Linux USB
 help / color / mirror / Atom feed
* [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init
@ 2026-11-17 13:00 Roadbobek via B4 Relay
  2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
                   ` (2 more replies)
  0 siblings, 3 replies; 7+ messages in thread
From: Roadbobek via B4 Relay @ 2026-11-17 13:00 UTC (permalink / raw)
  To: Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Philipp Zabel, Vinod Koul, Neil Armstrong,
	Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy,
	Roadbobek

When a bootloader such as U-Boot (for example with
CONFIG_USB_MUSB_HOST=y) initializes USB host mode and boots Linux
without stopping USB first, both the MUSB controller and USB0 PHY
reset lines are already deasserted when Linux probes them.

Because reset_control_deassert() alone is a no-op on an already
deasserted reset line, stale hardware register state persists across
boot. On Allwinner A33, MUSB stays wedged in a_wait_bcon with
DevCtl=0x99, Power=0xf0, and a frozen Frame counter even across
hot-unplug and replug.

This two-patch series briefly asserts reset before deasserting it in
both drivers so USB starts cleanly regardless of prior bootloader state:

- Patch 1 asserts glue->rst in sunxi_musb_init().
- Patch 2 asserts phy->reset in sun4i_usb_phy_init().

Tested on an Allwinner A33 Q8 tablet (Laser MID-786).

Signed-off-by: Roadbobek <admin@roadbobek.com>
---
Roadbobek (2):
      usb: musb: sunxi: Assert reset before deasserting on init
      phy: sun4i-usb: Assert reset before deasserting in phy init

 drivers/phy/allwinner/phy-sun4i-usb.c | 2 ++
 drivers/usb/musb/sunxi.c              | 3 +++
 2 files changed, 5 insertions(+)
---
base-commit: 666539eecdf4db623f5e3c3e55cf9f2dfad24aa4
change-id: 20261117-usb-sunxi-reset-fix-d1a0001440a9

Best regards,
--  
Roadbobek <admin@roadbobek.com>



^ permalink raw reply	[flat|nested] 7+ messages in thread

* [PATCH 1/2] usb: musb: sunxi: Assert reset before deasserting on init
  2026-11-17 13:00 [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Roadbobek via B4 Relay
@ 2026-11-17 13:00 ` Roadbobek via B4 Relay
  2026-11-17 13:05   ` Philipp Zabel
  2026-11-17 13:09   ` sashiko-bot
  2026-11-17 13:00 ` [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init Roadbobek via B4 Relay
  2026-11-17 13:04 ` [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Philipp Zabel
  2 siblings, 2 replies; 7+ messages in thread
From: Roadbobek via B4 Relay @ 2026-11-17 13:00 UTC (permalink / raw)
  To: Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Philipp Zabel, Vinod Koul, Neil Armstrong,
	Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy,
	Roadbobek

From: Roadbobek <admin@roadbobek.com>

When a bootloader such as U-Boot (for example with
CONFIG_USB_MUSB_HOST=y) initializes MUSB and boots Linux
without stopping USB first, the controller reset line is
already deasserted at probe time.

In that case reset_control_deassert() in sunxi_musb_init()
is a no-op and stale MUSB register state persists across boot.
On Allwinner A33, MUSB stays wedged in a_wait_bcon with
DevCtl=0x99, Power=0xf0, and a frozen Frame counter even
across hot-unplug and replug.

Assert glue->rst briefly before deasserting it when
SUNXI_MUSB_FL_HAS_RESET is set so MUSB starts cleanly.
Tested on an Allwinner A33 Q8 tablet (Laser MID-786).

