The error message:
Windows - Registry Recovery
One of the files containing the system's registry data had to be
recovered by use of a log or alternate copy. The recovery was
successful.

The system:
OS, Windows XP SP3 with patches
motherboard, Gigabyte GA-EP45-UD3L
CPU, Intel Q9550

The apparent cause:
BIOS settings
C2/C2E State Support
C4/C4E State Support

Enabling the subsetting (C4/C4E State Support) obviously corrupts
the Windows registry. Using Performance Monitor, the graph for
Physical Disk Idle Time tracks at about 30% instead of 100% when
the system is idle.

Enabling the prior setting (C2/C2E State Support) is not so
obvious (unless there is somewhere to look for the sign), but it
is deadly on my system. Enabling it apparently corrupts the
registry and/or other files and eventually can lead to much
crashing.

Reproducing the error or telling that the corruption exists has
been done by removing Device Manager IDE controllers and then
rebooting.

John Doe wrote:

> The error message:
> Windows - Registry Recovery
> One of the files containing the system's registry data had to be
> recovered by use of a log or alternate copy. The recovery was
> successful.
>
> The system:
> OS, Windows XP SP3 with patches
> motherboard, Gigabyte GA-EP45-UD3L
> CPU, Intel Q9550
>
> The apparent cause:
> BIOS settings
> C2/C2E State Support
> C4/C4E State Support
>
> Enabling the subsetting (C4/C4E State Support) obviously corrupts
> the Windows registry. Using Performance Monitor, the graph for
> Physical Disk Idle Time tracks at about 30% instead of 100% when
> the system is idle.
>
> Enabling the prior setting (C2/C2E State Support) is not so
> obvious (unless there is somewhere to look for the sign), but it
> is deadly on my system. Enabling it apparently corrupts the
> registry and/or other files and eventually can lead to much
> crashing.
>
> Reproducing the error or telling that the corruption exists has
> been done by removing Device Manager IDE controllers and then
> rebooting.

Don't use low-power (sleep) modes defined that did not exist or were not
supported at the time the OS was developed and originally released.

You didn't bother to mention WHERE in the BIOS (screen, settings group,
etc) you are changing these settings. I had to GUESS that you were
talking about the sleep mode and the processor states available in the
selected sleep mode. I don't even see the C4 power state defined in the
ACPI 4.0a spec that was released April 5, 2010 (See Processor Power
States, page 307). C3 is the max specified there.

http://www.acpi.info/DOWNLOADS/ACPIspec40a.pdf
See "Processor Power States", page 307.

Page 312:
8.1.5 Additional Processor Power States
ACPI introduced optional processor power states beyond C3 starting in
ACPI 2.0. These power states, C4… Cn, are conveyed to OSPM through the
_CST object defined in section 8.4.2.1, “_CST (C-States).” These
additional power states are characterized by equivalent operational
semantics to the C1 through C3 power states, as defined in the
previous sections, but with different entry/exit latencies and power
savings. See section 8.4.2.1, “_CST (C-States),” for more
information.

So what was the actual *default* value for this BIOS setting? Why did
you change it? Why are you playing with BIOS settings (to change away
from the defaults) that you don't understand?

VanguardLH <V nguard.LH> wrote:

> John Doe wrote:
>
>> The error message:
>> Windows - Registry Recovery
>> One of the files containing the system's registry data had to be
>> recovered by use of a log or alternate copy. The recovery was
>> successful.
>>
>> The system:
>> OS, Windows XP SP3 with patches
>> motherboard, Gigabyte GA-EP45-UD3L
>> CPU, Intel Q9550
>>
>> The apparent cause:
>> BIOS settings
>> C2/C2E State Support
>> C4/C4E State Support
>>
>> Enabling the subsetting (C4/C4E State Support) obviously
>> corrupts the Windows registry. Using Performance Monitor, the
>> graph for Physical Disk Idle Time tracks at about 30% instead
>> of 100% when the system is idle.
>>
>> Enabling the prior setting (C2/C2E State Support) is not so
>> obvious (unless there is somewhere to look for the sign), but
>> it is deadly on my system. Enabling it apparently corrupts the
>> registry and/or other files and eventually can lead to much
>> crashing.
>>
>> Reproducing the error or telling that the corruption exists has
>> been done by removing Device Manager IDE controllers and then
>> rebooting.
>
> Don't use low-power (sleep) modes defined that did not exist or
> were not supported at the time the OS was developed and
> originally released.
>
> You didn't bother to mention WHERE in the BIOS (screen, settings
> group, etc) you are changing these settings.

Do you have a Gigabyte GA-EP45-UD3L motherboard with an Intel
Q9550 CPU, MouthGuard? You shouldn't bother trying to figure it
out unless it matters to you. My post is meant for someone looking
for information on the subject.

