s/dirname/dirpath/ to avoid potentially shadowing dirname(3).

(mrg)

note rtwn(4).

(nonaka)

Added rtwn(4) for Realtek RTL8188CE/RTL8192CE PCIe 802.11b/g/n wireless network
devices.  Ported from OpenBSD.

(nonaka)

Don't make fifofs test depend on rump.

(rjs)

update c++/volatile/constexpr entry

(pooka)

Fix PTHREAD_FOO_INITIALIZER for C++ by not using volatile in the relevant
pthread types in C++ builds, attempt 2.

The problem with attempt 1 was making assumptions of what the MD
__cpu_simple_lock_t (declared volatile) looks like.  To get a same type
except non-volatile, we change the MD type to __cpu_simple_lock_nv_t
and typedef __cpu_simple_lock_t as a volatile __cpu_simple_lock_nv_t.
IMO, __cpu_simple_lock_t should not be volatile at all, but changing it
now is too risky.

Fixes at least Rumprun w/ gcc 5.1/5.2.  Furthermore, the mpd application
(and possibly others) will no longer require NetBSD-specific patches.

Tested: build.sh for i386, Rumprun for x86_64 w/ gcc 5.2.

Based on the patch from Christos in lib/49989.

(pooka)

+ copy symtab

(uebayasi)

Added some missing SPI device names.

(phx)

regen

(nonaka)

Added Realtek RTL8192CE.

(nonaka)

PR/50179: Timo Buhrmester: sh(1) variable expansion bug

(christos)

netbsd depends on $S/kern/kern_ksyms_buf.c; rarely updated but better safe.

(uebayasi)

Make dksubr use a spin-mutex again, since some drivers still call dk_done
from hardware interrupt. Instead, release mutex while calling start routine.

The buffer peek/use/get sequence which can no longer be atomic. So consume
the buffer directly and on error privately save and retry the buffer later.
The dk_drain function is used to flush such a deferred buffer together with
the buffer queue.
Adjust drivers to use dk_drain.

Fix an error path where dk_done was called while the lock was already held.

(mlelstv)

Tweak previous.

(uebayasi)

Echo exact objects (swap*.o etc.) when linking kernel.

(uebayasi)

Constify a little.

(uebayasi)

regen.

(msaitoh)

Add Intel PCIe NVMe SSD.

(msaitoh)

Mark the ssh test as expected failure pointing to PR lib/50174

(martin)

One more about linker script, another about genassym.

(uebayasi)

Use getfsspecname also when not found in /etc/fstab.

(mlelstv)

Have MI genassym.cf.

(uebayasi)

Don't allow MD to override the assym.h generation rule.

(Nothing defines it except i386/bioscall/Makefile, but it is not part of
kernel build.)

(uebayasi)

CID 1320410: Don't call free on uninitalized pointer.

(christos)

CID 1298707:  Change confusing code that makes coverity think that Length
is uninitalized (we want length to be zero so instead of setting it to 0,
we check if it not zero and then we set it if it is not?!?!)

(christos)

CID 1292510: Handle allocation failure.

(christos)

Commit stragglers.  Renew $ARFE$ in dt.c.

(dyoung)

initialize ncpuonline

(pooka)

remove mksysctls(), now provided by init_sysctl_base

(pooka)

Move a bunch of sysctl nodes from init_sysctl (kitchen sink sysctl file)
to init_sysctl_base (only base kernel defs).  Main motivation was to
fix sysconf(_SC_NPROCESSORS) for Rumprun.  As reported by neeraj on irc,
it returned -1 before this fix, so we were doing imaginary computing.

(pooka)

add ncpuonline

(pooka)

add cpu_getmodel()

(pooka)

hw nodes do not go under the kern tree, so add new init routine for them

(pooka)

Remember that dlsym() tends to fail on PowerPC during init (or at least
tended), so call rumphijack_dlsym() instead to be safe.

(pooka)

Revert 1.14 now that the arduous task of fixing rumphijack to allow
mmap() in early init has been completed.

(pooka)

allow mmap() to be called before init runs

(pooka)

I don't like -Ttext .

(uebayasi)

Replace a constant in ldscript.

(uebayasi)

In ttywait_timo(), break out of loop on all errors, not just
EWOULDBLOCK, as ttywait() did prior to 1.265.

(gson)

Rename variable to avoid -Wshadow warnings with some compilers.

(pooka)

+ kill makeoptions

(uebayasi)

Replace a constant in ldscript.

(uebayasi)

Fix previous; do include assym.h.

(uebayasi)

Replace COHERENCY_UNIT in ldscript.

(uebayasi)

Make <kernel>.ldscript ready to include assym.h.

(uebayasi)

Don't expand unwanted symbols by cpp(1).

(uebayasi)

Redo Rev. 1.188:
o Explicitly add dependency of `<kernel>' on `<kernel>.ldscript', instead of
  relying on ${SYSTEM_DEP}, which is evaluated much earlier than
  `sys/conf/Makefile.kern.inc' is included, to avoid evaluation ordering
  problems.
o Print a message when creating a `<kernel>.ldscript'.
o Trim empty lines in `<kernel>.ldscript'.

(uebayasi)

Print a message when creating a temporary ldscript.

(uebayasi)

Give a chance to send a DAD announce packet

ifconfig -w ensures IP addresses have left tentative state, however,
it doesn't guarantee that a DAD announce packet is sent. The kernel
clears tentative flag and then sends the packet so that ifconfig -w
can return before the kernel sends the packet.

(ozaki-r)

- pass the right address to bus_space_mmap() so this works on N32 kernels
- don't mess with XMAP9 when switching to graphics mode

(macallan)

some final -Ifoo/opt removal

(pooka)

add some _KERNEL_OPT as the finishing touch

(pooka)

Remove "#ifdef INET" code.  Por que?  Because opt_inet.h was not included
and "#ifdef INET" was just a fancy way of saying "#ifndef NetBSD".

(pooka)

convert auxio(4) to use a mutex for exclusion (this existing code didn't
really work for MP systems, anyway -- it only went to splhigh.)

various cleanups:
- remove unused AUXIO_SBUS
- make most functions static
- introduce auxio_{read,write}_led() frontends
- avoid a potential NULL deref in auxio_fd_control()

the text is 32 bytes smaller.

(mrg)

opt files were nop't

(pooka)

These "wip" components have been in progress for >5 years.  There hasn't
been much progress.  Let's just kill them, they're not even build-tested.

(pooka)

don't -I imaginary opt directories

(pooka)

would you like some freshly ground _KERNEL_OPT with that?
yes?  excellent choice, sir/madam.

(pooka)

purge rump/net of component-specific opt directories

(pooka)

continue consolidating non-modular option files

(pooka)

Remove individual opt_foo.h files, continue consolidating to opt_rumpkernel.h

(pooka)

to garnish, dust with _KERNEL_OPT

(pooka)

remove librump/rumpnet/opt, consolidate in rump/include/opt

(pooka)

Remove a bunch of opt files.

It's a cute idea to have component-specific opt files, but also a
completely stupid one since there's no way of knowing how options
transcend component boundaries, and therefore if a set of options is
conflicting or not.  So, just continue concentrating all of opt_foo.h in
opt_rumpkernel.h in accordance with the monolithic opt model.

(pooka)

sprinkle _KERNEL_OPT

(pooka)

+ include opt_inet.h for INET (or lack thereof)
+ include net/in_ether.h, not netinet/in_ether.h
  (did not cause a meltdown only because opt_inet.h was missing)

(pooka)

Add hardware checksumming support.

(rjs)

Fix logic for setting multicast addresses.

Setting CRC values for hashes is broken so all multicast addresses are
accepted for now.

(rjs)

Whitespace.

(rjs)

Remove unnecessary DPSRCS+=

-imacros automatically results in a dependency.

(pooka)

printf -> aprint.

(rjs)

Fix the build: provide compatibility with ports that define KERNLDSCRIPT to
a file?

(christos)

Consistently use iswap64() and 64bit intermediate variable, as well as
64bit formats for c_tapea. Fixes tape useage report for large filesystems.

(bouyer)

Default the read block size (-k default value) to kern.maxphys (usually
64k these days). This gives a noticable performance boost on large filesystems.

