sprinkle rump, gcc=12 etc.
Now sparc64 with HAVE_GCC=12 and MKRUMP=no builds again

(macallan)

sprinkle gcc=12 and rump attributes

(macallan)

allow userland to switch to 16bit colour

(macallan)

don't enable xrender support unless we have 24bit colour

(macallan)

allow mapping of blitter registers

(macallan)

turn LEDs off when shutting down

(macallan)

make gftfb_restore_palette() grab the default colour map from rasops instead
of just writing the driver's map into the hardware ( which may have been
modified by the likes of wsfb )
With this we get a readable console even when exiting X in a less than
graceful manner.

(macallan)

For some reason the drawing engine occasionally scribbles past the right
boundary when filling rectangles, especially annoying when we draw whitespaces
As a workaround we draw all rectangles less than 50 pixels wide by drawing
a 50 pixel rectangle into off-screen memory to the right of the visible fb and
then copy the portion we want. Keeps track of the colour and size of the
off-screen rectangle so we can avoid redrawing it whenever possible.

(macallan)

fix brainfart - only update fbi_fbsize, not the size of the visible fb...

(macallan)

in gftfb_ioctl():
- identify ourselves as WSDISPLAY_TYPE_STI
- return full fb geometry in WSDISPLAYIO_GET_FBINFO

(macallan)

add gftfb

(macallan)

remove accidentally left in debug goop

(macallan)

use uint8_t for all 8bit masks and data to avoid sign extending weirdness
found by running this on hppa with gcc12

(macallan)

if the framebuffer is wider than the visible area, use the full width for the
glyphcache

(macallan)

install stireg.h

(macallan)

support WSDISPLAYIO_SVIDEO so X can turn the monitor off
so far I only know how to turn off video output, not sync(s). Better than
nothing though.

(macallan)

add hardware cursor support, mostly for X

(macallan)

avoid one more instance of unnecessary blitter stalling

(macallan)

keep track of hw settings for blitter, fill of fb access to avoid unnecessary
register writes
while there, remove some accidentially left in debug goop

(macallan)

play the same lazy cursor (un)draw trick as cgsix and friends
visible speedup in things like systat

(macallan)

enable glyph cache, anti-aliased fonts etc.

(macallan)

remove check for unused bits in attr2idx() - this really wasn't helpful

(macallan)

blitter time!
everything except drawing characters i now done by hardware
next step - glyphcache

(macallan)

crude beginning of a native driver for PCI Visualize EG cards
so far it supports:
- colour
- virtual consoles
todo:
- mmap
- hardware acceleration
- STI refactoring

(macallan)

in sti_screen_setup() don't bother looking for fonts if STI_FBMODE is requested

(macallan)

Make this work properly on PCI Visualize EG:
- use bus_space_*_stream() for all register accesses in order to avoid
  byte order conversion on PCI devices
- when switching to framebuffer mode, request as many text planes as STI allows
- request access to both text and non-text planes in fb mode
With this and the bus_space commit a few days ago I have usable X on my C360
and a PCI Visualize EG.
This needs testing on quite literally everything else.

(macallan)

add missing bus_space_*_stream() methods instead of quietly aliasing them
to their potentially byte-swapping regular counterparts

ok skrll@

(macallan)

rearrange things so we:
- find machine_ledaddr on Cxxx and the like without lcd at mainbus
- build without options USELEDS

(macallan)

add support for doubling a font's size, with optional smoothing
- works only on fonts up to 8 pixels wide for now
- smoothing attempts to detect pixel staircases and add pixels to smooth
  them out

(macallan)

add xhci

(macallan)

defflag AUDIO_DEBUG

(macallan)

yet another tsleep/wakeup -> cv_*

(macallan)

don't abuse the interrupt lock for waits in the event thread

(macallan)

remove ancient #if 0 block
NFC

(macallan)

tsleep()/wakeup() -> cv_*

(macallan)

check for empty ranges in pmap_page_physload() itself

(macallan)

avoid calling pmap_page_physload() with an empty range
this avoids a panic in uvm_pmap_physload()
now my C360 boots again

(macallan)

include glyph name in -d output

(macallan)

document command line option

(macallan)