Signed-off-by: Roadbobek <admin@roadbobek.com>
---
 drivers/usb/musb/sunxi.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/usb/musb/sunxi.c b/drivers/usb/musb/sunxi.c
index a17b1c150626..3d16049e53b5 100644
--- a/drivers/usb/musb/sunxi.c
+++ b/drivers/usb/musb/sunxi.c
@@ -9,6 +9,7 @@
  */
 
 #include <linux/clk.h>
+#include <linux/delay.h>
 #include <linux/err.h>
 #include <linux/extcon.h>
 #include <linux/io.h>
@@ -239,6 +240,8 @@ static int sunxi_musb_init(struct musb *musb)
 		goto error_sram_release;
 
 	if (test_bit(SUNXI_MUSB_FL_HAS_RESET, &glue->flags)) {
+		reset_control_assert(glue->rst);
+		usleep_range(10, 20);
 		ret = reset_control_deassert(glue->rst);
 		if (ret)
 			goto error_clk_disable;

-- 
2.53.0



^ permalink raw reply related	[flat|nested] 7+ messages in thread

* [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init
  2026-11-17 13:00 [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Roadbobek via B4 Relay
  2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
@ 2026-11-17 13:00 ` Roadbobek via B4 Relay
  2026-11-17 13:04   ` Philipp Zabel
  2026-11-17 13:04 ` [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Philipp Zabel
  2 siblings, 1 reply; 7+ messages in thread
From: Roadbobek via B4 Relay @ 2026-11-17 13:00 UTC (permalink / raw)
  To: Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Philipp Zabel, Vinod Koul, Neil Armstrong,
	Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy,
	Roadbobek

From: Roadbobek <admin@roadbobek.com>

When a bootloader initializes USB before booting Linux without
first tearing down the controller, phy->reset may already be
deasserted when sun4i_usb_phy_init() runs.

Because reset_control_deassert() alone is a no-op on an
already deasserted reset line, stale USB PHY state can
persist across the bootloader-to-kernel handoff.

Assert phy->reset briefly before deasserting it in
sun4i_usb_phy_init() so the PHY starts cleanly regardless of
prior bootloader USB usage. Tested on an Allwinner A33 Q8
tablet (Laser MID-786).

Signed-off-by: Roadbobek <admin@roadbobek.com>
---
 drivers/phy/allwinner/phy-sun4i-usb.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/phy/allwinner/phy-sun4i-usb.c b/drivers/phy/allwinner/phy-sun4i-usb.c
index f26eed0db17e..05e6dcf035a1 100644
--- a/drivers/phy/allwinner/phy-sun4i-usb.c
+++ b/drivers/phy/allwinner/phy-sun4i-usb.c
@@ -270,6 +270,8 @@ static int sun4i_usb_phy_init(struct phy *_phy)
 		return ret;
 	}
 
+	reset_control_assert(phy->reset);
+	usleep_range(10, 20);
 	ret = reset_control_deassert(phy->reset);
 	if (ret) {
 		clk_disable_unprepare(phy->clk2);

-- 
2.53.0



^ permalink raw reply related	[flat|nested] 7+ messages in thread

* Re: [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init
  2026-11-17 13:00 [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Roadbobek via B4 Relay
  2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
  2026-11-17 13:00 ` [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init Roadbobek via B4 Relay
@ 2026-11-17 13:04 ` Philipp Zabel
  2 siblings, 0 replies; 7+ messages in thread
From: Philipp Zabel @ 2026-11-17 13:04 UTC (permalink / raw)
  To: admin, Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Vinod Koul, Neil Armstrong, Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy

On Di, 2026-11-17 at 13:00 +0000, Roadbobek via B4 Relay wrote:
> When a bootloader such as U-Boot (for example with
> CONFIG_USB_MUSB_HOST=y) initializes USB host mode and boots Linux
> without stopping USB first, both the MUSB controller and USB0 PHY
> reset lines are already deasserted when Linux probes them.
> 
> Because reset_control_deassert() alone is a no-op on an already
> deasserted reset line, stale hardware register state persists across
> boot. On Allwinner A33, MUSB stays wedged in a_wait_bcon with
> DevCtl=0x99, Power=0xf0, and a frozen Frame counter even across
> hot-unplug and replug.
> 
> This two-patch series briefly asserts reset before deasserting it in
> both drivers so USB starts cleanly regardless of prior bootloader state:
> 
> - Patch 1 asserts glue->rst in sunxi_musb_init().
> - Patch 2 asserts phy->reset in sun4i_usb_phy_init().

