| Message ID | Yo4gqLPtHO6XKMLn@kili |
|---|---|
| State | New |
| Headers | show |
| Series | USB: usbip: clean up mixed use of _irq() and _irqsave() | expand |
On 5/25/22 6:27 AM, Dan Carpenter wrote: > It generally doesn't make sense to use _irq() and _irqsave() in the same > function because either some of the callers have disabled IRQs or they > haven't. In this case, the v_recv_cmd_submit() appears to always be > called with IRQs enabled so the code works fine. That means I could > convert it to either _irq() or _irqsave() but I chose to use _irqsave() > because it's more conservative and easier to review. > > Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> > --- > drivers/usb/usbip/vudc_rx.c | 6 +++--- > 1 file changed, 3 insertions(+), 3 deletions(-) > > diff --git a/drivers/usb/usbip/vudc_rx.c b/drivers/usb/usbip/vudc_rx.c > index 1e8a23d92cb4..d4a2f30a7580 100644 > --- a/drivers/usb/usbip/vudc_rx.c > +++ b/drivers/usb/usbip/vudc_rx.c > @@ -104,18 +104,18 @@ static int v_recv_cmd_submit(struct vudc *udc, > if (pdu->base.direction == USBIP_DIR_IN) > address |= USB_DIR_IN; > > - spin_lock_irq(&udc->lock); > + spin_lock_irqsave(&udc->lock, flags); > urb_p->ep = vudc_find_endpoint(udc, address); > if (!urb_p->ep) { > /* we don't know the type, there may be isoc data! */ > dev_err(&udc->pdev->dev, "request to nonexistent endpoint"); > - spin_unlock_irq(&udc->lock); > + spin_unlock_irqrestore(&udc->lock, flags); > usbip_event_add(&udc->ud, VUDC_EVENT_ERROR_TCP); > ret = -EPIPE; > goto free_urbp; > } > urb_p->type = urb_p->ep->type; > - spin_unlock_irq(&udc->lock); > + spin_unlock_irqrestore(&udc->lock, flags); > > urb_p->new = 1; > urb_p->seqnum = pdu->base.seqnum; > Thank you for the patch. Reviewed-by: Shuah Khan <skhan@linuxfoundation.org> Greg, please pick this up. thanks, -- Shuah
diff --git a/drivers/usb/usbip/vudc_rx.c b/drivers/usb/usbip/vudc_rx.c index 1e8a23d92cb4..d4a2f30a7580 100644 --- a/drivers/usb/usbip/vudc_rx.c +++ b/drivers/usb/usbip/vudc_rx.c @@ -104,18 +104,18 @@ static int v_recv_cmd_submit(struct vudc *udc, if (pdu->base.direction == USBIP_DIR_IN) address |= USB_DIR_IN; - spin_lock_irq(&udc->lock); + spin_lock_irqsave(&udc->lock, flags); urb_p->ep = vudc_find_endpoint(udc, address); if (!urb_p->ep) { /* we don't know the type, there may be isoc data! */ dev_err(&udc->pdev->dev, "request to nonexistent endpoint"); - spin_unlock_irq(&udc->lock); + spin_unlock_irqrestore(&udc->lock, flags); usbip_event_add(&udc->ud, VUDC_EVENT_ERROR_TCP); ret = -EPIPE; goto free_urbp; } urb_p->type = urb_p->ep->type; - spin_unlock_irq(&udc->lock); + spin_unlock_irqrestore(&udc->lock, flags); urb_p->new = 1; urb_p->seqnum = pdu->base.seqnum;
It generally doesn't make sense to use _irq() and _irqsave() in the same function because either some of the callers have disabled IRQs or they haven't. In this case, the v_recv_cmd_submit() appears to always be called with IRQs enabled so the code works fine. That means I could convert it to either _irq() or _irqsave() but I chose to use _irqsave() because it's more conservative and easier to review. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> --- drivers/usb/usbip/vudc_rx.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)