Is it possible to force a paging out ?

NTDEV
1

For test purposes I would need to force some specific memory pages out to disk.
Is there any way to do so ?

Thanks.

2

Verifier in some modes forces all pageable memory to disk on each spinlock acquizition.

Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation
xxxxx@storagecraft.com
http://www.storagecraft.com

----- Original Message -----
From: I?aki Castillo
To: Windows System Software Devs Interest List
Sent: Monday, September 06, 2004 8:15 PM
Subject: [ntdev] Is it possible to force a paging out ?

For test purposes I would need to force some specific memory pages out to disk.
Is there any way to do so ?

Thanks.

Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com

3

I don’t think it actually forces them to disk - that would be incredibly slow. It can invalidate them so that any access by the raised-IRQL code causes a page fault, but this doesn’t require that they go to disk since the validity can be reestablished as the pages are accessed (unless it’s the processor at or above DISPATCH_LEVEL in which case the verifier bugchecks).

  • p

From: xxxxx@lists.osr.com [mailto:xxxxx@lists.osr.com] On Behalf Of Maxim S. Shatskih
Sent: Monday, September 06, 2004 1:22 PM
To: Windows System Software Devs Interest List
Subject: Re: [ntdev] Is it possible to force a paging out ?

Verifier in some modes forces all pageable memory to disk on each spinlock acquizition.

Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation
xxxxx@storagecraft.com
http://www.storagecraft.com

----- Original Message -----
From: I?aki Castillo mailto:xxxxx
To: Windows System Software Devs Interest List mailto:xxxxx
Sent: Monday, September 06, 2004 8:15 PM
Subject: [ntdev] Is it possible to force a paging out ?

For test purposes I would need to force some specific memory pages out to disk.
Is there any way to do so ?

Thanks.

—
Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com mailto:xxxxx

—
Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: unknown lmsubst tag argument: ‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com</mailto:xxxxx></mailto:xxxxx></mailto:xxxxx>

4

Thanks to Peter for the explanation of that. I have “used” the driver
verifier for quite some time, and find it a very handy tool for finding=

bugs in the code that would take a lot longer to discover (and correct)=
if
they only happened when it really goes wrong in the real world.

However, I wonder what purpose the original poster had in mind with
“forcing page out”. Is it to try something out in a disk filter driver,=
or
is it just to see what happens if the page is not accessible at higher
IRQL? If it’s the latter, obviously, Driver Verifier will do the job ni=
cely
(anything owned by the drivers that DV is told to care about will be
“inaccessible”). If it’s the former, it would need to be some special
function in Windows that causes this, and I don’t see what the general
usefulness of this would be, other than for specific test purposes (and=
I
don’t even see much point in that, but then I’ve been blamed for lack o=
f
imagination on occasion [whilst others think that my imagination runs a=
way
sometimes, you can’t win ;-)]).

So could the original poster explain what the actual purpose is? [Or sa=
y
“Ok, thanks, that’s what I was after” to one of the replies…]

–
Mats

xxxxx@lists.osr.com wrote on 09/07/2004 02:56:06 PM:

I don’t think it actually forces them to disk - that would be
incredibly slow. It can invalidate them so that any access by the
raised-IRQL code causes a page fault, but this doesn’t require that
they go to disk since the validity can be reestablished as the pages
are accessed (unless it’s the processor at or above DISPATCH_LEVEL
in which case the verifier bugchecks).

  • p

From: xxxxx@lists.osr.com [mailto:
xxxxx@lists.osr.com] On Behalf Of Maxim S. Shatskih
Sent: Monday, September 06, 2004 1:22 PM
To: Windows System Software Devs Interest List
Subject: Re: [ntdev] Is it possible to force a paging out ?

Verifier in some modes forces all pageable memory to disk on
each spinlock acquizition.

Maxim Shatskih, Windows DDK MVP
StorageCraft Corporation
xxxxx@storagecraft.com
http://www.storagecraft.com
----- Original Message -----
From: I=F1aki Castillo
To: Windows System Software Devs Interest List
Sent: Monday, September 06, 2004 8:15 PM
Subject: [ntdev] Is it possible to force a paging out ?

For test purposes I would need to force some specific memory pages
out to disk.
Is there any way to do so ?

Thanks.

Questions? First check the Kernel Driver FAQ at http://www.
osronline.com/article.cfm?id=3D256

You are currently subscribed to ntdev as: unknown lmsubst tag argumen=
t:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com=

Questions? First check the Kernel Driver FAQ at http://www.
osronline.com/article.cfm?id=3D256

You are currently subscribed to ntdev as: unknown lmsubst tag argumen=
t:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com=

Questions? First check the Kernel Driver FAQ at http://www.
osronline.com/article.cfm?id=3D256

You are currently subscribed to ntdev as: unknown lmsubst tag argumen=
t:
‘’
To unsubscribe send a blank email to xxxxx@lists.osr.com=

ForwardSourceID:NT00002DDE=

5

Iñaki Castillo wrote:

For test purposes I would need to force some specific memory pages out
to disk.

You need to be a bit more specific: Are you looking to ensure that the
memory pages are actually WRITTEN to disk, or is it enough that they are
just put on the list pending their writing to disk and possibly “soft
faulted” back in? Are these pages in your driver, in a section, or in
the system someplace else?

If you want to test IRQL handling and stuff, you can always do something
like MmPageEntireDriver(…)

Peter
OSR

6

Actually I don’t need specifically the page go back ‘to disk’.
It is enough for me if I can be sure it is not present, but is valid.

I don’t want to test IRQL handling.
I need to select specific pages, I don’t want to page out entire drivers.

Thanks.
Inaki.

-----Mensaje original-----
De: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com]En nombre de PeterGV
Enviado el: mi?rcoles, 08 de septiembre de 2004 0:57
Para: Windows System Software Devs Interest List
Asunto: Re:[ntdev] Is it possible to force a paging out ?

I?aki Castillo wrote:

For test purposes I would need to force some specific memory pages out
to disk.

You need to be a bit more specific: Are you looking to ensure that the
memory pages are actually WRITTEN to disk, or is it enough that they are
just put on the list pending their writing to disk and possibly “soft
faulted” back in? Are these pages in your driver, in a section, or in
the system someplace else?

If you want to test IRQL handling and stuff, you can always do something
like MmPageEntireDriver(…)

Peter
OSR

Questions? First check the Kernel Driver FAQ at http://www.osronline.com/article.cfm?id=256

You are currently subscribed to ntdev as: xxxxx@pandasoftware.es
To unsubscribe send a blank email to xxxxx@lists.osr.com