Hi Paul,

On Tue, May 06, 2014 at 11:54:19AM +0800, Pratyush ANAND wrote:
> All dwc3 based xhci host controller supports USB3.0 LPM functionality.
> Therefore enable it in platform data for all dwc3 based xhci device.
> 
> Signed-off-by: Pratyush Anand <pratyush.an...@st.com>
> ---
>  drivers/usb/dwc3/host.c | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/drivers/usb/dwc3/host.c b/drivers/usb/dwc3/host.c
> index 32db328..1fda6e8 100644
> --- a/drivers/usb/dwc3/host.c
> +++ b/drivers/usb/dwc3/host.c
> @@ -16,12 +16,14 @@
>   */
>  
>  #include <linux/platform_device.h>
> +#include <linux/usb/xhci_pdriver.h>
>  
>  #include "core.h"
>  
>  int dwc3_host_init(struct dwc3 *dwc)
>  {
>       struct platform_device  *xhci;
> +     struct usb_xhci_pdata   pdata;
>       int                     ret;
>  
>       xhci = platform_device_alloc("xhci-hcd", PLATFORM_DEVID_AUTO);
> @@ -46,6 +48,16 @@ int dwc3_host_init(struct dwc3 *dwc)
>               goto err1;
>       }
>  
> +     memset(&pdata, 0, sizeof(pdata));
> +
> +     pdata.usb3_lpm_capable = 1;

Even though all dwc3 platform is usb3 lpm capable, I would like to
make it optional, as I see some performance impact when capability is
enabled vs not enabled.

When enabled:
./testusb -a -t 2 -c 1000000 -s 1024
takes 198 sec to execute
vs when not enabled , it takes 169 sec to execute.

And this impact is with a gadget with maxburst and mult as 0.
Performance will be impacted more with higher maxburst and mult.

So making it optional would be a better choice. What do you say.

Pratyush


> +
> +     ret = platform_device_add_data(xhci, &pdata, sizeof(pdata));
> +     if (ret) {
> +             dev_err(dwc->dev, "couldn't add platform data to xHCI 
> device\n");
> +             goto err1;
> +     }
> +
>       ret = platform_device_add(xhci);
>       if (ret) {
>               dev_err(dwc->dev, "failed to register xHCI device\n");
> -- 
> 1.8.1.2
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to