void ath10k_core_register_work(struct work_struct *work)
> goto err_debug_destroy;
> }
>
> - status = ath10k_thermal_register(ar);
> - if (status) {
> - ath10k_err(ar, "could not register thermal device: %d\n",
> - status);
> - goto err_spectral_destroy;
> + if (test_bit(WMI_SERVICE_THERM_THROT, ar->wmi.svc_map)) {
> + status = ath10k_thermal_register(ar);
> + if (status) {
> + ath10k_err(ar, "could not register thermal device: %d\n",
> + status);
> + goto err_spectral_destroy;
> + }
> }
The code becomes cleaner if the test is in thermal.c so I moved it
there.
> @@ -2964,7 +2970,8 @@ void ath10k_core_unregister(struct ath10k *ar)
> if (!test_bit(ATH10K_FLAG_CORE_REGISTERED, &ar->dev_flags))
> return;
>
> - ath10k_thermal_unregister(ar);
> + if (test_bit(WMI_SERVICE_THERM_THROT, ar->wmi.svc_map))
> + ath10k_thermal_unregister(ar);
Same here.
> --- a/drivers/net/wireless/ath/ath10k/mac.c
> +++ b/drivers/net/wireless/ath/ath10k/mac.c
> @@ -4800,7 +4800,8 @@ static int ath10k_start(struct ieee80211_hw *hw)
> ath10k_regd_update(ar);
>
> ath10k_spectral_start(ar);
> - ath10k_thermal_set_throttling(ar);
> + if (test_bit(WMI_SERVICE_THERM_THROT, ar->wmi.svc_map))
> + ath10k_thermal_set_throttling(ar);
And here.
--
Kalle Valo
