Remove no longer needed explicit 32->64 sign extension.

This is the last 32-bit-on-disk item among those that were either
already tagged or readily discoverable.

(dholland)

Add byteswapping to the inode block-pointer accessors.

(dholland)

Use accessors for some more indirect block manipulations.

(dholland)

fix the same bug on the linux side, print the error, and return the -tive
error to mimick linux.

(christos)

Zero out the guard for bus_space_unmap before calling i915_dma_cleanup() which
calls i915_free_hws(), which then tries to unmap. Perhaps this fixes PR/50060.

(christos)

Provide and use some macros to access ehci_{xfer,pipe} from
usbd_{xfer,pipe}

No functional change

(skrll)

PR port-alpha/50325: cosmetic fix for cpu attach message
from Felix Deichmann.

(martin)

Remove the mach entry.

(maxv)

use a verb in the authors section for all entries

(mbalmer)

fox wording and typos

(mbalmer)

KNF

(skrll)

Add compile time asserts to make sure we have properly picked up types
for a 64bit cpu and 64bit openfirmware.

(martin)

Simplify, get rid of manual _LP64 define.
Do not pass a SUN4U define, as that is no longer used in the relevant headers.
Instead tell the compiler we are compiling for a v9 CPU (so it internally
defines everything we need).

(martin)

Ooops, spell the module name in full. "mqueue" vs "mq"

(pgoyette)

And yet another dependency, this time on mq

(pgoyette)

Add another dependency for nfsserver

(pgoyette)

Add an additional dependency on compat_sysv

(pgoyette)

Do not hard-code ingnoring of makeDependfile

(sjg)

Regen

(pgoyette)

Correctly mark syscalls 301, 302, 303 so that they will autoload the
compat_sysv module if not already built-in.

(pgoyette)

attaching controllers to serial ports is configured using btattach.conf
and not btdevctl.conf, so say that

(plunky)

Handle the contents of /usr/lib/dtrace so a MKDTRACE=yes release completes.

(riz)

PR/50322: Timo Buhrmester: tail -F <file> misbehaves with stdin closed
Compare fp with stdin not fileno(fp) with STDIN_FILENO, because if tail
is called with 0 closed, then we are not going to be setting event filters
for the file because we'll erroneously think it is stdin.

(christos)

new tzcode.

(christos)

Release 2015g - 2015-10-01 00:39:51 -0700

  Changes affecting code

    localtime no longer mishandles America/Anchorage after 2037.
    (Thanks to Bradley White for reporting the bug.)

    The localtime module allows the variables 'timezone', 'daylight',
    and 'altzone' to be in common storage shared with other modules,
    and declares them in case the system <time.h> does not.
    (Problems reported by Kees Dekker.)

    On platforms with tm_zone, strftime.c now assumes it is not NULL.
    This simplifies the code and is consistent with zdump.c.
    (Problem reported by Christos Zoulas.)

  Changes affecting documentation

  The tzfile man page now documents that transition times denote the
  starts (not the ends) of the corresponding time periods.
  (Ambiguity reported by Bill Seymour.)

(christos)

use c comments

(christos)

remove quick hack code to avoid shared IRQ issue.

(knakahara)

fix: "intrctl list" causes panic when the device using pci_intr_alloc() shares IRQ.

(knakahara)

Wrap/unwrap long lines

(skrll)

revert previous (turn off tcon when disabling video output).
I found it's better to do it in the hdmi driver, but the hdmi audio issue
needs to be fixed as well.

(bouyer)

PR/50312: Robert Elz: Unconditionally include <sys/exec.h> to fix compilation
when PAX_ASLR is not defined.

(christos)

Fix LLE_TRY_UPGRADE when GATEWAY

It's expected to return a value.

(ozaki-r)

note the release of NetBSD 7.0

(jnemeth)

Make ARFE understand numbers with decimal points.

(dyoung)

Don't check if an array is (not) NULL, check if the string is empty.

(joerg)

don't lie in the comment; this is not the re-entrant version

(christos)

Switch to a version that does not suffer from reporting the wrong length
if the file contains NULs. From OpenBSD.

