XU4 I2C snapshot

This is a complete, but untested, driver; except that it needs to be able
to configure gpios and, afaict, we don't yet have an fdtbus_gpio_* function
that will do that.

(marty)

XU4 interrupt combiner

Written but untested.  I'm still confused about how to handle two things:

1) at interrupt disestablishment, where do I get an interrupt number so
  that I can disable the interrupt on the combiner?

2) How is interrupt multiplexing handled?  I don't seem to have any sort of
  interrupt dispatch routine that takes the 1 interrupt that reaches the
  gic and turn it into one of eight combined interrupts to call the
  established interrupt for the original uncombined interrupt source.

(marty)

Fix output in an error case

(martin)

explicit casts to time_t, since we are going from unsigned -> signed.

(christos)

little size_t/ssize_t...

(christos)

add fenv.h

(christos)

- Add DDR4E, LPDDR3 and LPDDR4.
- Check NVDIMM hybrid.
- add DDR4's tWR and tWTR.
- Print cycle time with aprint_verbose on DDR4 as DDR[23].

(msaitoh)

add m68k, default types.

(christos)

fenv for m68k

(christos)

fenv for m68k

(christos)

add fenv for m68k

(christos)

use "watchdog" clock if not defined in fdt

(jmcneill)

rearrange...

(christos)

fix amd64/conf/ALL build failure

(knakahara)

fix the following softint parallel operation problem.

(0) softint handler "handler A" is established
(1) CPU#X does softint_schedule() for "handler A"
    - the softhand_t is set SOFTINT_PENDING flag
    - the softhand_t is NOT set SOFTINT_ACTIVE flag yet
(2) CPU#X begins other H/W interrupt processing
(3) CPU#Y does softint_disestablish() for "handler A"
    - waits until softhand_t's SOFTINT_ACTIVE of all CPUs is clear
    - the softhand_t is set not SOFTINT_ACTIVE but SOFTINT_PENDING,
      so CPU#Y does not wait
    - unset the function of "handler A"
(4) CPU#X does softint_execute()
    - the function of "handler A" is already clear, so panic

(knakahara)

fix whitespace

(christos)

fix comment, from kre

(christos)

XU4 GPIO Review Changes

Implement suggestions from Jared.

(marty)

Mark as using c++ using LIBISCXX instead of manual workaround.

The workaround was confused when MKLLVM was set, but HAVE_LLVM was not.

Solution suggested by joerg.

(wiz)

We need to check if the inode is initialized for ffsv2 when we translate a
filehandle to a vnode. This can come from nfs and it could be out of range.
In that case we read garbage from the disk, end up trying to free bogus data
when we put the vnode back and we crash.
XXX: pullup-7

(christos)

Use the original nfsd.

(christos)

Rump-ify.

(christos)

ggr, make debug static. I hate that the copy of this and the original
source has diverged so much!

(christos)

add get_net.c

(christos)

Split out get_net function so that it is easily testable. Restore
functionality where 1.2.3/24 -> 1.2.3.0/24 not 1.2.0.3/24 now that
getaddrinfo(3) parses old-style incomplete dotted quads as old
style addresses.

(christos)

sc_mtx is only ever taken in a usb task, i.e. thread context so IPL_NONE
is fine.

(skrll)

Implement OF_finddevice, now ofctl -p <path to node> works

(jmcneill)

get rid of HDAUDIO_FLAG_NO_STREAM_RESET hack

(jmcneill)

HDAUDIO_FLAG_NO_STREAM_RESET hack no longer required

(jmcneill)

fix divider calculations for hdmi, and treat clock ID 211 as pll_p_out0 instead of directly pll_p

(jmcneill)

explicitly set high cursor/winbuf address bits when setting modes

(jmcneill)

fdtbus depends on clk

(jmcneill)

David Binderman in PR 50578: fix resource leak in an error branch.

(martin)

Hardcode some dates to make this more MKREPRO safe.

(wiz)

Group flags together.  No functional change.

(skrll)

New files, in preparation for modularized version of raidframe driver.

(Driver changes to be committed on the weekend.)

(pgoyette)

Revert previous... I got ahead of myself

(skrll)

Transfer are only ever allocated from thread context now.  Change IPL_USB
to IPL_NONE in pool_cache_init for the xfer struct pool.

(skrll)

remove some stuff we don't have; rename rc command variable.

(christos)

Revert previous

(pgoyette)

If we attempt to autoload a driver module, make sure we return an error
if it fails.  Otherwise we might end up calling a builtin-but-disabled
driver module and that can generate all sorts of issues...

(pgoyette)

print the tag symbolically if we don't support it.

(christos)

OF_getprop -> of_getprop_bool

(jmcneill)

build fix

(jmcneill)

In usbd_transfer deal with errors from the HCD transfer (and start)
methods properly by removing the failed xfer from the pipe queue.

For example, an Apple keyboard and FS hub with attached ums(4) on
dwctwo(4) (rpo0/rpi model a) needs this fix otherwise probing
usbd_new_device will never complete after first failed attempt to get
the initial device descriptor.

(skrll)

Add fdtbus_gpio_{read,write}_raw, which tells the controller not to take
polarity into account. Tegra GPIO pin data includes pin polarity, but so
does a regulator-fixed node, so the end result was that the enable value
was being swapped twice. Change fregulator to use the raw APIs, and adapt
Tegra and Exynos GPIO drivers to support this flag.

(jmcneill)

tegra_car_* and tegra_i2c_dvc_write are no more

(jmcneill)

Switch Tegra over to fdt based clocks and reset controls.

(jmcneill)

Add support for fdt clock and reset controllers.

(jmcneill)

For reproducible builds, allow the source of the map to be set from the
environment, so that DW_at_producer is the same no matter what the actual
underlying source and destination directories are. See:
    https://gcc.gnu.org/ml/gcc-patches/2015-12/msg01168.html

(christos)

For reproducible builds, allow the source of the map to be set from the
environment, so that DW_at_producer is the same no matter what the actual
underlying source and destination directories are. See:
    https://gcc.gnu.org/ml/gcc-patches/2015-12/msg01168.html

(christos)

Add a prefix translation map for debugging symbols. Note the use of variable
indirection so that DW_at_producer that contains the compilation command
remains the same: readelf --debug-dump foo.debug

(christos)

Improve debug/comments (mostly from nhusb)

(skrll)

Stop the callout if the transfer didn't get enqueued

(skrll)

Report an error when return -ENODEV for the unsupported setup.

(skrll)

add __TEST_FENV

(christos)

Add __TEST_FENV

(christos)

Add __TEST_FENV

(christos)

Add __HAVE_FENV

(christos)

put have fenv stuff elsewhere.

(christos)

Put have fenv elsewhere.

(christos)

Put have fenv elsewhere.

(christos)

Can't use usbd_errstr in USBHIST_LOG as vmstat(1) doesn't like it

(skrll)

there is no longer a reason to have a separate hci_send function now
that the hci_usrreq function is disassembled, so merge hci_send_pcb
back into hci_send()

(plunky)

Be graceful about command timeouts when probing.

(mlelstv)

ignore timeouts for APP_CMD prefix as for the application command itself.

(mlelstv)

Sync list of fenv enabled architectures again (PR 48633), this time for mips
addition.

(martin)

remove dup definitions from fenv.h

(christos)

XU4 GPIO Driver

This is a moderately tested working gpio driver for the Exynos based ODROID
XU4.  To use this you have to edit the dtd file exynos54422-pinctrl.dtsi
and change the two occurances of 'gpz' to 'gpz0'.  Otherewise it will crash
on a lookup failure.

It certainly could use a code review.

(marty)

fix potentially bug

(knakahara)

fix incorrect KASSERT

(knakahara)

Tweak return value handling

rtrequest1 ensures to return an rtentry on success.

(ozaki-r)

fix label and variable name.

(christos)

more checking

(christos)

From FreeBSD:
- Whitespace cleanup
- Pass a filename rather than fd to libarchive (should work with 2.8+)
- Accept zipfiles from stdin
- Extract common code from extract()/extract_stdout() to extract2fd() (pending)

(christos)

Add mips fenv.h (From FreeBSD)

(christos)

Add mips fenv.c (From FreeBSD)

(christos)

Add fenv.h for mips

(christos)

Add fenv.h for mips.

