=================================================================== RCS file: /cvsrepo/anoncvs/cvs/www/sparc.html,v retrieving revision 1.14 retrieving revision 1.15 diff -u -r1.14 -r1.15 --- www/sparc.html 1997/01/29 01:54:37 1.14 +++ www/sparc.html 1997/04/19 23:58:16 1.15 @@ -67,10 +67,6 @@
-OpenBSD/sparc can be correctly described as NetBSD/sparc + improvements. -
- -The people working the most on OpenBSD/sparc currently consists of Theo de Raadt, Jason Downs, and probably Chuck Cranor again now that it once again works. @@ -79,10 +75,35 @@
An important note about OpenBSD/sparc is that it is designed so that a single kernel can run on *ALL* sparc machines. Whereas SunOS and -Solaris have always had `kernel architectures', ie. sun4, sun4c, and -sun4m, the same `GENERIC' OpenBSD kernel will run on all the models. +Solaris have always had separate `kernel architectures', ie. sun4, sun4c, +and sun4m, the same `GENERIC' OpenBSD kernel will run on all the supported +models.
++OpenBSD/sparc can be reasonably described as a continuation of the +NetBSD/sparc development with improvements. The current status is +somewhere between under development and ready for gneral use, in that +it does support many of the Sparc based systems and their peripherals, +and can be quite reliable depending on the system configuration and +usage. +
+ ++Most of the problems are believed to stem from the wide variety of sparc +processor and cache implementations along with their undocumented bugs, +rather then general kernal problems. +Feedback on which models do and do not work reliably is +appreciated, particularly with the newer sun4m implementations like the +Fujitsu TurboSparc and Ross HyperSparc based systems and upgrades. +
+ ++The installation tools and process as of the 2.0 release are currently not +in user-friendly form, as most developers short-cut the process from running +SunOS or OpenBSD systems. This situation is currently being improved. +
+Snapshots are made available from time to time. @@ -96,9 +117,8 @@
+OpenBSD/Sparc on the 4c/4m machines is criticaly dependent on configuration +information returned by the openboot prom. A sparc clone that differs +substantially from the Sun model as far as device names and properties will +require additional work in this area.