* [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-16 7:02 ` Krzysztof Kozlowski
2024-10-16 10:28 ` Dmitry Baryshkov
2024-10-15 18:26 ` [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441 Raj Kumar Bhagat
` (21 subsequent siblings)
22 siblings, 2 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Raj Kumar Bhagat
Add device-tree bindings for the ATH12K module found in the IPQ5332
device.
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
.../net/wireless/qcom,ath12k-ahb.yaml | 293 ++++++++++++++++++
1 file changed, 293 insertions(+)
create mode 100644 Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
diff --git a/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
new file mode 100644
index 000000000000..54784e396d7e
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
@@ -0,0 +1,293 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+# Copyright (c) 2024 Qualcomm Innovation Center, Inc. All rights reserved.
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/net/wireless/qcom,ath12k-ahb.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Qualcomm Technologies ath12k wireless devices (AHB)
+
+maintainers:
+ - Kalle Valo <kvalo@kernel.org>
+ - Jeff Johnson <jjohnson@kernel.org>
+
+description:
+ Qualcomm Technologies IEEE 802.11be AHB devices.
+
+properties:
+ compatible:
+ enum:
+ - qcom,ipq5332-wifi
+
+ reg:
+ maxItems: 1
+
+ clocks:
+ items:
+ - description: XO clock used for copy engine
+
+ clock-names:
+ items:
+ - const: gcc_xo_clk
+
+ interrupts:
+ items:
+ - description: Ready interrupt
+ - description: Spawn acknowledge interrupt
+ - description: Stop acknowledge interrupt
+ - description: misc-pulse1 interrupt events
+ - description: misc-latch interrupt events
+ - description: sw exception interrupt events
+ - description: interrupt event for ring CE0
+ - description: interrupt event for ring CE1
+ - description: interrupt event for ring CE2
+ - description: interrupt event for ring CE3
+ - description: interrupt event for ring CE4
+ - description: interrupt event for ring CE5
+ - description: interrupt event for ring CE6
+ - description: interrupt event for ring CE7
+ - description: interrupt event for ring CE8
+ - description: interrupt event for ring CE9
+ - description: interrupt event for ring CE10
+ - description: interrupt event for ring CE11
+ - description: interrupt event for ring host2wbm-desc-feed
+ - description: interrupt event for ring host2reo-re-injection
+ - description: interrupt event for ring host2reo-command
+ - description: interrupt event for ring host2rxdma-monitor-ring1
+ - description: interrupt event for ring reo2ost-exception
+ - description: interrupt event for ring wbm2host-rx-release
+ - description: interrupt event for ring reo2host-status
+ - description: interrupt event for ring reo2host-destination-ring4
+ - description: interrupt event for ring reo2host-destination-ring3
+ - description: interrupt event for ring reo2host-destination-ring2
+ - description: interrupt event for ring reo2host-destination-ring1
+ - description: interrupt event for ring rxdma2host-monitor-destination-mac3
+ - description: interrupt event for ring rxdma2host-monitor-destination-mac2
+ - description: interrupt event for ring rxdma2host-monitor-destination-mac1
+ - description: interrupt event for ring host2rxdma-host-buf-ring-mac3
+ - description: interrupt event for ring host2rxdma-host-buf-ring-mac2
+ - description: interrupt event for ring host2rxdma-host-buf-ring-mac1
+ - description: interrupt event for ring host2tcl-input-ring4
+ - description: interrupt event for ring host2tcl-input-ring3
+ - description: interrupt event for ring host2tcl-input-ring2
+ - description: interrupt event for ring host2tcl-input-ring1
+ - description: interrupt event for ring wbm2host-tx-completions-ring4
+ - description: interrupt event for ring wbm2host-tx-completions-ring3
+ - description: interrupt event for ring wbm2host-tx-completions-ring2
+ - description: interrupt event for ring wbm2host-tx-completions-ring1
+ - description: interrupt event for ring host2tx-monitor-ring1
+ - description: interrupt event for ring txmon2host-monitor-destination-mac3
+ - description: interrupt event for ring txmon2host-monitor-destination-mac2
+ - description: interrupt event for ring txmon2host-monitor-destination-mac1
+ - description: interrupt event for umac_reset
+
+ interrupt-names:
+ items:
+ - const: ready
+ - const: spawn
+ - const: stop-ack
+ - const: misc-pulse1
+ - const: misc-latch
+ - const: sw-exception
+ - const: ce0
+ - const: ce1
+ - const: ce2
+ - const: ce3
+ - const: ce4
+ - const: ce5
+ - const: ce6
+ - const: ce7
+ - const: ce8
+ - const: ce9
+ - const: ce10
+ - const: ce11
+ - const: host2wbm-desc-feed
+ - const: host2reo-re-injection
+ - const: host2reo-command
+ - const: host2rxdma-monitor-ring1
+ - const: reo2ost-exception
+ - const: wbm2host-rx-release
+ - const: reo2host-status
+ - const: reo2host-destination-ring4
+ - const: reo2host-destination-ring3
+ - const: reo2host-destination-ring2
+ - const: reo2host-destination-ring1
+ - const: rxdma2host-monitor-destination-mac3
+ - const: rxdma2host-monitor-destination-mac2
+ - const: rxdma2host-monitor-destination-mac1
+ - const: host2rxdma-host-buf-ring-mac3
+ - const: host2rxdma-host-buf-ring-mac2
+ - const: host2rxdma-host-buf-ring-mac1
+ - const: host2tcl-input-ring4
+ - const: host2tcl-input-ring3
+ - const: host2tcl-input-ring2
+ - const: host2tcl-input-ring1
+ - const: wbm2host-tx-completions-ring4
+ - const: wbm2host-tx-completions-ring3
+ - const: wbm2host-tx-completions-ring2
+ - const: wbm2host-tx-completions-ring1
+ - const: host2tx-monitor-ring1
+ - const: txmon2host-monitor-destination-mac3
+ - const: txmon2host-monitor-destination-mac2
+ - const: txmon2host-monitor-destination-mac1
+ - const: umac_reset
+
+ memory-region:
+ minItems: 1
+ description:
+ phandle to a node describing reserved memory (System RAM memory)
+ used by ath12k firmware (see bindings/reserved-memory/reserved-memory.txt)
+
+ qcom,rproc:
+ $ref: /schemas/types.yaml#/definitions/phandle
+ description:
+ DT entry of a WCSS node. WCSS node is the child node of q6 remoteproc driver.
+ (see bindings/remoteproc/qcom,multipd-pil.yaml)
+
+ qcom,smem-states:
+ $ref: /schemas/types.yaml#/definitions/phandle-array
+ description: States used by the AP to signal the remote processor
+ items:
+ - description: Shutdown WCSS pd
+ - description: Stop WCSS pd
+ - description: Spawn WCSS pd
+
+ qcom,smem-state-names:
+ description:
+ Names of the states used by the AP to signal the remote processor
+ items:
+ - const: shutdown
+ - const: stop
+ - const: spawn
+
+required:
+ - compatible
+ - reg
+ - clocks
+ - clock-names
+ - interrupts
+ - interrupt-names
+ - memory-region
+ - qcom,rproc
+ - qcom,smem-states
+ - qcom,smem-state-names
+
+additionalProperties: false
+
+examples:
+ - |
+
+ #include <dt-bindings/interrupt-controller/arm-gic.h>
+ #include <dt-bindings/clock/qcom,ipq5332-gcc.h>
+
+ wifi0: wifi@c0000000 {
+ compatible = "qcom,ipq5332-wifi";
+ reg = <0xc000000 0x1000000>;
+ clocks = <&gcc GCC_XO_CLK>;
+ clock-names = "gcc_xo_clk";
+ interrupts-extended = <&wcss_smp2p_in 9 IRQ_TYPE_NONE>,
+ <&wcss_smp2p_in 12 IRQ_TYPE_NONE>,
+ <&wcss_smp2p_in 11 IRQ_TYPE_NONE>,
+ <&intc GIC_SPI 559 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 560 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 561 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 422 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 423 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 424 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 425 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 426 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 427 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 428 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 429 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 430 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 431 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 432 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 433 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 491 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 495 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 493 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 544 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 457 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 466 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 497 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 454 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 453 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 452 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 451 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 488 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 488 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 484 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 554 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 554 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 549 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 507 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 500 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 499 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 498 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 450 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 449 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 448 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 447 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 543 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 486 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 486 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 482 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 419 IRQ_TYPE_EDGE_RISING>;
+ interrupt-names = "ready",
+ "spawn",
+ "stop-ack",
+ "misc-pulse1",
+ "misc-latch",
+ "sw-exception",
+ "ce0",
+ "ce1",
+ "ce2",
+ "ce3",
+ "ce4",
+ "ce5",
+ "ce6",
+ "ce7",
+ "ce8",
+ "ce9",
+ "ce10",
+ "ce11",
+ "host2wbm-desc-feed",
+ "host2reo-re-injection",
+ "host2reo-command",
+ "host2rxdma-monitor-ring1",
+ "reo2ost-exception",
+ "wbm2host-rx-release",
+ "reo2host-status",
+ "reo2host-destination-ring4",
+ "reo2host-destination-ring3",
+ "reo2host-destination-ring2",
+ "reo2host-destination-ring1",
+ "rxdma2host-monitor-destination-mac3",
+ "rxdma2host-monitor-destination-mac2",
+ "rxdma2host-monitor-destination-mac1",
+ "host2rxdma-host-buf-ring-mac3",
+ "host2rxdma-host-buf-ring-mac2",
+ "host2rxdma-host-buf-ring-mac1",
+ "host2tcl-input-ring4",
+ "host2tcl-input-ring3",
+ "host2tcl-input-ring2",
+ "host2tcl-input-ring1",
+ "wbm2host-tx-completions-ring4",
+ "wbm2host-tx-completions-ring3",
+ "wbm2host-tx-completions-ring2",
+ "wbm2host-tx-completions-ring1",
+ "host2tx-monitor-ring1",
+ "txmon2host-monitor-destination-mac3",
+ "txmon2host-monitor-destination-mac2",
+ "txmon2host-monitor-destination-mac1",
+ "umac_reset";
+
+ memory-region = <&q6_region>;
+ qcom,rproc = <&q6v5_wcss>;
+ qcom,smem-states = <&wcss_smp2p_out 8>,
+ <&wcss_smp2p_out 9>,
+ <&wcss_smp2p_out 10>;
+ qcom,smem-state-names = "shutdown",
+ "stop",
+ "spawn";
+ };
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-15 18:26 ` [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module Raj Kumar Bhagat
@ 2024-10-16 7:02 ` Krzysztof Kozlowski
2024-10-16 8:37 ` Raj Kumar Bhagat
2024-10-16 10:28 ` Dmitry Baryshkov
1 sibling, 1 reply; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-10-16 7:02 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Tue, Oct 15, 2024 at 11:56:16PM +0530, Raj Kumar Bhagat wrote:
> Add device-tree bindings for the ATH12K module found in the IPQ5332
> device.
>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
That's a v2, what changed?
Did you ignore entire review? Limited review follows because of that (I
am not going to do the same work twice).
> .../net/wireless/qcom,ath12k-ahb.yaml | 293 ++++++++++++++++++
> 1 file changed, 293 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
>
> diff --git a/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
> new file mode 100644
> index 000000000000..54784e396d7e
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
> @@ -0,0 +1,293 @@
> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
> +# Copyright (c) 2024 Qualcomm Innovation Center, Inc. All rights reserved.
> +%YAML 1.2
> +---
> +$id: http://devicetree.org/schemas/net/wireless/qcom,ath12k-ahb.yaml#
> +$schema: http://devicetree.org/meta-schemas/core.yaml#
> +
> +title: Qualcomm Technologies ath12k wireless devices (AHB)
> +
> +maintainers:
> + - Kalle Valo <kvalo@kernel.org>
> + - Jeff Johnson <jjohnson@kernel.org>
> +
> +description:
> + Qualcomm Technologies IEEE 802.11be AHB devices.
> +
> +properties:
> + compatible:
> + enum:
> + - qcom,ipq5332-wifi
> +
> + reg:
> + maxItems: 1
> +
> + clocks:
> + items:
> + - description: XO clock used for copy engine
> +
> + clock-names:
> + items:
> + - const: gcc_xo_clk
Drop _clk, drop gcc_. Look how this clock is called *everywhere* else.
> +
> + interrupts:
> + items:
> + - description: Ready interrupt
> + - description: Spawn acknowledge interrupt
> + - description: Stop acknowledge interrupt
> + - description: misc-pulse1 interrupt events
> + - description: misc-latch interrupt events
> + - description: sw exception interrupt events
> + - description: interrupt event for ring CE0
> + - description: interrupt event for ring CE1
> + - description: interrupt event for ring CE2
> + - description: interrupt event for ring CE3
> + - description: interrupt event for ring CE4
> + - description: interrupt event for ring CE5
> + - description: interrupt event for ring CE6
> + - description: interrupt event for ring CE7
> + - description: interrupt event for ring CE8
> + - description: interrupt event for ring CE9
> + - description: interrupt event for ring CE10
> + - description: interrupt event for ring CE11
> + - description: interrupt event for ring host2wbm-desc-feed
> + - description: interrupt event for ring host2reo-re-injection
> + - description: interrupt event for ring host2reo-command
> + - description: interrupt event for ring host2rxdma-monitor-ring1
> + - description: interrupt event for ring reo2ost-exception
> + - description: interrupt event for ring wbm2host-rx-release
> + - description: interrupt event for ring reo2host-status
> + - description: interrupt event for ring reo2host-destination-ring4
> + - description: interrupt event for ring reo2host-destination-ring3
> + - description: interrupt event for ring reo2host-destination-ring2
> + - description: interrupt event for ring reo2host-destination-ring1
> + - description: interrupt event for ring rxdma2host-monitor-destination-mac3
> + - description: interrupt event for ring rxdma2host-monitor-destination-mac2
> + - description: interrupt event for ring rxdma2host-monitor-destination-mac1
> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac3
> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac2
> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac1
> + - description: interrupt event for ring host2tcl-input-ring4
> + - description: interrupt event for ring host2tcl-input-ring3
> + - description: interrupt event for ring host2tcl-input-ring2
> + - description: interrupt event for ring host2tcl-input-ring1
> + - description: interrupt event for ring wbm2host-tx-completions-ring4
> + - description: interrupt event for ring wbm2host-tx-completions-ring3
> + - description: interrupt event for ring wbm2host-tx-completions-ring2
> + - description: interrupt event for ring wbm2host-tx-completions-ring1
> + - description: interrupt event for ring host2tx-monitor-ring1
> + - description: interrupt event for ring txmon2host-monitor-destination-mac3
> + - description: interrupt event for ring txmon2host-monitor-destination-mac2
> + - description: interrupt event for ring txmon2host-monitor-destination-mac1
> + - description: interrupt event for umac_reset
> +
> + interrupt-names:
> + items:
> + - const: ready
> + - const: spawn
> + - const: stop-ack
> + - const: misc-pulse1
> + - const: misc-latch
> + - const: sw-exception
> + - const: ce0
> + - const: ce1
> + - const: ce2
> + - const: ce3
> + - const: ce4
> + - const: ce5
> + - const: ce6
> + - const: ce7
> + - const: ce8
> + - const: ce9
> + - const: ce10
> + - const: ce11
> + - const: host2wbm-desc-feed
> + - const: host2reo-re-injection
> + - const: host2reo-command
> + - const: host2rxdma-monitor-ring1
> + - const: reo2ost-exception
> + - const: wbm2host-rx-release
> + - const: reo2host-status
> + - const: reo2host-destination-ring4
> + - const: reo2host-destination-ring3
> + - const: reo2host-destination-ring2
> + - const: reo2host-destination-ring1
> + - const: rxdma2host-monitor-destination-mac3
> + - const: rxdma2host-monitor-destination-mac2
> + - const: rxdma2host-monitor-destination-mac1
> + - const: host2rxdma-host-buf-ring-mac3
> + - const: host2rxdma-host-buf-ring-mac2
> + - const: host2rxdma-host-buf-ring-mac1
> + - const: host2tcl-input-ring4
> + - const: host2tcl-input-ring3
> + - const: host2tcl-input-ring2
> + - const: host2tcl-input-ring1
> + - const: wbm2host-tx-completions-ring4
> + - const: wbm2host-tx-completions-ring3
> + - const: wbm2host-tx-completions-ring2
> + - const: wbm2host-tx-completions-ring1
> + - const: host2tx-monitor-ring1
> + - const: txmon2host-monitor-destination-mac3
> + - const: txmon2host-monitor-destination-mac2
> + - const: txmon2host-monitor-destination-mac1
> + - const: umac_reset
> +
> + memory-region:
> + minItems: 1
upper constraint
> + description:
> + phandle to a node describing reserved memory (System RAM memory)
> + used by ath12k firmware (see bindings/reserved-memory/reserved-memory.txt)
> +
> + qcom,rproc:
> + $ref: /schemas/types.yaml#/definitions/phandle
> + description:
> + DT entry of a WCSS node. WCSS node is the child node of q6 remoteproc driver.
> + (see bindings/remoteproc/qcom,multipd-pil.yaml)
DT nodes are not children of drivers. But other DT nodes. Explain why
this phandle is needed, what is it for.
To me it looks like you incorrectly organized your nodes.
> +
> + qcom,smem-states:
> + $ref: /schemas/types.yaml#/definitions/phandle-array
> + description: States used by the AP to signal the remote processor
> + items:
> + - description: Shutdown WCSS pd
> + - description: Stop WCSS pd
> + - description: Spawn WCSS pd
> +
> + qcom,smem-state-names:
> + description:
> + Names of the states used by the AP to signal the remote processor
> + items:
> + - const: shutdown
> + - const: stop
> + - const: spawn
> +
> +required:
> + - compatible
> + - reg
> + - clocks
> + - clock-names
> + - interrupts
> + - interrupt-names
> + - memory-region
> + - qcom,rproc
> + - qcom,smem-states
> + - qcom,smem-state-names
> +
> +additionalProperties: false
> +
> +examples:
> + - |
> +
Stray blank line
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-16 7:02 ` Krzysztof Kozlowski
@ 2024-10-16 8:37 ` Raj Kumar Bhagat
2024-10-16 9:00 ` Krzysztof Kozlowski
0 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-16 8:37 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 10/16/2024 12:32 PM, Krzysztof Kozlowski wrote:
> On Tue, Oct 15, 2024 at 11:56:16PM +0530, Raj Kumar Bhagat wrote:
>> Add device-tree bindings for the ATH12K module found in the IPQ5332
>> device.
>>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>
> That's a v2, what changed?
>
> Did you ignore entire review? Limited review follows because of that (I
> am not going to do the same work twice).
>
Review comments in version 1 are addressed.
Per-patch version log is not added here. But we have consolidated version log
in the cover-letter. Will include per-patch version log from next version.
Below are the version log for v2:
- "qcom,board_id" property is dropped. This is not the direct dependency for Ath12k
AHB support, hence it can be taken up separately.
- "qcom,bdf-addr" property is dropped in device-tree and moved to ath12k driver.
- Currently we have only one compatible enum (qcom,ipq5332-wifi), hence conditional
if() check for defining the binding is removed.
- "reserved-memory" node is dropped from example DTS.
- "status" property is dropped in wifi node of example DTS.
>> .../net/wireless/qcom,ath12k-ahb.yaml | 293 ++++++++++++++++++
>> 1 file changed, 293 insertions(+)
>> create mode 100644 Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
>>
>> diff --git a/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
>> new file mode 100644
>> index 000000000000..54784e396d7e
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
>> @@ -0,0 +1,293 @@
>> +# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
>> +# Copyright (c) 2024 Qualcomm Innovation Center, Inc. All rights reserved.
>> +%YAML 1.2
>> +---
>> +$id: http://devicetree.org/schemas/net/wireless/qcom,ath12k-ahb.yaml#
>> +$schema: http://devicetree.org/meta-schemas/core.yaml#
>> +
>> +title: Qualcomm Technologies ath12k wireless devices (AHB)
>> +
>> +maintainers:
>> + - Kalle Valo <kvalo@kernel.org>
>> + - Jeff Johnson <jjohnson@kernel.org>
>> +
>> +description:
>> + Qualcomm Technologies IEEE 802.11be AHB devices.
>> +
>> +properties:
>> + compatible:
>> + enum:
>> + - qcom,ipq5332-wifi
>> +
>> + reg:
>> + maxItems: 1
>> +
>> + clocks:
>> + items:
>> + - description: XO clock used for copy engine
>> +
>> + clock-names:
>> + items:
>> + - const: gcc_xo_clk
>
> Drop _clk, drop gcc_. Look how this clock is called *everywhere* else.
>
Thanks, Based on other bindings example, will rename to "xo"
>> +
>> + interrupts:
>> + items:
>> + - description: Ready interrupt
>> + - description: Spawn acknowledge interrupt
>> + - description: Stop acknowledge interrupt
>> + - description: misc-pulse1 interrupt events
>> + - description: misc-latch interrupt events
>> + - description: sw exception interrupt events
>> + - description: interrupt event for ring CE0
>> + - description: interrupt event for ring CE1
>> + - description: interrupt event for ring CE2
>> + - description: interrupt event for ring CE3
>> + - description: interrupt event for ring CE4
>> + - description: interrupt event for ring CE5
>> + - description: interrupt event for ring CE6
>> + - description: interrupt event for ring CE7
>> + - description: interrupt event for ring CE8
>> + - description: interrupt event for ring CE9
>> + - description: interrupt event for ring CE10
>> + - description: interrupt event for ring CE11
>> + - description: interrupt event for ring host2wbm-desc-feed
>> + - description: interrupt event for ring host2reo-re-injection
>> + - description: interrupt event for ring host2reo-command
>> + - description: interrupt event for ring host2rxdma-monitor-ring1
>> + - description: interrupt event for ring reo2ost-exception
>> + - description: interrupt event for ring wbm2host-rx-release
>> + - description: interrupt event for ring reo2host-status
>> + - description: interrupt event for ring reo2host-destination-ring4
>> + - description: interrupt event for ring reo2host-destination-ring3
>> + - description: interrupt event for ring reo2host-destination-ring2
>> + - description: interrupt event for ring reo2host-destination-ring1
>> + - description: interrupt event for ring rxdma2host-monitor-destination-mac3
>> + - description: interrupt event for ring rxdma2host-monitor-destination-mac2
>> + - description: interrupt event for ring rxdma2host-monitor-destination-mac1
>> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac3
>> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac2
>> + - description: interrupt event for ring host2rxdma-host-buf-ring-mac1
>> + - description: interrupt event for ring host2tcl-input-ring4
>> + - description: interrupt event for ring host2tcl-input-ring3
>> + - description: interrupt event for ring host2tcl-input-ring2
>> + - description: interrupt event for ring host2tcl-input-ring1
>> + - description: interrupt event for ring wbm2host-tx-completions-ring4
>> + - description: interrupt event for ring wbm2host-tx-completions-ring3
>> + - description: interrupt event for ring wbm2host-tx-completions-ring2
>> + - description: interrupt event for ring wbm2host-tx-completions-ring1
>> + - description: interrupt event for ring host2tx-monitor-ring1
>> + - description: interrupt event for ring txmon2host-monitor-destination-mac3
>> + - description: interrupt event for ring txmon2host-monitor-destination-mac2
>> + - description: interrupt event for ring txmon2host-monitor-destination-mac1
>> + - description: interrupt event for umac_reset
>> +
>> + interrupt-names:
>> + items:
>> + - const: ready
>> + - const: spawn
>> + - const: stop-ack
>> + - const: misc-pulse1
>> + - const: misc-latch
>> + - const: sw-exception
>> + - const: ce0
>> + - const: ce1
>> + - const: ce2
>> + - const: ce3
>> + - const: ce4
>> + - const: ce5
>> + - const: ce6
>> + - const: ce7
>> + - const: ce8
>> + - const: ce9
>> + - const: ce10
>> + - const: ce11
>> + - const: host2wbm-desc-feed
>> + - const: host2reo-re-injection
>> + - const: host2reo-command
>> + - const: host2rxdma-monitor-ring1
>> + - const: reo2ost-exception
>> + - const: wbm2host-rx-release
>> + - const: reo2host-status
>> + - const: reo2host-destination-ring4
>> + - const: reo2host-destination-ring3
>> + - const: reo2host-destination-ring2
>> + - const: reo2host-destination-ring1
>> + - const: rxdma2host-monitor-destination-mac3
>> + - const: rxdma2host-monitor-destination-mac2
>> + - const: rxdma2host-monitor-destination-mac1
>> + - const: host2rxdma-host-buf-ring-mac3
>> + - const: host2rxdma-host-buf-ring-mac2
>> + - const: host2rxdma-host-buf-ring-mac1
>> + - const: host2tcl-input-ring4
>> + - const: host2tcl-input-ring3
>> + - const: host2tcl-input-ring2
>> + - const: host2tcl-input-ring1
>> + - const: wbm2host-tx-completions-ring4
>> + - const: wbm2host-tx-completions-ring3
>> + - const: wbm2host-tx-completions-ring2
>> + - const: wbm2host-tx-completions-ring1
>> + - const: host2tx-monitor-ring1
>> + - const: txmon2host-monitor-destination-mac3
>> + - const: txmon2host-monitor-destination-mac2
>> + - const: txmon2host-monitor-destination-mac1
>> + - const: umac_reset
>> +
>> + memory-region:
>> + minItems: 1
>
> upper constraint
>
>> + description:
>> + phandle to a node describing reserved memory (System RAM memory)
>> + used by ath12k firmware (see bindings/reserved-memory/reserved-memory.txt)
>> +
>> + qcom,rproc:
>> + $ref: /schemas/types.yaml#/definitions/phandle
>> + description:
>> + DT entry of a WCSS node. WCSS node is the child node of q6 remoteproc driver.
>> + (see bindings/remoteproc/qcom,multipd-pil.yaml)
>
> DT nodes are not children of drivers. But other DT nodes. Explain why
> this phandle is needed, what is it for.
>
> To me it looks like you incorrectly organized your nodes.
>
This phandle is required by wifi driver (ath12k) to retrieve the correct remote processor
(rproc_get_by_phandle()). Ath12k driver needs this rproc to interact with the remote
processor (example: booting-up remote processor).
In next version, will correct the description based on existing bindings (qcom,ath11k.yaml).
>> +
>> + qcom,smem-states:
>> + $ref: /schemas/types.yaml#/definitions/phandle-array
>> + description: States used by the AP to signal the remote processor
>> + items:
>> + - description: Shutdown WCSS pd
>> + - description: Stop WCSS pd
>> + - description: Spawn WCSS pd
>> +
>> + qcom,smem-state-names:
>> + description:
>> + Names of the states used by the AP to signal the remote processor
>> + items:
>> + - const: shutdown
>> + - const: stop
>> + - const: spawn
>> +
>> +required:
>> + - compatible
>> + - reg
>> + - clocks
>> + - clock-names
>> + - interrupts
>> + - interrupt-names
>> + - memory-region
>> + - qcom,rproc
>> + - qcom,smem-states
>> + - qcom,smem-state-names
>> +
>> +additionalProperties: false
>> +
>> +examples:
>> + - |
>> +
>
> Stray blank line
>
Will update in next version.
> Best regards,
> Krzysztof
>
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-16 8:37 ` Raj Kumar Bhagat
@ 2024-10-16 9:00 ` Krzysztof Kozlowski
2024-11-27 7:24 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-10-16 9:00 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 16/10/2024 10:37, Raj Kumar Bhagat wrote:
> On 10/16/2024 12:32 PM, Krzysztof Kozlowski wrote:
>> On Tue, Oct 15, 2024 at 11:56:16PM +0530, Raj Kumar Bhagat wrote:
>>> Add device-tree bindings for the ATH12K module found in the IPQ5332
>>> device.
>>>
>>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>>> ---
>>
>> That's a v2, what changed?
>>
>> Did you ignore entire review? Limited review follows because of that (I
>> am not going to do the same work twice).
>>
>
> Review comments in version 1 are addressed.
>
> Per-patch version log is not added here. But we have consolidated version log
> in the cover-letter. Will include per-patch version log from next version.
Hm? There is no such in cover letter. There is description, dependencies
and list of commits which indicates end (format-patch standard stuff).
>
> Below are the version log for v2:
> - "qcom,board_id" property is dropped. This is not the direct dependency for Ath12k
> AHB support, hence it can be taken up separately.
> - "qcom,bdf-addr" property is dropped in device-tree and moved to ath12k driver.
> - Currently we have only one compatible enum (qcom,ipq5332-wifi), hence conditional
> if() check for defining the binding is removed.
> - "reserved-memory" node is dropped from example DTS.
> - "status" property is dropped in wifi node of example DTS.
>>> +
>>> + clock-names:
>>> + items:
>>> + - const: gcc_xo_clk
>>
>> Drop _clk, drop gcc_. Look how this clock is called *everywhere* else.
>>
>
> Thanks, Based on other bindings example, will rename to "xo"
git grep gcc_xo_clk -> nothing like that!
...
>>> +
>>> + memory-region:
>>> + minItems: 1
>>
>> upper constraint
>>
>>> + description:
>>> + phandle to a node describing reserved memory (System RAM memory)
>>> + used by ath12k firmware (see bindings/reserved-memory/reserved-memory.txt)
>>> +
>>> + qcom,rproc:
>>> + $ref: /schemas/types.yaml#/definitions/phandle
>>> + description:
>>> + DT entry of a WCSS node. WCSS node is the child node of q6 remoteproc driver.
>>> + (see bindings/remoteproc/qcom,multipd-pil.yaml)
>>
>> DT nodes are not children of drivers. But other DT nodes. Explain why
>> this phandle is needed, what is it for.
>>
>> To me it looks like you incorrectly organized your nodes.
>>
>
> This phandle is required by wifi driver (ath12k) to retrieve the correct remote processor
> (rproc_get_by_phandle()). Ath12k driver needs this rproc to interact with the remote
> processor (example: booting-up remote processor).
That's driver aspect. Why does the hardware needs it?
WiFi is the remote processor, so I would expect this being a child. Or
just drop entirely.
You keep using here arguments how you designed your drivers, which is
not valid. Sorry, fix your drivers... or use arguments in terms of hardware.
>
> In next version, will correct the description based on existing bindings (qcom,ath11k.yaml).
Sorry, let's don't copy existing solutions just because they exist.
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-16 9:00 ` Krzysztof Kozlowski
@ 2024-11-27 7:24 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-11-27 7:24 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 10/16/2024 2:30 PM, Krzysztof Kozlowski wrote:
>>>> + description:
>>>> + phandle to a node describing reserved memory (System RAM memory)
>>>> + used by ath12k firmware (see bindings/reserved-memory/reserved-memory.txt)
>>>> +
>>>> + qcom,rproc:
>>>> + $ref: /schemas/types.yaml#/definitions/phandle
>>>> + description:
>>>> + DT entry of a WCSS node. WCSS node is the child node of q6 remoteproc driver.
>>>> + (see bindings/remoteproc/qcom,multipd-pil.yaml)
>>> DT nodes are not children of drivers. But other DT nodes. Explain why
>>> this phandle is needed, what is it for.
>>>
>>> To me it looks like you incorrectly organized your nodes.
>>>
>> This phandle is required by wifi driver (ath12k) to retrieve the correct remote processor
>> (rproc_get_by_phandle()). Ath12k driver needs this rproc to interact with the remote
>> processor (example: booting-up remote processor).
> That's driver aspect. Why does the hardware needs it?
>
> WiFi is the remote processor, so I would expect this being a child. Or
> just drop entirely.
>
> You keep using here arguments how you designed your drivers, which is
> not valid. Sorry, fix your drivers... or use arguments in terms of hardware.
>
>
>> In next version, will correct the description based on existing bindings (qcom,ath11k.yaml).
> Sorry, let's don't copy existing solutions just because they exist.
In the IPQ5332 platform, the WiFi component, known as the Wireless Control Subsystem (WCSS),
functions as the wireless controller. Additionally, there is a separate hardware module, the
Qualcomm Hexagon DSP(q6 remote processor), which is utilized for offloading WiFi processing
tasks. These two hardware modules operate in conjunction, necessitating a phandle in the WiFi
node that references the remote processor. Despite their interdependence, these modules are
distinct from a hardware perspective.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-15 18:26 ` [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module Raj Kumar Bhagat
2024-10-16 7:02 ` Krzysztof Kozlowski
@ 2024-10-16 10:28 ` Dmitry Baryshkov
2024-12-03 9:07 ` Raj Kumar Bhagat
1 sibling, 1 reply; 60+ messages in thread
From: Dmitry Baryshkov @ 2024-10-16 10:28 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Tue, Oct 15, 2024 at 11:56:16PM +0530, Raj Kumar Bhagat wrote:
> Add device-tree bindings for the ATH12K module found in the IPQ5332
> device.
