NKMEMPAGES_MAX_DEFAULT is no longer used. Remove it from param.h.
OK miod@

Unify and bump some of the NMBCLUSTERS defines. Some archs had it set to
4MB which is far too low especially when the platform is able to run MP.
New limits are, amd64 = 256M; arm64, mips64, sparc64

Unify and bump some of the NMBCLUSTERS defines. Some archs had it set to
4MB which is far too low especially when the platform is able to run MP.
New limits are, amd64 = 256M; arm64, mips64, sparc64 = 64M; alpha, arm,
hppa, i386, powerpc = 32M; m88k, sh = 8M
Still rather conservative numbers but much better than before. At least
some hangs of arm64 build boxes was caused by this.
OK kettenis@, visa@

show more ...

Increase the number of mbufs on most architectures. This is based
on a guess how much memory a typical machine has. If the value is
too high, users may run out of kernel memory. Then we will have

Increase the number of mbufs on most architectures. This is based
on a guess how much memory a typical machine has. If the value is
too high, users may run out of kernel memory. Then we will have
to adjust this again.
OK claudio@ deraadt@

show more ...

Only define KERNBASE if defined(_KERNEL), for nothing in userland needs it,
and only if not already defined. Allows for KERNBASE to be defined elsewhere.

PGSHIFT and PGOFSET are now contained inside the kernel namespace.

ALIGNBYTES/ALIGN/ALIGNED_POINTER can move to the MI file.

refactor sys/param.h and machine/param.h. A lot of #ifdef _KERNEL is added
to keep definitions our of user space. The MD files now follow a consistant
order -- all namespace intrusion is at the tai

refactor sys/param.h and machine/param.h. A lot of #ifdef _KERNEL is added
to keep definitions our of user space. The MD files now follow a consistant
order -- all namespace intrusion is at the tail can be cleaned up
independently. locore, bootblocks, and libkvm still see enough visibility to
build. Checked on 90% of platforms...

show more ...

Provide namespace-safe alignment macros in <machine/_types.h>, with
compat names kept in <machine/param.h>. In <sys/socket.h>, pull
in <sys/_types.h> instead of the namespace polluting <machine/para

Provide namespace-safe alignment macros in <machine/_types.h>, with
compat names kept in <machine/param.h>. In <sys/socket.h>, pull
in <sys/_types.h> instead of the namespace polluting <machine/param.h>
and completely eliminate __CMSG_ALIGN, replaced by _ALIGN

ok deraadt@

show more ...

Make sure the ALIGN() macro uses u_long on all platforms for consistency, and
update the comment block accordingly.

Normalize sentinel. Use _MACHINE_*_H_ and _<ARCH>_*_H_ properly and consitently.

Discussed and okay drahn@. Okay deraadt@.

Update comment to no longer mention a page is 4KB, since this may not be the
case.

bit less aggressive about hiding; ok miod

Protect more definitions userland doesn't need to care about with _KERNEL

Switch page size from 4KB to 16KB on R10k kernels without R5k support
(i.e. IP27 and IP30 sgi kernels).

Use a whole page for msgbuf if page size is larger than 4KB, instead of
the current 8KB.

Support for 16KB page size kernels; page size is now set in <machine/param.h>
rather than <mips64/param.h>.

For now, kernels are kept at 4KB to give people some time to build 16KB
compatible binarie

Support for 16KB page size kernels; page size is now set in <machine/param.h>
rather than <mips64/param.h>.

For now, kernels are kept at 4KB to give people some time to build 16KB
compatible binaries; this will change before the end of this release cycle.

Use of 16KB page size kernels yields a 18% speedup (which, offset by the
1.6% slowdown caused by the pmap changes, yields a 16.6% overall speedup).

show more ...

Use a pool to manage pmap pte pages and top level segment table, instead of
directly allocating pages from uvm; this will allow us to eventually use
a different kernel page size without having to alt

Use a pool to manage pmap pte pages and top level segment table, instead of
directly allocating pages from uvm; this will allow us to eventually use
a different kernel page size without having to alter the pmap structures
layout.
No functional change; measured slowdown of 1.6% for 4KB page kernels.

show more ...

Drop almost unused <machine/psl.h> on sgi; move USERMODE() definition from
there to trap.c which is its only user. This also cleans up multiple
inclusion of <machine/cpu.h> (because <machine/psl.h> i

Drop almost unused <machine/psl.h> on sgi; move USERMODE() definition from
there to trap.c which is its only user. This also cleans up multiple
inclusion of <machine/cpu.h> (because <machine/psl.h> includes it) in many
places.

show more ...

Fix signedness of comparison used to know whether we have already reached
the next scheduled clock interrupt; the comparison would before always be
true, causing the clock to really run at hz/2.

Whi

Fix signedness of comparison used to know whether we have already reached
the next scheduled clock interrupt; the comparison would before always be
true, causing the clock to really run at hz/2.

While there, remove unused nanodelay() and attempt to clean clock
initialization a bit.

show more ...

convert arm and mips64 platforms to ptoa/atop

tested by maja@

Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
mbuf constants from MD param.h to MI param.h.
Besides being the same on every arch, things will
most probly break if any arch has different values

Move the MSIZE, MCLSHIFT, MCLBYTES and the MCLOFSET
mbuf constants from MD param.h to MI param.h.
Besides being the same on every arch, things will
most probly break if any arch has different values
then the others.

The NMBCLUSTERS constants needs to be MD though;

ok miod@,krw@,claudio@

show more ...

Crank MAXPHYS to the regular 64KB.

delete userland DELAY() version; ok miod

remove unused bdbtofsb(bn) macro

found by drahn@

Revert 1.12 and go back to 32K MAXPHYS; 64k MAXPHYS exposes subtle bugs
and is not reliable enough.