Windows 2000 twice slower then XP for SCSI_PASS_THROUGH_DIRECT

Hello All,

I am using SCSI_PASS_THROUGH_DIRECT to write data on an optical media. The device is in CD-ROM
device stack. I do not use IRP_MJ_WRITE path, because I need sense info from the command.

I have observed a strange difference between Windows 2000 and Windows XP – writing time on Windows
2000 in ~2.5 times bigger then on Windows XP.

I checked this with True Time profiler (from CompuWare) and upper level filter driver. I put probe
point just before IoCallDriver and at beginning of completion routine for this request. I made the
measure on more than 10 000 writes. The write is done without interruption, there should be no
influence of device seek. The result is an average measure over all the commands.

Why is there such a big difference in execution time?! And of course, how could I get same speed
on Windows 2000 as I have on Windows XP?


Do you Yahoo!?
Yahoo! Mail - Helps protect you from nasty viruses.
http://promotions.yahoo.com/new_mail

Hi,

did you try using smaller buffer when sending WRITE10/WRITE12 to CD/DVD
burner? Decreasing
buffer to 32-60K usually helps.

Regards,
Anton A. Kolomyeytsev

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Friday, June 25, 2004 1:20 PM
To: Windows System Software Devs Interest List
Subject: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello All,

I am using SCSI_PASS_THROUGH_DIRECT to write data on an optical media.
The device is in CD-ROM device stack. I do not use IRP_MJ_WRITE path,
because I need sense info from the command.

I have observed a strange difference between Windows 2000 and Windows XP

  • writing time on Windows 2000 in ~2.5 times bigger then on Windows XP.

I checked this with True Time profiler (from CompuWare) and upper level
filter driver. I put probe point just before IoCallDriver and at
beginning of completion routine for this request. I made the measure on
more than 10 000 writes. The write is done without interruption, there
should be no influence of device seek. The result is an average measure
over all the commands.

Why is there such a big difference in execution time?! And of course,
how could I get same speed on Windows 2000 as I have on Windows XP?


Do you Yahoo!?
Yahoo! Mail - Helps protect you from nasty viruses.
http://promotions.yahoo.com/new_mail


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

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

Anton A. Kolomyeytsev> Hi,

Anton A. Kolomyeytsev> did you try using smaller buffer when sending WRITE10/WRITE12 to CD/DVD
Anton A. Kolomyeytsev> burner? Decreasing
Anton A. Kolomyeytsev> buffer to 32-60K usually helps.

Anton A. Kolomyeytsev> Regards,
Anton A. Kolomyeytsev> Anton A. Kolomyeytsev

I am using buffer size equal to 64K.
Hence buffer size not a problem.


Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail

Hello,

Additionally, when writing a packet-formatted media, you must write by
packet-size (64 K on CD, 32 K on DVD). The device will not accept smaller
size than a packet.

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Friday, June 25, 2004 5:39 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Anton A. Kolomyeytsev> Hi,

Anton A. Kolomyeytsev> did you try using smaller buffer when sending
WRITE10/WRITE12 to CD/DVD
Anton A. Kolomyeytsev> burner? Decreasing
Anton A. Kolomyeytsev> buffer to 32-60K usually helps.

Anton A. Kolomyeytsev> Regards,
Anton A. Kolomyeytsev> Anton A. Kolomyeytsev

I am using buffer size equal to 64K.
Hence buffer size not a problem.


Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!


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

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

David,

original poster told nothing about PAO recording method.

From what we've seen quite a lot of ATAPI devices get DMA problems with
larger buffers sizes
(report check condition with SCSI sense filled with zeros). Esp. if
ATAPI<->USB or FireWire
bridges used. Just my 2 cents ))

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of David Burg
Sent: Friday, June 25, 2004 9:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello,

Additionally, when writing a packet-formatted media, you must write by
packet-size (64 K on CD, 32 K on DVD). The device will not accept
smaller size than a packet.

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Friday, June 25, 2004 5:39 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Anton A. Kolomyeytsev> Hi,

Anton A. Kolomyeytsev> did you try using smaller buffer when sending
WRITE10/WRITE12 to CD/DVD Anton A. Kolomyeytsev> burner? Decreasing
Anton A. Kolomyeytsev> buffer to 32-60K usually helps.

Anton A. Kolomyeytsev> Regards,
Anton A. Kolomyeytsev> Anton A. Kolomyeytsev

I am using buffer size equal to 64K.
Hence buffer size not a problem.


Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!


