There's far more in there than just "iohsata.sys".
None of it is WHQL'ed though nor are the executables signed, although that is understandable under the circumstances.
Would that fact preclude that driver's use in Win2k3_x86 ?
i presume the "Precision Time Protocol (PTP)" aka "IEEE 1588" belongs in LAN along with the "Intel(R) Platform Controller Hub EG20T Gigabit Ethernet Controller - 8802". Others could be merged into Chipset, such as the "Intel(R) Platform Controller Hub EG20T I2C Controller - 8817" or the "Intel(R) Platform Controller Hub EG20T Packet Hub - 8801", or the SPI Bus driver.
Most, if not all, of these have somewhat equivalent "NullDrivers" to nullify their functionality (including the AHCI SATA Controller) for the following:
PCI\VEN_8086&DEV_8801.DeviceDesc = "Intel(R) Platform Controller Hub EG20T Packet Hub - 8801 !"
PCI\VEN_8086&DEV_8802.DeviceDesc = "Intel(R) Platform Controller Hub EG20T Gigabit Ethernet Controller - 8802 !"
PCI\VEN_8086&DEV_8803.DeviceDesc = "Intel(R) Platform Controller Hub EG20T General Purpose IO Controller - 8803 !"
PCI\VEN_8086&DEV_8808.DeviceDesc = "Intel(R) Platform Controller Hub EG20T USB Client Controller - 8808 !"
PCI\VEN_8086&DEV_880B.DeviceDesc = "Intel(R) Platform Controller Hub EG20T SATA AHCI Controller - 880B !"
PCI\VEN_8086&DEV_8810.DeviceDesc = "Intel(R) Platform Controller Hub EG20T DMA Controller #1 - 8810 !"
PCI\VEN_8086&DEV_8811.DeviceDesc = "Intel(R) Platform Controller Hub EG20T UART Controller - 8811 !"
PCI\VEN_8086&DEV_8812.DeviceDesc = "Intel(R) Platform Controller Hub EG20T UART Controller - 8812 !"
PCI\VEN_8086&DEV_8813.DeviceDesc = "Intel(R) Platform Controller Hub EG20T UART Controller - 8813 !"
PCI\VEN_8086&DEV_8814.DeviceDesc = "Intel(R) Platform Controller Hub EG20T UART Controller - 8814 !"
PCI\VEN_8086&DEV_8815.DeviceDesc = "Intel(R) Platform Controller Hub EG20T DMA Controller #2 - 8815"
PCI\VEN_8086&DEV_8816.DeviceDesc = "Intel(R) Platform Controller Hub EG20T Serial Peripheral Interface Bus - 8816 !"
PCI\VEN_8086&DEV_8817.DeviceDesc = "Intel(R) Platform Controller Hub EG20T I2C Controller - 8817 !"
PCI\VEN_8086&DEV_8818.DeviceDesc = "Intel(R) Platform Controller Hub EG20T Controller Area Network (CAN) Controller - 8818 !"
PCI\VEN_8086&DEV_8819.DeviceDesc = "Intel(R) Platform Controller Hub EG20T IEEE 1588 Hardware Assist - 8819 !"
Note the exclaimation mark at the end of the Description. I don't know it's significance, yet i'm noting it just the same.
I wonder what the preference should be for integration, that the remaining could be archived.
This reminds me of the HECI controllers aka "Intel(R) Management Engine" & their equivalent "NullDrivers".
I wonder what it is that has worked for you, whl0806, without "leaking like a sieve"?
We do have the option, since none of this is WHQL, to "fine-tune" this in the .Inf's.
You could use the "D-P-C Driver Latency Tester" as a guide concerning system latency and drivers detrimental to system latency (for example: legacy-AtherosWLAN, s478-P4-Prescott-HT-on-some-implimentations, ALi/ULi USB, et al?).
[off-topic]
I do like the sound of the "CAN Controller"!
All too often these days, it seems, we hear of those controllers that "CAN'T" or simply "won't" ...
[/off-topic]
Last edited by TechDud (2013-01-13 10:18:35)