(bouyer)

regen

(pooka)

<sys/mount.h> is necessary only with !RUMP_CLIENT, so include it only
in that case.  Fixes build for Hurd (which does not provide <sys/mount.h>)

from rmh@gnu.org

(pooka)

If a kernel linker script is provided, preprocess it by cpp(1).

Thus some constants (PAGE_SIZE, COHERENCY_UNIT, ...) can be replaced in linker
scripts without setting unnecessary temporary symbols, or doing ugly sed(1)
hacks.  No headers are included yet.

(uebayasi)

Fix vers.o dependency:
o Depend on ${SYSTEM_OBJ} instead of ${SYSTEM_DEP}.  The latter is obviously
  meant for kernels.
o Depend on Makefile instead of ${SYSTEM_SWAP_DEP}, which has existed since
  Rev. 1.1, but seems not used now.  Makefile is a file that must be updated
  after every config(1) run.  This is simpler than dependency on swap<kernel>.c.

(uebayasi)

Define ${LINKSCRIPT} in one place.

(uebayasi)

Add some missing obsolete markers

(martin)

Use BUILDRUMP_IMACROS if specified.

(pooka)

Change 0 to NULL for rtrequest's last argument (struct rtentry **ret_nrt)

(ozaki-r)

Use vfs_vnode_iterator for udf_do_sync.

- Build rb_tree to get an ordered list of nodes, sync them and
  clean the tree.

- Stop abusing mntvnode_lock to serialize udf_do_sync, use new
  mutex ump->sync_lock instead.

(hannken)

Use vfs_vnode_iterator for udf_unmount_sanity_check.

(hannken)

Remove dirtynodes_cv, it is only used for timed waits without any signals.

Replace the cv_timedwait with kpause.

(hannken)

Don't mention stab and DWARF sections, because these (poorly mtaintained)
lists only help to make them harder to read.

If those sections are found in inputs, they simply appear in outputs as
orphaned sections, sorted by section types and attributes.

(uebayasi)

handle blacklistd properly.

(christos)

Fix mbuf dmasyncing for extended MMU kernels.

(matt)

Add an assertion; if rtcache has an rtentry, its refcnt must be > 0

(ozaki-r)

Disable another tentative state check

It's too ephemeral to check robustly.

(ozaki-r)

Formatting and typo fixes.

(wiz)

Add man page for mcp48x1dac driver.

(rkujawa)

Quadruple GPT partition entry count limit (to 512 entries or 64KiB).

The UEFI 2.3.1 specification states that:
"A minimum of 16,384 bytes of space must be reserved for the GPT Partition Entry Array."
and [the size of a partition entry shall be a power of two greater than 128]
and that [the defined fields of a partition entry total 128 bytes].

Clamping the entries means that no partitions on the drive will be detected,
as this will result in an incorrect partition entry array CRC. This change
reduces the likelyhood of useless partitions, while still not allowing a
huge kernel memory allocation to load the partition entries into.

In the future this code should probably be reworked to checksum and evaluate
the partition array in chunks while still limiting the number of GPT
wedges added per drive to something reasonable.

(jakllsch)

if you change display types, the "Showing %s, refresh every %d seconds."
message is retained, and you can't get rid of it.  since it's dispalyed
in inverse, it feels like there's always an error happening (which is
really what the code has -- the above string is displayed via the error()
function.)  for several types of key press that largely are just "redraw"
operation (space, ^L, etc.), clear the error string as well.

(mrg)

move the obsolete list files into their respective sets.  this means
that if you install a new set and run "postinstall fix obsolete"
the expected things happen.

(hi 12.5 year old me who thought base was better than etc set, which
it really was, but this is better again :-)

(mrg)

whitespace

(jakllsch)

Support PCI_PRODUCT_INTEL_82Q45_KT_1 in addition to PCI_PRODUCT_INTEL_82Q45_KT.

(jakllsch)

pcidevs .h regen

(jakllsch)

Add another ID for Intel 82Q45 KT

(jakllsch)

swap the formats too, not just the args.

(christos)

Move rd assignment. Spotted by christos@. Thanks.

(skrll)

Do BUS_DMASYNC_POST{READ,WRITE} on transfer buffers appropriately

(skrll)

Expose usbd_xfer_isread

(skrll)

Use .word for data that is known to be aligned.