(christos)

switch to getline(3)

(christos)

Need to usbd_destroy_xfer the interrupt xfer that is automatically
provided using usbd_create_xfer

(skrll)

usbd_close_pipe should never be called with NULL for pipe.  Remove some
dead DIAGNOSTIC code when pipe == NULL

(skrll)

Our time.h defines clock_gettime with assembler name after including
sys/time.h. The Solaris compat code in sys/time.h wants to use it in an
inline function, but misses a prototype (and the relabeling). As stopgap
for getting consistent defines of clock_gettime duplicate the prototype
here.

(joerg)

Add .Pp for easy distinction between examples.

(ryoon)

Support for separate I/O thread adds new functions to dk_subr module.

Bump to 7.99.25

(mlelstv)

support deferral of diskstart to a separate thread.

(mlelstv)

more sanity checks

(mlelstv)

Fix locking botch

(skrll)

While PR 48633 is not properly fixed, keep the additional lists of
fenv-enabled architectures in sync.

(martin)

XU4 FDT final checkpoint

step N of N: get rid of baggage by removing whole files.  What's left is in
approximately the same shape as when the FDT update started, that is mostly
broken. What's missing is most of the 76 devices recognized in the dtd. In
other words, This is barely the start of a port.

Next up, gpio then i2c, followed either by straightening out usb or getting
the sdhc driver to work -- both probably require getting the interrupt
combiner to work first. A large chunk of work is left to do on the clocks.
I barely got them attaching to fdt and didn't do anything to take advantage
of the information in the dtd.

None of the other existing drivers, such as they are, properly request gpios,
i2c or clocks, and, of course power domains are off the table.

(marty)

Rewritten to take advantage of genfb(4).

(hkenken)

XU4 FDT Last drivers converted

There is a minimum conversion on the clock driver.

The USB driver needs reordering and is broken, but it was broken before.

Next up: tactical nuclear weapons

(marty)

Fix some more MKREPRO issues in cdrom creation. Now amd64 passes MKREPRO.

(christos)

more deterministic ffs for reproducible builds.

(christos)

XU4 FDT missed files

(marty)

XU4 FDT checkpoint

It is rather amazing that XU4 gets as far as it does, given how much of this
code simply doesn't work.  Focusing now on getting everything converted to
FDT. Next up USB and clocks.  After that nuke everything that's not needed
and start the port in earnest.

(marty)

Add timestamp support to the ffs image creation; needed for in kernel memory
images MKREPRO.

(christos)

Add MKREPRO_TIMESTAMP support to the makefs invocation.

(christos)

Suff_ClearSuffixes() needs to re-initialize suffNull,
otherwise its children retain old suffixes.
Have Suff_Init() call Suff_ClearSuffixes() for consistency.

(sjg)

Use MKREPRO_TIMESTAMP

(christos)

if we only asked for stopwords, use the original query.

(christos)

Powerpc now has fenv.h

(christos)

Powerpc fenv.c from FreeBSD

(christos)

new powerpc fenv.h

(christos)

Add fenv.h for powerpc (from FreeBSD)

(christos)

Add fenv for powerpc.

(christos)

XU4 debug cleanup

get rid of some unneeded debug chatter in machdep

(marty)

Remove this initial pass, since nothing further has been done with it for
six years.  It will soon be replaced with a working raid module.

(pgoyette)

If we process the callback immediately, don't bother adding it to the
list of future callbacks.  We've already processed the list (and removed
all the entries), and there's nothing in the future that will process
the list again.

This avoids the possibility of leaving an entry in the list that points
to an unloaded module's former address space.

(pgoyette)

give the kernel pax invocating the MKREPRO_TIMESTAMP treatment.

(christos)

PR/50573: Andreas Gustafsson: puffs can crash kernel for lack of argument
checking

(christos)

dd intel_old man page

(christos)

name the man page intel_old, so that we don't install it over the new
one depending on the build order.

(christos)

sort properly.

(christos)

Sort the subsections too... Hi Matt, you were right.

(christos)

Freeze the BUILDER and the BUILD_DATE for repeatable builds

(christos)

Obey MKREPRO_TIMESTAMP when creating sets (create all the files in the
sets with that date).

(christos)

XU4 GPIO FDT  missing file.

(marty)

XU4 GPIO FDT broken snapshot

This is broken. exynos_gpio_bank_config and the call to it are wrong, and
the acquire function doesn't work.

