Three cheers for the mighty C type system that doesn't recognize the
difference between an enum and an integer.  Due to wrong parameter order,
the lock backoff routine ended up sleeping 1 million seconds instead
of 1ms.  On the plus side, we got some sort of idea how often the shmif
spinlock backoff routine is hit during a standard test run.

(pooka)

Fix for problematic paths in /etc/daily and /etc/security reported in
PR/47645.

Add a separate file which contains the paths for the pkg_admin and
pkg_info utilities. This is called /etc/pkgpath.conf (to distinguish it
from pkg.conf).

Thanks also to Edgar Fuss for the sanity check.

(agc)

Fix feclearexcept return value (should be 0).

(matt)

Use MKSOFTFLOAT = no for including arm hardfloat files.
When arm and MKSOFTFLOAT = yes, make explicit dependency to .c files
so .S won't be used.

(matt)

Add eabi to compat dirs.
Add missing compat ld.elf_so-oabi

(matt)

Use the new endian=1234 to eliminate the rescue.ad.armeb

(matt)

Collapse all the ad.mips* set lists into a single ad.mips which works for
all variations of mips{,64}e{b,l}

(matt)

Provide scalbln weak alias. Fix argument order for STRONG_ALIAS.
Expect WEAK_ALIAS and STRONG_ALIAS to exist.

(joerg)

+reference

XXX: this manpage is hideously out-of-date

(pooka)

document the hypercall interface

(pooka)

Add sample code to suppress copyright message on interactive logins.

(mbalmer)

Add an arch64 flag for use by mips set lists (and powerpc)

(matt)

Re-order .PATH so that mknative headers are picked up first. Fixes
earm* builds.

(skrll)

Now that sets.subr understands endian=1234 just use ad.arm for everything
since it can support {,e}arm{,hf}{,eb}

(matt)

include assym.h

(skrll)

Explicitly make MKSOFTFLOAT a mkvar.
Add support for endian={1234,4321}
Add EABI flag for ARM.

(matt)

Flip parameter order in the clock hypercalls to make them consistent
with everything else wrt in/out parameters.

(pooka)

Don't hardcode offset in assembler. This ensures the return value in the
context is the correct value (0) on earm*

(skrll)

put optional modules on a separate SRCS line

(pooka)

Add protected visibility marker for things that we want to keep public
for historical ABI issues, but that are not allowed to be overriden.

(joerg)

Stub out anonmmap too.

This is a clever trick to enable me to avoid having to document
that interface.  I seriously doubt it's relevant beyond a POSIX
env where it's required by the proplib client-side implementation
which expects it can do munmap() on a return value from a proplib
call.

(pooka)

weak stubs for optional hypercalls

(pooka)

make this build/work again

(pooka)

In kthread_create(), pass the priority and cpu index down to the
rump kernel hypervisor (though the current implementation doesn't take
advantage of them).

(pooka)

Translate error codes between the hypervisor and rump kernel.

(pooka)

Don't hardcode offset in assembler. This ensures the return value in the
context is the correct value (0) on earm*

(skrll)

fix sed command

(pooka)

Some systems don't provide preadv/pwritev (hi Solaris!), so implement
that with lseek + readv/writev instead (and avoid rewinding the
pointer with soon-to-be documentation).

(pooka)

zero sc_iot and sc_memt before using them
from Phileas Fogg
while there, set sc_dev properly

(macallan)

for the TI AM335x, use the control_status register in the control module
to obtain the base frequency instead of assuming it's always 24MHz.

(matt)

Add minimal definition for control module control status register to obtain
the base frequency used for the system clock.

(matt)

Use !empty(LIBGCC_MACHINE_ARCH:Mearm*)

(matt)

Make this compile for earmhf

(matt)

Include softfloat-for-gcc.h for __ARM_EABI__ to get aeabi aliases.

(matt)

ARM EABI needs left shift logical too.

(matt)

Match against LIBC_MACHINE_ARCH for MKCOMPAT

(matt)

earm -> earm*

(matt)

Make that to process PATH_LOCALE the first time _find_category is
called. Use this to implement newlocale and provide duplocale/freelocale
as well. Based on patches by Takehiko Nozaki with simplications and fix
for the init order by myself.

(joerg)

Allow use from C++.

(joerg)

Explicitly disable standard C++ include paths to prevent mixing
includes.

(joerg)

fix non-NetBSD implementation

(pooka)

It is unclear whether cin is guaranteed to buffer the last input
character of a get() for ungetch() to work. Prefer putback() to make it
work with current implementations of cin in libc++. Tracked as
http://llvm.org/bugs/show_bug.cgi?id=15867 with test case.

