U-Boot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Mattijs Korpershoek <mkorpershoek@kernel.org>
To: Sam Day <me@samcday.com>, Lukasz Majewski <lukma@denx.de>,
	Mattijs Korpershoek <mkorpershoek@kernel.org>,
	Marek Vasut <marex@denx.de>, Tom Rini <trini@konsulko.com>,
	Neil Armstrong <neil.armstrong@linaro.org>,
	Sumit Garg <sumit.garg@kernel.org>,
	Stephan Gerhold <stephan.gerhold@linaro.org>,
	Casey Connolly <casey.connolly@linaro.org>
Cc: u-boot@lists.denx.de, u-boot-qcom@groups.io,
	"Quentin Schulz" <quentin.schulz@cherry.de>,
	"Kever Yang" <kever.yang@rock-chips.com>,
	"Peter Robinson" <pbrobinson@gmail.com>,
	"Jonas Karlman" <jonas@kwiboo.se>,
	"George Chan" <gchan9527@gmail.com>,
	"Balaji Selvanathan" <balaji.selvanathan@oss.qualcomm.com>,
	"Petr Beneš" <petr.benes@ysoft.com>,
	"Heiko Schocher" <hs@nabladev.com>,
	"Simon Glass" <sjg@chromium.org>,
	"Aswin Murugan" <aswin.murugan@oss.qualcomm.com>,
	"Paul Sajna" <sajattack@postmarketos.org>,
	"Marek Vasut" <marek.vasut+renesas@mailbox.org>,
	"Varadarajan Narayanan" <varadarajan.narayanan@oss.qualcomm.com>,
	"Svyatoslav Ryhel" <clamor95@gmail.com>,
	"Ilias Apalodimas" <ilias.apalodimas@linaro.org>,
	"Luca Weiss" <luca.weiss@fairphone.com>,
	"Sam Day" <me@samcday.com>
Subject: Re: [PATCH v2 1/3] usb: udc: ci: support USB gadget driver model
Date: Mon, 29 Jun 2026 14:20:40 +0200	[thread overview]
Message-ID: <87echpczc7.fsf@kernel.org> (raw)
In-Reply-To: <20260527-msm8916-usb-v2-1-343e1e32e77b@samcday.com>

Hi Sam,

Thank you for the patch.
Sorry for the delays in the review for this one.

On Wed, May 27, 2026 at 00:37, Sam Day <me@samcday.com> wrote:

> When CONFIG_DM_USB_GADGET is enabled, a UCLASS_USB_GADGET_GENERIC driver
> will be defined that wraps the ChipIdea UDC operations. The
> (dm_)?usb_gadget_.* symbols will no longer be defined (as these are now
> handled by the UDC uclass).
>
> If CONFIG_DM_USB_GADGET is not enabled, this driver behaves as it
> previously did.
>
> This new driver does not declare any compatibles of its own. It requires
> a glue driver to bind it. The ehci_msm driver will be updated in the
> following commit to do exactly that.
>
> Signed-off-by: Sam Day <me@samcday.com>
> ---
>  drivers/usb/gadget/Kconfig  |   1 -
>  drivers/usb/gadget/ci_udc.c | 118 ++++++++++++++++++++++++++++++++++++--------
>  2 files changed, 97 insertions(+), 22 deletions(-)
>
> diff --git a/drivers/usb/gadget/Kconfig b/drivers/usb/gadget/Kconfig
> index 5390878254a..6ba558cc292 100644
> --- a/drivers/usb/gadget/Kconfig
> +++ b/drivers/usb/gadget/Kconfig
> @@ -153,7 +153,6 @@ config USB_GADGET_OS_DESCRIPTORS
>  
>  config CI_UDC
>  	bool "ChipIdea device controller"
> -	depends on !DM_USB_GADGET
>  	select USB_GADGET_DUALSPEED
>  	help
>  	  Say Y here to enable device controller functionality of the
> diff --git a/drivers/usb/gadget/ci_udc.c b/drivers/usb/gadget/ci_udc.c
> index 4729570c525..42028e11d59 100644
> --- a/drivers/usb/gadget/ci_udc.c
> +++ b/drivers/usb/gadget/ci_udc.c
> @@ -10,6 +10,7 @@
>  #include <command.h>
>  #include <config.h>
>  #include <cpu_func.h>
> +#include <dm.h>
>  #include <net.h>
>  #include <malloc.h>
>  #include <wait_bit.h>
> @@ -94,8 +95,17 @@ static struct usb_request *
>  ci_ep_alloc_request(struct usb_ep *ep, unsigned int gfp_flags);
>  static void ci_ep_free_request(struct usb_ep *ep, struct usb_request *_req);
>  
> +#if CONFIG_IS_ENABLED(DM_USB_GADGET)
> +static int ci_udc_gadget_start(struct usb_gadget *g, struct usb_gadget_driver *driver);
> +static int ci_udc_gadget_stop(struct usb_gadget *g);
> +#endif
> +
>  static const struct usb_gadget_ops ci_udc_ops = {
>  	.pullup = ci_pullup,
> +#if CONFIG_IS_ENABLED(DM_USB_GADGET)
> +	.udc_start = ci_udc_gadget_start,
> +	.udc_stop = ci_udc_gadget_stop,
> +#endif
>  };
>  
>  static const struct usb_ep_ops ci_ep_ops = {
> @@ -978,7 +988,7 @@ void udc_irq(void)
>  	}
>  }
>  
> -int dm_usb_gadget_handle_interrupts(struct udevice *dev)
> +static int ci_udc_handle_interrupts(struct udevice *dev)
>  {
>  	struct ci_udc *udc = (struct ci_udc *)controller.ctrl->hcor;
>  	u32 value;
> @@ -1037,7 +1047,7 @@ static int ci_pullup(struct usb_gadget *gadget, int is_on)
>  static int ci_udc_probe(void)
>  {
>  	struct ept_queue_head *head;
> -	int i;
> +	int i, ret;
>  
>  	const int num = 2 * NUM_ENDPOINTS;
>  
> @@ -1046,6 +1056,14 @@ static int ci_udc_probe(void)
>  	const int eplist_raw_sz = num * sizeof(struct ept_queue_head);
>  	const int eplist_sz = roundup(eplist_raw_sz, ARCH_DMA_MINALIGN);
>  
> +	if (IS_ENABLED(CONFIG_DM_USB))
> +		ret = usb_setup_ehci_gadget(&controller.ctrl);
> +	else
> +		ret = usb_lowlevel_init(0, USB_INIT_DEVICE, (void **)&controller.ctrl);
> +
> +	if (ret)
> +		return ret;
> +

