From: Hans de Goede <hdegoede@redhat.com>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH 3/8] dm: usb: Store usb_device parent pointer in usb_device
Date: Fri, 01 May 2015 10:26:25 +0200 [thread overview]
Message-ID: <554338B1.9010504@redhat.com> (raw)
In-Reply-To: <CAPnjgZ1nuws_C1+_kwMt=N=pTnucYnhXgT6=YaDF+VQQyrMSdw@mail.gmail.com>
Hi,
On 01-05-15 06:11, Simon Glass wrote:
> Hi Hans,
>
> On 30 April 2015 at 08:35, Hans de Goede <hdegoede@redhat.com> wrote:
>> When calling into the hcd code in usb_scan_device() we do not yet have
>> the actual udevice for the device we are scanning, so we temporarily set
>> usb_device.dev to the parent.
>>
>> This means that we cannot use usb_device.dev to accurately determine our
>> place in the usb topology when reading descriptors, and this is necessary
>> to correctly program the usb-2 hub tt settings when a usb-1 device is
>> connected to a usb-2 hub.
>>
>> This commit (re)adds a direct parent usb_device pointer to struct usb_device
>> so that the usb hcd code can get to the usb_device parent without needing to
>> go through the dm.
>>
>> This fixes usb-1 devices plugged into usb-2 hubs not working with the driver
>> model.
>>
>> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
>> ---
>> drivers/usb/host/ehci-hcd.c | 24 +-----------------------
>> drivers/usb/host/usb-uclass.c | 5 ++---
>> include/usb.h | 2 +-
>> 3 files changed, 4 insertions(+), 27 deletions(-)
>>
>
> While I agree this is expedient, is there really no other way? How are
> we going to move to driver model if we add back in the
> non-driver-model pointers, etc.? We'll end up with a hybrid approach
> and a real mess.
>
> Is it possible to bug-fix the driver model code (presumably below) instead?
>
That is what I tried first, and failed todo. But after implementing the fix
as posted to the list I've been thinking about it some more and I think it should
be possible so I'll try again.
Regards,
Hans
>> diff --git a/drivers/usb/host/ehci-hcd.c b/drivers/usb/host/ehci-hcd.c
>> index 19f1e29..00c038c 100644
>> --- a/drivers/usb/host/ehci-hcd.c
>> +++ b/drivers/usb/host/ehci-hcd.c
>> @@ -292,7 +292,6 @@ static void ehci_update_endpt2_dev_n_port(struct usb_device *udev,
>> struct QH *qh)
>> {
>> struct usb_device *ttdev;
>> - int parent_devnum;
>>
>> if (udev->speed != USB_SPEED_LOW && udev->speed != USB_SPEED_FULL)
>> return;
>> @@ -302,35 +301,14 @@ static void ehci_update_endpt2_dev_n_port(struct usb_device *udev,
>> * the tt, so of the first upstream usb-2 hub, there may be usb-1 hubs
>> * in the tree before that one!
>> */
>> -#ifdef CONFIG_DM_USB
>> - struct udevice *parent;
>> -
>> - for (ttdev = udev; ; ) {
>> - struct udevice *dev = ttdev->dev;
>> -
>> - if (dev->parent &&
>> - device_get_uclass_id(dev->parent) == UCLASS_USB_HUB)
>> - parent = dev->parent;
>> - else
>> - parent = NULL;
>> - if (!parent)
>> - return;
>> - ttdev = dev_get_parentdata(parent);
>> - if (!ttdev->speed != USB_SPEED_HIGH)
>> - break;
>> - }
>> - parent_devnum = ttdev->devnum;
>> -#else
>> ttdev = udev;
>> while (ttdev->parent && ttdev->parent->speed != USB_SPEED_HIGH)
>> ttdev = ttdev->parent;
>> if (!ttdev->parent)
>> return;
>> - parent_devnum = ttdev->parent->devnum;
>> -#endif
>>
>> qh->qh_endpt2 |= cpu_to_hc32(QH_ENDPT2_PORTNUM(ttdev->portnr) |
>> - QH_ENDPT2_HUBADDR(parent_devnum));
>> + QH_ENDPT2_HUBADDR(ttdev->parent->devnum));
>> }
>>
>> static int
>> diff --git a/drivers/usb/host/usb-uclass.c b/drivers/usb/host/usb-uclass.c
>> index 95e371d..0157bc6 100644
>> --- a/drivers/usb/host/usb-uclass.c
>> +++ b/drivers/usb/host/usb-uclass.c
>> @@ -470,7 +470,6 @@ int usb_scan_device(struct udevice *parent, int port,
>> bool created = false;
>> struct usb_dev_platdata *plat;
>> struct usb_bus_priv *priv;
>> - struct usb_device *parent_udev;
>> int ret;
>> ALLOC_CACHE_ALIGN_BUFFER(struct usb_device, udev, 1);
>> struct usb_interface_descriptor *iface = &udev->config.if_desc[0].desc;
>> @@ -515,9 +514,9 @@ int usb_scan_device(struct udevice *parent, int port,
>> udev->devnum = priv->next_addr + 1;
>> udev->portnr = port;
>> debug("Calling usb_setup_device(), portnr=%d\n", udev->portnr);
>> - parent_udev = device_get_uclass_id(parent) == UCLASS_USB_HUB ?
>> + udev->parent = device_get_uclass_id(parent) == UCLASS_USB_HUB ?
>> dev_get_parentdata(parent) : NULL;
>> - ret = usb_setup_device(udev, priv->desc_before_addr, parent_udev, port);
>> + ret = usb_setup_device(udev, priv->desc_before_addr, udev->parent, port);
>> debug("read_descriptor for '%s': ret=%d\n", parent->name, ret);
>> if (ret)
>> return ret;
>> diff --git a/include/usb.h b/include/usb.h
>> index 1b667ff..7876df4 100644
>> --- a/include/usb.h
>> +++ b/include/usb.h
>> @@ -141,9 +141,9 @@ struct usb_device {
>> int act_len; /* transfered bytes */
>> int maxchild; /* Number of ports if hub */
>> int portnr; /* Port number, 1=first */
>> -#ifndef CONFIG_DM_USB
>> /* parent hub, or NULL if this is the root hub */
>> struct usb_device *parent;
>> +#ifndef CONFIG_DM_USB
>> struct usb_device *children[USB_MAXCHILDREN];
>> void *controller; /* hardware controller private data */
>> #endif
>> --
>> 2.3.6
>>
next prev parent reply other threads:[~2015-05-01 8:26 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-30 14:35 [U-Boot] [PATCH 0/8] usb: driver-model fixes and dm support sunxi-ehci.c Hans de Goede
2015-04-30 14:35 ` [U-Boot] [PATCH 1/8] dm: usb: Copy over usb_device values from usb_scan_device() to final usb_device Hans de Goede
2015-05-01 4:11 ` Simon Glass
2015-05-01 7:21 ` Hans de Goede
2015-05-01 8:03 ` Hans de Goede
2015-04-30 14:35 ` [U-Boot] [PATCH 2/8] dm: usb: Use controller_dev in dm ehci code Hans de Goede
2015-05-01 4:11 ` Simon Glass
2015-05-01 8:03 ` Hans de Goede
2015-04-30 14:35 ` [U-Boot] [PATCH 3/8] dm: usb: Store usb_device parent pointer in usb_device Hans de Goede
2015-05-01 4:11 ` Simon Glass
2015-05-01 8:26 ` Hans de Goede [this message]
2015-04-30 14:35 ` [U-Boot] [PATCH 4/8] dm: usb: Set desc_before_addr from ehci dm code Hans de Goede
2015-05-01 4:12 ` Simon Glass
2015-04-30 14:35 ` [U-Boot] [PATCH 5/8] dm: usb: Add support for interrupt queues to the dm usb code Hans de Goede
2015-05-01 4:12 ` Simon Glass
2015-04-30 14:35 ` [U-Boot] [PATCH 6/8] dm: usb: Prefix ehci interrupt-queue functions with _ehci_ Hans de Goede
2015-05-01 4:12 ` Simon Glass
2015-04-30 14:35 ` [U-Boot] [PATCH 7/8] dm: usb: Add support for interrupt queues to the dm ehci code Hans de Goede
2015-05-01 4:12 ` Simon Glass
2015-04-30 14:35 ` [U-Boot] [PATCH 8/8] sunxi: ehci: Convert to the driver-model Hans de Goede
2015-05-01 4:12 ` Simon Glass
2015-05-01 8:37 ` Hans de Goede
2015-05-02 14:03 ` Ian Campbell
2015-05-02 14:04 ` Ian Campbell
2015-05-04 14:16 ` Hans de Goede
2015-04-30 14:39 ` [U-Boot] [PATCH 0/8] usb: driver-model fixes and dm support sunxi-ehci.c Hans de Goede
2015-04-30 14:48 ` Simon Glass
2015-04-30 19:38 ` Hans de Goede
2015-04-30 22:04 ` Simon Glass
2015-05-01 7:17 ` Hans de Goede
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=554338B1.9010504@redhat.com \
--to=hdegoede@redhat.com \
--cc=u-boot@lists.denx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox