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.
We play mono samples on all 4 channels.
However, we get the volume settings for mono samples as a symmetic two-channel
setting... the other two channels used to stay at max volume...
I believe that something else is wrong here, but dont want to change MI
code (which in turn influences a couple of MD driver) thus late in the release
cycle.
- use board address space > 4 MB, instead of iszthreepa(), to detect Z3-mode
boards. We dont want the bus, but want the address configuration.
- s/CV64CONSOLE/CV3DCONSOLE/
- s/cv3d_zorroIII == 1/cv3d_zorroIII/ and s/cv3d_zorroIII != 1/!cv3d_zorroIII/
the CyberVision64 driver. Modified by Tobias Abt with some bugfixes
from Bernd Ernesti.
ZorroII is at the moment not supported and there is a small problem
with the Console driver where you just get a black screen, but the
system boots and you can use X11 without a problem.
The changes is to allow some limited mixer manipulation through
the audio device (instead of the mixer device).
This rendered 4 methods in audio_hw_if unused so garbage collect these.
pseudo-device rnd # /dev/random and in-kernel generator
in config files.
o Add declaration to all architectures.
o Clean up copyright message in rnd.c, rnd.h, and rndpool.c to include
that this code is derived in part from Ted Tyso's linux code.
This will be used by certain SCSI drivers to limit the length of DMA transfers
when the serial port is running at higher speeds. GVP and Phase5 DMA
interfers significantly with CPU execution, and significantly increases
receive buffer overruns (silo overflow). Reducing the DMA transfer length
can reduce the overruns.
- MIT syntax to Motorola syntax moveml register lists (for better readability)
- comment alignment
- remove commented out debug or alternate code from early DraCo support
development.
(I wanted to do this stuff for a long time, but always had "more important"
things to do.)
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.
NOTE: due to hardware requirement of the AGA chipset, the framebuffer width
has to be a multiple of 64 bits. An 800x600 display will use a memory width
of 832 bits, which the current X server can't handle. To get a usable
display, the width needs to be 832 or 768.
instead of bp->b_bcount. The I/O operation may have changed bp->b_bcount,
and not all of the kernel virtual memory will be freed. This will result
in the VM system putting pages back on the free list which are still
mapped. This may cause a deadlock later trying to enter pages into
the kernel pmap. [See PR#4120 for more details.]
* support chip clocks != COM_FREQ, by introducing sc_frequency (for the
mainline code) and adding a frequency parameter right after the rate
parameter to comcnattach() and com_kgdb_attach().
- Make com_isa and com_multi initialize sc_frequency to COM_FREQ.
- Make i386/machdep.c and alpha/dec_xxx.c call com*attach() with the freq.
parameter.
* supio_attach_args get two more fields: a sc_ipl and a sc_arg, both ints.
- com_supio uses the first for interupt establishment (all childs will, as
soon as they exist) and the 2nd for sc_frequency.
- drsupio passes sc_ipl alway as 5, and for the "com"s, sc_arg as 16*115200
- hyper will pass sc_ipl as 6, and sc_arg as 16 * 460800
without an 'a' partition. We open the RAW_PART partition instead.
Without this, a boot partition on the 'b' or later partition of such a disk
won't be found.
is for "standard PC i/o stuff" at known and constant locations, e.g. when
multi-io chips are used on non-ISA mainboards.
Implement drsupio.c, the DraCo version of this.
Attach the generic com.c to this bus.
Remove the old drcom hack.
Currently, this is only kind of a wrapper around the *_sicallback() function
family. This way, we'll only break the new driver if it doesn't work.
After more drivers are converted, we'll change it to be the other way round.
(currently only CD-ROM drives on i386). The sys/dev/scsipi system provides 2
busses to which devices can attach (scsibus and atapibus). This needed to
change some include files and structure names in the low level scsi drivers.
thorpej