npfctl_bpf_cidr: another buf fix in handling IPv6 masks (bug found on ARM).

(rmind)

Implement DTrace FBT provider for ARM

- Finding probable functions
- Replacing instructions
- Emulating instructions

It is tested only on ARMv7 CPUs yet, for example,
-m evbarm (-a earm) kernel=BEAGLEBONE.

(ozaki-r)

Enable invop handler on module load

It sets a fbt handler to dtrace_invop_jump_addr on module load,
which is used to call the dtrace probing function from the trap
handler.

(ozaki-r)

Make arm/dtrace_isa.c buildable

This is a band-aid for build, but it works anyway.
We have to do a real fix in the future.

(ozaki-r)

Add a trap handler for DTrace FBT provider

(ozaki-r)

Add breakpoint definitions for DTrace

(ozaki-r)

Deal with the issue that VM_MAP_INTRSAFE km_suballoc returns more than asked
for so the space needs for kmemusage accounting has to needs adjusted as well.

(matt)

There is some very odd code in s_exp2.c that only works if 'double'
  values are rounded to 53 bit mantissa in teh x87 registers.
This hasn't been true since Nov 11 2013.
Forcing a store-load for 'double' in STRICT_ASSIGN() seems to fix things
  (at least enough for the few random test cases).
I suspect the code could be changed so that the number of mantissa bits
  didn't matter.

(dsl)

remove various set-but-unused variables.

(mrg)

Make sure uvm_km_suballoc returned the correctly sized submap.

(matt)

Cleanup sscom to make more SoC invariant

(matt)

#include <arm/locore.h>

(matt)

#include <sys/evcnt.h>

(matt)

Sync bus error and address error handlers with other m68k implementation.

Tested on X68030.  Needs tests on 040 and 060turbo
(though it should work as other m68k ports).

The x68k port implemented 68060 support including these vector handlers
as early as amiga back in 1996, but even after amiga's locore.s was
improved several times (updating vectors at runtime to switch handlers
per CPU types etc.), x68k's one has not been updated.
After that, atari and mac68k pulled amiga's implementation,
hp300 pulled mac68k, and then most other m68k ports pulled hp300 ones.

Probably that's the reason why only x68k had different implementations
(i.e. no reason that avoids using common handler implementation),
and now it's time to prepare common arch/m68k/m68k/busaddrerr.s.

(tsutsui)

Add crtstuff for ia64. Compile tested only

(cherry)

Tweak the argument order passed to the elf executable entry point (usually "__start"),
so as to make call to the C function ___start() easier from assembler.
We setup the memory stack frame for ___start(), before calling it.

(cherry)

Add RCSID() macro

(cherry)

add missing break

(ginsbach)

Don't SSP a file that uses alloca.

(christos)

Note import of dhcpcd-6.3.2

(roy)

Sync

(roy)

NPF: add support for "stateful-ends".

(rmind)

Remember to bump revision for the "int my time VIF_DYING" change yesterday.

(pooka)

Remove unused variables

(martin)

Remove an unused variable

(martin)

Remove an unused variable

(martin)

Fixed wrong cast and invalid array access in the calibration framework, as
documented in PR kern/45872. The AA() macro accessed sample coordinates as
long integers, whereas they are really stored as signed integers.

Fixes calibration on my Wetab device.

(khorben)

If dlopen returns NULL assume we are statically linked and return,
rather than getting a warning when dlinfo fails.

(justin)

+ rumpkern_time

(pooka)

Add rumpkern_time for nanosleep, clock_gettime, setitimer, etc.

(pooka)

regen: time/timer related syscalls for rump kernels

(pooka)

Add rump kernel syscall wrapper flag for a bunch of time-related
syscalls (nanosleep, clock_gettime, etc.).  These are mostly intended
for situations where a rump kernel runs on an OS-less host.

(pooka)

make this compile again

(christos)

Add IST_MPSAFE flag

(matt)

Support MPSAFE interrupts.

(matt)

Add a commented out IST_MPSAFE

(matt)

Warn if a cpu failed to hatch.

(matt)

More CPUCFG registers

(matt)

No reason to attach if_virt as a RUMP_COMPONENT(), simply use MODULE().

(pooka)

remove unnecessary header inclusions

(pooka)

allow VIF_DYING() implementation to protest

(pooka)

For compilers without __COUNTER__, make the ctassert name contain
__INCLUDE_LEVEL__ ## _ ## __LINE__.  It's not perfect, but at least it's
better than just __LINE__ since it avoids collisions between .c's and .h's.

Incidentally, why does half of the kernel call CTASSERT() and the
other half __CTASSERT()?

(pooka)

Help the hack verifying malloc alignment expectations and force the
word size definition on sh* to be a compile time constant. We do not
provide a single library supporting both variants, so we pick the
smaller alignment for now.

(martin)

Rename rumpcomp_user.c -> virtif_user.c
Convert to new-style RUMPCOMP_USER_SRCS in Makefile.

(pooka)

Add __diagused.

(hannken)

Add __diagused.

(hannken)

Use -f with cp, for 444 sources.

(gdt)

Use cp -pf rather than cp -p.

Otherwise, those with sources that are 444 encounter failures.

(gdt)

remove compiler-rt duplicates to fix build

(pooka)

Fix non-armv6 case.

(matt)

It's nonliteral format strings it should not warn about.

(joerg)

Add -lrumpnet_netinet6 to the rump_server invocation, to stop
rpc.rquotad from printing the error message "rpc.rquotad:
_svc_tli_create: could not open connection for udp6: Address family
not supported by protocol family".  The error message currently ends
up in /dev/null because it is printed to stderr after calling daemon()
with noclose=0, but if the daemonization is moved to a later point to
fix the race condition of PR misc/48282, it will actually appear on
stderr and break the test, which is expecting stderr to be empty.

(gson)

Do not process linkmaps if there is exactly one link map present
(i.e. there is no ld.so).  Makes statically linked rump kernels work on
glibc (at least for me).

based on some discussion with Justin Cormack

(pooka)

No need to detach (and return EBUSY) if dk_openmask is non-zero.

(hannken)

When not built as host tool, define USE_WCHAR to get locale support.

(wiz)

Older gcc does not support -Wno-format-literal

(martin)

rename component.c -> netinet6_component.c

(pooka)

rename component.c -> bridge_component.c

(pooka)

rename component.c -> shmif_component.c

(pooka)

rename component.c -> netinet_component.c

(pooka)

rename component.c -> npf_component.c

(pooka)

rename component.c -> agr_component.c

(pooka)

rename component.c -> net_component.c

(pooka)

rename component.c -> netbt_component.c

(pooka)

rename component.c -> sockin_component.c

(pooka)

rename component.c -> netmpls_component.c

(pooka)

rename component.c -> local_component.c

(pooka)

rename component.c -> sys_sunos_component.c

(pooka)

rename component.c -> sys_linux_component.c

(pooka)

rename component.c -> sys_cygwin_component.c

(pooka)

rename component.c -> tty_component.c

(pooka)

rename component.c -> bpf_component.c

(pooka)

rename component.c -> fss_component.c

(pooka)

rename component.c -> scsitest_component.c

(pooka)

rename component.c -> scsipi_component.c

(pooka)

rename component.c -> audio_component.c

(pooka)

rename component.c -> netsmb_component.c

(pooka)

rename component.c -> drvctl_component.c

(pooka)

rename component.c -> umass_component.c

(pooka)

rename component.c -> dm_component.c

(pooka)

rename component.c -> putter_component.c

(pooka)

rename component.c -> pud_component.c

(pooka)

rename component.c -> rnd_component.c

(pooka)

rename component.c -> pad_component.c

(pooka)

rename component.c -> opencrypto_component.c

(pooka)

rename component.c -> cgd_component.c

(pooka)

Kill one colon that shouldn't be.

