Messages in this thread | | | Date | Thu, 27 May 2004 13:03:24 -0400 (EDT) | From | Arthur Perry <> | Subject | Re: GART error 11 |
| |
Hi Dave,
I have verified that booting the system with iommu=off (with vanilla kernel) does seem to make the problem go away.
I apologize for not being fully up to speed with knowing what relationship the iommu has with the gart, but I will find out.
I just wanted to post my findings.
Thanks again!
Best Regards,
Arthur Perry Lead Linux Developer / Linux Systems Architect Validation, CSU Celestica Sair/Linux Gnu Certified Professional Providing professional Linux solutions for 7+ years
On Thu, 27 May 2004, Dave Jones wrote:
> On Thu, May 27, 2004 at 11:02:10AM -0400, Arthur Perry wrote: > > > The question really comes down to: > > Is this problem an oversight of the distributors (silly! the agp driver should not be built into the kernel for server use!) > > or > > Kernel code implementation? (well, if no agp bus is present, then let's not go and set up the GART, right?) > > If you don't have an AGP graphics card, there'll be nothing really > set up, unless you are using the IOMMU feature of the x86-64 kernel. > Setting up of GART tables etc only happens when the graphics > drivers (DRI) asks it to. > > Does booting with iommu=off fix this for you? > It may be that some driver is doing something that it > shouldn't. If it does make the problem go away, what > devices do you have in the system (lspci output please) > > Dave > - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |