Commit Graph

13619 Commits

Author SHA1 Message Date
thorpej 8cbc5caabb The user segment table pointer is no longer kept here. Change it to
a spare field.
1998-01-06 06:46:05 +00:00
perry 2ec6474b31 RCSID Police. 1998-01-06 04:44:59 +00:00
jtk fe87ac4104 fix for changes in bioscall structure and include files 1998-01-06 00:35:42 +00:00
thorpej a8f80ad4fd Fix a bogosity apparently inherited from when the Utah 4.3BSD code base
was converted to use Mach VM for Net2/4.4BSD.  The user segment table
pointer was originally stored in the PCB.  When Mach VM came along,
however, it was also stored in the pmap, and loaded into the PCB in
pmap_activate().  pmap_activate() would then note that the PCB's USTP
was now in sync with the pmap's USTP, and the low-level context switch
code would use the value from the PCB.

However, pmap_activate() would also load the hardware MMU context if
the pmap was the current pmap (or, in the case where pmaps can be shared,
such as in NetBSD, if the proc was the current proc).  The low-level
context switch code would then reload the hardware _again_ using the
USTP from the PCB.

However, the optimization of not calling pmap_activate() if "stchanged"
was false ended up causing some processes to use stale USTP values from
the PCB when the low-level context switch code reloaded the hardware!
This was noticed by using a real vfork(2) (which worked for some time
before failing, surprisingly!)

Since I'm hard pressed to find any real optimization here (since the
hardware was always reloaded once, sometimes twice!), the code now always
calls pmap_activate(), which uses the correct USTP value (the one in the
pmap).  The PCB's USTP is now ignored, and should eventually be g/c'd.

Another optimization can actually be performed, and I have added a comment
describing what it is, but have not yet implemented it.

Also note that most of the loadustp() functions where actually incomplete.
This has been corrected.  These functions should probably be split up into
MMU-specific operations, and called indirectly, rather than doing constant
run-time decision making based on values that will never change during the
course of a boot's lifetime.
1998-01-05 23:16:21 +00:00
perry 3625dbcd48 RCSID Police. 1998-01-05 21:34:56 +00:00
perry 5646f55116 make script insert RCS ids into generated files 1998-01-05 21:28:21 +00:00
perry 086015d681 RCSID Police. 1998-01-05 21:13:51 +00:00
perry e464358f5f make script insert RCS ids into generated files 1998-01-05 21:12:34 +00:00
perry 49ad493e50 RCSID Police. 1998-01-05 21:11:15 +00:00
perry 489c2f7e06 make script insert RCS ids into generated files 1998-01-05 21:10:18 +00:00
perry b46484bb8a RCSID Police. 1998-01-05 20:51:25 +00:00
perry 015e898c02 RCSID Police. 1998-01-05 07:02:46 +00:00
perry 3e0fad1868 RCSID Police. 1998-01-05 06:28:44 +00:00
perry e13ab577cd RCSID Police. 1998-01-05 05:05:49 +00:00
thorpej b9f1b716f3 Now that all ports have pmap_activate(), and it has an identical interface,
prototype it in <vm/pmap.h>
1998-01-03 01:12:59 +00:00
thorpej b73f9750f6 Don't allocate PTEs in cpu_fork(); they were allocated when the new
process's pmap was created.  Instead, use pmap_activate() to load the
page {0,1} base and length registers into the PCB for the hardware's use.
1998-01-03 00:37:31 +00:00
thorpej aa58d4ac6b Adjust for the fact that the page {0,1} base and length registers are now
stored in the pmap structure.
1998-01-03 00:35:28 +00:00
thorpej f267fe868d - Adjust so that access to the PCB is not required in most pmap functions.
Store the page {0,1} base and length registers in the pmap structure,
  and implement pmap_activate(), which stores them in the PCB (for the
  hardware's use).
- Re-implement pmap_pinit(); allocate the PTEs here, not in cpu_fork().
1998-01-03 00:34:02 +00:00
thorpej 172a9381b6 - Remove the PCB pointer from the pmap structure. Instead, store the
page {0,1} base and length registers in the pmap structure.  They will
  be loaded in to the PCB when the process's address space is activated
  by way of pmap_activate().
- Remove pmap_pinit() macro; it's now a real function.
- Prototypes for pmap_pinit(), pmap_activate(), pmap_deactivate().
1998-01-03 00:28:43 +00:00
thorpej d8076f896b Grab the PCB directly from the u-area, not by indirecting through VM
structures.
1998-01-03 00:23:53 +00:00
thorpej 9ca174a51d Implement pmap_activate(). 1998-01-02 22:57:56 +00:00
thorpej a6f228e498 Make pmap_activate() take a struct proc *. 1998-01-02 22:36:33 +00:00
thorpej c162e92a71 Implement pmap_activate(). 1998-01-02 22:17:18 +00:00
thorpej f39e44955b Make pmap_activate() take a proc *. 1998-01-02 20:37:28 +00:00
thorpej d14e18faa9 Make pmap_activate() take a proc *. 1998-01-02 20:10:26 +00:00
ragge 5680284286 Fix problem in copy*str() when not emulating locc. 1998-01-02 19:33:27 +00:00
thorpej dd966fca05 Change an argument name to PMAP_ACTIVATE() to more accurately describe the
semantics of the argument.
1998-01-01 20:05:23 +00:00
thorpej 287b63b321 - Make pmap_activate() and pmap_deactivate() take a struct proc *.
- Define active_user_pmap() and use it in the appropriate places (from hp300
  port).
1998-01-01 19:52:50 +00:00
mikel 733dae35bc use ${SIZE} instead of "size" (somehow this one got missed) 1998-01-01 08:11:41 +00:00
thorpej 8b5e0b6e06 Remove a couple of unneeded VM-related headers. 1997-12-31 19:32:55 +00:00
cjs 89951259ed Add kernfs; remove more stuff we don't need. 1997-12-31 16:44:37 +00:00
thorpej 9c20093835 Make pmap_activate() and pmap_deactivate() take a struct proc *. 1997-12-31 10:12:55 +00:00
scottr fa1347ed5b Prompt for total sectors, and calculate sane defaults for sectors per
cylinder and total sectors.
1997-12-29 07:15:10 +00:00
scottr 89b10b928f Do better reporting of transmit errors. 1997-12-26 22:37:05 +00:00
oki ac86463ce3 Make compile without options MACHINE_NONCONTIG. 1997-12-24 17:48:10 +00:00
oki 49779c8c5a Add MDP_STACKADJ. 1997-12-24 17:46:08 +00:00
tv 89e7595756 Exclude /dev/fb code if no framebuffer is compiled into the system.
Adds a fb.h flag file with '#define NFB 1' if any such card is configured
in to the kernel.
1997-12-23 23:57:20 +00:00
thorpej ff7a4d0472 Add COMPAT_13. 1997-12-22 05:31:23 +00:00
jonathan 3550b22654 Commit patch for PMAG-A (mfb) pixel inversion accientally left out of 1.3.
See PR# port-pmax/4724.
1997-12-22 02:39:28 +00:00
hpeyerl 6563f769af update for correct SPU type and COMPAT_13. 1997-12-21 17:15:01 +00:00
perry 589a5e2c79 patch from is -- I don't know what it does, but he says it's critical. 1997-12-20 21:13:13 +00:00
pk bf8e38df4f Establish the interrupt handlers on sun4 `oclocks'; pointed out
by Chuck Cranor.
1997-12-19 22:53:10 +00:00
scottr 6a7a3fb133 spurintr() is a low-level interrupt handler, and must not
be called by C code!  Spotted by Bill Studenmund, who also provided
#the substance of this change.
1997-12-19 05:38:00 +00:00
pk 964a147839 Clean vers.c & vers.o. 1997-12-18 21:03:58 +00:00
mycroft 61475326e0 Add delays for lame keyboard controllers, as in the keyboard driver. Do a
better job of flushing input in various cases.
1997-12-18 16:49:10 +00:00
sakamoto 73a2fb9659 support simple stack trace. 1997-12-18 09:09:01 +00:00
sakamoto e2f98dba9e remove avail_end temporarily for compile. 1997-12-18 09:08:35 +00:00
sakamoto 15dd19ea29 use pmap.c of port-powerpc.
mem_regions support.
1997-12-18 09:07:58 +00:00
scottr e43a4bda4a Update for the NWO 1997-12-18 06:45:30 +00:00
scottr 0b62536fdf Make these more closely resemble the old configurations. 1997-12-18 06:44:15 +00:00