(joerg)

add missing demogorgon

(pooka)

comment out MEMSIZE (now that we probe memory size)
Add second sdhc controller for beaglebone black

(matt)

Map EMIF0 for TIAM335X and use to probe memory size.

(matt)

Add TI_AM335x EMIF SDRAM_CONFIG definitions

(matt)

few more conversions

(pooka)

Make hypercall calling conventions consistent: iff a hypercall can fail,
it returns an int containing the error value.

(pooka)

Update ticket 869.

(riz)

Pull up following revision(s) (requested by joerg in ticket #869):

Needed to fix build on alpha:

lib/libc/thread-stub/thread-stub.c: revision 1.25
Weak alias directly to the catchall stub and don't strong alias twice.

(riz)

Set the raidid and the softc together in both places where we initialize
raidPtr.

(christos)

Reduce hypercalls related to reading to essentially an amalgamation
of readv and preadv.  ditto for writing.  Hypercalls are so seldomly
used that it doesn't justify 3x the calls for syntactic sugar.

(pooka)

g/c unused

(pooka)

g/c unused

(pooka)

don't build in kernel build.

(christos)

don't compile the for kernel

(christos)

Make rumpuser_dl_bootstrap() optional for platforms which don't
provide dynamic linking.

(pooka)

Identityfile warnings fixes.
https://bugzilla.mindrot.org/show_bug.cgi?id=2084

(mlelstv)

remove accidentally committed bit

(pooka)

Replace the various "get info from hypervisor" interfaces with one
unified rumpuser_getparam(), and make it return a plist.  The
contents can come e.g. from the env or a config file.  Make
identifiers starting with an underscore denote system identifiers
which must be implemented by hypervisor. (yea, j/k about the plist bit)

(pooka)

The fsync hypercall is now used only with close, so get rid of it.
rumpuser_bio() should do the necessary syncing.

(pooka)

Give private hypercalls the ability to meddle with kthread contexts too.

(pooka)

gdb-7.6 out.

(wiz)

Create a kernel thread context for the bio upcalls.

(pooka)

remove routine not meant to be committed

(pooka)

Make the rump kernel upcalls which were previous available only to
the sysproxy module available for the entire hypervisor.

(pooka)

Fix PR/47123 for hp700 as well

(skrll)

Remove trailing whitespace.

(wiz)

regroup calls a bit and sprinkle comments
no functional change for a change

(pooka)

set default bio mode to threaded

(pooka)

unschedule around blocking i/o

(pooka)

Add WS{DISPLAY,KBD}_TYPE_* for epoc32.

(kiyohara)

Add info about LDOs and DCDCs in TPS65217.

(rkujawa)

RUMPUSER_OPEN_DIRECT is no longer necessary

(pooka)

changes to bio hypercalls, part 3/n:

retire the filemmap/memsync hypercalls, they're no longer used

(pooka)

add a private hypercall to map the bus into memory

(pooka)

rework bio hypercalls, part 2:

Nuke all the policy hacks (r/w, mmap, directio) from the paravirtualized
block driver and let the hypervisor decide how it wants to optimize
the I/O.  It can prepare for this based on if a file is opened with
the RUMPUSER_OPEN_BIO flag.

mmap was not faster than r/w except in a niche case (yes, it made a
good measurement), and directio was never on by default since
it was tricky at best to decide on the kernel side of things if
directio will do the right thing.

(pooka)

Remove a hack.

(skrll)

Rework how the bio hypercalls work, part 1/n:

Reduce the set of hypercalls to one: "do block i/o".  This not only
eliminates a lot of pseudo-duplicate code, it also gives the
hypervisor a lot more freedom on how to optimize the i/o.

(pooka)

Add some defines for epoc32.

(kiyohara)

libc was missing symbols for a while there. Re-run bintuils mknative
against a good libc.

(skrll)

Disable spin mutexes for now.  They need some more work in the case
where a spin mutex is used as an argument to cv_wait().  Plus, it
would be good to sprinkle some asserts to make sure that the cpu
context is not released while holding on to a spin mutex.  All in
all, should not be difficult, but needs careful testing and bravery
(the scheduler will bite your legs off).

(pooka)

Simplify. earm is dealt with elsewhere.

(skrll)

Welcome to 6.1_RC4!

(riz)

Tickets 877,878,880,881

(riz)

Tickets 878, 880, 881

(riz)

Pull up following revision(s) (requested by isaki in ticket #881):
sys/arch/x68k/dev/ite_tv.c: revision 1.16
Fix a kernel panic by printf(1).
The 2nd byte of ISO-2022-JP must be 0x21-0x7e.  Especially,
the character code less than 0x2121 causes SEGV, because a
font address points outside of the CGROM.
Reported by tsutsui@.  Should be pulled up to netbsd-6.