(christos)

fix build with INGENIC_DEBUG

(macallan)

add efuse driver

(macallan)

use the MAC address passed as a property if available instead of relying on
u-boot to program it into the chip for us ( which it may not do if we're not
netbooting )

(macallan)

add a driver for the chip's EFUSE interface, use it to find the MAC address
for the onboard ethernet controller

(macallan)

add mechanism to pass a MAC address to dme

(macallan)

note Lua update

(mbalmer)

Lua is at 5.3.1 now

(mbalmer)

fix kernel module build

(mbalmer)

update after conflict resolution

(mbalmer)

update docs after conflict resolution

(mbalmer)

Resolve conflicts.

(mbalmer)

Resolve conflicts.

(mbalmer)

Fix a bug that LSC's interrupt storm occured when MSI-X is used.
It was observed only on 82575.

(msaitoh)

Fix Lua version, add lneto as responsible person, too.

(mbalmer)

Create a temporary define involving IFF_STATICARP if we have it
instead of just testing for __FreeBSD__.
No functional change.

ok: ozaki-r@

(roy)

Remove unused struct member

(skrll)

Comments

(skrll)

- ACK Accelerate Disable in the RFCTL register is not bit 13 but 12.
  No binary change because this definition have not used yet.
- Add ACK data Disable bit's definition.

(msaitoh)

PR install/50311: missing gptmbr.bin on ramdisk based installs.

(martin)

Fix the -m32 compat build for sparc64 after recent sparc/include/types.h
changes. XXX I suspect this variant has not been tested in real life.

(martin)

Use __arraycount

(skrll)

Remove rt_ifa_localrequest().
In it's place, use rtrequest1() inside rt_ifa_addlocal() and
rtdeletemsg() inside rt_ifa_remlocal().

This removes the need for INET/INET6 specific code and allows
greater control over the creation of the local address route.

(roy)

Enqueue frames to a curcpu's pktqueue

Currently RX can run on a CPU other than CPU#0, so always enqueuing
to a pktqueue of CPU#0 makes no sense. Let's use a curcpu's pktqueue,
although bridge_foward softint doesn't run in parallel without
NET_MPSAFE.

This is a temporal solution. We need a fundamental solution.

(ozaki-r)

Add tests for assigining/deleting IP addresses

The tests help to find defects related to creation/deletion
of routes/llentries and assigning/deleting of IP addresses.

(ozaki-r)

make this kind of work.

(christos)

fix formatting

(snj)

fix the remaining problem with pcinfo; the dtracetoolkit execsnoop script
works.

(christos)

Adjust for NetBSD.

(christos)

Create an llentry after fixing an interface to store

In case of RTF_LOCAL routes, we change an output interface
of a route from original one to lo0ifp. An llentry also
has to be stored to lo0ifp in such cases.

Problem reported by roy@

(ozaki-r)

provide curthread for dtrace

(christos)

Expose struct kauth_cred for the benefit of the debugger. I can't convince gcc
to produce debug info for the structure if it does not appear in more than
one source file.

(christos)

almost fixed... Need to figure out what to do with curlwp.

(christos)

Move from usbd_{alloc,free}_xfer and usbd_{alloc,free}_buffer to
usbd_{create,destroy}_xfer.  The API change will allow future changes
to HCDs to simplify the transfer resource allocation and activation.

Several devices tested including ucom, umass, smsc, uvideo, and uaudio.

(skrll)

Do not use #ifdef SUN4U when testing for cpu features of post-v8 CPUs,
it is not good for SUN4V-only kernels. Instead use __sparc_v9__ (which
is also defined by the sparc compiler when called with cpu=ultrasparc).
ok: mrg@

(martin)

don't check for NULL before free() (Tilman Sauerbeck)

(christos)

Split auxio at ebus and auxio at sbus properly, so we can build kernels
without sbus support.

(martin)

CID/1325751: Avoid possible 32 bit overflow.

(christos)

CID/1325753: remove extra return.

(christos)

support hiding command timeout messages with a new command flag and use this
when probing for cards. Should fix PR 50302.

