Re: PCI resources above 4GB

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, Apr 10, 2012 at 3:53 AM, Steven Newbury <steve@xxxxxxxxxxxxxxx> wrote:
> On 10/04/12 01:51, Bjorn Helgaas wrote:
>> On Mon, Apr 9, 2012 at 4:49 AM, Steven Newbury <steve@xxxxxxxxxxxxxxx> wrote:
>>> Hi Bjorn,
>>>
>>> like many users here:
>>>
>>> https://bugs.launchpad.net/ubuntu/+source/nvidia-graphics-drivers-180/+bug/342926
>>>
>>> I've hit this bug:
>>> https://bugzilla.kernel.org/show_bug.cgi?id=10461
>>>
>>> I was wondering whether an option to enable an algorithm which
>>> allocated all 64bit capable devices above 4GB would at least allow
>>> systems where the BIOS has not provided a sufficient PCI hole to
>>> initialise all PCI devices?  I've been digging into the code, but I'm
>>> not yet comfortable to seriously attempt this myself, certainly not
>>> without guidance.
>>>
>>> It's a shame TJ seems to have disappeared, it would have been great to
>>> have had a look at his code, I'm sure many of these problems would have
>>> been resolved.
>>>
>>> Any help, workarounds, guidance or patches welcome! :)
>> Please attach complete dmesg logs with and without "pci=use_crs" to
>> the bugzilla.
>>
>>
> dmesg logs attached to bug, for both pci=use_crs and pci=nocrs.  In both
> cases assignment fails.

Can you please try attached patches with pci=nocrs?

for pci=use_crs, we need find safe place beyond _CRS, because your
_CRS limit under 4g.

Thanks

Yinghai

Attachment: pref_mem_64_only.patch
Description: Binary data

Attachment: allocate_high_at_first.patch
Description: Binary data


[Index of Archives]     [DMA Engine]     [Linux Coverity]     [Linux USB]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [Greybus]

  Powered by Linux