PERFORCE change 134449 for review

[Warner Losh]

http://perforce.freebsd.org/chv.cgi?CH=134449

Change 134449 by imp at imp_lighthouse on 2008/01/30 06:28:43

	Update todo list based on done items.

Affected files ...

.. //depot/projects/mips2-jnpr/src/sys/TODO#12 edit

Differences ...

==== //depot/projects/mips2-jnpr/src/sys/TODO#12 (text+ko) ====

@@ -7,17 +7,12 @@
 o unify cpu.h vs cpureg.h
 o Look for old vestiges of the array of args for p->p_regs[] and also comments
   about same.
-o cpu.c in mips2 appears to be done in machdep in jnpr.  reconcile
-  cpu.c in jmips is just for the cpu device driver.
 o Look at the differences between the trampoline code between the two versions
   as this may lead to problems.  Specifically, we have sigcode encoded
   into locore.S, and some code in sigreturn in pm_machdep.c.  Not sure where
   things are in libc.
 o What does that assmbler before mi_startup do in locore.S  Comment or
   remove.
-o juniper mips adds some mucking with the FPU in after the platform_init
-  call (well, it was called mips_init in Jmips).  Is this relevant to the
-  merged code, since mips2 did this earlier...
 o Don't forget John Baldwin's comments on the user switch stuff. XXX add more
   here XXX
 o What's defined in assym.s and param.h varies somewhat between the two
@@ -26,9 +21,6 @@
   variable setting, which is related to something that was omitted from
   the juniper-mips dump.
 o cpu init
-o I moved where we initialize the msgbuf.  We need to make sure that no
-  printfs happen.  This is where the mips2 initializes things, but I've
-  not audited things to make sure that's really sane.
 o Double check last merge from mips2 -> cavium code.  Something seems wrong
   in one of the edits.
 o In the juniper vendor tree, gdb support just showed up.  Merge it into
@@ -36,6 +28,9 @@
 o splhigh still around, is it needed?
 o PCPU stuff in exception.S appears to be 32-bit only, using only %hi and
   %lo.  Isn't it necessary to use %higher and %highest in the 64-bit case?
+o Adapt powerpc to new generic bs implementation.  It isn't a slam dunk because
+  it gratuitously does the stream methods differrently than arm.
+o octeon is a 64-bit kernel, so it should be in mips64, not mips32.
 
 ************ RRS (below) *********
 o obio.c is probably not a good name for the IQ80321 driver