While you are at it, could you also replace the
devm_reset_control_get() calls with devm_reset_control_get_exclusive(),
to make it clear the driver requires exclusive access to assert the
resets?

regards
Philipp

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init
  2026-11-17 13:00 ` [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init Roadbobek via B4 Relay
@ 2026-11-17 13:04   ` Philipp Zabel
  0 siblings, 0 replies; 7+ messages in thread
From: Philipp Zabel @ 2026-11-17 13:04 UTC (permalink / raw)
  To: admin, Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Vinod Koul, Neil Armstrong, Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy

On Di, 2026-11-17 at 13:00 +0000, Roadbobek via B4 Relay wrote:
> From: Roadbobek <admin@roadbobek.com>
> 
> When a bootloader initializes USB before booting Linux without
> first tearing down the controller, phy->reset may already be
> deasserted when sun4i_usb_phy_init() runs.
> 
> Because reset_control_deassert() alone is a no-op on an
> already deasserted reset line, stale USB PHY state can
> persist across the bootloader-to-kernel handoff.
> 
> Assert phy->reset briefly before deasserting it in
> sun4i_usb_phy_init() so the PHY starts cleanly regardless of
> prior bootloader USB usage. Tested on an Allwinner A33 Q8
> tablet (Laser MID-786).
> 
> Signed-off-by: Roadbobek <admin@roadbobek.com>
> ---
>  drivers/phy/allwinner/phy-sun4i-usb.c | 2 ++
>  1 file changed, 2 insertions(+)
> 
> diff --git a/drivers/phy/allwinner/phy-sun4i-usb.c b/drivers/phy/allwinner/phy-sun4i-usb.c
> index f26eed0db17e..05e6dcf035a1 100644
> --- a/drivers/phy/allwinner/phy-sun4i-usb.c
> +++ b/drivers/phy/allwinner/phy-sun4i-usb.c
> @@ -270,6 +270,8 @@ static int sun4i_usb_phy_init(struct phy *_phy)
>  		return ret;
>  	}
>  
> +	reset_control_assert(phy->reset);

No error handling?

> +	usleep_range(10, 20);
>  	ret = reset_control_deassert(phy->reset);
>  	if (ret) {
>  		clk_disable_unprepare(phy->clk2);

regards
Philipp

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH 1/2] usb: musb: sunxi: Assert reset before deasserting on init
  2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
@ 2026-11-17 13:05   ` Philipp Zabel
  2026-11-17 13:09   ` sashiko-bot
  1 sibling, 0 replies; 7+ messages in thread
From: Philipp Zabel @ 2026-11-17 13:05 UTC (permalink / raw)
  To: admin, Bin Liu, Greg Kroah-Hartman, Chen-Yu Tsai, Jernej Skrabec,
	Samuel Holland, Vinod Koul, Neil Armstrong, Manivannan Sadhasivam
  Cc: linux-usb, linux-arm-kernel, linux-sunxi, linux-kernel, linux-phy

On Di, 2026-11-17 at 13:00 +0000, Roadbobek via B4 Relay wrote:
> From: Roadbobek <admin@roadbobek.com>
> 
> When a bootloader such as U-Boot (for example with
> CONFIG_USB_MUSB_HOST=y) initializes MUSB and boots Linux
> without stopping USB first, the controller reset line is
> already deasserted at probe time.
> 
> In that case reset_control_deassert() in sunxi_musb_init()
> is a no-op and stale MUSB register state persists across boot.
> On Allwinner A33, MUSB stays wedged in a_wait_bcon with
> DevCtl=0x99, Power=0xf0, and a frozen Frame counter even
> across hot-unplug and replug.
> 
> Assert glue->rst briefly before deasserting it when
> SUNXI_MUSB_FL_HAS_RESET is set so MUSB starts cleanly.
> Tested on an Allwinner A33 Q8 tablet (Laser MID-786).
> 
> Signed-off-by: Roadbobek <admin@roadbobek.com>
> ---
>  drivers/usb/musb/sunxi.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/usb/musb/sunxi.c b/drivers/usb/musb/sunxi.c
> index a17b1c150626..3d16049e53b5 100644
> --- a/drivers/usb/musb/sunxi.c
> +++ b/drivers/usb/musb/sunxi.c
> @@ -9,6 +9,7 @@
>   */
>  
>  #include <linux/clk.h>
> +#include <linux/delay.h>
>  #include <linux/err.h>
>  #include <linux/extcon.h>
>  #include <linux/io.h>
> @@ -239,6 +240,8 @@ static int sunxi_musb_init(struct musb *musb)
>  		goto error_sram_release;
>  
>  	if (test_bit(SUNXI_MUSB_FL_HAS_RESET, &glue->flags)) {
> +		reset_control_assert(glue->rst);

