Add support eDMA for S32K344 #61311

Dat-NguyenDuy · 2023-08-09T02:33:31Z

This extends NXP current eDMA shim driver for supporting eDMA v3 on S32K344

zephyrbot · 2023-08-09T04:22:48Z

The following west manifest projects have been modified in this Pull Request:

Name	Old Revision	New Revision	Diff
hal_nxp	zephyrproject-rtos/hal_nxp@`8cc344e`	zephyrproject-rtos/hal_nxp@`6d91c17` (master)	zephyrproject-rtos/[email protected]

Note: This message is automatically posted and updated by the Manifest GitHub Action.

mmahadevan108 · 2023-08-15T16:07:45Z

@Dat-NguyenDuy, please take a look at the CI failures.

boards/arm/mr_canhubk3/doc/index.rst

danieldegrasse · 2023-08-16T01:57:27Z

drivers/dma/dma_mcux_edma.c

+#define DEV_DMAMUX_BASE(dev, idx) ((DMAMUX_Type *)DEV_CFG(dev)->dmamux_base[idx])
+#define DEV_DMAMUX_IDX(dev, ch)	(ch / DEV_CFG(dev)->channels_per_mux)
+
+#ifdef CONFIG_SOC_SERIES_S32K3_M7


Is there a reason this code needs to be SOC dependent?

This is SoC dependency, i'm considering it's worth to add a dt properties for it

I'd suggest to add a binding property to remove the SoC dependency in the driver, e.g.

#define DEV_DMAMUX_CHANNEL(dev, ch) ((ch % DEV_CFG(dev)->channels_per_mux) ^ (DEV_CFG(dev)->mux_channel_mask))

and use a default of 0 which would be the logical way to access the DMA CHCFGn registers. The SoCs that have a different arrangement can change the mux-channel-mask (feel free to suggest a better name) as needed, like in the case of K344 to 3.

I've updated

danieldegrasse · 2023-08-16T02:03:34Z

drivers/dma/dma_mcux_edma.c

-		IRQ_CONFIG(n, 16, dma_mcux_edma_error_irq_handler);   \
-								      \
-		LOG_DBG("install irq done");			      \
+#ifdef CONFIG_DMA_MCUX_SINGLE_IRQ_PER_CHANNEL


It seems like this macro will install an interrupt handler per each channel of the EDMA. Existing NXP SOCs using the legacy EDMA also have a per channel interrupt, but have a combined error interrupt. Would it be possible to enable per channel interrupts for all EDMA instances, and only enable the error interrupt servicing in the channel interrupt when this Kconfig was set?

Good idea, it's possible, i just re-pushed the update

@danieldegrasse could you pls revisit if this one looks fine to you

drivers/serial/uart_mcux_lpuart.c

hakehuang · 2023-09-26T09:11:38Z

I am trigger CI for all NXP related EDMA boards, will feadback once ready

regression test done. no issues found.

Dat-NguyenDuy · 2023-09-26T09:36:42Z

Minor update #ifdef DMA_MCUX_HAS_CHANNEL_GAP --> #if DMA_MCUX_HAS_CHANNEL_GAP

hakehuang

Test PASS on NXP regression for RT series

mmahadevan108 · 2023-09-26T15:09:03Z

@Dat-NguyenDuy , thank you for the PR. Overall this looks great. My request would be not leave the current EDMA driver Kconfigs and device tree binding names unchanged to maintain backwards compatibility.

mmahadevan108

Request to leave the EDMA Kconfigs and device tree binding names unchanged.

manuargue · 2023-09-26T15:53:25Z

Request to leave the EDMA Kconfigs and device tree binding names unchanged.

@mmahadevan108 I'm a bit puzzled, I had the impression we agreed naming existing binding v2 and the new one v3 to follow the corresponding DMA versions (#61311 (comment)). We should add this to release notes to make sure the changes are visible to users.

Add new dt binding for edma v3 that inherits whole dt properties from current version. One more property is added for SoCs that don't have separate error interrupt id, use same id with channel interrupt Signed-off-by: Dat Nguyen Duy <[email protected]>

On S32K344, the offset in memory map between each channel is 0x4000 for most channels, but there is specific case is between channel 11 and 12 which is 0x1D4000 instead. As a consequence, 32 channels are divided to two parts: one starts from channel 0 -> 11. The other is from channel 128 to 145. The channel gap is from 12 -> 127. For user and data structures in shim driver, the channel's value comes from 0 --> 31. Above constraint will be counted when interact with the mcux sdk Beside that, the DMAMUX register in this platform is very specific, not in identical with DMAMUX channel, so shim driver is updated to cover this case Signed-off-by: Dat Nguyen Duy <[email protected]>

Dat-NguyenDuy · 2023-09-27T03:41:45Z

Request to leave the EDMA Kconfigs and device tree binding names unchanged.

I re-pushed to leave current EDMA Kconfig and dts binding

mmahadevan108 · 2023-09-27T13:55:11Z

Please update west.yml

Changes made

Enable dma support for mr_canhubk3 board Signed-off-by: Dat Nguyen Duy <[email protected]>

Add devicetree overlay for enabling dma test for mr_canhubk3 board Signed-off-by: Dat Nguyen Duy <[email protected]>

Dat-NguyenDuy · 2023-09-27T14:05:35Z

Please update west.yml
I've updated

mmahadevan108 · 2023-09-27T14:14:21Z

Request to leave the EDMA Kconfigs and device tree binding names unchanged.

