| 1 |
Neat. Random guess, but it could be a bug in Bulldozer's memory controller |
| 2 |
or IOMMU. Try disabling IOMMU support in your kernel? |
| 3 |
On Apr 29, 2012 3:29 PM, "Markos Chandras" <hwoarang@g.o> wrote: |
| 4 |
|
| 5 |
> On 04/28/2012 01:24 AM, Matthew Marlowe wrote: |
| 6 |
> > On Fri, Apr 27, 2012 at 3:54 PM, Markos Chandras <hwoarang@g.o> |
| 7 |
> wrote: |
| 8 |
> >> On 04/27/2012 11:45 PM, Nikos Chantziaras wrote: |
| 9 |
> >>> On 27/04/12 22:35, Markos Chandras wrote: |
| 10 |
> >>>> I replaced my Phenom II cpu with a new 6-core AMD bulldozer. However, |
| 11 |
> I |
| 12 |
> >>>> noticed that all of my Gentoo virtual machines throw (compiler) |
| 13 |
> >>>> segmentation faults when building or running any application. |
| 14 |
> >>> |
| 15 |
> > |
| 16 |
> > I'm not familiar with virtualbox, but I've seen similar issues occur |
| 17 |
> > with vmware and the solution was to at least temporarily mask whatever |
| 18 |
> > new cpu flags the new hypervisor was passing to the guest. In vmware, |
| 19 |
> > one could limit the cpu flags to maintain compatibility with various |
| 20 |
> > cpu releases which was especially helpful in clusters.... Yes, your |
| 21 |
> > gentoo vms should have been fine ..but at least until you track down |
| 22 |
> > the issue, see if virtualbox has a similar feature? |
| 23 |
> > |
| 24 |
> Interestingly this seems to be caused when using my wireless card to |
| 25 |
> bridge the virtualbox interfaces onto. I can't reproduce (yet) any |
| 26 |
> segfaults when I use the onboard ethernet card. I have the following |
| 27 |
> wireless card (supported by the rtl8180 kernel module) |
| 28 |
> |
| 29 |
> 04:06.0 Ethernet controller: Belkin F5D7000 v7000 Wireless G Desktop |
| 30 |
> Card [Realtek RTL8185] (rev 20) |
| 31 |
> |
| 32 |
> -- |
| 33 |
> Regards, |
| 34 |
> Markos Chandras / Gentoo Linux Developer / Key ID: B4AFF2C2 |
| 35 |
> |
| 36 |
> |
Archives