>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
> .../net/wireless/qcom,ath12k-ahb.yaml | 293 ++++++++++++++++++
> 1 file changed, 293 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
Generic comment, please add qcom,ath12k-calibration-variant
--
With best wishes
Dmitry
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module
2024-10-16 10:28 ` Dmitry Baryshkov
@ 2024-12-03 9:07 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-03 9:07 UTC (permalink / raw)
To: Dmitry Baryshkov
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 10/16/2024 3:58 PM, Dmitry Baryshkov wrote:
> On Tue, Oct 15, 2024 at 11:56:16PM +0530, Raj Kumar Bhagat wrote:
>> Add device-tree bindings for the ATH12K module found in the IPQ5332
>> device.
>>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>> .../net/wireless/qcom,ath12k-ahb.yaml | 293 ++++++++++++++++++
>> 1 file changed, 293 insertions(+)
>> create mode 100644 Documentation/devicetree/bindings/net/wireless/qcom,ath12k-ahb.yaml
>
> Generic comment, please add qcom,ath12k-calibration-variant
>
Thanks, will include "qcom,ath12k-calibration-variant" in next version
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-16 6:58 ` Krzysztof Kozlowski
2024-10-15 18:26 ` [PATCH v2 03/22] wifi: ath12k: add ath12k_hw_params for IPQ5332 Raj Kumar Bhagat
` (20 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Raj Kumar Bhagat
RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
device.
Describe and add WiFi node for RDP441. Also, reserve the memory
required by IPQ5332 firmware.
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
arch/arm64/boot/dts/qcom/ipq5332-rdp441.dts | 59 ++++++++++-
arch/arm64/boot/dts/qcom/ipq5332.dtsi | 108 +++++++++++++++++++-
2 files changed, 165 insertions(+), 2 deletions(-)
diff --git a/arch/arm64/boot/dts/qcom/ipq5332-rdp441.dts b/arch/arm64/boot/dts/qcom/ipq5332-rdp441.dts
index 846413817e9a..699422299336 100644
--- a/arch/arm64/boot/dts/qcom/ipq5332-rdp441.dts
+++ b/arch/arm64/boot/dts/qcom/ipq5332-rdp441.dts
@@ -2,7 +2,7 @@
/*
* IPQ5332 AP-MI01.2 board device tree source
*
- * Copyright (c) 2022-2023 Qualcomm Innovation Center, Inc. All rights reserved.
+ * Copyright (c) 2022-2024 Qualcomm Innovation Center, Inc. All rights reserved.
*/
/dts-v1/;
@@ -12,6 +12,51 @@
/ {
model = "Qualcomm Technologies, Inc. IPQ5332 MI01.2";
compatible = "qcom,ipq5332-ap-mi01.2", "qcom,ipq5332";
+
+ /* Default Profile
+ * +============+==============+=====================+
+ * | | | |
+ * | Region | Start Offset | Size |
+ * | | | |
+ * +------------+--------------+---------------------+
+ * | | | |
+ * | | | |
+ * | | | |
+ * | WLAN Q6 | 0x4A900000 | 43MB |
+ * | | | |
+ * | | | |
+ * +------------+--------------+---------------------+
+ * | M3 Dump | 0x4D400000 | 1MB |
+ * +============+==============+=====================+
+ * | |
+ * | |
+ * | |
+ * | Rest of memory for Linux |
+ * | |
+ * | |
+ * | |
+ * +=================================================+
+ */
+
+ reserved-memory {
+ #address-cells = <2>;
+ #size-cells = <2>;
+ ranges;
+
+ q6_region: wcss@4a900000 {
+ reg = <0x0 0x4a900000 0x0 0x02b00000>;
+ no-map;
+ };
+
+ m3_dump: m3_dump@4d400000 {
+ reg = <0x0 0x4D400000 0x0 0x100000>;
+ no-map;
+ };
+
+ /* mlo_global_mem0: The MLO global memory is not
+ * enabled yet.
+ */
+ };
};
&blsp1_i2c1 {
@@ -21,6 +66,18 @@ &blsp1_i2c1 {
status = "okay";
};
+&wifi0 {
+ memory-region = <&q6_region>;
+ qcom,rproc = <&q6v5_wcss>;
+ qcom,smem-states = <&wcss_smp2p_out 8>,
+ <&wcss_smp2p_out 9>,
+ <&wcss_smp2p_out 10>;
+ qcom,smem-state-names = "shutdown",
+ "stop",
+ "spawn";
+ status = "okay";
+};
+
&sdhc {
bus-width = <4>;
max-frequency = <192000000>;
diff --git a/arch/arm64/boot/dts/qcom/ipq5332.dtsi b/arch/arm64/boot/dts/qcom/ipq5332.dtsi
index 06fe7c94ee37..b419edfea0cd 100644
--- a/arch/arm64/boot/dts/qcom/ipq5332.dtsi
+++ b/arch/arm64/boot/dts/qcom/ipq5332.dtsi
@@ -2,7 +2,7 @@
/*
* IPQ5332 device tree source
*
- * Copyright (c) 2022-2023 Qualcomm Innovation Center, Inc. All rights reserved.
+ * Copyright (c) 2022-2024 Qualcomm Innovation Center, Inc. All rights reserved.
*/
#include <dt-bindings/clock/qcom,apss-ipq.h>
@@ -517,6 +517,112 @@ glink-edge {
mboxes = <&apcs_glb 8>;
};
};
+
+ wifi0: wifi@c0000000 {
+ compatible = "qcom,ipq5332-wifi";
+ reg = <0xc000000 0x1000000>;
+ clocks = <&gcc GCC_XO_CLK>;
+ clock-names = "gcc_xo_clk";
+ interrupts-extended = <&wcss_smp2p_in 9 IRQ_TYPE_NONE>,
+ <&wcss_smp2p_in 12 IRQ_TYPE_NONE>,
+ <&wcss_smp2p_in 11 IRQ_TYPE_NONE>,
+ <&intc GIC_SPI 559 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 560 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 561 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 422 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 423 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 424 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 425 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 426 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 427 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 428 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 429 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 430 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 431 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 432 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 433 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 491 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 495 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 493 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 544 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 457 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 466 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 497 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 454 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 453 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 452 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 451 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 488 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 488 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 484 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 554 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 554 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 549 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 507 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 500 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 499 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 498 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 450 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 449 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 448 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 447 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 543 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 486 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 486 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 482 IRQ_TYPE_EDGE_RISING>,
+ <&intc GIC_SPI 419 IRQ_TYPE_EDGE_RISING>;
+
+ interrupt-names = "ready",
+ "spawn",
+ "stop-ack",
+ "misc-pulse1",
+ "misc-latch",
+ "sw-exception",
+ "ce0",
+ "ce1",
+ "ce2",
+ "ce3",
+ "ce4",
+ "ce5",
+ "ce6",
+ "ce7",
+ "ce8",
+ "ce9",
+ "ce10",
+ "ce11",
+ "host2wbm-desc-feed",
+ "host2reo-re-injection",
+ "host2reo-command",
+ "host2rxdma-monitor-ring1",
+ "reo2ost-exception",
+ "wbm2host-rx-release",
+ "reo2host-status",
+ "reo2host-destination-ring4",
+ "reo2host-destination-ring3",
+ "reo2host-destination-ring2",
+ "reo2host-destination-ring1",
+ "rxdma2host-monitor-destination-mac3",
+ "rxdma2host-monitor-destination-mac2",
+ "rxdma2host-monitor-destination-mac1",
+ "host2rxdma-host-buf-ring-mac3",
+ "host2rxdma-host-buf-ring-mac2",
+ "host2rxdma-host-buf-ring-mac1",
+ "host2tcl-input-ring4",
+ "host2tcl-input-ring3",
+ "host2tcl-input-ring2",
+ "host2tcl-input-ring1",
+ "wbm2host-tx-completions-ring4",
+ "wbm2host-tx-completions-ring3",
+ "wbm2host-tx-completions-ring2",
+ "wbm2host-tx-completions-ring1",
+ "host2tx-monitor-ring1",
+ "txmon2host-monitor-destination-mac3",
+ "txmon2host-monitor-destination-mac2",
+ "txmon2host-monitor-destination-mac1",
+ "umac_reset";
+
+ status = "disabled";
+ };
};
timer {
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-15 18:26 ` [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441 Raj Kumar Bhagat
@ 2024-10-16 6:58 ` Krzysztof Kozlowski
2024-10-16 8:48 ` Raj Kumar Bhagat
2024-10-16 10:30 ` Dmitry Baryshkov
0 siblings, 2 replies; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-10-16 6:58 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Tue, Oct 15, 2024 at 11:56:17PM +0530, Raj Kumar Bhagat wrote:
> RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
> device.
>
> Describe and add WiFi node for RDP441. Also, reserve the memory
> required by IPQ5332 firmware.
>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
Don't send one DTS patch in 22 patchset targetting different subsystem.
Imagine, how wireless maintainers are supposed to apply their bits? 21
commands instead of one command?
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-16 6:58 ` Krzysztof Kozlowski
@ 2024-10-16 8:48 ` Raj Kumar Bhagat
2024-10-16 10:30 ` Dmitry Baryshkov
1 sibling, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-16 8:48 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 10/16/2024 12:28 PM, Krzysztof Kozlowski wrote:
> On Tue, Oct 15, 2024 at 11:56:17PM +0530, Raj Kumar Bhagat wrote:
>> RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
>> device.
>>
>> Describe and add WiFi node for RDP441. Also, reserve the memory
>> required by IPQ5332 firmware.
>>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> Don't send one DTS patch in 22 patchset targetting different subsystem.
> Imagine, how wireless maintainers are supposed to apply their bits? 21
> commands instead of one command?
Sure, we can drop the DTS patch from this series. Will take this patch
separately once the binding (qcom,ath12k-ahb.yaml) is merged.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-16 6:58 ` Krzysztof Kozlowski
2024-10-16 8:48 ` Raj Kumar Bhagat
@ 2024-10-16 10:30 ` Dmitry Baryshkov
2024-10-16 10:55 ` Krzysztof Kozlowski
1 sibling, 1 reply; 60+ messages in thread
From: Dmitry Baryshkov @ 2024-10-16 10:30 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: Raj Kumar Bhagat, ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Wed, Oct 16, 2024 at 08:58:25AM +0200, Krzysztof Kozlowski wrote:
> On Tue, Oct 15, 2024 at 11:56:17PM +0530, Raj Kumar Bhagat wrote:
> > RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
> > device.
> >
> > Describe and add WiFi node for RDP441. Also, reserve the memory
> > required by IPQ5332 firmware.
> >
> > Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>
> Don't send one DTS patch in 22 patchset targetting different subsystem.
> Imagine, how wireless maintainers are supposed to apply their bits? 21
> commands instead of one command?
Huh? b4 shazam -P 1,3-22 should work. Or ideally the DTS should be the
last one, so applying all other patches should be obvious. As a reviewer
I find it troublesome to review bindindings / driver without an actual
DTS snippet.
--
With best wishes
Dmitry
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-16 10:30 ` Dmitry Baryshkov
@ 2024-10-16 10:55 ` Krzysztof Kozlowski
2024-10-16 11:13 ` Dmitry Baryshkov
0 siblings, 1 reply; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-10-16 10:55 UTC (permalink / raw)
To: Dmitry Baryshkov
Cc: Raj Kumar Bhagat, ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 16/10/2024 12:30, Dmitry Baryshkov wrote:
> On Wed, Oct 16, 2024 at 08:58:25AM +0200, Krzysztof Kozlowski wrote:
>> On Tue, Oct 15, 2024 at 11:56:17PM +0530, Raj Kumar Bhagat wrote:
>>> RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
>>> device.
>>>
>>> Describe and add WiFi node for RDP441. Also, reserve the memory
>>> required by IPQ5332 firmware.
>>>
>>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>>
>> Don't send one DTS patch in 22 patchset targetting different subsystem.
>> Imagine, how wireless maintainers are supposed to apply their bits? 21
>> commands instead of one command?
>
> Huh? b4 shazam -P 1,3-22 should work. Or ideally the DTS should be the
Hm indeed, it wasn't some time ago.
> last one, so applying all other patches should be obvious. As a reviewer
> I find it troublesome to review bindindings / driver without an actual
> DTS snippet.
Considering that patchsets for certain subsystem *have to skip DTS* (you
cannot include DTS in the series), then better get used to such
inconvenience.
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441
2024-10-16 10:55 ` Krzysztof Kozlowski
@ 2024-10-16 11:13 ` Dmitry Baryshkov
0 siblings, 0 replies; 60+ messages in thread
From: Dmitry Baryshkov @ 2024-10-16 11:13 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: Raj Kumar Bhagat, ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Wed, 16 Oct 2024 at 13:55, Krzysztof Kozlowski <krzk@kernel.org> wrote:
>
> On 16/10/2024 12:30, Dmitry Baryshkov wrote:
> > On Wed, Oct 16, 2024 at 08:58:25AM +0200, Krzysztof Kozlowski wrote:
> >> On Tue, Oct 15, 2024 at 11:56:17PM +0530, Raj Kumar Bhagat wrote:
> >>> RDP441 is based on IPQ5332. It has inbuilt AHB bus based IPQ5332 WiFi
> >>> device.
> >>>
> >>> Describe and add WiFi node for RDP441. Also, reserve the memory
> >>> required by IPQ5332 firmware.
> >>>
> >>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> >>
> >> Don't send one DTS patch in 22 patchset targetting different subsystem.
> >> Imagine, how wireless maintainers are supposed to apply their bits? 21
> >> commands instead of one command?
> >
> > Huh? b4 shazam -P 1,3-22 should work. Or ideally the DTS should be the
>
> Hm indeed, it wasn't some time ago.
>
> > last one, so applying all other patches should be obvious. As a reviewer
> > I find it troublesome to review bindindings / driver without an actual
> > DTS snippet.
>
> Considering that patchsets for certain subsystem *have to skip DTS* (you
> cannot include DTS in the series), then better get used to such
> inconvenience.
Yes, I'm getting used to that for some of the subsys.
--
With best wishes
Dmitry
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 03/22] wifi: ath12k: add ath12k_hw_params for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 01/22] dt-bindings: net: wireless: describe the ath12k AHB module Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 02/22] arm64: dts: qcom: add wifi node for IPQ5332 based RDP441 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 04/22] wifi: ath12k: refactor ath12k_hw_regs structure Raj Kumar Bhagat
` (19 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
Add ath12k_hw_params for new ath12k AHB based WiFi device IPQ5332.
Some hardware parameters like hw_ops, wmi_init & hal_ops are same
between IPQ5332 and QCN9274, hence use these same parameters for
IPQ5332. Other parameters are specific to IPQ5332, and those are
initially set to 0 or NULL, and will be populated in subsequent
patches.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/core.h | 3 +-
drivers/net/wireless/ath/ath12k/hw.c | 71 ++++++++++++++++++++++++++
drivers/net/wireless/ath/ath12k/qmi.h | 1 +
3 files changed, 74 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index ebfc1e370acc..126c9fed1fad 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -128,7 +128,8 @@ struct ath12k_skb_rxcb {
enum ath12k_hw_rev {
ATH12K_HW_QCN9274_HW10,
ATH12K_HW_QCN9274_HW20,
- ATH12K_HW_WCN7850_HW20
+ ATH12K_HW_WCN7850_HW20,
+ ATH12K_HW_IPQ5332_HW10,
};
enum ath12k_firmware_mode {
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index ec1bda95e555..7ca6ec99fd6e 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1085,6 +1085,77 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_aspm = false,
},
+ {
+ .name = "ipq5332 hw1.0",
+ .hw_rev = ATH12K_HW_IPQ5332_HW10,
+ .fw = {
+ .dir = "IPQ5332/hw1.0",
+ .board_size = 256 * 1024,
+ .cal_offset = 128 * 1024,
+ },
+ .max_radios = 1,
+ .single_pdev_only = false,
+ .qmi_service_ins_id = ATH12K_QMI_WLFW_SERVICE_INS_ID_V01_IPQ5332,
+ .internal_sleep_clock = false,
+
+ .hw_ops = &qcn9274_ops,
+ .regs = NULL,
+ .ring_mask = NULL,
+
+ .host_ce_config = NULL,
+ .ce_count = 0,
+ .target_ce_config = NULL,
+ .target_ce_count = 0,
+ .svc_to_ce_map = NULL,
+ .svc_to_ce_map_len = 0,
+
+ .hal_params = NULL,
+
+ .rxdma1_enable = false,
+ .num_rxdma_per_pdev = 1,
+ .num_rxdma_dst_ring = 0,
+ .rx_mac_buf_ring = false,
+ .vdev_start_delay = false,
+
+ .interface_modes = BIT(NL80211_IFTYPE_STATION) |
+ BIT(NL80211_IFTYPE_AP) |
+ BIT(NL80211_IFTYPE_MESH_POINT),
+ .supports_monitor = false,
+
+ .idle_ps = false,
+ .download_calib = true,
+ .supports_suspend = false,
+ .tcl_ring_retry = true,
+ .reoq_lut_support = false,
+ .supports_shadow_regs = false,
+
+ .num_tcl_banks = 48,
+ .max_tx_ring = 4,
+
+ .wmi_init = &ath12k_wmi_init_qcn9274,
+
+ .hal_ops = &hal_qcn9274_ops,
+
+ .qmi_cnss_feature_bitmap = BIT(CNSS_QDSS_CFG_MISS_V01),
+
+ .rfkill_pin = 0,
+ .rfkill_cfg = 0,
+ .rfkill_on_level = 0,
+
+ .rddm_size = 0,
+
+ .def_num_link = 0,
+ .max_mlo_peer = 256,
+
+ .otp_board_id_register = 0,
+
+ .supports_sta_ps = false,
+
+ .acpi_guid = NULL,
+ .supports_dynamic_smps_6ghz = false,
+ .iova_mask = 0,
+ .supports_aspm = false,
+ },
};
int ath12k_hw_init(struct ath12k_base *ab)
diff --git a/drivers/net/wireless/ath/ath12k/qmi.h b/drivers/net/wireless/ath/ath12k/qmi.h
index 0dfcbd8cb59b..6ce11060f1aa 100644
--- a/drivers/net/wireless/ath/ath12k/qmi.h
+++ b/drivers/net/wireless/ath/ath12k/qmi.h
@@ -21,6 +21,7 @@
#define ATH12K_QMI_WLFW_SERVICE_INS_ID_V01_WCN7850 0x1
#define ATH12K_QMI_WLFW_SERVICE_INS_ID_V01_QCN9274 0x07
+#define ATH12K_QMI_WLFW_SERVICE_INS_ID_V01_IPQ5332 0x2
#define ATH12K_QMI_WLANFW_MAX_TIMESTAMP_LEN_V01 32
#define ATH12K_QMI_RESP_LEN_MAX 8192
#define ATH12K_QMI_WLANFW_MAX_NUM_MEM_SEG_V01 52
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 04/22] wifi: ath12k: refactor ath12k_hw_regs structure
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (2 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 03/22] wifi: ath12k: add ath12k_hw_params for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332 Raj Kumar Bhagat
` (18 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S, Raj Kumar Bhagat
From: P Praneesh <quic_ppranees@quicinc.com>
IPQ5332 device have different register address values for the below
registers:
HAL_TCL1_RING_BASE_LSB
HAL_TCL1_RING_BASE_MSB
HAL_TCL2_RING_BASE_LSB
HAL_SEQ_WCSS_UMAC_CE0_SRC_REG
HAL_SEQ_WCSS_UMAC_CE0_DST_REG
HAL_SEQ_WCSS_UMAC_CE1_SRC_REG
HAL_SEQ_WCSS_UMAC_CE1_DST_REG
Hence, refactor ath12k_hw_regs structure to accommodate these changes
in IPQ5332.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hal.c | 82 +++++++++++++--------------
drivers/net/wireless/ath/ath12k/hal.h | 61 +++++++++++---------
drivers/net/wireless/ath/ath12k/hw.c | 27 +++++++++
drivers/net/wireless/ath/ath12k/hw.h | 8 +++
4 files changed, 110 insertions(+), 68 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/hal.c b/drivers/net/wireless/ath/ath12k/hal.c
index ca04bfae8bdc..d2ad0224fdfe 100644
--- a/drivers/net/wireless/ath/ath12k/hal.c
+++ b/drivers/net/wireless/ath/ath12k/hal.c
@@ -552,9 +552,9 @@ static int ath12k_hal_srng_create_config_qcn9274(struct ath12k_base *ab)
s->reg_start[1] = HAL_SEQ_WCSS_UMAC_REO_REG + HAL_REO_STATUS_HP;
s = &hal->srng_config[HAL_TCL_DATA];
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_BASE_LSB;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_BASE_LSB(ab);
s->reg_start[1] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_HP;
- s->reg_size[0] = HAL_TCL2_RING_BASE_LSB - HAL_TCL1_RING_BASE_LSB;
+ s->reg_size[0] = HAL_TCL2_RING_BASE_LSB(ab) - HAL_TCL1_RING_BASE_LSB(ab);
s->reg_size[1] = HAL_TCL2_RING_HP - HAL_TCL1_RING_HP;
s = &hal->srng_config[HAL_TCL_CMD];
@@ -566,29 +566,29 @@ static int ath12k_hal_srng_create_config_qcn9274(struct ath12k_base *ab)
s->reg_start[1] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL_STATUS_RING_HP;
s = &hal->srng_config[HAL_CE_SRC];
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG + HAL_CE_DST_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG + HAL_CE_DST_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG -
- HAL_SEQ_WCSS_UMAC_CE0_SRC_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG -
- HAL_SEQ_WCSS_UMAC_CE0_SRC_REG;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab) + HAL_CE_DST_RING_BASE_LSB;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab) + HAL_CE_DST_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab);
s = &hal->srng_config[HAL_CE_DST];
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_RING_BASE_LSB;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
s = &hal->srng_config[HAL_CE_DST_STATUS];
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG +
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) +
HAL_CE_DST_STATUS_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_STATUS_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_STATUS_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
s = &hal->srng_config[HAL_WBM_IDLE_LINK];
s->reg_start[0] = HAL_SEQ_WCSS_UMAC_WBM_REG + HAL_WBM_IDLE_LINK_RING_BASE_LSB(ab);
@@ -1371,9 +1371,9 @@ static int ath12k_hal_srng_create_config_wcn7850(struct ath12k_base *ab)
s = &hal->srng_config[HAL_TCL_DATA];
s->max_rings = 5;
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_BASE_LSB;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_BASE_LSB(ab);
s->reg_start[1] = HAL_SEQ_WCSS_UMAC_TCL_REG + HAL_TCL1_RING_HP;
- s->reg_size[0] = HAL_TCL2_RING_BASE_LSB - HAL_TCL1_RING_BASE_LSB;
+ s->reg_size[0] = HAL_TCL2_RING_BASE_LSB(ab) - HAL_TCL1_RING_BASE_LSB(ab);
s->reg_size[1] = HAL_TCL2_RING_HP - HAL_TCL1_RING_HP;
s = &hal->srng_config[HAL_TCL_CMD];
@@ -1386,31 +1386,31 @@ static int ath12k_hal_srng_create_config_wcn7850(struct ath12k_base *ab)
s = &hal->srng_config[HAL_CE_SRC];
s->max_rings = 12;
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG + HAL_CE_DST_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG + HAL_CE_DST_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG -
- HAL_SEQ_WCSS_UMAC_CE0_SRC_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG -
- HAL_SEQ_WCSS_UMAC_CE0_SRC_REG;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab) + HAL_CE_DST_RING_BASE_LSB;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab) + HAL_CE_DST_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_SRC_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab);
s = &hal->srng_config[HAL_CE_DST];
s->max_rings = 12;
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_RING_BASE_LSB;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
s = &hal->srng_config[HAL_CE_DST_STATUS];
s->max_rings = 12;
- s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG +
+ s->reg_start[0] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) +
HAL_CE_DST_STATUS_RING_BASE_LSB;
- s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG + HAL_CE_DST_STATUS_RING_HP;
- s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
- s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG -
- HAL_SEQ_WCSS_UMAC_CE0_DST_REG;
+ s->reg_start[1] = HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) + HAL_CE_DST_STATUS_RING_HP;
+ s->reg_size[0] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
+ s->reg_size[1] = HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) -
+ HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab);
s = &hal->srng_config[HAL_WBM_IDLE_LINK];
s->reg_start[0] = HAL_SEQ_WCSS_UMAC_WBM_REG + HAL_WBM_IDLE_LINK_RING_BASE_LSB(ab);
@@ -1756,7 +1756,7 @@ static void ath12k_hal_srng_src_hw_init(struct ath12k_base *ab,
HAL_TCL1_RING_BASE_MSB_RING_BASE_ADDR_MSB) |
u32_encode_bits((srng->entry_size * srng->num_entries),
HAL_TCL1_RING_BASE_MSB_RING_SIZE);
- ath12k_hif_write32(ab, reg_base + HAL_TCL1_RING_BASE_MSB_OFFSET, val);
+ ath12k_hif_write32(ab, reg_base + HAL_TCL1_RING_BASE_MSB_OFFSET(ab), val);
val = u32_encode_bits(srng->entry_size, HAL_REO1_RING_ID_ENTRY_SIZE);
ath12k_hif_write32(ab, reg_base + HAL_TCL1_RING_ID_OFFSET(ab), val);
diff --git a/drivers/net/wireless/ath/ath12k/hal.h b/drivers/net/wireless/ath/ath12k/hal.h
index 8a78bb9a10bc..69579f118b81 100644
--- a/drivers/net/wireless/ath/ath12k/hal.h
+++ b/drivers/net/wireless/ath/ath12k/hal.h
@@ -44,10 +44,14 @@ struct ath12k_base;
#define HAL_SEQ_WCSS_UMAC_OFFSET 0x00a00000
#define HAL_SEQ_WCSS_UMAC_REO_REG 0x00a38000
#define HAL_SEQ_WCSS_UMAC_TCL_REG 0x00a44000
-#define HAL_SEQ_WCSS_UMAC_CE0_SRC_REG 0x01b80000
-#define HAL_SEQ_WCSS_UMAC_CE0_DST_REG 0x01b81000
-#define HAL_SEQ_WCSS_UMAC_CE1_SRC_REG 0x01b82000
-#define HAL_SEQ_WCSS_UMAC_CE1_DST_REG 0x01b83000
+#define HAL_SEQ_WCSS_UMAC_CE0_SRC_REG(ab) \
+ ((ab)->hw_params->regs->hal_umac_ce0_src_reg_base)
+#define HAL_SEQ_WCSS_UMAC_CE0_DST_REG(ab) \
+ ((ab)->hw_params->regs->hal_umac_ce0_dest_reg_base)
+#define HAL_SEQ_WCSS_UMAC_CE1_SRC_REG(ab) \
+ ((ab)->hw_params->regs->hal_umac_ce1_src_reg_base)
+#define HAL_SEQ_WCSS_UMAC_CE1_DST_REG(ab) \
+ ((ab)->hw_params->regs->hal_umac_ce1_dest_reg_base)
#define HAL_SEQ_WCSS_UMAC_WBM_REG 0x00a34000
#define HAL_CE_WFSS_CE_REG_BASE 0x01b80000
@@ -57,8 +61,10 @@ struct ath12k_base;
/* SW2TCL(x) R0 ring configuration address */
#define HAL_TCL1_RING_CMN_CTRL_REG 0x00000020
#define HAL_TCL1_RING_DSCP_TID_MAP 0x00000240
-#define HAL_TCL1_RING_BASE_LSB 0x00000900
-#define HAL_TCL1_RING_BASE_MSB 0x00000904
+#define HAL_TCL1_RING_BASE_LSB(ab) \
+ ((ab)->hw_params->regs->hal_tcl1_ring_base_lsb)
+#define HAL_TCL1_RING_BASE_MSB(ab) \
+ ((ab)->hw_params->regs->hal_tcl1_ring_base_msb)
#define HAL_TCL1_RING_ID(ab) ((ab)->hw_params->regs->hal_tcl1_ring_id)
#define HAL_TCL1_RING_MISC(ab) \
((ab)->hw_params->regs->hal_tcl1_ring_misc)
@@ -76,30 +82,31 @@ struct ath12k_base;
((ab)->hw_params->regs->hal_tcl1_ring_msi1_base_msb)
#define HAL_TCL1_RING_MSI1_DATA(ab) \
((ab)->hw_params->regs->hal_tcl1_ring_msi1_data)
-#define HAL_TCL2_RING_BASE_LSB 0x00000978
+#define HAL_TCL2_RING_BASE_LSB(ab) \
+ ((ab)->hw_params->regs->hal_tcl2_ring_base_lsb)
#define HAL_TCL_RING_BASE_LSB(ab) \
((ab)->hw_params->regs->hal_tcl_ring_base_lsb)
-#define HAL_TCL1_RING_MSI1_BASE_LSB_OFFSET(ab) \
- (HAL_TCL1_RING_MSI1_BASE_LSB(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_MSI1_BASE_MSB_OFFSET(ab) \
- (HAL_TCL1_RING_MSI1_BASE_MSB(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_MSI1_DATA_OFFSET(ab) \
- (HAL_TCL1_RING_MSI1_DATA(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_BASE_MSB_OFFSET \
- (HAL_TCL1_RING_BASE_MSB - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_ID_OFFSET(ab) \
- (HAL_TCL1_RING_ID(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_CONSR_INT_SETUP_IX0_OFFSET(ab) \
- (HAL_TCL1_RING_CONSUMER_INT_SETUP_IX0(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_CONSR_INT_SETUP_IX1_OFFSET(ab) \
- (HAL_TCL1_RING_CONSUMER_INT_SETUP_IX1(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_TP_ADDR_LSB_OFFSET(ab) \
- (HAL_TCL1_RING_TP_ADDR_LSB(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_TP_ADDR_MSB_OFFSET(ab) \
- (HAL_TCL1_RING_TP_ADDR_MSB(ab) - HAL_TCL1_RING_BASE_LSB)
-#define HAL_TCL1_RING_MISC_OFFSET(ab) \
- (HAL_TCL1_RING_MISC(ab) - HAL_TCL1_RING_BASE_LSB)
+#define HAL_TCL1_RING_MSI1_BASE_LSB_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_MSI1_BASE_LSB(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_MSI1_BASE_MSB_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_MSI1_BASE_MSB(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_MSI1_DATA_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_MSI1_DATA(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_BASE_MSB_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_BASE_MSB(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_ID_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_ID(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_CONSR_INT_SETUP_IX0_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_CONSUMER_INT_SETUP_IX0(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_CONSR_INT_SETUP_IX1_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_CONSUMER_INT_SETUP_IX1(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_TP_ADDR_LSB_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_TP_ADDR_LSB(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_TP_ADDR_MSB_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_TP_ADDR_MSB(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
+#define HAL_TCL1_RING_MISC_OFFSET(ab) ({ typeof(ab) _ab = (ab); \
+ (HAL_TCL1_RING_MISC(_ab) - HAL_TCL1_RING_BASE_LSB(_ab)); })
/* SW2TCL(x) R2 ring pointers (head/tail) address */
#define HAL_TCL1_RING_HP 0x00002000
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index 7ca6ec99fd6e..c755cadb3d4d 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -615,6 +615,9 @@ static const struct ath12k_hw_regs qcn9274_v1_regs = {
.hal_tcl1_ring_msi1_base_msb = 0x0000094c,
.hal_tcl1_ring_msi1_data = 0x00000950,
.hal_tcl_ring_base_lsb = 0x00000b58,
+ .hal_tcl1_ring_base_lsb = 0x00000900,
+ .hal_tcl1_ring_base_msb = 0x00000904,
+ .hal_tcl2_ring_base_lsb = 0x00000978,
/* TCL STATUS ring address */
.hal_tcl_status_ring_base_lsb = 0x00000d38,
@@ -677,6 +680,12 @@ static const struct ath12k_hw_regs qcn9274_v1_regs = {
/* REO status ring address */
.hal_reo_status_ring_base = 0x00000a84,
+
+ /* CE base address */
+ .hal_umac_ce0_src_reg_base = 0x01b80000,
+ .hal_umac_ce0_dest_reg_base = 0x01b81000,
+ .hal_umac_ce1_src_reg_base = 0x01b82000,
+ .hal_umac_ce1_dest_reg_base = 0x01b83000,
};
static const struct ath12k_hw_regs qcn9274_v2_regs = {
@@ -691,6 +700,9 @@ static const struct ath12k_hw_regs qcn9274_v2_regs = {
.hal_tcl1_ring_msi1_base_msb = 0x0000094c,
.hal_tcl1_ring_msi1_data = 0x00000950,
.hal_tcl_ring_base_lsb = 0x00000b58,
+ .hal_tcl1_ring_base_lsb = 0x00000900,
+ .hal_tcl1_ring_base_msb = 0x00000904,
+ .hal_tcl2_ring_base_lsb = 0x00000978,
/* TCL STATUS ring address */
.hal_tcl_status_ring_base_lsb = 0x00000d38,
@@ -757,6 +769,12 @@ static const struct ath12k_hw_regs qcn9274_v2_regs = {
/* REO status ring address */
.hal_reo_status_ring_base = 0x00000aa0,
+
+ /* CE base address */
+ .hal_umac_ce0_src_reg_base = 0x01b80000,
+ .hal_umac_ce0_dest_reg_base = 0x01b81000,
+ .hal_umac_ce1_src_reg_base = 0x01b82000,
+ .hal_umac_ce1_dest_reg_base = 0x01b83000,
};
static const struct ath12k_hw_regs wcn7850_regs = {
@@ -771,6 +789,9 @@ static const struct ath12k_hw_regs wcn7850_regs = {
.hal_tcl1_ring_msi1_base_msb = 0x0000094c,
.hal_tcl1_ring_msi1_data = 0x00000950,
.hal_tcl_ring_base_lsb = 0x00000b58,
+ .hal_tcl1_ring_base_lsb = 0x00000900,
+ .hal_tcl1_ring_base_msb = 0x00000904,
+ .hal_tcl2_ring_base_lsb = 0x00000978,
/* TCL STATUS ring address */
.hal_tcl_status_ring_base_lsb = 0x00000d38,
@@ -833,6 +854,12 @@ static const struct ath12k_hw_regs wcn7850_regs = {
/* REO status ring address */
.hal_reo_status_ring_base = 0x00000a84,
+
+ /* CE base address */
+ .hal_umac_ce0_src_reg_base = 0x01b80000,
+ .hal_umac_ce0_dest_reg_base = 0x01b81000,
+ .hal_umac_ce1_src_reg_base = 0x01b82000,
+ .hal_umac_ce1_dest_reg_base = 0x01b83000,
};
static const struct ath12k_hw_hal_params ath12k_hw_hal_params_qcn9274 = {
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index 8d52182e28ae..36bec99e9e4d 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -293,6 +293,9 @@ struct ath12k_hw_regs {
u32 hal_tcl1_ring_msi1_base_msb;
u32 hal_tcl1_ring_msi1_data;
u32 hal_tcl_ring_base_lsb;
+ u32 hal_tcl1_ring_base_lsb;
+ u32 hal_tcl1_ring_base_msb;
+ u32 hal_tcl2_ring_base_lsb;
u32 hal_tcl_status_ring_base_lsb;
@@ -316,6 +319,11 @@ struct ath12k_hw_regs {
u32 pcie_qserdes_sysclk_en_sel;
u32 pcie_pcs_osc_dtct_config_base;
+ u32 hal_umac_ce0_src_reg_base;
+ u32 hal_umac_ce0_dest_reg_base;
+ u32 hal_umac_ce1_src_reg_base;
+ u32 hal_umac_ce1_dest_reg_base;
+
u32 hal_ppe_rel_ring_base;
u32 hal_reo2_ring_base;
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (3 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 04/22] wifi: ath12k: refactor ath12k_hw_regs structure Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 19:58 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 06/22] wifi: ath12k: add ath12k_hw_ring_mask " Raj Kumar Bhagat
` (17 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S, Raj Kumar Bhagat
From: P Praneesh <quic_ppranees@quicinc.com>
Add register addresses (ath12k_hw_regs) for new ath12k AHB based
WiFi device IPQ5332.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hw.c | 86 +++++++++++++++++++++++++++-
1 file changed, 85 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index c755cadb3d4d..7ef3c8ede34f 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -777,6 +777,90 @@ static const struct ath12k_hw_regs qcn9274_v2_regs = {
.hal_umac_ce1_dest_reg_base = 0x01b83000,
};
+static const struct ath12k_hw_regs ipq5332_regs = {
+ /* SW2TCL(x) R0 ring configuration address */
+ .hal_tcl1_ring_id = 0x00000918,
+ .hal_tcl1_ring_misc = 0x00000920,
+ .hal_tcl1_ring_tp_addr_lsb = 0x0000092c,
+ .hal_tcl1_ring_tp_addr_msb = 0x00000930,
+ .hal_tcl1_ring_consumer_int_setup_ix0 = 0x00000940,
+ .hal_tcl1_ring_consumer_int_setup_ix1 = 0x00000944,
+ .hal_tcl1_ring_msi1_base_lsb = 0x00000958,
+ .hal_tcl1_ring_msi1_base_msb = 0x0000095c,
+ .hal_tcl1_ring_base_lsb = 0x00000910,
+ .hal_tcl1_ring_base_msb = 0x00000914,
+ .hal_tcl1_ring_msi1_data = 0x00000960,
+ .hal_tcl2_ring_base_lsb = 0x00000988,
+ .hal_tcl_ring_base_lsb = 0x00000b68,
+
+ /* TCL STATUS ring address */
+ .hal_tcl_status_ring_base_lsb = 0x00000d48,
+
+ /* REO DEST ring address */
+ .hal_reo2_ring_base = 0x00000578,
+ .hal_reo1_misc_ctrl_addr = 0x00000b9c,
+ .hal_reo1_sw_cookie_cfg0 = 0x0000006c,
+ .hal_reo1_sw_cookie_cfg1 = 0x00000070,
+ .hal_reo1_qdesc_lut_base0 = 0x00000074,
+ .hal_reo1_qdesc_lut_base1 = 0x00000078,
+ .hal_reo1_ring_base_lsb = 0x00000500,
+ .hal_reo1_ring_base_msb = 0x00000504,
+ .hal_reo1_ring_id = 0x00000508,
+ .hal_reo1_ring_misc = 0x00000510,
+ .hal_reo1_ring_hp_addr_lsb = 0x00000514,
+ .hal_reo1_ring_hp_addr_msb = 0x00000518,
+ .hal_reo1_ring_producer_int_setup = 0x00000524,
+ .hal_reo1_ring_msi1_base_lsb = 0x00000548,
+ .hal_reo1_ring_msi1_base_msb = 0x0000054C,
+ .hal_reo1_ring_msi1_data = 0x00000550,
+ .hal_reo1_aging_thres_ix0 = 0x00000B28,
+ .hal_reo1_aging_thres_ix1 = 0x00000B2C,
+ .hal_reo1_aging_thres_ix2 = 0x00000B30,
+ .hal_reo1_aging_thres_ix3 = 0x00000B34,
+
+ /* REO Exception ring address */
+ .hal_reo2_sw0_ring_base = 0x000008c0,
+
+ /* REO Reinject ring address */
+ .hal_sw2reo_ring_base = 0x00000320,
+ .hal_sw2reo1_ring_base = 0x00000398,
+
+ /* REO cmd ring address */
+ .hal_reo_cmd_ring_base = 0x000002A8,
+
+ /* REO status ring address */
+ .hal_reo_status_ring_base = 0x00000aa0,
+
+ /* WBM idle link ring address */
+ .hal_wbm_idle_ring_base_lsb = 0x00000d3c,
+ .hal_wbm_idle_ring_misc_addr = 0x00000d4c,
+ .hal_wbm_r0_idle_list_cntl_addr = 0x00000240,
+ .hal_wbm_r0_idle_list_size_addr = 0x00000244,
+ .hal_wbm_scattered_ring_base_lsb = 0x00000250,
+ .hal_wbm_scattered_ring_base_msb = 0x00000254,
+ .hal_wbm_scattered_desc_head_info_ix0 = 0x00000260,
+ .hal_wbm_scattered_desc_head_info_ix1 = 0x00000264,
+ .hal_wbm_scattered_desc_tail_info_ix0 = 0x00000270,
+ .hal_wbm_scattered_desc_tail_info_ix1 = 0x00000274,
+ .hal_wbm_scattered_desc_ptr_hp_addr = 0x0000027c,
+
+ /* SW2WBM release ring address */
+ .hal_wbm_sw_release_ring_base_lsb = 0x0000037c,
+
+ /* WBM2SW release ring address */
+ .hal_wbm0_release_ring_base_lsb = 0x00000e08,
+ .hal_wbm1_release_ring_base_lsb = 0x00000e80,
+
+ /* PPE release ring address */
+ .hal_ppe_rel_ring_base = 0x0000046c,
+
+ /* CE base address */
+ .hal_umac_ce0_src_reg_base = 0x00740000,
+ .hal_umac_ce0_dest_reg_base = 0x00741000,
+ .hal_umac_ce1_src_reg_base = 0x00742000,
+ .hal_umac_ce1_dest_reg_base = 0x00743000,
+};
+
static const struct ath12k_hw_regs wcn7850_regs = {
/* SW2TCL(x) R0 ring configuration address */
.hal_tcl1_ring_id = 0x00000908,
@@ -1126,7 +1210,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.internal_sleep_clock = false,
.hw_ops = &qcn9274_ops,
- .regs = NULL,
+ .regs = &ipq5332_regs,
.ring_mask = NULL,
.host_ce_config = NULL,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332
2024-10-15 18:26 ` [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332 Raj Kumar Bhagat
@ 2024-10-18 19:58 ` Konrad Dybcio
2024-12-03 9:07 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 19:58 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: P Praneesh <quic_ppranees@quicinc.com>
>
> Add register addresses (ath12k_hw_regs) for new ath12k AHB based
> WiFi device IPQ5332.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
> Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
[...]