(joerg)

rename component.c -> raidframe_component.c

(pooka)

rename component.c -> wscons_component.c

(pooka)

rename component.c -> sysmon_component.c

(pooka)

rename component.c -> zfs_component.c

(pooka)

Add clrex when we get an abort.
Simplify undefined_entry to use the undefined stack to do transfer to the
undefined handler.

(matt)

rename component.c -> syspuffs_component.c

(pooka)

rename rumpcomp_user.* -> shmif_user.*

(pooka)

rename rumpcomp_user.* -> sockin_user.*

(pooka)

rename rumpcomp_user.* -> ugenhc_user.*

(pooka)

Allow multiple "rumpcomp_user" source modules to be specified by
introducing RUMPCOMP_USER_SRCS.  Make RUMPCOMP_USER issue a deprecation
warning, but for compat make it set RUMPCOMP_USER_SRCS=rumpcomp_user.c
for now.

(pooka)

libstdc++ doesn't like me using <cstdint> without requesting C++11, so
just do so.

(joerg)

Add some states not handled before. XXX: are both states in the | cases
possible?

(christos)

Add _Unwind_GetIPInfo for libstdc++.

(joerg)

component.c -> virtif_component.c, just to decrease risk of conflicts.
No runtime change.

(pooka)

Separate common bits for all interfaces implementing virtif hypercalls
from Makefile to Makefile.virtif.  The latter is meant to be .included.

(pooka)

Pass down data_base just in case something actually tries to use it.

(joerg)

remove some dead code, avoid set but unused variables.

(mrg)

Avoid double constants out of range

(martin)

Try to avoid using float constants out of range

(martin)

XXX This is all wrong and needs new tables and constants - but at least use
constants in the VAX double range for now, so it is compilable with newer
gcc.

(martin)

Mark a variable diagnostic-only, from John D. Baker.

(martin)

Remove trailing whitespace.

(wiz)

Rather than having long complicated asm statements, use inlines for
ldrex/strex and let the compiler arrange the code how it wants.

(matt)

When sending, only load mbufs which have a length.  This is especially
useful for fragmented datagrams where the tail of the first packet is
full of nothing.

(pooka)

Fix a prototype.

(jruoho)

Add more const to make GCC happy.

(joerg)

__ppc__ is not supported by GCC, so use __powerpc__.

(joerg)

add a mask for valid capabilities

also add a comment stating why capabilities start from 0x80

(pooka)

#ifdef a variable like its use

(martin)

Add a few dwarf/elf cat pages. Sort list.

(njoly)

Restructure union_lock() to always lock before testing for dead node.
Add two little helpers to lock or unlock a node.  Use "vp" for the
union node and "lockvp" for the node to be locked.  Use ISSET() to
test flags, add assertions.

(hannken)

Restructure layer_lock() to always lock before testing for dead node.
Use ISSET() to test flags, add assertions.

(hannken)

Restructure genfs_deadlock() and genfs_lock() to always lock before
testing for dead node.  Use ISSET() to test flags, add assertions.

Save the mount for fstrans_done() before genfs_unlock() unlocks the node.

(hannken)

ifdef variable like their use

(martin)

typo

(dholland)

Turn off literal format string warnings for one file

A better fix is coming

(agc)

Add quotes.

(joerg)

Do not include compiler-rt in librump.  librump is not a toolchain support
library, and toolchain support belongs in the host libc or libgcc or
equivalent entity that is actually supposed to complement the toolchain.

Fixes librump build on (a) Linux ARM (collisions with libgcc)

(pooka)

More readable condition.

(joerg)

Add a dummy element as explicit padding for PPC32. Fix DWARF enumeration
to match the values created by GCC. Fix DWARFish -> index conversion.

(joerg)

Rename IP_PSEUDO_REG to RETURN_REG. Fix PPC value. Sort.

(joerg)

GC unused enum.

(joerg)

0 is a valid LSDA encoding and can be seen in statically linked
programs. Initialize lsdaEncoding to DW_EH_PE_omit and check for that
value to decide whether a value should be decoded.

(joerg)

switch amd64 to gcc-4.8
Regression: ctf stuff does not work properly with some c++ (gcc)

(christos)

switch amd64 to gcc-4.8

(christos)

mark a diagnostic only variable

(para)

mark a diagnostic only variable

(para)

kill _RUMPKERNEL ifdef

(pooka)

kill undesirable #ifndef _RUMPKERNEL

(pooka)

kill undesirable "#ifdef _RUMPKERNEL"

(pooka)

Reinstance cpu_counter.h so that we don't have to work around the
MD implications everywhere else.

(pooka)

deduplicate uvm_object_printit() implementation

(pooka)

Add portability definitions in case they're not already defined

XXX - these need to be consolidated in one single header

(agc)

cast arguments to the correct type for format strings

(agc)

Annotate format strings. Remove string constant in front of real format
string.

(joerg)

Mark assertion handler as dead.

(joerg)

Use a struct for pair of frame name and whether it is optional. Add some
additional entries found in a -O0 build of ATF on PowerPC.

(joerg)

Remove trailing space.

(wiz)

avoid set but unused variables.
move variables under their usage #ifdef.

(mrg)

Mark a diagnostic only variable

(martin)

Remove unused variable found by gcc 4.8

(ozaki-r)

fix libelf.so sets

(christos)

Put the syncobjs in emul.c instead of locks.c so that they are
present for both locks.c and locks_up.c alike.

(pooka)

update comments

(pooka)

add a compat prototype for rump_getversion()

(pooka)

rump_pub_getversion -> rump_getversion

(pooka)

rump_getversion() does not require a thread context, so make it a straight
call instead of being wrapped around rump_pub_getversion().

(pooka)

Move the "is arch capable of loading native kernel modules into
rump kernel" clauses from bsd.own.mk to Makefile.rump.  Also,
add a rump_nativeabi_p() call to determine if rump kernel is
compiled with native ABI support.

(pooka)

Revert conversion to __thread as breaks on archs without TLS support

(justin)

updaet a little more:
- coldfire release works, but there are no kernels yet
- ia64 port gets much further now
- i386 port runs

(mrg)

Improve support for the CP2104. This particular device has a higher maximum
baud rate and thus the SLSA_R_SET_BAUDDIV sets the wrong rate.

Changed to code to use the SLSA_R_SET_BAUDRATE command and providing a
fallback for the old SLSA_R_SET_BAUDDIV command if the device doesn't
recognize the command. Unknown yet if the CP2101 has this command or not.

(reinoud)

update a bit.  delete ports that have switched to GCC 4.8.  with this
we still need significant work on these ports:

- mips*
- powerpc
- sh*
- vax
- m68k*

the x86 platforms are probably ready to switch, but i'm not in a good
position to test them.  these platforms are probably ready to switch,
as they're still mostly in bringup mode anyway:

- powerpc64
- coldfire
- ia64

(mrg)

pass $LD and $AS to the auto-build.h generation, so its configure
gets the right answers for many things.  in particular, it enables
HAVE_AS_TLS which the sparc port needs for TARGET_TLS to work.

(mrg)

add a nice CVS marker

(mbalmer)

note that kropla.com/dialcode.htm redirects to countrycode.org

(mbalmer)

reformat comment

(mbalmer)

add UGT

(mbalmer)

fix typo in previous commit

(agc)

Always print the MBR drive serial number, even if it's 0.

(jakllsch)

Avoid unused variable warning by borrowing sa1_cache_clean_addr briefly

(skrll)

Remove unused variables

(martin)

Remove unused variable

(martin)

Code style. No functional change.

(skrll)

Fix compat libelf version numbers

(martin)

Whtiespace.

(skrll)

Don't abuse usbd_status in xhci_init

(skrll)

Comment style.

(skrll)

-mcmodel option is CPU specific not machine specific.

(nakayama)

Regenerate for sparc's want64=true change.

Native toolchain fix for PR toolchain/48630.

(nakayama)

Enable 64-bit integer handling for gas even on 32-bit host since
sparc64 kernel now uses 64-bit constant with %hh operator.

Fix for PR toolchain/48630.

(nakayama)