Kaiser Page Pool!!!

[austinsnyc]

Well I believe I have found the offset for the Kaiser PagePool and it is set at the standard 6MB. here is the copy from HexWorkshop 4.2

FFFF FFFF 9B4F FFFF 64B0 0000 6400 0000 C874 0480 0000 6000 0000 0000

As you can see its almost the same as the Hermes which makes life easy if Oli or Pof would like to confirm my findings that would be great! The Offset is 21963880.. This should make for easy adjustment of the pagepool once we have the Imgfs corrected :)

[pof]

seems correct, but the pagepool offset may vary from one rom to another, see here (http://forum.xda-developers.com/showthread.php?t=295932) how to find it.

[austinsnyc]

Cool thanks for the confermation Pof. I have also spoke with Taz on getting the imgfs tools modded to work with the Kaiser. It will be a great day in Kaiser History when that happens!

[duttythroy]

Cool thanks for the confermation Pof. I have also spoke with Taz on getting the imgfs tools modded to work with the Kaiser. It will be a great day in Kaiser History when that happens!

Great news mate I cant wait to flash:)

[wizzzard]

Coming from a Wizard background where to me a 6mb page pool was my optimal, what size page pool is recomended with the128mb memory on our Kaisers. I must admit with the larger memory on the Kaiser I have stopped closing apps and just minimise most apps and let Windows get on with managing resources. It works much better than my Wizard in this respect.

[austinsnyc]

Coming from a Wizard background where to me a 6mb page pool was my optimal, what size page pool is recomended with the128mb memory on our Kaisers. I must admit with the larger memory on the Kaiser I have stopped closing apps and just minimise most apps and let Windows get on with managing resources. It works much better than my Wizard in this respect.

I am honestly not sure I might go down to a Zero PagePool and test that since there is so much RAM to be used :) Untill someone does some benchmarks with different page pool sizes I would have to say that 4MB and 12MB would probably be the fastest since that seems to be the case with most devices

[sirsycho]

Hey austinsync...

I'm not finding it at the offset you have above... For reference I'm finding it here:

0x003A6F80

That is the same location for both the ATT and HTC ROMs. Can you confirm? (or set me straight).

Cheers.

[austinsnyc]

yes give me a minute ok. Also are you just opening the OS.nb in Hex editor or the whole .nbh?

[sirsycho]

yes give me a minute ok. Also are you just opening the OS.nb in Hex editor or the whole .nbh?

no worries, already extracted the NBH so just dealing with the OS.nb.

Thanks.

Oh and I'm using Hex Workshop v5 (beta)

[rzanology]

okay guys....hold on let me put on my flame suit.

so i've been readying this post and i can't for the life of me figure out what the hell you guys are talking about. coming from the 8525, i know when i flash there are two files...and one of them is the file you guys refer to. How do i edit these files? where are you guys getting these tools? Where can i read up to better understand what does what.

[jcespi2005]

okay guys....hold on let me put on my flame suit.

so i've been readying this post and i can't for the life of me figure out what the hell you guys are talking about. coming from the 8525, i know when i flash there are two files...and one of them is the file you guys refer to. How do i edit these files? where are you guys getting these tools? Where can i read up to better understand what does what.

For better understand of the PagePool, take a deep reading in http://forum.xda-developers.com/showthread.php?t=295932 by Pof...

Sure helps you.

[austinsnyc]

what the factory set RAM? I set mine to 4MB and forgot what it used to be someone check there untouched Kaiser for me please thanks

[austinsnyc]

can no one look at there phone real fast and tell me what the Memory RAM is total??

[dan13l]

On my Vario III, the total program ram is 101.38. That help?

[austinsnyc]

thats perfect thank you!

[christonge]

I did something similar! I edited the OS.nb to 4mb pagepool and then flashed it but when I checked the total program RAM I didn't have a reference from the standard ROM, so I flashed back to check! jeje!
Anyway... I've noticed that there's no change in my total RAM after I changed to the 4mb pagepool..? Anyone else noticed this? Thanks....

[jcespi2005]

On my Vario III, the total program ram is 101.38. That help?

Same on my HTC TyTN II: 101,38 Mb memory and 134,29Mb Storage.