> + /* CE base address */
> + .hal_umac_ce0_src_reg_base = 0x00740000,
> + .hal_umac_ce0_dest_reg_base = 0x00741000,
> + .hal_umac_ce1_src_reg_base = 0x00742000,
> + .hal_umac_ce1_dest_reg_base = 0x00743000,
> +};
> +
> static const struct ath12k_hw_regs wcn7850_regs = {
> /* SW2TCL(x) R0 ring configuration address */
> .hal_tcl1_ring_id = 0x00000908,
> @@ -1126,7 +1210,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
> .internal_sleep_clock = false,
>
> .hw_ops = &qcn9274_ops,
> - .regs = NULL,
> + .regs = &ipq5332_regs,
This makes me believe the patches should be reordered (or perhaps
this should be squashed with "add ath12k_hw_params for IPQ5332"?)
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread* Re: [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332
2024-10-18 19:58 ` Konrad Dybcio
@ 2024-12-03 9:07 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-03 9:07 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S
On 10/19/2024 1:28 AM, Konrad Dybcio wrote:
> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>> From: P Praneesh <quic_ppranees@quicinc.com>
>>
>> Add register addresses (ath12k_hw_regs) for new ath12k AHB based
>> WiFi device IPQ5332.
>>
>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
>> Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
>> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>
> [...]
>
>> + /* CE base address */
>> + .hal_umac_ce0_src_reg_base = 0x00740000,
>> + .hal_umac_ce0_dest_reg_base = 0x00741000,
>> + .hal_umac_ce1_src_reg_base = 0x00742000,
>> + .hal_umac_ce1_dest_reg_base = 0x00743000,
>> +};
>> +
>> static const struct ath12k_hw_regs wcn7850_regs = {
>> /* SW2TCL(x) R0 ring configuration address */
>> .hal_tcl1_ring_id = 0x00000908,
>> @@ -1126,7 +1210,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
>> .internal_sleep_clock = false,
>>
>> .hw_ops = &qcn9274_ops,
>> - .regs = NULL,
>> + .regs = &ipq5332_regs,
>
> This makes me believe the patches should be reordered (or perhaps
> this should be squashed with "add ath12k_hw_params for IPQ5332"?)
>
Sure, in next version we will squash patch[2/22] to patch[8/22] into single patch that will add the complete hardware parameters.
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 06/22] wifi: ath12k: add ath12k_hw_ring_mask for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (4 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 05/22] wifi: ath12k: add ath12k_hw_regs for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 19:59 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 07/22] wifi: ath12k: add CE configurations " Raj Kumar Bhagat
` (16 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S, Raj Kumar Bhagat
From: P Praneesh <quic_ppranees@quicinc.com>
Add ath12k_hw_ring_mask for new ath12k AHB based WiFi device IPQ5332.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hw.c | 42 +++++++++++++++++++++++++++-
1 file changed, 41 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index 7ef3c8ede34f..20e4157683a7 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -573,6 +573,46 @@ static const struct ath12k_hw_ring_mask ath12k_hw_ring_mask_qcn9274 = {
},
};
+static const struct ath12k_hw_ring_mask ath12k_hw_ring_mask_ipq5332 = {
+ .tx = {
+ ATH12K_TX_RING_MASK_0,
+ ATH12K_TX_RING_MASK_1,
+ ATH12K_TX_RING_MASK_2,
+ ATH12K_TX_RING_MASK_3,
+ },
+ .rx_mon_dest = {
+ 0, 0, 0, 0, 0, 0, 0, 0,
+ ATH12K_RX_MON_RING_MASK_0,
+ },
+ .rx = {
+ 0, 0, 0, 0,
+ ATH12K_RX_RING_MASK_0,
+ ATH12K_RX_RING_MASK_1,
+ ATH12K_RX_RING_MASK_2,
+ ATH12K_RX_RING_MASK_3,
+ },
+ .rx_err = {
+ 0, 0, 0,
+ ATH12K_RX_ERR_RING_MASK_0,
+ },
+ .rx_wbm_rel = {
+ 0, 0, 0,
+ ATH12K_RX_WBM_REL_RING_MASK_0,
+ },
+ .reo_status = {
+ 0, 0, 0,
+ ATH12K_REO_STATUS_RING_MASK_0,
+ },
+ .host2rxdma = {
+ 0, 0, 0,
+ ATH12K_HOST2RXDMA_RING_MASK_0,
+ },
+ .tx_mon_dest = {
+ ATH12K_TX_MON_RING_MASK_0,
+ ATH12K_TX_MON_RING_MASK_1,
+ },
+};
+
static const struct ath12k_hw_ring_mask ath12k_hw_ring_mask_wcn7850 = {
.tx = {
ATH12K_TX_RING_MASK_0,
@@ -1211,7 +1251,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.hw_ops = &qcn9274_ops,
.regs = &ipq5332_regs,
- .ring_mask = NULL,
+ .ring_mask = &ath12k_hw_ring_mask_ipq5332,
.host_ce_config = NULL,
.ce_count = 0,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 06/22] wifi: ath12k: add ath12k_hw_ring_mask for IPQ5332
2024-10-15 18:26 ` [PATCH v2 06/22] wifi: ath12k: add ath12k_hw_ring_mask " Raj Kumar Bhagat
@ 2024-10-18 19:59 ` Konrad Dybcio
0 siblings, 0 replies; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 19:59 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: P Praneesh <quic_ppranees@quicinc.com>
>
> Add ath12k_hw_ring_mask for new ath12k AHB based WiFi device IPQ5332.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
> Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
Same comment as previous patch, it seems like patch 3 is incomplete
without this and a couple later ones
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 07/22] wifi: ath12k: add CE configurations for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (5 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 06/22] wifi: ath12k: add ath12k_hw_ring_mask " Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 08/22] wifi: ath12k: add ath12k_hw_hal_params " Raj Kumar Bhagat
` (15 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, P Praneesh,
Balamurugan S, Raj Kumar Bhagat
From: P Praneesh <quic_ppranees@quicinc.com>
Add below CE configurations for new ath12k AHB based WiFi device
IPQ5332:
- host ce config (ce_attr)
- target ce config (ce_pipe_config)
- target service to ce map (service_to_pipe)
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Co-developed-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ce.c | 90 +++++++++++
drivers/net/wireless/ath/ath12k/ce.h | 1 +
drivers/net/wireless/ath/ath12k/hw.c | 223 ++++++++++++++++++++++++++-
3 files changed, 308 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/ce.c b/drivers/net/wireless/ath/ath12k/ce.c
index be0d669d31fc..5ffe59c4cd9c 100644
--- a/drivers/net/wireless/ath/ath12k/ce.c
+++ b/drivers/net/wireless/ath/ath12k/ce.c
@@ -219,6 +219,96 @@ const struct ce_attr ath12k_host_ce_config_wcn7850[] = {
};
+const struct ce_attr ath12k_host_ce_config_ipq5332[] = {
+ /* CE0: host->target HTC control and raw streams */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 16,
+ .src_sz_max = 2048,
+ .dest_nentries = 0,
+ },
+ /* CE1: target->host HTT + HTC control */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 0,
+ .src_sz_max = 2048,
+ .dest_nentries = 512,
+ .recv_cb = ath12k_htc_rx_completion_handler,
+ },
+ /* CE2: target->host WMI */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 0,
+ .src_sz_max = 2048,
+ .dest_nentries = 128,
+ .recv_cb = ath12k_htc_rx_completion_handler,
+ },
+ /* CE3: host->target WMI */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 32,
+ .src_sz_max = 2048,
+ .dest_nentries = 0,
+ },
+ /* CE4: host->target HTT */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 2048,
+ .src_sz_max = 256,
+ .dest_nentries = 0,
+ },
+ /* CE5: target -> host PKTLOG */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 0,
+ .src_sz_max = 2048,
+ .dest_nentries = 512,
+ .recv_cb = ath12k_dp_htt_htc_t2h_msg_handler,
+ },
+ /* CE6: Target autonomous HIF_memcpy */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 0,
+ .src_sz_max = 0,
+ .dest_nentries = 0,
+ },
+ /* CE7: CV Prefetch */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 0,
+ .src_sz_max = 0,
+ .dest_nentries = 0,
+ },
+ /* CE8: Target HIF memcpy (Generic HIF memcypy) */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 0,
+ .src_sz_max = 0,
+ .dest_nentries = 0,
+ },
+ /* CE9: WMI logging/CFR/Spectral/Radar */
+ {
+ .flags = CE_ATTR_FLAGS,
+ .src_nentries = 0,
+ .src_sz_max = 2048,
+ .dest_nentries = 128,
+ },
+ /* CE10: Unused */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 0,
+ .src_sz_max = 0,
+ .dest_nentries = 0,
+ },
+ /* CE11: Unused */
+ {
+ .flags = CE_ATTR_FLAGS | CE_ATTR_DIS_INTR,
+ .src_nentries = 0,
+ .src_sz_max = 0,
+ .dest_nentries = 0,
+ },
+};
+
static int ath12k_ce_rx_buf_enqueue_pipe(struct ath12k_ce_pipe *pipe,
struct sk_buff *skb, dma_addr_t paddr)
{
diff --git a/drivers/net/wireless/ath/ath12k/ce.h b/drivers/net/wireless/ath/ath12k/ce.h
index 857bc5f9e946..1d9542e1c21a 100644
--- a/drivers/net/wireless/ath/ath12k/ce.h
+++ b/drivers/net/wireless/ath/ath12k/ce.h
@@ -164,6 +164,7 @@ struct ath12k_ce {
extern const struct ce_attr ath12k_host_ce_config_qcn9274[];
extern const struct ce_attr ath12k_host_ce_config_wcn7850[];
+extern const struct ce_attr ath12k_host_ce_config_ipq5332[];
void ath12k_ce_cleanup_pipes(struct ath12k_base *ab);
void ath12k_ce_rx_replenish_retry(struct timer_list *t);
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index 20e4157683a7..481b06613c1e 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -535,6 +535,217 @@ static const struct service_to_pipe ath12k_target_service_to_ce_map_wlan_wcn7850
},
};
+static const struct ce_pipe_config ath12k_target_ce_config_wlan_ipq5332[] = {
+ /* host->target HTC control and raw streams */
+ {
+ .pipenum = __cpu_to_le32(0),
+ .pipedir = __cpu_to_le32(PIPEDIR_OUT),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* target->host HTT */
+ {
+ .pipenum = __cpu_to_le32(1),
+ .pipedir = __cpu_to_le32(PIPEDIR_IN),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* target->host WMI + HTC control */
+ {
+ .pipenum = __cpu_to_le32(2),
+ .pipedir = __cpu_to_le32(PIPEDIR_IN),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* host->target WMI */
+ {
+ .pipenum = __cpu_to_le32(3),
+ .pipedir = __cpu_to_le32(PIPEDIR_OUT),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* host->target HTT */
+ {
+ .pipenum = __cpu_to_le32(4),
+ .pipedir = __cpu_to_le32(PIPEDIR_OUT),
+ .nentries = __cpu_to_le32(256),
+ .nbytes_max = __cpu_to_le32(256),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS | CE_ATTR_DIS_INTR),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* Target -> host PKTLOG */
+ {
+ .pipenum = __cpu_to_le32(5),
+ .pipedir = __cpu_to_le32(PIPEDIR_IN),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* Reserved for target autonomous HIF_memcpy */
+ {
+ .pipenum = __cpu_to_le32(6),
+ .pipedir = __cpu_to_le32(PIPEDIR_INOUT),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(16384),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* CE7 Reserved for CV Prefetch */
+ {
+ .pipenum = __cpu_to_le32(7),
+ .pipedir = __cpu_to_le32(PIPEDIR_OUT),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* CE8 Reserved for target generic HIF memcpy */
+ {
+ .pipenum = __cpu_to_le32(8),
+ .pipedir = __cpu_to_le32(PIPEDIR_INOUT),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(16384),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* CE9 WMI logging/CFR/Spectral/Radar/ */
+ {
+ .pipenum = __cpu_to_le32(9),
+ .pipedir = __cpu_to_le32(PIPEDIR_IN),
+ .nentries = __cpu_to_le32(32),
+ .nbytes_max = __cpu_to_le32(2048),
+ .flags = __cpu_to_le32(CE_ATTR_FLAGS),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* Unused TBD */
+ {
+ .pipenum = __cpu_to_le32(10),
+ .pipedir = __cpu_to_le32(PIPEDIR_NONE),
+ .nentries = __cpu_to_le32(0),
+ .nbytes_max = __cpu_to_le32(0),
+ .flags = __cpu_to_le32(0),
+ .reserved = __cpu_to_le32(0),
+ },
+ /* Unused TBD */
+ {
+ .pipenum = __cpu_to_le32(11),
+ .pipedir = __cpu_to_le32(PIPEDIR_NONE),
+ .nentries = __cpu_to_le32(0),
+ .nbytes_max = __cpu_to_le32(0),
+ .flags = __cpu_to_le32(0),
+ .reserved = __cpu_to_le32(0),
+ },
+};
+
+static const struct service_to_pipe ath12k_target_service_to_ce_map_wlan_ipq5332[] = {
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_VO),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(3),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_VO),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(2),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_BK),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(3),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_BK),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(2),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_BE),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(3),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_BE),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(2),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_VI),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(3),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_DATA_VI),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(2),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_CONTROL),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(3),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_CONTROL),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(2),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_RSVD_CTRL),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(0),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_RSVD_CTRL),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(1),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_TEST_RAW_STREAMS),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(0),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_TEST_RAW_STREAMS),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(1),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_HTT_DATA_MSG),
+ __cpu_to_le32(PIPEDIR_OUT),
+ __cpu_to_le32(4),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_HTT_DATA_MSG),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(1),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_PKT_LOG),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(5),
+ },
+ {
+ __cpu_to_le32(ATH12K_HTC_SVC_ID_WMI_CONTROL_DIAG),
+ __cpu_to_le32(PIPEDIR_IN),
+ __cpu_to_le32(9),
+ },
+ /* (Additions here) */
+
+ { /* must be last */
+ __cpu_to_le32(0),
+ __cpu_to_le32(0),
+ __cpu_to_le32(0),
+ },
+};
+
static const struct ath12k_hw_ring_mask ath12k_hw_ring_mask_qcn9274 = {
.tx = {
ATH12K_TX_RING_MASK_0,
@@ -1253,12 +1464,12 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.regs = &ipq5332_regs,
.ring_mask = &ath12k_hw_ring_mask_ipq5332,
- .host_ce_config = NULL,
- .ce_count = 0,
- .target_ce_config = NULL,
- .target_ce_count = 0,
- .svc_to_ce_map = NULL,
- .svc_to_ce_map_len = 0,
+ .host_ce_config = ath12k_host_ce_config_ipq5332,
+ .ce_count = 12,
+ .target_ce_config = ath12k_target_ce_config_wlan_ipq5332,
+ .target_ce_count = 12,
+ .svc_to_ce_map = ath12k_target_service_to_ce_map_wlan_ipq5332,
+ .svc_to_ce_map_len = 18,
.hal_params = NULL,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 08/22] wifi: ath12k: add ath12k_hw_hal_params for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (6 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 07/22] wifi: ath12k: add CE configurations " Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332 Raj Kumar Bhagat
` (14 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
Add ath12k_hw_hal_params for new ath12k AHB based WiFi device IPQ5332.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hw.c | 11 ++++++++++-
1 file changed, 10 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index 481b06613c1e..e5e2164c27d2 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1214,6 +1214,15 @@ static const struct ath12k_hw_hal_params ath12k_hw_hal_params_wcn7850 = {
HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW4_EN,
};
+static const struct ath12k_hw_hal_params ath12k_hw_hal_params_ipq5332 = {
+ .rx_buf_rbm = HAL_RX_BUF_RBM_SW3_BM,
+ .wbm2sw_cc_enable = HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW0_EN |
+ HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW1_EN |
+ HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW2_EN |
+ HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW3_EN |
+ HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW4_EN,
+};
+
static const struct ath12k_hw_params ath12k_hw_params[] = {
{
.name = "qcn9274 hw1.0",
@@ -1471,7 +1480,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.svc_to_ce_map = ath12k_target_service_to_ce_map_wlan_ipq5332,
.svc_to_ce_map_len = 18,
- .hal_params = NULL,
+ .hal_params = &ath12k_hw_hal_params_ipq5332,
.rxdma1_enable = false,
.num_rxdma_per_pdev = 1,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (7 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 08/22] wifi: ath12k: add ath12k_hw_hal_params " Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 20:00 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 10/22] wifi: ath12k: add new CMEM read-write ath12k_hif_ops Raj Kumar Bhagat
` (13 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
Current ath12k devices, QCN9274 and WCN7850, supports m3.bin firmware
download through ath12k driver. The new ath12k AHB based device
IPQ5332 supports m3 firmware download through remoteproc driver.
Hence, add new parameter (m3_fw_support) in ath12k_hw_params to avoid
m3 firmware download in IPQ5332.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hw.c | 8 ++++++++
drivers/net/wireless/ath/ath12k/hw.h | 2 ++
drivers/net/wireless/ath/ath12k/qmi.c | 28 ++++++++++++++++-----------
3 files changed, 27 insertions(+), 11 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index e5e2164c27d2..a4e0c21ac4b7 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1299,6 +1299,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.iova_mask = 0,
.supports_aspm = false,
+
+ .m3_fw_support = true,
},
{
.name = "wcn7850 hw2.0",
@@ -1379,6 +1381,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.iova_mask = ATH12K_PCIE_MAX_PAYLOAD_SIZE - 1,
.supports_aspm = true,
+
+ .m3_fw_support = true,
},
{
.name = "qcn9274 hw2.0",
@@ -1455,6 +1459,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.iova_mask = 0,
.supports_aspm = false,
+
+ .m3_fw_support = true,
},
{
.name = "ipq5332 hw1.0",
@@ -1526,6 +1532,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_dynamic_smps_6ghz = false,
.iova_mask = 0,
.supports_aspm = false,
+
+ .m3_fw_support = false,
},
};
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index 36bec99e9e4d..22bd0a91b3e9 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -220,6 +220,8 @@ struct ath12k_hw_params {
bool supports_dynamic_smps_6ghz;
u32 iova_mask;
+
+ bool m3_fw_support;
};
struct ath12k_hw_ops {
diff --git a/drivers/net/wireless/ath/ath12k/qmi.c b/drivers/net/wireless/ath/ath12k/qmi.c
index b93ce9f87f61..668232d83c10 100644
--- a/drivers/net/wireless/ath/ath12k/qmi.c
+++ b/drivers/net/wireless/ath/ath12k/qmi.c
@@ -2080,10 +2080,12 @@ static int ath12k_qmi_host_cap_send(struct ath12k_base *ab)
req.bdf_support_valid = 1;
req.bdf_support = 1;
- req.m3_support_valid = 1;
- req.m3_support = 1;
- req.m3_cache_support_valid = 1;
- req.m3_cache_support = 1;
+ if (ab->hw_params->m3_fw_support) {
+ req.m3_support_valid = 1;
+ req.m3_support = 1;
+ req.m3_cache_support_valid = 1;
+ req.m3_cache_support = 1;
+ }
req.cal_done_valid = 1;
req.cal_done = ab->qmi.cal_done;
@@ -2721,6 +2723,9 @@ static void ath12k_qmi_m3_free(struct ath12k_base *ab)
{
struct m3_mem_region *m3_mem = &ab->qmi.m3_mem;
+ if (!ab->hw_params->m3_fw_support)
+ return;
+
if (!m3_mem->vaddr)
return;
@@ -2799,15 +2804,16 @@ static int ath12k_qmi_wlanfw_m3_info_send(struct ath12k_base *ab)
struct qmi_txn txn;
int ret = 0;
- ret = ath12k_qmi_m3_load(ab);
- if (ret) {
- ath12k_err(ab, "failed to load m3 firmware: %d", ret);
- return ret;
+ if (ab->hw_params->m3_fw_support) {
+ ret = ath12k_qmi_m3_load(ab);
+ if (ret) {
+ ath12k_err(ab, "failed to load m3 firmware: %d", ret);
+ return ret;
+ }
+ req.addr = m3_mem->paddr;
+ req.size = m3_mem->size;
}
- req.addr = m3_mem->paddr;
- req.size = m3_mem->size;
-
ret = qmi_txn_init(&ab->qmi.handle, &txn,
qmi_wlanfw_m3_info_resp_msg_v01_ei, &resp);
if (ret < 0)
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332
2024-10-15 18:26 ` [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332 Raj Kumar Bhagat
@ 2024-10-18 20:00 ` Konrad Dybcio
2024-12-03 9:11 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 20:00 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: Balamurugan S <quic_bselvara@quicinc.com>
>
> Current ath12k devices, QCN9274 and WCN7850, supports m3.bin firmware
> download through ath12k driver. The new ath12k AHB based device
> IPQ5332 supports m3 firmware download through remoteproc driver.
>
> Hence, add new parameter (m3_fw_support) in ath12k_hw_params to avoid
> m3 firmware download in IPQ5332.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
> Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
> drivers/net/wireless/ath/ath12k/hw.c | 8 ++++++++
> drivers/net/wireless/ath/ath12k/hw.h | 2 ++
> drivers/net/wireless/ath/ath12k/qmi.c | 28 ++++++++++++++++-----------
> 3 files changed, 27 insertions(+), 11 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
> index e5e2164c27d2..a4e0c21ac4b7 100644
> --- a/drivers/net/wireless/ath/ath12k/hw.c
> +++ b/drivers/net/wireless/ath/ath12k/hw.c
> @@ -1299,6 +1299,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
> .iova_mask = 0,
>
> .supports_aspm = false,
> +
> + .m3_fw_support = true,
'support for m3 firmware' is not a fitting term.. maybe "needs_m3_fw"?
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread* Re: [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332
2024-10-18 20:00 ` Konrad Dybcio
@ 2024-12-03 9:11 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-03 9:11 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 10/19/2024 1:30 AM, Konrad Dybcio wrote:
> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>> From: Balamurugan S <quic_bselvara@quicinc.com>
>>
>> Current ath12k devices, QCN9274 and WCN7850, supports m3.bin firmware
>> download through ath12k driver. The new ath12k AHB based device
>> IPQ5332 supports m3 firmware download through remoteproc driver.
>>
>> Hence, add new parameter (m3_fw_support) in ath12k_hw_params to avoid
>> m3 firmware download in IPQ5332.
>>
>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
>> Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
>> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>> drivers/net/wireless/ath/ath12k/hw.c | 8 ++++++++
>> drivers/net/wireless/ath/ath12k/hw.h | 2 ++
>> drivers/net/wireless/ath/ath12k/qmi.c | 28 ++++++++++++++++-----------
>> 3 files changed, 27 insertions(+), 11 deletions(-)
>>
>> diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
>> index e5e2164c27d2..a4e0c21ac4b7 100644
>> --- a/drivers/net/wireless/ath/ath12k/hw.c
>> +++ b/drivers/net/wireless/ath/ath12k/hw.c
>> @@ -1299,6 +1299,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
>> .iova_mask = 0,
>>
>> .supports_aspm = false,
>> +
>> + .m3_fw_support = true,
>
> 'support for m3 firmware' is not a fitting term.. maybe "needs_m3_fw"?
