Re: Regression due to 59cf44575456 ("USB: core: Fix oversight in SuperSpeed initialization")

[Takashi Iwai <tiwai@suse.de>]

On Tue, 23 Apr 2024 21:29:27 +0200,
Alan Stern wrote:
> 
> On Mon, Apr 22, 2024 at 09:24:24PM +0200, Takashi Iwai wrote:
> > On Mon, 22 Apr 2024 20:03:46 +0200,
> > Alan Stern wrote:
> > > 
> > > On Mon, Apr 22, 2024 at 07:33:21PM +0200, Linux regression tracking (Thorsten Leemhuis) wrote:
> > > > Hi, Thorsten here, the Linux kernel's regression tracker. Top-posting
> > > > for once, to make this easily accessible to everyone.
> > > > 
> > > > Is anyone still working on fixing below regression? From here it looks
> > > > stalled, but I might have missed something.
> > > 
> > > I've been waiting to hear back from Oliver or Takashi.  A revised patch 
> > > taking my comments into account would be welcome; it should be a very 
> > > small change (just one or two lines of code).
> > 
> > As posted in another mail, it's a virtualized environment.
> > Details are found in the original bug report
> >   https://bugzilla.suse.com/show_bug.cgi?id=1220569
> 
> Hmmm.  If this is a virtualized device, isn't the best solution to fix the 
> emulation code for the device so that it presents a valid descriptor?

Honestly speaking, I don't know, but it smells like a hard way.  After
all, we brought some regression of the previously (even casually /
unintended) working "device"...


> > About the patch change: I appreciate if you cook it rather by
> > yourself since I'm not 100% sure what you suggested.  I can
> > provide the reporter a test kernel with the patch for confirmation, of
> > course.
> 
> Here's a condensed version of the patch you wrote.  But I would prefer not 
> to add this to the kernel if the problem can be fixed somewhere else.

Thanks, I asked the report for testing the patch now.


thanks,

Takashi

> 
> Alan Stern
> 
> 
> 
> Index: usb-devel/drivers/usb/core/hub.c
> ===================================================================
> --- usb-devel.orig/drivers/usb/core/hub.c
> +++ usb-devel/drivers/usb/core/hub.c
> @@ -5110,9 +5110,10 @@ hub_port_init(struct usb_hub *hub, struc
>  	}
>  	if (usb_endpoint_maxp(&udev->ep0.desc) == i) {
>  		;	/* Initial ep0 maxpacket guess is right */
> -	} else if ((udev->speed == USB_SPEED_FULL ||
> +	} else if (((udev->speed == USB_SPEED_FULL ||
>  				udev->speed == USB_SPEED_HIGH) &&
> -			(i == 8 || i == 16 || i == 32 || i == 64)) {
> +			(i == 8 || i == 16 || i == 32 || i == 64)) ||
> +			(udev->speed >= USB_SPEED_SUPER && i > 0)) {
>  		/* Initial guess is wrong; use the descriptor's value */
>  		if (udev->speed == USB_SPEED_FULL)
>  			dev_dbg(&udev->dev, "ep0 maxpacket = %d\n", i);
>