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

You are currently subscribed to ntdev as: xxxxx@nero.com
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=256

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

Hello Anton,

To clarify things:

  1. Yes, David Burg was right, I am using PAO recording method.
  2. I am using normal IDE device and this is DVD+RW device with DVD+RW media inserted.

Regards,
Al Shatilo

Friday, June 25, 2004, 10:43:56 PM, you wrote:

David,

original poster told nothing about PAO recording method.

From what we've seen quite a lot of ATAPI devices get DMA problems with
larger buffers sizes
(report check condition with SCSI sense filled with zeros). Esp. if
ATAPI<->USB or FireWire
bridges used. Just my 2 cents ))

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of David Burg
Sent: Friday, June 25, 2004 9:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello,

Additionally, when writing a packet-formatted media, you must write by
packet-size (64 K on CD, 32 K on DVD). The device will not accept
smaller size than a packet.

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Friday, June 25, 2004 5:39 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Anton A. Kolomyeytsev> Hi,

Anton A. Kolomyeytsev> did you try using smaller buffer when sending
WRITE10/WRITE12 to CD/DVD Anton A. Kolomyeytsev> burner? Decreasing
Anton A. Kolomyeytsev> buffer to 32-60K usually helps.

Anton A. Kolomyeytsev> Regards,
Anton A. Kolomyeytsev> Anton A. Kolomyeytsev

I am using buffer size equal to 64K.
Hence buffer size not a problem.


Do you Yahoo!?
Yahoo! Mail Address AutoComplete - You start. We finish.

Maybe next time you'd be so kind to tell everything by yourself and not
force others to GUESS :slight_smile:

You can send me your binary -- I can try to experiment with some of my
"hardware zoo" DVD+RW
burners and BusHound. You can also try it to see -- is this device
itself inserting delays into
command execution or SCSI command execution time is the same under all
of the OSes and
this is somebody (filter driver?) in the storage stack adding overhead.

From what I see now (we also have a CD/DVD burning engine working in NT
kernel) 2000 and XP
work identically -- no differences noticed. However we use TAO/SAO/DAO
and not PAO so the
situation is not equal.

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Saturday, June 26, 2004 4:45 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello Anton,

To clarify things:

  1. Yes, David Burg was right, I am using PAO recording method. 2. I am
    using normal IDE device and this is DVD+RW device with DVD+RW media
    inserted.

Regards,
Al Shatilo

Friday, June 25, 2004, 10:43:56 PM, you wrote:

David,

original poster told nothing about PAO recording method.

From what we've seen quite a lot of ATAPI devices get DMA problems with
larger buffers sizes
(report check condition with SCSI sense filled with zeros). Esp. if
ATAPI<->USB or FireWire
bridges used. Just my 2 cents ))

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of David Burg
Sent: Friday, June 25, 2004 9:21 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello,

Additionally, when writing a packet-formatted media, you must write by
packet-size (64 K on CD, 32 K on DVD). The device will not accept
smaller size than a packet.

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Friday, June 25, 2004 5:39 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Anton A. Kolomyeytsev> Hi,

Anton A. Kolomyeytsev> did you try using smaller buffer when sending
WRITE10/WRITE12 to CD/DVD Anton A. Kolomyeytsev> burner? Decreasing
Anton A. Kolomyeytsev> buffer to 32-60K usually helps.

Anton A. Kolomyeytsev> Regards,
Anton A. Kolomyeytsev> Anton A. Kolomyeytsev

I am using buffer size equal to 64K.
Hence buffer size not a problem.


Do you Yahoo!?
Yahoo! Mail Address AutoComplete - You start. We finish.


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

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

Hello Anton,

Saturday, June 26, 2004, 10:04:46 PM, you wrote:

Anton A. Kolomyeytsev> Maybe next time you’d be so kind to tell everything by yourself and not
Anton A. Kolomyeytsev> force others to GUESS :slight_smile:

I ask initial question and after add on-request additional information.
This is usual for any question.

Anton A. Kolomyeytsev> You can send me your binary – I can try to experiment with some of my
Anton A. Kolomyeytsev> “hardware zoo” DVD+RW
Anton A. Kolomyeytsev> burners and BusHound. You can also try it to see – is this device
Anton A. Kolomyeytsev> itself inserting delays into
Anton A. Kolomyeytsev> command execution or SCSI command execution time is the same under all
Anton A. Kolomyeytsev> of the OSes and
Anton A. Kolomyeytsev> this is somebody (filter driver?) in the storage stack adding overhead.

I have huge “hardware zoo” as well. My question not about testing on your side, I did tests on my
side already and put results into mailing list. I checked the issue on several burners - same
result.
Question is about different performance during SCSI_PASS_THROUGH_DIRECT request on Windows 2000
and Windows XP.

Anton A. Kolomyeytsev> From what I see now (we also have a CD/DVD burning engine working in NT
Anton A. Kolomyeytsev> kernel) 2000 and XP
Anton A. Kolomyeytsev> work identically – no differences noticed. However we use TAO/SAO/DAO
Anton A. Kolomyeytsev> and not PAO so the
Anton A. Kolomyeytsev> situation is not equal.

Do you use SCSI_PASS_THROUGH_DIRECT with buffers equal to 64K in your software?

Best regards,
Al Shatilo


Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail

yes

just rebuilt the code to check 2000 SP6, XP SP1 and 2003 on the same
hardware

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com – Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Tuesday, June 29, 2004 8:28 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello Anton,

Saturday, June 26, 2004, 10:04:46 PM, you wrote:

Anton A. Kolomyeytsev> Maybe next time you’d be so kind to tell
everything by yourself and not Anton A. Kolomyeytsev> force others to
GUESS :slight_smile:

I ask initial question and after add on-request additional information.
This is usual for any question.

Anton A. Kolomyeytsev> You can send me your binary – I can try to
experiment with some of my Anton A. Kolomyeytsev> “hardware zoo” DVD+RW
Anton A. Kolomyeytsev> burners and BusHound. You can also try it to see
– is this device Anton A. Kolomyeytsev> itself inserting delays into
Anton A. Kolomyeytsev> command execution or SCSI command execution time
is the same under all Anton A. Kolomyeytsev> of the OSes and Anton A.
Kolomyeytsev> this is somebody (filter driver?) in the storage stack
adding overhead.

I have huge “hardware zoo” as well. My question not about testing on
your side, I did tests on my side already and put results into mailing
list. I checked the issue on several burners - same result. Question is
about different performance during SCSI_PASS_THROUGH_DIRECT request on
Windows 2000 and Windows XP.

Anton A. Kolomyeytsev> From what I see now (we also have a CD/DVD
burning engine working in NT Anton A. Kolomyeytsev> kernel) 2000 and XP
Anton A. Kolomyeytsev> work identically – no differences noticed.
However we use TAO/SAO/DAO Anton A. Kolomyeytsev> and not PAO so the
Anton A. Kolomyeytsev> situation is not equal.

Do you use SCSI_PASS_THROUGH_DIRECT with buffers equal to 64K in your
software?

Best regards,
Al Shatilo


Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!
http://promotions.yahoo.com/new_mail


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

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

Hello Anton,

You mean Windows 2000 SP4? Yes, the defect is visible there. A rebuild does
not affect the defect - our build server do build from clean checkout each
new release. Al is testing always on the same computer (although the defect
is visible on multiple computers). We are not testing with Windows 2003
Server as the performance difference between XP and 2k is already enough for
us to need resolution.

(In case it was not yet obvious, I am working with Al Shatilo.)

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG phone: +49 (0)7248 911 862 (direct line)
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Anton A.
Kolomyeytsev (CoolDev.Com)
Sent: Tuesday, June 29, 2004 7:32 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

yes

just rebuilt the code to check 2000 SP6, XP SP1 and 2003 on the same
hardware

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Tuesday, June 29, 2004 8:28 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello Anton,

Saturday, June 26, 2004, 10:04:46 PM, you wrote:

Anton A. Kolomyeytsev> Maybe next time you'd be so kind to tell
everything by yourself and not Anton A. Kolomyeytsev> force others to
GUESS :slight_smile:

I ask initial question and after add on-request additional information.
This is usual for any question.

Anton A. Kolomyeytsev> You can send me your binary -- I can try to
experiment with some of my Anton A. Kolomyeytsev> "hardware zoo" DVD+RW
Anton A. Kolomyeytsev> burners and BusHound. You can also try it to see
-- is this device Anton A. Kolomyeytsev> itself inserting delays into
Anton A. Kolomyeytsev> command execution or SCSI command execution time
is the same under all Anton A. Kolomyeytsev> of the OSes and Anton A.
Kolomyeytsev> this is somebody (filter driver?) in the storage stack
adding overhead.

I have huge "hardware zoo" as well. My question not about testing on
your side, I did tests on my side already and put results into mailing
list. I checked the issue on several burners - same result. Question is
about different performance during SCSI_PASS_THROUGH_DIRECT request on
Windows 2000 and Windows XP.

Anton A. Kolomyeytsev> From what I see now (we also have a CD/DVD
burning engine working in NT Anton A. Kolomyeytsev> kernel) 2000 and XP
Anton A. Kolomyeytsev> work identically -- no differences noticed.
However we use TAO/SAO/DAO Anton A. Kolomyeytsev> and not PAO so the
Anton A. Kolomyeytsev> situation is not equal.