- make colour depth in fb mode configurable, default to 8bit
- support ioctl(FBIO*CMAP)

(macallan)

following a hunch...
- cache DEC and FG registers, only write them if the value actually changes
- wait for the engine to go idle before writing DEC
- wait for FIFO slots on everything else
with this we avoid waiting if possible and still avoid overlapping blit and
fill commands

(macallan)

wait for the engine to go idle before issuing rectfill commands
we get occasional overlap with blit commands if we just wait for fifo slots
needs further investigation, it is possible that not all writes to drawing
engine registers are pipelined and of course we don't have docs

(macallan)

fix tpyo - now the glyph cache can actually work...

(macallan)

- use sx_wait() to avoid stalling the MBus
- request 32bit alignment for glyphcache cells

(macallan)

add counter to periodically drain the instruction queue in order to avoid
stalling the MBus during long SX operations
adapted from xf86-video-suncg14

(macallan)

allow drivers to specify horizontal alignment of glyph cache cells
for things like SX which have alignment restrictions

(macallan)

use macros compatible with xf86-video-suncg14 to issue SX instructions
much more readable, alignment weirdness is handled automatically and code is
interchangable

(macallan)

check for proper error value from OF_finddevice()
should fix PR57394

(macallan)

apparently the 'ahb' and 'ths' clocks were renamed to 'bus' and 'mod' in the
fdt at some point, so look for those as well
with this my pinebook's sensors work again

(macallan)

initialize the diagnostic register with the value suggested by the SunOS
header. This sets a bunch of undocumented bits and yields a 10% speed increase
when rendering antialiased text.

(macallan)

add flags for drivers to requesr R2L bit/byte-ordered fonts, default to
L2R, chack them in vcons_load_font() instead of just trusting that we'd get
what we need

(macallan)

make vcons_putchar_buffer() return a flag indicating if anything actually
changed, skip the actual drawing op if nothing did

(macallan)

fix logic-reversing tpyo in putone()

(macallan)

do some cleanup
the gm driver has been commented out for 20 years, time to remove it
while there, remove ofb and the old ADB code which have been commented out
since 2007

(macallan)

don't attach on cascaded mpics for now - we don't really use them for anything
yet and apparently setup causes problems elsewhere, which needs investigation

(macallan)

be more consistent with debug options

(macallan)

one more skiplist entry...

(macallan)

don't complain about mpic and dart not being configured
mpic is handled elsewhere, dart is an iommu we don't use yet

(macallan)

suppress 'not configured' message for mpic on G5s

(macallan)

support cascading:
- attach on both /u3 and /u4
- install IPI handling only if we're primary
- link to primary PIC if we're cascaded

(macallan)

first step towards MSI support on G5:
- allow both openpic and u3_ht to attach
- make sure openpic comes first, if we have both then u3_ht is cascaded

(macallan)

add comment to explain why we zero SPRG0 here
NFC

(macallan)

#ifdef DEBUG the previous change, it seems to cause trouble on qemu which
needs further investigation

(macallan)

if we have the ROM font, make it available to wsdisplay
needs testing on non-macppc

(macallan)

Explicitly zero SPRG0 before doing anything else - it's used to store a pointer
to struct cpuinfo, this way we can catch attempts to use it before it's set
up properly

(macallan)

convert more seriously early output to ofprint
regular console output needs more of the kernel in working order now, and this
stuff happ0ens long before the banner

(macallan)

in ofprint() only append \r if the last character is \n

(macallan)

move ofprint() to powerpc/ofw_machdep.c and make it official
now that console output uses locks it needs more of the kernel to function, so
for now use direct OF calls for earliest debug output

(macallan)

remove workaround for old clang - it's not needed anymore and caused problems
elsewhere

(macallan)

don't KASSERT(kpreempt_disabled()) while cold
pserialize_read_*() can be called *very* early in kernel startup these days

(macallan)

This is X11's 75dpi/term14.bdf, use if you want your console to look like an
xterm. Unlike QVSS this font contains accented characters, umlauts etc.

(macallan)

put back quotation mark removal in FAMILI_NAME statement which got broken
in a previous commit

(macallan)

keep track of character names and dump them into the header file if known

(macallan)

do not rely on every STARTCHAR/ENDCHAR sequence starting with ENCODING - the
standard doesn't specify in which order each glyph's properties are listed

(macallan)

set AHCI_QUIRK_EXTRA_DELAY for ASMedia ASM1061 - now it reliably attaches old
1.5GBit/s only disks

(macallan)

rename a local variable so it doesn't conflict with a global one
NFC

(macallan)

remove debug printf()

(macallan)

error out on proportional fonts, provide option to override

(macallan)

add -N option to override the font's name

(macallan)

limit the comment width in the header output to the size of the character cell

(macallan)

if there are comments in the bdf copy them into the header output

(macallan)

use [] instead of # to represent pixels in the header output, this way pixels
are somewhat closer to square and the preview looks closer to the actual
output.
suggested by uwe@

(macallan)

add -h option to write font data as a header file that can be compiled
into a kernel

(macallan)

move \n where it's supposed to be
now fonts wider than 8 are dumped correctly

(macallan)

fix tpyo - now we draw characters with a smaller bounding box than the
character cell correctly, like term14.bdf

(macallan)

add -d option to dump rendered characters to the console

(macallan)

- convert remaining error output to err()
- suppress console output unless -v is specified

(macallan)

nuke pasto in gffb_init()

(macallan)

force software fallback for drawing rectangles into offscreen memory only
on NV_ARCH_40 and up

(macallan)

more xf86SetModeDefaultName( after xf86CVTMode()

(macallan)

more xf86SetModeDefaultName( after xf86CVTMode()

(macallan)

more xf86SetModeDefaultName( after xf86CVTMode()

(macallan)

more xf86SetModeDefaultName( after xf86CVTMode()

(macallan)

sprinkle xf86SetModeDefaultName() since apparently xf86CVTMode() no longer
sets a mode's .name
now this works again on macppc

(macallan)

if we add a mode for the DFP port, make sure it has a name so we don't crash
later on
now this works again on my 12" powerbook

(macallan)

default to apple layout for both adb and usb keyboards

(macallan)

a driver for Wacom Intuos drawing tablets, from Yorick Hardy
this has been sitting in my tree long enough and works fine with the hardware
I have access to

(macallan)

regen

(macallan)

add some Wacom devices

(macallan)

we also need WSDISPLAY_COMPAT_PCVT and _SYSCONS

(macallan)

uncomment WSDISPLAY_COMPAT_USL
while there, add obiofan

(macallan)

initialize a variable to appease clang

(macallan)

check platform-do-*-mute properties to find out the resp. GPIO's polarity
now this works properly on pmac7,3

(macallan)

be lazy about clearing the cursor since most likely we're going to overwrite
it anyway. Same logic as in cgsix.

(macallan)

fix tpyo
thanks rin@

(macallan)

treat PictOpOver with a solid source and an empty mask as simple fill
somewhat less glitchiness in gtk3 but we're not quite there yet

(macallan)

disable accidentally committed debug goop

(macallan)

make debug output a bit more manageable

(macallan)

don't pretend to support PictOpSrc - those can be quite a bit more complicated
than simple copy ops
With this gtk3 buttons are still occasionally glitchy but at least we don't
crash anymore

(macallan)

consistently use SX_RENDER_DEBUG, hide some more debug goop

(macallan)

be lazy about clearing the cursor - in most cases we're about to overwrite it
anyway. Same logic as with VCONS_DONT_READ.
visible speedup in things like systat and top

(macallan)

add locations where pkgsrc installs X11 fonts to the default font path

(macallan)

#if 0 now unused audio_track_is_record() to appease clang

(macallan)

the register is only one byte, act like it

(macallan)

KNF nits, NFC

(macallan)

we've been using the power key for its intended purpose for yesrs, which
makes breaking into DDB awkward, so let's do what everyone else does and
use Ctrl-Alt-Escape / Cmd-Escape depending on the layout variant

(macallan)

get machine_ledaddr from PDC if we find one
now USELEDS works on my c360

(macallan)

allow com at dino to actually attach

(macallan)

bump MSGBUFSIZE

(macallan)

Check the PROM for the amount of available VRAM instead of guessing, and use
it all, now that we know we can.

