fxload to restictive with address range

fxload to restictive with address range

[Kay Sievers]

Hello,
i have a DSL-USB-Modem:
http://www.devolo.de/produkte/adsl/mladslfunusb/index.php

inside is the 6485 Chip from Aanalog Devices:
http://www.analog.com/Analog_Root/productPage/productHome/0%2C2121%2CAD6485%2C00.html

A part of the chip is some kind of ezusb 8051.
(in the same chip is Ethernet and DSL-Pump)

I have extracted the firmware by sniffing Windows USB,
then converted to Intel-Hex-File.

http://vrfy.org/eagle6485.ihx

I've successfull loaded the firmware into the device with a modified fxload.
Unfortunately fxload does not support the address range of this chip.
I changed the following lines and everything is working well.



--- ezusb.c      2002-01-14 19:59:40.000000000 +0100
+++ ezusb.c.orig 2003-05-18 19:56:13.000000000 +0200
@@ -242,8 +242,8 @@
        /* with 8KB RAM, 0x0000-0x1b3f can be written
         * we can't tell if it's a 4KB device here
         */
-       if (off <= 0x1b3f) {
-           if ((off + len) > 0x1b40) {
+       if (off <= 0x7b3f) {
+           if ((off + len) > 0x7b40) {
                fprintf(stderr, need2stage, off);
                return -5;
            } /* else OK */


I expect more devices of this kind to appear in the future.
Maybe somebody can check, how to be prepared for larger RAM sizes.


Have a good day

Kay Sievers


-------------------------------------------------------
This SF.net email is sponsored by: If flattening out C++ or Java
code to make your application fit in a relational database is painful, 
don't do it! Check out ObjectStore. Now part of Progress Software.
http://www.objectstore.net/sourceforge
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

fxload to restictive with address range

[Kay Sievers]

---
Sorry, i diffed a old version. Here is the same mail with new diff.
btw: Are you expecting CR/LF in the record lines?
     It raises "record to short?". Please see patch.
---

Hello,
i have a DSL-USB-Modem:
http://www.devolo.de/produkte/adsl/mladslfunusb/index.php

inside is the 6485 Chip from Aanalog Devices:
http://www.analog.com/Analog_Root/productPage/productHome/0%2C2121%2CAD6485%2C00.html

A part of the chip is some kind of ezusb 8051.
(in the same chip is Ethernet and DSL-Pump)

I have extracted the firmware by sniffing Windows USB,
then converted to Intel-Hex-File.

http://vrfy.org/eagle6485.ihx

I've successfull loaded the firmware into the device with a modified fxload.
Unfortunately fxload does not support the address range of this chip.
I changed the following lines and everything is working well.


--- ezusb.c     2003-05-18 22:15:20.000000000 +0200
+++ ezusb.c.orig        2002-04-12 06:08:01.000000000 +0200
@@ -64,8 +64,8 @@
     /* with 8KB RAM, 0x0000-0x1b3f can be written
      * we can't tell if it's a 4KB device here
      */
-    if (addr <= 0x7b3f)
-     return ((addr + len) > 0x7b40);
+    if (addr <= 0x1b3f)
+     return ((addr + len) > 0x1b40);
 
     /* there may be more RAM; unclear if we can write it.
      * some bulk buffers may be unused, 0x1b3f-0x1f3f
@@ -361,7 +361,7 @@
          return -3;
      }
 
-     if ((len * 2) + 11 > strlen(buf)) {
+     if ((len * 2) + 11 >= strlen(buf)) {
          fprintf (stderr, "record too short?\n");
          return -4;
      }



I expect more devices of this kind to appear in the future.
Maybe somebody can check, how to be prepared for larger internal RAM sizes.


