From mboxrd@z Thu Jan  1 00:00:00 1970
From: Stephen Warren <swarren@wwwdotorg.org>
Date: Wed, 22 Aug 2012 10:06:31 -0600
Subject: [U-Boot] [PATCH 0/3] Tegra 2 USB ULPI series
In-Reply-To: <1345492773.6188.5.camel@selen>
References: <1345392496-28739-1-git-send-email-dev@lynxeye.de>
	<50327D70.5030504@wwwdotorg.org> <1345492773.6188.5.camel@selen>
Message-ID: <50350387.7050809@wwwdotorg.org>
List-Id: <u-boot.lists.denx.de>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: u-boot@lists.denx.de

On 08/20/2012 01:59 PM, Lucas Stach wrote:
> Am Montag, den 20.08.2012, 12:09 -0600 schrieb Stephen Warren:
>> On 08/19/2012 10:08 AM, Lucas Stach wrote:
>>> With this series we are able to initialize USB controllers
>>> using an external ULPI phy AKA USB2 on Tegra 2 devices.
>>>
>>> This was tested to work on a Toradex Colibri T20 board,
>>> where USB2 is used to access the ASIX ethernet chipset.
>>> Testing was done with "tegra20: usb: rework set_host_mode"
>>> applied. I did not spot any regressions on the UTMI ports.
>>>
>>> Patchset is based on top of u-boot-tegra/next
>>
>> Hmmm. I tried testing this on my local branch on Harmony, which has all
>> the patches necessary for multiple USB ports to work, and indeed they do
>> on Seaboard. However, this series (plus some Harmony-specific config and
>> .dts changes) didn't cause the ULPI port to work:-( Any idea why?
>>
> Could you try the attached hack and report back? If it doesn't help
> please build with DEBUG defined in
> arch/arm/cpu/armv7/tegra20/usb.c
> drivers/usb/host/ehci-hcd.c
> and provide me the output of the usb start command.

Oops. I had forgotten to enable ULPI/ULPI_VIEWPORT config options.

Now that I do that, U-Boot hangs when enabling the ULPI transceiver. I
see this with series v1, v1 plus the patch you included in the email I'm
replying to, and v2.

In particular, if I printf every line, the printf before the following
line appears, but not the one after it:

	setbits_le32(&usbctlr->susp_ctrl, ULPI_PHY_ENB);

If I comment that out, then the code runs farther, eventually hanging
inside ulpi_init(). Do you have any idea what's wrong?