(joerg)

Remove .proc 1, it has been ignored by gas for ages.

(joerg)

.align ensures alignment, so replace .uaword with .word.

(joerg)

regen

(skrll)

PR/50160: 5 duplicate entries in dev/usb/usbdevs

(skrll)

Fix reversed arguments to a print. nice and confusing...

(dholland)

Don't need to specify OUTPUT_FORMAT/OUTPUT_ARCH.

(uebayasi)

An adaptive mutex is sufficient, the data structures are accessed
by regular threads and by the biodone softint.

This allows diskstart routines to sleep (like cgd).

(mlelstv)

another wait test.

(christos)

Sync lfs changes with the x68k version of this code.

(dholland)

TODOs about kernel options.

(uebayasi)

Unfortunately no sh3 kernels (no longer?) use linker scripts, but I dare to
clean up .ctors/.dtors here too.

(uebayasi)

.interp and friends are unlikely to appear in kernels.

(uebayasi)

.rel/.rela should not be generated in kernels.

(uebayasi)

.init/.fini/.ctors/.dtors should be irrelevant to kernels.

(uebayasi)

According to matt@, .ARM.{extab,exidx} are no longer used.

(uebayasi)

+LTO

(dholland)

959

(snj)

Pull up following revision(s) (requested by jmcneill in ticket #959):
distrib/utils/embedded/conf/armv7.conf: revision 1.8 via patch
Cubieboard2 kernel is named netbsd-CUBIEBOARD.ub not netbsd-CUBIEBOARD2.ub

(snj)

Add kernel config for Tegra K1 "Nyan Big" board, as found in the Acer
Chromebook 13 (CB5-311).

(jmcneill)

Linker script TODO items.

(uebayasi)

fix include filename.

(christos)

Update for v1.10: use '(*)' instead of '#' as dagger replacement.

dholland finds the '#' unintuitive and hard to read.

(wiz)

report the signal that wait was interrupted by, which is not always SIGINT
anymore.

(christos)

Process pending signals while waiting for a job:
    $ cat << EOF > hup.sh
    #!/bin/sh
    trap 'echo SIGHUP; exit 1' 1
    sleep 10000 &
    wait
    EOF
    $ chmod +x ./hup.sh
    $ ./hup.sh &
    $ kill -HUP %1

(christos)

No longer access the disk driver directly.
If there is an open wedge, temporarily reference its vnode.
Otherwise try to open the block device.

(mlelstv)

revert the previous

(mlelstv)

fix bogus gcc unitialized warning

(christos)

Locate in both inputs hexadecimal numbers starting 0x and make them
"wild" in the alignments dt computes.  In dt, bitwise-AND the 0x-hex
numbers.  In it, bitwise-OR them.  Take care not to match a hexadecimal
with a decimal or vice versa!

TBD: identify hexadecimals that don't start 0x.

Remove a little dead code.

Split HB_DEBUG into HB_DEBUG and HB_ASSERT.  The latter just enables the
assertions.

Update old test results for the new treatment of 0x-hexadecimal.  Add
some new tests.

(dyoung)

Don't need to specify OUTPUT_FORMAT/OUTPUT_ARCH in linker scripts.

(uebayasi)

Assert we aren't trying to sync a read-only mbuf as PREREAD.

(matt)

No need to BUS_DMASYNC_PREREAD the transmit data buffer. Fixes "NFS writes
being corrupted?" issue on ODROID-C1 as seen on port-arm.

(jmcneill)

952, 957

(snj)

Pull up following revision(s) (requested by jnemeth in ticket #957):
usr.sbin/postinstall/postinstall: revision 1.200 via patch
PR/50158 - Simon Burge -- postinstall does not know about blacklistd

(snj)

Pull up following revision(s) (requested by mbalmer in ticket #952):
share/man/man7/hier.7: revision 1.110
Lua 5.1 -> Lua 5.3
From Kamil Rytarowski.

(snj)

Cubieboard2 kernel is named netbsd-CUBIEBOARD.ub not netbsd-CUBIEBOARD2.ub

(jmcneill)

add some more RTL8211F bits

(jmcneill)

fix bogus %m
use EXIT_FAILURE

(christos)