But I'm in over my head and I need to discuss this:

There is a problem with the dtd: it doesn't have addresses for the individual
gpios.  Do I add the addresses to it, or go back to the old version where I
have them hard coded in the driver.

There is a problem with creating the gpio device entries:  I suspect I really
need to treat the pinctrl devices as busses and create the gpios as attached
to those busses, but I'm not familiar with how to do that in NetBSD. At the
minimum, a pointer to a similar situation would give me code to follow.
This is different than the usual bus attachment in that the gpios aren't
devices in the dtd (they don't have "compatible" properties) so they don't
get an attach routine called.  An alternative to generating the bus
attachments might be to add "compatible" properties to the GPIO entries
in the dtd. so that they do get attached in the normal way.

If I'm going to modify the DTD, then it should be checked in, so a decision
on where to check them in would be nice, even if it does mean spreading them
all over because of license issues.  (This DTD is GPL v2)

(marty)

Add adm1026hm (MI).

(jdc)

Remove references to drivers (adt7463c, adt7467c, adm1030c) that were
superceded by dbcool.

(jdc)

strsep must be undefined first, certain glibc magic breaks a prototype
otherwise.

(joerg)

Remove trailing whitespace.

(wiz)

Use An in AUTHORS section.

(wiz)

Add the timestamp option to the pax front end.

(christos)

PR/50119: Thomas Klausner: Add --timestamp option to tar.

(christos)

Fix tools build.

(joerg)

C&P of change in rpi.conf (maybe factor out)

By default, RPI firmware sets the max CPU frequency to 600MHz. This can be
overridden by setting arm_freq in config.txt, but the default freq at boot
is still 600MHz.

Add logic to rc.local to compare the current vs. max CPU frequency; if they
differ, set the target frequency to the maximum.

(skrll)

Tweak CPUFLAGS to cover all CPUs

(skrll)

Deal with kernel builds where virtualisation isn't available

(skrll)

Missing field (was here before my change).

(maxv)

Update RPI firmware to version after the following commit

commit 224c75602b8bae1a9e942b4f1c7ed3aa8e0f0ec8
Author: popcornmix <popcornmix@gmail.com>
Date:  Tue Dec 8 14:48:57 2015 +0000

(skrll)

Use NULL instead of 0 for (function) pointer

(skrll)

Add some debug

(skrll)

KNF

(skrll)

Use KASSERT* instead of panic wrapped in #ifdef DIAGNOSTIC / #endif

(skrll)

use __diagused

(skrll)

Provide and use a MOTG_PIPE2MPIPE helper macro

(skrll)

Whitespace.

(skrll)

Whitespace.

(skrll)

Improve debug a little.

(skrll)

Fix up bus_space_map for latest RPI firmware which now passes the FB
address in the mailbox properties as a bus address.

(skrll)

Improve misleading variable name. Related to PR 50571.

XXX: also there should be real bounds-check logic in here.
XXX: if the on-disk data structure contains rubbish this code will
XXX: leak or trample arbitrary kernel memory.

(dholland)

add missing ,

(christos)

Forgot to commit this

(christos)

PR/50571: David Binderman: src/sys/fs/udf/udf_subr.c:6465: obvious
performance tidyup

(christos)

move sortinfo after pax

(christos)

Fix pasto.

(wiz)

Fix xrefs.

(wiz)

mention sortinfo

(christos)

add sortinfo

(christos)

Sort texinfo dir file for MKREPRO

(christos)

add sortinfo

(christos)

Add sortinfo

(christos)

chars is part of the library now.

(christos)

make buffers for parsing functions const.

(christos)

Adjust to the new mdocml

(christos)

more sources for compat.

(christos)

set Tx/Rx interrupts affinity to other than CPU#0 as much as possible

(knakahara)

Add missing LLE_WUNLOCK to nd6_free

(ozaki-r)

XU4 FDT sscon

This is a mess and I don't understand why part of it works, but it passes
the basic test.

There has to be some better way to get the info that I'm grabbing from the
exynos_uarts[] array.

I'm not sure how to deal with VA != PA when the DTB contains physical rather
than virtual addresses. This is going to be an issue for all of the drivers.

(marty)

EXYNOS FDT sscom

(marty)