Messages in this thread | | | From | Bjorn Helgaas <> | Date | Mon, 28 Jan 2013 19:50:04 -0700 | Subject | Re: [RFC PATCH v5 4/8] ACPI, PCI: avoid building pci_slot as module |
| |
On Mon, Jan 28, 2013 at 7:45 PM, Jiang Liu <jiang.liu@huawei.com> wrote: > On 2013-1-29 10:21, Yinghai Lu wrote: >> On Mon, Jan 28, 2013 at 6:07 PM, Jiang Liu <jiang.liu@huawei.com> wrote: >>> Could we use quirk to auto-disable PCIe native hotplug for >>> problematic platforms? >> >> that is some kind of boot command line way? > We could negotiate between acpiphp and pciehp if those problematic > platforms/chipsets could be identified by DMI info or PCI device ID.
That's a sub-optimal solution because it requires us to identify all the affected systems and also may require ongoing maintenance in the future. If we could get specifics on the issue, we might be able to design a better solution that works for everybody.
(BTW, this thread is specifically about pci_slot, which I think is a slightly different issue than the acpiphp/pciehp conflicts.)
| |