Have a good day

Kay Sievers


-------------------------------------------------------
This SF.net email is sponsored by: If flattening out C++ or Java
code to make your application fit in a relational database is painful, 
don't do it! Check out ObjectStore. Now part of Progress Software.
http://www.objectstore.net/sourceforge
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

Re: fxload to restictive with address range

[David Brownell]

Kay Sievers wrote:
> 
> I've successfull loaded the firmware into the device with a modified fxload.
> Unfortunately fxload does not support the address range of this chip.
> I changed the following lines and everything is working well.

Better to teach it about that particular chip,
rather than break every standard an21 or fx chip;
just add a new "-t type" option.  "fxload" is
100% correct for the devices it claims to support!

I glanced briefly at that analog.com webpage and
couldn't find any technical specs whatever; so it's
not clear how to handle your particular chip.

Or as a workaround, just use the two stage loader.
If it's EZ-USB compatible, "a3load.hex" should work.

- Dave



> --- ezusb.c     2003-05-18 22:15:20.000000000 +0200
> +++ ezusb.c.orig        2002-04-12 06:08:01.000000000 +0200
> @@ -64,8 +64,8 @@
>      /* with 8KB RAM, 0x0000-0x1b3f can be written
>       * we can't tell if it's a 4KB device here
>       */
> -    if (addr <= 0x7b3f)
> -     return ((addr + len) > 0x7b40);
> +    if (addr <= 0x1b3f)
> +     return ((addr + len) > 0x1b40);
>  
>      /* there may be more RAM; unclear if we can write it.
>       * some bulk buffers may be unused, 0x1b3f-0x1f3f
> @@ -361,7 +361,7 @@
>           return -3;
>       }
>  
> -     if ((len * 2) + 11 > strlen(buf)) {
> +     if ((len * 2) + 11 >= strlen(buf)) {
>           fprintf (stderr, "record too short?\n");
>           return -4;
>       }
> 
> 
> 
> I expect more devices of this kind to appear in the future.
> Maybe somebody can check, how to be prepared for larger internal RAM sizes.
> 
> 
> Have a good day
> 
> Kay Sievers
> 
> 





-------------------------------------------------------
This SF.net email is sponsored by: If flattening out C++ or Java
code to make your application fit in a relational database is painful, 
don't do it! Check out ObjectStore. Now part of Progress Software.
http://www.objectstore.net/sourceforge
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

Re: fxload to restictive with address range

[Kay Sievers]

On Sun, May 18, 2003 at 05:17:58PM -0700, David Brownell wrote:
> Kay Sievers wrote:
> >
> >I've successfull loaded the firmware into the device with a modified 
> >fxload.

> Or as a workaround, just use the two stage loader.
> If it's EZ-USB compatible, "a3load.hex" should work.

Yes, you are right.
Two stage loading is possible.
I didn't expected this, cause the Windows driver programms all at once.
So i'm happy now.