What happens if later down in ci_udc_probe() we fail to finish the
function sucessfully?
For example, memalign(ILIST_ALIGN, ILIST_SZ) can return -ENOMEM.

Doesn't that leaves us with an usb that's halfway configured?
Shouldn't we add some error handling / do some cleanup in that case?

>  	/* The QH list must be aligned to 4096 bytes. */
>  	controller.epts = memalign(eplist_align, eplist_sz);
>  	if (!controller.epts)
> @@ -1123,6 +1141,73 @@ static int ci_udc_probe(void)
>  	return 0;
>  }
>  
> +static void ci_udc_remove(void)
> +{
> +	if (IS_ENABLED(CONFIG_DM_USB)) {
> +		usb_remove_ehci_gadget(&controller.ctrl);
> +	} else {
> +		usb_lowlevel_stop(0);
> +		controller.ctrl = NULL;
> +	}
> +}
> +
> +static void ci_udc_stop(void)
> +{
> +	ci_ep_free_request(&controller.ep[0].ep, &controller.ep0_req->req);
> +	free(controller.items_mem);
> +	free(controller.epts);

It feels a bit weird to free the endpoint request in stop().
Shouldn't we do this in the remove() function to match the
ci_ep_alloc_request() which is done in ci_udc_probe()?

> +}
> +
> +static int ci_udc_gadget_start(struct usb_gadget *g, struct usb_gadget_driver *driver)
> +{
> +	controller.driver = driver;
> +	return 0;
> +}

Nitpick: could we please move this next to ci_udc_gadget_stop() ?
I know this would force you to write add controller.driver = driver in
usb_gadget_register_driver() but I don't this it's a problem given that
we do something similar in usb_gadget_unregister_driver().

> +
> +#if CONFIG_IS_ENABLED(DM_USB_GADGET)
> +static int ci_udc_generic_probe(struct udevice *dev)
> +{
> +	int ret = ci_udc_probe();
> +
> +	if (ret)
> +		return ret;
> +
> +	return usb_add_gadget_udc((struct device *)dev, &controller.gadget);
> +}
> +
> +static int ci_udc_generic_remove(struct udevice *dev)
> +{
> +	usb_del_gadget_udc(&controller.gadget);
> +
> +	ci_udc_remove();
> +
> +	return 0;
> +}
> +
> +static const struct usb_gadget_generic_ops ci_udc_generic_ops = {
> +	.handle_interrupts	= ci_udc_handle_interrupts,
> +};
> +
> +U_BOOT_DRIVER(ci_udc_generic) = {
> +	.name = "ci-udc",
> +	.id = UCLASS_USB_GADGET_GENERIC,
> +	.ops = &ci_udc_generic_ops,
> +	.probe = ci_udc_generic_probe,
> +	.remove = ci_udc_generic_remove,
> +};
> +
> +static int ci_udc_gadget_stop(struct usb_gadget *g)
> +{
> +	/* Avoid calling disconnect() twice; the UDC uclass already did it. */

There is no disconnect in udc-uclass.c? Did you mean "UDC core" here
instead of "UDC uclass" ?

