Messages in this thread | | | Date | Fri, 4 Jan 2013 19:02:06 -0700 | Subject | Re: [PATCH v7u1 26/31] x86: Don't enable swiotlb if there is not enough ram for it | From | Shuah Khan <> |
| |
On Fri, Jan 4, 2013 at 4:55 PM, Yinghai Lu <yinghai@kernel.org> wrote: > On Fri, Jan 4, 2013 at 3:21 PM, Shuah Khan <shuahkhan@gmail.com> wrote: > >> Please see attached dmesg for full log. I can do some testing on this >> system with your patch if you would like. > > That would be great. > > Please try > git://git.kernel.org/pub/scm/linux/kernel/git/yinghai/linux-yinghai.git > for-x86-boot > or just this patch. > > Too bad, I can not access AMD systems with IOMMU support. > > Thanks a lot.
I tried your patch on my AMD system. I did change the patch to print warning instead of panic() and it did trigger the condition for panic.:
[ 5.376654] AMD-Vi: Found IOMMU at 0000:00:00.2 cap 0x40 [ 5.376717] [ 5.376799] pci 0000:00:00.2: irq 72 for MSI/MSI-X
It would have panic'ed here:
[ 5.388858] AMD-Vi: can not enable swiotlb for unhandled devices by AMD iommu!
[ 5.388964] AMD-Vi: Lazy IO/TLB flushing enabled [ 5.389324] LVT offset 0 assigned for vector 0x400
I applied your patch to 3.6.11 and changed the panic() to pr_info() and also changed enough_mem_for_swiotlb() to always return false to simulate not enough memory condition as this system does have enough memory.
So at least on this AMD system, your patch will result in a panic.
Thanks, -- Shuah
| |