cas

Get your dangerous XP Cache Filter here

Recommended Posts

cas:

The reason I still converted to Dynamic Disk was because the patch only fixed my write speeds. My read scores were still horribly low....

After the XP cachefilter patch I hit 5800 121

9800 262

I stopeed after that becuse the games would install pretty damned quick then dog a lot trying to load. MAybe it is just this card. I have had the same results when this was in my P4 machine (ABIT TH7II and 1.6A@2.3)

Thanks everyone :D

Share this post


Link to post
Share on other sites

While I empathize with your problem Psyco, I would hate to confuse other readers. Psyco's read speeds are capped at 10MB/s. This issue is specific to his configuration, and not endemic to Windows XP.

Share this post


Link to post
Share on other sites

YEah, that soed suck. But thank god for the dynamic disk conversion. After that and the above tweaks aI can run 47.8-48.5 reads and writes. HD Tach rteports a 50234 peak transfer rate and an access time of 4.1ms with 4 percent cpu usage. Thanks CAS!!! :o

Share this post


Link to post
Share on other sites

Cas wrote,

vltk wrote:

Nope. After added the service entries to the registry and only in case of coping XPCacheFilter.sys to the system32dllcache folder I can load the driver (sow it in the Device Manager under Non-Plug and Play Drivers). But after rebooting WinXPpro hang with above mentioned code.

You added the service entries, but not the LowerFilters entry, and got a 0x7B?

Without the LowerFilters entry, there is nothing to associate the filter with the boot device, so this doesn't make very much sense.

Not so. Of course I added LowerFilters then reboot and have 0x7B.

Share this post


Link to post
Share on other sites
Thank you for your very generous offer Frank.

I am inclined to hold off on any specific configuration, until I can pin down what the failures have in common.  I really need to be able to reproduce the error, before I can fix it. 

In the meantime, a request directed at users of XPCacheFilter:

If you are successfully using the software, please post here.

If you were unable to use the software, please post here.

It would be very helpful if you could include the following information:

  • 1. CPU

2. Chipset

3. SCSI HBA Model(s)

4. A complete list of all of the devices hanging off of your SCSI bus(es)

5. A list of other storage devices in use ( ATA, IEEE1394, or USB disks)

6. File systems in use

7. Nation in which you bought XP

8. A list of any service packs or hox fixes you have installed

  • vltk’s post is a good example(thank you).

If you don’t want to provide specific information for privacy reasons, please, leave it blank rather than making something up.

After Christmas, I will look at writing a more proper installation routine, which may address the issue.

Hi,

I am, as mentioned above, unable to get the XPCacheFilter working, i.e. it BSODs when booting up. Below are the information of my configuration:

1. 2 x Intel Pentium III-S 1.4GHz

2. VIA Apollo Pro266T

3. Adaptec ASC-29160

4. 0 - Fujitsu MAM3184MC

1 - Seagate ST336752LC

2 - Seagate ST336752LC

3 - Seagate ST336752LC

4 - Seagate ST336732LC

5. Pioneer DVD-105 on VIA onboard Primary Master

Yamaha CRW2100E on VIA onboard Secondary Master

IBM 45GB 75GXP on Highpoint HPT370A Primary Master (RAID 0)

IBM 45GB 75GXP on Highpoint HPT370A Secondary Master (RAID 0)

6. FAT32 for 2 x IBM 45GB 75GXP RAID 0 array

FAT32 for Fujitsu MAM3184MC (XP Boot Drive)

NTFS for 4 x Seagate drives (Dynamic)

7. Not Sure

8. SP1 and any other patches up to 25/12/2002

Share this post


Link to post
Share on other sites

As requested

It would be very helpful if you could include the following information:

1. CPU 

2. Chipset 

3. SCSI HBA Model(s) 

4. A complete list of all of the devices hanging off of your SCSI bus(es) 

5. A list of other storage devices in use ( ATA, IEEE1394, or USB disks) 

6. File systems in use 

7. Nation in which you bought XP 

8. A list of any service packs or hox fixes you have installed

Unsuccessfull, No BSOD but spontanious reboot during XP load

1. CPU - Dual PIII 1GB (coppermine)

2. Chipset - Via Apollo Pro133A (N 694x, S 686B)

3. SCSI - Mobo integrated with LSI53C1010-33 Ultra160/Ultra3

4. IBM 36LZX (DDYS 18350N)

5. Maxtor DiamondMax Plus 40 Ultra ATA 66 (51024U2)

6. Fat32 and NTFS (both on SCSI drive, Fat32 is boot)

7. UK

8. SP1

Share this post


Link to post
Share on other sites

I have updated XPCacheFilter and XPCacheFilterFiles (v0.2).

