On Thu, Jan 09, 2014 at 09:23:18AM +0800, Peter Chen wrote: > On Wed, Jan 08, 2014 at 06:25:01AM -0800, Greg KH wrote: > > On Wed, Jan 08, 2014 at 12:54:25PM +0100, Marc Kleine-Budde wrote: > > > On 01/08/2014 01:23 AM, Greg KH wrote: > > > > On Mon, Jan 06, 2014 at 10:10:41AM +0800, Peter Chen wrote: > > > >> According to Freescale imx28 Errata, "ENGR119653 USB: ARM to USB > > > >> register error issue", All USB register write operations must > > > >> use the ARM SWP instruction. So, we implement a special ehci_write > > > >> for imx28. > > > >> > > > >> Discussion for it at below: > > > >> http://marc.info/?l=linux-usb&m=137996395529294&w=2 > > > >> > > > >> Signed-off-by: Peter Chen <peter.chen@xxxxxxxxxxxxx> > > > >> Acked-by: Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> > > > >> Signed-off-by: Marc Kleine-Budde <mkl@xxxxxxxxxxxxxx> > > > >> Tested-by: Marc Kleine-Budde <mkl@xxxxxxxxxxxxxx> > > > >> --- > > > >> drivers/usb/host/ehci.h | 18 +++++++++++++++++- > > > >> 1 files changed, 17 insertions(+), 1 deletions(-) > > > >> > > > >> diff --git a/drivers/usb/host/ehci.h b/drivers/usb/host/ehci.h > > > >> index c35a6e2b..e26099b 100644 > > > >> --- a/drivers/usb/host/ehci.h > > > >> +++ b/drivers/usb/host/ehci.h > > > >> @@ -225,6 +225,7 @@ struct ehci_hcd { /* one per controller */ > > > >> unsigned has_synopsys_hc_bug:1; /* Synopsys HC */ > > > >> unsigned frame_index_bug:1; /* MosChip (AKA NetMos) */ > > > >> unsigned need_oc_pp_cycle:1; /* MPC834X port power */ > > > >> + unsigned imx28_write_fix:1; /* For Freescale i.MX28 */ > > > >> > > > >> /* required for usb32 quirk */ > > > >> #define OHCI_CTRL_HCFS (3 << 6) > > > >> @@ -728,6 +729,18 @@ static inline unsigned int ehci_readl(const struct ehci_hcd *ehci, > > > >> #endif > > > >> } > > > >> > > > >> +#ifdef CONFIG_SOC_IMX28 > > > >> +static inline void imx28_ehci_writel(const unsigned int val, > > > >> + volatile __u32 __iomem *addr) > > > >> +{ > > > >> + __asm__ ("swp %0, %0, [%1]" : : "r"(val), "r"(addr)); > > > >> +} > > > >> +#else > > > >> +static inline void imx28_ehci_writel(const unsigned int val, > > > >> + volatile __u32 __iomem *addr) > > > >> +{ > > > >> +} > > > >> +#endif > > > >> static inline void ehci_writel(const struct ehci_hcd *ehci, > > > >> const unsigned int val, __u32 __iomem *regs) > > > >> { > > > >> @@ -736,7 +749,10 @@ static inline void ehci_writel(const struct ehci_hcd *ehci, > > > >> writel_be(val, regs) : > > > >> writel(val, regs); > > > >> #else > > > >> - writel(val, regs); > > > >> + if (IS_ENABLED(CONFIG_SOC_IMX28) && ehci->imx28_write_fix) > > > >> + imx28_ehci_writel(val, regs); > > > >> + else > > > >> + writel(val, regs); > > > >> #endif > > > > > > > > This IS_ENABLED() isn't needed at all, so please remove it. > > > > > > It's an optimisation for the hot path, if kernel isn't build for a mx28 > > > the newly added if() is completely optimized out. The same argument > > > applies to the next patch. > > > > If the kernel isn't built for mx28 then the if statment goes away > > without the IS_ENABLED() call as well, due to the empty inline > > function, right? > > > > With IS_ENABLED(), the non-imx28 platform don't need to judge the > condition of ehci->imx28_write_fix, if you don't think we need > to care one or two instructions for every other platforms, I can > delete it. Please just delete it, I doubt it matters at all. If it does, and is measurable, feel free to add it back at a later time. 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
