Jump to content

Get Windows XP x86 to recognize more than 4Gb with PAE?


AnX

Recommended Posts


Hey peeps,

everyone should be aware and understand the (probable) reasons why the (stupid) limitation has been put into place (besides commercial reasons).

What happened at the time was that most hardware (but not the high-grade hardware used on server machines) was somehow not fully compliant with the requisites for the PAE or the drivers (chipset/motherboard/etc.) were poorly designed/implemented.

The above led to have the first "user grade" systems with 4 Gb or more to be unstable.

To be fair, at the time very few "user grade" machine had (in 2002/2003) more than 1 or 2 Gb of Ram, so the issues were not that many.

With SP2 (2004) instead of actually tackling and solve the issues, the good MS guys decided to limit the amount of RAM and reserve the PAE , etc. to the Server OS (that are supposed to run on "better" hardware with "better" drivers).

There is no reason why this or that "patch" won't have very different results in practice on different systems, the *whatever* makes the OS unstable on a given hardware is not corrected, so someone will have the patch working perfectly on a specific system and someone else will experiment corruption or blue screens with the same patch on another system.

jaclaz
 

Link to comment
Share on other sites

On 14/06/2016 at 5:31 AM, jaclaz said:

Hey peeps,

everyone should be aware and understand the (probable) reasons why the (stupid) limitation has been put into place (besides commercial reasons).

What happened at the time was that most hardware (but not the high-grade hardware used on server machines) was somehow not fully compliant with the requisites for the PAE or the drivers (chipset/motherboard/etc.) were poorly designed/implemented.

The above led to have the first "user grade" systems with 4 Gb or more to be unstable.

To be fair, at the time very few "user grade" machine had (in 2002/2003) more than 1 or 2 Gb of Ram, so the issues were not that many.

With SP2 (2004) instead of actually tackling and solve the issues, the good MS guys decided to limit the amount of RAM and reserve the PAE , etc. to the Server OS (that are supposed to run on "better" hardware with "better" drivers).

There is no reason why this or that "patch" won't have very different results in practice on different systems, the *whatever* makes the OS unstable on a given hardware is not corrected, so someone will have the patch working perfectly on a specific system and someone else will experiment corruption or blue screens with the same patch on another system.

jaclaz
 

fRiends , You guys and gals wonder to know that MS also put some extra code to prevent you from using /PAE even you unlock it so you will face problem.

Donot believe me , Compare hal/krnl of XP SP1 with that of XPSP3

Link to comment
Share on other sites

  • 5 months later...

I ran some experiments to isolate the USB formatting problem. The problem is not in the Patches. It is in the USBPORT.SYS Driver. Using the HAL.DLL from SP1 did not solve the problem even though it supports >4GiB.

Microsoft was correct about buggy Drivers being a problem with using 64-Bit RAM. Of course they did not mention that they wrote the buggy Drivers.

As everyone knows by now, Microsoft's standard operating procedure is to limit functionality to avoid a problem rather than putting effort into fixing the cause.
At least until up-sell time.

The simple Patch may be overkill as it causes double buffering all the time, but it does not have much impact because Windows seems to use the 64-Bit memory first. Even with only 4GiB of total RAM (less than 1 GiB of 64-Bit RAM) the USB Driver was already being Double buffered.

Edited by rloew
Link to comment
Share on other sites

21 hours ago, rloew said:

I ran some experiments to isolate the USB formatting problem. The problem is not in the Patches. It is in the USBPORT.SYS Driver. Using the HAL.DLL from SP1 did not solve the problem even though it supports >4GiB.

Microsoft was correct about buggy Drivers being a problem with using 64-Bit RAM. Of course they did not mention that they wrote the buggy Drivers.

As everyone knows by now, Microsoft's standard operating procedure is to limit functionality to avoid a problem rather than putting effort into fixing the cause.
At least until up-sell time.

The simple Patch may be overkill as it causes double buffering all the time, but it does not have much impact because Windows seems to use the 64-Bit memory first. Even with only 4GiB of total RAM (less than 1 GiB of 64-Bit RAM) the USB Driver was already being Double buffered.

I also tried to fix format problem but failed.

It is better to use HP USB Format tool.

Thanks Rloew for your all effort.

Link to comment
Share on other sites

I would be curious to know something more about this "formatting" problem.

What it is related to exactly? :unsure:

1) formatting a volume on a USB partitioned device

2) formatting a USB "removable" device

3) partitioning a USB device

What is the error/problem?

Using disk manager or diskpart or format.com?

Does it affect all USB ports or just USB3 ones?

jaclaz
 

Link to comment
Share on other sites

3 hours ago, jaclaz said:

I would be curious to know something more about this "formatting" problem.