>
The name "m3_fw_support" is inherited from ath11k, for ath12k we could
change it to "needs_m3_fw".
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 10/22] wifi: ath12k: add new CMEM read-write ath12k_hif_ops
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (8 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 09/22] wifi: ath12k: avoid m3 firmware download in AHB device IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 11/22] wifi: ath12k: remap CMEM register space for IPQ5332 Raj Kumar Bhagat
` (12 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
In IPQ5332 AHB device, CMEM region is outside of WCSS register block.
Hence, add new ath12k_hif_ops (cmem_read32 and cmem_write32) for
accessing CMEM register.
In PCI devices (QCN9274 and WCN7850), these cmem_read32/cmem_write32
are same as read32/write32 ath12k_hif_ops. Hence, use the same
functions for these new ath12k_hif_ops cmem_read32/cmem_write32.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/dp.c | 10 ++++++----
drivers/net/wireless/ath/ath12k/hif.h | 13 +++++++++++++
drivers/net/wireless/ath/ath12k/pci.c | 2 ++
3 files changed, 21 insertions(+), 4 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/dp.c b/drivers/net/wireless/ath/ath12k/dp.c
index 2ab2a7d45be9..1e9139923838 100644
--- a/drivers/net/wireless/ath/ath12k/dp.c
+++ b/drivers/net/wireless/ath/ath12k/dp.c
@@ -1474,7 +1474,7 @@ static int ath12k_dp_cmem_init(struct ath12k_base *ab,
struct ath12k_dp *dp,
enum ath12k_dp_desc_type type)
{
- u32 cmem_base;
+ u32 cmem_base, data;
int i, start, end;
cmem_base = ab->qmi.dev_mem[ATH12K_QMI_DEVMEM_CMEM_INDEX].start;
@@ -1495,9 +1495,11 @@ static int ath12k_dp_cmem_init(struct ath12k_base *ab,
}
/* Write to PPT in CMEM */
- for (i = start; i < end; i++)
- ath12k_hif_write32(ab, cmem_base + ATH12K_PPT_ADDR_OFFSET(i),
- dp->spt_info[i].paddr >> ATH12K_SPT_4K_ALIGN_OFFSET);
+ for (i = start; i < end; i++) {
+ data = dp->spt_info[i].paddr >> ATH12K_SPT_4K_ALIGN_OFFSET;
+ ath12k_hif_cmem_write32(ab, cmem_base + ATH12K_PPT_ADDR_OFFSET(i),
+ data);
+ }
return 0;
}
diff --git a/drivers/net/wireless/ath/ath12k/hif.h b/drivers/net/wireless/ath/ath12k/hif.h
index 0e53ec269fa4..44e42065b551 100644
--- a/drivers/net/wireless/ath/ath12k/hif.h
+++ b/drivers/net/wireless/ath/ath12k/hif.h
@@ -12,6 +12,8 @@
struct ath12k_hif_ops {
u32 (*read32)(struct ath12k_base *ab, u32 address);
void (*write32)(struct ath12k_base *ab, u32 address, u32 data);
+ u32 (*cmem_read32)(struct ath12k_base *sc, u32 address);
+ void (*cmem_write32)(struct ath12k_base *sc, u32 address, u32 data);
void (*irq_enable)(struct ath12k_base *ab);
void (*irq_disable)(struct ath12k_base *ab);
int (*start)(struct ath12k_base *ab);
@@ -132,6 +134,17 @@ static inline void ath12k_hif_write32(struct ath12k_base *ab, u32 address,
ab->hif.ops->write32(ab, address, data);
}
+static inline u32 ath12k_hif_cmem_read32(struct ath12k_base *ab, u32 address)
+{
+ return ab->hif.ops->cmem_read32(ab, address);
+}
+
+static inline void ath12k_hif_cmem_write32(struct ath12k_base *ab, u32 address,
+ u32 data)
+{
+ ab->hif.ops->cmem_write32(ab, address, data);
+}
+
static inline int ath12k_hif_power_up(struct ath12k_base *ab)
{
if (!ab->hif.ops->power_up)
diff --git a/drivers/net/wireless/ath/ath12k/pci.c b/drivers/net/wireless/ath/ath12k/pci.c
index bd269aa1740b..0c393bc30f92 100644
--- a/drivers/net/wireless/ath/ath12k/pci.c
+++ b/drivers/net/wireless/ath/ath12k/pci.c
@@ -1316,6 +1316,8 @@ static const struct ath12k_hif_ops ath12k_pci_hif_ops = {
.stop = ath12k_pci_stop,
.read32 = ath12k_pci_read32,
.write32 = ath12k_pci_write32,
+ .cmem_read32 = ath12k_pci_read32,
+ .cmem_write32 = ath12k_pci_write32,
.power_down = ath12k_pci_power_down,
.power_up = ath12k_pci_power_up,
.suspend = ath12k_pci_hif_suspend,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 11/22] wifi: ath12k: remap CMEM register space for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (9 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 10/22] wifi: ath12k: add new CMEM read-write ath12k_hif_ops Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 20:09 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses Raj Kumar Bhagat
` (11 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
In IPQ5332 CMEM region is outside of WCSS register block. Hence, add
hardware param cmem_remap for IPQ5332. This parameter would be used
by Ath12k AHB driver to remap the CMEM registers to a new space for
accessing them.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/core.h | 7 +++++++
drivers/net/wireless/ath/ath12k/hal.h | 4 ++++
drivers/net/wireless/ath/ath12k/hw.c | 9 +++++++++
drivers/net/wireless/ath/ath12k/hw.h | 1 +
4 files changed, 21 insertions(+)
diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index 126c9fed1fad..535ad8aeb453 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -722,6 +722,11 @@ struct ath12k_pdev {
struct mlo_timestamp timestamp;
};
+struct cmem_remap {
+ u32 base;
+ u32 size;
+};
+
struct ath12k_fw_pdev {
u32 pdev_id;
u32 phy_id;
@@ -787,6 +792,8 @@ struct ath12k_base {
void __iomem *mem;
unsigned long mem_len;
+ void __iomem *mem_cmem;
+
struct {
enum ath12k_bus bus;
const struct ath12k_hif_ops *ops;
diff --git a/drivers/net/wireless/ath/ath12k/hal.h b/drivers/net/wireless/ath/ath12k/hal.h
index 69579f118b81..2184da129966 100644
--- a/drivers/net/wireless/ath/ath12k/hal.h
+++ b/drivers/net/wireless/ath/ath12k/hal.h
@@ -372,6 +372,10 @@ struct ath12k_base;
* ath12k_hal_rx_desc_get_err().
*/
+#define HAL_IPQ5332_CMEM_REG_BASE 0xC100000
+#define HAL_IPQ5332_CMEM_SIZE 0x40000
+#define HAL_IPQ5332_CMEM_BASE 0x100000
+
enum hal_srng_ring_id {
HAL_SRNG_RING_ID_REO2SW0 = 0,
HAL_SRNG_RING_ID_REO2SW1,
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index a4e0c21ac4b7..bf1b62debacf 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1223,6 +1223,11 @@ static const struct ath12k_hw_hal_params ath12k_hw_hal_params_ipq5332 = {
HAL_WBM_SW_COOKIE_CONV_CFG_WBM2SW4_EN,
};
+static const struct cmem_remap ath12k_cmem_remap_ipq5332 = {
+ .base = HAL_IPQ5332_CMEM_REG_BASE,
+ .size = HAL_IPQ5332_CMEM_SIZE,
+};
+
static const struct ath12k_hw_params ath12k_hw_params[] = {
{
.name = "qcn9274 hw1.0",
@@ -1301,6 +1306,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_aspm = false,
.m3_fw_support = true,
+ .cmem_remap = NULL,
},
{
.name = "wcn7850 hw2.0",
@@ -1383,6 +1389,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_aspm = true,
.m3_fw_support = true,
+ .cmem_remap = NULL,
},
{
.name = "qcn9274 hw2.0",
@@ -1461,6 +1468,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_aspm = false,
.m3_fw_support = true,
+ .cmem_remap = NULL,
},
{
.name = "ipq5332 hw1.0",
@@ -1534,6 +1542,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.supports_aspm = false,
.m3_fw_support = false,
+ .cmem_remap = &ath12k_cmem_remap_ipq5332,
},
};
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index 22bd0a91b3e9..e30bec688b74 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -222,6 +222,7 @@ struct ath12k_hw_params {
u32 iova_mask;
bool m3_fw_support;
+ const struct cmem_remap *cmem_remap;
};
struct ath12k_hw_ops {
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 11/22] wifi: ath12k: remap CMEM register space for IPQ5332
2024-10-15 18:26 ` [PATCH v2 11/22] wifi: ath12k: remap CMEM register space for IPQ5332 Raj Kumar Bhagat
@ 2024-10-18 20:09 ` Konrad Dybcio
0 siblings, 0 replies; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 20:09 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: Balamurugan S <quic_bselvara@quicinc.com>
>
> In IPQ5332 CMEM region is outside of WCSS register block. Hence, add
> hardware param cmem_remap for IPQ5332. This parameter would be used
> by Ath12k AHB driver to remap the CMEM registers to a new space for
> accessing them.
"would be used" isn't written in a very confident tone.
The commit title is misleading, this change doesn't actually do any
remapping.
Why is this region not described in the device tree instead?
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (10 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 11/22] wifi: ath12k: remap CMEM register space for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 20:02 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 13/22] wifi: ath12k: remap CE register space for IPQ5332 Raj Kumar Bhagat
` (10 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
In the current ath12k implementation, the CE addresses
CE_HOST_IE_ADDRESS and CE_HOST_IE_2_ADDRESS are incorrect. These
values were inherited from ath11k, but ath12k does not currently use
them.
However, the Ath12k AHB support relies on these addresses. Therefore,
corrects the CE addresses for ath12k.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ce.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/ce.h b/drivers/net/wireless/ath/ath12k/ce.h
index 1d9542e1c21a..46fd1f49ec4a 100644
--- a/drivers/net/wireless/ath/ath12k/ce.h
+++ b/drivers/net/wireless/ath/ath12k/ce.h
@@ -39,8 +39,8 @@
#define PIPEDIR_INOUT_H2H 4 /* bidirectional, host to host */
/* CE address/mask */
-#define CE_HOST_IE_ADDRESS 0x00A1803C
-#define CE_HOST_IE_2_ADDRESS 0x00A18040
+#define CE_HOST_IE_ADDRESS 0x75804C
+#define CE_HOST_IE_2_ADDRESS 0x758050
#define CE_HOST_IE_3_ADDRESS CE_HOST_IE_ADDRESS
#define CE_HOST_IE_3_SHIFT 0xC
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses
2024-10-15 18:26 ` [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses Raj Kumar Bhagat
@ 2024-10-18 20:02 ` Konrad Dybcio
2024-12-03 9:12 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 20:02 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: Balamurugan S <quic_bselvara@quicinc.com>
>
> In the current ath12k implementation, the CE addresses
> CE_HOST_IE_ADDRESS and CE_HOST_IE_2_ADDRESS are incorrect. These
> values were inherited from ath11k, but ath12k does not currently use
> them.
>
> However, the Ath12k AHB support relies on these addresses. Therefore,
> corrects the CE addresses for ath12k.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
This can be picked up independently of other patches in this patchset,
please try to position such changes at the beginning of series.
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses
2024-10-18 20:02 ` Konrad Dybcio
@ 2024-12-03 9:12 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-03 9:12 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S
On 10/19/2024 1:32 AM, Konrad Dybcio wrote:
> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>> From: Balamurugan S <quic_bselvara@quicinc.com>
>>
>> In the current ath12k implementation, the CE addresses
>> CE_HOST_IE_ADDRESS and CE_HOST_IE_2_ADDRESS are incorrect. These
>> values were inherited from ath11k, but ath12k does not currently use
>> them.
>>
>> However, the Ath12k AHB support relies on these addresses. Therefore,
>> corrects the CE addresses for ath12k.
>>
>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>
> This can be picked up independently of other patches in this patchset,
> please try to position such changes at the beginning of series.
>
Sure, will have this patch in start of the series in the next version.
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 13/22] wifi: ath12k: remap CE register space for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (11 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 12/22] wifi: ath12k: fix incorrect CE addresses Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 14/22] wifi: ath12k: add support for fixed QMI firmware memory Raj Kumar Bhagat
` (9 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
For IPQ5332 CE register space is moved out of WCSS region and the
space is not contiguous. Hence, add hardware params (ce_ie_addr &
ce_remap) for IPQ5332. These parameters would be used by Ath12k
AHB driver to remap the CE registers to a new space for accessing
them.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ce.h | 11 +++++++++++
drivers/net/wireless/ath/ath12k/core.h | 4 ++++
drivers/net/wireless/ath/ath12k/hal.h | 4 ++++
drivers/net/wireless/ath/ath12k/hw.c | 19 +++++++++++++++++++
drivers/net/wireless/ath/ath12k/hw.h | 2 ++
5 files changed, 40 insertions(+)
diff --git a/drivers/net/wireless/ath/ath12k/ce.h b/drivers/net/wireless/ath/ath12k/ce.h
index 46fd1f49ec4a..9dd71056e1c5 100644
--- a/drivers/net/wireless/ath/ath12k/ce.h
+++ b/drivers/net/wireless/ath/ath12k/ce.h
@@ -76,6 +76,17 @@ struct ce_pipe_config {
__le32 reserved;
};
+struct ce_ie_addr {
+ u32 ie1_reg_addr;
+ u32 ie2_reg_addr;
+ u32 ie3_reg_addr;
+};
+
+struct ce_remap {
+ u32 base;
+ u32 size;
+};
+
struct ce_attr {
/* CE_ATTR_* values */
unsigned int flags;
diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index 535ad8aeb453..ec95a9baf36a 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -794,6 +794,10 @@ struct ath12k_base {
void __iomem *mem_cmem;
+ void __iomem *mem_ce;
+ u32 ce_remap_base_addr;
+ bool ce_remap;
+
struct {
enum ath12k_bus bus;
const struct ath12k_hif_ops *ops;
diff --git a/drivers/net/wireless/ath/ath12k/hal.h b/drivers/net/wireless/ath/ath12k/hal.h
index 2184da129966..90756af4c4c2 100644
--- a/drivers/net/wireless/ath/ath12k/hal.h
+++ b/drivers/net/wireless/ath/ath12k/hal.h
@@ -11,6 +11,7 @@
#include "rx_desc.h"
struct ath12k_base;
+#define HAL_CE_REMAP_REG_BASE (ab->ce_remap_base_addr)
#define HAL_LINK_DESC_SIZE (32 << 2)
#define HAL_LINK_DESC_ALIGN 128
@@ -376,6 +377,9 @@ struct ath12k_base;
#define HAL_IPQ5332_CMEM_SIZE 0x40000
#define HAL_IPQ5332_CMEM_BASE 0x100000
+#define HAL_IPQ5332_CE_WFSS_REG_BASE 0x740000
+#define HAL_IPQ5332_CE_SIZE 0x200000
+
enum hal_srng_ring_id {
HAL_SRNG_RING_ID_REO2SW0 = 0,
HAL_SRNG_RING_ID_REO2SW1,
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index bf1b62debacf..4117a4b718e3 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1228,6 +1228,17 @@ static const struct cmem_remap ath12k_cmem_remap_ipq5332 = {
.size = HAL_IPQ5332_CMEM_SIZE,
};
+static const struct ce_ie_addr ath12k_ce_ie_addr_ipq5332 = {
+ .ie1_reg_addr = CE_HOST_IE_ADDRESS,
+ .ie2_reg_addr = CE_HOST_IE_2_ADDRESS,
+ .ie3_reg_addr = CE_HOST_IE_3_ADDRESS,
+};
+
+static const struct ce_remap ath12k_ce_remap_ipq5332 = {
+ .base = HAL_IPQ5332_CE_WFSS_REG_BASE,
+ .size = HAL_IPQ5332_CE_SIZE,
+};
+
static const struct ath12k_hw_params ath12k_hw_params[] = {
{
.name = "qcn9274 hw1.0",
@@ -1307,6 +1318,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.m3_fw_support = true,
.cmem_remap = NULL,
+ .ce_ie_addr = NULL,
+ .ce_remap = NULL,
},
{
.name = "wcn7850 hw2.0",
@@ -1390,6 +1403,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.m3_fw_support = true,
.cmem_remap = NULL,
+ .ce_ie_addr = NULL,
+ .ce_remap = NULL,
},
{
.name = "qcn9274 hw2.0",
@@ -1469,6 +1484,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.m3_fw_support = true,
.cmem_remap = NULL,
+ .ce_ie_addr = NULL,
+ .ce_remap = NULL,
},
{
.name = "ipq5332 hw1.0",
@@ -1543,6 +1560,8 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.m3_fw_support = false,
.cmem_remap = &ath12k_cmem_remap_ipq5332,
+ .ce_ie_addr = &ath12k_ce_ie_addr_ipq5332,
+ .ce_remap = &ath12k_ce_remap_ipq5332,
},
};
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index e30bec688b74..580c7be109e0 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -223,6 +223,8 @@ struct ath12k_hw_params {
bool m3_fw_support;
const struct cmem_remap *cmem_remap;
+ const struct ce_ie_addr *ce_ie_addr;
+ const struct ce_remap *ce_remap;
};
struct ath12k_hw_ops {
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 14/22] wifi: ath12k: add support for fixed QMI firmware memory
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (12 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 13/22] wifi: ath12k: remap CE register space for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter Raj Kumar Bhagat
` (8 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Raj Kumar Bhagat
IPQ5332 firmware supports only fixed QMI firmware memory.
Hence, add support to read reserved fixed memory region from
device-tree and provide the reserved memory segments for
firmware to use during QMI firmware memory request.
Note that the ability to set the fixed memory will be introduced in
a subsequent patch. Currently, the flag remains unset by default,
ensuring that existing chipsets are unaffected.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/core.h | 1 +
drivers/net/wireless/ath/ath12k/qmi.c | 133 +++++++++++++++++++++++--
2 files changed, 128 insertions(+), 6 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index ec95a9baf36a..75bc863ca381 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -212,6 +212,7 @@ enum ath12k_dev_flags {
ATH12K_FLAG_HTC_SUSPEND_COMPLETE,
ATH12K_FLAG_CE_IRQ_ENABLED,
ATH12K_FLAG_EXT_IRQ_ENABLED,
+ ATH12K_FLAG_FIXED_MEM_REGION,
};
struct ath12k_tx_conf {
diff --git a/drivers/net/wireless/ath/ath12k/qmi.c b/drivers/net/wireless/ath/ath12k/qmi.c
index 668232d83c10..ec8859031824 100644
--- a/drivers/net/wireless/ath/ath12k/qmi.c
+++ b/drivers/net/wireless/ath/ath12k/qmi.c
@@ -11,6 +11,8 @@
#include "debug.h"
#include <linux/of.h>
#include <linux/firmware.h>
+#include <linux/of_address.h>
+#include <linux/ioport.h>
#define SLEEP_CLOCK_SELECT_INTERNAL_BIT 0x02
#define HOST_CSTATE_BIT 0x04
@@ -2294,7 +2296,8 @@ static int ath12k_qmi_respond_fw_mem_request(struct ath12k_base *ab)
* failure to firmware and firmware then request multiple blocks of
* small chunk size memory.
*/
- if (ab->qmi.target_mem_delayed) {
+ if (!test_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags) &&
+ ab->qmi.target_mem_delayed) {
delayed = true;
ath12k_dbg(ab, ATH12K_DBG_QMI, "qmi delays mem_request %d\n",
ab->qmi.mem_seg_count);
@@ -2357,6 +2360,11 @@ static void ath12k_qmi_free_target_mem_chunk(struct ath12k_base *ab)
int i;
for (i = 0; i < ab->qmi.mem_seg_count; i++) {
+ if (test_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags) &&
+ ab->qmi.target_mem[i].v.ioaddr) {
+ iounmap(ab->qmi.target_mem[i].v.ioaddr);
+ ab->qmi.target_mem[i].v.ioaddr = NULL;
+ }
if (!ab->qmi.target_mem[i].v.addr)
continue;
@@ -2435,6 +2443,110 @@ static int ath12k_qmi_alloc_target_mem_chunk(struct ath12k_base *ab)
return 0;
}
+static int ath12k_qmi_assign_target_mem_chunk(struct ath12k_base *ab)
+{
+ struct device *dev = ab->dev;
+ struct device_node *mem_node, *dev_node;
+ struct resource res, m3_res;
+ int i, idx, ret;
+
+ for (i = 0, idx = 0; i < ab->qmi.mem_seg_count; i++) {
+ switch (ab->qmi.target_mem[i].type) {
+ case HOST_DDR_REGION_TYPE:
+ mem_node = of_parse_phandle(dev->of_node, "memory-region", 0);
+ if (!mem_node) {
+ ath12k_dbg(ab, ATH12K_DBG_QMI,
+ "memory-region not defined in device-tree\n");
+ ret = -ENODEV;
+ goto out;
+ }
+
+ ret = of_address_to_resource(mem_node, 0, &res);
+ of_node_put(mem_node);
+ if (ret) {
+ ath12k_dbg(ab, ATH12K_DBG_QMI,
+ "fail to get reg from memory-region\n");
+ goto out;
+ }
+
+ if (res.end - res.start + 1 < ab->qmi.target_mem[i].size) {
+ ath12k_dbg(ab, ATH12K_DBG_QMI,
+ "failed to assign mem type %d req size %d avail size %lld\n",
+ ab->qmi.target_mem[i].type,
+ ab->qmi.target_mem[i].size,
+ (res.end - res.start + 1));
+ ret = -EINVAL;
+ goto out;
+ }
+
+ ab->qmi.target_mem[idx].paddr = res.start;
+ ab->qmi.target_mem[idx].v.ioaddr =
+ ioremap(ab->qmi.target_mem[idx].paddr,
+ ab->qmi.target_mem[i].size);
+ if (!ab->qmi.target_mem[idx].v.ioaddr) {
+ ret = -EIO;
+ goto out;
+ }
+ ab->qmi.target_mem[idx].size = ab->qmi.target_mem[i].size;
+ ab->qmi.target_mem[idx].type = ab->qmi.target_mem[i].type;
+ idx++;
+ break;
+ case CALDB_MEM_REGION_TYPE:
+ /* Cold boot calibration is not enabled in Ath12k. Hence,
+ * assign paddr = 0.
+ * Once cold boot calibration is enabled add support to
+ * assign reserved memory from DT.
+ */
+ ab->qmi.target_mem[idx].paddr = 0;
+ ab->qmi.target_mem[idx].v.ioaddr = NULL;
+ ab->qmi.target_mem[idx].size = ab->qmi.target_mem[i].size;
+ ab->qmi.target_mem[idx].type = ab->qmi.target_mem[i].type;
+ idx++;
+ break;
+ case M3_DUMP_REGION_TYPE:
+ dev_node = of_find_node_by_name(NULL, "m3_dump");
+ if (!dev_node || of_address_to_resource(dev_node, 0, &m3_res)) {
+ ath12k_err(ab, "m3_dump not defined in device-tree\n");
+ ret = -EINVAL;
+ goto out;
+ }
+
+ if (m3_res.end - m3_res.start + 1 < ab->qmi.target_mem[i].size) {
+ ath12k_dbg(ab, ATH12K_DBG_QMI,
+ "failed to assign mem type %d req size %d avail size %lld\n",
+ ab->qmi.target_mem[i].type,
+ ab->qmi.target_mem[i].size,
+ (m3_res.end - m3_res.start + 1));
+ ret = -EINVAL;
+ goto out;
+ }
+
+ ab->qmi.target_mem[idx].paddr = m3_res.start;
+ ab->qmi.target_mem[idx].v.ioaddr =
+ ioremap(ab->qmi.target_mem[idx].paddr,
+ ab->qmi.target_mem[i].size);
+ if (!ab->qmi.target_mem[idx].v.ioaddr) {
+ ret = -EIO;
+ goto out;
+ }
+ ab->qmi.target_mem[idx].size = ab->qmi.target_mem[i].size;
+ ab->qmi.target_mem[idx].type = ab->qmi.target_mem[i].type;
+ idx++;
+ break;
+ default:
+ ath12k_warn(ab, "qmi ignore invalid mem req type %d\n",
+ ab->qmi.target_mem[i].type);
+ break;
+ }
+ }
+ ab->qmi.mem_seg_count = idx;
+
+ return 0;
+out:
+ ath12k_qmi_free_target_mem_chunk(ab);
+ return ret;
+}
+
static int ath12k_qmi_request_target_cap(struct ath12k_base *ab)
{
struct qmi_wlanfw_cap_req_msg_v01 req = {};
@@ -3185,11 +3297,20 @@ static void ath12k_qmi_msg_mem_request_cb(struct qmi_handle *qmi_hdl,
msg->mem_seg[i].type, msg->mem_seg[i].size);
}
- ret = ath12k_qmi_alloc_target_mem_chunk(ab);
- if (ret) {
- ath12k_warn(ab, "qmi failed to alloc target memory: %d\n",
- ret);
- return;
+ if (test_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags)) {
+ ret = ath12k_qmi_assign_target_mem_chunk(ab);
+ if (ret) {
+ ath12k_warn(ab, "failed to assign qmi target memory: %d\n",
+ ret);
+ return;
+ }
+ } else {
+ ret = ath12k_qmi_alloc_target_mem_chunk(ab);
+ if (ret) {
+ ath12k_warn(ab, "qmi failed to alloc target memory: %d\n",
+ ret);
+ return;
+ }
}
ath12k_qmi_driver_event_post(qmi, ATH12K_QMI_EVENT_REQUEST_MEM, NULL);
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (13 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 14/22] wifi: ath12k: add support for fixed QMI firmware memory Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-11-04 14:16 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts Raj Kumar Bhagat
` (7 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Raj Kumar Bhagat
The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
memory during QMI memory requests. This memory is part of the
HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
hardware parameter and provide this memory address to the firmware
during QMI memory requests.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/hw.c | 4 ++++
drivers/net/wireless/ath/ath12k/hw.h | 1 +
drivers/net/wireless/ath/ath12k/qmi.c | 8 ++++++++
3 files changed, 13 insertions(+)
diff --git a/drivers/net/wireless/ath/ath12k/hw.c b/drivers/net/wireless/ath/ath12k/hw.c
index 4117a4b718e3..8fc191b0b467 100644
--- a/drivers/net/wireless/ath/ath12k/hw.c
+++ b/drivers/net/wireless/ath/ath12k/hw.c
@@ -1320,6 +1320,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.cmem_remap = NULL,
.ce_ie_addr = NULL,
.ce_remap = NULL,
+ .bdf_addr = 0,
},
{
.name = "wcn7850 hw2.0",
@@ -1405,6 +1406,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.cmem_remap = NULL,
.ce_ie_addr = NULL,
.ce_remap = NULL,
+ .bdf_addr = 0,
},
{
.name = "qcn9274 hw2.0",
@@ -1486,6 +1488,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.cmem_remap = NULL,
.ce_ie_addr = NULL,
.ce_remap = NULL,
+ .bdf_addr = 0,
},
{
.name = "ipq5332 hw1.0",
@@ -1562,6 +1565,7 @@ static const struct ath12k_hw_params ath12k_hw_params[] = {
.cmem_remap = &ath12k_cmem_remap_ipq5332,
.ce_ie_addr = &ath12k_ce_ie_addr_ipq5332,
.ce_remap = &ath12k_ce_remap_ipq5332,
+ .bdf_addr = 0x4B500000,
},
};
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index 580c7be109e0..038fe1b30d11 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -225,6 +225,7 @@ struct ath12k_hw_params {
const struct cmem_remap *cmem_remap;
const struct ce_ie_addr *ce_ie_addr;
const struct ce_remap *ce_remap;
+ u32 bdf_addr;
};
struct ath12k_hw_ops {
diff --git a/drivers/net/wireless/ath/ath12k/qmi.c b/drivers/net/wireless/ath/ath12k/qmi.c
index ec8859031824..b5cad6656722 100644
--- a/drivers/net/wireless/ath/ath12k/qmi.c
+++ b/drivers/net/wireless/ath/ath12k/qmi.c
@@ -2491,6 +2491,14 @@ static int ath12k_qmi_assign_target_mem_chunk(struct ath12k_base *ab)
ab->qmi.target_mem[idx].type = ab->qmi.target_mem[i].type;
idx++;
break;
+ case BDF_MEM_REGION_TYPE:
+ ab->qmi.target_mem[idx].paddr = ab->hw_params->bdf_addr;
+ ab->qmi.target_mem[idx].v.ioaddr = NULL;
+ ab->qmi.target_mem[idx].size = ab->qmi.target_mem[i].size;
+ ab->qmi.target_mem[idx].type = ab->qmi.target_mem[i].type;
+ idx++;
+ break;
+
case CALDB_MEM_REGION_TYPE:
/* Cold boot calibration is not enabled in Ath12k. Hence,
* assign paddr = 0.
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-10-15 18:26 ` [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter Raj Kumar Bhagat
@ 2024-11-04 14:16 ` Konrad Dybcio
2024-12-03 9:18 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-11-04 14:16 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
> memory during QMI memory requests. This memory is part of the
> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
> hardware parameter and provide this memory address to the firmware
> during QMI memory requests.
Sounds like something to put in the device tree, no?
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-11-04 14:16 ` Konrad Dybcio
@ 2024-12-03 9:18 ` Raj Kumar Bhagat
2024-12-05 17:42 ` Konrad Dybcio
0 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-03 9:18 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>> memory during QMI memory requests. This memory is part of the
>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>> hardware parameter and provide this memory address to the firmware
>> during QMI memory requests.
>
> Sounds like something to put in the device tree, no?
>
This BDF memory address is the RAM offset. We did add this in device tree in
version 1. This is removed from device tree in v2 based on the review comment that
DT should not store RAM offset.
refer below link:
Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-03 9:18 ` Raj Kumar Bhagat
@ 2024-12-05 17:42 ` Konrad Dybcio
2024-12-06 4:34 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-12-05 17:42 UTC (permalink / raw)
To: Raj Kumar Bhagat, Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>> memory during QMI memory requests. This memory is part of the
>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>> hardware parameter and provide this memory address to the firmware
>>> during QMI memory requests.
>>
>> Sounds like something to put in the device tree, no?
>>
>
> This BDF memory address is the RAM offset. We did add this in device tree in
> version 1. This is removed from device tree in v2 based on the review comment that
> DT should not store RAM offset.
>
> refer below link:
> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
Right, I think this could be something under /reserved-memory instead
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-05 17:42 ` Konrad Dybcio
@ 2024-12-06 4:34 ` Raj Kumar Bhagat
2024-12-06 10:49 ` Konrad Dybcio
0 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-06 4:34 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 12/5/2024 11:12 PM, Konrad Dybcio wrote:
> On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
>> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>>> memory during QMI memory requests. This memory is part of the
>>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>>> hardware parameter and provide this memory address to the firmware
>>>> during QMI memory requests.
>>>
>>> Sounds like something to put in the device tree, no?
>>>
>>
>> This BDF memory address is the RAM offset. We did add this in device tree in
>> version 1. This is removed from device tree in v2 based on the review comment that
>> DT should not store RAM offset.
>>
>> refer below link:
>> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
>
> Right, I think this could be something under /reserved-memory instead
>
Thanks for the suggestion. However, the BDF_MEM_REGION_TYPE is already within the
memory reserved for HOST_DDR_REGION_TYPE through /reserved-memory. Therefore, reserving
the memory for BDF_MEM_REGION_TYPE again in the Device Tree (DT) will cause a warning
for 'overlapping memory reservation'.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-06 4:34 ` Raj Kumar Bhagat
@ 2024-12-06 10:49 ` Konrad Dybcio
2024-12-09 4:23 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-12-06 10:49 UTC (permalink / raw)
To: Raj Kumar Bhagat, Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 6.12.2024 5:34 AM, Raj Kumar Bhagat wrote:
> On 12/5/2024 11:12 PM, Konrad Dybcio wrote:
>> On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
>>> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>>>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>>>> memory during QMI memory requests. This memory is part of the
>>>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>>>> hardware parameter and provide this memory address to the firmware
>>>>> during QMI memory requests.
>>>>
>>>> Sounds like something to put in the device tree, no?
>>>>
>>>
>>> This BDF memory address is the RAM offset. We did add this in device tree in
>>> version 1. This is removed from device tree in v2 based on the review comment that
>>> DT should not store RAM offset.
