Re: [PATCH v4 3/4] usb: dwc3: dwc3-generic-plat: Add layerscape dwc3 support

[Thinh Nguyen <Thinh.Nguyen@synopsys.com>]

On Fri, Sep 26, 2025, Frank Li wrote:
> On Fri, Sep 26, 2025 at 10:43:03PM +0000, Thinh Nguyen wrote:
> > On Fri, Sep 26, 2025, Frank Li via B4 Relay wrote:
> > > From: Frank Li <Frank.Li@nxp.com>
> > >
> > > Add layerscape dwc3 support by using flatten dwc3 core library. Layerscape
> > > dwc3 need set gsbuscfg0-reqinfo as 0x2222 when dma-coherence set.
> > >
> > > Signed-off-by: Frank Li <Frank.Li@nxp.com>
> > > ---
> > > chagne in v4
> > > - rename gsbuscfg0 to gsbuscfg0_reqinfo
> > > - use flatten core's properties.
> > >
> > > change in v3 (skipped)
> > > ---
> > >  drivers/usb/dwc3/dwc3-generic-plat.c | 15 ++++++++++++++-
> > >  1 file changed, 14 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/usb/dwc3/dwc3-generic-plat.c b/drivers/usb/dwc3/dwc3-generic-plat.c
> > > index af95a527dcc27a7a14d38dcc887f74a888ed91e6..cfcefeffd7ad6c0a21ccc9aaa4483b9acb9dc4e0 100644
> > > --- a/drivers/usb/dwc3/dwc3-generic-plat.c
> > > +++ b/drivers/usb/dwc3/dwc3-generic-plat.c
> > > @@ -8,6 +8,7 @@
> > >   */
> > >
> > >  #include <linux/clk.h>
> > > +#include <linux/of_address.h>
> > >  #include <linux/platform_device.h>
> > >  #include <linux/reset.h>
> > >  #include "glue.h"
> > > @@ -29,6 +30,7 @@ static void dwc3_generic_reset_control_assert(void *data)
> > >
> > >  static int dwc3_generic_probe(struct platform_device *pdev)
> > >  {
> > > +	const struct dwc3_properties *properties;
> > >  	struct dwc3_probe_data probe_data = {};
> > >  	struct device *dev = &pdev->dev;
> > >  	struct dwc3_generic *dwc3g;
> > > @@ -75,7 +77,13 @@ static int dwc3_generic_probe(struct platform_device *pdev)
> > >  	probe_data.dwc = &dwc3g->dwc;
> > >  	probe_data.res = res;
> > >  	probe_data.ignore_clocks_and_resets = true;
> > > -	probe_data.properties = DWC3_DEFAULT_PROPERTIES;
> > > +
> > > +	properties = of_device_get_match_data(dev);
> > > +	if (of_dma_is_coherent(pdev->dev.of_node) && properties)
> >
> > Why do we need the check of_dma_is_coherent() here? Is this specific to
> > your platform to determine whether set gsbuscfg0_reqinfo only?
> 
> I remember someone's review feedback. Anyway, 0x2222 works for both.
> 

Yes. I believe gave a review feedback on that here:
https://lore.kernel.org/linux-usb/20240610202148.yamu4acezsijvpk7@synopsys.com/

This check should be platform specific and not a general check as being
done here. There will be other properties here where that check won't
be applicable.

Will there be a case where the dma coherent is set differently for the
same compatible string?

Regardless, I think we should not have that check here.

Thanks,
Thinh