Order of Update

[Keith Nuttle]

Does any one know how long MS things it will take to update all of the
computers?

What are they using to to determine the order of update. ie who is in
line first.

Alphabetical by windows store logon? Mac address? or other?

Desktop first? Laptops First?

I know it is not newest computers first, as my tablet has yet to be
updated and it is only a couple of months old.

[Brian Gregory]

On 02/08/2015 00:04, Keith Nuttle wrote:
> Does any one know how long MS things it will take to update all of the
> computers?
>
> What are they using to to determine the order of update. ie who is in
> line first.
>
> Alphabetical by windows store logon? Mac address? or other?
>
> Desktop first? Laptops First?
>
> I know it is not newest computers first, as my tablet has yet to be
> updated and it is only a couple of months old.
>

Devices that are left on a long time seem to be first.

--

Brian Gregory (in the UK).
To email me please remove all the letter vee from my email address.

[. . .winston]

Keith Nuttle wrote on 08/01/2015 7:04 PM:
> Does any one know how long MS things it will take to update all of the
> computers?
>
> What are they using to to determine the order of update. ie who is in
> line first.
>
> Alphabetical by windows store logon? Mac address? or other?
>
> Desktop first? Laptops First?
>
> I know it is not newest computers first, as my tablet has yet to be
> updated and it is only a couple of months old.
>
Compatibility data from the GWX icon carries the most significant
weight. As more data is collected from the upgrade...the new data
impacts the 'waiting to be upgraded' population.

As a result of the new data...the WU backend code is updated to ensure
that the waiting systems position in line is updated to allow 'best
case/most likely to succeed' installs first

i.e. How long ? Microsoft has only hinted at that - weeks, months....
--
...winston
msft mvp windows experience

[Paul]

Keith Nuttle wrote:
> Does any one know how long MS things it will take to update all of the
> computers?
>
> What are they using to to determine the order of update. ie who is in
> line first.
>
> Alphabetical by windows store logon? Mac address? or other?
>
> Desktop first? Laptops First?
>
> I know it is not newest computers first, as my tablet has yet to be
> updated and it is only a couple of months old.
>

Since the deployment isn't going all that
smoothly, expect the blocking feature on GWX to
hold off on your install for a while longer.

If a driver needed for your machine, is going to tip
over the install, there's no point going into an update
loop and driving you nuts. My guess is, the evidence is
there wasn't enough alpha-testing on the drivers, to make
good ones. For example, some RealTek drivers are making bad
sound right now.

Since the "deployment" is actually part of their
"test plan", and deployment time is being used
as test time, a user should not be all that surprised.
The question I have, is what a Dell purchaser would
see if buying a box today. Does their RealTek sound
work properly ? Is their NVIdia video card healthy ?

They have a whole year to make it work. That's
why the free upgrade interval is a year long. It's
not for tardy users, it's for MS cleanup :-)

Paul

[Al Drake]

On 8/2/2015 1:24 AM, . . .winston wrote:
> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>> Does any one know how long MS things it will take to update all of the
>> computers?
>>
>> What are they using to to determine the order of update. ie who is in
>> line first.
>>
>> Alphabetical by windows store logon? Mac address? or other?
>>
>> Desktop first? Laptops First?
>>
>> I know it is not newest computers first, as my tablet has yet to be
>> updated and it is only a couple of months old.
>>
> Compatibility data from the GWX icon carries the most significant
> weight. As more data is collected from the upgrade...the new data
> impacts the 'waiting to be upgraded' population.
>
> As a result of the new data...the WU backend code is updated to ensure
> that the waiting systems position in line is updated to allow 'best
> case/most likely to succeed' installs first
>
> i.e. How long ? Microsoft has only hinted at that - weeks, months....

I'm still trying to find out why one of my systems get a report of a
"compatibility issue: The CPU isn't supported"

I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM

[Paul]

Al Drake wrote:
> On 8/2/2015 1:24 AM, . . .winston wrote:
>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>> Does any one know how long MS things it will take to update all of the
>>> computers?
>>>
>>> What are they using to to determine the order of update. ie who is in
>>> line first.
>>>
>>> Alphabetical by windows store logon? Mac address? or other?
>>>
>>> Desktop first? Laptops First?
>>>
>>> I know it is not newest computers first, as my tablet has yet to be
>>> updated and it is only a couple of months old.
>>>
>> Compatibility data from the GWX icon carries the most significant
>> weight. As more data is collected from the upgrade...the new data
>> impacts the 'waiting to be upgraded' population.
>>
>> As a result of the new data...the WU backend code is updated to ensure
>> that the waiting systems position in line is updated to allow 'best
>> case/most likely to succeed' installs first
>>
>> i.e. How long ? Microsoft has only hinted at that - weeks, months....
>
> I'm still trying to find out why one of my systems get a report of a
> "compatibility issue: The CPU isn't supported"
>
> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM

How did your BIOS setup screen check go ?

Is NX/XD turned on ?

Paul

[Al Drake]

On 8/2/2015 1:56 PM, Paul wrote:
> Al Drake wrote:
>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>> Does any one know how long MS things it will take to update all of the
>>>> computers?
>>>>
>>>> What are they using to to determine the order of update. ie who is in
>>>> line first.
>>>>
>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>
>>>> Desktop first? Laptops First?
>>>>
>>>> I know it is not newest computers first, as my tablet has yet to be
>>>> updated and it is only a couple of months old.
>>>>
>>> Compatibility data from the GWX icon carries the most significant
>>> weight. As more data is collected from the upgrade...the new data
>>> impacts the 'waiting to be upgraded' population.
>>>
>>> As a result of the new data...the WU backend code is updated to ensure
>>> that the waiting systems position in line is updated to allow 'best
>>> case/most likely to succeed' installs first
>>>
>>> i.e. How long ? Microsoft has only hinted at that - weeks, months....
>>
>> I'm still trying to find out why one of my systems get a report of a
>> "compatibility issue: The CPU isn't supported"
>>
>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>
> How did your BIOS setup screen check go ?
>
> Is NX/XD turned on ?
>
> Paul
I have no idea what or where that setting is. I scoured BIOS and didn't
see anything that looked close to what you are asking.

