public inbox for u-boot@lists.denx.de
 help / color / mirror / Atom feed
From: Sean Anderson <sean.anderson@linux.dev>
To: Marek Vasut <marek.vasut@mailbox.org>,
	Michal Simek <michal.simek@amd.com>,
	u-boot@lists.denx.de,
	Ilias Apalodimas <ilias.apalodimas@linaro.org>
Subject: Re: [PATCH] usb: dwc3: Initialize Xilinx glue registers
Date: Fri, 16 Jan 2026 11:14:12 -0500	[thread overview]
Message-ID: <ce3b1b60-4c05-4255-9a46-e32751a269a9@linux.dev> (raw)
In-Reply-To: <8dad4173-4b2b-4ba7-8e73-838182570bc1@mailbox.org>

On 1/16/26 09:07, Marek Vasut wrote:
> On 1/16/26 9:19 AM, Michal Simek wrote:
>>
>>
>> On 1/6/26 23:17, Sean Anderson wrote:
>>> These registers may have been initialized by SPL/FSBL, but program them
>>> in case they have not. If they are left at their default values, the
>>> controller will never exit reset.
>>>
>>> Signed-off-by: Sean Anderson <sean.anderson@linux.dev>
>>>
>>> ---
>>>
>>>   drivers/usb/dwc3/dwc3-generic.c | 32 ++++++++++++++++++++++++++------
>>>   drivers/usb/dwc3/dwc3-generic.h |  2 ++
>>>   2 files changed, 28 insertions(+), 6 deletions(-)
>>>
>>> diff --git a/drivers/usb/dwc3/dwc3-generic.c b/drivers/usb/dwc3/dwc3- generic.c
>>> index c09014aec60..ce896db9532 100644
>>> --- a/drivers/usb/dwc3/dwc3-generic.c
>>> +++ b/drivers/usb/dwc3/dwc3-generic.c
>>> @@ -501,6 +501,27 @@ struct dwc3_glue_ops rk_ops = {
>>>       .glue_get_ctrl_dev = dwc3_rk_glue_get_ctrl_dev,
>>>   };
>>> +void dwc3_xlnx_glue_configure(struct udevice *dev, int index,
>>> +                  enum usb_dr_mode mode)
>>> +{
>>> +/* USB glue registers */
>>> +#define USB3_FPD_PIPE_CLK    0x7c
>>> +#define USB3_FPD_POWER_PRSNT    0x80
>>> +    struct dwc3_glue_data *glue = dev_get_plat(dev);
>>> +    void *base = map_physmem(glue->regs, 0x10000, MAP_NOCACHE);
>>> +
>>> +    if (generic_phy_valid(&glue->phy)) {
>>> +        writel(0, base + USB3_FPD_PIPE_CLK);
>>> +        writel(1, base + USB3_FPD_POWER_PRSNT);
>>> +    } else {
>>> +        writel(1, base + USB3_FPD_PIPE_CLK);
>>> +    }
>>> +}
>>> +
>>> +struct dwc3_glue_ops xlnx_ops = {
>>> +    .glue_configure = dwc3_xlnx_glue_configure,
>>> +};
>>> +
>>>   static int dwc3_glue_bind_common(struct udevice *parent, ofnode node)
>>>   {
>>>       const char *name = ofnode_get_name(node);
>>> @@ -621,11 +642,10 @@ int dwc3_glue_probe(struct udevice *dev)
>>>       struct udevice *child = NULL;
>>>       int index = 0;
>>>       int ret;
>>> -    struct phy phy;
>>> -    ret = generic_phy_get_by_name(dev, "usb3-phy", &phy);
>>> +    ret = generic_phy_get_by_name(dev, "usb3-phy", &glue->phy);
>>>       if (!ret) {
>>> -        ret = generic_phy_init(&phy);
>>> +        ret = generic_phy_init(&glue->phy);
>>>           if (ret)
>>>               return ret;
>>>       } else if (ret != -ENOENT && ret != -ENODATA) {
>>> @@ -643,8 +663,8 @@ int dwc3_glue_probe(struct udevice *dev)
>>>       if (ret)
>>>           return ret;
>>> -    if (generic_phy_valid(&phy)) {
>>> -        ret = generic_phy_power_on(&phy);
>>> +    if (generic_phy_valid(&glue->phy)) {
>>> +        ret = generic_phy_power_on(&glue->phy);
>>>           if (ret)
>>>               return ret;
>>>       }
>>> @@ -690,7 +710,7 @@ int dwc3_glue_remove(struct udevice *dev)
>>>   }
>>>   static const struct udevice_id dwc3_glue_ids[] = {
>>> -    { .compatible = "xlnx,zynqmp-dwc3" },
>>> +    { .compatible = "xlnx,zynqmp-dwc3", .data = (ulong)&xlnx_ops },
>>>       { .compatible = "xlnx,versal-dwc3" },
>>>       { .compatible = "ti,keystone-dwc3"},
>>>       { .compatible = "ti,dwc3", .data = (ulong)&ti_ops },
>>> diff --git a/drivers/usb/dwc3/dwc3-generic.h b/drivers/usb/dwc3/dwc3- generic.h
>>> index 40902c8923f..13f6391cae8 100644
>>> --- a/drivers/usb/dwc3/dwc3-generic.h
>>> +++ b/drivers/usb/dwc3/dwc3-generic.h
>>> @@ -10,12 +10,14 @@
>>>   #define __DRIVERS_USB_DWC3_GENERIC_H
>>>   #include <clk.h>
>>> +#include <generic-phy.h>
>>>   #include <reset.h>
>>>   #include <dwc3-uboot.h>
>>>   struct dwc3_glue_data {
>>>       struct clk_bulk        clks;
>>>       struct reset_ctl_bulk    resets;
>>> +    struct phy phy;
>>>       fdt_addr_t regs;
>>>       fdt_size_t size;
>>>   };
>>
>> I will let Marek to handle this. From my perspective we should take Jerome's patchset to update usb stack and then add this on the top.
> Since this fixes a real bug and is isolated to xilinx platform, it would be good if xilinx at least provided feedback on this patch.
> 
> +CC Ilias

It's not a bug fix (hence no fixes tag). See [1,2] for details.

--Sean

[1] https://lore.kernel.org/linux-usb/8f0e04a5-49df-4fe3-9c76-9a1c0bd822a0@linux.dev/
[2] https://lore.kernel.org/linux-usb/be8a566a-163f-4d20-b712-67de6f63a6b5@linux.dev/

  reply	other threads:[~2026-01-16 16:14 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-01-06 22:17 [PATCH] usb: dwc3: Initialize Xilinx glue registers Sean Anderson
2026-01-16  8:19 ` Michal Simek
2026-01-16 14:07   ` Marek Vasut
2026-01-16 16:14     ` Sean Anderson [this message]
2026-01-22 13:25       ` Michal Simek
2026-02-17  8:48         ` Kumar, Love
2026-01-29 10:13       ` Michal Simek
2026-02-17 11:06         ` Michal Simek

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=ce3b1b60-4c05-4255-9a46-e32751a269a9@linux.dev \
    --to=sean.anderson@linux.dev \
    --cc=ilias.apalodimas@linaro.org \
    --cc=marek.vasut@mailbox.org \
    --cc=michal.simek@amd.com \
    --cc=u-boot@lists.denx.de \
    /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