Do you use SCSI_PASS_THROUGH_DIRECT with buffers equal to 64K in your
software?

Best regards,
Al Shatilo


Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!


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

You are currently subscribed to ntdev as: xxxxx@cooldev.com
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=256

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

David,

of course SP4 (SP6 -- just my mind eclipse).

I was talking about our own code -- not yours. However as I've told
before we do not
use PAO (I've checked SAO and TAO). Burning speed is the same.

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of David Burg
Sent: Tuesday, June 29, 2004 8:52 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello Anton,

You mean Windows 2000 SP4? Yes, the defect is visible there. A rebuild
does not affect the defect - our build server do build from clean
checkout each new release. Al is testing always on the same computer
(although the defect is visible on multiple computers). We are not
testing with Windows 2003 Server as the performance difference between
XP and 2k is already enough for us to need resolution.

(In case it was not yet obvious, I am working with Al Shatilo.)

Best regards,

David Burg


David Burg
Software Development,
InCD Project Leader

Ahead Software AG phone: +49 (0)7248 911 862 (direct line)
Im Stoeckmaedle 18 fax: +49 (0)7248 911 888
76307 Karlsbad email: xxxxx@nero.com
Germany http://www.nero.com

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Anton A.
Kolomyeytsev (CoolDev.Com)
Sent: Tuesday, June 29, 2004 7:32 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

yes

just rebuilt the code to check 2000 SP6, XP SP1 and 2003 on the same
hardware

Regards,
Anton A. Kolomyeytsev

RocketDivision.Com -- Toolkits for Network and Storage Kernel Software
Developers

-----Original Message-----
From: xxxxx@lists.osr.com
[mailto:xxxxx@lists.osr.com] On Behalf Of Al Shatilo
Sent: Tuesday, June 29, 2004 8:28 PM
To: Windows System Software Devs Interest List
Subject: RE: [ntdev] Windows 2000 twice slower then XP for
SCSI_PASS_THROUGH_DIRECT

Hello Anton,

Saturday, June 26, 2004, 10:04:46 PM, you wrote:

Anton A. Kolomyeytsev> Maybe next time you'd be so kind to tell
everything by yourself and not Anton A. Kolomyeytsev> force others to
GUESS :slight_smile:

I ask initial question and after add on-request additional information.
This is usual for any question.

Anton A. Kolomyeytsev> You can send me your binary -- I can try to
experiment with some of my Anton A. Kolomyeytsev> "hardware zoo" DVD+RW
Anton A. Kolomyeytsev> burners and BusHound. You can also try it to see
-- is this device Anton A. Kolomyeytsev> itself inserting delays into
Anton A. Kolomyeytsev> command execution or SCSI command execution time
is the same under all Anton A. Kolomyeytsev> of the OSes and Anton A.
Kolomyeytsev> this is somebody (filter driver?) in the storage stack
adding overhead.

I have huge "hardware zoo" as well. My question not about testing on
your side, I did tests on my side already and put results into mailing
list. I checked the issue on several burners - same result. Question is
about different performance during SCSI_PASS_THROUGH_DIRECT request on
Windows 2000 and Windows XP.

Anton A. Kolomyeytsev> From what I see now (we also have a CD/DVD
burning engine working in NT Anton A. Kolomyeytsev> kernel) 2000 and XP
Anton A. Kolomyeytsev> work identically -- no differences noticed.
However we use TAO/SAO/DAO Anton A. Kolomyeytsev> and not PAO so the
Anton A. Kolomyeytsev> situation is not equal.

Do you use SCSI_PASS_THROUGH_DIRECT with buffers equal to 64K in your
software?

Best regards,
Al Shatilo


Do you Yahoo!?
Yahoo! Mail - 50x more storage than other providers!


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

You are currently subscribed to ntdev as: xxxxx@cooldev.com
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=256

You are currently subscribed to ntdev as: xxxxx@nero.com
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=256

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

The solution was simple as usually. I just start check driver versions on Windows 2000/XP and by
chance found the problem. It was setting for Secondary IDE Channel. Transfer mode on Windows 2000
was set, by default, to “PIO only”. On Windows XP the transfer mode for Secondary IDE Channel set,
by default, to “DMA if available”.
The execution time became same after I set Secondary IDE Channel on Windows 2000 into “DMA if
available”.


Do you Yahoo!?
New and Improved Yahoo! Mail - 100MB free storage!
http://promotions.yahoo.com/new_mail