(riz)

Pull up following revision(s) (requested by isaki in ticket #881):
sys/arch/x68k/dev/ite_tv.c: revision 1.16
Fix a kernel panic by printf(1).
The 2nd byte of ISO-2022-JP must be 0x21-0x7e.  Especially,
the character code less than 0x2121 causes SEGV, because a
font address points outside of the CGROM.
Reported by tsutsui@.  Should be pulled up to netbsd-6.

(riz)

Pull up following revision(s) (requested by tsutsui in ticket #880):
etc/etc.hp300/MAKEDEV.conf: revision 1.14
distrib/hp300/ramdisk/Makefile: revision 1.32
Fix yet another "openpty() failed" error during sysinst.
Prepare a new entry for ramdisk that includes ipty target in
etc.hp300/MAKEDEV.conf and use it in distrib/hp300/ramdisk/Makefile.
See also PR/46812, PR/47123, and PR/47774.
We could simply add "ipty" to MAKEDEVTARGETS as miniroot/Makefile does,
but it's better to shrink ramdisk size as much as possible.
Should be pulled up to both netbsd-6 and netbsd-6-0.

(riz)

Pull up following revision(s) (requested by tsutsui in ticket #880):
etc/etc.hp300/MAKEDEV.conf: revision 1.14
distrib/hp300/ramdisk/Makefile: revision 1.32
Fix yet another "openpty() failed" error during sysinst.
Prepare a new entry for ramdisk that includes ipty target in
etc.hp300/MAKEDEV.conf and use it in distrib/hp300/ramdisk/Makefile.
See also PR/46812, PR/47123, and PR/47774.
We could simply add "ipty" to MAKEDEVTARGETS as miniroot/Makefile does,
but it's better to shrink ramdisk size as much as possible.
Should be pulled up to both netbsd-6 and netbsd-6-0.

(riz)

Pull up following revision(s) (requested by jakllsch in ticket #878):
sys/dev/scsipi/ses.c: revision 1.45
Add detach support. From martin@.
"Works for me."
Addresses PR 44283.

(riz)

Pull up following revision(s) (requested by jakllsch in ticket #878):
sys/dev/scsipi/ses.c: revision 1.45
Add detach support. From martin@.
"Works for me."
Addresses PR 44283.

(riz)

Pull up following revision(s) (requested by enami in ticket #877):
lib/libpthread/pthread_condattr.3: revision 1.9
Fix pasto.

(riz)

Ticket 869, and adjust ticket 705.

(riz)

Pull up following revision(s) (requested by tsutsui in ticket #705):

Additional pullup needed to fix ticket 705:

etc/etc.x68k/MAKEDEV.conf: revision 1.14
Add 'bpf' to all ramdisk targets (*), or floppy targets if ramdisk calls that.
Sort bpf to come directly after 'std' where relevant.
* (apart from etc.evbppc who's ramdisk target *only* builds md0,
  which just doesn't make sense).
Should allow a fighting chance for dhcpcd to get an IP address.

(riz)

Pull up following revision(s) (requested by manu in ticket #869):
lib/libpthread/pthread_rwlock.c: revision 1.33
lib/libc/include/reentrant.h: revision 1.16
lib/libpthread/pthread_cond.c: revision 1.59
lib/libpthread/pthread_misc.c: revision 1.15
lib/libc/thread-stub/thread-stub.c: revision 1.23
lib/libpthread/pthread_cancelstub.c: revision 1.38
lib/libpthread/pthread_specific.c: revision 1.26
lib/libpthread/pthread_mutex.c: revision 1.56
lib/libpthread/pthread_tsd.c: revision 1.11
lib/libpthread/Makefile: revision 1.80
lib/libpthread/pthread.c: revision 1.143
lib/libpthread/pthread_int.h: revision 1.89
- Allow libpthread to be dlopened again, by providing libc stubs to
libpthread.
- Fail if the dlopened libpthread does pthread_create(). From manu@
- Discussed at length in the mailing lists; approved by core@
- This was chosen as the least intrusive patch that will provide
  the necessary functionality.
XXX: pullup to 6

(riz)

Use macros in machine/asm.h instead of locally defined, and
introduce CASPTRA for casxa/casa.

(nakayama)

Appearances count: spell "IPL_NONE" correctly.

(pooka)

One more mutex which is used as a spin mutex and therefore needs
to be tagged as one, lest the mighty assert smites thee.

(pooka)

fix locking order mountlist_lock -> mnt_unmounting.
Set IMNT_GONE early to protect against concurrent dounmount()
and vfs_busy() before the mountpoint is removed from
mount list.

(mlelstv)

Enable fenv support for arm.

(matt)

Add a C99 <fenv.h> implementation for arm.