No error handling?

> +		usleep_range(10, 20);
>  		ret = reset_control_deassert(glue->rst);
>  		if (ret)
>  			goto error_clk_disable;

regards
Philipp

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [PATCH 1/2] usb: musb: sunxi: Assert reset before deasserting on init
  2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
  2026-11-17 13:05   ` Philipp Zabel
@ 2026-11-17 13:09   ` sashiko-bot
  1 sibling, 0 replies; 7+ messages in thread
From: sashiko-bot @ 2026-11-17 13:09 UTC (permalink / raw)
  To: Roadbobek
  Cc: linux-usb, linux-sunxi, linux-phy, vkoul, neil.armstrong, olteanv

Thank you for your contribution! Sashiko AI review found 2 potential
issues to consider:
- [High] usb: musb: sunxi: Signed-off-by trailer uses single-word handle
  Roadbobek
- [Low] usb: musb: sunxi: missing Fixes trailer for sunxi_musb_init()
  reset fix
--

commit 6fb0d83bc22980ce7dcaa8d7c886a7bcf57b0905
Author: Roadbobek <admin@roadbobek.com>

> usb: musb: sunxi: Assert reset before deasserting on init
>
> When a bootloader such as U-Boot (for example with
> CONFIG_USB_MUSB_HOST=y) initializes MUSB and boots Linux
> without stopping USB first, the controller reset line is
> already deasserted at probe time.
>
> In that case reset_control_deassert() in sunxi_musb_init()
> is a no-op and stale MUSB register state persists across boot.
> On Allwinner A33, MUSB stays wedged in a_wait_bcon with
> DevCtl=0x99, Power=0xf0, and a frozen Frame counter even
> across hot-unplug and replug.
>
> Assert glue->rst briefly before deasserting it when
> SUNXI_MUSB_FL_HAS_RESET is set so MUSB starts cleanly.
> Tested on an Allwinner A33 Q8 tablet (Laser MID-786).
>
> Signed-off-by: Roadbobek <admin@roadbobek.com>

[Severity: High]
Per Documentation/process/submitting-patches.rst, the Author header and
Signed-off-by trailer must use a real full name rather than a single-word
handle or nickname.

[Severity: Low]
Since this fixes the reset initialization in sunxi_musb_init(), should it
include:

  Fixes: 132e23775779 ("usb: musb: sunxi: Add support for musb controller in A31 SoC")

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261117-usb-sunxi-reset-fix-v1-0-c2b76e0e84e1@roadbobek.com?part=1

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2026-11-17 13:09 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-17 13:00 [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Roadbobek via B4 Relay
2026-11-17 13:00 ` [PATCH 1/2] usb: musb: " Roadbobek via B4 Relay
2026-11-17 13:05   ` Philipp Zabel
2026-11-17 13:09   ` sashiko-bot
2026-11-17 13:00 ` [PATCH 2/2] phy: sun4i-usb: Assert reset before deasserting in phy init Roadbobek via B4 Relay
2026-11-17 13:04   ` Philipp Zabel
2026-11-17 13:04 ` [PATCH 0/2] usb/phy: sunxi: Assert reset before deasserting on init Philipp Zabel

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox