Someone else probably knows more than I, but here is my tiny bit ofAaron Martinez wrote:
Where this can cause a problem is newer boards with multiple PCI buses.
Usually it is something like one pair of slots are full PCI-X up to
133MHz and another pair is put aside for slower 66MHz cards or to allow
you to split the bandwidth. Board manufacturers do this because many
buses (including older pci) will run at the clock of the slowest card on
the bus, so giving you a separate bus allows you to have a slower card
in the system without slowing down the slots your faster cards are
plugged into. Unfortunately these multiple pci buses are not always set
up to be probed in a simple left-to-right order when looking at the
board. If you look at dmesg output you will probably see you have
multiple pci buses.
I may be wrong about this but these are my thoughts on the matter,
someone else feel free to correct me if I am wrong.
Best,
Chris
| Rafael J. Wysocki | 2.6.28-rc2-git7: Reported regressions from 2.6.27 |
| Dave Hansen | Re: [RFC/PATCH] Documentation of kernel messages |
| Jesper Juhl | Re: [RFD] Documentation/HOWTO translated into Japanese |
| Bart Van Assche | Integration of SCST in the mainstream Linux kernel |
git: | |
| Sander | 'struct task_struct' has no member named 'mems_allowed' (was: Re: 2.6.20-rc4-mm1) |
| Corey Minyard | [PATCH 3/3] Convert the UDP hash lock to RCU |
| Gerrit Renker | [PATCH 03/37] dccp: List management for new feature negotiation |
| David Miller | Re: [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
| David Miller | Re: [GIT]: Networking |
