drivers: udc_mcux_ehci: fix typo in transfer handler

[jfischer-no]

If the to-host data stage length is less than that requested by the
host, but equal to or a multiple of MPS, the device should send a ZLP,
not receive it.

Fixes: #84999
Tested on mimxrt1020_evk

udc_mcux_ip3511.c has similar type, but according to the reporter it would not fix similar issue there, so there should be something else.

[henrikbrixandersen]

If the to-host data stage length is less than that requested by the host, but equal to or a multiple of MPS, the device should send a ZLP, not receive it.

Fixes: #84999 Tested on mimxrt1020_evk

Confirmed, tested with mimxrt1014_evk.

udc_mcux_ip3511.c has similar type, so it should also Fixes: #83306, not tested.

My lpcxpresso55s16 board still doesn't enumerate on MS Windows even with this fix.

[jfischer-no]

udc_mcux_ip3511.c has similar type, so it should also Fixes: #83306, not tested.

My lpcxpresso55s16 board still doesn't enumerate on MS Windows even with this fix.

From your description it looks like similar issue, can you please add LOG_ERRs like

diff --git a/drivers/usb/udc/udc_mcux_ip3511.c b/drivers/usb/udc/udc_mcux_ip3511.c
index 9b1665168c5..8800e7a8976 100644
--- a/drivers/usb/udc/udc_mcux_ip3511.c
+++ b/drivers/usb/udc/udc_mcux_ip3511.c
@@ -354,10 +354,12 @@ static bool udc_mcux_handler_zlt(const struct device *dev, uint8_t ep, struct ne
        struct udc_mcux_data *priv = udc_get_private(dev);
 
        /* The whole transfer is already done by MCUX controller driver. */
+       LOG_ERR("ep %x mcux_len %u buf->len %u", ep, mcux_len, buf->len);
        if (mcux_len >= buf->len) {
                if (udc_ep_buf_has_zlp(buf)) {
                        usb_status_t status;
 
+                       LOG_ERR("zlp");
                        udc_ep_buf_clear_zlp(buf);
                        status = mcux_if->deviceSend(priv->mcux_device.controllerHandle,
                                        ep, NULL, 0);

and post the logs here?

[tmon-nordic]

Don't include two pretty much independent commits in a single PR. :samples: hid-keyboard: initialize report buffer at startup" is pretty much completely unrelated to "drivers: udc_mcux_ehci: fix typo in transfer handler".

[jfischer-no]

Don't include two pretty much independent commits in a single PR. :samples: hid-keyboard: initialize report buffer at startup" is pretty much completely unrelated to "drivers: udc_mcux_ehci: fix typo in transfer handler".

It is related to the MCUX UDC drivers. But it needs more investigation.

[henrikbrixandersen]

udc_mcux_ip3511.c has similar type, so it should also Fixes: #83306, not tested.

My lpcxpresso55s16 board still doesn't enumerate on MS Windows even with this fix.

From your description it looks like similar issue, can you please add LOG_ERRs like

diff --git a/drivers/usb/udc/udc_mcux_ip3511.c b/drivers/usb/udc/udc_mcux_ip3511.c
index 9b1665168c5..8800e7a8976 100644
--- a/drivers/usb/udc/udc_mcux_ip3511.c
+++ b/drivers/usb/udc/udc_mcux_ip3511.c
@@ -354,10 +354,12 @@ static bool udc_mcux_handler_zlt(const struct device *dev, uint8_t ep, struct ne
        struct udc_mcux_data *priv = udc_get_private(dev);
 
        /* The whole transfer is already done by MCUX controller driver. */
+       LOG_ERR("ep %x mcux_len %u buf->len %u", ep, mcux_len, buf->len);
        if (mcux_len >= buf->len) {
                if (udc_ep_buf_has_zlp(buf)) {
                        usb_status_t status;
 
+                       LOG_ERR("zlp");
                        udc_ep_buf_clear_zlp(buf);
                        status = mcux_if->deviceSend(priv->mcux_device.controllerHandle,
                                        ep, NULL, 0);

and post the logs here?

*** Booting Zephyr OS build v4.0.0-4411-g839bcdca6588 ***
[00:00:00.104,000] <inf> main: USBD message: Bus reset
[00:00:00.183,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:00.184,000] <err> udc_mcux: ep 80 mcux_len 0 buf->len 0
[00:00:00.194,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:00.203,000] <err> udc_mcux: ep 80 mcux_len 32 buf->len 32
[00:00:00.204,000] <err> udc_mcux: ep 80 mcux_len 64 buf->len 64
[00:00:00.204,000] <err> udc_mcux: zlp
[00:00:05.719,000] <err> udc_mcux: ep 80 mcux_len 65535 buf->len 64
[00:00:05.720,000] <inf> main: USBD message: Bus reset
[00:00:05.864,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:05.864,000] <err> usbd_core: UDC error event
[00:00:05.864,000] <inf> main: USBD message: Controller error
[00:00:10.874,000] <err> udc_mcux: ep 80 mcux_len 0 buf->len 0
[00:00:10.884,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:10.884,000] <err> usbd_core: UDC error event
[00:00:10.884,000] <inf> main: USBD message: Controller error
[00:00:15.912,000] <err> udc_mcux: ep 80 mcux_len 32 buf->len 32
[00:00:15.912,000] <err> usbd_core: UDC error event
[00:00:15.912,000] <inf> main: USBD message: Controller error
[00:00:21.418,000] <inf> main: USBD message: Bus reset
[00:00:21.563,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:21.563,000] <err> usbd_core: UDC error event
[00:00:21.563,000] <inf> main: USBD message: Controller error
[00:00:26.572,000] <err> udc_mcux: ep 80 mcux_len 0 buf->len 0
[00:00:26.582,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:26.583,000] <err> usbd_core: UDC error event
[00:00:26.583,000] <inf> main: USBD message: Controller error
[00:00:31.598,000] <err> udc_mcux: ep 80 mcux_len 32 buf->len 32
[00:00:31.598,000] <err> usbd_core: UDC error event
[00:00:31.599,000] <inf> main: USBD message: Controller error
[00:00:37.107,000] <inf> main: USBD message: Bus reset
[00:00:37.251,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:37.251,000] <err> usbd_core: UDC error event
[00:00:37.252,000] <inf> main: USBD message: Controller error
[00:00:42.254,000] <err> udc_mcux: ep 80 mcux_len 0 buf->len 0
[00:00:42.264,000] <err> udc_mcux: ep 80 mcux_len 18 buf->len 18
[00:00:42.264,000] <err> usbd_core: UDC error event
[00:00:42.265,000] <inf> main: USBD message: Controller error
[00:00:47.286,000] <err> udc_mcux: ep 80 mcux_len 32 buf->len 32
[00:00:47.286,000] <err> usbd_core: UDC error event
[00:00:47.286,000] <inf> main: USBD message: Controller error
...

[henrikbrixandersen]

Blocking to prevent merge until the fix is working.

[jfischer-no]

@MarkWangChinese udc_mcux_ip3511.c has similar typo, but according to the reporter it would not fix it, I do not have hardware to investigate it further, can you please look at the #83306? Thank you.

[jfischer-no]

Blocking to prevent merge until the fix is working.

@henrikbrixandersen I reduced the commit to fix #84999 only. This is tested and should not be blocked.

[MarkWangChinese]

@MarkWangChinese udc_mcux_ip3511.c has similar typo, but according to the reporter it would not fix it, I do not have hardware to investigate it further, can you please look at the #83306? Thank you.

OK, I will look at it later.