Quoting Sean Hefty <[EMAIL PROTECTED]>: > Subject: [PATCH] add support for large RMPP transfers > > Add support for sending and receiving large RMPP transfers. > Current code supports transfers only as large as a single memory allocation. > This patch uses linked list of memory buffers when sending and receiving > data to avoid needing contiguous pages for larger transfers. > > Signed-off-by: Sean Hefty <[EMAIL PROTECTED]> > > Also integrates changes from Michael and Jack.
Sean, the patch seems to match relatively closely Jack's design, checked in by r5437 (svn log below). The patch comment does not make it clear, however. Could you please correct this? I think you also need to add Jack's S.O.B. line (svn log below). -------------------------- r5437 | halr | 2006-02-16 21:25:26 +0200 (Thu, 16 Feb 2006) | 17 lines Large RMPP support Receive side: copy the arriving MADs to chunks instead of coalescing to one large buffer in kernel space Send side: split a multipacket MAD buffer to a list of segments, (multipacket_list) and send these using a gather list of size 2. Also, save pointer to last sent segment, and retrieve requested segments by walking list starting at last sent segment. Finally, save pointer to last-acked segment. When retrying, retrieve segments for resending relative to this pointer. When updating last ack, start at this pointer. Signed-off-by: Jack Morgenstein <[EMAIL PROTECTED]> -- Michael S. Tsirkin Staff Engineer, Mellanox Technologies _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