[Paul]

Al Drake wrote:
> On 8/2/2015 1:56 PM, Paul wrote:
>> Al Drake wrote:
>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>> Does any one know how long MS things it will take to update all of the
>>>>> computers?
>>>>>
>>>>> What are they using to to determine the order of update. ie who is in
>>>>> line first.
>>>>>
>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>
>>>>> Desktop first? Laptops First?
>>>>>
>>>>> I know it is not newest computers first, as my tablet has yet to be
>>>>> updated and it is only a couple of months old.
>>>>>
>>>> Compatibility data from the GWX icon carries the most significant
>>>> weight. As more data is collected from the upgrade...the new data
>>>> impacts the 'waiting to be upgraded' population.
>>>>
>>>> As a result of the new data...the WU backend code is updated to ensure
>>>> that the waiting systems position in line is updated to allow 'best
>>>> case/most likely to succeed' installs first
>>>>
>>>> i.e. How long ? Microsoft has only hinted at that - weeks, months....
>>>
>>> I'm still trying to find out why one of my systems get a report of a
>>> "compatibility issue: The CPU isn't supported"
>>>
>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>
>> How did your BIOS setup screen check go ?
>>
>> Is NX/XD turned on ?
>>
>> Paul
> I have no idea what or where that setting is. I scoured BIOS and didn't
> see anything that looked close to what you are asking.

Give me the motherboard make and model number.

Paul

[Al Drake]

On 8/2/2015 10:04 PM, Paul wrote:
> Al Drake wrote:
>> On 8/2/2015 1:56 PM, Paul wrote:
>>> Al Drake wrote:
>>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>>> Does any one know how long MS things it will take to update all of
>>>>>> the
>>>>>> computers?
>>>>>>
>>>>>> What are they using to to determine the order of update. ie who is in
>>>>>> line first.
>>>>>>
>>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>>
>>>>>> Desktop first? Laptops First?
>>>>>>
>>>>>> I know it is not newest computers first, as my tablet has yet to be
>>>>>> updated and it is only a couple of months old.
>>>>>>
>>>>> Compatibility data from the GWX icon carries the most significant
>>>>> weight. As more data is collected from the upgrade...the new data
>>>>> impacts the 'waiting to be upgraded' population.
>>>>>
>>>>> As a result of the new data...the WU backend code is updated to ensure
>>>>> that the waiting systems position in line is updated to allow 'best
>>>>> case/most likely to succeed' installs first
>>>>>
>>>>> i.e. How long ? Microsoft has only hinted at that - weeks, months....
>>>>
>>>> I'm still trying to find out why one of my systems get a report of a
>>>> "compatibility issue: The CPU isn't supported"
>>>>
>>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>>
>>> How did your BIOS setup screen check go ?
>>>
>>> Is NX/XD turned on ?
>>>
>>> Paul
>> I have no idea what or where that setting is. I scoured BIOS and
>> didn't see anything that looked close to what you are asking.
>
> Give me the motherboard make and model number.
>
> Paul

Gigabyte P35-S3G F2 BIOS

[Paul]

Al Drake wrote:
> On 8/2/2015 10:04 PM, Paul wrote:
>> Al Drake wrote:
>>> On 8/2/2015 1:56 PM, Paul wrote:
>>>> Al Drake wrote:
>>>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>>>> Does any one know how long MS things it will take to update all of
>>>>>>> the
>>>>>>> computers?
>>>>>>>
>>>>>>> What are they using to to determine the order of update. ie who
>>>>>>> is in
>>>>>>> line first.
>>>>>>>
>>>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>>>
>>>>>>> Desktop first? Laptops First?
>>>>>>>
>>>>>>> I know it is not newest computers first, as my tablet has yet to be
>>>>>>> updated and it is only a couple of months old.
>>>>>>>
>>>>>> Compatibility data from the GWX icon carries the most significant
>>>>>> weight. As more data is collected from the upgrade...the new data
>>>>>> impacts the 'waiting to be upgraded' population.
>>>>>>
>>>>>> As a result of the new data...the WU backend code is updated to
>>>>>> ensure
>>>>>> that the waiting systems position in line is updated to allow 'best
>>>>>> case/most likely to succeed' installs first
>>>>>>
>>>>>> i.e. How long ? Microsoft has only hinted at that - weeks, months....
>>>>>
>>>>> I'm still trying to find out why one of my systems get a report of a
>>>>> "compatibility issue: The CPU isn't supported"
>>>>>
>>>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>>>
>>>> How did your BIOS setup screen check go ?
>>>>
>>>> Is NX/XD turned on ?
>>>>
>>>> Paul
>>> I have no idea what or where that setting is. I scoured BIOS and
>>> didn't see anything that looked close to what you are asking.
>>
>> Give me the motherboard make and model number.
>>
>> Paul
>
> Gigabyte P35-S3G F2 BIOS

Page 35, No Execute memory protection. Default - Enabled.

http://download1.gigabyte.us/Files/Manual/motherboard_manual_ga-p35-s3g_e.pdf

Paul

[Al Drake]

