Fix diagnostic trap to be correct.  PSL_EE@l would be sign-extended and this
could never match the result of the andi. before it.

(matt)

Correct last, ha+l uses lis+addi, so use h+l.

(joerg)

Build 32bit constant with lis+ori, not lis+addi. The instructions
differ on the sign extension handling of the immediate.

(joerg)

Apply some errata workarounds

(skrll)

Second argument of config_rootfound is void *, so discard const
explicitly.

(joerg)

Sync

(roy)

Fix ALTERNATE_GLOBALS so it at least works on sun4u - using a scratch register is not a good idea...

(palle)

Call etherinit from ifinit1 only when it is required

This unbreaks the builds of kernels that don't build if_ethersubr.c.

(ozaki-r)

Use lower case.  No functional change.

(skrll)

fix tpyo

(macallan)

regen.

(msaitoh)

Add Intel Xeon E3-1200 v3 Host Bridge, DRAM.

(msaitoh)

pq3pci_config_addr_read is only used in a #if 0 block, so hide it under
the same condition.

(joerg)

split PRU_CONNECT function out of pr_generic() usrreq switches and put
into seaparate functions

  xxx_listen(struct socket *, struct mbuf *)

  - always KASSERT(solocked(so)) and KASSERT(nam != NULL)
  - replace calls to pr_generic() with req = PRU_CONNECT with
    pr_connect()
  - rename existin {l2cap,sco,rfcomm}_connect() to
    {l2cap,sco,rfcomm}_connect_pcb() respectively to permit
    naming consistency with other protocols functions.
  - drop struct lwp * parameter from unp_connect() and at_pcbconnect()
    and use curlwp instead where appropriate.

patch reviewed by rmind

(rtr)

Fix typo in comment

(ozaki-r)

Fixup some comments.

(skrll)

Trailing whitespace.

(skrll)

put boilerplate extraction of inpcb or in6pcb and tcpcb performed in tcp
usrreqs into a function that can be called instead of cut & pasting it
to every single usrreq function.

tcp_getpcb(struct socket *, struct inpcb **, struct in6pcb **, struct tcpcb **)

  * examines the family of the provided socket and fills in either inpcb
    or in6pcb and tcpcb.
  * if the pcb is not present for the family of the socket EINVAL is
    returned, if the family is not AF_INET{,6} EAFNOSUPPORT is returned.

signature provided by and patch reviewed by rmind

(rtr)

GC openpic_read.

(joerg)

make sure all the KMS code is only ever run when in KMS mode.  fixes
problems with the UMS (aka existing) driver type when KMS is enabled.

(mrg)

Convert cobalt to generic mips bus_dma.

(skrll)

In the background subshell that flushes partial output every few
seconds, detach from the controlling terminal.  Without this, pressing
control-T (to send SIGINFO from the terminal) would almost always result
in the sleep process printing "about N seconds left of the original 3",
which is useless information.

(apb)

Official PPC ISA documentation has no SP register, so use R1 directly.

(joerg)

Linux wq locks are taken with spin locks held, so need IPL_VM.

(riastradh)

cmpl wants four operands, so use the shorter cmplw instead.

(joerg)

Posix requires the va_size of a symlink to be pathlength for symbolic links.
This fixes yet another atf case.

(reinoud)

Fix debug printout of udf_write_metadata_partition_spacetable(), it referred
to an old variable.

(reinoud)

Access time should only be set on accessing the data or messing with the
contents, not on inode access rights and ownership changes. Should address
PR kern/49033 for UDF.

Test results now come clean for bugs related to this issue in the ATF.

(reinoud)

Mark the udf_renamerace test case (but not udf_renamerace_dirs) as an
expected failure again, now with a reference to PR kern/49046.
Since the test only fails part of the time, force failure to
avoid failure reports reports due to unexpected success.

(gson)

Add microcode file missing from update.

(rjs)

Remove very old and incorrect comment.

(skrll)

Use IFADDR_FOREACH

(ozaki-r)

Use if_free instead of free

(ozaki-r)

Update radeon microcode.

(rjs)

update for reality.

(mrg)

rs6000/sysv.h must be after netbsd.h so we use the ppc CC1_SPEC

(matt)

Fix parenthesis on __ha macro

(matt)

Add a mutex for global variables of if_ethersubr.c

To initialize the mutex, we introduce etherinit that is called from ifinit1.

