Enabling DMA Support for frdm_mcxw7x #91869

EmilioCBen · 2025-06-18T19:30:18Z

Enabling DMA Support for frdm_mcxw7x

sonarqubecloud · 2025-06-20T15:12:41Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

teburd · 2025-07-28T15:55:23Z

re-assigned to platform owner, no changes to DMA API in PR, all soc, dts, and driver changes from what I can tell.

Add DMA nodes for MCXW7X SOC DTS. This SOC used TRIGMUX instead of DMAMUX. Signed-off-by: Emilio Benavente <emilio.benavente@nxp.com>

Updating the dma_mcux_edma driver to support transfers that do not use Always-On channels to transfer data yet expected hardware to continue to move a buffer of data. E.G. tests/drivers/dma/loop_transfer tests. Signed-off-by: Emilio Benavente <emilio.benavente@nxp.com>

Enable EDMAv3 for the frdm_mcxw71 and frdm_mcxw72 platforms. Signed-off-by: Emilio Benavente <emilio.benavente@nxp.com>

sonarqubecloud · 2025-07-28T16:31:54Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

decsny · 2025-07-28T19:59:29Z

dts/arm/nxp/nxp_mcxw7x_common.dtsi

+		dmas = <&edma 0 32>, <&edma 1 31>;
+		dma-names = "tx", "rx";


was this tested with uart async api test?

Testing lpuart_async_api fails, debugging the issue at the moment.

decsny · 2025-07-28T20:00:05Z

modules/hal_nxp/mcux/mcux-sdk-ng/drivers/drivers.cmake

@@ -202,6 +201,14 @@ if(CONFIG_SOC_FAMILY_MCXW OR CONFIG_SOC_MCXN947 OR CONFIG_SOC_MCXN547)
  set_variable_ifdef(CONFIG_SOC_FLASH_MCUX CONFIG_MCUX_COMPONENT_driver.flash_k4)
 endif()

+if(CONFIG_SOC_FAMILY_MCXW OR CONFIG_SOC_MCXN547)


why or mcxn547?

Typo, meant MCXN947, will be fixed in the next push.

decsny · 2025-07-28T20:00:47Z

modules/hal_nxp/mcux/mcux-sdk-ng/drivers/drivers.cmake

+if(${MCUX_DEVICE} MATCHES "S32K3")
+  if(CONFIG_DMA)
+    zephyr_include_directories(${MCUX_SDK_NG_DIR}/drivers/dmamux)
+    set(CONFIG_MCUX_COMPONENT_driver.dmamux ON)
+  endif()
+endif()


are you adding for s32 also?

decsny · 2025-07-28T20:02:17Z

drivers/dma/dma_mcux_edma.c

+	if (data->transferConfig.majorLoopCounts < 2) {
+		return -ENOSYS;
+	}


i dont understand this check and error code, maybe you can put a comment

A comment will be added on the next push.

decsny · 2025-07-28T20:07:15Z

drivers/dma/dma_mcux_edma.c

@@ -613,6 +621,9 @@ static int dma_mcux_edma_start(const struct device *dev, uint32_t channel)
 #endif
 	data->busy = true;
 	EDMA_StartTransfer(DEV_EDMA_HANDLE(dev, channel));
+#if defined(CONFIG_DMA_MCUX_EDMA_V3)
+	EDMA_TriggerChannelStart(DEV_BASE(dev), channel);


this API exists on edma v4 and v2 as well, why is it only needed all the sudden for v3 ?

This will be fixed on the next update, this should actually be dependent on if the transfer is MemoryToMemory
AON feature on the channel is not enabled.

decsny · 2025-07-28T20:08:16Z

drivers/dma/dma_mcux_edma.c

+	config->source_burst_length = block_config->block_size;
+	config->dest_burst_length = block_config->block_size;


why are you writing to the config struct the user passed in?

No longer needed, will be updated on the next push.

decsny · 2025-07-28T20:09:32Z

drivers/dma/dma_mcux_edma.c

+#if (defined(FSL_FEATURE_DMAMUX_HAS_A_ON) && FSL_FEATURE_DMAMUX_HAS_A_ON)
 	data->transfer_settings.source_burst_length = config->source_burst_length;
 	data->transfer_settings.dest_burst_length = config->dest_burst_length;
+#else
+	data->transfer_settings.source_burst_length = block_config->block_size;
+	data->transfer_settings.dest_burst_length = block_config->block_size;


commit message needs better explanation for these values or even a comment in the code maybe

decsny · 2025-07-28T20:09:44Z

drivers/dma/dma_mcux_edma.c

+	data->transfer_settings.source_data_size = config->source_data_size;
+	data->transfer_settings.dest_data_size = config->dest_data_size;


ditto commit message needs better explanation for these values or even a comment in the code maybe

EmilioCBen added area: DMA Direct Memory Access platform: NXP NXP labels Jun 18, 2025

github-actions bot added platform: NXP Drivers NXP Semiconductors, drivers area: Clock Control labels Jun 18, 2025

github-actions bot requested review from bperseghetti, butok, dbaluta, decsny, DerekSnell, dleach02, George-Stefan, manuargue, MaochenWang1, mmahadevan108, nordic-krch, PetervdPerk-NXP, Raymond0225 and teburd June 18, 2025 19:31

github-actions bot assigned teburd Jun 18, 2025

EmilioCBen force-pushed the feature/frdm_mcxw7x_dma branch 3 times, most recently from 7043218 to 8fb1415 Compare June 20, 2025 14:47

EmilioCBen force-pushed the feature/frdm_mcxw7x_dma branch from 8fb1415 to 22d8690 Compare July 23, 2025 16:08

zephyrbot requested a review from axelnxp July 23, 2025 16:09

EmilioCBen force-pushed the feature/frdm_mcxw7x_dma branch 2 times, most recently from 90c083d to 2cdd5f9 Compare July 28, 2025 14:43

teburd assigned dleach02 and unassigned teburd Jul 28, 2025

EmilioCBen added 3 commits July 28, 2025 11:26

dts: arm: nxp: nxp_mcxw7x: Add DMA Node

c5ba2b1

Add DMA nodes for MCXW7X SOC DTS. This SOC used TRIGMUX instead of DMAMUX. Signed-off-by: Emilio Benavente <emilio.benavente@nxp.com>

boards: nxp: frdm_mcxw7x: Enabled DMA for frdm_mcxw7x

fbe2d4c

Enable EDMAv3 for the frdm_mcxw71 and frdm_mcxw72 platforms. Signed-off-by: Emilio Benavente <emilio.benavente@nxp.com>

EmilioCBen force-pushed the feature/frdm_mcxw7x_dma branch from 2cdd5f9 to fbe2d4c Compare July 28, 2025 16:28

decsny requested changes Jul 28, 2025

View reviewed changes

		config->source_burst_length = block_config->block_size;
		config->dest_burst_length = block_config->block_size;

		data->transfer_settings.source_data_size = config->source_data_size;
		data->transfer_settings.dest_data_size = config->dest_data_size;

Enabling DMA Support for frdm_mcxw7x #91869

Are you sure you want to change the base?

Enabling DMA Support for frdm_mcxw7x #91869

Conversation

EmilioCBen commented Jun 18, 2025

Uh oh!

sonarqubecloud bot commented Jun 20, 2025

Quality Gate passed

Uh oh!

teburd commented Jul 28, 2025

Uh oh!

sonarqubecloud bot commented Jul 28, 2025

Quality Gate passed

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!