On 8/3/2015 12:03 AM, Paul wrote:
> Al Drake wrote:
>> On 8/2/2015 10:04 PM, Paul wrote:
>>> Al Drake wrote:
>>>> On 8/2/2015 1:56 PM, Paul wrote:
>>>>> Al Drake wrote:
>>>>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>>>>> Does any one know how long MS things it will take to update all of
>>>>>>>> the
>>>>>>>> computers?
>>>>>>>>
>>>>>>>> What are they using to to determine the order of update. ie who
>>>>>>>> is in
>>>>>>>> line first.
>>>>>>>>
>>>>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>>>>
>>>>>>>> Desktop first? Laptops First?
>>>>>>>>
>>>>>>>> I know it is not newest computers first, as my tablet has yet to be
>>>>>>>> updated and it is only a couple of months old.
>>>>>>>>
>>>>>>> Compatibility data from the GWX icon carries the most significant
>>>>>>> weight. As more data is collected from the upgrade...the new data
>>>>>>> impacts the 'waiting to be upgraded' population.
>>>>>>>
>>>>>>> As a result of the new data...the WU backend code is updated to
>>>>>>> ensure
>>>>>>> that the waiting systems position in line is updated to allow 'best
>>>>>>> case/most likely to succeed' installs first
>>>>>>>
>>>>>>> i.e. How long ? Microsoft has only hinted at that - weeks,
>>>>>>> months....
>>>>>>
>>>>>> I'm still trying to find out why one of my systems get a report of a
>>>>>> "compatibility issue: The CPU isn't supported"
>>>>>>
>>>>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>>>>
>>>>> How did your BIOS setup screen check go ?
>>>>>
>>>>> Is NX/XD turned on ?
>>>>>
>>>>> Paul
>>>> I have no idea what or where that setting is. I scoured BIOS and
>>>> didn't see anything that looked close to what you are asking.
>>>
>>> Give me the motherboard make and model number.
>>>
>>> Paul
>>
>> Gigabyte P35-S3G F2 BIOS
>
> Page 35, No Execute memory protection. Default - Enabled.
>
> http://download1.gigabyte.us/Files/Manual/motherboard_manual_ga-p35-s3g_e.pdf
>
>
> Paul

I finally Googled to learn what that was. I was confused when I was
directed to look for a "security' setting. I can't say I remember any of
my boards having that feature. I was more than a little confused when I
tried to connect that heading to my CPU as it deals with memory.

Thank you very much for taking the time to find what you did.

I guess having that turned on is a benefit to successfully upgrading to
Windows 10.

I'm going to go ahead and see if I can upgrade this system in spite of
that report. At least I'll know and there won't be any ill effects
otherwise.

[Paul]

Al Drake wrote:
> On 8/3/2015 12:03 AM, Paul wrote:
>> Al Drake wrote:
>>> On 8/2/2015 10:04 PM, Paul wrote:
>>>> Al Drake wrote:
>>>>> On 8/2/2015 1:56 PM, Paul wrote:
>>>>>> Al Drake wrote:
>>>>>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>>>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>>>>>> Does any one know how long MS things it will take to update all of
>>>>>>>>> the
>>>>>>>>> computers?
>>>>>>>>>
>>>>>>>>> What are they using to to determine the order of update. ie who
>>>>>>>>> is in
>>>>>>>>> line first.
>>>>>>>>>
>>>>>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>>>>>
>>>>>>>>> Desktop first? Laptops First?
>>>>>>>>>
>>>>>>>>> I know it is not newest computers first, as my tablet has yet
>>>>>>>>> to be
>>>>>>>>> updated and it is only a couple of months old.
>>>>>>>>>
>>>>>>>> Compatibility data from the GWX icon carries the most significant
>>>>>>>> weight. As more data is collected from the upgrade...the new data
>>>>>>>> impacts the 'waiting to be upgraded' population.
>>>>>>>>
>>>>>>>> As a result of the new data...the WU backend code is updated to
>>>>>>>> ensure
>>>>>>>> that the waiting systems position in line is updated to allow 'best
>>>>>>>> case/most likely to succeed' installs first
>>>>>>>>
>>>>>>>> i.e. How long ? Microsoft has only hinted at that - weeks,
>>>>>>>> months....
>>>>>>>
>>>>>>> I'm still trying to find out why one of my systems get a report
>>>>>>> of a
>>>>>>> "compatibility issue: The CPU isn't supported"
>>>>>>>
>>>>>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>>>>>
>>>>>> How did your BIOS setup screen check go ?
>>>>>>
>>>>>> Is NX/XD turned on ?
>>>>>>
>>>>>> Paul
>>>>> I have no idea what or where that setting is. I scoured BIOS and
>>>>> didn't see anything that looked close to what you are asking.
>>>>
>>>> Give me the motherboard make and model number.
>>>>
>>>> Paul
>>>
>>> Gigabyte P35-S3G F2 BIOS
>>
>> Page 35, No Execute memory protection. Default - Enabled.
>>
>> http://download1.gigabyte.us/Files/Manual/motherboard_manual_ga-p35-s3g_e.pdf
>>
>>
>>
>> Paul
>
> I finally Googled to learn what that was. I was confused when I was
> directed to look for a "security' setting. I can't say I remember any of
> my boards having that feature. I was more than a little confused when I
> tried to connect that heading to my CPU as it deals with memory.
>
> Thank you very much for taking the time to find what you did.
>
> I guess having that turned on is a benefit to successfully upgrading to
> Windows 10.
>
> I'm going to go ahead and see if I can upgrade this system in spite of
> that report. At least I'll know and there won't be any ill effects
> otherwise.

You have around six items to check for, and
one of them has a BIOS switch to check as well.

Some of the others can be checked with Coreinfo
from Sysinternals.

