Annotation of wikisrc/projects/project/buffer-queue.mdwn, revision 1.2

1.1       jmmv        1: [[!template id=project
                      3: title="More intelligent buffer queue management"
                      5: contact="""
                      6: [tech-kern](
                      7: """
                      9: category="kernel"
1.2     ! jmmv       10: difficulty="medium"
1.1       jmmv       11: duration="2-3 months"
                     13: description="""
                     14: NetBSD has a fixed, kernel-wide upper limit on transfer size, which is currently
                     15: set to 64k on most port.  This is too small to have good performances on modern
                     16: IDE and SCSI hardware; on the other hand some devices can't do more than 64k,
                     17: and in some case are even limited to less (the Xen virtual block device for
                     18: example).  Software RAID will also cause requests to be split in multiple
                     19: smaller requests.
                     21: NetBSD would greatly benefit from a more intelligent buffer queue management
                     22: between the block device drivers and the higher levels (the framework here
                     23: currently only applies some selectable algorithm to sort the queue).  This
                     24: framework should be able to split buffers too large for a device into smaller
                     25: ones, or aggregate multiple contiguous requests into a larger one.  This will
                     26: most probably require change to at last some block device drivers.
                     27: """
                     28: ]]

CVSweb for NetBSD wikisrc <> software: FreeBSD-CVSweb