dma access:
if_mn Nope
if_pcn Nope
if_sf Nope
if_sk Nope
if_ste Nope
if_tl Nope
if_vr Nope
if_wb Nope
amd Nope (Is this OK due to CAM?)
ncr Nope (Is this OK due to CAM?)
meteor Nope
Also: the simos.c driver uses vtophys(), but is documented as being
broken, so isn't listed.
uio interfaces are present on alpha, but not yet tested due to lack
of mainstream network interface drivers (hopefully soon forthcoming).
Note there are separate issues here with large memory machines, and
Drew doesn't claim responsibility for that :-).
o Move two VFS locking cleanups/tasks into the WIP sections from the
New section, update timestamps. Jeff has been actively working on
VFS locking for a while now, and has addressed sizable chunks of
these tasks.
o Add kan as an owner on the i386 lazy context switch task.
o Add a sparc64 lazy context switch task and assign to Jake.
o Remove green from VM task, leaving just alc, update timestamp,
and mark as a WIP.
if_lge Nope
if_lmc Nope
if_my Nope
if_nge Nope
if_pdq Nope
if_tx Nope
iir Nope
isp OK
mlx OK
mly OK
mpt OK
pst Nope
stg OK
sym Nope
trm Nope
musycc Nope
ubsec Nope
usb Nope
wds OK
Also, make sure we have ';' after entities and related notions.
notes field (generally kvtop(), vtophys()).
Add/change the following conditions:
if_ed Nope
if_en Nope
if_idt Nope
if_ie Nope
aha OK
aic OK
asr Nope
buslogic Nope
dpt Nope
ida OK
ct Nope
digi Nope
drm Nope
exca OK
fb Nope
firewire Nope
hea Nope
hfa Nope
hifn OK
src/sys/dev/*/if* for use of dma. Most drivers don't use busdma.
if_lnc uses busdma, but also uses vtophys() so is likely bogus.
if_hme and if_gem appear to use older busdma interfaces and may
require updating.
a first pass with a brief description, and some initial status. Status
is split into two areas: platform support (i386, sparc64, etc), and
driver support (network interfaces, storage, ...). For each, the level
of support is in the process of being documented, and for drivers under
active work, that is listed along with the task owner. Providing
support for PAE on i386, as well as high performance driver support on
non-i386 platforms, will require the conversion of many of the important
system drivers. We'll need to inventory the relevant drivers, and then
have maintainers update them to use the proper interfaces.
him with the other members on this page.
As of last night, Kris has also joined the team. He's also done a lot
of work from the portmgr@ side for the last few releases, and adding
him to our alias should save me from CCing portmgr@ so often during
each release cycle.
Also, remove myself from "lead RE" position. We are moving to a
per-release team lead rather than a perennial team lead.
* Fix spelling nits
* Promote consistant use of <tt> tags
* Re-word those things that in their current state may to some appear
in some ways as possibly overly wordy
* Note the applicability of some GNOMENG options for GNOME 2 ports
* Move the discussion of the deprecated USE_* macros from the top to the
bottom
Approved by: pat