(matt)

Spello

(skrll)

G/C M_VCPAGELIST

(skrll)

One more missing cast.

(joerg)

Fix value ranges.

(joerg)

Drop sys/ctype_bits.h again.

(joerg)

Include at least ctype_bits.h, even when not providing the macros.

(joerg)

Add const.

(dholland)

shorten fileops test further
add new test that verifies locking of the mount list.

(mlelstv)

Fix typo.

(joerg)

Drop s_scalbnl.c (and s_scalbln.S as side effect) and provide scalbnl in
s_scalbln.c.

(joerg)

Fix a kernel panic by printf(1).
The 2nd byte of ISO-2022-JP must be 0x21-0x7e.  Especially,
the character code less than 0x2121 causes SEGV, because a
font address points outside of the CGROM.
Reported by tsutsui@.  Should be pulled up to netbsd-6.

(isaki)

If we have both wedge and partition info, use the partition info in the
wedge case too. From mlelstv.
XXX: pullup-6

(christos)

some more headers are required by the third implementation

(pooka)

don't need event.h here anymore

(pooka)

include event.h only where used

(pooka)

Change rumpuser_cv_timedwait() from absolute time to relative time.
It's then the hypervisor's problem to translate it accordingly.
Now we no longer have to worry about the kernel having to know the
hypervisor's time and vice versa.

(pooka)

Don't forget space between -mcmodel and -cxx-isystem on SPARC and
SPARC64.

(joerg)

Improve the time-related hypercalls so that's it's possible to
sleep until an absolute time on the host's monotonic clock (should
something like that be supported).

(pooka)

fix "build.sh rumptest" wrt private hypercalls

(pooka)

Bump date for previous.

(wiz)

Add EXIT STATUS section. From Ryan Kavanagh via jmc@openbsd.
Bump date. Remove unnecessary quotes.

XXX: The behaviour is now documented as-is, but it is weird.
"biff n; biff y" has an exit status of 1 because biff reports the
notification status when it was started, not after it has finished.
{Free,Open}BSD's biff behave the same, but it's unclear if it's intended
or a bug.

(wiz)

Note NetBSD/epoc32.

(kiyohara)

Add new port NetBSD/epoc32.

(kiyohara)

Add support SoC CL PS-711x.

(kiyohara)

Fix ID ARM710a for CL PS-711x.

(kiyohara)

be a bit less sloppy with errno

(pooka)

* remove the unused "opaque" parameter
* fix some typos in the previous

(pooka)

make the file monitoring hypercalls private to shmif

(pooka)

problem with ioconf.c solved, reenable

(pooka)

Build only rumpcomp_user.c as part of the rumpcomp_user rules
(otherwise ioconf.c wants to be part of .ALLSRC for whatever reason
I'm not really interested in, but which the make gurus will surely
explain to me)

(pooka)

need errno

(pooka)

Add slhci at zbus.

(rkujawa)

duh, the untypoed version doesn't work with bsd.ioconf.mk.  comment it
out while i figure it out

(pooka)

typo

(pooka)

The ioctl hypercall is used only by the usb host controller driver,
so make it exclusive to that component.  I'm happy to see it gone.

(pooka)

Fix yet another "openpty() failed" error during sysinst.

Prepare a new entry for ramdisk that includes ipty target in
etc.hp300/MAKEDEV.conf and use it in distrib/hp300/ramdisk/Makefile.
See also PR/46812, PR/47123, and PR/47774.

We could simply add "ipty" to MAKEDEVTARGETS as miniroot/Makefile does,
but it's better to shrink ramdisk size as much as possible.

Should be pulled up to both netbsd-6 and netbsd-6-0.

(tsutsui)

Don't forget to include bsd.own.mk.

(joerg)

Make all earm variants use earm.

(matt)

Add missing build glue for libc++.

(matt)

MKLIBCXX and MKLIBSTDCXX

(joerg)

Introduce MKLIBSTDCXX, defaulting to yes.

(joerg)

no need for raid.h

(christos)

put needs flag back in a bunch of drivers because we need the file for #ifdefs

(christos)

make this work on PowerMac G5
from Phileas Fogg

(macallan)

put the one out-of-order register info in its proper sequence

(jakllsch)

remove PMAC_G5 kludge - fixed this elsewhere

(macallan)

some macppc boxes have onboard devices where parts or the whole PCI interrupt
register is hardwired to zero - fix this up here

(macallan)

make this work on G5
from Phileas Fogg, interrupt fixup from me

(macallan)

remove trailing whitespace

(jakllsch)

Update to LLVM/Clang snapshot r180686. Allows building without cxxabi.h.

(joerg)

Typo

(joerg)

Optionally recurse into libc++.

(joerg)

Add build glue for libc++.

(joerg)