> I had to GUESS that you were talking about the sleep mode and
> the processor states available in the selected sleep mode.

You would not have to guess if you had the same system and were
experiencing the same problem, MouthGuard, if you had the
motherboard manual, and if you knew how to do a search with
Ctrl+F.

> I don't even see the C4 power state defined in the ACPI 4.0a
> spec that was released April 5, 2010 (See Processor Power
> States, page 307). C3 is the max specified there.

Maybe the C4/C4E State Support BIOS setting really is from hell...

> So what was the actual *default* value for this BIOS setting?

Download the motherboard manual and see for yourself, MouthGuard.

> Why did you change it? Why are you playing with BIOS settings
> (to change away from the defaults) that you don't understand?

Because it was there? What difference does it make to you,
MouthGuard?

The reason behind getting into trouble is irrelevant at this time.
The problem has apparently been resolved. I posted for the
conceivable benefit of anyone else who might have taken the same
path. Posting a solution is a well accepted practice here on
UseNet in technical discussion groups, at least as common as
asking questions.
--

















> Path: news.astraweb.com!border1.newsrouter.astraweb.com!nntp.club.cc.cmu.edu!feeder.erje.net!news2.arglkargh.de!news.albasani.net!not-for-mail
> From: VanguardLH <V nguard.LH>
> Newsgroups: microsoft.public.windowsxp.general
> Subject: Re: >> C2/C2E State Support
> Date: Sun, 14 Nov 2010 15:12:44 -0600
> Organization: 0.35MOA sniper at loose in Usenet
> Lines: 60
> Message-ID: <ibpjb8$ii4$1 news.albasani.net>
> References: <4cdf6681$0$18677$c3e8da3$38634283 news.astraweb.com>
> Mime-Version: 1.0
> Content-Type: text/plain; charset="utf-8"
> Content-Transfer-Encoding: 8bit
> X-Trace: news.albasani.net +7hIWj8uMicvtd6h+cJL79RFWQrz63j8HYz2p0HoQ9LMVrbtwnuChFXle8YYEpNbH3nrzqbhaQOUD6pq3hH8APfwn0LtMBN8fpCRSKbIbvSa7PXT6BluAqMp5Lpuytrl
> NNTP-Posting-Date: Sun, 14 Nov 2010 21:12:08 +0000 (UTC)
> Keywords: VanguardLH VLH811
> Cancel-Lock: sha1:L1Kg2jr/mBXr7EXTw+5yT8wb0cI=
> User-Agent: 40tude_Dialog/2.0.15.41
> Injection-Info: news.albasani.net; logging-data="BlZ/aABq82hPht5DgH7PekVtL5WKWuabVwflhjMF8Vimf7qgPMyKnyzzSdz+/YLPpSOloAuOMdg4o5uP7HuVqkLHga+Wduk/SHMxycWGkaaHXSICvMf0ENj3/he8prEI"; mail-complaints-to="abuse albasani.net"
> X-MOTD: It is wiser to remain silent and let others ponder your ignorance than to speak and remove all doubt
>

John Doe wrote:

> VanguardLH <V nguard.LH> wrote:
>
>> John Doe wrote:
>>
>>> The error message:
>>> Windows - Registry Recovery
>>> One of the files containing the system's registry data had to be
>>> recovered by use of a log or alternate copy. The recovery was
>>> successful.
>>>
>>> The system:
>>> OS, Windows XP SP3 with patches
>>> motherboard, Gigabyte GA-EP45-UD3L
>>> CPU, Intel Q9550
>>>
>>> The apparent cause:
>>> BIOS settings
>>> C2/C2E State Support
>>> C4/C4E State Support
>>>
>>> Enabling the subsetting (C4/C4E State Support) obviously
>>> corrupts the Windows registry. Using Performance Monitor, the
>>> graph for Physical Disk Idle Time tracks at about 30% instead
>>> of 100% when the system is idle.
>>>
>>> Enabling the prior setting (C2/C2E State Support) is not so
>>> obvious (unless there is somewhere to look for the sign), but
>>> it is deadly on my system. Enabling it apparently corrupts the
>>> registry and/or other files and eventually can lead to much
>>> crashing.
>>>
>>> Reproducing the error or telling that the corruption exists has
>>> been done by removing Device Manager IDE controllers and then
>>> rebooting.
>>
>> Don't use low-power (sleep) modes defined that did not exist or
>> were not supported at the time the OS was developed and
>> originally released.
>>
>> You didn't bother to mention WHERE in the BIOS (screen, settings
>> group, etc) you are changing these settings.
>
> Do you have a Gigabyte GA-EP45-UD3L motherboard with an Intel
> Q9550 CPU, MouthGuard? You shouldn't bother trying to figure it
> out unless it matters to you. My post is meant for someone looking
> for information on the subject.
>
>> I had to GUESS that you were talking about the sleep mode and
>> the processor states available in the selected sleep mode.
>
> You would not have to guess if you had the same system and were
> experiencing the same problem, MouthGuard, if you had the
> motherboard manual, and if you knew how to do a search with
> Ctrl+F.
>
>> I don't even see the C4 power state defined in the ACPI 4.0a
>> spec that was released April 5, 2010 (See Processor Power
>> States, page 307). C3 is the max specified there.
>
> Maybe the C4/C4E State Support BIOS setting really is from hell...
>
>> So what was the actual *default* value for this BIOS setting?
>
> Download the motherboard manual and see for yourself, MouthGuard.
>
>> Why did you change it? Why are you playing with BIOS settings
>> (to change away from the defaults) that you don't understand?
>
> Because it was there? What difference does it make to you,
> MouthGuard?
>
> The reason behind getting into trouble is irrelevant at this time.
> The problem has apparently been resolved. I posted for the
> conceivable benefit of anyone else who might have taken the same
> path. Posting a solution is a well accepted practice here on
> UseNet in technical discussion groups, at least as common as
> asking questions.

