On Wed, Oct 06, 2010 at 06:10:16PM +0530, Kalliguddi, Hema wrote: > Hi, > > >-----Original Message----- > >From: Ming Lei [mailto:tom.leiming@xxxxxxxxx] > >Sent: Wednesday, October 06, 2010 2:54 PM > >To: gregkh@xxxxxxx > >Cc: Balbi, Felipe; greg@xxxxxxxxx; Kalliguddi, Hema; > >linux-usb@xxxxxxxxxxxxxxx > >Subject: Re: patch "usb: musb: gadget: Unmapping the dma > >buffer when switching to PIO mode" added to gregkh-2.6 tree > > > >2010/10/6 <gregkh@xxxxxxx>: > >> > > > >> @@ -298,7 +336,7 @@ static void txstate(struct musb *musb, s > >> csr); > >> > >> #ifndef CONFIG_MUSB_PIO_ONLY > >> - if (is_dma_capable() && musb_ep->dma) { > >> + if (is_dma_capable() && !musb_ep->dma && > >musb->dma_controller) { > > > >As I said before, the change above is absolutely unnecessary, and does > >cause performance regression for usbtest #6. > > > > Due to wrong check, Tx transfers are happening in pio mode so the regression. > > Greg, > > Please drop this patch, I will send corrected patch again. Now dropped. And the one following this that fixed it up a bit. thanks, greg k-h -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
