As long as the offset is page-aligned, it’s certainly legal to send down
both read and write paging I/O request with non-aligned lengths (as long
as offset+length>=ValidDataLength for writes). If the standard
filesystems can handle it, then so should any filter layered on top of
them.
-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Tony Mason
Sent: Friday, January 10, 2003 2:17 PM
To: File Systems Developers
Subject: [ntfsd] RE: Can I presume the size of buffer of
Paging IO is mult iple of 4K?
This does not square with my observations of behavior - I
have seen smaller paging I/O operations than 4KB, but never
smaller than 512 bytes. A non-cached I/O smaller than sector
size isn’t really valid anyway.
Regards,
Tony
Tony Mason
Consulting Partner
OSR Open Systems Resources, Inc.
http://www.osr.com
-----Original Message-----
From: Cary Fu [mailto:xxxxx@authenex.com]
Sent: Friday, January 10, 2003 2:08 PM
To: File Systems Developers
Subject: [ntfsd] Can I presume the size of buffer of Paging
IO is multiple of 4K?
Hi all,
I am developing a file system filter driver and it seems
to work very well in most cases. In my driver, I always
presume the size of Read/Write buffer size is multiple of 4K
when paging IO. Is that right? If not, can I force IO
manager always creates multiple of 4K buffer?
Thanks in advance.
Cary
You are currently subscribed to ntfsd as: xxxxx@osr.com
To unsubscribe send a blank email to xxxxx@lists.osr.com
You are currently subscribed to ntfsd as: xxxxx@nryan.com
To unsubscribe send a blank email to xxxxx@lists.osr.com
The Mm (MiWaitForInPageComplete) postprocesses the paging read as
following:
- if it returned STATUS_END_OF_FILE - treat as success, and zero all
pages
- if it returned Information < requested length - then zero the tail.
Max
----- Original Message -----
From: “Nicholas Ryan”
To: “File Systems Developers”
Sent: Monday, January 13, 2003 9:15 AM
Subject: [ntfsd] RE: Can I presume the size of buffer of Paging IO is
mult iple of 4K?
> As long as the offset is page-aligned, it’s certainly legal to send
down
> both read and write paging I/O request with non-aligned lengths (as
long
> as offset+length>=ValidDataLength for writes). If the standard
> filesystems can handle it, then so should any filter layered on top
of
> them.
>
> - Nicholas Ryan
>
>
>
> > -----Original Message-----
> > From: xxxxx@lists.osr.com
> > [mailto:xxxxx@lists.osr.com] On Behalf Of Tony Mason
> > Sent: Friday, January 10, 2003 2:17 PM
> > To: File Systems Developers
> > Subject: [ntfsd] RE: Can I presume the size of buffer of
> > Paging IO is mult iple of 4K?
> >
> >
> > This does not square with my observations of behavior - I
> > have seen smaller paging I/O operations than 4KB, but never
> > smaller than 512 bytes. A non-cached I/O smaller than sector
> > size isn’t really valid anyway.
> >
> > Regards,
> >
> > Tony
> >
> > Tony Mason
> > Consulting Partner
> > OSR Open Systems Resources, Inc.
> > http://www.osr.com
> >
> >
> > -----Original Message-----
> > From: Cary Fu [mailto:xxxxx@authenex.com]
> > Sent: Friday, January 10, 2003 2:08 PM
> > To: File Systems Developers
> > Subject: [ntfsd] Can I presume the size of buffer of Paging
> > IO is multiple of 4K?
> >
> > Hi all,
> > I am developing a file system filter driver and it seems
> > to work very well in most cases. In my driver, I always
> > presume the size of Read/Write buffer size is multiple of 4K
> > when paging IO. Is that right? If not, can I force IO
> > manager always creates multiple of 4K buffer?
> >
> > Thanks in advance.
> >
> > Cary
> >
> >
> >
> > —
> > You are currently subscribed to ntfsd as: xxxxx@osr.com
> > To unsubscribe send a blank email to
xxxxx@lists.osr.com
> >
> > —
> > You are currently subscribed to ntfsd as: xxxxx@nryan.com
> > To unsubscribe send a blank email to
xxxxx@lists.osr.com
> >
>
>
>
> —
> You are currently subscribed to ntfsd as: xxxxx@storagecraft.com
> To unsubscribe send a blank email to xxxxx@lists.osr.com
>