Since I have not successfully reproduced the problem, I have no idea if my changes will make a difference. The new driver is 25% smaller, which will save you a whopping ~1K!

Because I am lazy, updating the driver properly requires that you uninstall, reboot, install, reboot.

If you have been unsuccessful before, give the new version a try. Who knows, it might have some effect.

Share this post


Link to post
Share on other sites
System :

XP Sp1

Abit TH7II

512 ram

P4 1.6@2.4

Adaptec 26160 

Quantum Atlas V 9.1 GB

Works like a charm, cas I thank you so very much.

are you running your rdram at 3x or 4x?

Share this post


Link to post
Share on other sites
I'm running it at 3x.  I can get my Samsung to 125 x 4 but not to stable.

1.6@2.4=150fsb, 1.6=16x. 150*3=450, 150*4=600 - so it's either pc900 or pc1200, correct? where are you getting the 125?

Share this post


Link to post
Share on other sites

Yes you are correct, currently I'm at16x150=2400 Rdram at 3x. I have run my system at 16x125=2000 with Rdram at 4x, but the ram would crash xp at that speed.

Share this post


Link to post
Share on other sites
I have updated XPCacheFilter and XPCacheFilterFiles (v0.2). 

Since I have not successfully reproduced the problem, I have no idea if my changes will make a difference.  The new driver is 25% smaller, which will save you a whopping ~1K!

Because I am lazy, updating the driver properly requires that you uninstall, reboot, install, reboot. 

If you have been unsuccessful before, give the new version a try.  Who knows, it might have some effect.

Don’t help. The same BSOD. More over now I can’t load the driver to make it be a part of the storage stack. Don’t see it under Non-Plug and Play Drivers. Coping XPCF.sys driver into system32dllcache folder didn’t help with new driver.

Cas, but what is your WinXP-SCSI system configuration? May be it will be possible for me to change my configuration making it closer to yours.

Share this post


Link to post
Share on other sites

And oh, the shame

(He was ashamed!)

Thoughta changin' [its] name

But didn’t get around to it.

What an embarrassment. XPCacheFilter.zip has been updated, and should now work for those who insist on using FAT32 for their system volume.

Share this post


Link to post
Share on other sites

A hearty thanks to ebe, Psyco, vltk, Helpme, ALC-X, samiraj, Tualatin, nuanda, kbird, and those who sent me email. You guys were the key to identifying the source of the problem.

You should now look under Disk Drives in the Device Manager for evidence of installation; look under Driver Details.

Share this post


Link to post
Share on other sites

Just for everyones info if it has not been mentioned before....

I just installed .NetServer and they have addressed the problems and added another tab on the scsi disk in the polices tab under device manager. Its something snappy like "Advanced performance".....And with that checked we get the scsi performance back with basic disks. This was not that tough for them to fix. Hopefully XP/Win2k will get the same treatment.

Cheers

Tex

Share this post


Link to post
Share on other sites
A hearty thanks to ebe, Psyco, vltk, Helpme, ALC-X, samiraj, Tualatin, nuanda, kbird, and those who sent me email.  You guys were the key to identifying the source of the problem.

You should now look under Disk Drives in the Device Manager for evidence of installation; look under Driver Details.

No, Thank you for making it!

Should we use the never version of the filter driver if we're still using the inital release?

Share this post


Link to post
Share on other sites
A hearty thanks to ebe, Psyco, vltk, Helpme, ALC-X, samiraj, Tualatin, nuanda, kbird, and those who sent me email.  You guys were the key to identifying the source of the problem.

You should now look under Disk Drives in the Device Manager for evidence of installation; look under Driver Details.

Happy New Year for all in this forum! Much thanks for your Cas, you made a good work. At last it works for me. My ATTO scores W=R=58700 from 8 to 1024Kb. So now judging from ATTO we have normal (to be more accurate - equal to W2000) writes with blocks from 0.5 to 1024kb testing with 32Mb file under WinXP.

After testing with ATTO I tried with my dvd.vob file disk-to-self copying. And again have (the same as before) much worse (bigger time copying) results for *.vob file copying for SCSI 15K disk compare to 7200 IDE disk.

WinXPpro

C:! to C:=3m07s (~5.5Mb/s)

D:! to D:=2m13s (~7.9Mb/s) – pay attention, 7200 IDE disk 1.4 times quicker then 15.3 SCSI one.

