![]() | |||||||||||||
Messages in this thread |
MOLNAR Ingo wrote: > hm, if DMA channels are being mucked with from within an IRQ handler, then > we should spinlock the DMA controller itself. It's only in recent kernels I've thought about this too. Yes, sound among others need to get the residue count out of the DMA controller at interrupt time. The problem with the 8237 design is the "intrinsic" state shared between all the channels on a particular chip, especially the "byte pointer flip flop". This FF is affected whenever one reads or writes values, be it base address or counter values. So yes, the cleanest thing would be a global spinlock for the DMA controller. > the clean solution IMO is to add a global spinlock to DMA controllers. Tom - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/faq.html | ||||||||||||
| Last update: 2005-03-22 12:44 [from the cache] ©2003-2008 | |||||||||||||