@mmahadevan108 I'm a bit puzzled, I had the impression we agreed naming existing binding v2 and the new one v3 to follow the corresponding DMA versions (#61311 (comment)). We should add this to release notes to make sure the changes are visible to users.

@manuargue , the general guidance from the Zephyr project is to limit breaking backward compatibility as much as possible. In this case, we are sticking with the SDMA naming convention for the driver folders.

hakehuang · 2023-09-27T14:55:30Z

regression test is done. no issue found in rt platforms with this change

Dat-NguyenDuy requested review from nashif, mmahadevan108, dleach02, manuargue, tbursztyka, MaureenHelm and galak as code owners August 9, 2023 02:33

zephyrbot added area: UART Universal Asynchronous Receiver-Transmitter platform: NXP NXP area: SPI SPI bus area: DMA Direct Memory Access area: Devicetree Binding PR modifies or adds a Device Tree binding labels Aug 9, 2023

zephyrbot requested review from danieldegrasse, dbaluta, decsny, DerekSnell, EmilioCBen, iuliana-prodan, teburd and yvanderv August 9, 2023 02:34

zephyrbot assigned dleach02 Aug 9, 2023

zephyrbot added manifest manifest-hal_nxp labels Aug 9, 2023

zephyrbot added the DNM This PR should not be merged (Do Not Merge) label Aug 9, 2023

mmahadevan108 requested review from hakehuang and removed request for decsny August 15, 2023 16:06

danieldegrasse reviewed Aug 16, 2023

View reviewed changes

hakehuang reviewed Aug 16, 2023

View reviewed changes

drivers/serial/uart_mcux_lpuart.c Outdated Show resolved Hide resolved

manuargue previously approved these changes Sep 26, 2023

View reviewed changes

Dat-NguyenDuy dismissed manuargue’s stale review via 59d9f69 September 26, 2023 09:35

hakehuang previously approved these changes Sep 26, 2023

View reviewed changes

manuargue previously approved these changes Sep 26, 2023

View reviewed changes

mmahadevan108 previously requested changes Sep 26, 2023

View reviewed changes

Dat-NguyenDuy added 2 commits September 27, 2023 09:12

Dat-NguyenDuy dismissed stale reviews from manuargue and hakehuang via dc50609 September 27, 2023 03:37

Dat-NguyenDuy added 2 commits September 27, 2023 21:02

boards: arm: mr_canhubk3: enable dma support

dadd440

Enable dma support for mr_canhubk3 board Signed-off-by: Dat Nguyen Duy <[email protected]>

tests: drivers: dma: enable dma test for mr_canhubk3

69dd303

Add devicetree overlay for enabling dma test for mr_canhubk3 board Signed-off-by: Dat Nguyen Duy <[email protected]>

zephyrbot removed the DNM This PR should not be merged (Do Not Merge) label Sep 27, 2023

manuargue approved these changes Sep 27, 2023

View reviewed changes

mmahadevan108 approved these changes Sep 27, 2023

View reviewed changes

hakehuang self-requested a review September 27, 2023 14:55

hakehuang approved these changes Sep 27, 2023

View reviewed changes

dleach02 merged commit 11f5eb5 into zephyrproject-rtos:main Sep 27, 2023

manuargue deleted the s32k3-dma branch September 27, 2023 23:33

danieldegrasse mentioned this pull request Nov 27, 2023

drivers: dma: Introduce driver for NXP's eDMA IP #65671

Merged

ubieda mentioned this pull request Aug 15, 2024

NXP's IMX RT1010 triggers assert on SPI Loopback test when EDMA enabled #77129

Closed

Add support eDMA for S32K344 #61311

Add support eDMA for S32K344 #61311

Uh oh!

Conversation

Dat-NguyenDuy commented Aug 9, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zephyrbot commented Aug 9, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mmahadevan108 commented Aug 15, 2023

Uh oh!

Uh oh!

danieldegrasse Aug 16, 2023

Choose a reason for hiding this comment

Uh oh!

Dat-NguyenDuy Aug 25, 2023

Choose a reason for hiding this comment

Uh oh!

manuargue Sep 22, 2023

Choose a reason for hiding this comment

Uh oh!

Dat-NguyenDuy Sep 22, 2023

Choose a reason for hiding this comment

Uh oh!

danieldegrasse Aug 16, 2023

Choose a reason for hiding this comment

Uh oh!

Dat-NguyenDuy Aug 25, 2023

Choose a reason for hiding this comment

Uh oh!

Dat-NguyenDuy Sep 13, 2023

Choose a reason for hiding this comment

Uh oh!

Uh oh!

hakehuang commented Sep 26, 2023

Uh oh!

Dat-NguyenDuy commented Sep 26, 2023 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

hakehuang left a comment

Choose a reason for hiding this comment

Uh oh!

mmahadevan108 commented Sep 26, 2023

Uh oh!

mmahadevan108 left a comment

Choose a reason for hiding this comment

Uh oh!

manuargue commented Sep 26, 2023

Uh oh!

Dat-NguyenDuy commented Sep 27, 2023

Uh oh!

mmahadevan108 commented Sep 27, 2023

Uh oh!

Dat-NguyenDuy commented Sep 27, 2023

Uh oh!

mmahadevan108 commented Sep 27, 2023

Uh oh!

hakehuang commented Sep 27, 2023

Uh oh!

Uh oh!

Dat-NguyenDuy commented Aug 9, 2023 •

edited

Loading

zephyrbot commented Aug 9, 2023 •

edited

Loading

Dat-NguyenDuy commented Sep 26, 2023 •

edited

Loading