PCIe during resume from sleep

NTDEV
1

I have a question about what happens when the computer is awoken from sleep
state. We have noticed that on XP our device seems to disappear and our
driver unloaded but when we access our device it is present and functional.
Under Vista the driver is unloaded, but is reloaded soon thereafter. The
same device connected via a PCIe slot is never detected as being removed. I
was thinking that maybe the LOM chip is on one of the first PCIe connections
and something is not stable early on. Any one have any suggestions? Since
it is a LOM we can’t get a CATC trace to work and there is no Jtag for an
ICE.

2

Your BIOS is doing this. They do it because they can save more power by
disconnecting the device from the bus.

–
Jake Oshins
Hyper-V I/O Architect
Windows Kernel Team

This post implies no warranties and confers no rights.

“David Craig” wrote in message news:xxxxx@ntdev…
> I have a question about what happens when the computer is awoken from
> sleep state. We have noticed that on XP our device seems to disappear and
> our driver unloaded but when we access our device it is present and
> functional. Under Vista the driver is unloaded, but is reloaded soon
> thereafter. The same device connected via a PCIe slot is never detected
> as being removed. I was thinking that maybe the LOM chip is on one of the
> first PCIe connections and something is not stable early on. Any one have
> any suggestions? Since it is a LOM we can’t get a CATC trace to work and
> there is no Jtag for an ICE.
>
>

3

Jake,

Thanks. We are setting up a test to track vaux and vmain to our chip and to
trigger on disk access to see if when Windows resumes if our power has not
been restored. Our NIC does not exhibit the problem if WOL is active and it
only fails about once in 300 suspend/resume sequences.

Thanks,
David Craig
Broadcom Corporation

“Jake Oshins” wrote in message
news:xxxxx@ntdev…
> Your BIOS is doing this. They do it because they can save more power by
> disconnecting the device from the bus.
>
> –
> Jake Oshins
> Hyper-V I/O Architect
> Windows Kernel Team
>
> This post implies no warranties and confers no rights.
>
> --------------------------------------------
>
> “David Craig” wrote in message
> news:xxxxx@ntdev…
>> I have a question about what happens when the computer is awoken from
>> sleep state. We have noticed that on XP our device seems to disappear
>> and our driver unloaded but when we access our device it is present and
>> functional. Under Vista the driver is unloaded, but is reloaded soon
>> thereafter. The same device connected via a PCIe slot is never detected
>> as being removed. I was thinking that maybe the LOM chip is on one of
>> the first PCIe connections and something is not stable early on. Any one
>> have any suggestions? Since it is a LOM we can’t get a CATC trace to
>> work and there is no Jtag for an ICE.
>>
>>
>