Avoid unnecessary (and pointless) references to pcb->pcb_savefpu.
They were all in asserts.

(dsl)

Add fpu_set_default_cw() and use it in the emulations to set the default
  x87 control word.
This means that nothing outside fpu.c cares about the internals of the
  fpu save area.
New kernel modules won't load with the old kernel - but that won't matter.

(dsl)

#include <arm/locore.h>

(matt)

fix typo

(martin)

Use Aux32Info not AuxInfo. Noticed by Maxime Villard.

(njoly)

use aprint.

(christos)

Update a comment

(skrll)

Bump MD size. Verified working.

(skrll)

Determine whether the cpu supports xsave (and hence AVX).
The result is only written to sysctl nodes at the moment.
I see:
machdep.fpu_save = 3 (implies xsaveopt)
machdep.xsave_size = 832
machdep.xsave_features = 7
Completely common up the i386 and amd64 machdep sysctl creation.

(dsl)

Use cprng_strong32 for LINUX_AT_RANDOM on amd64 too.

(njoly)

Fix wrong KASSERTs. Do not compare size vs. entries count.

(njoly)

Fix obvious typo

(martin)

ttioctl always gets a valid lwp reference. Replace attempt to handle a NULL
reference in only one place with a regular assertion.

(mlelstv)

skip unparseable addresses.

(mlelstv)

Provide a prototype for xputchar

(martin)

Remove misplaced #endif

(martin)

add */po and itm/testsuite.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

remove the testsuite and po subdirectories.
27292 files are removed in this series of commits.

(mrg)

don't look in libiberty/testsuite for anything, it's going away.

(mrg)

tmux-1.9a out.

(wiz)

Add way_size to arm_cache_info
Fix arm67_tlb_purge prototype

(matt)

Add L1_TABLE_SIZE_REAL

(matt)

Don't manipulate the pte to get uncached memory, use PMAP_NOCACHE instead.
Convert footbring to kenter_pa/kremove

(matt)

Trim include files

(matt)

Drop empty priority lists, not the full ones. Fixes kern/48611.

(mlelstv)

Deal with non-4KB page sizes

(matt)

Deal with non-4KB page sizes

(matt)

include <arm/locore.h>

(matt)

Deal with non-4K page sizes.

(matt)

use kenter_pa/kremove

(matt)

Allow kernels without tsciic to build.

(martin)

Re-use the unused ci_cpu_serial[3] to save the highest cpuid values
  for the normal and extended leafs.
(The 'normal' one might be luring in the global cpulevel.)
Read the 'extended feature' from cpuid.80000001.%ecx/edx into
    ci_feat_val[3/2] just after saving cpuid.1.%ecx/dx in ci_feat_val[1/0]
    instead of doing it separately for amd k678 and via c3 processors
    in their probe functions and repeating it for all cpus a few instructions
    later when x86_cpu_topology() is called.
x86_cpu_topology() is only called from cpu_probe() and really doesn't
  deserve its own source file. Chasing the setup code is bad enough anyway.

(dsl)

Fix examples, from Henning Petersen in PR misc/48612.

(dholland)

pipe2 requires fcntl.h for its flags; PR 48614 from Steffen Daode Nurpmeso.
Also fix the wording for EINVAL as suggested by Robert Elz.

(dholland)

fix typo, from Henning Petersen in PR misc/48613

(dholland)

Missed one __sync_* op (or gcc4.8 does inline it, while 4.5 does not?)

(martin)

Try to hide the C runtime implementation specific __sync_* ops from librump,
to avoid duplicates.

(martin)

Explicitly include armreg.h (for the interrupt enable bit)

(martin)

Add missing __sunc_* ops for sun2

(martin)

fix -w output

(mlelstv)

add u option to getopt again.

(mlelstv)

Add missing __sync_* ops to libc

(martin)

Use more markup. Add "flag" in a sentence.

(wiz)

Mark up NULL with Dv.

(wiz)

Add the missing __sync_* ops to libc

(skrll)

Update arguments of vrecycle(), description of getnewvnode() and
the vnode flags.

(hannken)

Update the description of vfs_busy() and vfs_unbusy().

(hannken)

Add the missing __sync_* ops to libc

(martin)

Another layout change

(skrll)

Layout changes - no functional change.

(skrll)

Move the __sync_* ops added in the previous change to a libc-only section

(martin)

Define __HAVE_ATOMIC64_OPS if EABI && ARMv6 or later.

(matt)

Simplify error path.

ok christos@

(maxv)

Match *INSTALL* (so we can match *INSTALL.MP)

(matt)

Rearrange KERNEL_BASE_PHYS

(matt)

Rework PIC method to be simplier.  Change be more cortex neutral.

(matt)

Add EMAC and CPUCFG registers

(matt)

Remove unnecessary  struct simplelock forward declaration.

(skrll)

Remove struct simplelock forward declaration.

