On Tuesday 01 February 2005 23:54, Daniel Phillips wrote: [...] > What is wrong with memory-mapping the DMA buffer? This would not give > access to the DMA control registers themselves.
Nothing per se. In fact, that's what I try to achieve too. However, if you mmap in userspace directly the direct DMA buffer area, then only one process can do it. That's pretty X centric. (It's not a criticism btw, just a remark.). Also, you need kernel-specific mechanisms to ensure that this is locked memory (never goes to swap, never gets remapped) and/or to have multiple contiguous pages. Feasible, but maybe not so fancy after all. Plus you also have issues wrt to starting DMA and waiting for it to end (all via ioctl()). My idea is more to let the kernel driver reassemble the regular pages submitted by (possible several) processes for DMA execution using the indirect DMA mode. That makes the indirect mode the norm, but it seems to me it integrates more naturally with kernels virtual memory mechanisms. Rodolphe _______________________________________________ Open-graphics mailing list [email protected] http://lists.duskglow.com/mailman/listinfo/open-graphics List service provided by Duskglow Consulting, LLC (www.duskglow.com)