What it is related to exactly? :unsure:

1) formatting a volume on a USB partitioned device

2) formatting a USB "removable" device

3) partitioning a USB device

What is the error/problem?

Using disk manager or diskpart or format.com?

Does it affect all USB ports or just USB3 ones?

jaclaz
 

USB 3.0 never face it. Drive format also stuck like usb 2.0 . Bit confusing.

Link to comment
Share on other sites

The problem occurs, as reported previously, when quick formatting a USB Drive from My Computer.
There is some activity. Then it appears to trash the first Sector of the PBR. Then it hangs.
Cancelling the Format produces an error and leaves the device locked out. System will not shutdown properly afterwards.

I have not determined the precise nature of the problem. Reading and Writing from anywhere in my 32GB of RAM works fine. It may have to do with one of the other SCSI commands.

The only solution I know of is to replace USBPORT.SYS with the Windows 2003 Version as reported earlier.
USBHUB.SYS does not seem to need replacement.

Link to comment
Share on other sites

3 hours ago, rloew said:

The problem occurs, as reported previously, when quick formatting a USB Drive from My Computer.
There is some activity. Then it appears to trash the first Sector of the PBR. Then it hangs.
Cancelling the Format produces an error and leaves the device locked out. System will not shutdown properly afterwards.

I have not determined the precise nature of the problem. Reading and Writing from anywhere in my 32GB of RAM works fine. It may have to do with one of the other SCSI commands.

The only solution I know of is to replace USBPORT.SYS with the Windows 2003 Version as reported earlier.
USBHUB.SYS does not seem to need replacement.

Yep :), but I was asking if this happened on a non-partitioned device (a "Normal" USB stick which normally comes from factory as "removable") i.e. a device which first sector is the PBR or it it happened with USB hard disks (that are always a "fixed" device) i.e. a device that has one or more  partition created and that has the MBR as first sector and the PBR at a given offset (typically 63 sectors) or if it happened with a "removable" but partitioned device.

The reference by Dybia to the HP USB Format utility is "queer" as that tool does not do the same things as the built-in Windows XP tools.

By default in Windows XP:

1) if the device is "removable", the USB device is given "directly" a drive letter and is formatted (by format.com) as "superfloppy".
2) if the device is "fixed", the USB device needs to be partitioned (by diskpart or disk manager) then format.com is used on the partition(s)

The HP USB format tool creates (often with wrong, unbalanced CHS/LBA data) a partition on the ("removable") device, then formats it, the result is DIFFERENT from what you can do with built-in tools.

If you prefer, maybe  using a filter driver like cfadisk.sys or dummydisk.sys or diskmod.sys would change something (or maybe it won't) :unsure:.

Anyway, I failed to understand that once USBPORT.SYS from Windows 2003 is used the problem is solved :thumbup.

jaclaz
 

Link to comment
Share on other sites

3 hours ago, Dibya said:

I will try replacing with srv2003 version of cfadisk.sys or dummydisk.sys or diskmod.sys

Replace WHAT? :w00t:

With WHICH version? :ph34r:

It seems like you are not very familiar with the concepts of partitioning and formatting (USB) devices and with the ideas behind the mentioned filter drivers. :dubbio:

jaclaz
 

Link to comment
Share on other sites

11 minutes ago, jaclaz said:

Replace WHAT? :w00t:

With WHICH version? :ph34r:

It seems like you are not very familiar with the concepts of partitioning and formatting (USB) devices and with the ideas behind the mentioned filter drivers. :dubbio:

jaclaz
 

cfadisk.sys with Server2003 Sp2 version.

I never partitioned any usb device . MY expandable HDD is 80GB old one.

Link to comment
Share on other sites

1 hour ago, Dibya said:

cfadisk.sys with Server2003 Sp2 version.

I never partitioned any usb device . MY expandable HDD is 80GB old one.

There must be a communication problem of some kind.

If you use the HP USB format tool, it will make a PARTITIONED USB stick (unlike the "normal" Windows XP, where you CANNOT partition a "removable" device), so if you used that tool, you did partition the USB stick (evidently without knowing that you did so).

cfadisk.sys is a THIRD PARTY filter driver and it doesn't exist in XP, or in Server 2003.

What (the heck) is an "expandable" HDD? :w00t:

Does it expand (and possibly shrink)? :ph34r:

Anyway what is the relevance of it being 80 Gb (and old) ? :dubbio:

jaclaz


 

Link to comment
Share on other sites

I quick formatted a 4GB Giga Bank Hard Drive with one Partition from My Computer.
It produced the problem so I had not tried any other combinations.

I just tried using the FORMAT Command and a "removable" drive. Same problem.

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...