Annotation of wikisrc/projects/project/buffer-queue.mdwn, revision 1.7

1.1       jmmv        1: [[!template id=project
                      2: 
1.6       dholland    3: title="Buffer queue coalescing and splitting"
1.1       jmmv        4: 
                      5: contact="""
                      6: [tech-kern](mailto:tech-kern@NetBSD.org)
                      7: """
                      8: 
1.5       dholland    9: category="filesystems"
1.2       jmmv       10: difficulty="medium"
1.1       jmmv       11: duration="2-3 months"
                     12: 
                     13: description="""
1.6       dholland   14: Currently the buffer handling logic only sorts the buffer queue (aka
                     15: disksort).
                     16: In an ideal world it would be able to coalesce adjacent small
                     17: requests, as this can produce considerable speedups.
                     18: It might also be worthwhile to split large requests into smaller
                     19: chunks on the fly as needed by hardware or lower-level software.
                     20: 
                     21: Note that the latter interacts nontrivially with the ongoing
1.7     ! dholland   22: [dynamic MAXPHYS](../maxphys-dynamic) project and might not be
1.6       dholland   23: worthwhile.
                     24: Coalescing adjacent small requests (up to some potentially arbitrary
                     25: MAXPHYS limit) is worthwhile regardless, though.
1.1       jmmv       26: """
                     27: ]]

CVSweb for NetBSD wikisrc <wikimaster@NetBSD.org> software: FreeBSD-CVSweb