| Age | Commit message (Collapse) | Author |
|---|
|
are emulated by disabling interrupts which requires a lot more header files
than a simple atomic operation should be need. ok deraadt@
|
|
|
|
recommit pmemrange:
physmem allocator: change the view of free memory from single
free pages to free ranges. Classify memory based on region with
associated use-counter (which is used to construct a priority
list of where to allocate memory).
Based on code from tedu@, help from many.
Useable now that bugs have been found and fixed in most architecture's
pmap.c
ok by everyone who has done a pmap or uvm commit in the last year.
|
|
ok miod@
|
|
board-specific defines to prevent any risk of collision. This also adds
clock support for AV530 family, and timecounter support (cio code sync'ed
with mvme88k). And various bugs fixed in the process.
This is enough to get models 4600 and 530 to run multiuser with a Hawk
Ethernet VME card (the onboard Ethernet is not supported yet, coming soon).
There is no way to share a disk with DG/UX yet, the kernel (and fdisk(8))
needs to become aware of its ways.
|
|
|
|
|
|
|
|
|
|
like it is everything.h
|
|
like it is everything.h
ok tedu
|
|
|
|
ok tedu
|
|
OK tedu@
|
|
|
|
|
|
sysctl.h was reliant on this particular include, and many drivers included
sysctl.h unnecessarily. remove sysctl.h or add proc.h as needed.
ok deraadt
|
|
by returning ENXIO instead of ENOENT, to essentially indicate hotplug
sensor that has gone away. Accessing beyond the end of the sensordev
list still returns ENOENT, so that you can see there are no further devices.
ok kettenis oga
|
|
|
|
chip initialization values (SCNTL0, DMODE, DWT, CTEST7).
No functional change on hppa; required by another upcoming attachment.
|
|
HARD uncached on anything since the P3), use just PCD which is UC-
(uncached but mtrrs can override it to write combining).
Now kernel mappings of WC-mtrred memory actually get write combining
memory.
the speed up is very very nice.
ok kettenis@, deraadt@, beck@ (who likes his scrollbars non-wibbly)
|
|
ok kettenis@
|
|
ok kettenis@
|
|
ramdisks.
OK marco@
|
|
ramdisks.
OK marco@
|
|
Also features support for {awkw,bast}ard 6:1 CMMU:CPU configurations (4I2D).
Tested on model 4605, which runs up to cpu_initclocks(), which is not written
for this system family yet. No regression on model 4300.
|
|
is missing.
While there, provide more information when the system is not recognized.
|
|
to be 00:00:00:00:00:00 if `-z' is passed; this allows netboot to work on a
machine which has lost its nvram (despite the ethernet address not being
stored in nvram at all, stupid prom can't be bothered to get it right...).
Crank netboot version.
|
|
on vax, since interrupts on syscon are edge-triggered. Fixes the onboard
lance getting stuck after a while.
|
|
|
|
|
|
to sometimes fail, by only using cache operation sequences known not to
trigger the errata, at the expense of a small (barely noticeable) performance
hit.
This workaround is only enabled when compiling a MULTIPROCESSOR kernel, since
the problem will not arise when using only one processor.
This change, refined over a few months, finally makes GENERIC.MP run stably
on 197DP systems. Of course, just writing this is enough to invoke Murphy's
law and make it fail again in an even more obscure and rare way...
|
|
ok miod@
|
|
|
|
all the interrupt handles when reprioritizing the interrupt on reception,
but always run the handler at the desired priority. Make sure
ci_handled_intr_level is set correctly. Gets rid of splassert warnings
seem on many of the PCIe systems with mpi(4).
tested by deraadt@, jbg@
|
|
bit set. While there, add an #ifdef MULTIPROCESSOR to make the diff between
amd64 and i386 smaller.
ok art@
|
|
|
|
|
|
paths. Spotted by Mike Belopuhov.
|
|
and skip function codes 132 to 135, inclusive, on models 362 and 382 - this
is a fake DIO device probably intended to let non-SGC aware (i.e. legacy)
HP-UX lowlevel code be able to ``detect'' a frame buffer.
This change (as well as letting sgc attach) is untested due to the lack of any
3x2 hardware, but shouldn't make things worse.
|
|
bitsavers confirm the long suspicion the onboard frame buffer on these
machines is an SGC device.
|
|
every ip35 but Fuel), when fetching spdmem records, so the DIMM number
reported matches the numbering on the PIMM.
|
|
twice.
From Ted Patterson.
|
|
|
|
a process instead of using curproc. ok deraadt
|
|
robert@.
|
|
|
|
doing if (p != NULL) pmap_extract() else vtophys() in a loop, just do
pmap_extract unconditionally.
ok miod@ (he found a typo, all hail miod!)
|
|
!= 0 bit. Makes GCC 4.2.1 happy.
|
|
recent-ish macs with macgpio(4). Basically dropping one to DDB when pushed.
...and not to be confused by the "reset" button ;-)
feedback and ok drahn@
|
