Hi,
On 07-06-18 15:30, Andy Shevchenko wrote:
On Thu, Jun 7, 2018 at 1:38 PM, Hans de Goede <hdegoede@xxxxxxxxxx> wrote:
On some Bay Trail (BYT) systems the firmware does not enable the
ULPI Refclk.
This commit adds a helper which checks and if necessary enabled the Refclk
and calls this helper for BYT machines.
+static void dwc3_pci_enable_ulpi_refclock(struct pci_dev *pci)
+{
+ void __iomem *reg;
+ struct resource res;
+ struct device *dev = &pci->dev;
+ u32 value;
+
+ res.start = pci_resource_start(pci, 1);
+ res.end = pci_resource_end(pci, 1);
+ res.name = "dwc_usb3_bar1";
+ res.flags = IORESOURCE_MEM;
+
+ reg = devm_ioremap_resource(dev, &res);
+ if (IS_ERR(reg)) {
+ dev_err(dev, "cannot check GP_RWREG1 to assert ulpi refclock\n");
+ return;
+ }
I'm not sure I understand what's wrong with simple
pci_iomap() & Co (perhaps pcim_iomap() / pcim_iomap_regions() and others)
Good point, I took this from the crufty Android X86 patched which
Intel maintains here:
https://github.com/01org/ProductionKernelQuilts
And did not realize I could simplify this, will fix for v2.
+
+ value = readl(reg + GP_RWREG1);
+ if (!(value & GP_RWREG1_ULPI_REFCLK_DISABLE))
+ return; /* ULPI refclk already enabled */
+
+ dev_warn(dev, "ULPI refclock is disabled from the BIOS, enabling it\n");
+ value &= ~GP_RWREG1_ULPI_REFCLK_DISABLE;
+ writel(value, reg + GP_RWREG1);
+ msleep(100);
This has to be explained.
Erm, this comes 1:1 from Intels Android X86 patches I've no
idea why it is there, I believe it is better to leave this
uncommented then making something up.
Regards,
Hans
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
