On Thu, Dec 25, 2025 at 10:16:29AM +0000, Alice Guo (OSS) wrote:
> > -----邮件原件-----
> > 发件人: Quentin Schulz <[email protected]>
> > 发送时间: 2025年12月16日 17:50
> > 收件人: Alice Guo (OSS) <[email protected]>; dl-uboot-imx
> > <[email protected]>; [email protected]
> > 抄送: Stefano Babic <[email protected]>; Fabio Estevam
> > <[email protected]>; Tom Rini <[email protected]>; Marek Vasut
> > <[email protected]>; Marek Vasut <[email protected]>; Ye Li
> > <[email protected]>; [email protected]; Peng Fan <[email protected]>;
> > Mattijs Korpershoek <[email protected]>; Lukasz Majewski
> > <[email protected]>; Sherry Sun <[email protected]>; Kever Yang
> > <[email protected]>; Jonas Karlman <[email protected]>; George
> > Chan <[email protected]>; Simon Glass <[email protected]>; David Zang
> > <[email protected]>; Alice Guo <[email protected]>
> > 主题: Re: [PATCH v1 5/7] usb: Kconfig: imply DM_USB_GADGET
> > 
> > Hi Alice,
> > 
> > On 12/16/25 7:38 AM, [email protected] wrote:
> > > From: Sherry Sun <[email protected]>
> > >
> > > Since the ci_udc driver has been converted to DM driver, the
> > > CONFIG_DM_USB_GADGET should also be selected when CONFIG_CI_UDC is
> > > enabled, then we can use DM ci_udc driver.
> > >
> > 
> > Is DM_USB_GADGET *required* for CI_UDC to work? "imply" allows to have
> > DM_USB_GADGET=y AND CI_UDC=n, which is probably not what you want. You
> > probably simply want "depends on DM_USB_GADGET"
> 
> There are 2 implementations supported in drivers/usb/gadget/ci_udc.c: a 
> non-DM path and a DM-based path. The current commit log wording might be 
> ambiguous.

OK. But the key here is that we shouldn't keep code around for the case
of both non-DM and full U-Boot. It's valid to have non-DM and xPL, or DM
and xPL or DM and full U-Boot. We just need to be clear and sure that
we're doing this in both code and commit message.

-- 
Tom

Attachment: signature.asc
Description: PGP signature

Reply via email to