>>>
>>> refer below link:
>>> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
>>
>> Right, I think this could be something under /reserved-memory instead
>>
>
> Thanks for the suggestion. However, the BDF_MEM_REGION_TYPE is already within the
> memory reserved for HOST_DDR_REGION_TYPE through /reserved-memory. Therefore, reserving
> the memory for BDF_MEM_REGION_TYPE again in the Device Tree (DT) will cause a warning
> for 'overlapping memory reservation'.
Then you can grab a handle to it with of_reserved_mem_lookup()
and of_reserved_mem_device_init_by_idx()
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-06 10:49 ` Konrad Dybcio
@ 2024-12-09 4:23 ` Raj Kumar Bhagat
2024-12-13 0:18 ` Konrad Dybcio
0 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-09 4:23 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 12/6/2024 4:19 PM, Konrad Dybcio wrote:
> On 6.12.2024 5:34 AM, Raj Kumar Bhagat wrote:
>> On 12/5/2024 11:12 PM, Konrad Dybcio wrote:
>>> On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
>>>> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>>>>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>>>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>>>>> memory during QMI memory requests. This memory is part of the
>>>>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>>>>> hardware parameter and provide this memory address to the firmware
>>>>>> during QMI memory requests.
>>>>>
>>>>> Sounds like something to put in the device tree, no?
>>>>>
>>>>
>>>> This BDF memory address is the RAM offset. We did add this in device tree in
>>>> version 1. This is removed from device tree in v2 based on the review comment that
>>>> DT should not store RAM offset.
>>>>
>>>> refer below link:
>>>> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
>>>
>>> Right, I think this could be something under /reserved-memory instead
>>>
>>
>> Thanks for the suggestion. However, the BDF_MEM_REGION_TYPE is already within the
>> memory reserved for HOST_DDR_REGION_TYPE through /reserved-memory. Therefore, reserving
>> the memory for BDF_MEM_REGION_TYPE again in the Device Tree (DT) will cause a warning
>> for 'overlapping memory reservation'.
>
> Then you can grab a handle to it with of_reserved_mem_lookup()
> and of_reserved_mem_device_init_by_idx()
>
The memory HOST_DDR_REGION_TYPE is a bigger memory around 43MB, while the memory
BDF_MEM_REGION_TYPE is smaller around 256KB within HOST_DDR_REGION_TYPE, Using the
above mentioned API we still have to store the offset in ath12k to point at memory
BDF_MEM_REGION_TYPE from the start of HOST_DDR_REGION_TYPE.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-09 4:23 ` Raj Kumar Bhagat
@ 2024-12-13 0:18 ` Konrad Dybcio
2024-12-13 4:29 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Konrad Dybcio @ 2024-12-13 0:18 UTC (permalink / raw)
To: Raj Kumar Bhagat, Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 9.12.2024 5:23 AM, Raj Kumar Bhagat wrote:
> On 12/6/2024 4:19 PM, Konrad Dybcio wrote:
>> On 6.12.2024 5:34 AM, Raj Kumar Bhagat wrote:
>>> On 12/5/2024 11:12 PM, Konrad Dybcio wrote:
>>>> On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
>>>>> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>>>>>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>>>>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>>>>>> memory during QMI memory requests. This memory is part of the
>>>>>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>>>>>> hardware parameter and provide this memory address to the firmware
>>>>>>> during QMI memory requests.
>>>>>>
>>>>>> Sounds like something to put in the device tree, no?
>>>>>>
>>>>>
>>>>> This BDF memory address is the RAM offset. We did add this in device tree in
>>>>> version 1. This is removed from device tree in v2 based on the review comment that
>>>>> DT should not store RAM offset.
>>>>>
>>>>> refer below link:
>>>>> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
>>>>
>>>> Right, I think this could be something under /reserved-memory instead
>>>>
>>>
>>> Thanks for the suggestion. However, the BDF_MEM_REGION_TYPE is already within the
>>> memory reserved for HOST_DDR_REGION_TYPE through /reserved-memory. Therefore, reserving
>>> the memory for BDF_MEM_REGION_TYPE again in the Device Tree (DT) will cause a warning
>>> for 'overlapping memory reservation'.
>>
>> Then you can grab a handle to it with of_reserved_mem_lookup()
>> and of_reserved_mem_device_init_by_idx()
>>
>
> The memory HOST_DDR_REGION_TYPE is a bigger memory around 43MB, while the memory
> BDF_MEM_REGION_TYPE is smaller around 256KB within HOST_DDR_REGION_TYPE, Using the
> above mentioned API we still have to store the offset in ath12k to point at memory
> BDF_MEM_REGION_TYPE from the start of HOST_DDR_REGION_TYPE.
That's still way better than hardcoding platform specifics in the common
driver
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter
2024-12-13 0:18 ` Konrad Dybcio
@ 2024-12-13 4:29 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-13 4:29 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 12/13/2024 5:48 AM, Konrad Dybcio wrote:
> On 9.12.2024 5:23 AM, Raj Kumar Bhagat wrote:
>> On 12/6/2024 4:19 PM, Konrad Dybcio wrote:
>>> On 6.12.2024 5:34 AM, Raj Kumar Bhagat wrote:
>>>> On 12/5/2024 11:12 PM, Konrad Dybcio wrote:
>>>>> On 3.12.2024 10:18 AM, Raj Kumar Bhagat wrote:
>>>>>> On 11/4/2024 7:46 PM, Konrad Dybcio wrote:
>>>>>>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>>>>>>> The Ath2k AHB device (IPQ5332) firmware requests BDF_MEM_REGION_TYPE
>>>>>>>> memory during QMI memory requests. This memory is part of the
>>>>>>>> HOST_DDR_REGION_TYPE. Therefore, add the BDF memory address to the
>>>>>>>> hardware parameter and provide this memory address to the firmware
>>>>>>>> during QMI memory requests.
>>>>>>>
>>>>>>> Sounds like something to put in the device tree, no?
>>>>>>>
>>>>>>
>>>>>> This BDF memory address is the RAM offset. We did add this in device tree in
>>>>>> version 1. This is removed from device tree in v2 based on the review comment that
>>>>>> DT should not store RAM offset.
>>>>>>
>>>>>> refer below link:
>>>>>> Link: https://lore.kernel.org/all/f8cd9c3d-47e1-4709-9334-78e4790acef0@kernel.org/
>>>>>
>>>>> Right, I think this could be something under /reserved-memory instead
>>>>>
>>>>
>>>> Thanks for the suggestion. However, the BDF_MEM_REGION_TYPE is already within the
>>>> memory reserved for HOST_DDR_REGION_TYPE through /reserved-memory. Therefore, reserving
>>>> the memory for BDF_MEM_REGION_TYPE again in the Device Tree (DT) will cause a warning
>>>> for 'overlapping memory reservation'.
>>>
>>> Then you can grab a handle to it with of_reserved_mem_lookup()
>>> and of_reserved_mem_device_init_by_idx()
>>>
>>
>> The memory HOST_DDR_REGION_TYPE is a bigger memory around 43MB, while the memory
>> BDF_MEM_REGION_TYPE is smaller around 256KB within HOST_DDR_REGION_TYPE, Using the
>> above mentioned API we still have to store the offset in ath12k to point at memory
>> BDF_MEM_REGION_TYPE from the start of HOST_DDR_REGION_TYPE.
>
> That's still way better than hardcoding platform specifics in the common
> driver
>
Sure, I agree. I'll update in latest version to store the offset for BDF_MEM_REGION_TYPE.
Thanks!
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (14 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 15/22] wifi: ath12k: add BDF address in hardware parameter Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-21 9:06 ` Kalle Valo
2024-10-15 18:26 ` [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332 Raj Kumar Bhagat
` (6 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Raj Kumar Bhagat
Currently in Ath12k, tasklet is used to handle the BH context of CE
interrupts. However the tasklet is marked deprecated and has some
design flaws. To replace tasklets, BH workqueue support has been
added. BH workqueue behaves similarly to regular workqueues except
that the queued work items are executed in the BH context.
Hence, convert the tasklet to BH workqueue for handling CE interrupts
in the BH context.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ce.h | 2 +-
drivers/net/wireless/ath/ath12k/pci.c | 14 +++++++-------
2 files changed, 8 insertions(+), 8 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/ce.h b/drivers/net/wireless/ath/ath12k/ce.h
index 9dd71056e1c5..75c9f8f1f7c6 100644
--- a/drivers/net/wireless/ath/ath12k/ce.h
+++ b/drivers/net/wireless/ath/ath12k/ce.h
@@ -159,7 +159,7 @@ struct ath12k_ce_pipe {
void (*send_cb)(struct ath12k_ce_pipe *pipe);
void (*recv_cb)(struct ath12k_base *ab, struct sk_buff *skb);
- struct tasklet_struct intr_tq;
+ struct work_struct intr_wq;
struct ath12k_ce_ring *src_ring;
struct ath12k_ce_ring *dest_ring;
struct ath12k_ce_ring *status_ring;
diff --git a/drivers/net/wireless/ath/ath12k/pci.c b/drivers/net/wireless/ath/ath12k/pci.c
index 0c393bc30f92..9a50d813e9b4 100644
--- a/drivers/net/wireless/ath/ath12k/pci.c
+++ b/drivers/net/wireless/ath/ath12k/pci.c
@@ -425,9 +425,9 @@ static void ath12k_pci_sync_ce_irqs(struct ath12k_base *ab)
}
}
-static void ath12k_pci_ce_tasklet(struct tasklet_struct *t)
+static void ath12k_pci_ce_workqueue(struct work_struct *work)
{
- struct ath12k_ce_pipe *ce_pipe = from_tasklet(ce_pipe, t, intr_tq);
+ struct ath12k_ce_pipe *ce_pipe = from_work(ce_pipe, work, intr_wq);
int irq_idx = ATH12K_PCI_IRQ_CE0_OFFSET + ce_pipe->pipe_num;
ath12k_ce_per_engine_service(ce_pipe->ab, ce_pipe->pipe_num);
@@ -449,7 +449,7 @@ static irqreturn_t ath12k_pci_ce_interrupt_handler(int irq, void *arg)
disable_irq_nosync(ab->irq_num[irq_idx]);
- tasklet_schedule(&ce_pipe->intr_tq);
+ queue_work(system_bh_wq, &ce_pipe->intr_wq);
return IRQ_HANDLED;
}
@@ -675,7 +675,7 @@ static int ath12k_pci_config_irq(struct ath12k_base *ab)
irq_idx = ATH12K_PCI_IRQ_CE0_OFFSET + i;
- tasklet_setup(&ce_pipe->intr_tq, ath12k_pci_ce_tasklet);
+ INIT_WORK(&ce_pipe->intr_wq, ath12k_pci_ce_workqueue);
ret = request_irq(irq, ath12k_pci_ce_interrupt_handler,
ab_pci->irq_flags, irq_name[irq_idx],
@@ -962,7 +962,7 @@ static void ath12k_pci_aspm_restore(struct ath12k_pci *ab_pci)
PCI_EXP_LNKCTL_ASPMC);
}
-static void ath12k_pci_kill_tasklets(struct ath12k_base *ab)
+static void ath12k_pci_cancel_workqueue(struct ath12k_base *ab)
{
int i;
@@ -972,7 +972,7 @@ static void ath12k_pci_kill_tasklets(struct ath12k_base *ab)
if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
continue;
- tasklet_kill(&ce_pipe->intr_tq);
+ cancel_work_sync(&ce_pipe->intr_wq);
}
}
@@ -980,7 +980,7 @@ static void ath12k_pci_ce_irq_disable_sync(struct ath12k_base *ab)
{
ath12k_pci_ce_irqs_disable(ab);
ath12k_pci_sync_ce_irqs(ab);
- ath12k_pci_kill_tasklets(ab);
+ ath12k_pci_cancel_workqueue(ab);
}
int ath12k_pci_map_service_to_pipe(struct ath12k_base *ab, u16 service_id,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts
2024-10-15 18:26 ` [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts Raj Kumar Bhagat
@ 2024-10-21 9:06 ` Kalle Valo
2024-10-22 7:13 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Kalle Valo @ 2024-10-21 9:06 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
Raj Kumar Bhagat <quic_rajkbhag@quicinc.com> writes:
> Currently in Ath12k, tasklet is used to handle the BH context of CE
> interrupts. However the tasklet is marked deprecated and has some
> design flaws. To replace tasklets, BH workqueue support has been
> added. BH workqueue behaves similarly to regular workqueues except
> that the queued work items are executed in the BH context.
>
> Hence, convert the tasklet to BH workqueue for handling CE interrupts
> in the BH context.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
22 patches is a lot and I'm not going to look at this in detail, please
reduce your patchset size. 10-12 patches is recommended. For example,
this could be easily submitted separately.
--
https://patchwork.kernel.org/project/linux-wireless/list/
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts
2024-10-21 9:06 ` Kalle Valo
@ 2024-10-22 7:13 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-22 7:13 UTC (permalink / raw)
To: Kalle Valo
Cc: ath12k, linux-wireless, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm
On 10/21/2024 2:36 PM, Kalle Valo wrote:
> Raj Kumar Bhagat <quic_rajkbhag@quicinc.com> writes:
>
>> Currently in Ath12k, tasklet is used to handle the BH context of CE
>> interrupts. However the tasklet is marked deprecated and has some
>> design flaws. To replace tasklets, BH workqueue support has been
>> added. BH workqueue behaves similarly to regular workqueues except
>> that the queued work items are executed in the BH context.
>>
>> Hence, convert the tasklet to BH workqueue for handling CE interrupts
>> in the BH context.
>>
>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>
> 22 patches is a lot and I'm not going to look at this in detail, please
> reduce your patchset size. 10-12 patches is recommended. For example,
> this could be easily submitted separately.
>
Sure Kalle,
We are working on other review comments in this series, in the next version
we will reduce the number of patches.
This patch we will be sending separately as v3.
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (15 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 16/22] wifi: ath12k: convert tasklet to BH workqueue for CE interrupts Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-18 20:29 ` Konrad Dybcio
2024-10-15 18:26 ` [PATCH v2 18/22] wifi: ath12k: Power up root PD Raj Kumar Bhagat
` (5 subsequent siblings)
22 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
Add Initial Ath12k AHB driver support for IPQ5332. IPQ5332 is AHB
based IEEE802.11be 2 GHz 2x2 WiFi device.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Co-developed-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ahb.c | 950 +++++++++++++++++++++++++
drivers/net/wireless/ath/ath12k/ahb.h | 37 +
drivers/net/wireless/ath/ath12k/core.h | 4 +
drivers/net/wireless/ath/ath12k/hw.h | 1 +
4 files changed, 992 insertions(+)
create mode 100644 drivers/net/wireless/ath/ath12k/ahb.c
create mode 100644 drivers/net/wireless/ath/ath12k/ahb.h
diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
new file mode 100644
index 000000000000..7049efdae38c
--- /dev/null
+++ b/drivers/net/wireless/ath/ath12k/ahb.c
@@ -0,0 +1,950 @@
+// SPDX-License-Identifier: BSD-3-Clause-Clear
+/*
+ * Copyright (c) 2018-2019 The Linux Foundation. All rights reserved.
+ * Copyright (c) 2022-2024 Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+
+#include <linux/dma-mapping.h>
+#include <linux/iommu.h>
+#include <linux/module.h>
+#include <linux/of.h>
+#include <linux/of_address.h>
+#include <linux/of_device.h>
+#include <linux/platform_device.h>
+#include <linux/remoteproc.h>
+#include <linux/soc/qcom/smem.h>
+#include <linux/soc/qcom/smem_state.h>
+#include "ahb.h"
+#include "debug.h"
+#include "hif.h"
+
+static const struct of_device_id ath12k_ahb_of_match[] = {
+ { .compatible = "qcom,ipq5332-wifi",
+ .data = (void *)ATH12K_HW_IPQ5332_HW10,
+ },
+ { }
+};
+
+MODULE_DEVICE_TABLE(of, ath12k_ahb_of_match);
+
+#define ATH12K_IRQ_CE0_OFFSET 4
+
+static const char *irq_name[ATH12K_IRQ_NUM_MAX] = {
+ "misc-pulse1",
+ "misc-latch",
+ "sw-exception",
+ "watchdog",
+ "ce0",
+ "ce1",
+ "ce2",
+ "ce3",
+ "ce4",
+ "ce5",
+ "ce6",
+ "ce7",
+ "ce8",
+ "ce9",
+ "ce10",
+ "ce11",
+ "host2wbm-desc-feed",
+ "host2reo-re-injection",
+ "host2reo-command",
+ "host2rxdma-monitor-ring3",
+ "host2rxdma-monitor-ring2",
+ "host2rxdma-monitor-ring1",
+ "reo2ost-exception",
+ "wbm2host-rx-release",
+ "reo2host-status",
+ "reo2host-destination-ring4",
+ "reo2host-destination-ring3",
+ "reo2host-destination-ring2",
+ "reo2host-destination-ring1",
+ "rxdma2host-monitor-destination-mac3",
+ "rxdma2host-monitor-destination-mac2",
+ "rxdma2host-monitor-destination-mac1",
+ "ppdu-end-interrupts-mac3",
+ "ppdu-end-interrupts-mac2",
+ "ppdu-end-interrupts-mac1",
+ "rxdma2host-monitor-status-ring-mac3",
+ "rxdma2host-monitor-status-ring-mac2",
+ "rxdma2host-monitor-status-ring-mac1",
+ "host2rxdma-host-buf-ring-mac3",
+ "host2rxdma-host-buf-ring-mac2",
+ "host2rxdma-host-buf-ring-mac1",
+ "rxdma2host-destination-ring-mac3",
+ "rxdma2host-destination-ring-mac2",
+ "rxdma2host-destination-ring-mac1",
+ "host2tcl-input-ring4",
+ "host2tcl-input-ring3",
+ "host2tcl-input-ring2",
+ "host2tcl-input-ring1",
+ "wbm2host-tx-completions-ring4",
+ "wbm2host-tx-completions-ring3",
+ "wbm2host-tx-completions-ring2",
+ "wbm2host-tx-completions-ring1",
+ "tcl2host-status-ring",
+};
+
+enum ext_irq_num {
+ host2wbm_desc_feed = 16,
+ host2reo_re_injection,
+ host2reo_command,
+ host2rxdma_monitor_ring3,
+ host2rxdma_monitor_ring2,
+ host2rxdma_monitor_ring1,
+ reo2host_exception,
+ wbm2host_rx_release,
+ reo2host_status,
+ reo2host_destination_ring4,
+ reo2host_destination_ring3,
+ reo2host_destination_ring2,
+ reo2host_destination_ring1,
+ rxdma2host_monitor_destination_mac3,
+ rxdma2host_monitor_destination_mac2,
+ rxdma2host_monitor_destination_mac1,
+ ppdu_end_interrupts_mac3,
+ ppdu_end_interrupts_mac2,
+ ppdu_end_interrupts_mac1,
+ rxdma2host_monitor_status_ring_mac3,
+ rxdma2host_monitor_status_ring_mac2,
+ rxdma2host_monitor_status_ring_mac1,
+ host2rxdma_host_buf_ring_mac3,
+ host2rxdma_host_buf_ring_mac2,
+ host2rxdma_host_buf_ring_mac1,
+ rxdma2host_destination_ring_mac3,
+ rxdma2host_destination_ring_mac2,
+ rxdma2host_destination_ring_mac1,
+ host2tcl_input_ring4,
+ host2tcl_input_ring3,
+ host2tcl_input_ring2,
+ host2tcl_input_ring1,
+ wbm2host_tx_completions_ring4,
+ wbm2host_tx_completions_ring3,
+ wbm2host_tx_completions_ring2,
+ wbm2host_tx_completions_ring1,
+ tcl2host_status_ring,
+};
+
+static u32 ath12k_ahb_cmem_read32(struct ath12k_base *ab, u32 offset)
+{
+ offset = offset - HAL_IPQ5332_CMEM_BASE;
+ return ioread32(ab->mem_cmem + offset);
+}
+
+static void ath12k_ahb_cmem_write32(struct ath12k_base *ab, u32 offset,
+ u32 value)
+{
+ offset = offset - HAL_IPQ5332_CMEM_BASE;
+ iowrite32(value, ab->mem_cmem + offset);
+}
+
+static u32 ath12k_ahb_read32(struct ath12k_base *ab, u32 offset)
+{
+ if (ab->ce_remap && offset < HAL_SEQ_WCSS_UMAC_OFFSET) {
+ offset = offset - HAL_CE_REMAP_REG_BASE;
+ return ioread32(ab->mem_ce + offset);
+ }
+ return ioread32(ab->mem + offset);
+}
+
+static void ath12k_ahb_write32(struct ath12k_base *ab, u32 offset,
+ u32 value)
+{
+ if (ab->ce_remap && offset < HAL_SEQ_WCSS_UMAC_OFFSET) {
+ offset = offset - HAL_CE_REMAP_REG_BASE;
+ iowrite32(value, ab->mem_ce + offset);
+ } else {
+ iowrite32(value, ab->mem + offset);
+ }
+}
+
+static void ath12k_ahb_cancel_workqueue(struct ath12k_base *ab)
+{
+ int i;
+
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ struct ath12k_ce_pipe *ce_pipe = &ab->ce.ce_pipe[i];
+
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+
+ cancel_work_sync(&ce_pipe->intr_wq);
+ }
+}
+
+static void ath12k_ahb_ext_grp_disable(struct ath12k_ext_irq_grp *irq_grp)
+{
+ int i;
+
+ for (i = 0; i < irq_grp->num_irq; i++)
+ disable_irq_nosync(irq_grp->ab->irq_num[irq_grp->irqs[i]]);
+}
+
+static void __ath12k_ahb_ext_irq_disable(struct ath12k_base *ab)
+{
+ int i;
+
+ for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
+ struct ath12k_ext_irq_grp *irq_grp = &ab->ext_irq_grp[i];
+
+ ath12k_ahb_ext_grp_disable(irq_grp);
+ if (irq_grp->napi_enabled) {
+ napi_synchronize(&irq_grp->napi);
+ napi_disable(&irq_grp->napi);
+ irq_grp->napi_enabled = false;
+ }
+ }
+}
+
+static void ath12k_ahb_ext_grp_enable(struct ath12k_ext_irq_grp *irq_grp)
+{
+ int i;
+
+ for (i = 0; i < irq_grp->num_irq; i++)
+ enable_irq(irq_grp->ab->irq_num[irq_grp->irqs[i]]);
+}
+
+static void ath12k_ahb_setbit32(struct ath12k_base *ab, u8 bit, u32 offset)
+{
+ u32 val;
+
+ val = ath12k_ahb_read32(ab, offset);
+ ath12k_ahb_write32(ab, offset, val | BIT(bit));
+}
+
+static void ath12k_ahb_clearbit32(struct ath12k_base *ab, u8 bit, u32 offset)
+{
+ u32 val;
+
+ val = ath12k_ahb_read32(ab, offset);
+ ath12k_ahb_write32(ab, offset, val & ~BIT(bit));
+}
+
+static void ath12k_ahb_ce_irq_enable(struct ath12k_base *ab, u16 ce_id)
+{
+ const struct ce_attr *ce_attr;
+ const struct ce_ie_addr *ce_ie_addr = ab->hw_params->ce_ie_addr;
+ u32 ie1_reg_addr, ie2_reg_addr, ie3_reg_addr;
+
+ ie1_reg_addr = ce_ie_addr->ie1_reg_addr;
+ ie2_reg_addr = ce_ie_addr->ie2_reg_addr;
+ ie3_reg_addr = ce_ie_addr->ie3_reg_addr;
+
+ ce_attr = &ab->hw_params->host_ce_config[ce_id];
+ if (ce_attr->src_nentries)
+ ath12k_ahb_setbit32(ab, ce_id, ie1_reg_addr);
+
+ if (ce_attr->dest_nentries) {
+ ath12k_ahb_setbit32(ab, ce_id, ie2_reg_addr);
+ ath12k_ahb_setbit32(ab, ce_id + CE_HOST_IE_3_SHIFT,
+ ie3_reg_addr);
+ }
+}
+
+static void ath12k_ahb_ce_irq_disable(struct ath12k_base *ab, u16 ce_id)
+{
+ const struct ce_attr *ce_attr;
+ const struct ce_ie_addr *ce_ie_addr = ab->hw_params->ce_ie_addr;
+ u32 ie1_reg_addr, ie2_reg_addr, ie3_reg_addr;
+
+ ie1_reg_addr = ce_ie_addr->ie1_reg_addr;
+ ie2_reg_addr = ce_ie_addr->ie2_reg_addr;
+ ie3_reg_addr = ce_ie_addr->ie3_reg_addr;
+
+ ce_attr = &ab->hw_params->host_ce_config[ce_id];
+ if (ce_attr->src_nentries)
+ ath12k_ahb_clearbit32(ab, ce_id, ie1_reg_addr);
+
+ if (ce_attr->dest_nentries) {
+ ath12k_ahb_clearbit32(ab, ce_id, ie2_reg_addr);
+ ath12k_ahb_clearbit32(ab, ce_id + CE_HOST_IE_3_SHIFT,
+ ie3_reg_addr);
+ }
+}
+
+static void ath12k_ahb_sync_ce_irqs(struct ath12k_base *ab)
+{
+ int i;
+ int irq_idx;
+
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+
+ irq_idx = ATH12K_IRQ_CE0_OFFSET + i;
+ synchronize_irq(ab->irq_num[irq_idx]);
+ }
+}
+
+static void ath12k_ahb_sync_ext_irqs(struct ath12k_base *ab)
+{
+ int i, j;
+ int irq_idx;
+
+ for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
+ struct ath12k_ext_irq_grp *irq_grp = &ab->ext_irq_grp[i];
+
+ for (j = 0; j < irq_grp->num_irq; j++) {
+ irq_idx = irq_grp->irqs[j];
+ synchronize_irq(ab->irq_num[irq_idx]);
+ }
+ }
+}
+
+static void ath12k_ahb_ce_irqs_enable(struct ath12k_base *ab)
+{
+ int i;
+
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+ ath12k_ahb_ce_irq_enable(ab, i);
+ }
+}
+
+static void ath12k_ahb_ce_irqs_disable(struct ath12k_base *ab)
+{
+ int i;
+
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+ ath12k_ahb_ce_irq_disable(ab, i);
+ }
+}
+
+static int ath12k_ahb_start(struct ath12k_base *ab)
+{
+ ath12k_ahb_ce_irqs_enable(ab);
+ ath12k_ce_rx_post_buf(ab);
+
+ return 0;
+}
+
+static void ath12k_ahb_ext_irq_enable(struct ath12k_base *ab)
+{
+ struct ath12k_ext_irq_grp *irq_grp;
+ int i;
+
+ for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
+ irq_grp = &ab->ext_irq_grp[i];
+ if (!irq_grp->napi_enabled) {
+ napi_enable(&irq_grp->napi);
+ irq_grp->napi_enabled = true;
+ }
+ ath12k_ahb_ext_grp_enable(irq_grp);
+ }
+}
+
+static void ath12k_ahb_ext_irq_disable(struct ath12k_base *ab)
+{
+ __ath12k_ahb_ext_irq_disable(ab);
+ ath12k_ahb_sync_ext_irqs(ab);
+}
+
+static void ath12k_ahb_stop(struct ath12k_base *ab)
+{
+ if (!test_bit(ATH12K_FLAG_CRASH_FLUSH, &ab->dev_flags))
+ ath12k_ahb_ce_irqs_disable(ab);
+ ath12k_ahb_sync_ce_irqs(ab);
+ ath12k_ahb_cancel_workqueue(ab);
+ del_timer_sync(&ab->rx_replenish_retry);
+ ath12k_ce_cleanup_pipes(ab);
+}
+
+static void ath12k_ahb_init_qmi_ce_config(struct ath12k_base *ab)
+{
+ struct ath12k_qmi_ce_cfg *cfg = &ab->qmi.ce_cfg;
+
+ cfg->tgt_ce_len = ab->hw_params->target_ce_count;
+ cfg->tgt_ce = ab->hw_params->target_ce_config;
+ cfg->svc_to_ce_map_len = ab->hw_params->svc_to_ce_map_len;
+ cfg->svc_to_ce_map = ab->hw_params->svc_to_ce_map;
+ ab->qmi.service_ins_id = ab->hw_params->qmi_service_ins_id;
+}
+
+static void ath12k_ahb_free_ext_irq(struct ath12k_base *ab)
+{
+ int i, j;
+
+ for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
+ struct ath12k_ext_irq_grp *irq_grp = &ab->ext_irq_grp[i];
+
+ for (j = 0; j < irq_grp->num_irq; j++)
+ free_irq(ab->irq_num[irq_grp->irqs[j]], irq_grp);
+
+ netif_napi_del(&irq_grp->napi);
+ free_netdev(irq_grp->napi_ndev);
+ }
+}
+
+static void ath12k_ahb_free_irq(struct ath12k_base *ab)
+{
+ int irq_idx;
+ int i;
+
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+ irq_idx = ATH12K_IRQ_CE0_OFFSET + i;
+ free_irq(ab->irq_num[irq_idx], &ab->ce.ce_pipe[i]);
+ }
+
+ ath12k_ahb_free_ext_irq(ab);
+}
+
+static void ath12k_ahb_ce_workqueue(struct work_struct *work)
+{
+ struct ath12k_ce_pipe *ce_pipe = from_work(ce_pipe, work, intr_wq);
+
+ ath12k_ce_per_engine_service(ce_pipe->ab, ce_pipe->pipe_num);
+
+ ath12k_ahb_ce_irq_enable(ce_pipe->ab, ce_pipe->pipe_num);
+}
+
+static irqreturn_t ath12k_ahb_ce_interrupt_handler(int irq, void *arg)
+{
+ struct ath12k_ce_pipe *ce_pipe = arg;
+
+ /* last interrupt received for this CE */
+ ce_pipe->timestamp = jiffies;
+
+ ath12k_ahb_ce_irq_disable(ce_pipe->ab, ce_pipe->pipe_num);
+
+ queue_work(system_bh_wq, &ce_pipe->intr_wq);
+
+ return IRQ_HANDLED;
+}
+
+static int ath12k_ahb_ext_grp_napi_poll(struct napi_struct *napi, int budget)
+{
+ struct ath12k_ext_irq_grp *irq_grp = container_of(napi,
+ struct ath12k_ext_irq_grp,
+ napi);
+ struct ath12k_base *ab = irq_grp->ab;
+ int work_done;
+
+ work_done = ath12k_dp_service_srng(ab, irq_grp, budget);
+ if (work_done < budget) {
+ napi_complete_done(napi, work_done);
+ ath12k_ahb_ext_grp_enable(irq_grp);
+ }
+
+ if (work_done > budget)
+ work_done = budget;
+
+ return work_done;
+}
+
+static irqreturn_t ath12k_ahb_ext_interrupt_handler(int irq, void *arg)
+{
+ struct ath12k_ext_irq_grp *irq_grp = arg;
+
+ /* last interrupt received for this group */
+ irq_grp->timestamp = jiffies;
+
+ ath12k_ahb_ext_grp_disable(irq_grp);
+
+ napi_schedule(&irq_grp->napi);
+
+ return IRQ_HANDLED;
+}
+
+static int ath12k_ahb_config_ext_irq(struct ath12k_base *ab)
+{
+ struct ath12k_ext_irq_grp *irq_grp;
+ const struct hal_ops *hal_ops;
+ int i, j, irq, irq_idx, ret;
+ u32 num_irq;
+
+ hal_ops = ab->hw_params->hal_ops;
+ for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
+ irq_grp = &ab->ext_irq_grp[i];
+ num_irq = 0;
+
+ irq_grp->ab = ab;
+ irq_grp->grp_id = i;
+
+ irq_grp->napi_ndev = alloc_netdev_dummy(0);
+ if (!irq_grp->napi_ndev)
+ return -ENOMEM;
+
+ netif_napi_add(irq_grp->napi_ndev, &irq_grp->napi,
+ ath12k_ahb_ext_grp_napi_poll);
+
+ for (j = 0; j < ATH12K_EXT_IRQ_NUM_MAX; j++) {
+ if (ab->hw_params->ring_mask->tx[i] &&
+ j <= ATH12K_MAX_TCL_RING_NUM &&
+ (ab->hw_params->ring_mask->tx[i] &
+ BIT(hal_ops->tcl_to_wbm_rbm_map[j].wbm_ring_num))) {
+ irq_grp->irqs[num_irq++] =
+ wbm2host_tx_completions_ring1 - j;
+ }
+
+ if (ab->hw_params->ring_mask->rx[i] & BIT(j)) {
+ irq_grp->irqs[num_irq++] =
+ reo2host_destination_ring1 - j;
+ }
+
+ if (ab->hw_params->ring_mask->rx_err[i] & BIT(j))
+ irq_grp->irqs[num_irq++] = reo2host_exception;
+
+ if (ab->hw_params->ring_mask->rx_wbm_rel[i] & BIT(j))
+ irq_grp->irqs[num_irq++] = wbm2host_rx_release;
+
+ if (ab->hw_params->ring_mask->reo_status[i] & BIT(j))
+ irq_grp->irqs[num_irq++] = reo2host_status;
+
+ if (ab->hw_params->ring_mask->rx_mon_dest[i] & BIT(j))
+ irq_grp->irqs[num_irq++] =
+ rxdma2host_monitor_destination_mac1;
+ }
+
+ irq_grp->num_irq = num_irq;
+
+ for (j = 0; j < irq_grp->num_irq; j++) {
+ irq_idx = irq_grp->irqs[j];
+
+ irq = platform_get_irq_byname(ab->pdev,
+ irq_name[irq_idx]);
+ ab->irq_num[irq_idx] = irq;
+ irq_set_status_flags(irq, IRQ_NOAUTOEN | IRQ_DISABLE_UNLAZY);
+ ret = request_irq(irq, ath12k_ahb_ext_interrupt_handler,
+ IRQF_TRIGGER_RISING,
+ irq_name[irq_idx], irq_grp);
+ if (ret) {
+ ath12k_err(ab, "failed request_irq for %d\n",
+ irq);
+ }
+ }
+ }
+
+ return 0;
+}
+
+static int ath12k_ahb_config_irq(struct ath12k_base *ab)
+{
+ int irq, irq_idx, i;
+ int ret;
+
+ /* Configure CE irqs */
+ for (i = 0; i < ab->hw_params->ce_count; i++) {
+ struct ath12k_ce_pipe *ce_pipe = &ab->ce.ce_pipe[i];
+
+ if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
+ continue;
+
+ irq_idx = ATH12K_IRQ_CE0_OFFSET + i;
+
+ INIT_WORK(&ce_pipe->intr_wq, ath12k_ahb_ce_workqueue);
+ irq = platform_get_irq_byname(ab->pdev, irq_name[irq_idx]);
+ ret = request_irq(irq, ath12k_ahb_ce_interrupt_handler,
+ IRQF_TRIGGER_RISING, irq_name[irq_idx],
+ ce_pipe);
+ if (ret)
+ return ret;
+
+ ab->irq_num[irq_idx] = irq;
+ }
+
+ /* Configure external interrupts */
+ ret = ath12k_ahb_config_ext_irq(ab);
+
+ return ret;
+}
+
+static int ath12k_ahb_map_service_to_pipe(struct ath12k_base *ab, u16 service_id,
+ u8 *ul_pipe, u8 *dl_pipe)
+{
+ const struct service_to_pipe *entry;
+ bool ul_set = false, dl_set = false;
+ int i;
+
+ for (i = 0; i < ab->hw_params->svc_to_ce_map_len; i++) {
+ entry = &ab->hw_params->svc_to_ce_map[i];
+
+ if (__le32_to_cpu(entry->service_id) != service_id)
+ continue;
+
+ switch (__le32_to_cpu(entry->pipedir)) {
+ case PIPEDIR_NONE:
+ break;
+ case PIPEDIR_IN:
+ WARN_ON(dl_set);
+ *dl_pipe = __le32_to_cpu(entry->pipenum);
+ dl_set = true;
+ break;
+ case PIPEDIR_OUT:
+ WARN_ON(ul_set);
+ *ul_pipe = __le32_to_cpu(entry->pipenum);
+ ul_set = true;
+ break;
+ case PIPEDIR_INOUT:
+ WARN_ON(dl_set);
+ WARN_ON(ul_set);
+ *dl_pipe = __le32_to_cpu(entry->pipenum);
+ *ul_pipe = __le32_to_cpu(entry->pipenum);
+ dl_set = true;
+ ul_set = true;
+ break;
+ }
+ }
+
+ if (WARN_ON(!ul_set || !dl_set))
+ return -ENOENT;
+
+ return 0;
+}
+
+static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
+ .start = ath12k_ahb_start,
+ .stop = ath12k_ahb_stop,
+ .read32 = ath12k_ahb_read32,
+ .write32 = ath12k_ahb_write32,
+ .cmem_read32 = ath12k_ahb_cmem_read32,
+ .cmem_write32 = ath12k_ahb_cmem_write32,
+ .irq_enable = ath12k_ahb_ext_irq_enable,
+ .irq_disable = ath12k_ahb_ext_irq_disable,
+ .map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
+};
+
+static int ath12k_ahb_clock_init(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ int ret;
+
+ ab_ahb->xo_clk = devm_clk_get(ab->dev, "gcc_xo_clk");
+ if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
+ ath12k_err(ab, "failed to get gcc_xo_clk: %d\n",
+ PTR_ERR_OR_ZERO(ab_ahb->xo_clk));
+ ret = ab_ahb->xo_clk ? PTR_ERR(ab_ahb->xo_clk) : -ENODEV;
+ ab_ahb->xo_clk = NULL;
+ return ret;
+ }
+
+ return 0;
+}
+
+static void ath12k_ahb_clock_deinit(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+
+ devm_clk_put(ab->dev, ab_ahb->xo_clk);
+ ab_ahb->xo_clk = NULL;
+}
+
+static int ath12k_ahb_clock_enable(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ int ret;
+
+ if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
+ ath12k_err(ab, "clock is not initialized\n");
+ return -EIO;
+ }
+
+ ret = clk_prepare_enable(ab_ahb->xo_clk);
+ if (ret) {
+ ath12k_err(ab, "failed to enable gcc_xo_clk: %d\n", ret);
+ return ret;
+ }
+
+ return 0;
+}
+
+static void ath12k_ahb_clock_disable(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+
+ clk_disable_unprepare(ab_ahb->xo_clk);
+}
+
+static int ath12k_ahb_resource_init(struct ath12k_base *ab)
+{
+ struct platform_device *pdev = ab->pdev;
+ struct resource *mem_res;
+ void __iomem *mem;
+ int ret;
+
+ mem = devm_platform_get_and_ioremap_resource(pdev, 0, &mem_res);
+ if (IS_ERR(mem)) {
+ dev_err(&pdev->dev, "ioremap error\n");
+ ret = PTR_ERR(mem);
+ goto out;
+ }
+
+ ab->mem = mem;
+ ab->mem_len = resource_size(mem_res);
+
+ if (ab->hw_params->ce_remap) {
+ const struct ce_remap *ce_remap = ab->hw_params->ce_remap;
+ /* ce register space is moved out of wcss and the space is not
+ * contiguous, hence remapping the CE registers to a new space
+ * for accessing them.
+ */
+ ab->mem_ce = ioremap(ce_remap->base, ce_remap->size);
+ if (IS_ERR(ab->mem_ce)) {
+ dev_err(&pdev->dev, "ce ioremap error\n");
+ ret = -ENOMEM;
+ goto err_mem_unmap;
+ }
+ ab->ce_remap = true;
+ ab->ce_remap_base_addr = HAL_IPQ5332_CE_WFSS_REG_BASE;
+ }
+
+ if (ab->hw_params->cmem_remap) {
+ const struct cmem_remap *cmem_remap = ab->hw_params->cmem_remap;
+ /* For device like IPQ5332 CMEM region is outside WCSS block.
+ * Allocate separate I/O remap to access CMEM address.
+ */
+ ab->mem_cmem = ioremap(cmem_remap->base, cmem_remap->size);
+ if (IS_ERR(ab->mem_cmem)) {
+ dev_err(&pdev->dev, "cmem ioremap error\n");
+ ret = -ENOMEM;
+ goto err_mem_ce_unmap;
+ }
+ }
+
+ ret = ath12k_ahb_clock_init(ab);
+ if (ret)
+ goto err_mem_cmem_unmap;
+
+ ret = ath12k_ahb_clock_enable(ab);
+ if (ret)
+ goto err_clock_deinit;
+
+ return 0;
+
+err_clock_deinit:
+ ath12k_ahb_clock_deinit(ab);
+
+err_mem_cmem_unmap:
+ if (ab->hw_params->cmem_remap)
+ iounmap(ab->mem_cmem);
+
+err_mem_ce_unmap:
+ ab->mem_cmem = NULL;
+ if (ab->hw_params->ce_remap)
+ iounmap(ab->mem_ce);
+
+err_mem_unmap:
+ ab->mem_ce = NULL;
+ devm_iounmap(ab->dev, ab->mem);
+
+out:
+ ab->mem = NULL;
+ return ret;
+}
+
+static void ath12k_ahb_resource_deinit(struct ath12k_base *ab)
+{
+ if (ab->mem)
+ devm_iounmap(ab->dev, ab->mem);
+
+ if (ab->mem_ce)
+ iounmap(ab->mem_ce);
+
+ if (ab->mem_cmem)
+ iounmap(ab->mem_cmem);
+
+ ab->mem = NULL;
+ ab->mem_ce = NULL;
+ ab->mem_cmem = NULL;
+
+ ath12k_ahb_clock_disable(ab);
+ ath12k_ahb_clock_deinit(ab);
+}
+
+static enum ath12k_hw_rev ath12k_ahb_get_hw_rev(struct platform_device *pdev)
+{
+ const struct of_device_id *of_id;
+
+ of_id = of_match_device(ath12k_ahb_of_match, &pdev->dev);
+ if (!of_id) {
+ dev_err(&pdev->dev, "Failed to find matching device tree id\n");
+ return -EINVAL;
+ }
+
+ return (enum ath12k_hw_rev)of_id->data;
+}
+
+static int ath12k_ahb_probe(struct platform_device *pdev)
+{
+ struct ath12k_base *ab;
+ const struct ath12k_hif_ops *hif_ops;
+ struct device_node *mem_node;
+ enum ath12k_hw_rev hw_rev;
+ u32 addr;
+ int ret;
+
+ hw_rev = ath12k_ahb_get_hw_rev(pdev);
+ switch (hw_rev) {
+ case ATH12K_HW_IPQ5332_HW10:
+ hif_ops = &ath12k_ahb_hif_ops_ipq5332;
+ break;
+ default:
+ dev_err(&pdev->dev, "Unsupported device type %d\n", hw_rev);
+ return -EOPNOTSUPP;
+ }
+
+ ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
+ if (ret) {
+ dev_err(&pdev->dev, "Failed to set 32-bit consistent dma\n");
+ return ret;
+ }
+
+ ab = ath12k_core_alloc(&pdev->dev, sizeof(struct ath12k_ahb),
+ ATH12K_BUS_AHB);
+ if (!ab) {
+ dev_err(&pdev->dev, "failed to allocate ath12k base\n");
+ return -ENOMEM;
+ }
+
+ ab->hif.ops = hif_ops;
+ ab->pdev = pdev;
+ ab->hw_rev = hw_rev;
+ platform_set_drvdata(pdev, ab);
+
+ /* Set fixed_mem_region to true for platforms that support fixed memory
+ * reservation from DT. If memory is reserved from DT for FW, ath12k driver
+ * need not to allocate memory.
+ */
+ if (!of_property_read_u32(ab->dev->of_node, "memory-region", &addr)) {
+ set_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags);
+ mem_node = of_find_node_by_name(NULL, "mlo_global_mem_0");
+ if (!mem_node)
+ ab->mlo_capable_flags = 0;
+ }
+
+ ret = ath12k_core_pre_init(ab);
+ if (ret)
+ goto err_core_free;
+
+ ret = ath12k_ahb_resource_init(ab);
+ if (ret)
+ goto err_core_free;
+
+ ret = ath12k_hal_srng_init(ab);
+ if (ret)
+ goto err_resource_deinit;
+
+ ret = ath12k_ce_alloc_pipes(ab);
+ if (ret) {
+ ath12k_err(ab, "failed to allocate ce pipes: %d\n", ret);
+ goto err_hal_srng_deinit;
+ }
+
+ ath12k_ahb_init_qmi_ce_config(ab);
+
+ ret = ath12k_ahb_config_irq(ab);
+ if (ret) {
+ ath12k_err(ab, "failed to configure irq: %d\n", ret);
+ goto err_ce_free;
+ }
+
+ ret = ath12k_core_init(ab);
+ if (ret) {
+ ath12k_err(ab, "failed to init core: %d\n", ret);
+ goto err_ce_free;
+ }
+
+ return 0;
+
+err_ce_free:
+ ath12k_ce_free_pipes(ab);
+
+err_hal_srng_deinit:
+ ath12k_hal_srng_deinit(ab);
+
+err_resource_deinit:
+ ath12k_ahb_resource_deinit(ab);
+
+err_core_free:
+ ath12k_core_free(ab);
+ platform_set_drvdata(pdev, NULL);
+
+ return ret;
+}
+
+static void ath12k_ahb_remove_prepare(struct ath12k_base *ab)
+{
+ unsigned long left;
+
+ if (test_bit(ATH12K_FLAG_RECOVERY, &ab->dev_flags)) {
+ left = wait_for_completion_timeout(&ab->driver_recovery,
+ ATH12K_AHB_RECOVERY_TIMEOUT);
+ if (!left)
+ ath12k_warn(ab, "failed to receive recovery response completion\n");
+ }
+
+ set_bit(ATH12K_FLAG_UNREGISTERING, &ab->dev_flags);
+ cancel_work_sync(&ab->restart_work);
+ cancel_work_sync(&ab->qmi.event_work);
+}
+
+static void ath12k_ahb_free_resources(struct ath12k_base *ab)
+{
+ struct platform_device *pdev = ab->pdev;
+
+ ath12k_ahb_free_irq(ab);
+ ath12k_hal_srng_deinit(ab);
+ ath12k_ce_free_pipes(ab);
+ ath12k_ahb_resource_deinit(ab);
+ ath12k_core_free(ab);
+ platform_set_drvdata(pdev, NULL);
+}
+
+static void ath12k_ahb_remove(struct platform_device *pdev)
+{
+ struct ath12k_base *ab = platform_get_drvdata(pdev);
+
+ if (test_bit(ATH12K_FLAG_QMI_FAIL, &ab->dev_flags)) {
+ ath12k_qmi_deinit_service(ab);
+ goto qmi_fail;
+ }
+
+ ath12k_ahb_remove_prepare(ab);
+ ath12k_core_deinit(ab);
+
+qmi_fail:
+ ath12k_ahb_free_resources(ab);
+}
+
+static void ath12k_ahb_shutdown(struct platform_device *pdev)
+{
+ struct ath12k_base *ab = platform_get_drvdata(pdev);
+
+ /* platform shutdown() & remove() are mutually exclusive.
+ * remove() is invoked during rmmod & shutdown() during
+ * system reboot/shutdown.
+ */
+ ath12k_ahb_remove_prepare(ab);
+
+ if (!(test_bit(ATH12K_FLAG_REGISTERED, &ab->dev_flags)))
+ goto free_resources;
+
+ ath12k_core_deinit(ab);
+
+free_resources:
+ ath12k_ahb_free_resources(ab);
+}
+
+static struct platform_driver ath12k_ahb_driver = {
+ .driver = {
+ .name = "ath12k_ahb",
+ .of_match_table = ath12k_ahb_of_match,
+ },
+ .probe = ath12k_ahb_probe,
+ .remove = ath12k_ahb_remove,
+ .shutdown = ath12k_ahb_shutdown,
+};
+
+int ath12k_ahb_init(void)
+{
+ return platform_driver_register(&ath12k_ahb_driver);
+}
+
+void ath12k_ahb_exit(void)
+{
+ platform_driver_unregister(&ath12k_ahb_driver);
+}
diff --git a/drivers/net/wireless/ath/ath12k/ahb.h b/drivers/net/wireless/ath/ath12k/ahb.h
new file mode 100644
index 000000000000..bd0366a79587
--- /dev/null
+++ b/drivers/net/wireless/ath/ath12k/ahb.h
@@ -0,0 +1,37 @@
+/* SPDX-License-Identifier: BSD-3-Clause-Clear */
+/*
+ * Copyright (c) 2018-2019 The Linux Foundation. All rights reserved.
+ * Copyright (c) 2022-2024, Qualcomm Innovation Center, Inc. All rights reserved.
+ */
+#ifndef ATH12K_AHB_H
+#define ATH12K_AHB_H
+
+#include <linux/clk.h>
+#include "core.h"
+
+#define ATH12K_AHB_RECOVERY_TIMEOUT (3 * HZ)
+
+#define ATH12K_AHB_SMP2P_SMEM_MSG GENMASK(15, 0)
+#define ATH12K_AHB_SMP2P_SMEM_SEQ_NO GENMASK(31, 16)
+#define ATH12K_AHB_SMP2P_SMEM_VALUE_MASK 0xFFFFFFFF
+#define ATH12K_PCI_CE_WAKE_IRQ 2
+#define ATH12K_PCI_IRQ_CE0_OFFSET 3
+
+enum ath12k_ahb_smp2p_msg_id {
+ ATH12K_AHB_POWER_SAVE_ENTER = 1,
+ ATH12K_AHB_POWER_SAVE_EXIT,
+};
+
+struct ath12k_base;
+
+struct ath12k_ahb {
+ struct rproc *tgt_rproc;
+ struct clk *xo_clk;
+};
+
+static inline struct ath12k_ahb *ath12k_ab_to_ahb(struct ath12k_base *ab)
+{
+ return (struct ath12k_ahb *)ab->drv_priv;
+}
+
+#endif
diff --git a/drivers/net/wireless/ath/ath12k/core.h b/drivers/net/wireless/ath/ath12k/core.h
index 75bc863ca381..a815a23a7274 100644
--- a/drivers/net/wireless/ath/ath12k/core.h
+++ b/drivers/net/wireless/ath/ath12k/core.h
@@ -142,6 +142,7 @@ enum ath12k_firmware_mode {
#define ATH12K_IRQ_NUM_MAX 57
#define ATH12K_EXT_IRQ_NUM_MAX 16
+#define ATH12K_MAX_TCL_RING_NUM 3
struct ath12k_ext_irq_grp {
struct ath12k_base *ab;
@@ -149,6 +150,7 @@ struct ath12k_ext_irq_grp {
u32 num_irq;
u32 grp_id;
u64 timestamp;
+ bool napi_enabled;
struct napi_struct napi;
struct net_device *napi_ndev;
};
@@ -1063,6 +1065,8 @@ static inline const char *ath12k_bus_str(enum ath12k_bus bus)
switch (bus) {
case ATH12K_BUS_PCI:
return "pci";
+ case ATH12K_BUS_AHB:
+ return "ahb";
}
return "unknown";
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index 038fe1b30d11..b3e2a5d7e0e4 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -121,6 +121,7 @@ enum ath12k_hw_rate_ofdm {
enum ath12k_bus {
ATH12K_BUS_PCI,
+ ATH12K_BUS_AHB,
};
#define ATH12K_EXT_IRQ_GRP_NUM_MAX 11
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332
2024-10-15 18:26 ` [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332 Raj Kumar Bhagat
@ 2024-10-18 20:29 ` Konrad Dybcio
2024-12-06 9:56 ` Raj Kumar Bhagat
2024-12-06 10:00 ` Raj Kumar Bhagat
0 siblings, 2 replies; 60+ messages in thread
From: Konrad Dybcio @ 2024-10-18 20:29 UTC (permalink / raw)
To: Raj Kumar Bhagat, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
> From: Balamurugan S <quic_bselvara@quicinc.com>
>
> Add Initial Ath12k AHB driver support for IPQ5332. IPQ5332 is AHB
> based IEEE802.11be 2 GHz 2x2 WiFi device.
>
> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>
> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
> Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
> Co-developed-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
> ---
[...]
> +enum ext_irq_num {
> + host2wbm_desc_feed = 16,
> + host2reo_re_injection,
Why?
> +static u32 ath12k_ahb_cmem_read32(struct ath12k_base *ab, u32 offset)
> +{
> + offset = offset - HAL_IPQ5332_CMEM_BASE;
> + return ioread32(ab->mem_cmem + offset);
return ioread32(ab->mem_cmem + offset - HAL_IPQ5332_CMEM_BASE)?
Or maybe the mem_cmem base should be moved?
> +static int ath12k_ahb_start(struct ath12k_base *ab)
> +{
> + ath12k_ahb_ce_irqs_enable(ab);
> + ath12k_ce_rx_post_buf(ab);
> +
> + return 0;
> +}
Neither this nor ath12k_pci returns anything useful - perhaps make this void?
> +static void ath12k_ahb_free_ext_irq(struct ath12k_base *ab)
Any reason we're not using devm APIs?
> +static int ath12k_ahb_config_ext_irq(struct ath12k_base *ab)
> +{
> + struct ath12k_ext_irq_grp *irq_grp;
> + const struct hal_ops *hal_ops;
> + int i, j, irq, irq_idx, ret;
> + u32 num_irq;
> +
> + hal_ops = ab->hw_params->hal_ops;
> + for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
> + irq_grp = &ab->ext_irq_grp[i];
> + num_irq = 0;
> +
> + irq_grp->ab = ab;
> + irq_grp->grp_id = i;
> +
> + irq_grp->napi_ndev = alloc_netdev_dummy(0);
> + if (!irq_grp->napi_ndev)
> + return -ENOMEM;
> +
> + netif_napi_add(irq_grp->napi_ndev, &irq_grp->napi,
> + ath12k_ahb_ext_grp_napi_poll);
> +
> + for (j = 0; j < ATH12K_EXT_IRQ_NUM_MAX; j++) {
> + if (ab->hw_params->ring_mask->tx[i] &&
> + j <= ATH12K_MAX_TCL_RING_NUM &&
> + (ab->hw_params->ring_mask->tx[i] &
> + BIT(hal_ops->tcl_to_wbm_rbm_map[j].wbm_ring_num))) {
> + irq_grp->irqs[num_irq++] =
> + wbm2host_tx_completions_ring1 - j;
> + }
This is unreadable
> +
> + if (ab->hw_params->ring_mask->rx[i] & BIT(j)) {
Consider taking a pointer to ring_mask so that the lines are shorter
> + irq_grp->irqs[num_irq++] =
> + reo2host_destination_ring1 - j;
> + }
> +
> + if (ab->hw_params->ring_mask->rx_err[i] & BIT(j))
> + irq_grp->irqs[num_irq++] = reo2host_exception;
> +
> + if (ab->hw_params->ring_mask->rx_wbm_rel[i] & BIT(j))
> + irq_grp->irqs[num_irq++] = wbm2host_rx_release;
> +
> + if (ab->hw_params->ring_mask->reo_status[i] & BIT(j))
> + irq_grp->irqs[num_irq++] = reo2host_status;
> +
> + if (ab->hw_params->ring_mask->rx_mon_dest[i] & BIT(j))
> + irq_grp->irqs[num_irq++] =
> + rxdma2host_monitor_destination_mac1;
> + }
> +
> + irq_grp->num_irq = num_irq;
> +
> + for (j = 0; j < irq_grp->num_irq; j++) {
> + irq_idx = irq_grp->irqs[j];
> +
> + irq = platform_get_irq_byname(ab->pdev,
> + irq_name[irq_idx]);
> + ab->irq_num[irq_idx] = irq;
> + irq_set_status_flags(irq, IRQ_NOAUTOEN | IRQ_DISABLE_UNLAZY);
> + ret = request_irq(irq, ath12k_ahb_ext_interrupt_handler,
> + IRQF_TRIGGER_RISING,
> + irq_name[irq_idx], irq_grp);
> + if (ret) {
> + ath12k_err(ab, "failed request_irq for %d\n",
> + irq);
> + }
> + }
Instead of doing all this magic, can we request the IRQs manually, as we
have interrupt-names in dt?
> + }
> +
> + return 0;
> +}
> +
> +static int ath12k_ahb_config_irq(struct ath12k_base *ab)
> +{
> + int irq, irq_idx, i;
> + int ret;
> +
> + /* Configure CE irqs */
> + for (i = 0; i < ab->hw_params->ce_count; i++) {
> + struct ath12k_ce_pipe *ce_pipe = &ab->ce.ce_pipe[i];
> +
> + if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
> + continue;
> +
> + irq_idx = ATH12K_IRQ_CE0_OFFSET + i;
> +
> + INIT_WORK(&ce_pipe->intr_wq, ath12k_ahb_ce_workqueue);
> + irq = platform_get_irq_byname(ab->pdev, irq_name[irq_idx]);
> + ret = request_irq(irq, ath12k_ahb_ce_interrupt_handler,
> + IRQF_TRIGGER_RISING, irq_name[irq_idx],
> + ce_pipe);
> + if (ret)
> + return ret;
> +
> + ab->irq_num[irq_idx] = irq;
> + }
> +
> + /* Configure external interrupts */
> + ret = ath12k_ahb_config_ext_irq(ab);
> +
> + return ret;
> +}
> +
> +static int ath12k_ahb_map_service_to_pipe(struct ath12k_base *ab, u16 service_id,
> + u8 *ul_pipe, u8 *dl_pipe)
> +{
> + const struct service_to_pipe *entry;
> + bool ul_set = false, dl_set = false;
> + int i;
> +
> + for (i = 0; i < ab->hw_params->svc_to_ce_map_len; i++) {
> + entry = &ab->hw_params->svc_to_ce_map[i];
> +
> + if (__le32_to_cpu(entry->service_id) != service_id)
> + continue;
> +
> + switch (__le32_to_cpu(entry->pipedir)) {
> + case PIPEDIR_NONE:
> + break;
> + case PIPEDIR_IN:
> + WARN_ON(dl_set);
> + *dl_pipe = __le32_to_cpu(entry->pipenum);
> + dl_set = true;
> + break;
> + case PIPEDIR_OUT:
> + WARN_ON(ul_set);
> + *ul_pipe = __le32_to_cpu(entry->pipenum);
> + ul_set = true;
> + break;
> + case PIPEDIR_INOUT:
> + WARN_ON(dl_set);
> + WARN_ON(ul_set);
> + *dl_pipe = __le32_to_cpu(entry->pipenum);
> + *ul_pipe = __le32_to_cpu(entry->pipenum);
> + dl_set = true;
> + ul_set = true;
> + break;
> + }
if pipedir == PIPEDIR_IN || pipedir == PIPEDIR_INOUT
if pipedir == PIPEDIR_OUT || pipedir == PIPE_INOUT
?
> + }
> +
> + if (WARN_ON(!ul_set || !dl_set))
> + return -ENOENT;
> +
> + return 0;
> +}
> +
> +static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
> + .start = ath12k_ahb_start,
> + .stop = ath12k_ahb_stop,
> + .read32 = ath12k_ahb_read32,
> + .write32 = ath12k_ahb_write32,
> + .cmem_read32 = ath12k_ahb_cmem_read32,
> + .cmem_write32 = ath12k_ahb_cmem_write32,
> + .irq_enable = ath12k_ahb_ext_irq_enable,
> + .irq_disable = ath12k_ahb_ext_irq_disable,
> + .map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
> +};
> +
> +static int ath12k_ahb_clock_init(struct ath12k_base *ab)
> +{
> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
> + int ret;
> +
> + ab_ahb->xo_clk = devm_clk_get(ab->dev, "gcc_xo_clk");
> + if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
> + ath12k_err(ab, "failed to get gcc_xo_clk: %d\n",
> + PTR_ERR_OR_ZERO(ab_ahb->xo_clk));
> + ret = ab_ahb->xo_clk ? PTR_ERR(ab_ahb->xo_clk) : -ENODEV;
> + ab_ahb->xo_clk = NULL;
> + return ret;
> + }
Won't clk core print errors in both of these cases?
> +
> + return 0;
> +}
> +
> +static void ath12k_ahb_clock_deinit(struct ath12k_base *ab)
> +{
> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
> +
> + devm_clk_put(ab->dev, ab_ahb->xo_clk);
> + ab_ahb->xo_clk = NULL;
> +}
> +
> +static int ath12k_ahb_clock_enable(struct ath12k_base *ab)
> +{
> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
> + int ret;
> +
> + if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
> + ath12k_err(ab, "clock is not initialized\n");
> + return -EIO;
> + }
> +
> + ret = clk_prepare_enable(ab_ahb->xo_clk);
> + if (ret) {
> + ath12k_err(ab, "failed to enable gcc_xo_clk: %d\n", ret);
> + return ret;
> + }
> +
> + return 0;
> +}
> +
> +static void ath12k_ahb_clock_disable(struct ath12k_base *ab)
> +{
> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
> +
> + clk_disable_unprepare(ab_ahb->xo_clk);
> +}
> +
> +static int ath12k_ahb_resource_init(struct ath12k_base *ab)
> +{
> + struct platform_device *pdev = ab->pdev;
> + struct resource *mem_res;
> + void __iomem *mem;
> + int ret;
> +
> + mem = devm_platform_get_and_ioremap_resource(pdev, 0, &mem_res);
> + if (IS_ERR(mem)) {
> + dev_err(&pdev->dev, "ioremap error\n");
> + ret = PTR_ERR(mem);
> + goto out;
If you assign ab->mem directly, you can get rid of the line below
and return the error here
> + }
> +
> + ab->mem = mem;
> + ab->mem_len = resource_size(mem_res);
> +
> + if (ab->hw_params->ce_remap) {
> + const struct ce_remap *ce_remap = ab->hw_params->ce_remap;
> + /* ce register space is moved out of wcss and the space is not
> + * contiguous, hence remapping the CE registers to a new space
> + * for accessing them.
> + */
Please capitalize words consistently
> + ab->mem_ce = ioremap(ce_remap->base, ce_remap->size);
> + if (IS_ERR(ab->mem_ce)) {
> + dev_err(&pdev->dev, "ce ioremap error\n");
> + ret = -ENOMEM;
> + goto err_mem_unmap;
> + }
> + ab->ce_remap = true;
> + ab->ce_remap_base_addr = HAL_IPQ5332_CE_WFSS_REG_BASE;
> + }
> +
> + if (ab->hw_params->cmem_remap) {
> + const struct cmem_remap *cmem_remap = ab->hw_params->cmem_remap;
> + /* For device like IPQ5332 CMEM region is outside WCSS block.
IPQ5332 is not a 'device'
> + * Allocate separate I/O remap to access CMEM address.
> + */
> + ab->mem_cmem = ioremap(cmem_remap->base, cmem_remap->size);
> + if (IS_ERR(ab->mem_cmem)) {
> + dev_err(&pdev->dev, "cmem ioremap error\n");
> + ret = -ENOMEM;
> + goto err_mem_ce_unmap;
> + }
> + }
[...]
> + ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
> + if (ret) {
> + dev_err(&pdev->dev, "Failed to set 32-bit consistent dma\n");
s/consistent/coherent
You're setting the mask, not the DMA itself
[...]
> + /* Set fixed_mem_region to true for platforms that support fixed memory
> + * reservation from DT. If memory is reserved from DT for FW, ath12k driver
> + * need not to allocate memory.
> + */
> + if (!of_property_read_u32(ab->dev->of_node, "memory-region", &addr)) {
> + set_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags);
> + mem_node = of_find_node_by_name(NULL, "mlo_global_mem_0");
This is not mentioned or documented anywhere.
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread* Re: [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332
2024-10-18 20:29 ` Konrad Dybcio
@ 2024-12-06 9:56 ` Raj Kumar Bhagat
2024-12-06 11:46 ` Konrad Dybcio
2024-12-06 10:00 ` Raj Kumar Bhagat
1 sibling, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-06 9:56 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 10/19/2024 1:59 AM, Konrad Dybcio wrote:
> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>> From: Balamurugan S <quic_bselvara@quicinc.com>
>>
>> Add Initial Ath12k AHB driver support for IPQ5332. IPQ5332 is AHB
>> based IEEE802.11be 2 GHz 2x2 WiFi device.
>>
>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>
>> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
>> Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
>> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
>> Co-developed-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>> ---
>
> [...]
>
>> +enum ext_irq_num {
>> + host2wbm_desc_feed = 16,
>> + host2reo_re_injection,
>
> Why?
>
This enum is used as a IRQ number for Ath12k AHB. Based on this enum
we can get the IRQ name from irq_name[]. This helps to request the original
IRQ number from the DT.
It is starting from 16 becasue, in irq_name[], the name for ext IRQ starts
from 16 index.
>
>> +static u32 ath12k_ahb_cmem_read32(struct ath12k_base *ab, u32 offset)
>> +{
>> + offset = offset - HAL_IPQ5332_CMEM_BASE;
>> + return ioread32(ab->mem_cmem + offset);
>
> return ioread32(ab->mem_cmem + offset - HAL_IPQ5332_CMEM_BASE)?
>
> Or maybe the mem_cmem base should be moved?
>
sure, will update in next version.
>> +static int ath12k_ahb_start(struct ath12k_base *ab)
>> +{
>> + ath12k_ahb_ce_irqs_enable(ab);
>> + ath12k_ce_rx_post_buf(ab);
>> +
>> + return 0;
>> +}
>
> Neither this nor ath12k_pci returns anything useful - perhaps make this void?
>
>> +static void ath12k_ahb_free_ext_irq(struct ath12k_base *ab)
>
> Any reason we're not using devm APIs?
>
Thanks, we will move devm APIs in next version.
>> +static int ath12k_ahb_config_ext_irq(struct ath12k_base *ab)
>> +{
>> + struct ath12k_ext_irq_grp *irq_grp;
>> + const struct hal_ops *hal_ops;
>> + int i, j, irq, irq_idx, ret;
>> + u32 num_irq;
>> +
>> + hal_ops = ab->hw_params->hal_ops;
>> + for (i = 0; i < ATH12K_EXT_IRQ_GRP_NUM_MAX; i++) {
>> + irq_grp = &ab->ext_irq_grp[i];
>> + num_irq = 0;
>> +
>> + irq_grp->ab = ab;
>> + irq_grp->grp_id = i;
>> +
>> + irq_grp->napi_ndev = alloc_netdev_dummy(0);
>> + if (!irq_grp->napi_ndev)
>> + return -ENOMEM;
>> +
>> + netif_napi_add(irq_grp->napi_ndev, &irq_grp->napi,
>> + ath12k_ahb_ext_grp_napi_poll);
>> +
>> + for (j = 0; j < ATH12K_EXT_IRQ_NUM_MAX; j++) {
>> + if (ab->hw_params->ring_mask->tx[i] &&
>> + j <= ATH12K_MAX_TCL_RING_NUM &&
>> + (ab->hw_params->ring_mask->tx[i] &
>> + BIT(hal_ops->tcl_to_wbm_rbm_map[j].wbm_ring_num))) {
>> + irq_grp->irqs[num_irq++] =
>> + wbm2host_tx_completions_ring1 - j;
>> + }
>
> This is unreadable
>
In next version we will use a ring_mask pointer and also add comments
to make it more readable.
>> +
>> + if (ab->hw_params->ring_mask->rx[i] & BIT(j)) {
>
> Consider taking a pointer to ring_mask so that the lines are shorter
>
Thansk for the suggestion, will update in next version.
>> + irq_grp->irqs[num_irq++] =
>> + reo2host_destination_ring1 - j;
>> + }
>> +
>> + if (ab->hw_params->ring_mask->rx_err[i] & BIT(j))
>> + irq_grp->irqs[num_irq++] = reo2host_exception;
>> +
>> + if (ab->hw_params->ring_mask->rx_wbm_rel[i] & BIT(j))
>> + irq_grp->irqs[num_irq++] = wbm2host_rx_release;
>> +
>> + if (ab->hw_params->ring_mask->reo_status[i] & BIT(j))
>> + irq_grp->irqs[num_irq++] = reo2host_status;
>> +
>> + if (ab->hw_params->ring_mask->rx_mon_dest[i] & BIT(j))
>> + irq_grp->irqs[num_irq++] =
>> + rxdma2host_monitor_destination_mac1;
>> + }
>> +
>> + irq_grp->num_irq = num_irq;
>> +
>> + for (j = 0; j < irq_grp->num_irq; j++) {
>> + irq_idx = irq_grp->irqs[j];
>> +
>> + irq = platform_get_irq_byname(ab->pdev,
>> + irq_name[irq_idx]);
>> + ab->irq_num[irq_idx] = irq;
>> + irq_set_status_flags(irq, IRQ_NOAUTOEN | IRQ_DISABLE_UNLAZY);
>> + ret = request_irq(irq, ath12k_ahb_ext_interrupt_handler,
>> + IRQF_TRIGGER_RISING,
>> + irq_name[irq_idx], irq_grp);
>> + if (ret) {
>> + ath12k_err(ab, "failed request_irq for %d\n",
>> + irq);
>> + }
>> + }
>
> Instead of doing all this magic, can we request the IRQs manually, as we
> have interrupt-names in dt?
>
I'm not sure if I fully understood this comment.
If we manually request IRQs using their names from the DT, we won't be able to
group the IRQs. Grouping the IRQs is one of our main objectives here. Additionally,
we are not using all the IRQ names defined in the DT, so the logic in this function
is crucial for grouping and requesting the IRQs according to the ring-mask.
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static int ath12k_ahb_config_irq(struct ath12k_base *ab)
>> +{
>> + int irq, irq_idx, i;
>> + int ret;
>> +
>> + /* Configure CE irqs */
>> + for (i = 0; i < ab->hw_params->ce_count; i++) {
>> + struct ath12k_ce_pipe *ce_pipe = &ab->ce.ce_pipe[i];
>> +
>> + if (ath12k_ce_get_attr_flags(ab, i) & CE_ATTR_DIS_INTR)
>> + continue;
>> +
>> + irq_idx = ATH12K_IRQ_CE0_OFFSET + i;
>> +
>> + INIT_WORK(&ce_pipe->intr_wq, ath12k_ahb_ce_workqueue);
>> + irq = platform_get_irq_byname(ab->pdev, irq_name[irq_idx]);
>> + ret = request_irq(irq, ath12k_ahb_ce_interrupt_handler,
>> + IRQF_TRIGGER_RISING, irq_name[irq_idx],
>> + ce_pipe);
>> + if (ret)
>> + return ret;
>> +
>> + ab->irq_num[irq_idx] = irq;
>> + }
>> +
>> + /* Configure external interrupts */
>> + ret = ath12k_ahb_config_ext_irq(ab);
>> +
>> + return ret;
>> +}
>> +
>> +static int ath12k_ahb_map_service_to_pipe(struct ath12k_base *ab, u16 service_id,
>> + u8 *ul_pipe, u8 *dl_pipe)
>> +{
>> + const struct service_to_pipe *entry;
>> + bool ul_set = false, dl_set = false;
>> + int i;
>> +
>> + for (i = 0; i < ab->hw_params->svc_to_ce_map_len; i++) {
>> + entry = &ab->hw_params->svc_to_ce_map[i];
>> +
>> + if (__le32_to_cpu(entry->service_id) != service_id)
>> + continue;
>> +
>> + switch (__le32_to_cpu(entry->pipedir)) {
>> + case PIPEDIR_NONE:
>> + break;
>> + case PIPEDIR_IN:
>> + WARN_ON(dl_set);
>> + *dl_pipe = __le32_to_cpu(entry->pipenum);
>> + dl_set = true;
>> + break;
>> + case PIPEDIR_OUT:
>> + WARN_ON(ul_set);
>> + *ul_pipe = __le32_to_cpu(entry->pipenum);
>> + ul_set = true;
>> + break;
>> + case PIPEDIR_INOUT:
>> + WARN_ON(dl_set);
>> + WARN_ON(ul_set);
>> + *dl_pipe = __le32_to_cpu(entry->pipenum);
>> + *ul_pipe = __le32_to_cpu(entry->pipenum);
>> + dl_set = true;
>> + ul_set = true;
>> + break;
>> + }
>
> if pipedir == PIPEDIR_IN || pipedir == PIPEDIR_INOUT
> if pipedir == PIPEDIR_OUT || pipedir == PIPE_INOUT
>
> ?
>
Thanks for this logic simplification. Will use this in next version.
>> + }
>> +
>> + if (WARN_ON(!ul_set || !dl_set))
>> + return -ENOENT;
>> +
>> + return 0;
>> +}
>> +
>> +static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
>> + .start = ath12k_ahb_start,
>> + .stop = ath12k_ahb_stop,
>> + .read32 = ath12k_ahb_read32,
>> + .write32 = ath12k_ahb_write32,
>> + .cmem_read32 = ath12k_ahb_cmem_read32,
>> + .cmem_write32 = ath12k_ahb_cmem_write32,
>> + .irq_enable = ath12k_ahb_ext_irq_enable,
>> + .irq_disable = ath12k_ahb_ext_irq_disable,
>> + .map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
>> +};
>> +
>> +static int ath12k_ahb_clock_init(struct ath12k_base *ab)
>> +{
>> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
>> + int ret;
>> +
>> + ab_ahb->xo_clk = devm_clk_get(ab->dev, "gcc_xo_clk");
>> + if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
>> + ath12k_err(ab, "failed to get gcc_xo_clk: %d\n",
>> + PTR_ERR_OR_ZERO(ab_ahb->xo_clk));
>> + ret = ab_ahb->xo_clk ? PTR_ERR(ab_ahb->xo_clk) : -ENODEV;
>> + ab_ahb->xo_clk = NULL;
>> + return ret;
>> + }
>
> Won't clk core print errors in both of these cases?
>
Did not see any print error form clk core during validation. Hence added
the error print for ath12k.
>> +
>> + return 0;
>> +}
>> +
>> +static void ath12k_ahb_clock_deinit(struct ath12k_base *ab)
>> +{
>> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
>> +
>> + devm_clk_put(ab->dev, ab_ahb->xo_clk);
>> + ab_ahb->xo_clk = NULL;
>> +}
>> +
>> +static int ath12k_ahb_clock_enable(struct ath12k_base *ab)
>> +{
>> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
>> + int ret;
>> +
>> + if (IS_ERR_OR_NULL(ab_ahb->xo_clk)) {
>> + ath12k_err(ab, "clock is not initialized\n");
>> + return -EIO;
>> + }
>> +
>> + ret = clk_prepare_enable(ab_ahb->xo_clk);
>> + if (ret) {
>> + ath12k_err(ab, "failed to enable gcc_xo_clk: %d\n", ret);
>> + return ret;
>> + }
>> +
>> + return 0;
>> +}
>> +
>> +static void ath12k_ahb_clock_disable(struct ath12k_base *ab)
>> +{
>> + struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
>> +
>> + clk_disable_unprepare(ab_ahb->xo_clk);
>> +}
>> +
>> +static int ath12k_ahb_resource_init(struct ath12k_base *ab)
>> +{
>> + struct platform_device *pdev = ab->pdev;
>> + struct resource *mem_res;
>> + void __iomem *mem;
>> + int ret;
>> +
>> + mem = devm_platform_get_and_ioremap_resource(pdev, 0, &mem_res);
>> + if (IS_ERR(mem)) {
>> + dev_err(&pdev->dev, "ioremap error\n");
>> + ret = PTR_ERR(mem);
>> + goto out;
>
> If you assign ab->mem directly, you can get rid of the line below
> and return the error here
>
Thanks, will update in next version.
>> + }
>> +
>> + ab->mem = mem;
>> + ab->mem_len = resource_size(mem_res);
>> +
>> + if (ab->hw_params->ce_remap) {
>> + const struct ce_remap *ce_remap = ab->hw_params->ce_remap;
>> + /* ce register space is moved out of wcss and the space is not
>> + * contiguous, hence remapping the CE registers to a new space
>> + * for accessing them.
>> + */
>
> Please capitalize words consistently
>
sure, will update in next version.
>> + ab->mem_ce = ioremap(ce_remap->base, ce_remap->size);
>> + if (IS_ERR(ab->mem_ce)) {
>> + dev_err(&pdev->dev, "ce ioremap error\n");
>> + ret = -ENOMEM;
>> + goto err_mem_unmap;
>> + }
>> + ab->ce_remap = true;
>> + ab->ce_remap_base_addr = HAL_IPQ5332_CE_WFSS_REG_BASE;
>> + }
>> +
>> + if (ab->hw_params->cmem_remap) {
>> + const struct cmem_remap *cmem_remap = ab->hw_params->cmem_remap;
>> + /* For device like IPQ5332 CMEM region is outside WCSS block.
>
> IPQ5332 is not a 'device'
>
>> + * Allocate separate I/O remap to access CMEM address.
>> + */
>> + ab->mem_cmem = ioremap(cmem_remap->base, cmem_remap->size);
>> + if (IS_ERR(ab->mem_cmem)) {
>> + dev_err(&pdev->dev, "cmem ioremap error\n");
>> + ret = -ENOMEM;
>> + goto err_mem_ce_unmap;
>> + }
>> + }
>
> [...]
>
>> + ret = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
>> + if (ret) {
>> + dev_err(&pdev->dev, "Failed to set 32-bit consistent dma\n");
>
> s/consistent/coherent
>
> You're setting the mask, not the DMA itself
>
Thanks will update in next version.
> [...]
>
>> + /* Set fixed_mem_region to true for platforms that support fixed memory
>> + * reservation from DT. If memory is reserved from DT for FW, ath12k driver
>> + * need not to allocate memory.
>> + */
>> + if (!of_property_read_u32(ab->dev->of_node, "memory-region", &addr)) {
>> + set_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags);
>> + mem_node = of_find_node_by_name(NULL, "mlo_global_mem_0");
>
> This is not mentioned or documented anywhere.
>
In next version, will document the below info:
"If the platform supports fixed memory, then it should define/reserve
MLO global memory in DT to support Multi Link Operation.
If MLO global memory is not reserved in fixed memory mode, then
MLO cannot be supported."
^ permalink raw reply [flat|nested] 60+ messages in thread* Re: [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332
2024-12-06 9:56 ` Raj Kumar Bhagat
@ 2024-12-06 11:46 ` Konrad Dybcio
0 siblings, 0 replies; 60+ messages in thread
From: Konrad Dybcio @ 2024-12-06 11:46 UTC (permalink / raw)
To: Raj Kumar Bhagat, Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 6.12.2024 10:56 AM, Raj Kumar Bhagat wrote:
> On 10/19/2024 1:59 AM, Konrad Dybcio wrote:
>> On 15.10.2024 8:26 PM, Raj Kumar Bhagat wrote:
>>> From: Balamurugan S <quic_bselvara@quicinc.com>
>>>
>>> Add Initial Ath12k AHB driver support for IPQ5332. IPQ5332 is AHB
>>> based IEEE802.11be 2 GHz 2x2 WiFi device.
>>>
>>> Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
>>> Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
>>>
>>> Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
>>> Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
>>> Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
>>> Co-developed-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>>> Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
>>> ---
>>
>> [...]
>>
>>> +enum ext_irq_num {
>>> + host2wbm_desc_feed = 16,
>>> + host2reo_re_injection,
>>
>> Why?
>>
>
> This enum is used as a IRQ number for Ath12k AHB. Based on this enum
> we can get the IRQ name from irq_name[]. This helps to request the original
> IRQ number from the DT.
> It is starting from 16 becasue, in irq_name[], the name for ext IRQ starts
> from 16 index.
[...]
>
>>> + irq_grp->irqs[num_irq++] =
>>> + reo2host_destination_ring1 - j;
>>> + }
>>> +
>>> + if (ab->hw_params->ring_mask->rx_err[i] & BIT(j))
>>> + irq_grp->irqs[num_irq++] = reo2host_exception;
>>> +
>>> + if (ab->hw_params->ring_mask->rx_wbm_rel[i] & BIT(j))
>>> + irq_grp->irqs[num_irq++] = wbm2host_rx_release;
>>> +
>>> + if (ab->hw_params->ring_mask->reo_status[i] & BIT(j))
>>> + irq_grp->irqs[num_irq++] = reo2host_status;
>>> +
>>> + if (ab->hw_params->ring_mask->rx_mon_dest[i] & BIT(j))
>>> + irq_grp->irqs[num_irq++] =
>>> + rxdma2host_monitor_destination_mac1;
>>> + }
>>> +
>>> + irq_grp->num_irq = num_irq;
>>> +
>>> + for (j = 0; j < irq_grp->num_irq; j++) {
>>> + irq_idx = irq_grp->irqs[j];
>>> +
>>> + irq = platform_get_irq_byname(ab->pdev,
>>> + irq_name[irq_idx]);
>>> + ab->irq_num[irq_idx] = irq;
>>> + irq_set_status_flags(irq, IRQ_NOAUTOEN | IRQ_DISABLE_UNLAZY);
>>> + ret = request_irq(irq, ath12k_ahb_ext_interrupt_handler,
>>> + IRQF_TRIGGER_RISING,
>>> + irq_name[irq_idx], irq_grp);
>>> + if (ret) {
>>> + ath12k_err(ab, "failed request_irq for %d\n",
>>> + irq);
>>> + }
>>> + }
>>
>> Instead of doing all this magic, can we request the IRQs manually, as we
>> have interrupt-names in dt?
>>
>
> I'm not sure if I fully understood this comment.
> If we manually request IRQs using their names from the DT, we won't be able to
> group the IRQs. Grouping the IRQs is one of our main objectives here. Additionally,
> we are not using all the IRQ names defined in the DT, so the logic in this function
> is crucial for grouping and requesting the IRQs according to the ring-mask.
Surely you can name these "foo_bar_ring%d" in DT and use the OF APIs
[...]
>>
>>> + /* Set fixed_mem_region to true for platforms that support fixed memory
>>> + * reservation from DT. If memory is reserved from DT for FW, ath12k driver
>>> + * need not to allocate memory.
>>> + */
>>> + if (!of_property_read_u32(ab->dev->of_node, "memory-region", &addr)) {
>>> + set_bit(ATH12K_FLAG_FIXED_MEM_REGION, &ab->dev_flags);
>>> + mem_node = of_find_node_by_name(NULL, "mlo_global_mem_0");
>>
>> This is not mentioned or documented anywhere.
>>
>
> In next version, will document the below info:
>
> "If the platform supports fixed memory, then it should define/reserve
> MLO global memory in DT to support Multi Link Operation.
> If MLO global memory is not reserved in fixed memory mode, then
> MLO cannot be supported."
You should also explain what Multi Link Operation means
Konrad
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332
2024-10-18 20:29 ` Konrad Dybcio
2024-12-06 9:56 ` Raj Kumar Bhagat
@ 2024-12-06 10:00 ` Raj Kumar Bhagat
1 sibling, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-06 10:00 UTC (permalink / raw)
To: Konrad Dybcio, ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh
On 10/19/2024 1:59 AM, Konrad Dybcio wrote:
>> +static int ath12k_ahb_start(struct ath12k_base *ab)
>> +{
>> + ath12k_ahb_ce_irqs_enable(ab);
>> + ath12k_ce_rx_post_buf(ab);
>> +
>> + return 0;
>> +}
> Neither this nor ath12k_pci returns anything useful - perhaps make this void?
Thanks for the comment, this requires changes to be done also in ath12k pci and
hif_ops, hence, will take is up separately.
^ permalink raw reply [flat|nested] 60+ messages in thread
* [PATCH v2 18/22] wifi: ath12k: Power up root PD
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (16 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 17/22] wifi: ath12k: add AHB driver support for IPQ5332 Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 19/22] wifi: ath12k: Register various userPD interrupts and save SMEM entries Raj Kumar Bhagat
` (4 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Sowmiya Sree Elavalagan,
Raj Kumar Bhagat
From: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Q6 processor acts as rootPD, other hardware like IPQ5332 which are
attached to Q6 act as userPDs. WCSS driver handles loading and booting
of rootPD, while the ath12k driver boots the userPD.
Get the rproc handle from the DTS entry and boot the rootPD if it
is not already powered on. Register to the rproc notifier to monitor
the rproc state, this allows ath12k driver to know power up/down
sequence of the rootPD. Power up the rootPD and wait for a power-up
notification from the notifier callback before powering up the userPDs.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Signed-off-by: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ahb.c | 128 ++++++++++++++++++++++++++
drivers/net/wireless/ath/ath12k/ahb.h | 7 ++
2 files changed, 135 insertions(+)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
index 7049efdae38c..6c3943b63ac5 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.c
+++ b/drivers/net/wireless/ath/ath12k/ahb.c
@@ -608,6 +608,124 @@ static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
.map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
};
+static int ath12k_ahb_root_pd_state_notifier(struct notifier_block *nb,
+ const unsigned long event, void *data)
+{
+ struct ath12k_ahb *ab_ahb = container_of(nb, struct ath12k_ahb, root_pd_nb);
+ struct ath12k_base *ab = ab_ahb->ab;
+
+ if (event == ATH12K_RPROC_AFTER_POWERUP) {
+ ath12k_dbg(ab, ATH12K_DBG_AHB, "Root PD is UP\n");
+ complete(&ab_ahb->rootpd_ready);
+ }
+
+ return 0;
+}
+
+static int ath12k_ahb_register_rproc_notifier(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+
+ ab_ahb->root_pd_nb.notifier_call = ath12k_ahb_root_pd_state_notifier;
+ init_completion(&ab_ahb->rootpd_ready);
+
+ ab_ahb->root_pd_notifier = qcom_register_ssr_notifier(ab_ahb->tgt_rproc->name,
+ &ab_ahb->root_pd_nb);
+
+ if (!ab_ahb->root_pd_notifier)
+ return -EINVAL;
+
+ return 0;
+}
+
+static void ath12k_ahb_unregister_rproc_notifier(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+
+ if (!ab_ahb->root_pd_notifier) {
+ ath12k_err(ab, "Rproc notifier not registered\n");
+ return;
+ }
+
+ qcom_unregister_ssr_notifier(ab_ahb->root_pd_notifier,
+ &ab_ahb->root_pd_nb);
+}
+
+static int ath12k_ahb_get_rproc(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ struct device *dev = ab->dev;
+ struct rproc *prproc;
+ phandle rproc_phandle;
+
+ if (of_property_read_u32(dev->of_node, "qcom,rproc", &rproc_phandle)) {
+ ath12k_err(ab, "failed to get q6_rproc handle\n");
+ return -ENOENT;
+ }
+
+ prproc = rproc_get_by_phandle(rproc_phandle);
+ if (!prproc) {
+ ath12k_err(ab, "failed to get rproc\n");
+ return -EINVAL;
+ }
+ ab_ahb->tgt_rproc = prproc;
+
+ return 0;
+}
+
+static int ath12k_ahb_boot_root_pd(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ unsigned long time_left;
+ int ret;
+
+ ret = rproc_boot(ab_ahb->tgt_rproc);
+ if (ret < 0) {
+ ath12k_err(ab, "RootPD boot failed\n");
+ return ret;
+ }
+
+ time_left = wait_for_completion_timeout(&ab_ahb->rootpd_ready,
+ ATH12K_ROOTPD_READY_TIMEOUT);
+ if (!time_left) {
+ ath12k_err(ab, "RootPD ready wait timed out\n");
+ return -ETIMEDOUT;
+ }
+
+ return 0;
+}
+
+static int ath12k_ahb_configure_rproc(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ int ret;
+
+ ret = ath12k_ahb_get_rproc(ab);
+ if (ret < 0) {
+ ath12k_err(ab, "failed to get rproc: %d\n", ret);
+ return ret;
+ }
+
+ ret = ath12k_ahb_register_rproc_notifier(ab);
+ if (ret < 0) {
+ ath12k_err(ab, "failed to register rproc notifier\n");
+ return ret;
+ }
+
+ if (ab_ahb->tgt_rproc->state != RPROC_RUNNING) {
+ ret = ath12k_ahb_boot_root_pd(ab);
+ if (ret < 0) {
+ ath12k_err(ab, "failed to boot the remote processor Q6\n");
+ goto unreg_notifier;
+ }
+ }
+ return 0;
+
+unreg_notifier:
+ ath12k_ahb_unregister_rproc_notifier(ab);
+ return ret;
+}
+
static int ath12k_ahb_clock_init(struct ath12k_base *ab)
{
struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
@@ -773,6 +891,7 @@ static int ath12k_ahb_probe(struct platform_device *pdev)
struct ath12k_base *ab;
const struct ath12k_hif_ops *hif_ops;
struct device_node *mem_node;
+ struct ath12k_ahb *ab_ahb;
enum ath12k_hw_rev hw_rev;
u32 addr;
int ret;
@@ -804,6 +923,8 @@ static int ath12k_ahb_probe(struct platform_device *pdev)
ab->pdev = pdev;
ab->hw_rev = hw_rev;
platform_set_drvdata(pdev, ab);
+ ab_ahb = ath12k_ab_to_ahb(ab);
+ ab_ahb->ab = ab;
/* Set fixed_mem_region to true for platforms that support fixed memory
* reservation from DT. If memory is reserved from DT for FW, ath12k driver
@@ -836,6 +957,12 @@ static int ath12k_ahb_probe(struct platform_device *pdev)
ath12k_ahb_init_qmi_ce_config(ab);
+ ret = ath12k_ahb_configure_rproc(ab);
+ if (ret) {
+ ath12k_err(ab, "failed to configure rproc: %d\n", ret);
+ goto err_ce_free;
+ }
+
ret = ath12k_ahb_config_irq(ab);
if (ret) {
ath12k_err(ab, "failed to configure irq: %d\n", ret);
@@ -880,6 +1007,7 @@ static void ath12k_ahb_remove_prepare(struct ath12k_base *ab)
set_bit(ATH12K_FLAG_UNREGISTERING, &ab->dev_flags);
cancel_work_sync(&ab->restart_work);
cancel_work_sync(&ab->qmi.event_work);
+ ath12k_ahb_unregister_rproc_notifier(ab);
}
static void ath12k_ahb_free_resources(struct ath12k_base *ab)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.h b/drivers/net/wireless/ath/ath12k/ahb.h
index bd0366a79587..d1fc63091fb0 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.h
+++ b/drivers/net/wireless/ath/ath12k/ahb.h
@@ -7,6 +7,7 @@
#define ATH12K_AHB_H
#include <linux/clk.h>
+#include <linux/remoteproc/qcom_rproc.h>
#include "core.h"
#define ATH12K_AHB_RECOVERY_TIMEOUT (3 * HZ)
@@ -16,6 +17,8 @@
#define ATH12K_AHB_SMP2P_SMEM_VALUE_MASK 0xFFFFFFFF
#define ATH12K_PCI_CE_WAKE_IRQ 2
#define ATH12K_PCI_IRQ_CE0_OFFSET 3
+#define ATH12K_ROOTPD_READY_TIMEOUT (5 * HZ)
+#define ATH12K_RPROC_AFTER_POWERUP QCOM_SSR_AFTER_POWERUP
enum ath12k_ahb_smp2p_msg_id {
ATH12K_AHB_POWER_SAVE_ENTER = 1,
@@ -25,8 +28,12 @@ enum ath12k_ahb_smp2p_msg_id {
struct ath12k_base;
struct ath12k_ahb {
+ struct ath12k_base *ab;
struct rproc *tgt_rproc;
struct clk *xo_clk;
+ struct completion rootpd_ready;
+ struct notifier_block root_pd_nb;
+ void *root_pd_notifier;
};
static inline struct ath12k_ahb *ath12k_ab_to_ahb(struct ath12k_base *ab)
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 19/22] wifi: ath12k: Register various userPD interrupts and save SMEM entries
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (17 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 18/22] wifi: ath12k: Power up root PD Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 20/22] wifi: ath12k: Power up userPD Raj Kumar Bhagat
` (3 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Sowmiya Sree Elavalagan,
Raj Kumar Bhagat
From: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Q6 and ath12k driver communicates using SMEM and IRQs. Spawn interrupt
is triggered once the userPD thread is spawned. Ready interrupts denotes
userPD is completely powered up and ready. Stop-ack is to acknowledge
the ath12k driver that userPD is stopped. Ath12k driver needs to set spawn
bit in SMEM to instruct Q6 to spawn a userPD. Similarly stop bit is
set when userPD needs to be stopped.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Signed-off-by: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ahb.c | 84 ++++++++++++++++++++++++++-
drivers/net/wireless/ath/ath12k/ahb.h | 16 +++++
drivers/net/wireless/ath/ath12k/hw.h | 1 +
3 files changed, 99 insertions(+), 2 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
index 6c3943b63ac5..7b898f01c255 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.c
+++ b/drivers/net/wireless/ath/ath12k/ahb.c
@@ -28,6 +28,11 @@ static const struct of_device_id ath12k_ahb_of_match[] = {
MODULE_DEVICE_TABLE(of, ath12k_ahb_of_match);
#define ATH12K_IRQ_CE0_OFFSET 4
+#define ATH12K_MAX_UPDS 1
+#define ATH12K_UPD_IRQ_WRD_LEN 18
+static const char ath12k_userpd_irq[][9] = {"spawn",
+ "ready",
+ "stop-ack"};
static const char *irq_name[ATH12K_IRQ_NUM_MAX] = {
"misc-pulse1",
@@ -608,6 +613,78 @@ static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
.map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
};
+static irqreturn_t ath12k_userpd_irq_handler(int irq, void *data)
+{
+ struct ath12k_base *ab = data;
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+
+ if (irq == ab_ahb->userpd_irq_num[ATH12K_USERPD_SPAWN_IRQ]) {
+ complete(&ab_ahb->userpd_spawned);
+ } else if (irq == ab_ahb->userpd_irq_num[ATH12K_USERPD_READY_IRQ]) {
+ complete(&ab_ahb->userpd_ready);
+ } else if (irq == ab_ahb->userpd_irq_num[ATH12K_USERPD_STOP_ACK_IRQ]) {
+ complete(&ab_ahb->userpd_stopped);
+ } else {
+ ath12k_err(ab, "Invalid userpd interrupt\n");
+ return IRQ_NONE;
+ }
+
+ return IRQ_HANDLED;
+}
+
+static int ath12k_ahb_config_rproc_irq(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ int i, ret;
+ char *upd_irq_name;
+
+ for (i = 0; i < ATH12K_USERPD_MAX_IRQ; i++) {
+ ab_ahb->userpd_irq_num[i] = platform_get_irq_byname(ab->pdev,
+ ath12k_userpd_irq[i]);
+ if (ab_ahb->userpd_irq_num[i] < 0) {
+ ath12k_err(ab, "Failed to get %s irq: %d", ath12k_userpd_irq[i],
+ ab_ahb->userpd_irq_num[i]);
+ return -EINVAL;
+ }
+
+ upd_irq_name = devm_kzalloc(&ab->pdev->dev, ATH12K_UPD_IRQ_WRD_LEN,
+ GFP_KERNEL);
+ if (!upd_irq_name)
+ return -ENOMEM;
+
+ scnprintf(upd_irq_name, ATH12K_UPD_IRQ_WRD_LEN, "UserPD%u-%s",
+ ab_ahb->userpd_id, ath12k_userpd_irq[i]);
+ ret = devm_request_threaded_irq(&ab->pdev->dev, ab_ahb->userpd_irq_num[i],
+ NULL, ath12k_userpd_irq_handler,
+ IRQF_TRIGGER_RISING | IRQF_ONESHOT,
+ upd_irq_name, ab);
+ if (ret) {
+ ath12k_err(ab, "Request %s irq failed: %d\n",
+ ath12k_userpd_irq[i], ret);
+ return ret;
+ }
+ }
+
+ ab_ahb->spawn_state = devm_qcom_smem_state_get(&ab->pdev->dev, "spawn",
+ &ab_ahb->spawn_bit);
+ if (IS_ERR(ab_ahb->spawn_state)) {
+ ath12k_err(ab, "Failed to acquire spawn state\n");
+ return PTR_ERR(ab_ahb->spawn_state);
+ }
+
+ ab_ahb->stop_state = devm_qcom_smem_state_get(&ab->pdev->dev, "stop",
+ &ab_ahb->stop_bit);
+ if (IS_ERR(ab_ahb->stop_state)) {
+ ath12k_err(ab, "Failed to acquire stop state\n");
+ return PTR_ERR(ab_ahb->stop_state);
+ }
+
+ init_completion(&ab_ahb->userpd_spawned);
+ init_completion(&ab_ahb->userpd_ready);
+ init_completion(&ab_ahb->userpd_stopped);
+ return 0;
+}
+
static int ath12k_ahb_root_pd_state_notifier(struct notifier_block *nb,
const unsigned long event, void *data)
{
@@ -719,7 +796,8 @@ static int ath12k_ahb_configure_rproc(struct ath12k_base *ab)
goto unreg_notifier;
}
}
- return 0;
+
+ return ath12k_ahb_config_rproc_irq(ab);
unreg_notifier:
ath12k_ahb_unregister_rproc_notifier(ab);
@@ -893,13 +971,14 @@ static int ath12k_ahb_probe(struct platform_device *pdev)
struct device_node *mem_node;
struct ath12k_ahb *ab_ahb;
enum ath12k_hw_rev hw_rev;
- u32 addr;
+ u32 addr, userpd_id;
int ret;
hw_rev = ath12k_ahb_get_hw_rev(pdev);
switch (hw_rev) {
case ATH12K_HW_IPQ5332_HW10:
hif_ops = &ath12k_ahb_hif_ops_ipq5332;
+ userpd_id = ATH12K_IPQ5332_USERPD_ID;
break;
default:
dev_err(&pdev->dev, "Unsupported device type %d\n", hw_rev);