(mlelstv)

print wedge announcement in one line instead of two

(jmcneill)

Update comments to match previous change (avoid panic in SIOCGNATL)

(prlw1)

quick hack for shared IRQ issue.

(knakahara)

Remove dubious vhold()/holdrele() from write_logvol_dscr strategies.

The udf_node already gets locked with UDF_LOCK_NODE() and udf_reclaim()
will wait for the lock to clear.

(hannken)

Make this build for me.

(riz)

Start matching some of the field names to NetBSD's
XXX: needs work

(christos)

Arrange to install the dtrace files (some are broken, sets will need to be
fixed)

(christos)

load the kernel for NetBSD (for now)

(christos)

Add definitions for uid_t and gid_t for NetBSD because we are screwing up
with the names in the kernel and the ctf sections end up having only __uid_t
and __gid_t.

(christos)

Previous code was mis-indented but correct; simplify to make the intent clear.

(christos)

When disabling video ouput also turn off the tcon. This will cause the
HDMI output to stop too, and the monitor will go to sleep.

(bouyer)

Catch up with switch to use Xorg server.

(rjs)

Fix arplookup logic

It should first lookup and then create an entry if not found (and if
creation is requested).

(ozaki-r)

Use satosin6 instead of its own macro

(ozaki-r)

put back module filename finding code.

(christos)

Minor tweaks, from Michael McConville

(pgoyette)

Cache CPU index in the non-preemptible part otherwise it can be
unreliable (and report a CPU as patched while it was not).

(jym)

Use sysmon_taskq instead of our own workqueue for submitting button events.

(jmcneill)

kmem_free() the address returned by kmem_alloc().  found by Brainy.
use the newly aligned location if we needed it.  found by kre.

(mrg)

Bump date.

(wiz)

Adapt to usbroothub and use usbroothub defaults for some requests
rather than returning an error.

(skrll)

Fix commented out USB HC debug options.  RPI doesn't have [eo]hci.

(skrll)

Drop unused (inline) function.

(joerg)

panic takes a format string.

(joerg)

Unsigned values can not be negative.

(joerg)

Use pointer computation for references outside an object.

(joerg)

Don't check unsigned values for negativity.

(joerg)

Arrays are never NULL.

(joerg)

Avoid left-shifting negative values.

(joerg)

Panic takes a format string.

(joerg)

Catch up with the current, generated Makefile format.

(uebayasi)

changes for tnftp 20151004

(lukem)

regen for tnftp 20151004

(lukem)

Release tnftp 20151003

* Add  -x xferbufsize  to set xferbuf size.
* Add Server Name Indication (SNI) support for https.
* Increase buffer limit used for response handling.

(lukem)

Merge NetBSD-20151003 import

(lukem)

Report as USB 3.0

(skrll)

From t-hash
    + Import port speed detection code from OpenBSD.
      Check port power bit instead of speed bits. It's a cool idea.

(skrll)

Fix format string.

(joerg)

Remove dubious vhold()/holdrele() from lfs_reserve().
The vnodes are always referenced on entry.

If we changed ulfs_remove() and ulfs_rmdir() to return the locked dvp
the vnodes were always locked on entry.

Remove an outdated comment from lfs_reserveavail(), unlocking/relocking
the vnode was removed in rev 1.49.

(hannken)

The per-inode state 'id_entryno' is used by pass1 for a block count,
so widen it to 'long long'. pass2 uses it for the number of entries in
a directory (IIUC) which does not need to be wider than int, but for
now let's not try to split into two fields. FUTURE...

(dholland)

Use daddr_t instead of uint32_t for the in-memory copy of the inode
block pointers.

(dholland)

Drop an explicit sign-extension in fsck that shouldn't be needed any
more.

(dholland)

Add lfs_checkword type for reading checksum data out of structures.
This is always uint32_t, but having a name for it both makes things
clearer and avoids confusion about whether it should be 32 or 64 bit.

Note: deployed in only one place (that was erroneously tagged
ondisk32) so far.

(dholland)

Fix hardwired 32-bit stuff in fsck:
  - compute the maximum file size using LFS_BLKPTRSIZE()
  - use the new IINFO in pass 6 instead of uint32_t pointers
  - use accessors to read and write indirect blocks

(dholland)

Remove a dead data structure.

(dholland)

Use accessors for another batch of indirect block accesses, this time
in dumplfs.

(dholland)

Apply IINFO in the other userland tools.

(dholland)

Use IINFO in the cleaner.

(dholland)

Use IINFO in lfs_writeinode().
(both the kernel and the userland copies)

(dholland)

Use the new IINFO in the rfw code, eliminating hardwired 32-bit values.

(dholland)

Add an IINFO struct, which is like the FINFO struct but for the inode
blocks portion of the segment summary.

A segment summary block begins with a header (SEGSUM); the rest of the
block contains FINFO structures describing file blocks growing upward
from the bottom (after the header), and IINFO structures describing
inode blocks grown downward from the end of the block. (When they meet
the segment is full regardless of how many blocks might be left.)

IINFO contains just a block number, and until now this information was
handled by just using uint32_t*; switching to a structure will make
the code a lot easier to read, and also make it easier to have 32-bit
and 64-bit versions without making a mess.

This commit just adds the structures and accessors; they'll be
deployed into the code in subsequent commits.

(dholland)

Fix up setlists for new debug compat libs.

(riz)

Restore some defines lost in last update, and don't include sys/cpuvar.h

MKDTRACE=yes now builds for arm platforms again.

(riz)

Update signature of unimplemented function to match prototype.

(riz)

Make usage static and dead.

(joerg)

Support per-target AFLAGS.

(joerg)

PR/50295: Fix clock_nanosleep return value documentation.

(christos)

PR/50295: clock_nanotime() should not set errno, but return the error.

(christos)

put back the DTRACE macros

(christos)

we don't need any of this now.

(christos)

Merge conflicts

(christos)

Adapt to NetBSD modules and link set access.
XXX: We don't have an event system so that we can issue callback that
    will be run on construction or destruction to enable and disable probes
XXX: We don't have a way to use link_sets across modules properly.

(christos)

Change SDT (Statically Defined Tracing) probes to use link sets so that it
is easier to add probes. (From FreeBSD)

(christos)

Based on suggestion from jared, redo previous using a specific print callback
to avoid the extra line. Now output looks like:
gpio0 at awingpio0 port B: 18 pins
gpio1 at awingpio0 port C: 25 pins
gpio2 at awingpio0 port D: 28 pins
gpio3 at awingpio0 port E: 12 pins
gpio4 at awingpio0 port G: 12 pins
gpio5 at awingpio0 port H: 28 pins
gpio6 at awingpio0 port I: 22 pins

(bouyer)

Print the port name when gpio(4) attaches, makes it much easier to find
what port is connected to what gpio instance:
gpio0 at awingpio0: 18 pins
gpio0: port B
gpio1 at awingpio0: 25 pins
gpio1: port C
gpio2 at awingpio0: 28 pins
gpio2: port D
gpio3 at awingpio0: 12 pins
gpio3: port E
gpio4 at awingpio0: 12 pins
gpio4: port G
gpio5 at awingpio0: 28 pins
gpio5: port H
gpio6 at awingpio0: 22 pins
gpio6: port I

(bouyer)

Fix comiple error. plum(4) is not PCI...

(msaitoh)

Fix case.

(mbalmer)

Fix a bug that the TPH ST Table is wrongly decoded. Found by llvm.

(msaitoh)

x86: Add PCI Extended Configuration Space support.
pci(3): Decode Extended Capability in PCI Extended Configuration Space.