Oh, so we are to be thrilled that you choose to share you daily diary
entry here in Usenet. You didn't want help. You definitely don't seem
to be providing any to anyone specific. So you just spewed out your
diary entry. I guess I was wrong that a child wouldn't be putzing
around in the BIOS settings but your response has definitely proven you
are a child.

Like a troll, you thought that I'd miss that you added unrelated
newsgroups to your reply and I'd be posting to somewhere else. No
thanks.

Oh yes, your post was ever so helpful to others using the same mobo and
same OS. Uh huh, like your guessing that the C4 CPU state caused the
problem was laced with so much good evidence ... not! Yeah, figuring
you're screwing with other BIOS settings you don't understand, that you
are having problems with your host really isn't a surprise.

VanguardLH <V nguard.LH> wrote:

> John Doe wrote:
>> VanguardLH <V nguard.LH> wrote:
>>> John Doe wrote:
>>>
>>>> The error message:
>>>> Windows - Registry Recovery
>>>> One of the files containing the system's registry data had to be
>>>> recovered by use of a log or alternate copy. The recovery was
>>>> successful.
>>>>
>>>> The system:
>>>> OS, Windows XP SP3 with patches
>>>> motherboard, Gigabyte GA-EP45-UD3L
>>>> CPU, Intel Q9550
>>>>
>>>> The apparent cause:
>>>> BIOS settings
>>>> C2/C2E State Support
>>>> C4/C4E State Support
>>>>
>>>> Enabling the subsetting (C4/C4E State Support) obviously
>>>> corrupts the Windows registry. Using Performance Monitor, the
>>>> graph for Physical Disk Idle Time tracks at about 30% instead
>>>> of 100% when the system is idle.
>>>>
>>>> Enabling the prior setting (C2/C2E State Support) is not so
>>>> obvious (unless there is somewhere to look for the sign), but
>>>> it is deadly on my system. Enabling it apparently corrupts
>>>> the registry and/or other files and eventually can lead to
>>>> much crashing.
>>>>
>>>> Reproducing the error or telling that the corruption exists
>>>> has been done by removing Device Manager IDE controllers and
>>>> then rebooting.
>>>
>>> Don't use low-power (sleep) modes defined that did not exist
>>> or were not supported at the time the OS was developed and
>>> originally released.
>>>
>>> You didn't bother to mention WHERE in the BIOS (screen,
>>> settings group, etc) you are changing these settings.
>>
>> Do you have a Gigabyte GA-EP45-UD3L motherboard with an Intel
>> Q9550 CPU, MouthGuard? You shouldn't bother trying to figure it
>> out unless it matters to you. My post is meant for someone
>> looking for information on the subject.
>>
>>> I had to GUESS that you were talking about the sleep mode and
>>> the processor states available in the selected sleep mode.
>>
>> You would not have to guess if you had the same system and were
>> experiencing the same problem, MouthGuard, if you had the
>> motherboard manual, and if you knew how to do a search with
>> Ctrl+F.
>>
>>> I don't even see the C4 power state defined in the ACPI 4.0a
>>> spec that was released April 5, 2010 (See Processor Power
>>> States, page 307). C3 is the max specified there.
>>
>> Maybe the C4/C4E State Support BIOS setting really is from
>> hell...
>>
>>> So what was the actual *default* value for this BIOS setting?
>>
>> Download the motherboard manual and see for yourself,
>> MouthGuard.
>>
>>> Why did you change it? Why are you playing with BIOS settings
>>> (to change away from the defaults) that you don't understand?
>>
>> Because it was there? What difference does it make to you,
>> MouthGuard?
>>
>> The reason behind getting into trouble is irrelevant at this
>> time. The problem has apparently been resolved. I posted for
>> the conceivable benefit of anyone else who might have taken the
>> same path. Posting a solution is a well accepted practice here
>> on UseNet in technical discussion groups, at least as common as
>> asking questions.
>
> Oh, so we are to be thrilled that you choose to share you daily
> diary entry here in Usenet.

