ARM: sunxi: Fix USB host init fail on v5.1-rc1
Subsystems (4)
Chen-Yu Tsai
maintainercover ARM: sunxi: Fix USB host init fail on v5.1-rc1
From: Chen-Yu Tsai <[email protected]>
Hi everyone,
As previously reported [1], USB EHCI/OHCI hosts are broken on Allwinner SoCs on v5.1-rc1. This is partially triggered by commit b97a31348379 ("usb: core: comply to PHY framework"), and partially due to how the Allwinner USB PHY driver handles phy_set_mode for non-OTG PHYs.
This series fixes this in both places.
Patch 1 makes phy-sun4i-usb accept PHY_MODE_USB_HOST for non-OTG PHYs.
Patch 2 makes the usb core fall back to setting the mode PHY_MODE_USB_HOST if set_mode with PHY_MODE_USB_HOST_SS fails. If that fails then the failure path is the same as before. This should make it so existing USB 3.0 drivers are affected.
The patches don't have any dependencies on each other, and could go in through separate branches. However this affects usability of USB input devices and Ethernet dongles, I'd rather they go in sooner than later.
Regards
ChenYu
[1] https://lkml.org/lkml/2019/3/18/74
Chen-Yu Tsai (2):
phy: sun4i-usb: Support set_mode to USB_HOST for non-OTG PHYs
usb: core: Try generic PHY_MODE_USB_HOST if usb_phy_roothub_set_mode
fails
drivers/phy/allwinner/phy-sun4i-usb.c | 5 ++++- drivers/usb/core/hcd.c | 3 +++ 2 files changed, 7 insertions(+), 1 deletion(-)
Signature · 1 line
2.20.1
Chen-Yu Tsai
maintainerPATCH 1/2 phy: sun4i-usb: Support set_mode to USB_HOST for non-OTG PHYs
From: Chen-Yu Tsai <[email protected]>
While only the first PHY supports mode switching, the remaining PHYs work in USB host mode. They should support set_mode with mode=USB_HOST instead of failing. This is especially needed now that the USB core does set_mode for all USB ports, which was added in commit b97a31348379 ("usb:
core: comply to PHY framework").
Make set_mode with mode=USB_HOST a no-op instead of failing for the non-OTG USB PHYs.
--- drivers/phy/allwinner/phy-sun4i-usb.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/phy/allwinner/phy-sun4i-usb.c +4 -1
@@ -485,8 +485,11 @@ static int sun4i_usb_phy_set_mode(struct phy *_phy,struct sun4i_usb_phy_data *data =to_sun4i_usb_phy_data (phy);int new_mode; -if (phy->index !=0 ) +if (phy->index !=0 ) { +if (mode ==PHY_MODE_USB_HOST ) +return 0 ;return -EINVAL ; + }switch (mode) {case PHY_MODE_USB_HOST :
Chen-Yu Tsai
maintainerPATCH 2/2 usb: core: Try generic PHY_MODE_USB_HOST if usb_phy_roothub_set_mode fails
In reply to Chen-Yu TsaiFrom: Chen-Yu Tsai <[email protected]>
Some PHYs do not support PHY_MODE_USB_HOST_SS, i.e. USB 3.0 or higher. Fall back and try the more generic PHY_MODE_USB_HOST if it fails.
--- drivers/usb/core/hcd.c | 3 +++ 1 file changed, 3 insertions(+)
drivers/usb/core/hcd.c +3 -0
@@ -2741,6 +2741,9 @@ int usb_add_hcd(struct usb_hcd *hcd, retval =usb_phy_roothub_set_mode (hcd->phy_roothub,PHY_MODE_USB_HOST_SS ); +if (retval) + retval =usb_phy_roothub_set_mode (hcd->phy_roothub, +PHY_MODE_USB_HOST );if (retval) goto err_usb_phy_roothub_power_on;
Re: [PATCH 2/2] usb: core: Try generic PHY_MODE_USB_HOST if usb_phy_roothub_set_mode fails
Quoted 17 lines · Chen-Yu Tsai
From: Chen-Yu Tsai <[email protected]>
Some PHYs do not support
PHY_MODE_USB_HOST_SS, i.e. USB 3.0 or higher. Fall back and try the more genericPHY_MODE_USB_HOSTif it fails.Fixes b97a31348379 ("usb: core: comply to PHY framework")Signed-off-by Chen-Yu Tsai <[email protected]>--- drivers/usb/core/hcd.c | 3 +++ 1 file changed, 3 insertions(+)drivers/usb/core/hcd.c +0
-0
drivers/usb/core/hcd.c int usb_add_hcd(struct usb_hcd *hcd, · line 2749
Quoted 9 lines
drivers/usb/core/hcd.c +3
-0@@ -2741,6 +2741,9 @@ int usb_add_hcd(struct usb_hcd *hcd, retval =usb_phy_roothub_set_mode (hcd->phy_roothub,PHY_MODE_USB_HOST_SS ); +if (retval) + retval =usb_phy_roothub_set_mode (hcd->phy_roothub, +PHY_MODE_USB_HOST );if (retval) goto err_usb_phy_roothub_power_on;
This fixes USB on Amlogic GXL/GXL, tested on a Nexbox A1 with a S912 SoC running v5.1-rc1
Error was :
[ 4.851477] phy phy-d0078080.phy.3: unsupported PHY mode 5 [ 4.856266] xhci-hcd: probe of xhci-hcd.0.auto failed with error -22
Neil
Quoted 13 lines · Chen-Yu Tsai
From: Chen-Yu Tsai <[email protected]>
Hi everyone,
As previously reported [1], USB EHCI/OHCI hosts are broken on Allwinner SoCs on v5.1-rc1. This is partially triggered by commit
b97a31348379("usb: core: comply to PHY framework"), and partially due to how the Allwinner USB PHY driver handlesphy_set_modefor non-OTG PHYs.This series fixes this in both places.
Patch 1 makes phy-sun4i-usb accept
PHY_MODE_USB_HOSTfor non-OTG PHYs.
Quoted 8 lines
Patch 2 makes the usb core fall back to setting the mode
PHY_MODE_USB_HOSTifset_modewithPHY_MODE_USB_HOST_SSfails. If that fails then the failure path is the same as before. This should make it so existing USB 3.0 drivers are affected.The patches don't have any dependencies on each other, and could go in through separate branches. However this affects usability of USB input devices and Ethernet dongles, I'd rather they go in sooner than later.
Maxime
--
Maxime Ripard, Bootlin
Embedded Linux and Kernel engineering
https://bootlin.com
Greg Kroah-Hartman
maintainerQuoted 13 lines · Chen-Yu Tsai
From: Chen-Yu Tsai <[email protected]>
Hi everyone,
As previously reported [1], USB EHCI/OHCI hosts are broken on Allwinner SoCs on v5.1-rc1. This is partially triggered by commit
b97a31348379("usb: core: comply to PHY framework"), and partially due to how the Allwinner USB PHY driver handlesphy_set_modefor non-OTG PHYs.This series fixes this in both places.
Patch 1 makes phy-sun4i-usb accept
PHY_MODE_USB_HOSTfor non-OTG PHYs.
Quoted 8 lines
Patch 2 makes the usb core fall back to setting the mode
PHY_MODE_USB_HOSTifset_modewithPHY_MODE_USB_HOST_SSfails. If that fails then the failure path is the same as before. This should make it so existing USB 3.0 drivers are affected.The patches don't have any dependencies on each other, and could go in through separate branches. However this affects usability of USB input devices and Ethernet dongles, I'd rather they go in sooner than later.
I'll take them both now, thanks.
greg k-h