Re: [ts-7000] TS-7250 booting 2.6-git

["Mike Taht" <>]

On 4/27/06, Lennert Buytenhek <> wrote:

On Thu, Apr 27, 2006 at 04:38:34PM +0700, Eko Didik Widianto wrote:

> I have the same problem when loading linux-2.6.16-rc5 with derevo-16
> patch as following:
>
> Using base address 0x00218000 and length 0x0014ba18
> Uncompressing Linux.......................................................................................... done, booting the kernel.
>
> Error: unrecognized/unsupported machine ID (r1 = 0x00000163).
>
> Available machine support:
>
> ID (hex)        NAME
> 000003be        Glomation GESBC-9312-sx
> 000002a1        Technologic Systems TS-72xx SBC
>
> Please check your kernel config and/or bootloader.
>
> So, i check linux/arch/arm/tools/mach-types file, and i note this:
> cx861xx      ARCH_CX861XX           CX861XX                 355 (0x163)
> ts72xx         MACH_TS72XX             TS72XX                  673
>
> Does my TS-7200 redboot need to be updated?

I don't know whether newer TS-7200 Redboots actually pass the right
value (673) or not..

The older one doesn't.

What I did to hack around the bootloader passing a wrong machine ID
was to add a bit of code to arch/arm/boot/compressed/head.S:

Index: linux-2.6.17-rc1/arch/arm/boot/compressed/head.S
===================================================================
--- linux-2.6.17-rc1.orig/arch/arm/boot/compressed/head.S
+++ linux-2.6.17-rc1/arch/arm/boot/compressed/head.S
@@ -825,3 +827,8 @@ reloc_end:
                .align
                .section ".stack", "w"
user_stack:    .space  4096
+
+               .section        ".start", "ax"
+
+               mov     r7, #(673 & 0xff00)
+               orr     r7, r7, #(673 & 0x00ff)

This makes your kernel unsuitable for running on other ep93xx boards,
though.

Easier to (temporarily) flip mach-types.h's definitions of 355 and 673...

Anyway, I've tried every suggestion so far and so far fail to get past
"done, booting the kernel." I pulled down latest git this morning. My total current diff
is below. I note that there are some interesting things going on in the non-git local.diff and
serial-uartbaud.diff patches from svn. (I am under the impression that the TS7250 did slightly odd things with the uart) Quilting those in now....

One other side note is that I'm using the gcc-4.0.1-glibc-2.3.5/arm-unknown-linux-gnu/  compiler with no options. I remember vaguely that using a crunch enabled compiler on the kernel was a bad thing in the bad olddays

diff --git a/drivers/serial/amba-pl010.c b/drivers/serial/amba-pl010.c
index 1631414..b133769 100644
--- a/drivers/serial/amba-pl010.c
+++ b/drivers/serial/amba-pl010.c
@@ -585,7 +585,7 @@ pl010_console_get_options(struct uart_po
 static int __init pl010_console_setup(struct console *co, char *options)
 {
        struct uart_port *port;
-       int baud = 38400;
+       int baud = 115200;
        int bits = 8;
        int parity = 'n';
        int flow = 'n';
diff --git a/kernel/printk.c b/kernel/printk.c
index c056f33..4c6e617 100644
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -507,7 +507,7 @@ asmlinkage int vprintk(const char *fmt,

        /* Emit the output into the temporary buffer */
        printed_len = vscnprintf(printk_buf, sizeof(printk_buf), fmt, args);
-
+       printascii(printk_buf);
        /*
         * Copy the output into log_buf.  If the caller didn't provide
         * appropriate log level tags, we insert them here

-- 
Mike Taht
PostCards From the Bleeding Edge
http://the-edge.blogspot.com

---

YAHOO! GROUPS LINKS

•  Visit your group "ts-7000" on the web.
   
•  To unsubscribe from this group, send an email to:
   =Unsubscribe
   
•  Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.

---