Skip to contentSkip to conversation

ath11k: QCA6390 stability fixes

PATCH In review Kalle Valo 2 participants 12 messages Dec 10, 2020 – Dec 12, 2020 Qualcomm Atheros ATH11K wireless driver

lore
Share & tools
lore

Kalle Valo

author
#0 lore raw

cover ath11k: QCA6390 stability fixes

Few fixes to improve statibility on QCA6390. Please review.

Kalle

Carl Huang (4):
ath11k: put hw to DBS using WMI_PDEV_SET_HW_MODE_CMDID
ath11k: pci: fix hot reset stability issues
ath11k: pci: fix L1ss clock unstable problem
ath11k: pci: disable VDD4BLOW

Kalle Valo (1):
ath11k: mhi: print a warning if firmware crashed

Pradeep Kumar Chitrapu (1):
ath11k: Fix incorrect tlvs in scan start command

 drivers/net/wireless/ath/ath11k/core.c |  12 ++-
 drivers/net/wireless/ath/ath11k/hw.h   |   6 --
 drivers/net/wireless/ath/ath11k/mhi.c  |   9 +++
 drivers/net/wireless/ath/ath11k/pci.c  | 132 +++++++++++++++++++++++++++++++--
 drivers/net/wireless/ath/ath11k/pci.h  |  24 ++++++
 drivers/net/wireless/ath/ath11k/wmi.c  |  69 +++++++++--------
 drivers/net/wireless/ath/ath11k/wmi.h  |   3 +
 7 files changed, 206 insertions(+), 49 deletions(-)
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#1 lore raw

PATCH 1/6 ath11k: mhi: print a warning if firmware crashed

There was no way to detect if the firmware crashed so add a warning. At the moment the firmware is not restarted or anything like that, so when this happens ath11k modules need to be reloaded.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

---
 drivers/net/wireless/ath/ath11k/mhi.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/net/wireless/ath/ath11k/mhi.c +9 -0
index 47a1ce1bee4f..4cf093718888 100644
@@ -190,6 +190,15 @@ static void ath11k_mhi_op_runtime_put(struct mhi_controller *mhi_cntrl)
 static void ath11k_mhi_op_status_cb(struct mhi_controller *mhi_cntrl,
 				    enum mhi_callback cb)
 {
+	struct ath11k_base *ab = dev_get_drvdata(mhi_cntrl->cntrl_dev);
+
+	switch (cb) {
+	case MHI_CB_SYS_ERROR:
+		ath11k_warn(ab, "firmware crashed: MHI_CB_SYS_ERROR\n");
+		break;
+	default:
+		break;
+	}
 }
 
 static int ath11k_mhi_op_read_reg(struct mhi_controller *mhi_cntrl,
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#2 lore raw

Re: [PATCH 1/6] ath11k: mhi: print a warning if firmware crashed

Quoted 7 lines · Kalle Valo

There was no way to detect if the firmware crashed so add a warning. At the moment the firmware is not restarted or anything like that, so when this happens ath11k modules need to be reloaded.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

6 patches applied to ath-next branch of ath.git, thanks.

fc46e1b2a24a ath11k: mhi: print a warning if firmware crashed 43ed15e1ee01 ath11k: put hw to DBS using WMI_PDEV_SET_HW_MODE_CMDID babb0ced6acd ath11k: pci: fix hot reset stability issues 0699940755e9 ath11k: pci: fix L1ss clock unstable problem 0ccdf4398827 ath11k: pci: disable VDD4BLOW
f57ad6a9885e ath11k: Fix incorrect tlvs in scan start command

Signature · 9 lines
https://patchwork.kernel.org/project/linux-wireless/patch/[email protected]/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#3 lore raw

Re: [PATCH 1/6] ath11k: mhi: print a warning if firmware crashed

In reply to Kalle Valo
Quoted 7 lines · Kalle Valo

There was no way to detect if the firmware crashed so add a warning. At the moment the firmware is not restarted or anything like that, so when this happens ath11k modules need to be reloaded.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

6 patches applied to ath-next branch of ath.git, thanks.

fc46e1b2a24a ath11k: mhi: print a warning if firmware crashed 43ed15e1ee01 ath11k: put hw to DBS using WMI_PDEV_SET_HW_MODE_CMDID babb0ced6acd ath11k: pci: fix hot reset stability issues 0699940755e9 ath11k: pci: fix L1ss clock unstable problem 0ccdf4398827 ath11k: pci: disable VDD4BLOW
f57ad6a9885e ath11k: Fix incorrect tlvs in scan start command

Signature · 3 lines
https://patchwork.kernel.org/project/linux-wireless/patch/[email protected]/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
author
#4 lore raw

PATCH 2/6 ath11k: put hw to DBS using WMI_PDEV_SET_HW_MODE_CMDID

In reply to Kalle Valo

From: Carl Huang <[email protected]>

It's recommended to use wmi command WMI_PDEV_SET_HW_MODE_CMDID to put hardware to dbs mode instead of wmi_init command. This fixes a few strange stability issues.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

---
 drivers/net/wireless/ath/ath11k/core.c | 12 ++++++++---
 drivers/net/wireless/ath/ath11k/hw.h   |  6 ------
 drivers/net/wireless/ath/ath11k/wmi.c  | 38 ++++++++++++++++++++++++++++++----
 drivers/net/wireless/ath/ath11k/wmi.h  |  3 +++
 4 files changed, 46 insertions(+), 13 deletions(-)
drivers/net/wireless/ath/ath11k/core.c +9 -3
index da20ed752ca5..3c28e7ad7d51 100644
@@ -51,7 +51,6 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
 		.svc_to_ce_map = ath11k_target_service_to_ce_map_wlan_ipq8074,
 		.svc_to_ce_map_len = 21,
 		.single_pdev_only = false,
-		.needs_band_to_mac = true,
 		.rxdma1_enable = true,
 		.num_rxmda_per_pdev = 1,
 		.rx_mac_buf_ring = false,
@@ -89,7 +88,6 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
 		.svc_to_ce_map = ath11k_target_service_to_ce_map_wlan_ipq6018,
 		.svc_to_ce_map_len = 19,
 		.single_pdev_only = false,
-		.needs_band_to_mac = true,
 		.rxdma1_enable = true,
 		.num_rxmda_per_pdev = 1,
 		.rx_mac_buf_ring = false,
@@ -127,7 +125,6 @@ static const struct ath11k_hw_params ath11k_hw_params[] = {
 		.svc_to_ce_map = ath11k_target_service_to_ce_map_wlan_qca6390,
 		.svc_to_ce_map_len = 14,
 		.single_pdev_only = true,
-		.needs_band_to_mac = false,
 		.rxdma1_enable = false,
 		.num_rxmda_per_pdev = 2,
 		.rx_mac_buf_ring = true,
@@ -646,6 +643,15 @@ static int ath11k_core_start(struct ath11k_base *ab,
 		goto err_reo_cleanup;
 	}
 
+	/* put hardware to DBS mode */
+	if (ab->hw_params.single_pdev_only) {
+		ret = ath11k_wmi_set_hw_mode(ab, WMI_HOST_HW_MODE_DBS);
+		if (ret) {
+			ath11k_err(ab, "failed to send dbs mode: %d\n", ret);
+			goto err_hif_stop;
+		}
+	}
+
 	ret = ath11k_dp_tx_htt_h2t_ver_req_msg(ab);
 	if (ret) {
 		ath11k_err(ab, "failed to send htt version request message: %d\n",
drivers/net/wireless/ath/ath11k/hw.h +0 -6
index 17c4560b1c73..f33a458a465f 100644
@@ -143,12 +143,6 @@ struct ath11k_hw_params {
 
 	bool single_pdev_only;
 
-	/* For example on QCA6390 struct
-	 * wmi_init_cmd_param::band_to_mac_config needs to be false as the
-	 * firmware creates the mapping.
-	 */
-	bool needs_band_to_mac;
-
 	bool rxdma1_enable;
 	int num_rxmda_per_pdev;
 	bool rx_mac_buf_ring;
drivers/net/wireless/ath/ath11k/wmi.c +34 -4
index 59e179d86f38..79b7f991c612 100644
@@ -3479,6 +3479,35 @@ int ath11k_wmi_wait_for_unified_ready(struct ath11k_base *ab)
 	return 0;
 }
 