(ozaki-r)

Take rdev->fence_lock around the wakeup.

(riastradh)

GCC sets up u?int_fast8_t to be int, so be consistent with it.

(joerg)

add missing KASSERT()s at the top of sockin_usrreq(), req shall not be
either of these operations.

>      KASSERT(req != PRU_BIND);
>      KASSERT(req != PRU_LISTEN);

(rtr)

SPEC_MAXOFFSET_T doesn't exist anywhere else, so don't use it here
either.

(joerg)

PR/49045: Tetsuya Isaki: remove remaining caddr_t in ptrace(2)

(christos)

Enable net.bpf.jit only if MODULAR and BPFJIT. Tweak a warning about postponed
jit activation.

(alnsn)

Use two different mutexes for tx and rx

This change splits the mutex of wm into two: one for tx and the other
for rx. By doing so, lock contentions can be reduced. We lock both for
other operations that need locking, e.g., init, stop and ioctl.

The modification doesn't change the behavior of the driver.

(ozaki-r)

radeon depends on firmload.

Noted by rjs@.

(riastradh)

disable KMS for now; it breaks UMS + (old) radeondrm.

(mrg)

callout_ack returns void, not bool.

(riastradh)

More fixes as a step towards ARM_MMU_EXTENDED on RPI.

- don't set CPU_CONTROL_SYST_ENABLE in arm11x6_setup for ARM_MMU_EXTENDED

- Use the new MMU defines for V6N

- pull in arm/cpuconf.h in vmparam s that VM_MAXUSER_ADDRESS gets set
  correctly.

(skrll)

sun4v: Update TODO list with relevant sun4v stuff

(palle)

Change procfs from hashlist to vcache.
- Key is (type, pid, fd)
- Remove argument "p" from procfs_allocvp().  It is only used
  when "type == PFSfd".  Lookup the proc with proc_find() when
  procfs_loadvnode() needs it.
- Use a vfs_vnode_iterator for procfs_revoke_vnodes().

(hannken)

sun4v: Introduce a new macro ALTERNATE_GLOBALS (matching NORMAL_GLOBALS). Two new helper macros ENABLE_INTERRUPTS and DISABLE_INTERRUPTS are also created. With these changes the GENERIC.UP kernel will reach rootmount on sun4v systems - ok martin@

(palle)

Run scripts at build time via ${HOST_SH}

(apb)

Linux work is queued in intr context, so block intrs when locking.

(Yes, this getting out of hand.)

(riastradh)

Add usr/sbin/sysinst

(martin)

Prepare for ARM_MMU_EXTENDED

(skrll)

unifdef -U__OLD_INTERRUPT_CODE

(skrll)

In rc.subr, add _have_rc_postprocessor function and use it instead of
inline tests like [ -n "${_rc_postprocessor_fd}" ]. The new function
performs a few new tests, including verifying that /etc/rc is still
running (using a new _rc_pid variable set by /etc/rc).

This is intended to deal with the case that a script run from /etc/rc
spawns a background process, then /etc/rc exits, but the background
process still has environment variables inherited from /etc/rc.

Fixes PR 46546.

(apb)

Fix illegal snprintf usage noted in PR 47976 by simplifying it away:
instead of consing up a string and printing it to stdout, just print
to stdout. Not as compositionally tidy, but much simpler and perfectly
adequate here.

(dholland)

Fix broken snprintf usage; noted in PR 47976 from Henning Petersen.
As a bonus, this removes the nonliteral format arguments.

(dholland)

Remove broken snprintf usage (noted in PR 47976) and replace with
something simpler.

(dholland)

Fix snprintf usage; from Henning Petersen in PR 47976.

(dholland)

Fix snprintf usage and, while here, don't trundle off the end of an
array. Part of PR 47976 from Henning Petersen.

(dholland)

Update the bug report URL as per PR 48712. I assume in order for this
to propagate to where it'll be seen we need to rerun mknative, which
I'm not in a position to do right now; but this way it'll update the
next time someone else does that.

(dholland)

Fix misplaced parenthesis, from Henning Petersen in PR 44804.