Paul

[Al Drake]

On 8/3/2015 5:15 AM, Paul wrote:
> Al Drake wrote:
>> On 8/3/2015 12:03 AM, Paul wrote:
>>> Al Drake wrote:
>>>> On 8/2/2015 10:04 PM, Paul wrote:
>>>>> Al Drake wrote:
>>>>>> On 8/2/2015 1:56 PM, Paul wrote:
>>>>>>> Al Drake wrote:
>>>>>>>> On 8/2/2015 1:24 AM, . . .winston wrote:
>>>>>>>>> Keith Nuttle wrote on 08/01/2015 7:04 PM:
>>>>>>>>>> Does any one know how long MS things it will take to update
>>>>>>>>>> all of
>>>>>>>>>> the
>>>>>>>>>> computers?
>>>>>>>>>>
>>>>>>>>>> What are they using to to determine the order of update. ie who
>>>>>>>>>> is in
>>>>>>>>>> line first.
>>>>>>>>>>
>>>>>>>>>> Alphabetical by windows store logon? Mac address? or other?
>>>>>>>>>>
>>>>>>>>>> Desktop first? Laptops First?
>>>>>>>>>>
>>>>>>>>>> I know it is not newest computers first, as my tablet has yet
>>>>>>>>>> to be
>>>>>>>>>> updated and it is only a couple of months old.
>>>>>>>>>>
>>>>>>>>> Compatibility data from the GWX icon carries the most significant
>>>>>>>>> weight. As more data is collected from the upgrade...the new data
>>>>>>>>> impacts the 'waiting to be upgraded' population.
>>>>>>>>>
>>>>>>>>> As a result of the new data...the WU backend code is updated to
>>>>>>>>> ensure
>>>>>>>>> that the waiting systems position in line is updated to allow
>>>>>>>>> 'best
>>>>>>>>> case/most likely to succeed' installs first
>>>>>>>>>
>>>>>>>>> i.e. How long ? Microsoft has only hinted at that - weeks,
>>>>>>>>> months....
>>>>>>>>
>>>>>>>> I'm still trying to find out why one of my systems get a report
>>>>>>>> of a
>>>>>>>> "compatibility issue: The CPU isn't supported"
>>>>>>>>
>>>>>>>> I have a Pentium Dual-Core E6800 @ 3.33GHz with 8 Gigs RAM
>>>>>>>
>>>>>>> How did your BIOS setup screen check go ?
>>>>>>>
>>>>>>> Is NX/XD turned on ?
>>>>>>>
>>>>>>> Paul
>>>>>> I have no idea what or where that setting is. I scoured BIOS and
>>>>>> didn't see anything that looked close to what you are asking.
>>>>>
>>>>> Give me the motherboard make and model number.
>>>>>
>>>>> Paul
>>>>
>>>> Gigabyte P35-S3G F2 BIOS
>>>
>>> Page 35, No Execute memory protection. Default - Enabled.
>>>
>>> http://download1.gigabyte.us/Files/Manual/motherboard_manual_ga-p35-s3g_e.pdf
>>>
>>>
>>>
>>> Paul
>>
>> I finally Googled to learn what that was. I was confused when I was
>> directed to look for a "security' setting. I can't say I remember any
>> of my boards having that feature. I was more than a little confused
>> when I tried to connect that heading to my CPU as it deals with memory.
>>
>> Thank you very much for taking the time to find what you did.
>>
>> I guess having that turned on is a benefit to successfully upgrading
>> to Windows 10.
>>
>> I'm going to go ahead and see if I can upgrade this system in spite
>> of that report. At least I'll know and there won't be any ill effects
>> otherwise.
>
> You have around six items to check for, and
> one of them has a BIOS switch to check as well.
>
> Some of the others can be checked with Coreinfo
> from Sysinternals.
>
> Paul
>
What would I be looking for that I would need and what am I do do with
what I find? What BIOS switch needs to be checked and what changes need
to be made?

[Paul]

Al Drake wrote:

> What would I be looking for that I would need and what am I do do with
> what I find? What BIOS switch needs to be checked and what changes need
> to be made?
>

The task definition was:

Find out what hardware requirement your CPU does not
meet. Even though your CPU is not of the crusty P4 generation
and is Core2. There is no excuse for the OS to not install!

As resources, you have Coreinfo. Try running it in Command Prompt.

https://technet.microsoft.com/en-us/sysinternals/cc835722

And the Win8+Win8.1 requirements were:

PAE - Physical address extension, for NX bit support.
PAE is used, because the translation table definition
leaves room for them to store an NX bit in the table.
Not sure this applies to 64 bit installs.
NX - Prevents malware from writing to code space.
Can be disabled in the BIOS.
SSE2 - Instruction set extension, used extensively for
block data movement.

PrefetchW - "Prefetch at least a 32-byte line into L1 data cache"
Possibly created by a cache hint pragma in source code.

CMPXCHG16b - "allows for atomic operations on octal words"
"perform a 128-bit locked compare and exchange"
(The processor may be 64 bit, but you're doing a 128 bit
operation in an uninterruptible way.) Some AMD64
lacked this. Coreinfo abbreviation is different.

LAHF/SAHF - "LAHF stands for Load AH from Flags and SAHF
stands for Store AH into Flags."
"used for... floating-point condition handling"

To run Coreinfo:

1) Download it and unpack ("Expand") with any ZIP tool you've
got.
2) Say the coreinfo.exe file ends up in C:\downloads ...
3) Open a command prompt window.

cd /d C:\Downloads
coreinfo
coreinfo > save_result_in_this_text_file.txt
notepad save_result_in_this_text_file.txt

4) Scroll through the results, looking for things like
CX16 (CMPXCHG16b).

This is my CPU. An asterisk in the middle column means "Yes".

Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
x86 Family 6 Model 23 Stepping 10, GenuineIntel

PAE * Supports > 32-bit physical addresses
NX * Supports no-execute page protection
[Must be switched on in BIOS too...]
SSE2 * Supports Streaming SIMD Extensions 2
PREFETCHW * Supports PREFETCHW instruction
CX16 * Supports CMPXCHG16B instruction
LAHF-SAHF * Supports LAHF/SAHF instructions in 64-bit mode

So I have an asterisk for all six of them. And
as far as I know, so should you. If NX was switched
off in the BIOS, that should be sufficient to cause
a silent failure. I don't think the other five are
in the BIOS.

The BIOS has some power state controls such as C1E,
but I hope the OS won't fall over if those have
been adjusted. Switching off all the C states, forces
the machine to stay at 3.33GHz all the time, and run
a bit warmer.

******* hardware config *******

E6800 CPU Pentium Processor 2M Cache, 3.33GHz FSB1066 (Q3'2010)

Gigabyte P35-S3G motherboard

Paul

[Al Drake]

On 8/3/2015 7:25 AM, Paul wrote:
> Al Drake wrote:
>
>> What would I be looking for that I would need and what am I do do
>> with what I find? What BIOS switch needs to be checked and what
>> changes need to be made?
>>
>
> The task definition was:
>
> Find out what hardware requirement your CPU does not
> meet. Even though your CPU is not of the crusty P4 generation
> and is Core2. There is no excuse for the OS to not install!
>
> As resources, you have Coreinfo. Try running it in Command Prompt.
>
> https://technet.microsoft.com/en-us/sysinternals/cc835722
>
> And the Win8+Win8.1 requirements were:
>
> PAE - Physical address extension, for NX bit support.
> PAE is used, because the translation table definition
> leaves room for them to store an NX bit in the table.
> Not sure this applies to 64 bit installs.
> NX - Prevents malware from writing to code space.
> Can be disabled in the BIOS.
> SSE2 - Instruction set extension, used extensively for
> block data movement.
>
> PrefetchW - "Prefetch at least a 32-byte line into L1 data cache"
> Possibly created by a cache hint pragma in source code.
>
> CMPXCHG16b - "allows for atomic operations on octal words"
> "perform a 128-bit locked compare and exchange"
> (The processor may be 64 bit, but you're doing a 128 bit
> operation in an uninterruptible way.) Some AMD64
> lacked this. Coreinfo abbreviation is different.
>
> LAHF/SAHF - "LAHF stands for Load AH from Flags and SAHF
> stands for Store AH into Flags."
> "used for... floating-point condition handling"
>
> To run Coreinfo:
>
> 1) Download it and unpack ("Expand") with any ZIP tool you've
> got.
> 2) Say the coreinfo.exe file ends up in C:\downloads ...
> 3) Open a command prompt window.
>
> cd /d C:\Downloads
> coreinfo
> coreinfo > save_result_in_this_text_file.txt
> notepad save_result_in_this_text_file.txt
>
> 4) Scroll through the results, looking for things like
> CX16 (CMPXCHG16b).
>
> This is my CPU. An asterisk in the middle column means "Yes".
>
> Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
> x86 Family 6 Model 23 Stepping 10, GenuineIntel
>
> PAE * Supports > 32-bit physical addresses
> NX * Supports no-execute page protection
> [Must be switched on in BIOS too...]
> SSE2 * Supports Streaming SIMD Extensions 2
> PREFETCHW * Supports PREFETCHW instruction
> CX16 * Supports CMPXCHG16B instruction
> LAHF-SAHF * Supports LAHF/SAHF instructions in 64-bit mode
>
> So I have an asterisk for all six of them. And
> as far as I know, so should you. If NX was switched
> off in the BIOS, that should be sufficient to cause
> a silent failure. I don't think the other five are
> in the BIOS.
>
> The BIOS has some power state controls such as C1E,
> but I hope the OS won't fall over if those have
> been adjusted. Switching off all the C states, forces
> the machine to stay at 3.33GHz all the time, and run
> a bit warmer.
>
> ******* hardware config *******
>
> E6800 CPU Pentium Processor 2M Cache, 3.33GHz FSB1066 (Q3'2010)
>
> Gigabyte P35-S3G motherboard
>
> Paul

Everything you have showing the "*" is what shows here except for the
CX16.

[Paul]