+int ath11k_wmi_set_hw_mode(struct ath11k_base *ab,
+			   enum wmi_host_hw_mode_config_type mode)
+{
+	struct wmi_pdev_set_hw_mode_cmd_param *cmd;
+	struct sk_buff *skb;
+	struct ath11k_wmi_base *wmi_ab = &ab->wmi_ab;
+	int len;
+	int ret;
+
+	len = sizeof(*cmd);
+
+	skb = ath11k_wmi_alloc_skb(wmi_ab, len);
+	cmd = (struct wmi_pdev_set_hw_mode_cmd_param *)skb->data;
+
+	cmd->tlv_header = FIELD_PREP(WMI_TLV_TAG, WMI_TAG_PDEV_SET_HW_MODE_CMD) |
+			  FIELD_PREP(WMI_TLV_LEN, sizeof(*cmd) - TLV_HDR_SIZE);
+
+	cmd->pdev_id = WMI_PDEV_ID_SOC;
+	cmd->hw_mode_index = mode;
+
+	ret = ath11k_wmi_cmd_send(&wmi_ab->wmi[0], skb, WMI_PDEV_SET_HW_MODE_CMDID);
+	if (ret) {
+		ath11k_warn(ab, "failed to send WMI_PDEV_SET_HW_MODE_CMDID\n");
+		dev_kfree_skb(skb);
+	}
+
+	return ret;
+}
+
 int ath11k_wmi_cmd_init(struct ath11k_base *ab)
 {
 	struct ath11k_wmi_base *wmi_sc = &ab->wmi_ab;
@@ -3497,10 +3526,11 @@ int ath11k_wmi_cmd_init(struct ath11k_base *ab)
 	init_param.hw_mode_id = wmi_sc->preferred_hw_mode;
 	init_param.mem_chunks = wmi_sc->mem_chunks;
 
-	if (ab->hw_params.needs_band_to_mac) {
-		init_param.num_band_to_mac = ab->num_radios;
-		ath11k_fill_band_to_mac_param(ab, init_param.band_to_mac);
-	}
+	if (ab->hw_params.single_pdev_only)
+		init_param.hw_mode_id = WMI_HOST_HW_MODE_MAX;
+
+	init_param.num_band_to_mac = ab->num_radios;
+	ath11k_fill_band_to_mac_param(ab, init_param.band_to_mac);
 
 	return ath11k_init_cmd_send(&wmi_sc->wmi[0], &init_param);
 }
drivers/net/wireless/ath/ath11k/wmi.h +3 -0
index 05d5f6d30072..62840717a946 100644
@@ -5171,4 +5171,7 @@ int ath11k_wmi_fils_discovery(struct ath11k *ar, u32 vdev_id, u32 interval,
 			      bool unsol_bcast_probe_resp_enabled);
 int ath11k_wmi_probe_resp_tmpl(struct ath11k *ar, u32 vdev_id,
 			       struct sk_buff *tmpl);
+int ath11k_wmi_set_hw_mode(struct ath11k_base *ab,
+			   enum wmi_host_hw_mode_config_type mode);
+
 #endif
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#5 lore raw

PATCH 3/6 ath11k: pci: fix hot reset stability issues

In reply to Kalle Valo

From: Carl Huang <[email protected]>

For QCA6390, host needs to reset some registers before MHI power up to fix PCI link unstable issue if hot reset happened. Also clear all pending interrupts during power up.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

---
 drivers/net/wireless/ath/ath11k/pci.c | 52 +++++++++++++++++++++++++++++++----
 drivers/net/wireless/ath/ath11k/pci.h | 10 +++++++
 2 files changed, 57 insertions(+), 5 deletions(-)
drivers/net/wireless/ath/ath11k/pci.c +47 -5
index c9ab1e9fc0fa..9b6d4bb40e6c 100644
@@ -239,15 +239,57 @@ static void ath11k_pci_clear_dbg_registers(struct ath11k_base *ab)
 	ath11k_dbg(ab, ATH11K_DBG_PCI, "soc reset cause:%d\n", val);
 }
 
+static void ath11k_pci_enable_ltssm(struct ath11k_base *ab)
+{
+	u32 val;
+	int i;
+
+	val = ath11k_pci_read32(ab, PCIE_PCIE_PARF_LTSSM);
+
+	/* PCIE link seems very unstable after the Hot Reset*/
+	for (i = 0; val != PARM_LTSSM_VALUE && i < 5; i++) {
+		if (val == 0xffffffff)
+			mdelay(5);
+
+		ath11k_pci_write32(ab, PCIE_PCIE_PARF_LTSSM, PARM_LTSSM_VALUE);
+		val = ath11k_pci_read32(ab, PCIE_PCIE_PARF_LTSSM);
+	}
+
+	ath11k_dbg(ab, ATH11K_DBG_PCI, "pci ltssm 0x%x\n", val);
+
+	val = ath11k_pci_read32(ab, GCC_GCC_PCIE_HOT_RST);
+	val |= GCC_GCC_PCIE_HOT_RST_VAL | 0x10;
+	ath11k_pci_write32(ab, GCC_GCC_PCIE_HOT_RST, val);
+	val = ath11k_pci_read32(ab, GCC_GCC_PCIE_HOT_RST);
+
+	ath11k_dbg(ab, ATH11K_DBG_PCI, "pci pcie_hot_rst 0x%x\n", val);
+
+	mdelay(5);
+}
+
+static void ath11k_pci_clear_all_intrs(struct ath11k_base *ab)
+{
+	/* This is a WAR for PCIE Hotreset.
+	 * When target receive Hotreset, but will set the interrupt.
+	 * So when download SBL again, SBL will open Interrupt and
+	 * receive it, and crash immediately.
+	 */
+	ath11k_pci_write32(ab, PCIE_PCIE_INT_ALL_CLEAR, PCIE_INT_CLEAR_ALL);
+}
+
 static void ath11k_pci_force_wake(struct ath11k_base *ab)
 {
 	ath11k_pci_write32(ab, PCIE_SOC_WAKE_PCIE_LOCAL_REG, 1);
 	mdelay(5);
 }
 
-static void ath11k_pci_sw_reset(struct ath11k_base *ab)
+static void ath11k_pci_sw_reset(struct ath11k_base *ab, bool power_on)
 {
-	ath11k_pci_soc_global_reset(ab);
+	if (power_on) {
+		ath11k_pci_enable_ltssm(ab);
+		ath11k_pci_clear_all_intrs(ab);
+	}
+
 	ath11k_mhi_clear_vector(ab);
 	ath11k_pci_soc_global_reset(ab);
 	ath11k_mhi_set_mhictrl_reset(ab);
@@ -770,7 +812,7 @@ static int ath11k_pci_power_up(struct ath11k_base *ab)
 
 	ab_pci->register_window = 0;
 	clear_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags);
-	ath11k_pci_sw_reset(ab_pci->ab);
+	ath11k_pci_sw_reset(ab_pci->ab, true);
 
 	ret = ath11k_mhi_start(ab_pci);
 	if (ret) {
@@ -785,10 +827,10 @@ static void ath11k_pci_power_down(struct ath11k_base *ab)
 {
 	struct ath11k_pci *ab_pci = ath11k_pci_priv(ab);
 
+	ath11k_pci_force_wake(ab_pci->ab);
 	ath11k_mhi_stop(ab_pci);
 	clear_bit(ATH11K_PCI_FLAG_INIT_DONE, &ab_pci->flags);
-	ath11k_pci_force_wake(ab_pci->ab);
-	ath11k_pci_sw_reset(ab_pci->ab);
+	ath11k_pci_sw_reset(ab_pci->ab, false);
 }
 
 static void ath11k_pci_kill_tasklets(struct ath11k_base *ab)
drivers/net/wireless/ath/ath11k/pci.h +10 -0
index 9c503a17c237..296f1ca6b3ec 100644
@@ -24,6 +24,16 @@
 /* register used for handshake mechanism to validate UMAC is awake */
 #define PCIE_SOC_WAKE_PCIE_LOCAL_REG		0x3004
 
+#define PCIE_PCIE_PARF_LTSSM			0x1e081b0
+#define PARM_LTSSM_VALUE			0x111
+
+#define GCC_GCC_PCIE_HOT_RST			0x1e402bc
+#define GCC_GCC_PCIE_HOT_RST_VAL		0x10
+
+#define PCIE_PCIE_INT_ALL_CLEAR			0x1e08228
+#define PCIE_SMLH_REQ_RST_LINK_DOWN		0x2
+#define PCIE_INT_CLEAR_ALL			0xffffffff
+
 struct ath11k_msi_user {
 	char *name;
 	int num_vectors;
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#6 lore raw

PATCH 4/6 ath11k: pci: fix L1ss clock unstable problem

In reply to Kalle Valo

From: Carl Huang <[email protected]>

For QCA6390, one PCI related clock drifts sometimes, and it makes PCI link difficult to quit L1ss. Fix it by writing some registers which are known to fix the problem.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

---
 drivers/net/wireless/ath/ath11k/pci.c | 70 +++++++++++++++++++++++++++++++++++
 drivers/net/wireless/ath/ath11k/pci.h | 11 ++++++
 2 files changed, 81 insertions(+)
drivers/net/wireless/ath/ath11k/pci.c +70 -0
index 9b6d4bb40e6c..064c35052697 100644
@@ -239,6 +239,75 @@ static void ath11k_pci_clear_dbg_registers(struct ath11k_base *ab)
 	ath11k_dbg(ab, ATH11K_DBG_PCI, "soc reset cause:%d\n", val);
 }
 
+static int ath11k_pci_set_link_reg(struct ath11k_base *ab,
+				   u32 offset, u32 value, u32 mask)
+{
+	u32 v;
+	int i;
+
+	v = ath11k_pci_read32(ab, offset);
+	if ((v & mask) == value)
+		return 0;
+
+	for (i = 0; i < 10; i++) {
+		ath11k_pci_write32(ab, offset, (v & ~mask) | value);
+
+		v = ath11k_pci_read32(ab, offset);
+		if ((v & mask) == value)
+			return 0;
+
+		mdelay(2);
+	}
+
+	ath11k_warn(ab, "failed to set pcie link register 0x%08x: 0x%08x != 0x%08x\n",
+		    offset, v & mask, value);
+
+	return -ETIMEDOUT;
+}
+
+static int ath11k_pci_fix_l1ss(struct ath11k_base *ab)
+{
+	int ret;
+
+	ret = ath11k_pci_set_link_reg(ab,
+				      PCIE_QSERDES_COM_SYSCLK_EN_SEL_REG,
+				      PCIE_QSERDES_COM_SYSCLK_EN_SEL_VAL,
+				      PCIE_QSERDES_COM_SYSCLK_EN_SEL_MSK);
+	if (!ret) {
+		ath11k_warn(ab, "failed to set sysclk: %d\n", ret);
+		return ret;
+	}
+
+	ret = ath11k_pci_set_link_reg(ab,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG1_REG,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG1_VAL,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG_MSK);
+	if (!ret) {
+		ath11k_warn(ab, "failed to set dtct config1 error: %d\n", ret);
+		return ret;
+	}
+
+	ret = ath11k_pci_set_link_reg(ab,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG2_REG,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG2_VAL,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG_MSK);
+	if (!ret) {
+		ath11k_warn(ab, "failed to set dtct config2: %d\n", ret);
+		return ret;
+	}
+
+	ret = ath11k_pci_set_link_reg(ab,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG4_REG,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG4_VAL,
+				      PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG_MSK);
+	if (!ret) {
+		ath11k_warn(ab, "failed to set dtct config4: %d\n", ret);
+		return ret;
+	}
+
+	return 0;
+}
+
 static void ath11k_pci_enable_ltssm(struct ath11k_base *ab)
 {
 	u32 val;
@@ -288,6 +357,7 @@ static void ath11k_pci_sw_reset(struct ath11k_base *ab, bool power_on)
 	if (power_on) {
 		ath11k_pci_enable_ltssm(ab);
 		ath11k_pci_clear_all_intrs(ab);
+		ath11k_pci_fix_l1ss(ab);
 	}
 
 	ath11k_mhi_clear_vector(ab);
drivers/net/wireless/ath/ath11k/pci.h +11 -0
index 296f1ca6b3ec..6e7cc8904c3d 100644
@@ -34,6 +34,17 @@
 #define PCIE_SMLH_REQ_RST_LINK_DOWN		0x2
 #define PCIE_INT_CLEAR_ALL			0xffffffff
 