(msaitoh)

PCI Extended Configuration stuff written by nonaka@:
- Add PCI Extended Configuration Space support into x86.
- Check register offset of pci_conf_read() in MD part. It returns (pcireg_t)-1
  if it isn't accessible.
- Decode Extended Capability in PCI Extended Configuration Space.
  Currently the following extended capabilities are decoded:
  - Advanced Error Reporting
  - Virtual Channel
  - Device Serial Number
  - Power Budgeting
  - Root Complex Link Declaration
  - Root Complex Event Collector Association
  - Access Control Services
  - Alternative Routing-ID Interpretation
  - Address Translation Services
  - Single Root IO Virtualization
  - Page Request
  - TPH Requester
  - Latency Tolerance Reporting
  - Secondary PCI Express
  - Process Address Space ID
  - LN Requester
  - L1 PM Substates
  The following extended capabilities are not decoded yet:
  - Root Complex Internal Link Control
  - Multi-Function Virtual Channel
  - RCRB Header
  - Vendor Unique
  - Configuration Access Correction
  - Multiple Root IO Virtualization
  - Multicast
  - Resizable BAR
  - Dynamic Power Allocation
  - Protocol Multiplexing
  - Downstream Port Containment
  - Precision Time Management
  - M-PCIe
  - Function Reading Status Queueing
  - Readiness Time Reporting
  - Designated Vendor-Specific

(msaitoh)

Fix typo

(ozaki-r)

List source name, not object name to unbreak AFLAGS.*.

(joerg)

Redo previous so as to not break non-arm MKDTRACE=yes builds.

(riz)

Only declare 'ldn' where it's actually used.  Fixes build on arm
with MKDTRACE=yes.

(riz)

Build Xorg server and wsfb driver for mac68k.

(rjs)

Add proc:::exit
XXX: I am going to kill that soon and make the providers register automagically
with linker sets. This can't scale.

(christos)

recognize the proc provider as belonging to sdt

(christos)

Reduce MAXDSIZ from 416MB back to 224MB.
Due to limitations by the current pmap implementation our virtual address
space is limited to 512MB. A MAXDSIZ of 416MB would leave only a little bit
more than 30MB for shared objects and heap allocations, so 224MB seems like
a good compromise.

(phx)

don't  MIPS_PHYS_TO_KSEG1() the hardware address for O2's serial ports
now serial console works again

(macallan)

pass the port's hardware address in COM_INIT_REGS() so the com driver can
figure out wether this is the console or not

(macallan)

Disable CFI in the GENERIC kernel again, as currently only cmdset 1.3
is supported, which is not present in all flash chips.
Add lmtemp(4), an LM75 temperature sensor, which is found in some
Synology NAS models.

(phx)

Now more than a single I2C device per model may be configured directly.
An additional flags argument makes the configuration depend on the model-
flags in an optional BTINFO_MODEL bootinfo node.
Print the model name together with the vendor name, when known.

(phx)

Pass a name and flags precisely describing the current model and its
features via a new bootinfo node (currently Synology only). This allows
for example the configuration of model-specific temperature I2C sensors
and fan control.
Support for the Synology DS207 and DS209 drive LEDs and wait until the
2nd drive is completely powered up.

(phx)

fix IVARs offset for 82580 and newer.

(knakahara)

Make GATEWAY (fastforward) work again

With GATEWAY (fastforward), the whole forwarding processing runs in
hardware interrupt context. So we cannot use rwlock for lltable and
llentry in that case.

This change replaces rwlock with mutex(IPL_NET) for lltable and llentry
when GATEWAY is enabled. We need to tweak locking only around rtree
in lltable_free. Other than that, what we need to do is to change macros
for locks.

I hope fastforward runs in softint some day in the future...

(ozaki-r)

Remove extra opt_gateway.h

(ozaki-r)

Add delay to improve stability of test

(ozaki-r)

-  Add workaround for 82574 Errata 25 and 82583 Errata 12 "Dropped RX packets"
  and for 82573 (unknown). Set GCR_L1_ACT_WITHOUT_L0S_RX bit. The NVM Image
  version 2.1.4 and newer have this workaround.
- Print the NVM image version on 82583, too.

Pointed by joerg@.

(msaitoh)

Add lockdebug_barrier

ok pooka@

(ozaki-r)

Remove redundant UNLOCKED and LOCKED

UNLOCKED and LOCKED are done inside mutex_exit and mutex_enter respectively
so we don't need to do them outside mutex_exit and mutex_enter.

Reviewed by pooka@

(ozaki-r)

We switched to newer versions in elftoolchain a while ago.

(christos)

do the stack protector stuff like the other files.

(christos)