Who the **** is "we", MouthGuard? And what does "thrilled" have
anything to do with it? I posted 30 lines of text here, get over
it.

> You didn't want help.

By golly, I think he's got it.

To imagine that we are not supposed to post insights/answers here
on UseNet is just ridiculous.

> You definitely don't seem to be providing any to anyone
> specific.

I helped your mother just the other night, MouthGuard.

> So you just spewed out your diary entry.

Hardly. I went through serious trouble figuring out what my
problems were. Besides, MouthGuard, what I post here on UseNet is
none of your damn business (unless you want it to be).

> I guess I was wrong that a child wouldn't be putzing around in
> the BIOS settings but your response has definitely proven you

> are a child.

Just because you run around UseNet playing daddy, MouthGuard, does
not mean others are children (as if being a child is a bad thing).

> Like a troll,

Your mother is a troll, MouthGuard.

> you thought that I'd miss that you added unrelated newsgroups to
> your reply and I'd be posting to somewhere else. No thanks.

I could not care less, MouthGuard... If you think you are such a
big thing on UseNet, try stopping Google Groups from spewing spam
all over the place instead of wasting your time trying to prevent
regular posters from adding information to the archive.

> Oh yes, your post was ever so helpful to others using the same
> mobo and same OS.

Why in the past tense, MouthGuard? So far, UseNet has been
extremely long-lived. People find help here by searching the
Internet or the archives. UseNet is widely propagated and
information posted here is easy to find.

> Uh huh, like your guessing that the C4 CPU state caused the
> problem was laced with so much good evidence ... not!

And there is MouthGuard admitting that he is trolling for answers
instead of asking straightforward questions like a grown-up would.
That is what I thought.

Good evidence does not have to be provided along with an answer to
a question/problem, it just has to work. And whether it works for
anyone else, time will tell.
--


















> Yeah, figuring you're screwing with other BIOS settings you
> don't understand, that you are having problems with your host
> really isn't a surprise.
>
>
>

> Path: news.astraweb.com!border1.newsrouter.astraweb.com!feed.news.qwest.net!mpls-nntp-08.inet.qwest.net!news.glorb.com!weretis.net!feeder4.news.weretis.net!news.albasani.net!not-for-mail
> From: VanguardLH <V nguard.LH>
> Newsgroups: microsoft.public.windowsxp.general
> Subject: Re: >> C2/C2E State Support
> Date: Sun, 14 Nov 2010 22:16:41 -0600
> Organization: 0.35MOA sniper at loose in Usenet
> Lines: 97
> Message-ID: <ibqc59$knq$1 news.albasani.net>
> References: <4cdf6681$0$18677$c3e8da3$38634283 news.astraweb.com> <ibpjb8$ii4$1 news.albasani.net> <4ce05797$0$18344$c3e8da3$e408f015 news.astraweb.com>
> Mime-Version: 1.0
> Content-Type: text/plain; charset="us-ascii"
> Content-Transfer-Encoding: 7bit
> X-Trace: news.albasani.net gkFh7Fa5n0nUib6svkuD8WqXbzzl0YcCvOqlZZuaDtRPv9hX0KYHXMTty9oKK2b/ei6GkV0JLu9+H44jbSQcPMbmUJxddstpmC/Xd0eEaKF/EMlHrChO5h3GrkB7+veU
> NNTP-Posting-Date: Mon, 15 Nov 2010 04:15:38 +0000 (UTC)
> Keywords: VanguardLH VLH811
> Cancel-Lock: sha1:SzvUSGFZJsSC1TlseKOzB6Ghuik=
> User-Agent: 40tude_Dialog/2.0.15.41
> Injection-Info: news.albasani.net; logging-data="dMJWZF4LXAs4/WBexrNBxwEhpgk+p2gKHWiX8SqMe8pMLVsdRid4zH9U2ljpf3QwGLfk4xddY+Zf6zq4HbkCtCuRzNXXm3E+qn2CAo1BSds0jx6l6Xp5j9EFFcrMsZ5c"; mail-complaints-to="abuse albasani.net"
> X-MOTD: It is wiser to remain silent and let others ponder your ignorance than to speak and remove all doubt
>