[U-Boot] u-boot driver model

Simon Glass sjg at chromium.org
Thu Oct 22 01:24:47 CEST 2015


+U-Boot

Hi Marco,

On 20 October 2015 at 06:56, Hoefle Marco <Marco.Hoefle at nanotronic.ch> wrote:
>
> Hello Simon,
> I read your presentation regarding driver model for u-boot. We want to port our old u-boot for Microblaze to an up to date one.
> I am struggling with the device tree -> driver model mapping.
>
> The problem is identical to the thread here:
> http://lists.denx.de/pipermail/u-boot/2015-April/210754.html
>  Invalid bus 0 (err=-19)
> *** Warning - spi_flash_probe() failed, using default environment
>
> Unfortunately dm tree crashes so it is hard to debug further:
> U-Boot> dm tree
> Hardware exception at 0x4 address
> Return address from exception 0x8
> Illegal op-code exception
> Unaligned halfword access
> Unaligned load access
> Register R0
> ### ERROR ### Please RESET the board ###
>
> This is the device tree compiled into u-boot:
>
> U-Boot> bdinfo
> DRAM bank   = 0x00000000
> -> start    = 0x80000000
> -> size     = 0x10000000
> flash start    = 0x00000000
> flash size     = 0x00000000
> flash offset   = 0x00000000
> eth0name    = Xelite.40e00000
> ethaddr     = 00:0A:35:02:C7:19
> current eth = Xelite.40e00000
> ip_addr     = 192.168.0.3
> baudrate    = 115200 bps
> relocaddr   = 0x8FF9B000
> reloc off   = 0x00F9B000
> fdt_blob    = 0x8FEB8A7C
> new_fdt     = 0x8FEB8A7C
> fdt_size    = 0x000024A0
>
> U-Boot> fdt addr 0x8FEB8A7C
> U-Boot> fdt print
> / {
>         #address-cells = <0x00000001>;
>         #size-cells = <0x00000001>;
>         aliases {
>                 serial0 = "serial at 40600000";
>                 spi0 = "/axi_quad_spi at 44a00000";
>         };
>         memory {
>                 device_type = "memory";
>                 reg = <0x80000000 0x10000000>;
>         };
>         axi_quad_spi at 44a00000 {
>                 compatible = "xlnx,xps-spi-2.00.a";
>                 reg = <0x44a00000 0x00010000>;
>                 spi-max-frequency = <0x423a35c7>;
>                 #address-cells = <0x00000001>;
>                 #size-cells = <0x00000000>;
>                 status = "okay";
>                 is-dual = <0x00000000>;
>                 num-cs = <0x00000001>;
>                 flash at 0 {
>                         compatible = "n25q128a11";
>                         reg = <0x00000000>;
>                         spi-tx-bus-width = <0x00000001>;
>                         spi-rx-bus-width = <0x00000004>;
>                         spi-max-frequency = <0x0d3ed78e>;
>                         #address-cells = <0x00000001>;
>                         #size-cells = <0x00000001>;
>                         partition at qspi-fsbl-uboot {
>                                 label = "qspi-fsbl-uboot";
>                                 reg = <0x00000000 0x00100000>;
>                         };
>                         partition at qspi-linux {
>                                 label = "qspi-linux";
>                                 reg = <0x00100000 0x00500000>;
>                         };
>                         partition at qspi-device-tree {
>                                 label = "qspi-device-tree";
>                                 reg = <0x00600000 0x00020000>;
>                         };
>                         partition at qspi-rootfs {
>                                 label = "qspi-rootfs";
>                                 reg = <0x00620000 0x005e0000>;
>                         };
>                         partition at qspi-bitstream {
>                                 label = "qspi-bitstream";
>                                 reg = <0x00c00000 0x00400000>;
>                         };
>                 };
>         };
> };
>
>
>
> Do you have any hint on how to debug this further?

Can you try with no SPI and see if you can get 'dm tree' to work? If
that doesn't work it isn't worth trying to figure out SPI.

I suggest starting with what access is happening in 'dm tree' that
causes a problem. You can probably add some debugging to
drivers/core/dump.c.

> Thanks and best regards,
> Marco

Regards,
Simon


More information about the U-Boot mailing list