On 2017/6/3 18:23, Marek Vasut wrote:
On 06/01/2017 01:25 PM, Meng Dongyang wrote:
Fix macro error of dwc2 driver, add macro definition to config force mode and HNP/SRP capability.
Can you please elaborate some more on the purpose of this patch ? It seems wrong, since the driver works on systems using DWC2 as is and this patch changes constants used in the driver.
OK, I will think about making it works as before. But I found that some config need to be change on some platform.
Signed-off-by: Meng Dongyang daniel.meng@rock-chips.com
Changes in v2:
Splited from patch [07/08] of v1
drivers/usb/host/dwc2.c | 9 ++++++++- drivers/usb/host/dwc2.h | 10 ++++++---- 2 files changed, 14 insertions(+), 5 deletions(-)
diff --git a/drivers/usb/host/dwc2.c b/drivers/usb/host/dwc2.c index 0e5df15..51619d6 100644 --- a/drivers/usb/host/dwc2.c +++ b/drivers/usb/host/dwc2.c @@ -370,7 +370,7 @@ static void dwc_otg_core_init(struct dwc2_priv *priv) usbcfg &= ~DWC2_GUSBCFG_DDRSEL; #endif } else { /* UTMI+ interface */ -#if (CONFIG_DWC2_UTMI_PHY_WIDTH == 16) +#if (CONFIG_DWC2_UTMI_WIDTH == 16)
The macro definition of UTMI width is CONFIG_DWC2_UTMI_WIDTH, yet here is CONFIG_DWC2_UTMI_PHY_WIDTH. So maybe it's better to change them to the same one.
usbcfg |= DWC2_GUSBCFG_PHYIF;#endif } @@ -394,6 +394,11 @@ static void dwc_otg_core_init(struct dwc2_priv *priv) usbcfg |= DWC2_GUSBCFG_ULPI_CLK_SUS_M; } #endif
+#ifdef CONFIG_DWC2_FORCE_HOST_MODE
- usbcfg |= DWC2_GUSBCFG_FORCEHOSTMODE;
+#endif
I found that on the platform not support HNP/SRP capability, the dr_mode will still be otg after init and the host mode can't work. So may be a config of force_mode is necessary.
writel(usbcfg, ®s->gusbcfg);
/* Program the GAHBCFG Register. */ @@ -422,8 +427,10 @@ static void dwc_otg_core_init(struct dwc2_priv *priv)
writel(ahbcfg, ®s->gahbcfg);
+#ifdef CONFIG_DWC2_HNP_SRP_CAP /* Program the GUSBCFG register for HNP/SRP. */ setbits_le32(®s->gusbcfg, DWC2_GUSBCFG_HNPCAP | DWC2_GUSBCFG_SRPCAP); +#endif
#ifdef CONFIG_DWC2_IC_USB_CAP setbits_le32(®s->gusbcfg, DWC2_GUSBCFG_IC_USB_CAP); diff --git a/drivers/usb/host/dwc2.h b/drivers/usb/host/dwc2.h index 4482dc6..3c62e9b 100644 --- a/drivers/usb/host/dwc2.h +++ b/drivers/usb/host/dwc2.h @@ -771,11 +771,11 @@ struct dwc2_core_regs { #define CONFIG_DWC2_MAX_TRANSFER_SIZE 65535 #define CONFIG_DWC2_MAX_PACKET_COUNT 511
-#define DWC2_PHY_TYPE_FS 0 -#define DWC2_PHY_TYPE_UTMI 1 -#define DWC2_PHY_TYPE_ULPI 2 +#define DWC2_PHY_TYPE_UTMI 0 +#define DWC2_PHY_TYPE_ULPI 1 +#define DWC2_PHY_TYPE_FS 2
According to table 5-10 of "DesignWare Cores USB 2.0 Hi-Speed On-The-Go(OTG)" 3.10a databook,
ULPI or UTMI+ Select (ULPI_UTMI_Sel) The application uses this bit to select either a UTMI+ interface or ULPI Interface. ■ 1'b0: UTMI+ Interface ■ 1'b1: ULPI Interface
But the macor definition in current code would make a opposite config on ULPI_UTMI_Sel bit, (in the function "dwc_otg_core_init") 1: UTMI+ Interface 0: ULPI Interface
It seems wrong.
#define CONFIG_DWC2_PHY_TYPE DWC2_PHY_TYPE_UTMI /* PHY type */ -#define CONFIG_DWC2_UTMI_WIDTH 8 /* UTMI bus width (8/16) */ +#define CONFIG_DWC2_UTMI_WIDTH 16 /* UTMI bus width (8/16) */
So basically you now force 16bit UTMI bus width instead of 8 ?
Three changes basically 1. force 16bit UTMI bus width instead of 8 2. config force mode on the platform without HNP/SRP capability. 3. change macro definition about PHY type (ULPI_UTMI_Sel).
This breaks every single platform in U-Boot using the DWC2.
#undef CONFIG_DWC2_PHY_ULPI_DDR /* ULPI PHY uses DDR mode */ #define CONFIG_DWC2_PHY_ULPI_EXT_VBUS /* ULPI PHY controls VBUS */ @@ -785,5 +785,7 @@ struct dwc2_core_regs { #undef CONFIG_DWC2_THR_CTL /* Threshold control */ #define CONFIG_DWC2_TX_THR_LENGTH 64 #undef CONFIG_DWC2_IC_USB_CAP /* IC Cap */ +#define CONFIG_DWC2_FORCE_HOST_MODE /* Force host mode */ +#undef CONFIG_DWC2_HNP_SRP_CAP /* HNP/SRP Cap */
#endif /* __DWC2_H__ */