(skrll)

Remove unnecessary include of sys/simplelock.h

(skrll)

Provide the missing __sync_* ops for earlier arm versions

(martin)

Use code model small for LLVM, it should be equivalent to GCC's medlow.

(joerg)

Use small code model for LLVM on SPARC64, it should be equivalent to
GCC's medlow.

(joerg)

Provide all __sync_* ops in libc.

(martin)

sun4v: Rename TLB_ defines to SUN4U_TLB_ so entries created using TSB_DATA() are properly setup for sun4u and sun4v. Relocate the cputyp variable from autoconf.c to locore.s and make it const in param.h so optimized code can be generated. Parts from OpenBSD. Optimization suggested by nakayama@. OK martin@, mrg@, nakayama@

(palle)

Provide the missing __sync_* ops for sparc.

(martin)

No need to include sys/simplelock.h here.

(skrll)

Add a few missing namespace includes

(martin)

Declare 16 and 8 bit atomic CAS - while it is not usually sane to expect
these functions to work in MI code, we need them to provide some of the
__sync_* functions gcc and other compilers expect nowadays on some
architectures.

(martin)

Provide 8 and 16 bit sync ops (using 16 bit and 8 bit cas)

(martin)

Fix previous: move the alias declaration outside of a #ifdef _LP64

(martin)

Ticket #1033

(sborrill)

Pull up the following revisions(s) (requested by tsutsui in ticket #1033):
sys/arch/atari/dev/ite.c: revision 1.72

Fix kernel crash when a user tries to switch to nonexistent terminal.
Fixes PR/48599.

(sborrill)

Ticket #1033

(sborrill)

Pull up the following revisions(s) (requested by tsutsui in ticket #1033):
sys/arch/atari/dev/ite.c: revision 1.72

Fix kernel crash when a user tries to switch to nonexistent terminal.
Fixes PR/48599.

(sborrill)

Add the tsciic manual pages to the sets lists.

(jdc)

Add manual page for tsciic and cross reference it from related pages.

(jdc)

Add tsciic (alpha MI).

(jdc)

Add tsciic and more (commented out) DS20L i2c devices.

(jdc)

Add tsciic, a driver for the DECchip 21272 Core Logic chipset I2C controller.
Tested on DS20L.

(jdc)

Ticket #1033

(sborrill)

Pull up the following revisions(s) (requested by tsutsui in ticket #1033):
sys/arch/atari/dev/ite.c: revision 1.72

Fix kernel crash when a user tries to switch to nonexistent terminal.
Fixes PR/48599.

(sborrill)

Provide 32bit __sync ops based on atomic_cas_32.

(martin)

Ticket #1032

(sborrill)

Pull up the following revisions(s) (requested by mlelstv in ticket #1032):
sys/dev/usb/umass.c: revision 1.148
sys/dev/usb/umassvar.h: revision 1.35
sys/dev/usb/usb_mem.c: revision 1.64

Fix problem where umass might allocate buffers in interrupt context which
causes it to fail or a DIAGNOSTIC kernel to panic.

(sborrill)

Add missing __sync_lock_test_and_set_* aliases for PR lib/48601

(martin)

Remove excemption for i386 (PR lib/48601) - missing aliases will be
added to the C code.

(martin)

Fix typo in fix for PR/48606

(skrll)

Revert rev1.38. The header already begins with EXEC_SCRIPT_MAGIC="#!".
So it can't be ELFMAG="\177ELF" at the same time.

ok christos@

(maxv)

Increase LINUX32_ELF_AUX_ENTRIES to avoid overrun in linux32/. Also,
add comments and KASSERTs to make sure people don't forget to increase
XX_AUX_ENTRIES's when adding vectors.

Reported by martin@ (CV), with suggestions from chs@.

ok martin@ chs@

(maxv)

Properly check the section size to avoid out-of-bound reads. The
computed size must be the exact same size that is indicated in
sh_size.

ok agc@ christos@

(maxv)

defparam KERNHIST_DELAY

(matt)

For now, don't reset arm_cache_prefer_mask unless both l1 caches are PIPT

(matt)

Add -X (don't cross mount points when recursing) from tls@

(christos)

PR/48606: Lloyd Parkes: Drivers not using ifp->if_input but using ether_input
directly.

(christos)

Update to version 20140220

Changes from previous version:

+ portability fixes from xtraeme for his Linux distribution:
+ add search for ar(1) into autoconf
+ define __printflike if it's not already defined
+ fix missing asprintf (I fixed this differently)

(agc)

Keep track of what each cache is (VIVT/VIPT/PIPT).

cpu0: 32KB/32B 2-way L1 VIPT Instruction cache
cpu0: 32KB/64B 4-way write-back-locking-C L1 PIPT Data cache
cpu0: 256KB/64B 8-way write-through L2 PIPT Unified cache

(matt)

tmux-1.9 out.

(wiz)

#include <sys/param.h>

(matt)

Include gtmr_var.h

(matt)

Add support for running BE

(matt)

Add prototype for awin_cpu_hatch

(matt)

Add missing __sync_* ops

(martin)

Use right register when setting ttbr

(matt)

(Belatedly) add lmenv.