thanks foo bar

(macallan)

turns out we can map the full amount of VRAM on a 4MB board even when in
double buffer mode

thanks foo bar

(macallan)

fix calculation for start of off-screen memory in 24bit colour

(macallan)

unmask interrupt bits leading to other interrupt registers
now hardware interrupts actually fire instead of being found pending by the
timer interrupt
with this SCSI throughput on my c360 is what you'd expect from an UW esiop

(macallan)

include pmap_pvt.c in bridge mode as well

(macallan)

more __diagused to appease clang

(macallan)

initialize ntag with the old dmatag so we never end up with an invalid one
suggested by jmcneill@

(macallan)

bump MAXTSIZ
now clang runs again

(macallan)

shut up an unused variable warning
apparently clang gets confused by bus_dmatag_subregion() being EOPNOTSUPP
on powerpc

(macallan)

when enumerating ports and endpoints treat missing 'reg' properties as zero
ok jmcneill:
Looking at Linux. If port or endpoint are missing a 'reg' property it defaults to 0.
Please make our code do the same.

see https://github.com/devicetree-org/dt-schema/blob/main/dtschema/schemas/graph.yaml

with this my pinebook has a usable console again

(macallan)

don't blindly enable memory and IO access on macppc

(macallan)

add flag to disable blindly enabling memory and IO access on everything we
find

(macallan)

more KASSERT vs. DIAGNOSTIC fallout

(macallan)

make this compile again by moving a KASSERT into the #ifdef DIAGNOSTIC block
it depends on

(macallan)

add some comments

(macallan)

make ALU instructions use the same format as load/store instructions
same binary generated

(macallan)

add macros to simplify issuing SX instructions, hide the address alignment /
displacement shenanigans SX needs, and make the code more resemble the
inline assembler source it arguably is

(macallan)

always drain the pipeline before writing SX_ROP_CONTROL

(macallan)

only write the mask register if we're actually changing it

(macallan)

cg14_invert():
- use only 32bit accesses
- use byte mask / ROPs for partial writes

(macallan)

break down large, unaligned copies into chunks that CG14Copy8_short_*()
can handle, make sure destinations are aligned whenever possible
now we only fall back to byte-wise access for really small copies, which
we probably shouldn't bother throwing at SX in the first place

(macallan)

skip reading destination fb if we're going to overwrite it, only read words
we need to partially write
while there write mask registers only if we're actually going to need them
another 20% speedup

(macallan)

CG14Copy8_short_rop(): skip the funnel shifter if source and destination are
aligned. Small but measurable speedup.

(macallan)

add another Copy8() variant:
- supports unaligned source and destination
- uses all 32bit accesses
- supports copies up to 124 pixels wide so an entire line fits into registers
  and we can ignore x direction
... mostly an exercise in learning how to use the funnel shifter
TODO:
- skip the funnel shifter if source and destination are aligned
- skip fb reads where possible, like straight GXcopy

(macallan)

be more consistent with underscores in instruction names

(macallan)

add SX_ROPB instruction

(macallan)

add shift instructions

(macallan)

shutup debug goop

(macallan)

start optimizing Copy8() operations
- only go right to left if srcY == dstY in the same pixmap
- special case copy where src and dst X have the same alignment
- special case the above where we don't need to read dst
scrolling should now be about as fast as SX can go

(macallan)

use 32bit accesses for non-trivial cases in Solid8() as well
while there make GXset and GXclear special cases of GXcopy
with this we only ever read the destination pixmap if we really need to

(macallan)

draw very narrow rectangles correctly

(macallan)

fix shadow fb geometry calculations
... not that a shadow fb helps much on this hardware.

