On 10/25/2010 08:50 PM, Vladislav Bolkhovitin wrote: > Boaz Harrosh, on 10/24/2010 01:55 PM wrote: >> On 10/22/2010 07:32 PM, Vladislav Bolkhovitin wrote: >>> Boaz Harrosh, on 10/19/2010 06:13 PM wrote: >>>> This is a trivial addition to the SG API that can receive kernel >>>> pointers. It is only used by the out-of-tree test module. So >>>> it's immediate need is questionable. For maintenance ease it might >>>> just get in, as it's very small. >>> >>> This patch doesn't handle queue's restrictions (DMA, etc.). >>> >> >> Please for give me. I know that the name is misleading. The >> Scatter-List here is the osd scatter-list. Which describes >> extents on the object, not of the memory. The proper name would >> be extents API. The reason we named it SG is because this is the >> way the OSD standard names it. We keep very close with the STD >> naming convention, so it ended up this way. > > OK, but the queue's restrictions aren't handled by it anyway. > > Vlad Yes they are! add_pc_page does the limits. blk_make_request() receives a bio-chain and does bouncing. But all this does not matter, because osd is a char-device not block. Meaning osd objects can be written at char alignment (offset and length). The way we work with multi segmented "out" and "in" buffers, anything not char aligned like iscsi or fb will not work very good any way. (Not very good I mean, everything will bounce) The libosd at the API level and internally works with bio-chains the block-request is only made much later. blk_make_request(bio-chain) should make everything work. If not it should be fixed. Boaz -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
