Re: Linux acpi compatibility / iasl compiler

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



2009/9/30 Moore, Robert <robert.moore@xxxxxxxxx>:
> What exactly was the repair/change that you made to the DSDT?
>
>
> -----Original Message-----
> From: Joshua C. [mailto:joshuacov@xxxxxxxxxxxxxx]
> Sent: Wednesday, September 30, 2009 6:09 AM
> To: Moore, Robert
> Cc: Lin, Ming M; linux-acpi@xxxxxxxxxxxxxxx
> Subject: Re: Linux acpi compatibility / iasl compiler
>
> 2009/9/25 Joshua C. <joshuacov@xxxxxxxxxxxxxx>:
>> 2009/9/25 Moore, Robert <robert.moore@xxxxxxxxx>:
>>>
>>>> 1. Can I substitute the original dsdt (compiled with ms asl 2) with
>>>> the one compiled and optimazed with Intel iasl?
>>>>
>>>
>>> Yes. It should work the same as the original.
>>>
>>>> 2. Is there such a "big" diffference between the code produced by
>>>> these apps?
>>>
>>> iASL is an optimizing compiler. It will optimize all namepaths to their minimum lengths, it will minimize integer sizes, and will perform constant folding. These optimizations can shrink the code considerably.
>>>
>>>
>>>> 3. How should windows react to the new dsdt recompiled with intel's
>>>> iasl?
>>>
>>> The DSDT should perform identically. It has simply been optimized by iASL.
>>> I can't speak to the other ACPI tables on your machine (although, usually any other ASL/AML tables are SSDTs)
>>>
>>>
>>> 4) You should let the kernel override the DSDT with your DSDT before you even think about flashing the bios.
>>>
>>  I tried it. Linux works perfectly. As for windows:
>> -win7 doesn't show any visible signs of problems. I did my best to
>> load the table in the registry. I even tried it without UAC.It looks
>> like it works with the optimazed dsdt.
>> However this is closed-source acpi driver and maybe it still reads
>> (from the bios) the original dsdt (the ms one). Therefore I'm not that
>> sure about it.
>> -winXP: after loading the new optimized dsdt in the registry and
>> restarting the pc almost nothing works. random shutdowns with bsod and
>> some of the hardware isn't even seen! Only a new install is the
>> solution.
>> I know that the winxp acpi driver should be rather old but I didn't
>> expect such a downturn.
>>
>> To sum up:
>>
>> Linux works great, but from what i saw I can't say that windows (at
>> least xp, not sure for 7) likes the iasl compiled code. This once
>> again proved the "obvious" luck of standards in windows.
>>
>> Thanx to Robert Moore for the reply. I'll test some more times with
>> different iasl options before finally flashing the modded bios.
>>
>
> I flashed the bios with the recompiled DSDT (with iasl-20090903,
> optimized) table some days ago. Here are the results:
>
> Linux:  No problems at all.Works great and nothing extra should be done. Great.
>
> Windows7: It didn't like it the first time. After flashing the bios
> wouldn't boot so I needed to reinstall it. After this it works fine.
> The whole story makes me think that windows caches the dsdt in the
> registry and compares it with the one provided by the bios during
> boot. On mismatch it won't boot.
>
> WindowsXPSP3: It doesn't like intel compiler's optimazations. I tried
> all options (-of -oi -on) but I wouldn't boot. It coulnd't be
> installed either. Neither does it like the MS asl v.3. Obviously their
> acpi driver is quite old. I coundn't test the compatibility mode of
> iasl because the compiled dsdt was 13 bytes bigger than the available
> space (in the bios file). Therefore I cannot comment on this but I bet
> it wouldn't like it either.
>
> Anyway, I'm back to linux and it has no problems with my new (repaired) DSDT.
>

There is a broken/missing object in the original dsdt (\SB_.PSHR).
Take a look at kernel bug #10237. There you'll find 2 dsdt tables. The
first on is from ykzhao (uploaded on 2008-05-09 03:09) and the second
one is the one I'm using right now ("my custom DSDT" uploaded on
2008-05-12 07:06). I just added the missing PSHR object
(\SB_.PCI0.LPC0.PHSR (0x0D, 0x00), TJ85) and fixed the warnungs.
--
To unsubscribe from this list: send the line "unsubscribe linux-acpi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux IBM ACPI]     [Linux Power Management]     [Linux Kernel]     [Linux Laptop]     [Kernel Newbies]     [Share Photos]     [Security]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Samba]     [Video 4 Linux]     [Device Mapper]     [Linux Resources]

  Powered by Linux