(macallan)

first step towards acceleration in 8 bit colour
- Copy() uses byte mode
- Solid() uses 32bit writes where possible for plain fills
... mostly works, but could be quite a bit faster

(macallan)

support 8 bit colour
no acceleration yet, the code is riddled with 32bit per pixel assumptions

(macallan)

remove old guesswork code to determine VRAM size - we can rely on OBP now

(macallan)

fix license

(macallan)

crude beginnings of a driver for the Southland Media / Quantum 3D MGX
basic acceleration mostly works in 24bit colour
needs native hw cursor code
windowmaker produces visible artifacts, but it's good enough to run amiwm and
a few xterms

(macallan)

provide an endian-flipped view of the framebuffer via mmap() if we know how
for now this is sparc64 only

(macallan)

remove accidentially committed debug goop
thanks ryo@

(macallan)

actually mmap() the blitter registers when asked to, while there do some
magic number reduction

(macallan)

add mgxreg.h

(macallan)

install mgxreg.h so we can use it from Xorg

(macallan)

recognize Southland Media MGX

(macallan)

return 0 in mgx_ioctl()::FBIOG*
now Xorg can find us

(macallan)

use quirk bit for devices that need sc_alwayson instead of adding yet another
if (uha->uiaa->uiaa_vendor == USB_VENDOR_..., remove entries now covered
by UQ_ALWAYS_ON

(macallan)

add a quirk for devices that need ums::sc_alwayson

(macallan)

regen

(macallan)

add PixArt Raspberry Pi Mouse

(macallan)

explain why we skip over dev 1f func 7 in config space access functions
no functional change

(macallan)

fix license copypasta
NFC

(macallan)

this driver allows to read the speeds of fans controlled via mac-io/fans/*
found in some G5, for example my 7,3
based on guesswork and observation

(macallan)

switch smu to common fan control code

(macallan)

set EXA_MIXED_PIXMAPS so we don't get pixmaps we can't handle in NVSolid()
while there shut up some related debug output and remove unused variables

(macallan)

pass sensor locations to i2c devices if we can find them
tested on 2nd gen Mini

(macallan)

make thermal zone parameters configurable by sysctl

(macallan)

check temperatures a bit more often
while there, explain why the delay is variable

(macallan)

adjust comments, break a long line
NFC

(macallan)

first step towards abstracting thermal zone management out of the fcu driver
( and eventually, the smu driver )
todo:
- add sysctl()s to set zone parameters
- handle envsys
- adapt smu

(macallan)

move the fcu driver into arch/macppc where it belongs

(macallan)

don't put CPUs in separate zones, turns out there really isn't much that
keeps the airflow apart around them

(macallan)

fix off by one which resulted in all idle time reported as interrupt time
final fix from nick@

(macallan)

do ii2 locking dance in psoc_dump()

(macallan)

use opt_psoc.h, suppress some debug output in !PSOC_DEBUG

(macallan)

use opt_lmu.h
while there, remove an obsolete comment

(macallan)

defflag LMU_DEBUG and PSOC_DEBUG

(macallan)

add \n to attach output

(macallan)

don't prop_object_retain(NULL)

(macallan)

fix wsdisplay attachment in the New Order Of Things

(macallan)

use probed VRAM size to determine where to put the cursor instead of blindly
assuming 16MB
also make sure the glyph cache can't overlap with the cursor

(macallan)

add -l option to dump a list of available fonts

(macallan)

add an ioctl() to get a list of fonts currently available via wsfont

(macallan)

clarify comment in previous
( this is a hack, should go away when the root cause is fixed etc. )
no functional change

(macallan)

avoid mapping 0xfffff000 - my beige G3 DSIs on it
with this my the machine boots again
tested on a variety of G4 and G5 models with no problems

(macallan)

try WSDISPLAYIO_GET_EDID if DDC2 fails
now my iMac G5 correctly identifies its display

(macallan)

add basic support for iMac G5 audio:
- match AOAShasta
- deal with headphone gpios labeled as lineout in OF
- set sc_mode to SNAPPER_IS_PCM3052 - OF makes it look like an unlabeled
  TAS3004
TODO:
- actually support pcm3052 instead of treating it like a sw codec

(macallan)

switch G5 iMacs to full speed

(macallan)

add support for the CPU temperature sensor found in iMac G5s

(macallan)

add gffb
tested on 20" 1.8GHz iMac G5

(macallan)

more node name adjustments
also, pass sensor names if we can find them
now we find and currectly use the hard drive temperature sensor on my iMac G5

(macallan)

deal with node name inconsistencies between PowerMac10,x and 8,x
Now we find fans and iic devices on iMac G5

(macallan)

match Geforce FX5200 Go found in some iMac G5 models

(macallan)