(dholland)

"internally consistency checks" -> "internal consistency checks"

(dholland)

Fix typo in previous. Eesh.

(dholland)

Revert ttm_bo.c 1.3 and try to fix the double-free properly.

bo->destroy doesn't ttm_tt_destroy or ttm_tt_fini anything.  Trouble
is that ttm_tt_fini expects the swap storage to still be there, and
ttm_tt_destroy -- which is what calls ttm_tt_fini by way of
ttm->func->destroy -- has already nulled it out.

(riastradh)

PR 49036: net.inet6 has not been created when the sysctl constructor
for net.inet6.multicast is run.

(joerg)

Break overlong lines.

(riastradh)

Don't forget to decrement the acquired count too.

(riastradh)

Make sure the low bits, not the high bits, of the offset are zero!

(riastradh)

Let the drm driver decide what part of what object gets mmapped.

Pass the file around too so radeon/ttm can get at it to verify access.

Add drm_gem_or_legacy_mmap_object for drivers to choose the previous
behaviour, like i915.

(riastradh)

drmkms_ttm is not split into a module (yet).

(riastradh)

Don't destroy mutexes because the module can't be unload.

(alnsn)

Add d_discard to a cdevsw I apparently missed.

(dholland)

Sort sections.

(wiz)

Add a dummy sysinst architecture for ports that don't have their own yet.

This makes the presence of sysinst in /usr/sbin MI (as intended) --
remember that one of the goals is for sysinst to be able to install
chroots, and installing a chroot is a sensible operation on all ports,
even those for which an installer per se might not make sense.

Currently this dummy architecture fails instead of doing anything, but
as they say, one thing at a time.

(dholland)

Add /usr/sbin/sysinst.

(dholland)

Don't descend into sysinst.

(dholland)

Adjust for sysinst move.

(dholland)

Descend into sysinst.

(dholland)

Update for building in usr.sbin.

(dholland)

Move sysinst sources to usr.sbin.

This commit only physically moves the sources - there are no other
changes, to maximize the probability that this will be treated as a
rename if we ever do manage to migrate away from CVS.

Moving sysinst has been discussed on and off for years and has two
goals: making it easier to work on sysinst, and also making sysinst
available on running systems for use installing chroots and VM images
and other such things. None of the latter is possible yet, but as they
say, one thing at a time.

Doing this now was approved in an impromptu fashion by mrg, riz,
riastradh, me, and groo.

(dholland)

Move CONFIG_X86 from files.i915drmkms to drm_os_netbsd.h.

(riastradh)

Comment out options and devices to sync GENERIC rev 1.90.

(tsutsui)

Shrink GENERIC kernel to fit to bootloader's limit (2MB).

Maybe now we can't call it GENERIC, but it still boots on TME.

(tsutsui)

Move sun3 bootloader address to allow loading growing GENERIC kernels.

(tsutsui)

Improved the NORMAL_GLOBALS macro in the sun4u case, so only the relevant bit (AG) in %pstate is affected - ok martin@

(palle)

npf_conn_conkey: fix a comment.

(rmind)

Avoid needless #ifdef __i386__ with judicious use of MIN.

(riastradh)

Fix __raw_i915_read64 on i386.  From rjs@.

(riastradh)

Use tvtohz instead of ad-hacking it myself, suggested by apb@.

(riastradh)

Reorganize and bump VM related constants.

- sync VM_MAXUSER_ADDRESS and VM_MAX_ADDRESS with other hp300 derived ports
  and use same definitions for USRSTACK as other m68k ports
  (no worth to have sun3/sunos compat in these days)
- remove unused KUSER_AREA
- bump MAXTSIZ, MAXDSIZ, and MAXSSIZ

These changes allow my 64MB TT030 build pkgsrc/textproc/icu
which seems to require >200MB VA space.
(note our current 040/060 pmap implementation can't handle >224MB VA size)

(tsutsui)

Sync VM_MAXUSER_ADDRESS and VM_MAX_ADDRESS with other hp300 derived ports.

There is no particular reason to have different values among m68k ports.
Briefly checked on LUNA-II.

(tsutsui)

Fix non debug build.

(rjs)

Don't use saved EREG registers because sljit 0.91 can generate
bogus code on amd64. The A and X registers are saved on the stack.

