name of the new 'flags' argument, added as part of the new pmap_enter() API.
Added a comment which explains why there are still sun3x-specific flags
being passed in the unused bits of the physical address, and how these flags
are independent of the new 'flags' argument.
* Map the message buffer with access_type = VM_PROT_READ|VM_PROT_WRITE `just
because'.
* Map the file system buffers with access_type = VM_PROT_READ|VM_PROT_WRITE to
avoid possible problems with pagemove().
* Do not use VM_PROT_EXEC with either of the above.
* Map pages for /dev/mem with access_type = prot. Also, DO NOT use
pmap_kenter() for this, as we DO NOT want to lose modification information.
* Map pages in dumpsys() with VM_PROT_READ.
* Map pages in m68k mappedcopyin()/mappedcopyout() and writeback() with
access_type = prot.
* For now, bus_dma*(), pmap_map(), vmapbuf(), and similar functions still use
access_type = 0. This should probably be revisited.
DVMA segment (for the page reserved by the PROM). If we copy all
the mappings, pmap_enter_kernel may complain about finding valid
PTEs that are not recorded in our PV lists. Make pmap_remove_pte
do something reasonable if it finds a PTE for a managed page that
has not been recorded in the PV lists. (Plus other minor stuff).
msgbuf. Note that old 'dmesg' and 'syslogd' binaries will continue running,
though old 'dmesg' binaries will output a few bytes of junk at the start of
the buffer, and will miss a few bytes at the end of the buffer.
all interrupts which might cause re-entrancy.
Allow virtual addresses which are not managed by the MI VM system to
participate in the PV system.
Remove a few frivoulous TLB cache flushes.
allow the PROM monitor to see mappings in kernel space. This also
fixes a problem with exiting to the monitor with the stack pointer
set to some place in kernel space. Also fix NUM_KERN_PTES, and
change get_pte() so it works on any virtual address.
one page too many (worse, it unmapped our stack page... Ouch!)
Also, use loadcrp() and such instead of calling pmap_activate
in various places so we have more control over TLB flush.
Now pmap_activate() is called only by locore (and that call is
done only if PMAP_DEBUG is defined). Other minor clean-up...
jeremy