Al Drake wrote:
> On 8/3/2015 7:25 AM, Paul wrote:
>> Al Drake wrote:
>>
>>> What would I be looking for that I would need and what am I do do
>>> with what I find? What BIOS switch needs to be checked and what
>>> changes need to be made?
>>>
>>
>> The task definition was:
>>
>> Find out what hardware requirement your CPU does not
>> meet. Even though your CPU is not of the crusty P4 generation
>> and is Core2. There is no excuse for the OS to not install!
>>
>> As resources, you have Coreinfo. Try running it in Command Prompt.
>>
>> https://technet.microsoft.com/en-us/sysinternals/cc835722
>>
>> And the Win8+Win8.1 requirements were:
>>
>> PAE - Physical address extension, for NX bit support.
>> PAE is used, because the translation table definition
>> leaves room for them to store an NX bit in the table.
>> Not sure this applies to 64 bit installs.
>> NX - Prevents malware from writing to code space.
>> Can be disabled in the BIOS.
>> SSE2 - Instruction set extension, used extensively for
>> block data movement.
>>
>> PrefetchW - "Prefetch at least a 32-byte line into L1 data cache"
>> Possibly created by a cache hint pragma in source code.
>>
>> CMPXCHG16b - "allows for atomic operations on octal words"
>> "perform a 128-bit locked compare and exchange"
>> (The processor may be 64 bit, but you're doing a 128 bit
>> operation in an uninterruptible way.) Some AMD64
>> lacked this. Coreinfo abbreviation is different.
>>
>> LAHF/SAHF - "LAHF stands for Load AH from Flags and SAHF
>> stands for Store AH into Flags."
>> "used for... floating-point condition handling"
>>
>> To run Coreinfo:
>>
>> 1) Download it and unpack ("Expand") with any ZIP tool you've
>> got.
>> 2) Say the coreinfo.exe file ends up in C:\downloads ...
>> 3) Open a command prompt window.
>>
>> cd /d C:\Downloads
>> coreinfo
>> coreinfo > save_result_in_this_text_file.txt
>> notepad save_result_in_this_text_file.txt
>>
>> 4) Scroll through the results, looking for things like
>> CX16 (CMPXCHG16b).
>>
>> This is my CPU. An asterisk in the middle column means "Yes".
>>
>> Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
>> x86 Family 6 Model 23 Stepping 10, GenuineIntel
>>
>> PAE * Supports > 32-bit physical addresses
>> NX * Supports no-execute page protection
>> [Must be switched on in BIOS too...]
>> SSE2 * Supports Streaming SIMD Extensions 2
>> PREFETCHW * Supports PREFETCHW instruction
>> CX16 * Supports CMPXCHG16B instruction
>> LAHF-SAHF * Supports LAHF/SAHF instructions in 64-bit mode
>>
>> So I have an asterisk for all six of them. And
>> as far as I know, so should you. If NX was switched
>> off in the BIOS, that should be sufficient to cause
>> a silent failure. I don't think the other five are
>> in the BIOS.
>>
>> The BIOS has some power state controls such as C1E,
>> but I hope the OS won't fall over if those have
>> been adjusted. Switching off all the C states, forces
>> the machine to stay at 3.33GHz all the time, and run
>> a bit warmer.
>>
>> ******* hardware config *******
>>
>> E6800 CPU Pentium Processor 2M Cache, 3.33GHz FSB1066 (Q3'2010)
>>
>> Gigabyte P35-S3G motherboard
>>
>> Paul
>
> Everything you have showing the "*" is what shows here except for the
> CX16.

Thanks for working through it.

I found this.

http://superuser.com/questions/187254/how-prevalent-are-old-x64-processors-lacking-the-cmpxchg16b-instruction

"Support in the CPU may not actually be enough as
some Core 2 users found out on the launch of Windows 8.1

http://www.pcworld.com/article/2058683/new-windows-8-1-requirements-strand-some-users-on-windows-8.html

particularly those having an Intel DP35DP motherboard.
I'm not yet sure what in the motherboard could prevent
CMPXCHG16B from working. Probably the feature detection
bits can be disabled by BIOS."

I would start checking the BIOS release notes for the
BIOS updates, and see if there are any comments on
the subject. That's about all I can think of. Pretty
weird that the person mentioned P35, as yours is P35
chipset too.

The instruction executes inside the CPU, and should have
nothing to do with the chipset. The BIOS contributes
microcode patches, and it could be that some early
microcode patch had disabled either feature detection
or the actual instruction itself. Microcode patches
allow patching off parts of the CPU, and are how
all these processors with 100 errata in their errata
sheets, work so well for us. A number of the errata
are discovered in subsequent testing, months after
a processor is released.

No CPU has ever left the factory, free of errata.
And only some bugs are pretty serious from a performance
point of view, like the TLB issue on one of the AMD
processors (9500?).

Paul

[Al Drake]

On 8/3/2015 10:10 PM, Paul wrote:
> Al Drake wrote:
>> On 8/3/2015 7:25 AM, Paul wrote:
>>> Al Drake wrote:
>>>
>>>> What would I be looking for that I would need and what am I do do
>>>> with what I find? What BIOS switch needs to be checked and what
>>>> changes need to be made?
>>>>
>>>
>>> The task definition was:
>>>
>>> Find out what hardware requirement your CPU does not
>>> meet. Even though your CPU is not of the crusty P4 generation
>>> and is Core2. There is no excuse for the OS to not install!
>>>
>>> As resources, you have Coreinfo. Try running it in Command Prompt.
>>>
>>> https://technet.microsoft.com/en-us/sysinternals/cc835722
>>>
>>> And the Win8+Win8.1 requirements were:
>>>
>>> PAE - Physical address extension, for NX bit support.
>>> PAE is used, because the translation table definition
>>> leaves room for them to store an NX bit in the table.
>>> Not sure this applies to 64 bit installs.
>>> NX - Prevents malware from writing to code space.
>>> Can be disabled in the BIOS.
>>> SSE2 - Instruction set extension, used extensively for
>>> block data movement.
>>>
>>> PrefetchW - "Prefetch at least a 32-byte line into L1 data cache"
>>> Possibly created by a cache hint pragma in source code.
>>>
>>> CMPXCHG16b - "allows for atomic operations on octal words"
>>> "perform a 128-bit locked compare and exchange"
>>> (The processor may be 64 bit, but you're doing a 128
>>> bit
>>> operation in an uninterruptible way.) Some AMD64
>>> lacked this. Coreinfo abbreviation is different.
>>>
>>> LAHF/SAHF - "LAHF stands for Load AH from Flags and SAHF
>>> stands for Store AH into Flags."
>>> "used for... floating-point condition handling"
>>>
>>> To run Coreinfo:
>>>
>>> 1) Download it and unpack ("Expand") with any ZIP tool you've
>>> got.
>>> 2) Say the coreinfo.exe file ends up in C:\downloads ...
>>> 3) Open a command prompt window.
>>>
>>> cd /d C:\Downloads
>>> coreinfo
>>> coreinfo > save_result_in_this_text_file.txt
>>> notepad save_result_in_this_text_file.txt
>>>
>>> 4) Scroll through the results, looking for things like
>>> CX16 (CMPXCHG16b).
>>>
>>> This is my CPU. An asterisk in the middle column means "Yes".
>>>
>>> Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
>>> x86 Family 6 Model 23 Stepping 10, GenuineIntel
>>>
>>> PAE * Supports > 32-bit physical addresses
>>> NX * Supports no-execute page protection
>>> [Must be switched on in BIOS too...]
>>> SSE2 * Supports Streaming SIMD Extensions 2
>>> PREFETCHW * Supports PREFETCHW instruction
>>> CX16 * Supports CMPXCHG16B instruction
>>> LAHF-SAHF * Supports LAHF/SAHF instructions in 64-bit mode
>>>
>>> So I have an asterisk for all six of them. And
>>> as far as I know, so should you. If NX was switched
>>> off in the BIOS, that should be sufficient to cause
>>> a silent failure. I don't think the other five are
>>> in the BIOS.
>>>
>>> The BIOS has some power state controls such as C1E,
>>> but I hope the OS won't fall over if those have
>>> been adjusted. Switching off all the C states, forces
>>> the machine to stay at 3.33GHz all the time, and run
>>> a bit warmer.
>>>
>>> ******* hardware config *******
>>>
>>> E6800 CPU Pentium Processor 2M Cache, 3.33GHz FSB1066 (Q3'2010)
>>>
>>> Gigabyte P35-S3G motherboard
>>>
>>> Paul
>>
>> Everything you have showing the "*" is what shows here except for the
>> CX16.
>
> Thanks for working through it.
>
> I found this.
>
> http://superuser.com/questions/187254/how-prevalent-are-old-x64-processors-lacking-the-cmpxchg16b-instruction
>
>
> "Support in the CPU may not actually be enough as
> some Core 2 users found out on the launch of Windows 8.1
>
>
> http://www.pcworld.com/article/2058683/new-windows-8-1-requirements-strand-some-users-on-windows-8.html
>
>
> particularly those having an Intel DP35DP motherboard.
> I'm not yet sure what in the motherboard could prevent
> CMPXCHG16B from working. Probably the feature detection
> bits can be disabled by BIOS."
>
> I would start checking the BIOS release notes for the
> BIOS updates, and see if there are any comments on
> the subject. That's about all I can think of. Pretty
> weird that the person mentioned P35, as yours is P35
> chipset too.
>
> The instruction executes inside the CPU, and should have
> nothing to do with the chipset. The BIOS contributes
> microcode patches, and it could be that some early
> microcode patch had disabled either feature detection
> or the actual instruction itself. Microcode patches
> allow patching off parts of the CPU, and are how
> all these processors with 100 errata in their errata
> sheets, work so well for us. A number of the errata
> are discovered in subsequent testing, months after
> a processor is released.
>
> No CPU has ever left the factory, free of errata.
> And only some bugs are pretty serious from a performance
> point of view, like the TLB issue on one of the AMD
> processors (9500?).
>
> Paul
>
Maybe it's time to retire this old Box. After all I only use it for
browsing and email these days. I just like the case it's in. SO easy to
work with.

[Paul]

