usb: ohci-at91: Fix the unhandle interrupt when resume
authorGuiting Shen <aarongt.shen@gmail.com>
Mon, 26 Jun 2023 15:27:13 +0000 (23:27 +0800)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Thu, 3 Aug 2023 08:24:10 +0000 (10:24 +0200)
commit c55afcbeaa7a6f4fffdbc999a9bf3f0b29a5186f upstream.

The ohci_hcd_at91_drv_suspend() sets ohci->rh_state to OHCI_RH_HALTED when
suspend which will let the ohci_irq() skip the interrupt after resume. And
nobody to handle this interrupt.

According to the comment in ohci_hcd_at91_drv_suspend(), it need to reset
when resume from suspend(MEM) to fix by setting "hibernated" argument of
ohci_resume().

Signed-off-by: Guiting Shen <aarongt.shen@gmail.com>
Cc: stable <stable@kernel.org>
Reviewed-by: Alan Stern <stern@rowland.harvard.edu>
Link: https://lore.kernel.org/r/20230626152713.18950-1-aarongt.shen@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
drivers/usb/host/ohci-at91.c

index 533537ef3c21db8a7c54d4116b2c8cb2dcd2d599..360680769494bc9f4460fd441f15505f8f742476 100644 (file)
@@ -673,7 +673,13 @@ ohci_hcd_at91_drv_resume(struct device *dev)
        else
                at91_start_clock(ohci_at91);
 
-       ohci_resume(hcd, false);
+       /*
+        * According to the comment in ohci_hcd_at91_drv_suspend()
+        * we need to do a reset if the 48Mhz clock was stopped,
+        * that is, if ohci_at91->wakeup is clear. Tell ohci_resume()
+        * to reset in this case by setting its "hibernated" flag.
+        */
+       ohci_resume(hcd, !ohci_at91->wakeup);
 
        return 0;
 }