On Wednesday 04 February 2009, Sarah Sharp wrote: > I'm a bit confused as to why a host controller would want to limit the > number of entries in an sg list (instead of say, the size of the overall > transfer or some other metric). One example: the hardware may use the kind of DMA engine where DMA descriptors are stored in dedicated SRAM, not main memory. There can be a hard limit ... e.g. no more than 64 DMA descriptors per transaction, even assuming no transfers to other endpoints (or non-USB devices) would need to happen at the same time. If there's no natural limit, use UINT_MAX or whatever. The block layer *also* allows per-segment limits, for much the same reason: DMA descriptors may have limits. Maybe a 16-bit counter for the number of DMAREQ responses that can be described, which for USB might map to packet counters but for hardware that's less FIFO-rich might be individual words. - Dave -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
