On 2014-08-12 at 15:31:51 +0200, Matthias Brugger [off-list ref] wrote:

2014-08-08 14:28 GMT+02:00 Tobias Klauser [off-list ref]:

quoted

On 2014-08-08 at 13:32:03 +0200, Matthias Brugger [off-list ref] wrote:

quoted

This patch adds support for the UART block found on Mediatek SoCs.
The device has a highspeed register which influences the calcualtion of the
divisor. The chip lacks support for some baudrates. When requested, we set the
divisor to the next smaller baudrate and adjust the c_cflag accordingly.

Signed-off-by: Matthias Brugger <matthias.bgg@gmail.com>
---
 drivers/tty/serial/8250/8250_mtk.c |  296 ++++++++++++++++++++++++++++++++++++
 drivers/tty/serial/8250/Kconfig    |    7 +
 drivers/tty/serial/8250/Makefile   |    1 +
 3 files changed, 304 insertions(+)
 create mode 100644 drivers/tty/serial/8250/8250_mtk.c

diff --git a/drivers/tty/serial/8250/8250_mtk.c b/drivers/tty/serial/8250/8250_mtk.c
new file mode 100644
index 0000000..d63080b
--- /dev/null
+++ b/drivers/tty/serial/8250/8250_mtk.c

@@ -0,0 +1,296 @@

[...]

quoted

+static int mtk8250_probe(struct platform_device *pdev)
+{
+     struct uart_8250_port uart = {};
+     struct resource *regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+     struct resource *irq = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
+     struct mtk8250_data *data;
+     int err;
+
+     if (!regs || !irq) {
+             dev_err(&pdev->dev, "no registers/irq defined\n");
+             return -EINVAL;
+     }
+
+     spin_lock_init(&uart.port.lock);
+     uart.port.mapbase = regs->start;
+     uart.port.irq = irq->start;
+     uart.port.pm = mtk8250_do_pm;
+     uart.port.type = PORT_16550;
+     uart.port.flags = UPF_BOOT_AUTOCONF | UPF_FIXED_PORT;
+     uart.port.dev = &pdev->dev;
+
+     uart.port.membase = devm_ioremap(&pdev->dev, regs->start,
+                                      resource_size(regs));
+     if (!uart.port.membase)
+             return -ENOMEM;

You can use devm_ioremap_resource here and get rid of the check for
!regs above, since devm_ioremap_resource already does that.

        regs = platform_get_resource(pdev, IORESOURCE_MEM, 0);

        ...

        uart.port.membase = devm_ioremap_resource(&pdev->dev, regs);
        if (IS_ERR(uart.port.membase))
                return PTR_ERR(uart.port.membase);

devm_ioremap_resource creates a busy resource region in the
iomem_resource. This leads the UART to silently fail.
I suppose that's why 8250_dw.c uses devm_ioremap instead of
devm_ioremap_resource. The 8250_dw has the same issue.

Ah yes, of course. Sorry about that.

Thanks
Tobias