+#define PCIE_QSERDES_COM_SYSCLK_EN_SEL_REG	0x01e0c0ac
+#define PCIE_QSERDES_COM_SYSCLK_EN_SEL_VAL	0x10
+#define PCIE_QSERDES_COM_SYSCLK_EN_SEL_MSK	0xffffffff
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG1_REG	0x01e0c628
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG1_VAL	0x02
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG2_REG	0x01e0c62c
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG2_VAL	0x52
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG4_REG	0x01e0c634
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG4_VAL	0xff
+#define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG_MSK	0x000000ff
+
 struct ath11k_msi_user {
 	char *name;
 	int num_vectors;
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#7 lore raw

PATCH 5/6 ath11k: pci: disable VDD4BLOW

In reply to Kalle Valo

From: Carl Huang <[email protected]>

It's recommended to disable VDD4BLOW during initialisation.

Tested-on: QCA6390 hw2.0 PCI WLAN.HST.1.0.1-01740-QCAHSTSWPLZ_V2_TO_X86-1

---
 drivers/net/wireless/ath/ath11k/pci.c | 10 ++++++++++
 drivers/net/wireless/ath/ath11k/pci.h |  3 +++
 2 files changed, 13 insertions(+)
drivers/net/wireless/ath/ath11k/pci.c +10 -0
index 064c35052697..818e37c32a85 100644
@@ -346,6 +346,15 @@ static void ath11k_pci_clear_all_intrs(struct ath11k_base *ab)
 	ath11k_pci_write32(ab, PCIE_PCIE_INT_ALL_CLEAR, PCIE_INT_CLEAR_ALL);
 }
 
+static void ath11k_pci_set_wlaon_pwr_ctrl(struct ath11k_base *ab)
+{
+	u32 val;
+
+	val = ath11k_pci_read32(ab, WLAON_QFPROM_PWR_CTRL_REG);
+	val &= ~QFPROM_PWR_CTRL_VDD4BLOW_MASK;
+	ath11k_pci_write32(ab, WLAON_QFPROM_PWR_CTRL_REG, val);
+}
+
 static void ath11k_pci_force_wake(struct ath11k_base *ab)
 {
 	ath11k_pci_write32(ab, PCIE_SOC_WAKE_PCIE_LOCAL_REG, 1);
@@ -357,6 +366,7 @@ static void ath11k_pci_sw_reset(struct ath11k_base *ab, bool power_on)
 	if (power_on) {
 		ath11k_pci_enable_ltssm(ab);
 		ath11k_pci_clear_all_intrs(ab);
+		ath11k_pci_set_wlaon_pwr_ctrl(ab);
 		ath11k_pci_fix_l1ss(ab);
 	}
 
drivers/net/wireless/ath/ath11k/pci.h +3 -0
index 6e7cc8904c3d..0432a702416b 100644
@@ -45,6 +45,9 @@
 #define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG4_VAL	0xff
 #define PCIE_USB3_PCS_MISC_OSC_DTCT_CONFIG_MSK	0x000000ff
 
+#define WLAON_QFPROM_PWR_CTRL_REG		0x01f8031c
+#define QFPROM_PWR_CTRL_VDD4BLOW_MASK		0x4
+
 struct ath11k_msi_user {
 	char *name;
 	int num_vectors;
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
#8 lore raw

Re: [PATCH 5/6] ath11k: pci: disable VDD4BLOW

Quoted 3 lines · Kalle Valo

From: Carl Huang <[email protected]>

It's recommended to disable VDD4BLOW during initialisation.

Can you explain more about "recommended by whom and why"?

