write throttle, memory mapped files, ...

NTFSD
1

Gentlefolk

The cc/mm throttle cached changes (writes) which come through the file
system to files when the number of available pages is low in order to avoid
the system panic which arises when there are no available pages. It appears
however that there is no mechanism to throttle cached changes which do not
come through the file system because files are memory mapped. It might
therefore be possible to write some user mode programs which will cause the
system panic; this is possible and can be demonstrated.

The question I have is how does mm/cc decide whether to raise NO_MORE_PAGES
or DIRTY_MAPPED_PAGES_CONGESTION?

Cheers
Lyndon

2

This will never cause panic, only the performance degradation. The thread
which caused a page fault will be stalled till there will be some available
pages.

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

----- Original Message -----
From: “Lyndon J Clarke”
Newsgroups: ntfsd
To: “Windows File Systems Devs Interest List”
Sent: Wednesday, April 27, 2005 1:22 PM
Subject: [ntfsd] write throttle, memory mapped files, …

> Gentlefolk
>
> The cc/mm throttle cached changes (writes) which come through the file
> system to files when the number of available pages is low in order to avoid
> the system panic which arises when there are no available pages. It appears
> however that there is no mechanism to throttle cached changes which do not
> come through the file system because files are memory mapped. It might
> therefore be possible to write some user mode programs which will cause the
> system panic; this is possible and can be demonstrated.
>
> The question I have is how does mm/cc decide whether to raise NO_MORE_PAGES
> or DIRTY_MAPPED_PAGES_CONGESTION?
>
> Cheers
> Lyndon
>
>
>
> —
> Questions? First check the IFS FAQ at
https://www.osronline.com/article.cfm?id=17
>
> You are currently subscribed to ntfsd as: xxxxx@storagecraft.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com

3

Beg to differ; I am able to force bugcheck NO_MORE_PAGES.

“Maxim S. Shatskih” wrote in message
news:xxxxx@ntfsd…
> This will never cause panic, only the performance degradation. The
> thread
> which caused a page fault will be stalled till there will be some
> available
> pages.
>
> Maxim Shatskih, Windows DDK MVP
> StorageCraft Corporation
> xxxxx@storagecraft.com
> http://www.storagecraft.com
>
> ----- Original Message -----
> From: “Lyndon J Clarke”
> Newsgroups: ntfsd
> To: “Windows File Systems Devs Interest List”
> Sent: Wednesday, April 27, 2005 1:22 PM
> Subject: [ntfsd] write throttle, memory mapped files, …
>
>
>> Gentlefolk
>>
>> The cc/mm throttle cached changes (writes) which come through the file
>> system to files when the number of available pages is low in order to
>> avoid
>> the system panic which arises when there are no available pages. It
>> appears
>> however that there is no mechanism to throttle cached changes which do
>> not
>> come through the file system because files are memory mapped. It might
>> therefore be possible to write some user mode programs which will cause
>> the
>> system panic; this is possible and can be demonstrated.
>>
>> The question I have is how does mm/cc decide whether to raise
>> NO_MORE_PAGES
>> or DIRTY_MAPPED_PAGES_CONGESTION?
>>
>> Cheers
>> Lyndon
>>
>>
>>
>> —
>> Questions? First check the IFS FAQ at
> https://www.osronline.com/article.cfm?id=17
>>
>> You are currently subscribed to ntfsd as: xxxxx@storagecraft.com
>> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>

4

> Beg to differ; I am able to force bugcheck NO_MORE_PAGES.

Do you have a testprogram for it ?
If yes, could you send me it ?

L.

5

I think Maxim if you say this cannot cause panic I must have made an error;
rechecking …

“Maxim S. Shatskih” wrote in message
news:xxxxx@ntfsd…
> This will never cause panic, only the performance degradation. The
> thread
> which caused a page fault will be stalled till there will be some
> available
> pages.
>
> Maxim Shatskih, Windows DDK MVP
> StorageCraft Corporation
> xxxxx@storagecraft.com
> http://www.storagecraft.com
>
> ----- Original Message -----
> From: “Lyndon J Clarke”
> Newsgroups: ntfsd
> To: “Windows File Systems Devs Interest List”
> Sent: Wednesday, April 27, 2005 1:22 PM
> Subject: [ntfsd] write throttle, memory mapped files, …
>
>
>> Gentlefolk
>>
>> The cc/mm throttle cached changes (writes) which come through the file
>> system to files when the number of available pages is low in order to
>> avoid
>> the system panic which arises when there are no available pages. It
>> appears
>> however that there is no mechanism to throttle cached changes which do
>> not
>> come through the file system because files are memory mapped. It might
>> therefore be possible to write some user mode programs which will cause
>> the
>> system panic; this is possible and can be demonstrated.
>>
>> The question I have is how does mm/cc decide whether to raise
>> NO_MORE_PAGES
>> or DIRTY_MAPPED_PAGES_CONGESTION?
>>
>> Cheers
>> Lyndon
>>
>>
>>
>> —
>> Questions? First check the IFS FAQ at
> https://www.osronline.com/article.cfm?id=17
>>
>> You are currently subscribed to ntfsd as: xxxxx@storagecraft.com
>> To unsubscribe send a blank email to xxxxx@lists.osr.com
>
>