[sleeping]

Same on my HTC TyTN II: 101,38 Mb memory and 134,29Mb Storage.

On my virgin AT&T Tilt it's 101.38 MB Program and 124.91 MB Storage

[austinsnyc]

Yes I have noticed this also guys no matter what you change the PagePool to in Hex it always reverts back to 6MB so this is some very interesting stuff. Maybe that offset isn't for the Kaiser or maybe the Kaiser doesn't have a Pagepool given the amaount of RAM it has from the factory I'm not sure myself until more testing is done.

[jcespi2005]

Yes I have noticed this also guys no matter what you change the PagePool to in Hex it always reverts back to 6MB so this is some very interesting stuff. Maybe that offset isn't for the Kaiser or maybe the Kaiser doesn't have a Pagepool given the amaount of RAM it has from the factory I'm not sure myself until more testing is done.

Perhaps the PagePool is fixed parameter in Kaiser?

[DiscoSmurf]

Yes I have noticed this also guys no matter what you change the PagePool to in Hex it always reverts back to 6MB so this is some very interesting stuff. Maybe that offset isn't for the Kaiser or maybe the Kaiser doesn't have a Pagepool given the amaount of RAM it has from the factory I'm not sure myself until more testing is done.

Perhaps the PagePool is fixed parameter in Kaiser?

it appears the recently cooked NEW Ported AKU.0.7.0 Kernel & OS has a modified page pool of 12MB. see my attached images.

the NEW Ported AKU.0.7.0 Kernel & OS should be first the image if you hover over it and labeled "helmi" followed by the htc and at&t respectfully.

-ds

[austinsnyc]

well thats the thing even though he has changed the number there is still the same amount of memory

[dan13l]

You're absolutely right: running the AKU.0.7.0 I've got 132.91 Storage, and 101.38 Program.

[jcespi2005]

well thats the thing even though he has changed the number there is still the same amount of memory

Other posibility is the PagePool is using (reserved) memory above or bellow the accesible RAM for programs so you can see visible changes when you change the pagepool... can be this possible? or am i saying nonsenses :)

[austinsnyc]

Yes which leads me to believe one of two things.
A: the Kaiser has no pagepool at all.
B: this is not the location of the Kaiser pagepool even though it seems like it is.

[jcespi2005]

Yes which leads me to believe one of two things.
A: the Kaiser has no pagepool at all.
B: this is not the location of the Kaiser pagepool even though it seems like it is.

How can we test and verify if A or B or C...? any idea on the how to? i read and reread the Pof post about pagepool, but can't find any method to see the pagepool use in the running system. Seems the only effect is the amount of memory.

[christonge]

Finally got connected after day in the country.... this issue is bugging me. I have been running the "4mb pagepool" ROM now for 24 hours and I have to say it feels quicker but of course this could all be in my head :p.... as the RAM shows no number change.
I think a way to see would be to run a benchmark for each setting. Although it´s not completely conclusive it would give us some indication. I will download the Spb version tomorrow and give it a try... lets see if we can get some answers.:confused:

[UnicornKaz]

My TyTN II original ( factory stock ) showed: (in the "start -->> settings -->> system-->> memory)

133.91 MB Storage total ( also same at "start -->> settings -->> system -->> device info-->> hardware" )
101.38 MB Program total

but in the "start -->> settings -->> about" ... it showed 101.63 MB :confused:

the Program amount seems to match with other reports but NOT the Storage .... but why the difference at 2 different places for my own "Program" ??

could someone explain ??

Thanks

[jackieboy]

Why would I want a 4mb pagepool vs a 12mb or 6mb etc?

[cuboosh]

higher the number the better the speed.. however the lower the number the more the program memory..so the story goes

[Sleuth255]

Pagepool settings on a Kaiser appear to be change the amount of available program/storage memory however.

It appears to me that the 4MB size is slower than the 6MB size btw.

edit: confirmed. 4MB pagepool size really dogged out my Tilt. I'm back to 6MB.

[jcespi2005]

Pagepool settings on a Kaiser appear to be change the amount of available program/storage memory however.

It appears to me that the 4MB size is slower than the 6MB size btw.

edit: confirmed. 4MB pagepool size really dogged out my Tilt. I'm back to 6MB.

