Re: [PATCH 1/2] swiotlb: Rate-limit printing when running out of SW-IOMMU space

2016-11-05 Thread Konrad Rzeszutek Wilk
On Mon, Oct 31, 2016 at 04:45:03PM +0100, Geert Uytterhoeven wrote: > If the system runs out of SW-IOMMU space, changes are high successive > requests will fail, too, flooding the kernel log. This is true > especially for streaming DMA, which is typically used repeatedly outside > the driver's

Re: [PATCH 1/2] swiotlb: Rate-limit printing when running out of SW-IOMMU space

2016-10-31 Thread Sergei Shtylyov
On 10/31/2016 06:45 PM, Geert Uytterhoeven wrote: If the system runs out of SW-IOMMU space, changes are high successive s/changes/chances/? requests will fail, too, flooding the kernel log. This is true especially for streaming DMA, which is typically used repeatedly outside the

[PATCH 1/2] swiotlb: Rate-limit printing when running out of SW-IOMMU space

2016-10-31 Thread Geert Uytterhoeven
If the system runs out of SW-IOMMU space, changes are high successive requests will fail, too, flooding the kernel log. This is true especially for streaming DMA, which is typically used repeatedly outside the driver's initialization routine. Add rate-limiting to fix this. While at it, get rid