[Archive] TomTom v9 Firmware and Maps

[dsa8310]

Originally Posted by dsa8310

Mode (25) should be different from the Self mode, otherwise tests are irrelevant!

If usure what your Self mode is, echo it just before setting it to a different value (immediately after line reading reset_barcelona), with (a) line(s) reading:

echo
echo -n "Self is "
cat /proc/barcelona/modelid
echo

Then check file log in the root of storage.

[dsa8310]

Oh yes, sound=1 means answer with Self to the first reading of modelid (don't use it!); sound=2 means answer with Self to the second reading of modelid; and sound=3 means answer with Self to the third reading of modelid.
Other values won't matter.

[keessie74]

I want to but don't own a Start or XL2 device......

[dsa8310]

Originally Posted by dsa8310

If usure what your Self mode is, echo it just before setting it to a different value (immediately after line reading reset_barcelona), with (a) line(s) reading:

echo
echo -n "Self is "
cat /proc/barcelona/modelid
echo

Then check file log in the root of storage.

After learnig the value of Self, remove its echo from ttn (especially line: cat /proc/barcelona/modelid) !

[dsa8310]

Originally Posted by keessie74

I want to but don't own a Start or XL2 device......

Yes, thanks. Nonetheless you could give another try to your loquendo issue: put sound=0 to avoid any interference (or don't use the sound parameter at all, it's the same, I think).

BTW, code is in file dsa.app/Contents/Resources/Java/firmvare/dsa.c - with lines commented out activated in the root dsa.ko module.

[dsa8310]

Updates:

dsa-23August2012.zip

ttn.zip


dsa-23August2012.zip is identical to dsa-22August2012.zip with the exception of file dsa.c (source).

This ttn.zip has got a better coded dsa.ko, and contains its source dsa.c too.

[dsa8310]

Please DISREGARD links in post #3876 above.
Here are updated links:

dsa-23August2012.zip

ttn.zip


Both have improved dsa.ko (and source).

[keessie74]

Bravo! It's a miracle!!

With dsa-23August2012 you solved the LIVE-subsciption and Loquendo problem on the XL LIVE IQ Routes device!! I patched the device the same way as before with SLF and 36-XL LIVE modes using the 9.430.2-2008.le.CAB-navcore-LE cabinet. I'm just waiting on the LIVE-services password to do some advanced tests. You definitely changed something important in the patching method, but what?

BTW: I didn't set any sound parameters in neither the new dsa.c and ttn. Do you want to test me with the sound parameter?

[keessie74]

This weekend I will test the ttn patching as well on this device (to see or freezing at boot won't happen anymore).

[keessie74]

After entering the LIVE-subscription password both LIVE and Loquendo still work.
Many thanks again....!!

[dsa8310]

I am glad it's sorted out.
But there was no miracle, just a sloppy implementation of the older dsa.ko...

Previously, I was creating extra entries in the proc filesystem which were duplicates of the originals, but writable, which allowed me to overwrite parameters such loquendo, modelid, etc.
My duplicates were placed before the originals and so they were taking precedence over them, and all seemed OK.
Unfortunately, in reality, access to proc might have been optimized, in that the search for a parameter (like loquendo) could start not from the beginning but from the last accessed position, and this could be beyond my parameters; as a result, the return could have had the original value, not patched...

Now, there are no more duplicate entries, because I simply overwrite the read/write functions of the existing, unique entry of each parameter (of interest), with my own functions (actually coded as before) which make the entry writable, aka patchable.

I have also taken care of the eof (end of file) status, updated correctly now (before it was ignored) which has eliminated multiple unnecessary accesses, and helped me to determine that there are only three readings of modelid, etc.

Finally, the original parameter values are copied internally during module initialization, as it should have happened in the first place.

I could infer that there are other devices which would benefit from this dsa update, perhaps in the very same area of the computer voices (loquendo), so dsa-23August2012.zip from post #3877 is a mandatory upgrade.


---
Now, about ttn: it contains an extended dsa.ko kernel module with one parameter, named sound, which works as described in post #3872.
My hope is that by answering with self to the second or third inquiry about the value of modelid, that value would be used for setting out the device sound. It might not be the case at all, but using dsa.ko for patching the sound, instead of HS, would be so nice...
(first modelid reading is for skins, but it might be used for sound too, in which case HS is the only solution - many thanks again to rickgtx for its invaluable contribution!)

[keessie74]

I indeed checked the scan.c source and saw the great disassembly work.

[dsa8310]

Originally Posted by keessie74

I indeed checked the scan.c source and saw the great disassembly work.

Yes, he pointed me in the right direction.

[keessie74]

Are you ready for the next challenge?

[dsa8310]

I just hope to have the sound tested out, conclusively, and I'm done. Again.