@@ -925,6 +1004,7 @@ static int ath12k_ahb_probe(struct platform_device *pdev)
platform_set_drvdata(pdev, ab);
ab_ahb = ath12k_ab_to_ahb(ab);
ab_ahb->ab = ab;
+ ab_ahb->userpd_id = userpd_id;
/* Set fixed_mem_region to true for platforms that support fixed memory
* reservation from DT. If memory is reserved from DT for FW, ath12k driver
diff --git a/drivers/net/wireless/ath/ath12k/ahb.h b/drivers/net/wireless/ath/ath12k/ahb.h
index d1fc63091fb0..0999e2bbe970 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.h
+++ b/drivers/net/wireless/ath/ath12k/ahb.h
@@ -25,6 +25,13 @@ enum ath12k_ahb_smp2p_msg_id {
ATH12K_AHB_POWER_SAVE_EXIT,
};
+enum ath12k_ahb_userpd_irq {
+ ATH12K_USERPD_SPAWN_IRQ,
+ ATH12K_USERPD_READY_IRQ,
+ ATH12K_USERPD_STOP_ACK_IRQ,
+ ATH12K_USERPD_MAX_IRQ,
+};
+
struct ath12k_base;
struct ath12k_ahb {
@@ -34,6 +41,15 @@ struct ath12k_ahb {
struct completion rootpd_ready;
struct notifier_block root_pd_nb;
void *root_pd_notifier;
+ struct qcom_smem_state *spawn_state;
+ struct qcom_smem_state *stop_state;
+ struct completion userpd_spawned;
+ struct completion userpd_ready;
+ struct completion userpd_stopped;
+ u32 userpd_id;
+ u32 spawn_bit;
+ u32 stop_bit;
+ int userpd_irq_num[ATH12K_USERPD_MAX_IRQ];
};
static inline struct ath12k_ahb *ath12k_ab_to_ahb(struct ath12k_base *ab)
diff --git a/drivers/net/wireless/ath/ath12k/hw.h b/drivers/net/wireless/ath/ath12k/hw.h
index b3e2a5d7e0e4..cd034ed14ce3 100644
--- a/drivers/net/wireless/ath/ath12k/hw.h
+++ b/drivers/net/wireless/ath/ath12k/hw.h
@@ -97,6 +97,7 @@
#define ATH12K_REGDB_FILE_NAME "regdb.bin"
#define ATH12K_PCIE_MAX_PAYLOAD_SIZE 128
+#define ATH12K_IPQ5332_USERPD_ID 1
enum ath12k_hw_rate_cck {
ATH12K_HW_RATE_CCK_LP_11M = 0,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 20/22] wifi: ath12k: Power up userPD
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (18 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 19/22] wifi: ath12k: Register various userPD interrupts and save SMEM entries Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 21/22] wifi: ath12k: Power down userPD Raj Kumar Bhagat
` (2 subsequent siblings)
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Sowmiya Sree Elavalagan,
Raj Kumar Bhagat
From: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
UserPD firmware image is loaded and booted by ath12k driver. Get the userPD
memory region from DTS and load the firmware for userPD from pre-defined
path into io-remapped address of this region. Authenticate this image
using pasid which is a peripheral ID. Set the spawn bit to instruct Q6
to spawn userPD thread. Wait for userPD to spawn which is indicated by
spawn interrupt. Ready interrupt is triggered once the userPD is powered
up completely.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Signed-off-by: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ahb.c | 138 ++++++++++++++++++++++++++
drivers/net/wireless/ath/ath12k/ahb.h | 10 +-
2 files changed, 147 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
index 7b898f01c255..fa3927295c4e 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.c
+++ b/drivers/net/wireless/ath/ath12k/ahb.c
@@ -5,13 +5,17 @@
*/
#include <linux/dma-mapping.h>
+#include <linux/firmware.h>
+#include <linux/firmware/qcom/qcom_scm.h>
#include <linux/iommu.h>
#include <linux/module.h>
#include <linux/of.h>
#include <linux/of_address.h>
#include <linux/of_device.h>
+#include <linux/of_reserved_mem.h>
#include <linux/platform_device.h>
#include <linux/remoteproc.h>
+#include <linux/soc/qcom/mdt_loader.h>
#include <linux/soc/qcom/smem.h>
#include <linux/soc/qcom/smem_state.h>
#include "ahb.h"
@@ -357,6 +361,139 @@ static void ath12k_ahb_stop(struct ath12k_base *ab)
ath12k_ce_cleanup_pipes(ab);
}
+static int ath12k_ahb_power_up(struct ath12k_base *ab)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ u32 pasid;
+ char fw_name[ATH12K_USERPD_FW_NAME_LEN];
+ char fw2_name[ATH12K_USERPD_FW_NAME_LEN];
+ struct device *dev = ab->dev;
+ const struct firmware *fw, *fw2;
+ struct device_node *fw_mem_node;
+ struct reserved_mem *rmem = NULL;
+ unsigned long time_left;
+ phys_addr_t mem_phys;
+ void *mem_region;
+ size_t mem_size;
+ int ret;
+
+ fw_mem_node = of_parse_phandle(dev->of_node, "memory-region", 0);
+ if (fw_mem_node)
+ rmem = of_reserved_mem_lookup(fw_mem_node);
+
+ of_node_put(fw_mem_node);
+
+ if (!rmem) {
+ ath12k_err(ab, "Failed to acquire Q6 memory-region\n");
+ return -EINVAL;
+ }
+
+ mem_phys = rmem->base;
+ mem_size = rmem->size;
+ mem_region = (void *)devm_ioremap_wc(dev, mem_phys, mem_size);
+ if (!mem_region) {
+ ath12k_err(ab, "unable to map memory region: %pa+%pa\n",
+ &rmem->base, &rmem->size);
+ return -ENOMEM;
+ }
+
+ snprintf(fw_name, sizeof(fw_name), "%s/%s/%s%d%s", ATH12K_FW_DIR,
+ ab->hw_params->fw.dir, ATH12K_AHB_FW_PREFIX, ab_ahb->userpd_id,
+ ATH12K_AHB_FW_SUFFIX);
+
+ ret = request_firmware(&fw, fw_name, dev);
+ if (ret < 0) {
+ ath12k_err(ab, "request_firmware failed\n");
+ return ret;
+ }
+
+ ath12k_dbg(ab, ATH12K_DBG_AHB, "Booting fw image %s, size %zd\n", fw_name,
+ fw->size);
+
+ if (!fw->size) {
+ ath12k_err(ab, "Invalid firmware size\n");
+ ret = -EINVAL;
+ goto err_fw;
+ }
+
+ pasid = (u32_encode_bits(ab_ahb->userpd_id, ATH12K_USERPD_ID_MASK)) |
+ ATH12K_AHB_UPD_SWID;
+
+ /* Load FW image to a reserved memory location */
+ ret = qcom_mdt_load(dev, fw, fw_name, pasid, mem_region, mem_phys, mem_size,
+ &mem_phys);
+ if (ret) {
+ ath12k_err(ab, "Failed to load MDT segments: %d\n", ret);
+ goto err_fw;
+ }
+
+ snprintf(fw2_name, sizeof(fw2_name), "%s/%s/%s", ATH12K_FW_DIR,
+ ab->hw_params->fw.dir, ATH12K_AHB_FW2);
+
+ ret = request_firmware(&fw2, fw2_name, dev);
+ if (ret < 0) {
+ ath12k_err(ab, "request_firmware failed\n");
+ goto err_fw;
+ }
+
+ ath12k_dbg(ab, ATH12K_DBG_AHB, "Booting fw image %s, size %zd\n", fw2_name,
+ fw2->size);
+
+ if (!fw2->size) {
+ ath12k_err(ab, "Invalid firmware size\n");
+ ret = -EINVAL;
+ goto err_fw2;
+ }
+
+ ret = qcom_mdt_load_no_init(dev, fw2, fw2_name, pasid, mem_region, mem_phys,
+ mem_size, &mem_phys);
+ if (ret) {
+ ath12k_err(ab, "Failed to load MDT segments: %d\n", ret);
+ goto err_fw2;
+ }
+
+ /* Authenticate FW image using peripheral ID */
+ ret = qcom_scm_pas_auth_and_reset(pasid);
+ if (ret) {
+ ath12k_err(ab, "failed to boot the remote processor %d\n", ret);
+ goto err_fw2;
+ }
+
+ /* Instruct Q6 to spawn userPD thread */
+ ret = qcom_smem_state_update_bits(ab_ahb->spawn_state, BIT(ab_ahb->spawn_bit),
+ BIT(ab_ahb->spawn_bit));
+ if (ret) {
+ ath12k_err(ab, "Failed to update spawn state %d\n", ret);
+ goto err_fw2;
+ }
+
+ time_left = wait_for_completion_timeout(&ab_ahb->userpd_spawned,
+ ATH12K_USERPD_SPAWN_TIMEOUT);
+ if (!time_left) {
+ ath12k_err(ab, "UserPD spawn wait timed out\n");
+ ret = -ETIMEDOUT;
+ goto err_fw2;
+ }
+
+ time_left = wait_for_completion_timeout(&ab_ahb->userpd_ready,
+ ATH12K_USERPD_READY_TIMEOUT);
+ if (!time_left) {
+ ath12k_err(ab, "UserPD ready wait timed out\n");
+ ret = -ETIMEDOUT;
+ goto err_fw2;
+ }
+
+ qcom_smem_state_update_bits(ab_ahb->spawn_state, BIT(ab_ahb->spawn_bit), 0);
+
+ ath12k_info(ab, "UserPD%d is now UP\n", ab_ahb->userpd_id);
+
+err_fw2:
+ release_firmware(fw2);
+err_fw:
+ release_firmware(fw);
+ return ret;
+}
+
static void ath12k_ahb_init_qmi_ce_config(struct ath12k_base *ab)
{
struct ath12k_qmi_ce_cfg *cfg = &ab->qmi.ce_cfg;
@@ -611,6 +748,7 @@ static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
.irq_enable = ath12k_ahb_ext_irq_enable,
.irq_disable = ath12k_ahb_ext_irq_disable,
.map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
+ .power_up = ath12k_ahb_power_up,
};
static irqreturn_t ath12k_userpd_irq_handler(int irq, void *data)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.h b/drivers/net/wireless/ath/ath12k/ahb.h
index 0999e2bbe970..0dbbbfd45eab 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.h
+++ b/drivers/net/wireless/ath/ath12k/ahb.h
@@ -19,7 +19,15 @@
#define ATH12K_PCI_IRQ_CE0_OFFSET 3
#define ATH12K_ROOTPD_READY_TIMEOUT (5 * HZ)
#define ATH12K_RPROC_AFTER_POWERUP QCOM_SSR_AFTER_POWERUP
-
+#define ATH12K_AHB_FW_PREFIX "q6_fw"
+#define ATH12K_AHB_FW_SUFFIX ".mdt"
+#define ATH12K_AHB_FW2 "iu_fw.mdt"
+#define ATH12K_AHB_UPD_SWID 0x12
+#define ATH12K_USERPD_SPAWN_TIMEOUT (5 * HZ)
+#define ATH12K_USERPD_READY_TIMEOUT (10 * HZ)
+#define ATH12K_USERPD_STOP_TIMEOUT (5 * HZ)
+#define ATH12K_USERPD_ID_MASK GENMASK(9, 8)
+#define ATH12K_USERPD_FW_NAME_LEN 35
enum ath12k_ahb_smp2p_msg_id {
ATH12K_AHB_POWER_SAVE_ENTER = 1,
ATH12K_AHB_POWER_SAVE_EXIT,
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 21/22] wifi: ath12k: Power down userPD
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (19 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 20/22] wifi: ath12k: Power up userPD Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-15 18:26 ` [PATCH v2 22/22] wifi: ath12k: enable ath12k AHB support Raj Kumar Bhagat
2024-10-16 6:57 ` [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Krzysztof Kozlowski
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Sowmiya Sree Elavalagan,
Raj Kumar Bhagat
From: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Set the stop bit in SMEM to power down the userPD. Wait for stop-ack IRQ
to indicate power down completion. Release the userPD firmware using its
peripheral ID.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Signed-off-by: Sowmiya Sree Elavalagan <quic_ssreeela@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/ahb.c | 30 +++++++++++++++++++++++++++
1 file changed, 30 insertions(+)
diff --git a/drivers/net/wireless/ath/ath12k/ahb.c b/drivers/net/wireless/ath/ath12k/ahb.c
index fa3927295c4e..50907cf648f8 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.c
+++ b/drivers/net/wireless/ath/ath12k/ahb.c
@@ -494,6 +494,34 @@ static int ath12k_ahb_power_up(struct ath12k_base *ab)
return ret;
}
+static void ath12k_ahb_power_down(struct ath12k_base *ab, bool is_suspend)
+{
+ struct ath12k_ahb *ab_ahb = ath12k_ab_to_ahb(ab);
+ unsigned long time_left;
+ u32 pasid;
+ int ret;
+
+ qcom_smem_state_update_bits(ab_ahb->stop_state, BIT(ab_ahb->stop_bit),
+ BIT(ab_ahb->stop_bit));
+
+ time_left = wait_for_completion_timeout(&ab_ahb->userpd_stopped,
+ ATH12K_USERPD_STOP_TIMEOUT);
+ if (!time_left) {
+ ath12k_err(ab, "UserPD stop wait timed out\n");
+ return;
+ }
+
+ qcom_smem_state_update_bits(ab_ahb->stop_state, BIT(ab_ahb->stop_bit), 0);
+
+ pasid = (u32_encode_bits(ab_ahb->userpd_id, ATH12K_USERPD_ID_MASK)) |
+ ATH12K_AHB_UPD_SWID;
+ /* Release the firmware */
+ ret = qcom_scm_pas_shutdown(pasid);
+ if (ret)
+ ath12k_err(ab, "scm pas shutdown failed for userPD%d\n",
+ ab_ahb->userpd_id);
+}
+
static void ath12k_ahb_init_qmi_ce_config(struct ath12k_base *ab)
{
struct ath12k_qmi_ce_cfg *cfg = &ab->qmi.ce_cfg;
@@ -749,6 +777,7 @@ static const struct ath12k_hif_ops ath12k_ahb_hif_ops_ipq5332 = {
.irq_disable = ath12k_ahb_ext_irq_disable,
.map_service_to_pipe = ath12k_ahb_map_service_to_pipe,
.power_up = ath12k_ahb_power_up,
+ .power_down = ath12k_ahb_power_down,
};
static irqreturn_t ath12k_userpd_irq_handler(int irq, void *data)
@@ -1245,6 +1274,7 @@ static void ath12k_ahb_remove(struct platform_device *pdev)
struct ath12k_base *ab = platform_get_drvdata(pdev);
if (test_bit(ATH12K_FLAG_QMI_FAIL, &ab->dev_flags)) {
+ ath12k_ahb_power_down(ab, false);
ath12k_qmi_deinit_service(ab);
goto qmi_fail;
}
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* [PATCH v2 22/22] wifi: ath12k: enable ath12k AHB support
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (20 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 21/22] wifi: ath12k: Power down userPD Raj Kumar Bhagat
@ 2024-10-15 18:26 ` Raj Kumar Bhagat
2024-10-16 6:57 ` [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Krzysztof Kozlowski
22 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-10-15 18:26 UTC (permalink / raw)
To: ath12k
Cc: linux-wireless, Kalle Valo, Rob Herring, Krzysztof Kozlowski,
Conor Dooley, Jeff Johnson, Bjorn Andersson, Konrad Dybcio,
devicetree, linux-kernel, linux-arm-msm, Balamurugan S,
P Praneesh, Raj Kumar Bhagat
From: Balamurugan S <quic_bselvara@quicinc.com>
Currently only PCI devices is supported in Ath12k driver. Refactor
Ath12k module_init and module_exit to include Ath12k AHB support.
Add Ath12k AHB support in Kconfig with dependency on Remoteproc
driver. Ath12k AHB support relies on remoteproc driver for firmware
download, power up/down etc.
Tested-on: IPQ5332 hw1.0 AHB WLAN.WBE.1.3.1-00130-QCAHKSWPL_SILICONZ-1
Tested-on: QCN9274 hw2.0 PCI WLAN.WBE.1.1.1-00210-QCAHKSWPL_SILICONZ-1
Signed-off-by: Balamurugan S <quic_bselvara@quicinc.com>
Co-developed-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: P Praneesh <quic_ppranees@quicinc.com>
Signed-off-by: Raj Kumar Bhagat <quic_rajkbhag@quicinc.com>
---
drivers/net/wireless/ath/ath12k/Kconfig | 6 ++++
drivers/net/wireless/ath/ath12k/Makefile | 1 +
drivers/net/wireless/ath/ath12k/ahb.h | 12 ++++++++
drivers/net/wireless/ath/ath12k/core.c | 35 ++++++++++++++++++++++--
drivers/net/wireless/ath/ath12k/pci.c | 12 ++------
drivers/net/wireless/ath/ath12k/pci.h | 2 ++
6 files changed, 56 insertions(+), 12 deletions(-)
diff --git a/drivers/net/wireless/ath/ath12k/Kconfig b/drivers/net/wireless/ath/ath12k/Kconfig
index f64e7c322216..58837a6dec9a 100644
--- a/drivers/net/wireless/ath/ath12k/Kconfig
+++ b/drivers/net/wireless/ath/ath12k/Kconfig
@@ -15,6 +15,12 @@ config ATH12K
If you choose to build a module, it'll be called ath12k.
+config ATH12K_AHB
+ bool "QTI ath12k AHB support"
+ depends on ATH12K && REMOTEPROC && QCOM_Q6V5_WCSS_SEC
+ help
+ Enable support for Ath12k AHB bus chipsets, example IPQ5332.
+
config ATH12K_DEBUG
bool "ath12k debugging"
depends on ATH12K
diff --git a/drivers/net/wireless/ath/ath12k/Makefile b/drivers/net/wireless/ath/ath12k/Makefile
index 5a1ed20d730e..4b7b9404d0c6 100644
--- a/drivers/net/wireless/ath/ath12k/Makefile
+++ b/drivers/net/wireless/ath/ath12k/Makefile
@@ -23,6 +23,7 @@ ath12k-y += core.o \
fw.o \
p2p.o
+ath12k-$(CONFIG_ATH12K_AHB) += ahb.o
ath12k-$(CONFIG_ATH12K_DEBUGFS) += debugfs.o debugfs_htt_stats.o
ath12k-$(CONFIG_ACPI) += acpi.o
ath12k-$(CONFIG_ATH12K_TRACING) += trace.o
diff --git a/drivers/net/wireless/ath/ath12k/ahb.h b/drivers/net/wireless/ath/ath12k/ahb.h
index 0dbbbfd45eab..fc21211c01dd 100644
--- a/drivers/net/wireless/ath/ath12k/ahb.h
+++ b/drivers/net/wireless/ath/ath12k/ahb.h
@@ -65,4 +65,16 @@ static inline struct ath12k_ahb *ath12k_ab_to_ahb(struct ath12k_base *ab)
return (struct ath12k_ahb *)ab->drv_priv;
}
+#ifdef CONFIG_ATH12K_AHB
+int ath12k_ahb_init(void);
+void ath12k_ahb_exit(void);
+#else
+static inline int ath12k_ahb_init(void)
+{
+ return 0;
+}
+
+static inline void ath12k_ahb_exit(void) {};
+#endif
+
#endif
diff --git a/drivers/net/wireless/ath/ath12k/core.c b/drivers/net/wireless/ath/ath12k/core.c
index 9cd485ed42ab..97450726b59d 100644
--- a/drivers/net/wireless/ath/ath12k/core.c
+++ b/drivers/net/wireless/ath/ath12k/core.c
@@ -9,15 +9,18 @@
#include <linux/remoteproc.h>
#include <linux/firmware.h>
#include <linux/of.h>
+#include "ahb.h"
#include "core.h"
#include "dp_tx.h"
#include "dp_rx.h"
#include "debug.h"
-#include "hif.h"
-#include "fw.h"
#include "debugfs.h"
+#include "fw.h"
+#include "hif.h"
+#include "pci.h"
#include "wow.h"
+static int ahb_err, pci_err;
unsigned int ath12k_debug_mask;
module_param_named(debug_mask, ath12k_debug_mask, uint, 0644);
MODULE_PARM_DESC(debug_mask, "Debugging mask");
@@ -1341,5 +1344,31 @@ struct ath12k_base *ath12k_core_alloc(struct device *dev, size_t priv_size,
return NULL;
}
-MODULE_DESCRIPTION("Core module for Qualcomm Atheros 802.11be wireless LAN cards.");
+static int ath12k_init(void)
+{
+ ahb_err = ath12k_ahb_init();
+ if (ahb_err)
+ pr_warn("Failed to initialize ath12k AHB device: %d\n", ahb_err);
+
+ pci_err = ath12k_pci_init();
+ if (pci_err)
+ pr_warn("Failed to initialize ath12k PCI device: %d\n", pci_err);
+
+ /* If both failed, return one of the failures (arbitrary) */
+ return ahb_err && pci_err ? ahb_err : 0;
+}
+
+static void ath12k_exit(void)
+{
+ if (!pci_err)
+ ath12k_pci_exit();
+
+ if (!ahb_err)
+ ath12k_ahb_exit();
+}
+
+module_init(ath12k_init)
+module_exit(ath12k_exit)
+
+MODULE_DESCRIPTION("Driver support for Qualcomm Technologies 802.11be WLAN devices");
MODULE_LICENSE("Dual BSD/GPL");
diff --git a/drivers/net/wireless/ath/ath12k/pci.c b/drivers/net/wireless/ath/ath12k/pci.c
index 9a50d813e9b4..f4a868c2275e 100644
--- a/drivers/net/wireless/ath/ath12k/pci.c
+++ b/drivers/net/wireless/ath/ath12k/pci.c
@@ -1627,7 +1627,7 @@ static struct pci_driver ath12k_pci_driver = {
.driver.pm = &ath12k_pci_pm_ops,
};
-static int ath12k_pci_init(void)
+int ath12k_pci_init(void)
{
int ret;
@@ -1638,16 +1638,10 @@ static int ath12k_pci_init(void)
return ret;
}
- return 0;
+ return ret;
}
-module_init(ath12k_pci_init);
-static void ath12k_pci_exit(void)
+void ath12k_pci_exit(void)
{
pci_unregister_driver(&ath12k_pci_driver);
}
-
-module_exit(ath12k_pci_exit);
-
-MODULE_DESCRIPTION("Driver support for Qualcomm Technologies PCIe 802.11be WLAN devices");
-MODULE_LICENSE("Dual BSD/GPL");
diff --git a/drivers/net/wireless/ath/ath12k/pci.h b/drivers/net/wireless/ath/ath12k/pci.h
index 31584a7ad80e..18648ca11dfc 100644
--- a/drivers/net/wireless/ath/ath12k/pci.h
+++ b/drivers/net/wireless/ath/ath12k/pci.h
@@ -145,4 +145,6 @@ void ath12k_pci_stop(struct ath12k_base *ab);
int ath12k_pci_start(struct ath12k_base *ab);
int ath12k_pci_power_up(struct ath12k_base *ab);
void ath12k_pci_power_down(struct ath12k_base *ab, bool is_suspend);
+int ath12k_pci_init(void);
+void ath12k_pci_exit(void);
#endif /* ATH12K_PCI_H */
--
2.34.1
^ permalink raw reply related [flat|nested] 60+ messages in thread* Re: [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332
2024-10-15 18:26 [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Raj Kumar Bhagat
` (21 preceding siblings ...)
2024-10-15 18:26 ` [PATCH v2 22/22] wifi: ath12k: enable ath12k AHB support Raj Kumar Bhagat
@ 2024-10-16 6:57 ` Krzysztof Kozlowski
2024-12-06 11:07 ` Raj Kumar Bhagat
22 siblings, 1 reply; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-10-16 6:57 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On Tue, Oct 15, 2024 at 11:56:15PM +0530, Raj Kumar Bhagat wrote:
> Currently, Ath12k driver only supports WiFi devices that are based on
> PCI bus. New Ath12k device IPQ5332 is based on AHB bus. Hence, add
> Ath12k AHB support for IPQ5332.
>
> IPQ5332 is IEEE802.11be 2 GHz 2x2 Wifi device. To bring-up IPQ5332
> device:
> - Add hardware parameters for IPQ5332.
> - CE and CMEM register address space in IPQ5332 is separate from WCSS
> register space. Hence, add logic to remap CE and CMEM register
> address.
> - Add support for fixed QMI firmware memory for IPQ5332.
> - Support userPD handling for WCSS secure PIL driver to enable ath12k
> AHB support.
>
> Depends-On: [PATCH V7 0/5] remove unnecessary q6 clocks
> Depends-On: [PATCH V2 0/4] Add new driver for WCSS secure PIL loading
> Link: https://lore.kernel.org/all/20240820055618.267554-1-quic_gokulsri@quicinc.com/
> Link: https://lore.kernel.org/all/20240829134021.1452711-1-quic_gokulsri@quicinc.com/
These are series targetting other subsystems. I do not understand why
you created such dependency. It does not look needed and for sure is not
good: nothing here can be tested, nothing can be applied.
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread* Re: [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332
2024-10-16 6:57 ` [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332 Krzysztof Kozlowski
@ 2024-12-06 11:07 ` Raj Kumar Bhagat
2024-12-06 12:55 ` Krzysztof Kozlowski
0 siblings, 1 reply; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-06 11:07 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 10/16/2024 12:27 PM, Krzysztof Kozlowski wrote:
> On Tue, Oct 15, 2024 at 11:56:15PM +0530, Raj Kumar Bhagat wrote:
>> Currently, Ath12k driver only supports WiFi devices that are based on
>> PCI bus. New Ath12k device IPQ5332 is based on AHB bus. Hence, add
>> Ath12k AHB support for IPQ5332.
>>
>> IPQ5332 is IEEE802.11be 2 GHz 2x2 Wifi device. To bring-up IPQ5332
>> device:
>> - Add hardware parameters for IPQ5332.
>> - CE and CMEM register address space in IPQ5332 is separate from WCSS
>> register space. Hence, add logic to remap CE and CMEM register
>> address.
>> - Add support for fixed QMI firmware memory for IPQ5332.
>> - Support userPD handling for WCSS secure PIL driver to enable ath12k
>> AHB support.
>>
>> Depends-On: [PATCH V7 0/5] remove unnecessary q6 clocks
>> Depends-On: [PATCH V2 0/4] Add new driver for WCSS secure PIL loading
>> Link: https://lore.kernel.org/all/20240820055618.267554-1-quic_gokulsri@quicinc.com/
>> Link: https://lore.kernel.org/all/20240829134021.1452711-1-quic_gokulsri@quicinc.com/
>
> These are series targetting other subsystems. I do not understand why
> you created such dependency. It does not look needed and for sure is not
> good: nothing here can be tested, nothing can be applied.
To validate this series, the dependencies mentioned above were necessary, which
is why they were included.
Currently, the "[PATCH V7 0/5] remove unnecessary q6 clocks" has been merged,
so this dependency will not be required in the next version.
The "[PATCH V2 0/4] Add new driver for WCSS secure PIL loading" series is still
under review and is required for validation.
However, this series can still be applied and compiled without these dependencies.
Please let us know if we should remove the dependency in the next version.
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332
2024-12-06 11:07 ` Raj Kumar Bhagat
@ 2024-12-06 12:55 ` Krzysztof Kozlowski
2024-12-08 13:45 ` Raj Kumar Bhagat
0 siblings, 1 reply; 60+ messages in thread
From: Krzysztof Kozlowski @ 2024-12-06 12:55 UTC (permalink / raw)
To: Raj Kumar Bhagat
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 06/12/2024 12:07, Raj Kumar Bhagat wrote:
> On 10/16/2024 12:27 PM, Krzysztof Kozlowski wrote:
>> On Tue, Oct 15, 2024 at 11:56:15PM +0530, Raj Kumar Bhagat wrote:
>>> Currently, Ath12k driver only supports WiFi devices that are based on
>>> PCI bus. New Ath12k device IPQ5332 is based on AHB bus. Hence, add
>>> Ath12k AHB support for IPQ5332.
>>>
>>> IPQ5332 is IEEE802.11be 2 GHz 2x2 Wifi device. To bring-up IPQ5332
>>> device:
>>> - Add hardware parameters for IPQ5332.
>>> - CE and CMEM register address space in IPQ5332 is separate from WCSS
>>> register space. Hence, add logic to remap CE and CMEM register
>>> address.
>>> - Add support for fixed QMI firmware memory for IPQ5332.
>>> - Support userPD handling for WCSS secure PIL driver to enable ath12k
>>> AHB support.
>>>
>>> Depends-On: [PATCH V7 0/5] remove unnecessary q6 clocks
>>> Depends-On: [PATCH V2 0/4] Add new driver for WCSS secure PIL loading
>>> Link: https://lore.kernel.org/all/20240820055618.267554-1-quic_gokulsri@quicinc.com/
>>> Link: https://lore.kernel.org/all/20240829134021.1452711-1-quic_gokulsri@quicinc.com/
>>
>> These are series targetting other subsystems. I do not understand why
>> you created such dependency. It does not look needed and for sure is not
>> good: nothing here can be tested, nothing can be applied.
>
> To validate this series, the dependencies mentioned above were necessary, which
> is why they were included.
What does it mean "validate"? You are supposed to describe how upstream
can consume this.
>
> Currently, the "[PATCH V7 0/5] remove unnecessary q6 clocks" has been merged,
> so this dependency will not be required in the next version.
>
> The "[PATCH V2 0/4] Add new driver for WCSS secure PIL loading" series is still
> under review and is required for validation.
>
> However, this series can still be applied and compiled without these dependencies.
> Please let us know if we should remove the dependency in the next version.
So write proper cover letter not bringing up fake dependencies.
Otherwise answer is: this cannot be tested, thus it will not be reviewed.
Best regards,
Krzysztof
^ permalink raw reply [flat|nested] 60+ messages in thread
* Re: [PATCH v2 00/22] wifi: ath12k: add Ath12k AHB driver support for IPQ5332
2024-12-06 12:55 ` Krzysztof Kozlowski
@ 2024-12-08 13:45 ` Raj Kumar Bhagat
0 siblings, 0 replies; 60+ messages in thread
From: Raj Kumar Bhagat @ 2024-12-08 13:45 UTC (permalink / raw)
To: Krzysztof Kozlowski
Cc: ath12k, linux-wireless, Kalle Valo, Rob Herring,
Krzysztof Kozlowski, Conor Dooley, Jeff Johnson, Bjorn Andersson,
Konrad Dybcio, devicetree, linux-kernel, linux-arm-msm
On 12/6/2024 6:25 PM, Krzysztof Kozlowski wrote:
> On 06/12/2024 12:07, Raj Kumar Bhagat wrote:
>> On 10/16/2024 12:27 PM, Krzysztof Kozlowski wrote:
>>> On Tue, Oct 15, 2024 at 11:56:15PM +0530, Raj Kumar Bhagat wrote:
>>>> Currently, Ath12k driver only supports WiFi devices that are based on
>>>> PCI bus. New Ath12k device IPQ5332 is based on AHB bus. Hence, add
>>>> Ath12k AHB support for IPQ5332.
>>>>
>>>> IPQ5332 is IEEE802.11be 2 GHz 2x2 Wifi device. To bring-up IPQ5332
>>>> device:
>>>> - Add hardware parameters for IPQ5332.
>>>> - CE and CMEM register address space in IPQ5332 is separate from WCSS
>>>> register space. Hence, add logic to remap CE and CMEM register
>>>> address.
>>>> - Add support for fixed QMI firmware memory for IPQ5332.
>>>> - Support userPD handling for WCSS secure PIL driver to enable ath12k
>>>> AHB support.
>>>>
>>>> Depends-On: [PATCH V7 0/5] remove unnecessary q6 clocks
>>>> Depends-On: [PATCH V2 0/4] Add new driver for WCSS secure PIL loading
>>>> Link: https://lore.kernel.org/all/20240820055618.267554-1-quic_gokulsri@quicinc.com/
>>>> Link: https://lore.kernel.org/all/20240829134021.1452711-1-quic_gokulsri@quicinc.com/
>>>
>>> These are series targetting other subsystems. I do not understand why
>>> you created such dependency. It does not look needed and for sure is not
>>> good: nothing here can be tested, nothing can be applied.
>>
>> To validate this series, the dependencies mentioned above were necessary, which
>> is why they were included.
>
> What does it mean "validate"? You are supposed to describe how upstream
> can consume this.
>
"validate" here means building an image, bring-up DUT in AP, STA or Mesh mode,
associate Station and run bi-directional data traffic.
^ permalink raw reply [flat|nested] 60+ messages in thread