some updates.

This commit is contained in:
yamt 2007-02-23 18:31:42 +00:00
parent aa221faf9e
commit d2bcbf783e

View File

@ -1,4 +1,4 @@
# $NetBSD: ROADMAP,v 1.12 2006/11/29 21:21:41 xtraeme Exp $ # $NetBSD: ROADMAP,v 1.13 2007/02/23 18:31:42 yamt Exp $
A high-level roadmap for NetBSD A high-level roadmap for NetBSD
@ -80,7 +80,7 @@ aa. Modular scheduler supporting CPU affinity and some soft RT features.
multi-processor systems could benefit from taking processor affinity multi-processor systems could benefit from taking processor affinity
into account (including support for HT CPUs). into account (including support for HT CPUs).
Responsible: matt (possibly) Responsible: matt (possibly)
ETA: ? ETA: ? yamt-idlelwp branch is a first step of this.
ab. Reduction of the giant lock ab. Reduction of the giant lock
There are several proposals for the best way forward on this, but There are several proposals for the best way forward on this, but
@ -136,8 +136,8 @@ ai. Complete support for LWPs
aj. PTHREAD_CONCURRENCY > 1 support aj. PTHREAD_CONCURRENCY > 1 support
A single process that uses threads should be able to reliably A single process that uses threads should be able to reliably
use more than one CPU. use more than one CPU.
Responsible: TBD Responsible: ad
ETA: ? ETA: 5.0 (1:1 pthread come with newlock2)
ak. AIO support ak. AIO support
POSIX aio_*() with full support for Asynchronous I/O (AIO) in the POSIX aio_*() with full support for Asynchronous I/O (AIO) in the
@ -161,8 +161,8 @@ an. Update the locking mechanisms in the kernel
now-archaic "newlock" branch, from thorpej. It requires some now-archaic "newlock" branch, from thorpej. It requires some
overhaul of cpu_switch/scheduler so that mutex_*(9) and ltsleep(9) overhaul of cpu_switch/scheduler so that mutex_*(9) and ltsleep(9)
can interlock. can interlock.
Responsible: TBD Responsible: ad
ETA: ? ETA: 5.0 (newlock2)
ao. Review TCP/IP developments ao. Review TCP/IP developments
Fix NewReno Fix NewReno