C: to D: =36s (~28.4Mb/s)

So driver didn’t solve general problem – terrible disk-to-self file transfer SCSI performance compare to IDE under all Win NT cores. At last for the big (bigger then or 32Mb) continuous files. Don’t tell me that this is not very frequent case. All disk-to-self unraring, mpg4 and mpg3 encoding, video capture will suffer from this.

I want to remind shortly part of my former results for different OS before implying Cas filter.

W2000 and DOS7. W2K installed without SP on IDE disk, but with all latest chipset, AGP and devices drivers. DOS7(W98SE-command.com) installed on bootable diskette with himem.sys, Emm386.exe, smartdrive.exe /X /L 2048 2048 enabled in config.sys and autoexec.bat, the file manager - Volkov Commander (analog of Norton Commander) that include very general drivers for SCSI and IDE. For DOS tests I don’t make restarts and defragmentation after each copying, get tired. The rest details of my setup you can see above at this page.

ATTO SCSI drive results for W2K after installing it on IDE disk – Writes ~ 40Mb/s(curve with slope) and Reads ~ 56Mb/s. Results don’t change after installing of 4in1 VIA’s driver. Nothing changed after installing VIA’s Raid patch. But after installing of GB Latency Patch v.020 beta throughput increase for Writes to ~ 59.3 Mb/s and for Reads to ~ 59.4 Mb/s with straight curve from 4 to 1024Kb. ATTO was used with default parameters except file testing length set to 32Mb. So we come to ideal picture that people have testing performance for SCSI file transfer under W2K before installing SP3( I don’t do anything with different SP’s because don’t have them). Now look at the “real life” test results with 1Gb dvd.vob. To remind, disk C: - SCSI, disk D: - IDE. OS is always on a bootable disk.

WinXPpro

C:! to C:=3m04s (~5.6Mb/s)

D:! to D:=2m25s (~7.1Mb/s) – pay attention,7200 IDE disk quicker then 15.3 SCSI one.

D: to C: =34s (~30.8Mb/s) – throughput only limited by velocity of sequential read of IDE disk.

Win2000pro

C:! to C:=3m14s (~5.3Mb/s)

D:! to D:=2m14s (~6.7Mb/s) – pay attention,7200 IDE disk quicker then 15.3 SCSI one.

D: to C: =39s (~27.2Mb/s)

DOS, floppy boot.

C:! to C:=1m33s (~11.0Mb/s)

D:! to D:=2m25s (~7.1Mb/s) – VICE VERSA and CORRECT. 7200 IDE disk slower then 15.3 SCSI one.

D: to C: =53s (~19.3Mb/s)

I don’t see any drops in read-write SCSI performance under WinXPpro for disk to disk file transfer data. The other peoples file transfer results from one SCSI 15K disk to another SCSI 15K disk attached to two separate SCSI adapters, to two independent channels of one adapter or in worst case to one channel of adapter don’t have any performance hits.

For the test of SCSI one disk file transfer (disk-to-self ) anyone can find out that results will be 3-5 times worse then for disk to disk file transfer under any windows operation systems (understandable – disk’s head cannot simultaneously read and write plus separate time needed for head’s positioning). The same for IDE. In general WinXP make SCSI disk-to-self operation quicker then W2K. But both systems terribly slow compare to the same for IDE disk-to-self file transfer. But in real we have in 1.5 time worse throughput for SCSI 15K disk compare to IDE 7.2K one. And in DOS case with non specific very general SCSI driver we have throughput near 2 times better for disk-to-self SCSI file transfer (compare to WinXP) and 1.55 times better then disk-to-self IDE file transfer. It is possible that in SCSI case MS sacrifice with performance of “not often utilized operation” self-to-disk BIG file copying (bigger then 1Mb) on NT core + I/O interface in name of security, reliability and ets. But if we have user with only one big SCSI disk it will be for him “very often utilized” class of operations.

I add Sandra’s results. They quite normal for test file 1023Mb and sequential reads and writes.

Sandra’s Pro details for SCSI disk (Cas filter don’t have any affect on them for me):

Windows Disk Cache Used : No

Use Overlapped I/O : Yes

Command Queue Depth : 4 command(s)

Test File Size : 1023MB

File Server Optimized : No

Benchmark Breakdown

Buffered Read : 94 MB/s

Sequential Read : 56 MB/s

Random Read : 13 MB/s

Buffered Write : 53 MB/s

Sequential Write : 56 MB/s

Random Write : 19 MB/s

Average Access Time : 4 ms (estimated)

It’s very simple to check these results for all who have at last one SCSI and one IDE disk.

Cas, what do you think about all this. Is it possible to write some fix to make disk.sys behave the same for SCSI as for IDE foer disk-to-self copying. Or may be you have some other ideas.

What think about this MS – all WinXP (W2000, possible NT) SCSI users suffer from this from the very begin of existing of these OS. And what can tell us MS – reliability by expense of performance or what?

Share this post


Link to post
Share on other sites
While I empathize with your problem Psyco, I would hate to confuse other readers.  Psyco's read speeds are capped at 10MB/s.  This issue is specific to his configuration, and not endemic to Windows XP.
:cry:

I really wish I could figure out this problem Psyco has, as I share it! Both writes and reads capped at around 10-11 MB/s, although using the cachefilter driver [bless You, cas] cured the writes after a fashion -- putting the max writes up around 50MB/s in ATTO. Incidentally, I went ahead and uninstalled, as I am an avid gamer off hours and two of my currrent favorite 3D games developed severe graphics problems after the cachefilter install and one would refuse to load and dump me back to the desktop each time [FYI if you run into it, in Operation Flash Point the message is "cannot memory-map file RESAddonsO.pbo"]. The uninstall immediately cured the problems [and so might a reinstall of the games, I suppose]. So I may use cas' driver as needed if it provides sufficient advantages. I am at a total loss as to why the crippled reads. I am either going to go back to a multiple boot with Win2000 or 98SE or upgrade to XP Pro -- my XP Home [free with a HDD and worth every penny] with no dynamic disk option sucks.

Just for the record:

1. CPU PentiumIII 933 fcpga

2. Chipset i815e on ASUS CUSL2-C

3. SCSI HBA Model(s) Adaptec 19160

4. SCSI devices: 2@ Maxtor 10KIII 18G U160 [1 with O/S], Quantum10K 18G U160, + Kenwood CD-ROM & Toshiba DVD

5. other storage devices in use: Maxtor D740X 80G ATA133 & Plextor CD-RW

6. FAT & FAT32 (hard-to-wean DriveImage junkie)

7. USA

8. XP Home: SP1

Share this post


Link to post
Share on other sites
Cas, what do you think about all this. Is it possible to write some fix to make disk.sys behave the same for SCSI as for IDE foer disk-to-self copying. Or may be you have some other ideas.

I really don't know.

Without having independently confirmed the issue, I would guess that differences in the way requests were broken up in the disk or port drivers, could have some effect on the number of seeks required for a disk to self copy.

You may want to try cascopy to set a baseline.

Share this post


Link to post
Share on other sites
... two of my currrent favorite 3D games developed severe graphics problems after the cachefilter install ... The uninstall immediately cured the problems

Ouch.

Let me look into this.

Share this post


Link to post
Share on other sites
While I empathize with your problem Psyco' date=' I would hate to confuse other readers.  Psyco's read speeds are capped at 10MB/s.  This issue is specific to his configuration, and not endemic to Windows XP  

I really wish I could figure out this problem Psyco has, as I share it!  Both writes and reads capped at around 10-11 MB/s, although using the cachefilter driver [bless You, cas'] cured the writes after a fashion -- putting the max writes up around 50MB/s in ATTO.  .

:cry:

My LSI u160 was also like yours in Win XP. Reads always even with the sorry writes and I had seen no one that had my prob till now.

I emailed cas about moding his filter for the reads also. Untill .Net Server comes or MS decides to implement the same fix in XP for us... There is only the dynamic disk to rescue us... (grin) and the funny thing is they also have the dynamic disk bug fixed in .Net server so changing the disk to dynamic won't fix the prob... But we have the new "advanced performance " tab to our rescue.

The good news is MS is still taking volunteers to Uhhh.. "test" .Net Advanced server in the beta test program and it has looked nice so far.

I like it. Its like a win XP with less of the XP gui and more of the win2k interface with very cool new wizards. Before you walked thru all the steps to set up Win2k .. domain... dhcp... dns..rounting Active Directory...

Now they have a wizrd that takes 3 minutes and asks you what you want to name the Domain... And if you want it to install rounting for ya..Jus. Al the rest is automatic. Just sweet.

So far .Net Server has been very sweet for me. I love the cleartype fonts and had hopped I wouldn't have to spend considerable time on a win2k deaktop again but since we do Oracle work it was gonna happen... (grin) But .Net Server may be my answer... And unlike the XP beta I received this one is time bombed to die after a year not 3 months.

Tex

Share this post


Link to post
Share on other sites

vltk

I'm wondering if your problem is swap file related. cas could probably advise if I'm off the mark here, but I believe copying files involves the swap file, so if the swap file is on the disk that's performing the same-disk copy it may impact performance? Try moving your swap file to the IDE disk and run the same tests. See how much impact it has on both the SCSI and IDE results.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now