I'm interested to know if total amount of memory changes with thouse pagepool... think when you change the pagepool, the total memory must change too, but in Kaiser seems different. Do you know how pagepool works in Kaiser?

THanks ;)

[Sleuth255]

Interestingly, when I patched to 4MB, my available program memory increased by about 1.3megs. Even stranger, my available storage memory also increased.... :confused:

I'm wondering if this means that the pagepool is dynamically allocated from program memory in a Kaiser at boottime. 6MB is absolutely faster than 4MB too. I haven't tried even larger values yet.

[jcespi2005]

Interestingly, when I patched to 4MB, my available program memory increased by about 1.3megs. Even stranger, my available storage memory also increased.... :confused:

I'm wondering if this means that the pagepool is dynamically allocated from program memory in a Kaiser at boottime. 6MB is absolutely faster than 4MB too. I haven't tried even larger values yet.

But although the pagepool was dynamically allocated, this must take effect on total memory...unless the memory used for pagepool was over the memory reserved for the system, but is difficult to know where...

Perhaps the memory for storage and for programs can be shared, like on the olds 2003 VM versions?

Perhaps something like real swap file like on Windows NT/2000/XP/...?

Perhaps the Qualcomm processor have some special feature to allow the pagepool on other area...

Too many "perhaps".....

Cheers.

[1adam12]

What tool are you guys using to set the Page Pool size? This thing definitely needs a boost and I'd like to try some different values. 'Perceived' speed is as good as real speed ;)

[om-by]

What tool are you guys using to set the Page Pool size? This thing definitely needs a boost and I'd like to try some different values. 'Perceived' speed is as good as real speed ;)

Not a big problem - how to change and by what tool... look here (http://forum.xda-developers.com/showthread.php?t=323269).
Another question - about effect... try, for example, to compare 1 and 64 :)

[Moe5508]

Great - thanks for making this available for Kaiser!

[UnicornKaz]

can the author of this thread or the experts out there confirm on the legit result of this report by "tom_codon" and what he said regarding the page pool is INEFFECTIVE ???

http://forum.xda-developers.com/showthread.php?p=1891111&highlight=Niki%2CPolaris%2Ckaiser#post1891111

Thank you !!!

[GSLEON3]

What we know about WM6 builds & the Kaise is that there are two page pools with two settings each. This is common MSDN knowledge. Look for my PP post from yesterday.

What the tool is changing is the ~Max that can set by the OEM through the OEM adaptation layer. The actual target PP size looks like it's in xip & nobody's found the address yet. It's low level & loaded before the reg.

[GSLEON3]

I found some more interesting information in the MSDN resources.

Acording to MS you should be able to change th PP without altering code or the OAL.

The kernel has two pools: one for loader pages (EXE and DLL pages) and another for the file cache and memory-mapped files. The two pools can be controlled independently, and can be disabled by setting their sizes to zero. When a pool is disabled, there are no constraints on memory usage. The loader or file cache can consume pages up to the limit of free physical memory.

The pool parameters are configurable with the Config.bib file FIXUPVAR values, so it is possible to modify these values without modifying code or rebuilding the OEM adaptation layer (OAL).

Copy Codeconst volatile DWORD LoaderPoolTarget;
const volatile DWORD LoaderPoolMaximum;
const volatile DWORD LoaderPoolReleaseIncrement;
const volatile DWORD LoaderPoolCriticalIncrement;
const volatile DWORD LoaderPoolNormalPriority256;
const volatile DWORD LoaderPoolCriticalPriority256;
const volatile DWORD FilePoolTarget;
const volatile DWORD FilePoolMaximum;
const volatile DWORD FilePoolReleaseIncrement;
const volatile DWORD FilePoolCriticalIncrement;
const volatile DWORD FilePoolNormalPriority256;
const volatile DWORD FilePoolCriticalPriority256;
OEMs can also modify the values at run-time, by calling an OAL I/O control, IOCTL_HAL_GET_POOL_PARAMETERS, and passing a NKPagePoolParameters structure.



It also seems that you can create a state where the page pool is dynamic & will page out as far as is needed just by setting the value to 0. Anyone know where the config.bib resides???

[Michaelr219]

no, but if you figure this out... youd be legendary lol... this would really help out with performance