| Age | Commit message (Collapse) | Author |
|---|
|
MI float.h which pulls in and defines the values that are needed from
there, and repair sys/limits.h so that it defines the values it needs
as well (depending on POSIX version, XPG version, etc). guenther has
a more exact selection of that coming for limits.h.
this also fixes a few mistakes for the vax.
reviewed by kettenis and guenther.
|
|
(level 6). Will be used shortly.
|
|
|
|
there is no way we can tell models 360 and 362 apart by probing for an SGC
bus, or for a frodo(4) chip since of the `lil' black swan' 3x2 models, only
model 382 has frodo. So don't bother trying to tell models 360 and 362
apart, and use a common 36X symbol. No functional change except for useless
code removal.
|
|
- add nearbyint, nearbyintf and nearbyintl implemented using fenv
|
|
Discussed and okay drahn@. Okay deraadt@.
|
|
|
|
Okay guenther@, millert@.
|
|
|
|
blocking other cleanups
ok miod@
|
|
fallback definition in <sys/sched.h>, so that there is no hidden include
ordering requirement between <machine/intr.h> and <sys/sched.h>.
ok deraadt@ tedu@
|
|
This is a noop on all arches but hp300, where this fixes (at least) IPL_BIO
mutexes.
|
|
corresponding physical memory; makes procmap(8) happy on hp300.
|
|
MD code would free resources that couldn't be freed until we were no
longer running in that processor. However, it's is unused on all
architectures since mikeb@'s tss changes on x86 earlier in the year.
ok miod@
|
|
|
|
Right now when mi_switch picks up the same proc, we didn't clear the
flag which would mean that every time we service an AST we would attempt
a context switch. For some architectures, amd64 being probably the
most extreme, that meant attempting to context switch for every
trap and interrupt.
Now we clear_resched explicitly after every context switch, even if it
didn't do anything. Which also allows us to remove some more code
in cpu_switchto (not done yet).
miod@ ok
|
|
of making it dynamic and the smallest value above the former three. Idea
from NetBSD.
|
|
Not sure what's more surprising: how long it took for NetBSD to
catch up to the rest of the BSDs (including UCB), or the amount of
code that NetBSD has claimed for itself without attributing to the
actual authors.
OK deraadt@
|
|
match the real hardware level used for them).
So instead of keeping a few loose variables to store the runtime computed
psr bits to use for the logical IPL levels, just keep a complete array of
all psr values. Thus we can have a correct splassert_check().
|
|
|
|
|
|
cpu_disklabel can go away, since nothing anymore needs to use it; ok miod
|
|
|
|
to support hotplug media on most architectures. disklabel setup and
verification done using new helper functions. Disklabels must *always*
have a correct checksum now. Same code paths are used to learn on-disk
location disklabels, to avoid new errors sneaking in. Tested on almost all
cases, testing help from todd, kettenis, krw, otto, dlg, robert, gwk, drahn
|
|
|
|
|
|
help miod@, art@
ok deraadt@
|
|
device_register() function -- even if it does nothing. reduces the
cpp-based blather different between architectures
idea ok'd by miod; tested on all architectures (except a few miod will
need to cleanup because he has them)
|
|
unconditionnaly.
|
|
__BUS_SPACE_ALIGNED_ADDRESS.
|
|
it's a good idea to use atomic.h operations on it. This mechanic
change updates all bit operations on p_flag to atomic_{set,clear}bits_int.
Only exception is that P_OWEUPC is set by MI code before calling
need_proftick and it's automatically cleared by ADDUPC. There's
no reason for MD handling of that flag since everyone handles it the
same way.
kettenis@ ok
|
|
rt and such; tested and ok miod drahn
|
|
right now that are supposed to be atomic with respect to interrupts and
SMP: atomic_setbits_int and atomic_clearbits_int.
All architectures other than i386 and amd64 get dummy implementations
since at first we'll be replacing operations that are done with
"a |= bit" and "a &= ~bit" today. More proper implementations will follow
kettenis@, miod@ ok
|
|
use of MD spl code bowels. No functional change.
|
|
pointers, to be used for the various bus_space operations.
intio devices no longer need to pass an intiobase-relative address to
bus_space_map.
|
|
|
|
for cpu_swapin() on hppa* which is kept).
|
|
faster uvm initialization.
|
|
registers to program an 1GB untranslated window over the physical memory,
and use it to define __HAVE_PMAP_DIRECT for faster pool operation.
|
|
which bites the dust.
|
|
declarations in <machine/{cpu,param}.h> and elsewhere.
|
|
m68k-based platforms.
|
|
makes uvm_page.o smaller.
|
|
ok miod@
|
|
rename the types using the __ protected namespace (e.g. __size_t).
Idea from FreeBSD.
|
|
|
|
When not running on serial console, the PROM will pick the ``internal''
graphics device, and if it does not exist, the DIO device with the lowest
select code, and will resort to SGC devices if no internal or DIO device
was found.
However, the current logic would search for a certain frame buffer type first,
then for its best hardware location, following the order they are listed
in conf.c.
By replacing gazillions of almost-exactly-duplicated frame buffer code with
one single instance, which is device type-agnostic, we can match the PROM
(and the bootblocks) logic again. Plus this saves a few KB of code!
|
|
This means we are no longer aborting the console scans even if the final
console has been found, but on the other hand makes some logic simpler;
plus it will be necessary for upcoming scode vs device loop order reversal.
No user-visible change.
|
|
No functional change.
|
|
cruft for old compilers and environments. there may be a 2nd round of
polishing after this; ok miod and others
|