(jdc)

Use .Nm for 'ls'.

(wiz)

Add -O (only leaf files) and -P (print full path), from tls@

(christos)

This doesn't need fpu.h, but should include ucontext.h

(dsl)

Move the amd64 and i386 pcb to the bottom of the uarea, and move the
  kernel stack to the top.
Change the pcb layouts so that fpu save area is at the end and is
  64byte aligned ready for xsave (saving the ymm registers).
Welcome to 6.99.32

(dsl)

This needs stdint.h in userspace (for uint64_t)

(dsl)

cpu_cortex doesn't need the arm11 cpufuncs any more

(matt)

armv7 doens't need to use the arm11 routines anymore.

(matt)

Add armv7 versions of tlb routines.

(matt)

Refactor and cleanup a bit.  Prepare for ASIDs.

(matt)

Make the tlb_flush*SE take a vaddr_t, not a u_int.

(matt)

Simplify the variants returning a bool

(martin)

for non-of-power-2 sets, we need to round to next power of 2.  (which is
simply doubling the value and then subtracting 1).

(matt)

Make sure to clear MDLWP_NOALIGNFLT on !aapcs or !(armv6 && armv7)

(matt)

Zero the block on failure, not just the first pointer in it.

(joerg)

Add prototype for main.

(joerg)

Make sure AFLT_ENABLE in the cpuctrl mask for armv7_setup.

(matt)

Add dhcpcd(8) to luna68k sysinst.  Also bump ramdisk size.

(tsutsui)

Fix formatting errors in last commit

(sborrill)

Ticket #1898

(sborrill)

Pull up the following revisions(s) (requested by prlw1 in ticket #1898):
lib/libpthread/pthread_cond.c: revision 1.62
lib/libpthread/pthread_mutex.c: revision 1.57,1.59

Partial fix for thread deadlock commonly observed with named.
Also address PR/44756.

(sborrill)

Bail out in case m_pulldown failed.

(joerg)

Ticket #1029

(sborrill)

Pull up the following revisions(s) (requested by prlw1 in ticket #1029):
lib/libpthread/pthread_cond.c: revision 1.62
lib/libpthread/pthread_mutex.c: revision 1.57,1.59

Partial fix for thread deadlock commonly observed with named.
Also address PR/44756.

(sborrill)

Don't dereference NULL pointers.

(joerg)

Don't use conditional format strings, if they aren't equivalent in the
arguments they use.

(joerg)

Correct the implementation of mvderwin, it now works as specified by
SUSv2.

(blymn)

Don't print an empty line after the debug message "Enabled Dynamic
Window Scaling".

(gson)

Since fpu.c is contained in every kernel, and it calls routines in
convert_xmm_s87.c we need to include convert_xmm_s87.c in all kernels.

Fixes building of kernels that do not have COMPAT_NETBSD32 defined.

(pgoyette)

Use sys/endian.h for bswapXX.

(joerg)

Call memset in a loop, not loop around doing nothing and calling memset
once.

(joerg)

Mark radeonfb_unclip as unused.

(joerg)

Use a format string.

(joerg)

remember to commit this file too

(pooka)

+ resettodr

(pooka)

make implementation of rumpuser_kill() match the documentation

(pooka)

document rumpuser_kill()

(pooka)

Translate signal numbers between rump kernel and (POSIX) host.

(pooka)

Remove the "host" signal model.  I'm not sure it ever made any sense.
I'm also quite sure nobody ever really wanted to use it.

Change the "raise" sigmodel to include the rump kernel PID in the call
to rumpuser_kill().  The purpose is to provide some hint for non-POSIX
platforms which don't sport signals and multiplex many clients in a
single address space.

While this is technically an ABI change, see above.  Also, since
rumpuser_kill() was not documented, I'll just claim I'm fixing a bug in
the implementation (and will be sure to document the interface correctly
very soon).

(pooka)

Casting to void is the canonical way of marking a variable as
potentially unused.

(joerg)

fix typo.

(matt)

Fix dmaranges for BCM563XX
Conditionalize call to arml2cc_init

(matt)

Fix coherent range for BCM563XX

(matt)

arml2cc -> needs-flag

(matt)

#include <sys/param.h>

(matt)

Conditionalize call to arml2cc_init

(matt)

Add needs-flag to arml2cc

(matt)

Explicitly include x86/fpu.h instead of relying in x86/frame.h including it.

(dsl)

Add explicit #include <x86/fpu.h> instead of relying on pcb.h including it.

(dsl)

Don't rely on pcb.h including x86/include/sysarch.h

(dsl)

Remove some pointless inclusions os sys/user.h

(dsl)

Userspace shouldn't (and doesn't seem to) need machine/pcb.h.
Unfortunately the file is included by loads of stuff, mostly in
src/external, so can't be killed outright.

(dsl)