The most recent version of sljit fixes bogus code generation but
it's not backward compatible with sljit 0.91.

(alnsn)

Recover from channel halt errors by using the 3 strikes xacterr rule.

Adapted from the RaspberryPI linux driver.

This allows the rum(4) I was sent to be somewhat usable. Need to
investigate further what is causing the problem in the first place.

Might help PR/49019 (RPI: interrupt storm when url0 is up)

(skrll)

enable KMS on x86.

(mrg)

link against libdrm_radeon, from @riastradh

(mrg)

Report the same PCI domain (cf_unit) as everything else in NetBSD.

(riastradh)

Rename radeondrmkms to just plain radeon.

radeondrmkmsfb will stay as is until the name radeonfb is available
(currently taken by legacy dev/pci/radeonfb.c).

(riastradh)

Defer real radeon attachment until mountroot so we can load firmware.

(riastradh)

Fix radeon framebuffer mmapping.

The framebuffer is not a system-RAM-backed, TTM-managed object;
rather, it is a VRAM-backed object.  So we have to use its bus
address, which is a little less sketchy with bus_space_mmap anyway.

(riastradh)

Fix botched conversion from ALIGN to round_up (not DIV_ROUND_UP!).

(riastradh)

Don't ttm_tt_destroy bo->ttm if bo->destroy will destroy it too.

(riastradh)

Fill in i2c_bit_add_bus.

(riastradh)

Make Linux usecs_to_jiffies work with hz up to 2000.

(riastradh)

fix MKPAM=no build

(dholland)

Fix some dumb thinkos in radeon_gart.c.

(riastradh)

Fix little arithmetic mistakes in radeon_gart.c.

Unlikely to fix the failing ring tests on boot, though.

(riastradh)

Fix dtrace kernel module build.
Add nodiscard to cdevsw.
They are not fixed in fallocate/fdiscard changes.

(ryoon)

Another __diagused

(pgoyette)

npf_conn_conkey: adjust to return the key length and add a comment
describing the key layout.

(rmind)

npf_mk_connlist: destroy the connections on error path.

(rmind)

Mark isc __diagused in agp_i810_unbind_memory.

(riastradh)

Disable reusing graphics stolen memory in i915drmkms for now.

Sometimes the GTT is located in graphics stolen memory, and who knows
what else the BIOS might have done with it.  Upstream has a patch to
sneak around the GTT, but disabling it altogether is simpler and
probably safer.

(riastradh)

Add generic versions of machine/int_*.h for compilers providing
appropiate macros for all necessary types.

(joerg)

Bump SYMTAB_SPACE for clang build.

(joerg)

UREAD4 is unused.

(joerg)

Regen.

(msaitoh)

Add some Intel Core internal devices.

(msaitoh)

npf_tcp_inwindow: enable strict RST check by default.

(rmind)

Output CPU manufacturer, implementation and mask code.

(nakayama)

- Call wm_set_pcie_completion_timeout() on I350, I354, I210 and I211, too.
  Same as FreeBSD and OpenBSD.
- Drop PHPM_GO_LINK_D bit in WMREG_PHPM on some chips. From FreeBSD.

(msaitoh)

Use clockfreq() to print a system tick frequency.

The stick-frequency of Fire V100 is 5555555Hz and printing it as
5.555MHz instead of 5MHz is more precise.

(nakayama)

fix some case of reference to uninitialized tp->snd_fack.
This bug causes dropping FIN mistekenly.
pointed out in PR/48283 by YASUOKA Masahiko, thanks!

(ryo)

Backout using humanize_number(9) in clockfreq().
humanize_number(9) cannot handle a fraction part and doesn't
match the intention of clockfreq().

Also backout the changes caused by the fallout of clockfreq().

Fixed outputs:

-cpu0 at mainbus0: MB86904 @ 170  MHz, MB86910 or WTL1164/5 FPU
+cpu0 at mainbus0: MB86904 @ 170 MHz, MB86910 or WTL1164/5 FPU

-sbus0 at iommu0: clock = 21250  MHz
+sbus0 at iommu0: clock = 21.250 MHz

-cpu0 at mainbus0: SUNW,UltraSPARC-II @ 449  MHz, CPU id 0
+cpu0 at mainbus0: SUNW,UltraSPARC-II @ 449.971 MHz, CPU id 0

(nakayama)

Implement WSDISPLAYIO_SMODE ioctl for proper mmap and colormap handling.

