Minor updates
git-svn-id: svn://svn.rockbox.org/rockbox/trunk@54 a1c6a512-1295-4272-9138-f99709370657
This commit is contained in:
parent
5c983fc7c5
commit
68ee35c679
2 changed files with 20 additions and 19 deletions
22
www/main.t
22
www/main.t
|
@ -113,9 +113,11 @@ To subscribe, send a message to <a href="mailto:majordomo@cool.haxx.se">majordom
|
||||||
|
|
||||||
<h2>About the hardware</h2>
|
<h2>About the hardware</h2>
|
||||||
|
|
||||||
<p>I wrote a <a href="/isd200/archos.html">"dissection" page</a> some months ago,
|
<p>I wrote a <a href="internals/bjorn.html">"dissection" page</a> some months ago,
|
||||||
showing the inside of the Archos and listing the main components.
|
showing the inside of the Archos and listing the main components.
|
||||||
I have also collected a couple of <a href="docs/">data sheets</a>.
|
I have also collected a couple of <a href="docs/">data sheets</a>.
|
||||||
|
Also, don't miss the <a href="notes.html">research notes</a>
|
||||||
|
from my reverse-engineering the firmware.
|
||||||
|
|
||||||
<h2>About the software</h2>
|
<h2>About the software</h2>
|
||||||
|
|
||||||
|
@ -125,24 +127,6 @@ The first thing this version does after boot is to look for a file called
|
||||||
If it exists, it is loaded into RAM and started.
|
If it exists, it is loaded into RAM and started.
|
||||||
This is how firmware upgrades are loaded.
|
This is how firmware upgrades are loaded.
|
||||||
|
|
||||||
<h3>File format</h3>
|
|
||||||
<p>The archos.mod file is scrambled, but luckily not using encryption.
|
|
||||||
|
|
||||||
<p>Each data byte is inverted and ROLed 1 bit.
|
|
||||||
The data is then spread over four memory segments. The two least significant bits of the address is used as segment number and the rest as offset in the segment. So, basically:
|
|
||||||
|
|
||||||
<ul>
|
|
||||||
<li>segment number = address % 4
|
|
||||||
<li>segment offset = address / 4
|
|
||||||
<li>segment length = imgsize / 4
|
|
||||||
</ul>
|
|
||||||
|
|
||||||
<p>A 6-byte header is added to the beginning of the scrambled image:
|
|
||||||
<ul>
|
|
||||||
<li>32 bit length (big-endian)
|
|
||||||
<li>16 bit checksum
|
|
||||||
</ul>
|
|
||||||
|
|
||||||
<h2>Dreams</h2>
|
<h2>Dreams</h2>
|
||||||
<p>Ok, forget about reality, what could we do with this?
|
<p>Ok, forget about reality, what could we do with this?
|
||||||
|
|
||||||
|
|
17
www/tools.t
17
www/tools.t
|
@ -3,6 +3,23 @@
|
||||||
|
|
||||||
<h2>Descrambler / Scrambler</h2>
|
<h2>Descrambler / Scrambler</h2>
|
||||||
|
|
||||||
|
<p>The archos.mod file is scrambled, but luckily not using encryption.
|
||||||
|
|
||||||
|
<p>Each data byte is inverted and ROLed 1 bit.
|
||||||
|
The data is then spread over four memory segments. The two least significant bits of the address is used as segment number and the rest as offset in the segment. So, basically:
|
||||||
|
|
||||||
|
<ul>
|
||||||
|
<li>segment number = address % 4
|
||||||
|
<li>segment offset = address / 4
|
||||||
|
<li>segment length = imgsize / 4
|
||||||
|
</ul>
|
||||||
|
|
||||||
|
<p>A 6-byte header is added to the beginning of the scrambled image:
|
||||||
|
<ul>
|
||||||
|
<li>32 bit length (big-endian)
|
||||||
|
<li>16 bit checksum
|
||||||
|
</ul>
|
||||||
|
|
||||||
<p>I've written a small utility to descramble the firmware files:
|
<p>I've written a small utility to descramble the firmware files:
|
||||||
<ul>
|
<ul>
|
||||||
<li><a href="descramble.c">descramble.c</a> - 1835 bytes - The source code (pure ANSI C, should work everywhere). GPL licensed.
|
<li><a href="descramble.c">descramble.c</a> - 1835 bytes - The source code (pure ANSI C, should work everywhere). GPL licensed.
|
||||||
|
|
Loading…
Reference in a new issue