Thank you, for your help.
Kay


pim:/home/kay/dslmodem# fxload -v -s /usr/share/usb/a3load.hex -I /usr/share/usb/eagle6485.ihx -D /proc/bus/usb/002/111 
microcontroller type: fx
1st stage:  load 2nd stage loader
open RAM hexfile image /usr/share/usb/a3load.hex
stop CPU
write on-chip, addr 0x0357 len   23 (0x0017)
write on-chip, addr 0x01b5 len  297 (0x0129)
write on-chip, addr 0x0080 len  309 (0x0135)
write on-chip, addr 0x02de len   71 (0x0047)
write on-chip, addr 0x0003 len    3 (0x0003)
write on-chip, addr 0x0325 len   50 (0x0032)
write on-chip, addr 0x0043 len    3 (0x0003)
write on-chip, addr 0x0400 len    4 (0x0004)
write on-chip, addr 0x0000 len    3 (0x0003)
write on-chip, addr 0x036e len   12 (0x000c)
... WROTE: 775 bytes, 10 segments, avg 77
reset CPU
open RAM hexfile image /usr/share/usb/eagle6485.ihx
2nd stage:  write external memory
write external, addr 0x3184 len  695 (0x02b7)
write external, addr 0x343b len  813 (0x032d)
write external, addr 0x3800 len 1010 (0x03f2)
write external, addr 0x3bf2 len  184 (0x00b8)
write external, addr 0x3caa len  646 (0x0286)
write external, addr 0x3000 len  388 (0x0184)
stop CPU
2nd stage:  write on-chip memory
write on-chip, addr 0x109e len  149 (0x0095)
write on-chip, addr 0x09ee len  389 (0x0185)
write on-chip, addr 0x0f15 len  164 (0x00a4)
write on-chip, addr 0x0800 len  494 (0x01ee)
write on-chip, addr 0x0033 len    3 (0x0003)
write on-chip, addr 0x1597 len    4 (0x0004)
write on-chip, addr 0x0063 len    3 (0x0003)
write on-chip, addr 0x1426 len   45 (0x002d)
write on-chip, addr 0x0fb9 len   60 (0x003c)
write on-chip, addr 0x0e6c len  169 (0x00a9)
write on-chip, addr 0x0722 len  198 (0x00c6)
write on-chip, addr 0x1000 len  158 (0x009e)
write on-chip, addr 0x07ff len    1 (0x0001)
write on-chip, addr 0x02e3 len  583 (0x0247)
write on-chip, addr 0x0080 len  611 (0x0263)
write on-chip, addr 0x14a5 len   36 (0x0024)
write on-chip, addr 0x14ed len   34 (0x0022)
write on-chip, addr 0x11bf len  116 (0x0074)
write on-chip, addr 0x13c2 len   50 (0x0032)
write on-chip, addr 0x0ffe len    1 (0x0001)
write on-chip, addr 0x14c9 len   36 (0x0024)
write on-chip, addr 0x0fff len    1 (0x0001)
write on-chip, addr 0x12fe len    2 (0x0002)
write on-chip, addr 0x159b len   13 (0x000d)
write on-chip, addr 0x147f len   38 (0x0026)
write on-chip, addr 0x12c5 len   57 (0x0039)
write on-chip, addr 0x1530 len   30 (0x001e)
write on-chip, addr 0x0ff5 len    8 (0x0008)
write on-chip, addr 0x0b73 len  269 (0x010d)
write on-chip, addr 0x1586 len    9 (0x0009)
write on-chip, addr 0x1233 len   81 (0x0051)
write on-chip, addr 0x052a len  504 (0x01f8)
write on-chip, addr 0x150f len   16 (0x0010)
write on-chip, addr 0x07e8 len   23 (0x0017)
write on-chip, addr 0x158f len    8 (0x0008)
write on-chip, addr 0x1358 len   54 (0x0036)
write on-chip, addr 0x13f4 len   50 (0x0032)
write on-chip, addr 0x138e len   52 (0x0034)
write on-chip, addr 0x004b len    3 (0x0003)
write on-chip, addr 0x0c80 len  262 (0x0106)
write on-chip, addr 0x1453 len   44 (0x002c)
write on-chip, addr 0x1284 len   65 (0x0041)
write on-chip, addr 0x154e len   56 (0x0038)
write on-chip, addr 0x151f len   17 (0x0011)
write on-chip, addr 0x0043 len    3 (0x0003)
write on-chip, addr 0x1300 len   88 (0x0058)
write on-chip, addr 0x0000 len    3 (0x0003)
write on-chip, addr 0x1133 len   12 (0x000c)
write on-chip, addr 0x0d86 len  230 (0x00e6)
write on-chip, addr 0x113f len  128 (0x0080)
write on-chip, addr 0x0ffd len    1 (0x0001)
... WROTE: 9167 bytes, 57 segments, avg 160
reset CPU
pim:/home/kay/dslmodem# 


-------------------------------------------------------
This SF.net email is sponsored by: If flattening out C++ or Java
code to make your application fit in a relational database is painful, 
don't do it! Check out ObjectStore. Now part of Progress Software.
http://www.objectstore.net/sourceforge
_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel