You are not logged in.
I have been working on a custom DSDT file following the guide at LessWatts :
http://www.lesswatts.org/projects/acpi/ … ngDSDT.php
in conjuction with the Kernel Compilation guide Arch Wiki at:
https://wiki.archlinux.org/index.php/Ke … ild_System
I believe I have a workable DSDT file, however it seems that my custom kernel overrides it:
(from dmesg.log )
0.008216] TOSHIBA Satellite detected - force copy of DSDT to local memory
0.008361] ACPI: Forced DSDT copy: length 0x071F8 copied locally,original unmapped
If I read this properly (below) it seems like all Toshiba Satellites have "forced copy of DSDT to local memory".
https://bugzilla.kernel.org/attachment.cgi?id=31922
The question is : how could I remove the "patch" blacklisting my Toshiba Satellite L300D ?
The kernel is compiled with no errors, installs fine, but on boot the DSDT table from the BIOS is sitting at /sys/firmware/acpi/tables/DSDT and my custom DSDT is nowhere to be seen !
Thanks in advance for any light that can be shone on this !
Note added later: I found the c module refered to in the bugzilla attachment above on my computer at :
/abs/linux/src/linux-3.1/drivers/acpi/bus.c
So now the question is : If I just go into this file, and change the references from "Toshiba Satellite" to "Boshiba Rattellie" ,
and then change the md5 checksum would that work ? ( Answer added later : Yes ! )
Last edited by Ceyx (2012-01-11 20:57:52)
Offline
I gave up on the DSDT thing for that machine, instead I upgraded the BIOS and that solved all the probs. Read my wiki entry here:
https://wiki.archlinux.org/index.php/To … llite_L300
Philosophy is looking for a black cat in a dark room. Metaphysics is looking for a black cat in a dark room that isn't there. Religion is looking for a black cat in a dark room that isn't there and shouting "I found it!". Science is looking for a black cat in a dark room with a flashlight.
Offline
Hey lagagnon,
I did read your wiki many moons ago, it did help. Unfortunately for me, my Toshiba L300D has an ATI video system installed - not an Intel like yours. My BIOS is up to date....the keyboard is remapped and everything works - except for the fan. It is alway on at whatever speed it woke up at, and sometimes if doing CPU intensive things like kernel compilation, it will overheat to the point that the computer just craps out - shuts itself down. Invariably this is just 2 seconds before the compile is complete.
It really isn't a big deal to get the DSDT going, but it is like a stone in my shoe - irritating every once in a while ( for the last 4 years ! ) I also have found that I have learned more about Linux trying to fix the unfixable than I have from machines that work out of the box.
I'll post the results of my research here for anyone else who may have similar problems.
Thanks !
Last edited by Ceyx (2012-01-09 18:59:26)
Offline
What version of BIOS do you have then?
Philosophy is looking for a black cat in a dark room. Metaphysics is looking for a black cat in a dark room that isn't there. Religion is looking for a black cat in a dark room that isn't there and shouting "I found it!". Science is looking for a black cat in a dark room with a flashlight.
Offline
1.9 from Toshiba Canada.
Offline
I honestly think you should flash the latest, which I believe is 2.1 or better.
Philosophy is looking for a black cat in a dark room. Metaphysics is looking for a black cat in a dark room that isn't there. Religion is looking for a black cat in a dark room that isn't there and shouting "I found it!". Science is looking for a black cat in a dark room with a flashlight.
Offline
This is bizarre - I had the latest until 10 days ago when Toshiba posted an update.
My custom dsdt / kernel is compiling as I write this. If it works it may be more interesting to play with. I'll Flash the Bios with the new update and have a look to see if there is any Linux support for it inside.
Thanks for the prod !
Offline
Update : actually I have the latest version for my model ( 1.90) There is something wonky about Toshiba Canada's site.
There is an update on the main page for another similar model ( PSK0QC ) which is 2.10.
Last edited by Ceyx (2012-01-11 04:11:24)
Offline
...
Note added later: I found the c module refered to in the bugzilla attachment above on my computer at :/abs/linux/src/linux-3.1/drivers/acpi/bus.c
So now the question is : If I just go into this file, and change the references from "Toshiba Satellite" to "Boshiba Rattellie" ,
and then change the md5 checksum would that work ? ( Answer added later : Yes ! )
Hello, I have the same problem. I have to recompile the DSDT file because my fan control doesn't work. But on my arch system my Satellite L300D blacklisted, too.
Well, did you get it to work with the changes in the bus.c as you mentioned? If so, then I have to recompile the kernel each time it is upgraded from the arch repos?
Thank you!
Offline
Yes I did get the changes to work in bus.c, and the kernel did compile with my custom DSDT file, with no blacklist.
If you " Set the PACKAGER variable in /etc/makepkg.conf " ( reference https://wiki.archlinux.org/index.php/Ar … kepkg.conf )
then it will not overwrite your kernel.
I got sidetracked from this project, but where I left it :
The DSDT was still missing too many important components when captured on a Linux Machine. I built a Vista System on a spare 2.5 inch drive and swapped it in and out of the Toshiba L300D. There are ways to get the dsdt from Vista using the actual BIOS firmware embedded on the mainboard ( I am sure you know ) and also capturing the additions that MS adds to its Registry. At this point I just sort of left it. but it might be fun to pick it up again.....
How far along are you ?
Offline
I have been battling DSDT on my Toshiba Qosmio 505-887. Is there a benefit from getting it from a running Win install? My DSDT seems almost impossible to recompile though, it starts at the max 200 errors, I have been able to figure out how to solve the first 10ish or so but yet the amount of errors never goes down on each fix. So mine must be super buggy or something. Have you found a good site that discusses fixes for DSDT compile errors? Some of the errors I got on my last try I had no clue about and could not find a thing on the net.
Offline
Yes vwyodajl, I thought there was distict advantages to dissassembly on a Windows installation. From what I can gather, Microsoft sort of fleshes out or pads the skeleton ( buggered ) BIOS using the Registry. In particular, the ACPI:FACS and HPET portions of ACPI in the Regstry are not available under Linux.
The Hackintosh people are all over this, getting OSX to run on Intel machines....
An Italian hacker put together a Java DSDT editor that runs on all OS's and may ( can't recall right now ) access the Registry:
http://olarila.com/forum/viewtopic.php?f=19&t=62
As to your question about any good sites the answer ( for me ) was looking thru GB's of really loosy goosey BS before I found a very small nugget of anything worth while !
This thing really irritates me when I think of it - Stallman is right - "vendor lockin".
Keep me posted !
Last edited by Ceyx (2012-02-18 21:30:05)
Offline
I tried asking for help with it on some toshiba dev forums I found and the responses I got were, it runs on windows and that is the OS it shipped with so whats the problem. LOL
I am planning on trying some more work relating to my DSDT in a few days on my next 39 hours of travel time I have. I will post if I find anything major, going to check out the link you posted and see if getting it from a running windows will be different. I agree finding good info on working with it seems to be very far in between.
Offline