Thanks,

Peter

Signature · 3 lines
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#9 lore raw

Re: [PATCH 5/6] ath11k: pci: disable VDD4BLOW

Quoted 6 lines · Peter Oh
Quoted 3 lines · Kalle Valo

From: Carl Huang <[email protected]>

It's recommended to disable VDD4BLOW during initialisation.

Can you explain more about "recommended by whom and why"?

It's coming from the internal teams, I understood the internal driver does the same.

Signature · 8 lines
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#10 lore raw

PATCH 6/6 ath11k: Fix incorrect tlvs in scan start command

In reply to Kalle Valo

From: Pradeep Kumar Chitrapu <[email protected]>

Currently 6G specific tlvs have duplicate entries which is causing scan failures. Fix this by removing the duplicate entries of the same tlv. This also fixes out-of-bound memory writes caused due to adding tlvs when num_hint_bssid and num_hint_s_ssid are ZEROs.

Tested-on: QCN9074 hw1.0 PCI WLAN.HK.2.4.0.1-01386-QCAHKSWPL_SILICONZ-1

Fixes 74601ecfef6e ("ath11k: Add support for 6g scan hint")
---
 drivers/net/wireless/ath/ath11k/wmi.c | 31 -------------------------------
 1 file changed, 31 deletions(-)
drivers/net/wireless/ath/ath11k/wmi.c +0 -31
index 79b7f991c612..95ebb105dfd7 100644
@@ -2213,37 +2213,6 @@ int ath11k_wmi_send_scan_start_cmd(struct ath11k *ar,
 		}
 	}
 
-	len = params->num_hint_s_ssid * sizeof(struct hint_short_ssid);
-	tlv = ptr;
-	tlv->header = FIELD_PREP(WMI_TLV_TAG, WMI_TAG_ARRAY_FIXED_STRUCT) |
-		      FIELD_PREP(WMI_TLV_LEN, len);
-	ptr += TLV_HDR_SIZE;
-	if (params->num_hint_s_ssid) {
-		s_ssid = ptr;
-		for (i = 0; i < params->num_hint_s_ssid; ++i) {
-			s_ssid->freq_flags = params->hint_s_ssid[i].freq_flags;
-			s_ssid->short_ssid = params->hint_s_ssid[i].short_ssid;
-			s_ssid++;
-		}
-	}
-	ptr += len;
-
-	len = params->num_hint_bssid * sizeof(struct hint_bssid);
-	tlv = ptr;
-	tlv->header = FIELD_PREP(WMI_TLV_TAG, WMI_TAG_ARRAY_FIXED_STRUCT) |
-		      FIELD_PREP(WMI_TLV_LEN, len);
-	ptr += TLV_HDR_SIZE;
-	if (params->num_hint_bssid) {
-		hint_bssid = ptr;
-		for (i = 0; i < params->num_hint_bssid; ++i) {
-			hint_bssid->freq_flags =
-				params->hint_bssid[i].freq_flags;
-			ether_addr_copy(&params->hint_bssid[i].bssid.addr[0],
-					&hint_bssid->bssid.addr[0]);
-			hint_bssid++;
-		}
-	}
-
 	ret = ath11k_wmi_cmd_send(wmi, skb,
 				  WMI_START_SCAN_CMDID);
 	if (ret) {
Signature · 7 lines
2.7.4


-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k
author
#11 lore raw
In reply to Kalle Valo
Quoted 10 lines · Kalle Valo

Few fixes to improve statibility on QCA6390. Please review.

Kalle

Carl Huang (4):
ath11k: put hw to DBS using WMI_PDEV_SET_HW_MODE_CMDID
ath11k: pci: fix hot reset stability issues
ath11k: pci: fix L1ss clock unstable problem
ath11k: pci: disable VDD4BLOW

Quoted 5 lines

Kalle Valo (1):
ath11k: mhi: print a warning if firmware crashed

Pradeep Kumar Chitrapu (1):
ath11k: Fix incorrect tlvs in scan start command

Forgot to mention that these fixes are just cleaned up versions from ath11k-qca6390-bringup branch prepared to go to v5.11, so nothing new here for QCA6390 testing point of view.

Signature · 8 lines
https://patchwork.kernel.org/project/linux-wireless/list/

https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches

-- 
ath11k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath11k