diff mbox

[U-Boot,v5,1/1] tegra: usb: Fix device enumeration problem of USB1

Message ID 4B9C9637D5087840A465BDCB251780E9E2D6EDA3E9@HKMAIL02.nvidia.com
State Accepted
Commit 8b3f7bf7ec95816a7d2f89331e8877a351b3ec7c
Headers show

Commit Message

Jim Lin June 25, 2012, 6:40 a.m. UTC
A known hardware issue of USB1 port where bit 1 (connect status
change) of PORTSC register will be set after issuing Port Reset
(like "usb reset" in u-boot command line).
This will be treated as an error and stops later device enumeration.

Therefore we clear that bit after Port Reset in order to proceed
later device enumeration.

Signed-off-by: Jim Lin <jilin@nvidia.com>
---
To reproduce this issue, you can modify board .dts file to set
as the following to build u-boot binary.
"
 usb0 = "/usb@c5000000";
 usb1 = "/usb@c5008000";
"
Install device on USB1 port (address at 0xc5000000).
And run "usb reset" in u-boot console to enumerate device.

Before adding this patch, we could see problem every time.
After adding, tried 10 times of "usb reset", "usb tree", "usb stop"
, without seeing issue.

Changes in v5:
- Define USB address mask value in header file

Changes in v4:
- Add comment to describe replacing weak function ehci_powerup_fixup of ehci-hcd.c
- Remove using variable my_reg

Changes in v3:
- Move patch for USB1 controller into ehci_powerup_fixup of ehci-tegra.c
- Update copyright year to 2012

Changes in v2:
- Change config name
- Add a callback function at the end of ehci_submit_root() function

 arch/arm/include/asm/arch-tegra2/tegra2.h |    1 +
 drivers/usb/host/ehci-tegra.c             |   18 +++++++++++++++++-
 2 files changed, 18 insertions(+), 1 deletions(-)

Comments

Stephen Warren June 25, 2012, 3:48 p.m. UTC | #1
On 06/25/2012 12:40 AM, Jim Lin wrote:
> A known hardware issue of USB1 port where bit 1 (connect status
> change) of PORTSC register will be set after issuing Port Reset
> (like "usb reset" in u-boot command line).
> This will be treated as an error and stops later device enumeration.
> 
> Therefore we clear that bit after Port Reset in order to proceed
> later device enumeration.
> 
> Signed-off-by: Jim Lin <jilin@nvidia.com>

Acked-by: Stephen Warren <swarren@wwwdotorg.org>
Marek Vasut June 25, 2012, 11:47 p.m. UTC | #2
Dear Stephen Warren,

> On 06/25/2012 12:40 AM, Jim Lin wrote:
> > A known hardware issue of USB1 port where bit 1 (connect status
> > change) of PORTSC register will be set after issuing Port Reset
> > (like "usb reset" in u-boot command line).
> > This will be treated as an error and stops later device enumeration.
> > 
> > Therefore we clear that bit after Port Reset in order to proceed
> > later device enumeration.
> > 
> > Signed-off-by: Jim Lin <jilin@nvidia.com>
> 
> Acked-by: Stephen Warren <swarren@wwwdotorg.org>

Ok, it's icky nvidia stuff :-)

Will you apply it? I'm good with the current shape.

Best regards,
Marek Vasut
Marek Vasut June 26, 2012, 1:05 a.m. UTC | #3
Dear Tom Warren,

> Marek,
> 
> > -----Original Message-----
> > From: Marek Vasut [mailto:marek.vasut@gmail.com]
> > Sent: Monday, June 25, 2012 4:48 PM
> > To: Stephen Warren
> > Cc: Jim Lin; 'u-boot@lists.denx.de'; Tom Warren
> > Subject: Re: [PATCH v5 1/1] tegra: usb: Fix device enumeration problem of
> > USB1
> > 
> > Dear Stephen Warren,
> > 
> > > On 06/25/2012 12:40 AM, Jim Lin wrote:
> > > > A known hardware issue of USB1 port where bit 1 (connect status
> > > > change) of PORTSC register will be set after issuing Port Reset
> > > > (like "usb reset" in u-boot command line).
> > > > This will be treated as an error and stops later device enumeration.
> > > > 
> > > > Therefore we clear that bit after Port Reset in order to proceed
> > > > later device enumeration.
> > > > 
> > > > Signed-off-by: Jim Lin <jilin@nvidia.com>
> > > 
> > > Acked-by: Stephen Warren <swarren@wwwdotorg.org>
> > 
> > Ok, it's icky nvidia stuff :-)
> > 
> > Will you apply it? I'm good with the current shape.
> 
> This will be applied to u-boot-tegra/next as soon as I'm done validating
> Allen's SPL changes.

Good, I'm sure I'm leaving it in good hands :)

Thanks Tom!

> Tom
> 
> > Best regards,
> > Marek Vasut

Best regards,
Marek Vasut
diff mbox

Patch

diff --git a/arch/arm/include/asm/arch-tegra2/tegra2.h b/arch/arm/include/asm/arch-tegra2/tegra2.h
index d4ada10..710334b 100644
--- a/arch/arm/include/asm/arch-tegra2/tegra2.h
+++ b/arch/arm/include/asm/arch-tegra2/tegra2.h
@@ -45,6 +45,7 @@ 
 #define NV_PA_CSITE_BASE	0x70040000
 #define TEGRA_USB1_BASE		0xC5000000
 #define TEGRA_USB3_BASE		0xC5008000
+#define TEGRA_USB_ADDR_MASK	0xFFFFC000
 
 #define TEGRA2_SDRC_CS0		NV_PA_SDRAM_BASE
 #define LOW_LEVEL_SRAM_STACK	0x4000FFFC
diff --git a/drivers/usb/host/ehci-tegra.c b/drivers/usb/host/ehci-tegra.c
index a7e105b..4646b29 100644
--- a/drivers/usb/host/ehci-tegra.c
+++ b/drivers/usb/host/ehci-tegra.c
@@ -1,5 +1,5 @@ 
 /*
- * Copyright (c) 2009 NVIDIA Corporation
+ * Copyright (c) 2009-2012 NVIDIA Corporation
  *
  * See file CREDITS for list of people who contributed to this
  * project.
@@ -29,6 +29,22 @@ 
 #include <asm/errno.h>
 #include <asm/arch/usb.h>
 
+/*
+ * A known hardware issue where Connect Status Change bit of PORTSC register
+ * of USB1 controller will be set after Port Reset.
+ * We have to clear it in order for later device enumeration to proceed.
+ * This ehci_powerup_fixup overrides the weak function ehci_powerup_fixup
+ * in "ehci-hcd.c".
+ */
+void ehci_powerup_fixup(uint32_t *status_reg, uint32_t *reg)
+{
+	mdelay(50);
+	if (((u32) status_reg & TEGRA_USB_ADDR_MASK) != TEGRA_USB1_BASE)
+		return;
+	/* For EHCI_PS_CSC to be cleared in ehci_hcd.c */
+	if (ehci_readl(status_reg) & EHCI_PS_CSC)
+		*reg |= EHCI_PS_CSC;
+}
 
 /*
  * Create the appropriate control structures to manage