On 2022-12-14 14:57, Dave Voutila wrote:
Mischa <open...@mlst.nl> writes:

On 2022-12-13 20:29, Dave Voutila wrote:
Dave Voutila <d...@sisu.io> writes:

tech@,
The below diff tweaks how vmd and vmm define memory ranges (adding
a
"type" attribute) so we can properly build an e820 memory map to
hand to
things like SeaBIOS or the OpenBSD ramdisk kernel (when direct booting
bsd.rd).
Why do it? We've been carrying a few patches to SeaBIOS in the
ports
tree to hack around how vmd articulates some memory range details. By
finally implementing a proper bios memory map table we can drop
some of
those patches. (Diff to ports@ coming shortly.)
Bonus is it cleans up how we were hacking a bios memory map for
direct
booting ramdisk kernels.
Note: the below diff *will* work with the current SeaBIOS
(vmm-firmware), so you do *not* need to build the port.
You will, however, need to:
- build, install, & reboot into a new kernel
- make sure you update /usr/include/amd64/vmmvar.h with a copy of
  symlink to sys/arch/amd64/include/vmmvar.h
- rebuild & install vmctl
- rebuild & install vmd
This should *not* result in any behavioral changes of current vmd
guests. If you notice any, especially guests failing to start, please
rebuild a kernel with VMM_DEBUG to help diagnose the regression.

Updated diff to fix some accounting issues with guest memory. (vmctl
should report the correct max mem now.)

Booted... The memory display in vmctl show is normal again.

root@current:~ # vmctl show
   ID   PID VCPUS  MAXMEM  CURMEM     TTY        OWNER    STATE NAME
    4 56252     1    1.0G    989M   ttyp4     runbsd04  running vm04
    3 60536     1    8.0G    2.2G   ttyp3       runbsd  running vm03
    2 20642     1   16.0G    3.4G   ttyp2       runbsd  running vm02
    1 81947     1   30.0G    5.6G   ttyp1       runbsd  running vm01

All seems to running normal. Anything specific I need to look out for?


Other than the above, no not really. Going to keep this diff out on
tech@ a few days to allow folks with a variety of guests to test before
I ask for OK's to commit.

The next change will be SeaBIOS (vmm-firmware) once this lands.

Perfect! Will do some more tests and will let you know if I find something.

Mischa

Reply via email to