- allow mmap framebuffer memories only in WSDISPLAYIO_DUMBFB
- initialize palette for ANSI text colors on back to WSDISPLAYIO_MODE_EMUL

Tested on LUNA with 4bpp framebuffer.

(tsutsui)

The lua(4) device is experimental.
Incompatible changes might be made in the future.

(apb)

Prettify intelfb/radeondrmkmsfb autoconf output.

(riastradh)

'result' -> 'error'

(maxv)

Remove ELF_ROUND and ELF_TRUNC (unused). Found by my code scanner.

(maxv)

This case is not supposed to fall through!

(riastradh)

Bump date for previous to avoid wizd(1) invocation.

(gdt)

Add PPS support to ucom(4).

This is basically cribbed from regular serial ports, and just adds
hooks to call the pps support routines.

Also, note in the ucom(4) man page that there is about 1 ms of
latency.  Discussed on tech-kern in October of 2013, with the only
concern being that someone who didn't know what they were doing might
set up a stratum 1 server, and that somehow might have worse
timekeeping than whatever else that person might have done; the man
page comment is a mitigation for this.

This patch has been live-tested in netbsd-5/i386 and netbsd-6/i386,
and has been running on machines without a USB-serial GPS device for
most of a year with no adverse consequences (very little happens if
the PPS ioctls are not invoked).

(gdt)

fix compile error.

(hkenken)

note fallocate/fdiscard

(dholland)

Kobo touch

(hkenken)

New firmware to support the model B+

commit 59c7325f2d49603faad398677f339dfd7870ecfd
Author: popcornmix <popcornmix@gmail.com>
Date:  Thu Jul 24 17:58:30 2014 +0100

(skrll)

PTE_SYNC_RANGE a newly allocated L1 page for ARM_MMU_EXTENDED

(matt)

Added ELECOM WDC-150SU2M support.

(nonaka)

Fix description of DIAGNOSTIC in x86 kernels.

(dholland)

Modify p3_get_bus_clock():
- Intel SDM says 06_17H is the same as 06_0fH. Same as OpenBSD.
- Add some Silvermont models.
- For Slivermont architecture, 0x011 is not 166.67MHz but 116.67MHz.
- Print model name not in decimal but in hexadecimal
- Cleanup code.

(msaitoh)

More update:
- Future Atom E3000, Z3000 (0x4a, 0x5a, 0x5d)
- Atom C2000 (0x4d)

(msaitoh)

Clean up error handling in rumpuser_init()

(justin)

Update table for processor families and processor number series from the
latest Intel SDM.
- Atom Z3000 (0x37)
- Core M based on Broadwell (0x3d)
- Next gen Xeon based on Haswell (0x3f)

(msaitoh)

Remove atf_tc_expect_fail() calls for udf file-system.  These tests are
currently passing.  As discussed on current-users.  Any new failures
should be reported via send-pr.

(pgoyette)

dctx->obj may still be null here!

(riastradh)

Forgot this file for radeondrmkmsfb.

(riastradh)

Declare radeon_device_private.

Not used right now, but harmless and needed by a local patch and I
accidentally committed the definition already.

(riastradh)

Hook up radeondrmkmsfb code.

Builds, not testable yet.

(riastradh)

Don't forget to set ret on the way out.

(riastradh)

PR kern/49033: try to provide better diagnostics if the udf_lookup_complex
test randomly fails.

(martin)

Add intelfb to i386/DRMKMS too.

(riastradh)

bump VCPROP_MAXCMDLINE to 1024, use this instead of MAX_BOOT_STRING for sizing boot_args, and allow for overriding VCPROP_MAXCMDLINE in kernel config. Fixes cmdline parsing with newer firmware (required for B+)

(jmcneill)

Initial commit of support for Kobo touch. (i.MX507 or i.MX508)
(http://www.kobo.com)

(hkenken)

Serial comma, more markup, remove trailing whitespace.

(wiz)

tty is an array and can't be null; thanks joerg@clang

(christos)

Sort errors. Whitespace nits.

(wiz)

Sort SEE ALSO.

(wiz)