Al Drake wrote:
> On 8/3/2015 10:10 PM, Paul wrote:
>> Al Drake wrote:
>>> On 8/3/2015 7:25 AM, Paul wrote:
>>>> Al Drake wrote:
>>>>
>>>>> What would I be looking for that I would need and what am I do do
>>>>> with what I find? What BIOS switch needs to be checked and what
>>>>> changes need to be made?
>>>>>
>>>>
>>>> The task definition was:
>>>>
>>>> Find out what hardware requirement your CPU does not
>>>> meet. Even though your CPU is not of the crusty P4 generation
>>>> and is Core2. There is no excuse for the OS to not install!
>>>>
>>>> As resources, you have Coreinfo. Try running it in Command Prompt.
>>>>
>>>> https://technet.microsoft.com/en-us/sysinternals/cc835722
>>>>
>>>> And the Win8+Win8.1 requirements were:
>>>>
>>>> PAE - Physical address extension, for NX bit support.
>>>> PAE is used, because the translation table definition
>>>> leaves room for them to store an NX bit in the table.
>>>> Not sure this applies to 64 bit installs.
>>>> NX - Prevents malware from writing to code space.
>>>> Can be disabled in the BIOS.
>>>> SSE2 - Instruction set extension, used extensively for
>>>> block data movement.
>>>>
>>>> PrefetchW - "Prefetch at least a 32-byte line into L1 data cache"
>>>> Possibly created by a cache hint pragma in source code.
>>>>
>>>> CMPXCHG16b - "allows for atomic operations on octal words"
>>>> "perform a 128-bit locked compare and exchange"
>>>> (The processor may be 64 bit, but you're doing a 128
>>>> bit
>>>> operation in an uninterruptible way.) Some AMD64
>>>> lacked this. Coreinfo abbreviation is different.
>>>>
>>>> LAHF/SAHF - "LAHF stands for Load AH from Flags and SAHF
>>>> stands for Store AH into Flags."
>>>> "used for... floating-point condition handling"
>>>>
>>>> To run Coreinfo:
>>>>
>>>> 1) Download it and unpack ("Expand") with any ZIP tool you've
>>>> got.
>>>> 2) Say the coreinfo.exe file ends up in C:\downloads ...
>>>> 3) Open a command prompt window.
>>>>
>>>> cd /d C:\Downloads
>>>> coreinfo
>>>> coreinfo > save_result_in_this_text_file.txt
>>>> notepad save_result_in_this_text_file.txt
>>>>
>>>> 4) Scroll through the results, looking for things like
>>>> CX16 (CMPXCHG16b).
>>>>
>>>> This is my CPU. An asterisk in the middle column means "Yes".
>>>>
>>>> Intel(R) Core(TM)2 Duo CPU E8400 @ 3.00GHz
>>>> x86 Family 6 Model 23 Stepping 10, GenuineIntel
>>>>
>>>> PAE * Supports > 32-bit physical addresses
>>>> NX * Supports no-execute page protection
>>>> [Must be switched on in BIOS too...]
>>>> SSE2 * Supports Streaming SIMD Extensions 2
>>>> PREFETCHW * Supports PREFETCHW instruction
>>>> CX16 * Supports CMPXCHG16B instruction
>>>> LAHF-SAHF * Supports LAHF/SAHF instructions in 64-bit mode
>>>>
>>>> So I have an asterisk for all six of them. And
>>>> as far as I know, so should you. If NX was switched
>>>> off in the BIOS, that should be sufficient to cause
>>>> a silent failure. I don't think the other five are
>>>> in the BIOS.
>>>>
>>>> The BIOS has some power state controls such as C1E,
>>>> but I hope the OS won't fall over if those have
>>>> been adjusted. Switching off all the C states, forces
>>>> the machine to stay at 3.33GHz all the time, and run
>>>> a bit warmer.
>>>>
>>>> ******* hardware config *******
>>>>
>>>> E6800 CPU Pentium Processor 2M Cache, 3.33GHz FSB1066 (Q3'2010)
>>>>
>>>> Gigabyte P35-S3G motherboard
>>>>
>>>> Paul
>>>
>>> Everything you have showing the "*" is what shows here except for the
>>> CX16.
>>
>> Thanks for working through it.
>>
>> I found this.
>>
>> http://superuser.com/questions/187254/how-prevalent-are-old-x64-processors-lacking-the-cmpxchg16b-instruction
>>
>>
>>
>> "Support in the CPU may not actually be enough as
>> some Core 2 users found out on the launch of Windows 8.1
>>
>>
>> http://www.pcworld.com/article/2058683/new-windows-8-1-requirements-strand-some-users-on-windows-8.html
>>
>>
>>
>> particularly those having an Intel DP35DP motherboard.
>> I'm not yet sure what in the motherboard could prevent
>> CMPXCHG16B from working. Probably the feature detection
>> bits can be disabled by BIOS."
>>
>> I would start checking the BIOS release notes for the
>> BIOS updates, and see if there are any comments on
>> the subject. That's about all I can think of. Pretty
>> weird that the person mentioned P35, as yours is P35
>> chipset too.
>>
>> The instruction executes inside the CPU, and should have
>> nothing to do with the chipset. The BIOS contributes
>> microcode patches, and it could be that some early
>> microcode patch had disabled either feature detection
>> or the actual instruction itself. Microcode patches
>> allow patching off parts of the CPU, and are how
>> all these processors with 100 errata in their errata
>> sheets, work so well for us. A number of the errata
>> are discovered in subsequent testing, months after
>> a processor is released.
>>
>> No CPU has ever left the factory, free of errata.
>> And only some bugs are pretty serious from a performance
>> point of view, like the TLB issue on one of the AMD
>> processors (9500?).
>>
>> Paul
>>
> Maybe it's time to retire this old Box. After all I only use it for
> browsing and email these days. I just like the case it's in. SO easy to
> work with.

The purpose of reviewing BIOS release notes,
is to see whether upgrading the BIOS chip makes
sense or not.

Microcode patches are done at two points in time. The
BIOS has around eight microcode patches (on average)
stored inside the BIOS chip. These handle all the different
families of CPUs in the CPU support chart (maybe
40 CPUs). The CPU actually checks the header of the
microcode patch, and rejects patches loaded into
it, which do not match.

Windows also has a Microcode patcher. It's a service
that runs at startup, and attempts to install a
microcode patch.

The purpose of the BIOS patcher, is to keep the
computer stable enough to finish booting. Then,
if Windows wants to install a later patch,
it can be done early in the session.

There have been CPU issues, which prevent boot
from finishing. And that would be an example of
a case, where a BIOS update is mandatory to
finish the computer build and testing. As a general
rule, people no longer see that kind of issue, and
the motherboard designers and teams at Intel
are earning their pay.

Also, the microcode patch applied to the CPU,
ends up displayed as the CPU "revision". So you
can trace, with a Windows level utility, what
patch level you're at. There were some CPUs
on one of the OS upgrades, where the CPU revision
was "zero". Which means neither effort to patch
was working.

https://support.microsoft.com/en-us/kb/842465

"Download and run Intel Processor Identification Utility

Write down the following CPU information for each processor:

CPU Family
CPU Model
CPU Stepping
CPU Revision

The CPU Family, CPU Model, and CPU Stepping values
identify the specific type of processor.

The CPU Revision value identifies the revision
of the microcode update that is applied."

For example, Intel PIU identified my CPU as

CPU Type 0
CPU Family 6
CPU Model 17
CPU Stepping A
CPU Revision A07

and the A should be stripped off to give "Microcode patch 07".

The main benefit of doing that sort of check, is to
see if the CPU Revision is 00, as that would be
bad. And implies the OS had a problem that needs
to be fixed. You could check that while you're in Windows 7,
just to see whether it's been patched or not.

Looking at the available info, I think there are
newer BIOS files available, but I don't see
a strong incentive to install them. The issue is
not specifically identified. And looking at a
Spec Update PDF from Intel right now, there's no
"strong" indication there of a problem either.
There is a TLB bug which mentions the class of instruction,
but it also mentions just about every other instruction
that sets an EFLAG. The bug is not specific enough
to point a finger.

If you tried a BIOS update, it would be to see
if an updated microcode patch made any difference.
The Microsoft installer DVD should really be using
a microcode patcher of its own, before the install starts.
If it did not, and relied on the BIOS patcher/version,
that would be a mistake. And then a BIOS update might
help.

Paul