> +	controller.driver = NULL;
> +	udc_disconnect();
> +
> +	ci_udc_stop();
> +
> +	return 0;
> +}
> +
> +#else
>  int usb_gadget_register_driver(struct usb_gadget_driver *driver)
>  {
>  	int ret;
> @@ -1132,14 +1217,6 @@ int usb_gadget_register_driver(struct usb_gadget_driver *driver)
>  	if (!driver->bind || !driver->setup || !driver->disconnect)
>  		return -EINVAL;
>  
> -#if CONFIG_IS_ENABLED(DM_USB)
> -	ret = usb_setup_ehci_gadget(&controller.ctrl);
> -#else
> -	ret = usb_lowlevel_init(0, USB_INIT_DEVICE, (void **)&controller.ctrl);
> -#endif
> -	if (ret)
> -		return ret;
> -
>  	ret = ci_udc_probe();
>  	if (ret) {
>  		DBG("udc probe failed, returned %d\n", ret);
> @@ -1151,7 +1228,8 @@ int usb_gadget_register_driver(struct usb_gadget_driver *driver)
>  		DBG("driver->bind() returned %d\n", ret);
>  		return ret;
>  	}
> -	controller.driver = driver;
> +
> +	ci_udc_gadget_start(&controller.gadget, driver);
>  
>  	return 0;
>  }
> @@ -1163,20 +1241,18 @@ int usb_gadget_unregister_driver(struct usb_gadget_driver *driver)
>  	driver->unbind(&controller.gadget);
>  	controller.driver = NULL;
>  
> -	ci_ep_free_request(&controller.ep[0].ep, &controller.ep0_req->req);
> -	free(controller.items_mem);
> -	free(controller.epts);
> -
> -#if CONFIG_IS_ENABLED(DM_USB)
> -	usb_remove_ehci_gadget(&controller.ctrl);
> -#else
> -	usb_lowlevel_stop(0);
> -	controller.ctrl = NULL;
> -#endif
> +	ci_udc_stop();
> +	ci_udc_remove();
>  
>  	return 0;
>  }
>  
> +int dm_usb_gadget_handle_interrupts(struct udevice *dev)
> +{
> +	return ci_udc_handle_interrupts(dev);
> +}
> +#endif
> +
>  bool dfu_usb_get_reset(void)
>  {
>  	struct ci_udc *udc = (struct ci_udc *)controller.ctrl->hcor;
>
> -- 
> 2.54.0

  reply	other threads:[~2026-06-29 12:20 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-27  0:37 [PATCH v2 0/3] Support MSM8916 USB PHY + UDC in qcom_defconfig Sam Day
2026-05-27  0:37 ` [PATCH v2 1/3] usb: udc: ci: support USB gadget driver model Sam Day
2026-06-29 12:20   ` Mattijs Korpershoek [this message]
2026-05-27  0:37 ` [PATCH v2 2/3] usb: host: ehci-msm: Register ChipIdea UDC from glue wrapper Sam Day
2026-06-29 12:33   ` Mattijs Korpershoek
2026-05-27  0:38 ` [PATCH v2 3/3] qcom_defconfig: Enable MSM8916 USB PHY + ChipIdea UDC Sam Day
2026-06-29 12:34   ` Mattijs Korpershoek
2026-06-30  1:45   ` Balaji Selvanathan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87echpczc7.fsf@kernel.org \
    --to=mkorpershoek@kernel.org \
    --cc=aswin.murugan@oss.qualcomm.com \
    --cc=balaji.selvanathan@oss.qualcomm.com \
    --cc=casey.connolly@linaro.org \
    --cc=clamor95@gmail.com \
    --cc=gchan9527@gmail.com \
    --cc=hs@nabladev.com \
    --cc=ilias.apalodimas@linaro.org \
    --cc=jonas@kwiboo.se \
    --cc=kever.yang@rock-chips.com \
    --cc=luca.weiss@fairphone.com \
    --cc=lukma@denx.de \
    --cc=marek.vasut+renesas@mailbox.org \
    --cc=marex@denx.de \
    --cc=me@samcday.com \
    --cc=neil.armstrong@linaro.org \
    --cc=pbrobinson@gmail.com \
    --cc=petr.benes@ysoft.com \
    --cc=quentin.schulz@cherry.de \
    --cc=sajattack@postmarketos.org \
    --cc=sjg@chromium.org \
    --cc=stephan.gerhold@linaro.org \
    --cc=sumit.garg@kernel.org \
    --cc=trini@konsulko.com \
    --cc=u-boot-qcom@groups.io \
    --cc=u-boot@lists.denx.de \
    --cc=varadarajan.narayanan@oss.qualcomm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox