From: marex@denx.de (Marek Vasut)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 09/11] MXS: Add USB PHY driver
Date: Mon, 23 Apr 2012 04:17:12 +0200 [thread overview]
Message-ID: <201204230417.13182.marex@denx.de> (raw)
In-Reply-To: <F281D0F91ED19E4D8E63A7504E8A649803B793D9@039-SN2MPN1-023.039d.mgd.msft.net>
Dear Chen Peter-B29397,
> > +
> > +#define HW_USBPHY_PWD 0x00
> > +
> > +#define HW_USBPHY_CTRL 0x30
> > +#define HW_USBPHY_CTRL_SET 0x34
> > +#define HW_USBPHY_CTRL_CLR 0x38
> > +#define HW_USBPHY_CTRL_TOG 0x3c
> > +
> > +#define BM_USBPHY_CTRL_SFTRST (1 << 31)
> > +#define BM_USBPHY_CTRL_CLKGATE (1 << 30)
> > +#define BM_USBPHY_CTRL_ENUTMILEVEL3 (1 << 15)
> > +#define BM_USBPHY_CTRL_ENUTMILEVEL2 (1 << 14)
> > +#define BM_USBPHY_CTRL_ENIRQDEVPLUGIN (1 << 11)
> > +#define BM_USBPHY_CTRL_ENOTGIDDETECT (1 << 7)
> > +#define BM_USBPHY_CTRL_ENDEVPLUGINDETECT (1 << 4)
> > +#define BM_USBPHY_CTRL_HOSTDISCONDETECT_IRQ (1 << 3)
> > +#define BM_USBPHY_CTRL_ENHOSTDISCONDETECT (1 << 1)
> > +
> > +#define HW_USBPHY_STATUS 0x40
> > +
> > +#define BM_USBPHY_STATUS_OTGID_STATUS (1 << 8)
> > +#define BM_USBPHY_STATUS_DEVPLUGIN_STATUS (1 << 6)
> > +#define BM_USBPHY_STATUS_HOSTDISCON_STATUS (1 << 3)
> > +
> > +/* FIXME */
> > +#define M28EVK_USBOTG_ENABLE MXS_GPIO_NR(3, 12)
> > +
>
> This is platform code, should not put at driver.
>
> > +struct mxs_usb_phy {
> > + struct usb_phy phy;
> > +
> > + struct work_struct work;
> > +
> > + struct clk *clk;
> > + struct resource *mem_res;
> > + void __iomem *mem;
> > + int irq;
> > + int gpio_vbus;
> > + int gpio_vbus_inverted;
> > +
> > + enum usb_otg_state new_state;
> > +
> > + bool host_mode;
> > +};
> > +
> > +static int mxs_usb_set_host(struct usb_otg *otg, struct usb_bus *host)
>
> should be at imx-otg.c
I don't understand why. This is clearly specific to this PHY driver.
>
> > +
> > +static int mxs_usb_set_peripheral(struct usb_otg *otg, struct usb_gadget
> > *gg)
>
> gg may not a good name, how about just gadget?
> should be at imx-otg.c
DTTO
> > +static void mxs_usb_work(struct work_struct *w)
>
> It is OTG switch routine, and should be put imx-otg.c. Sascha also
> suggested before.
Fine by me, but until someone can clearly explain to me why PHY specific code
should be in a shared file (even if it's completely unrelated to it), I see no
point moving this.
> > +{
> > + struct mxs_usb_phy *phy = container_of(w, struct mxs_usb_phy, work);
> > + struct usb_otg *otg = phy->phy.otg;
> > + struct usb_hcd *hcd;
> > +
> > + switch (otg->phy->state) {
> > + case OTG_STATE_A_HOST:
> > + if (phy->new_state == OTG_STATE_UNDEFINED) {
> > + hcd = bus_to_hcd(otg->host);
> > + usb_remove_hcd(hcd);
> > + otg->phy->state = phy->new_state;
> > + /* Turn off VBUS */
> > + gpio_set_value(phy->gpio_vbus,
> > + phy->gpio_vbus_inverted);
> > + }
> > + break;
> > + case OTG_STATE_B_PERIPHERAL:
> > + if (phy->new_state == OTG_STATE_UNDEFINED) {
> > + usb_del_gadget_udc(otg->gadget);
> > + otg->phy->state = phy->new_state;
> > + }
> > + break;
> > + case OTG_STATE_UNDEFINED:
> > + /* Check desired state. */
> > + switch (phy->new_state) {
> > + case OTG_STATE_A_HOST:
> > + if (!otg->host)
> > + break;
> > + otg->phy->state = phy->new_state;
> > +
> > + /* Turn on VBUS */
> > + gpio_set_value(phy->gpio_vbus,
> > + !phy->gpio_vbus_inverted);
> > +
> > + hcd = bus_to_hcd(otg->host);
> > + usb_add_hcd(hcd, hcd->irq, IRQF_SHARED);
> > + break;
> > + case OTG_STATE_B_PERIPHERAL:
> > + if (!otg->gadget)
> > + break;
> > + otg->phy->state = phy->new_state;
> > + usb_add_gadget_udc(phy->phy.dev, otg->gadget);
> > + break;
> > + default:
> > + break;
> > + }
> > + break;
> > +
> > + default:
> > + break;
> > + }
> > +}
> > +
>
> Seems you forget one otg state: OTG_STATE_B_IDLE.
>
> at device mode:
> OTG_STATE_UNDEFINED -> OTG_STATE_B_IDLE (nothing on port)
> OTG_STATE_B_PERIPHERAL -> OTG_STATE_B_IDLE (disconnect with host)
This OTG biz really needs further work. So far, we only support Host/Device.
>
> > +static int mxs_usb_phy_init(struct usb_phy *x)
> > +{
> > + struct mxs_usb_phy *phy = container_of(x, struct mxs_usb_phy, phy);
> > + uint32_t val;
> > +
> > + /* Enable clock to the PHY. */
> > + clk_enable(phy->clk);
> > +
> > + /* Reset the PHY block. */
> > + mxs_reset_block(phy->mem + HW_USBPHY_CTRL);
> > +
> > + /* Power up the PHY. */
> > + writel(0, phy->mem + HW_USBPHY_PWD);
> > +
> > + /* Enable FS/LS compatibility. */
> > + val = BM_USBPHY_CTRL_ENUTMILEVEL2 | BM_USBPHY_CTRL_ENUTMILEVEL3;
> > +
> > + if (phy->host_mode)
> > + val |= BM_USBPHY_CTRL_ENHOSTDISCONDETECT;
> > + else
> > + val |= BM_USBPHY_CTRL_ENIRQDEVPLUGIN |
> > + BM_USBPHY_CTRL_ENOTGIDDETECT |
> > + BM_USBPHY_CTRL_ENDEVPLUGINDETECT;
>
> Please explain the meaning of above code?
>
> > +static void mxs_usb_phy_shutdown(struct usb_phy *x)
> > +{
> > + struct mxs_usb_phy *phy = container_of(x, struct mxs_usb_phy, phy);
> > +
>
> You need to set portsc.phcd=1 and power down phy using set 0xffffffff to
> HW_USBPHY_PWD.
>
> > + /* Gate off the PHY. */
> > + writel(BM_USBPHY_CTRL_CLKGATE, phy->mem + HW_USBPHY_CTRL_SET);
> > +
> > + /* Disable clock to the PHY. */
> > + clk_disable(phy->clk);
> > +}
> > +
> > +static irqreturn_t mxs_phy_irq(int irq, void *data)
> > +{
> > + struct mxs_usb_phy *phy = data;
> > + uint32_t status = readl(phy->mem + HW_USBPHY_STATUS);
> > +
> > + if (status & BM_USBPHY_CTRL_HOSTDISCONDETECT_IRQ)
> > + writel(BM_USBPHY_CTRL_ENHOSTDISCONDETECT,
> > + phy->mem + HW_USBPHY_CTRL_CLR);
> > + else
> > + writel(BM_USBPHY_CTRL_ENHOSTDISCONDETECT,
> > + phy->mem + HW_USBPHY_CTRL_SET);
> > +
>
> Have you tested several connect/disconnect operation with high speed
> device? - where you clear BM_USBPHY_CTRL_HOSTDISCONDETECT_IRQ?
> - This code will always run at every usb interrupt, in fact, high speed
> detector only needs to be set/clear with connect/disconnect.
I have yet to figure out how to actually detect host connect/disconnect in OTG
mode while also detect Device connect/disconnect. Can you elaborate on what
needs to be set in the CTRL register to generate these interrupts?
> > + return IRQ_NONE;
> > +}
> > +
> >
> >
> > +
> > +static int __devexit mxs_phy_remove(struct platform_device *pdev)
> > +{
> > + struct mxs_usb_phy *phy = platform_get_drvdata(pdev);
> > +
> > + /* Stop the PHY work. */
> > + cancel_work_sync(&phy->work);
> > +
> > + /* Power down the PHY. */
> > + mxs_usb_phy_shutdown(&phy->phy);
> > +
> > + /* Remove the transceiver. */
> > + usb_set_transceiver(NULL);
> > +
> > + /* Shut off VBUS. */
> > + gpio_set_value(phy->gpio_vbus, phy->gpio_vbus_inverted);
> > + gpio_free(phy->gpio_vbus);
> > +
> > + /* Free the IRQ requested by this device. */
> > + free_irq(phy->irq, phy);
> > +
> > + /* Free the IRQ requested by this device. */
> > + free_irq(phy->irq, phy);
> > +
>
> Two free_irq???
Rebase issue
>
> > + /* Stop the PHY clock. */
> > + clk_disable_unprepare(phy->clk);
> > + clk_put(phy->clk);
> > +
> > + /* Free the rest. */
> > + platform_set_drvdata(pdev, NULL);
> > +
> > + return 0;
> > +}
> > +
next prev parent reply other threads:[~2012-04-23 2:17 UTC|newest]
Thread overview: 151+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-04-17 10:15 [RFC PATCH 0/8] MXS: Add i.MX28 USB Host driver Marek Vasut
2012-04-17 10:15 ` [PATCH 1/8] MXS: Make clk_disable return integer Marek Vasut
2012-04-17 10:15 ` [PATCH 2/8] MXS: Add USB EHCI and USB PHY clock handling Marek Vasut
2012-04-17 18:01 ` Sascha Hauer
2012-04-17 10:15 ` [PATCH 3/8] MXS: Fixup i.MX233 USB base address name Marek Vasut
2012-04-17 10:15 ` [PATCH 4/8] MXS: Add platform registration hooks for USB EHCI Marek Vasut
2012-04-17 10:15 ` [PATCH 5/8] MXS: Add USB PHY driver Marek Vasut
2012-04-17 17:51 ` Sascha Hauer
2012-04-19 11:40 ` Arnd Bergmann
2012-04-17 10:15 ` [PATCH 6/8] MXS: Add separate MXS EHCI HCD driver Marek Vasut
2012-04-17 10:15 ` [PATCH 7/8] MXS: Enable USB on M28EVK Marek Vasut
2012-04-17 17:40 ` Sergei Shtylyov
2012-04-17 10:15 ` [PATCH 8/8] MXS: Enable USB on MX28EVK Marek Vasut
2012-04-17 10:37 ` [RFC PATCH 0/8] MXS: Add i.MX28 USB Host driver Dirk Behme
2012-04-17 11:24 ` Marek Vasut
2012-04-18 1:00 ` Chen Peter-B29397
2012-04-17 11:29 ` Sascha Hauer
2012-04-17 11:45 ` Marek Vasut
2012-04-17 12:18 ` Sascha Hauer
2012-04-17 20:29 ` Marek Vasut
2012-04-18 7:02 ` Sascha Hauer
2012-04-18 8:05 ` Chen Peter-B29397
2012-04-17 12:36 ` Heikki Krogerus
2012-04-18 4:31 ` Chen Peter-B29397
2012-04-18 7:40 ` Sascha Hauer
2012-04-20 13:10 ` 答复: " Liu JunJie-B08287
2012-04-20 16:25 ` Marek Vasut
2012-04-23 2:02 ` Liu JunJie-B08287
2012-04-18 2:07 ` [RFC PATCH 00/10 V2] " Marek Vasut
2012-04-18 2:07 ` [PATCH 01/10] MXS: Make clk_disable return integer Marek Vasut
2012-04-18 2:07 ` [PATCH 02/10] MXS: Add USB EHCI and USB PHY clock handling Marek Vasut
2012-04-18 2:07 ` [PATCH 03/10] MXS: Fixup i.MX233 USB base address name Marek Vasut
2012-04-18 2:07 ` [PATCH 04/10] MXS: Add data shared between imx-usb, PHY and EHCI driver Marek Vasut
2012-04-18 2:07 ` [PATCH 05/10] MXS: Add platform registration hooks for USB EHCI Marek Vasut
2012-04-18 2:07 ` [PATCH 06/10] MXS: Add composite imx-usb driver Marek Vasut
2012-04-18 2:07 ` [PATCH 07/10] MXS: Add USB PHY driver Marek Vasut
2012-04-18 2:07 ` [PATCH 08/10] MXS: Add separate MXS EHCI HCD driver Marek Vasut
2012-04-18 7:01 ` Chen Peter-B29397
2012-04-18 8:40 ` Sascha Hauer
2012-04-18 9:18 ` Chen Peter-B29397
2012-04-18 9:45 ` Sascha Hauer
2012-04-19 2:54 ` Chen Peter-B29397
2012-04-19 6:34 ` Sascha Hauer
2012-04-18 2:07 ` [PATCH 09/10] MXS: Enable USB on M28EVK Marek Vasut
2012-04-18 12:08 ` Sergei Shtylyov
2012-04-18 12:24 ` Marek Vasut
2012-04-18 2:07 ` [PATCH 10/10] MXS: Enable USB on MX28EVK Marek Vasut
2012-04-18 2:09 ` [RFC PATCH 00/10 V2] MXS: Add i.MX28 USB Host driver Marek Vasut
2012-04-18 7:52 ` Chen Peter-B29397
2012-04-18 17:49 ` Marek Vasut
2012-04-19 2:37 ` Chen Peter-B29397
2012-04-18 17:46 ` [RFC PATCH 00/10 V3] " Marek Vasut
2012-04-18 17:46 ` [PATCH 01/10] MXS: Make clk_disable return integer Marek Vasut
2012-04-18 17:46 ` [PATCH 02/10] MXS: Add USB EHCI and USB PHY clock handling Marek Vasut
2012-04-18 17:46 ` [PATCH 03/10] MXS: Fixup i.MX233 USB base address name Marek Vasut
2012-04-18 17:46 ` [PATCH 04/10] MXS: Add data shared between imx-usb and EHCI driver Marek Vasut
2012-04-18 17:46 ` [PATCH 05/10] MXS: Add platform registration hooks for USB EHCI Marek Vasut
2012-04-19 11:51 ` Arnd Bergmann
2012-04-19 14:24 ` Shawn Guo
2012-04-19 21:32 ` Marek Vasut
2012-04-20 0:40 ` Shawn Guo
2012-04-20 0:56 ` Marek Vasut
2012-04-20 1:34 ` Shawn Guo
2012-04-20 1:40 ` Marek Vasut
2012-04-18 17:46 ` [PATCH 06/10] MXS: Add imx-usb driver Marek Vasut
2012-04-18 17:46 ` [PATCH 07/10] MXS: Add USB PHY driver Marek Vasut
2012-04-18 19:36 ` Sascha Hauer
2012-04-18 20:02 ` Marek Vasut
2012-04-18 20:16 ` Sascha Hauer
2012-04-19 22:06 ` Marek Vasut
2012-04-20 2:35 ` Chen Peter-B29397
2012-04-18 17:46 ` [PATCH 08/10] MXS: Add separate MXS EHCI HCD driver Marek Vasut
2012-04-18 19:50 ` Sascha Hauer
2012-04-18 20:07 ` Marek Vasut
2012-04-18 20:43 ` Sascha Hauer
2012-04-18 17:46 ` [PATCH 09/10] MXS: Enable USB on M28EVK Marek Vasut
2012-04-18 17:46 ` [PATCH 10/10] MXS: Enable USB on MX28EVK Marek Vasut
2012-04-20 2:13 ` [RFC PATCH 00/10 V3] MXS: Add i.MX28 USB Host driver Marek Vasut
2012-04-20 9:48 ` Sascha Hauer
2012-04-20 12:26 ` Peter Chen
2012-04-20 12:44 ` Marek Vasut
2012-04-20 13:00 ` Robert Schwebel
2012-04-20 13:26 ` Sascha Hauer
2012-04-20 16:36 ` Michael Grzeschik
2012-04-21 7:42 ` Peter Chen
2012-04-21 8:17 ` Sascha Hauer
2012-04-21 13:53 ` Subodh Nijsure
2012-04-21 15:37 ` Marek Vasut
2012-04-22 12:59 ` [RFC PATCH 00/11 V4] " Marek Vasut
2012-04-22 12:59 ` [PATCH 01/11] MXS: Make clk_disable return integer Marek Vasut
2012-04-22 12:59 ` [PATCH 02/11] MXS: Add USB EHCI and USB PHY clock handling Marek Vasut
2012-04-22 14:42 ` Shawn Guo
2012-04-22 15:42 ` Marek Vasut
2012-04-22 16:34 ` Shawn Guo
2012-04-22 12:59 ` [PATCH 03/11] MXS: Fixup i.MX233 USB base address name Marek Vasut
2012-04-22 12:59 ` [PATCH 04/11] MXS: Add data shared between imx-otg and EHCI driver Marek Vasut
2012-04-22 14:39 ` Shawn Guo
2012-04-22 15:42 ` Marek Vasut
2012-04-22 16:38 ` Shawn Guo
2012-04-22 12:59 ` [PATCH 05/11] MXS: Modify the ci13xxx_udc to avoid adding UDC Marek Vasut
2012-04-22 12:59 ` [PATCH 06/11] MXS: Add small registration glue for ci13xxx_udc Marek Vasut
2012-04-23 1:46 ` Chen Peter-B29397
2012-04-22 12:59 ` [PATCH 07/11] MXS: Add separate MXS EHCI HCD driver Marek Vasut
2012-04-23 1:44 ` Chen Peter-B29397
2012-04-23 2:11 ` Marek Vasut
2012-04-22 12:59 ` [PATCH 08/11] MXS: Add imx-otg driver Marek Vasut
2012-04-23 6:39 ` Sascha Hauer
2012-04-23 9:38 ` Marek Vasut
2012-04-22 12:59 ` [PATCH 09/11] MXS: Add USB PHY driver Marek Vasut
2012-04-23 1:42 ` Chen Peter-B29397
2012-04-23 2:17 ` Marek Vasut [this message]
2012-04-23 12:20 ` Chen Peter-B29397
2012-04-22 12:59 ` [PATCH 10/11] MXS: Add platform registration hooks for USB EHCI Marek Vasut
2012-04-22 12:59 ` [PATCH 11/11] MXS: Enable USB on M28EVK Marek Vasut
2012-04-23 1:57 ` Chen Peter-B29397
2012-04-23 2:18 ` Marek Vasut
2012-04-23 2:01 ` [RFC PATCH 00/11 V4] MXS: Add i.MX28 USB Host driver Chen Peter-B29397
2012-04-23 2:18 ` Marek Vasut
2012-04-23 11:59 ` Chen Peter-B29397
2012-04-23 12:09 ` Marek Vasut
2012-04-23 12:25 ` Chen Peter-B29397
2012-04-23 12:40 ` Marek Vasut
2012-04-23 2:37 ` Marek Vasut
2012-04-24 3:18 ` [RFC PATCH 00/11 V5] " Marek Vasut
2012-04-24 3:18 ` [PATCH 01/11] MXS: Make clk_disable return integer Marek Vasut
2012-04-24 3:18 ` [PATCH 02/11] MXS: Add USB EHCI and USB PHY clock handling Marek Vasut
2012-04-24 3:18 ` [PATCH 03/11] MXS: Fixup i.MX233 USB base address name Marek Vasut
2012-04-24 3:18 ` [PATCH 04/11] MXS: Add data shared between imx-otg and EHCI driver Marek Vasut
2012-04-24 3:18 ` [PATCH 05/11] MXS: Modify the ci13xxx_udc to avoid adding UDC Marek Vasut
2012-04-29 8:26 ` Chen Peter-B29397
2012-04-24 3:18 ` [PATCH 06/11] MXS: Add small registration glue for ci13xxx_udc Marek Vasut
2012-04-24 3:18 ` [PATCH 07/11] MXS: Add separate MXS EHCI HCD driver Marek Vasut
2012-04-29 8:28 ` Chen Peter-B29397
2012-04-24 3:18 ` [PATCH 08/11] MXS: Add imx-otg driver Marek Vasut
2012-04-24 14:48 ` Lothar Waßmann
2012-04-24 14:50 ` Sascha Hauer
2012-04-24 16:13 ` Lothar Waßmann
2012-04-24 16:47 ` Sascha Hauer
2012-04-24 17:49 ` Marek Vasut
2012-04-24 20:49 ` Sascha Hauer
2012-04-24 20:58 ` Marek Vasut
2012-04-25 0:17 ` Chen Peter-B29397
2012-04-24 3:18 ` [PATCH 09/11] MXS: Add USB PHY driver Marek Vasut
2012-04-24 3:18 ` [PATCH 10/11] MXS: Add platform registration hooks for USB EHCI Marek Vasut
2012-04-24 3:18 ` [PATCH 11/11] MXS: Enable USB on M28EVK Marek Vasut
2012-04-24 20:02 ` [RFC PATCH 00/11 V5] MXS: Add i.MX28 USB Host driver Russell King - ARM Linux
2012-04-25 12:27 ` Marek Vasut
2012-04-25 12:41 ` Russell King - ARM Linux
2012-04-25 12:43 ` Marek Vasut
-- strict thread matches above, loose matches on Subject: below --
2012-04-29 22:34 [RFC PATCH 00/11 V6] " Marek Vasut
2012-04-29 22:34 ` [PATCH 09/11] MXS: Add USB PHY driver Marek Vasut
2012-05-01 1:55 [RFC PATCH 00/11 V7] MXS: Add i.MX28 USB Host driver Marek Vasut
2012-05-01 1:56 ` [PATCH 09/11] MXS: Add USB PHY driver Marek Vasut
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=201204230417.13182.marex@denx.de \
--to=marex@denx.de \
--cc=linux-arm-kernel@lists.infradead.org \
/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;
as well as URLs for NNTP newsgroup(s).