s/Black/Block

(christos)

Nix trailing whitespace.

(riastradh)

Nix trailing whitespace.

(riastradh)

Use sys/cpu.h so that curcpu defined in terms of curlwp->l_cpu works too.

(ad)

lookup_fastforward():

- If the root vnode of a mount is being reclaimed concurrent to a lookup,
  it's possbile to become confounded and bail out of the loop with both
  foundobj=NULL and searchdir=NULL (causing a NULL pointer deref).  If that
  happens everything should be rolled back to the start for retry.  Problem
  found and debugged by hannken@.

- If the terminal node was !VDIR then searchdir was needlessly referenced.
  No functional impact.

(ad)

Fix a comment.

(ad)

Note import of dhcpcd-9.1.2

(roy)

dhcpcd: Build privsep resource limited sandbox.

(roy)

Sync

(roy)

only install space handler and enable interrupt for EC if ACPI
claims the device is actually present, it's not enough when there
is entry for it in the ACPI tables

fixes interrupt storm triggered on Dell PowerEdge R220 by enabling
GPE interrupt on a non-enabled EC (for which _REG call later
fails), reported and fix tested by Dima Veselov:
http://mail-index.netbsd.org/netbsd-users/2020/03/02/msg024166.html

XXX pullup netbsd-9

(jdolecek)

make(1): fix performance problem in specially crafted :M modifier

This fix was previously suspected to make the vax build fail.  The next
build succeeded though, and it started 2 hours before this fix was
reverted.

(rillig)

fix reversed mv, pointed out by wiz@

(christos)

deal with blacklist -> blocklist

(christos)

language sensitivity

(christos)

Follow ipfilter -> npf changes.

(tsutsui)

Add show subcommand that displays the codec configuration in
human-readable tabulated form

(sborrill)

A revision of "The purpose of computing is insight, not numbers" by
Richard Hamming.
From The Art of Doing Science and Engineering

(sevan)

Serialize rdtsc using with lfence, mfence or cpuid to read TSC more precisely.

x86/x86/tsc.c rev. 1.67 reduced cache problem and got big improvement, but it
still has room. I measured the effect of lfence, mfence, cpuid and rdtscp.
The impact to TSC skew and/or drift is:

AMD:  mfence > rdtscp > cpuid > lfence-serialize > lfence = nomodify
Intel: lfence > rdtscp > cpuid > nomodify

So, mfence is the best on AMD and lfence is the best on Intel. If it has no
SSE2, we can use cpuid.

NOTE:
  - An AMD's document says DE_CFG_LFENCE_SERIALIZE bit can be used for
    serializing, but it's not so good.
  - On Intel i386(not amd64), it seems the improvement is very little.
  - rdtscp instruct can be used as serializing instruction + rdtsc, but
    it's not good as [lm]fence. Both Intel and AMD's document say that
    the latency of rdtscp is bigger than rdtsc, so I suspect the difference
    of the result comes from it.

(msaitoh)

KNF- wrap some long lines.

(simonb)

Finish CPU core support for Octeon Cavium CN70XX:
- decode actual CPU name
- per CPU core reset logic (partially adapted from OpenBSD)
- handle Octeon 3 ioclock rate differences to other cores (from OpenBSD)

(simonb)

mention blacklist -> blocklist rename

(christos)

missed the capitalized ones.

(christos)

Rename blacklist -> blocklist

(christos)

Count down bits of entropy, not bits of data, in x86 cpu_rng.

Fixes logic in this loop for XSTORERNG on VIA CPUs, which are deemed
to have half the entropy per bit of data as RDSEED on Intel CPUs, so
that it gathers enough entropy on the first request, not on the
second request.

(riastradh)

Use x86_read_psl/x86_disable_intr/x86_read_psl to defer interrupts.

Using x86_disable_intr/x86_enable_intr causes a bit of a snag when we
try it early at boot before we're ready to handle interrupts, because
it has the effect of enabling interrupts!

Fixes instant reset at boot on VIA CPUs.  The instant reset on boot
is new since the entropy rework, which initialized the x86 CPU RNG
earlier than before, but in principle this could also cause other
problems while not early at boot too.

XXX pullup

(riastradh)

remove error(1) comments

(christos)

new file(1)

(christos)

bump libmagic

(christos)

merge conflicts

(christos)

Remove mips32r2 error introduced in previous commit; there no MP support
at all for 32-bit MIPSNN kernels.

(simonb)

Remove old compat include of rump_syscallshotgun.h

It was separated in 2016 and is no longer needed.

(kamil)

mvcesa(4): Don't use prev msg's last block as IV for next msg in CBC.

This violates the security contract of the CBC construction, which
requires that the IV be unpredictable in advance; an adaptive adversary
can exploit this to verify plaintext guesses.

XXX Compile-tested only.

(riastradh)

tsc_get_timecount(): disable the "clock goes backwards" check on i386 for
the moment since it requires 64-bit store to be atomic because of nesting
via interrupt.

(ad)

swcrypto(4): Simplify iv generation logic with cprng_fast.

(riastradh)

qat(4): Simplify iv generation logic with cprng_fast.

(riastradh)

ubsec(4): Don't use prev msg's last block as IV for next msg in CBC.

This violates the security contract of the CBC construction, which
requires that the IV be unpredictable in advance; an adaptive adversary
can exploit this to verify plaintext guesses.

XXX Compile-tested only.

(riastradh)

padlock(4): Don't use prev msg's last block as IV for next msg in CBC.

This violates the security contract of the CBC construction, which
requires that the IV be unpredictable in advance; an adaptive adversary
can exploit this to verify plaintext guesses.

XXX Compile-tested only.

(riastradh)

glxsb(4): Don't use prev msg's last block as IV for next msg in CBC.

This violates the security contract of the CBC construction, which
requires that the IV be unpredictable in advance; an adaptive adversary
can exploit this to verify plaintext guesses.

XXX Compile-tested only.

(riastradh)

Arithmetic error in previous.

(ad)

usr.bin/make: revert performance improvement

That change might be the cause of a build failure for vax.
http://releng.netbsd.org/builds/HEAD/202006131940Z/ builds fine.
http://releng.netbsd.org/builds/HEAD/202006141020Z/ doesn't.

The build fails with:
/home/source/ab/HEAD/src/external/gpl3/gcc/dist/gcc/machmode.h:524:28:
error: 'mode_size_inline' was not declared in this scope

(rillig)

is -> are

(rkujawa)

g/c vm_page_zero_enable

(ad)

Fix header compatibility with C++

This file is OS-agnostic and cannot include OS specific headers.

The fallback for NetBSD specific symbol __BEGIN_DECLS didn't work as it was
misspelled as _BEGIN_DECLS.

(kamil)

- Fix a lock order reversal in pmap_page_protect().

- Make sure pmap is always locked when updating stats; atomics no longer
  needed to do that.

- Remove unneeded traversal of pv list in pmap_enter_pv().

- Shrink struct vm_page from 136 to 128 bytes (cache line sized) and struct
  pv_entry from 48 to 32 bytes (power of 2 sized).

- Embed a pv_entry in each vm_page.  This means PV entries don't need to
  be allocated for private anonymous memory / COW pages / most UBC mappings.
  Dynamic PV entries are then used only for stuff like shared libraries and
  shared memory.

Proposed on port-arm@.

(ad)

Remove PG_ZERO.  It worked brilliantly on x86 machines from the mid-90s but
having spent an age experimenting with it over the last 6 months on various
machines and with different use cases it's always either break-even or a
slight net loss for me.

(ad)

pool_cache:

- make all counters per-CPU and make cache layer do its work with atomic ops.
- conserve memory by caching empty groups globally.

(ad)

Another bug.  The CAS loop in pthread_cond_signal() could race against the
thread it is trying to awake.  The thread could exit the condvar and then
reinsert itself at the head of the list with a new waiter behind it.  It's
likely possible to fix this in a way that's wait-free but for now just fix
the bug.

(ad)

Don't need to ignore ESRCH from _lwp_park() any more.

(ad)

add/fix linted comments

(christos)

Fix incorrect type (found by lint), and add linted comments for the
long long -> long (uintmax_t on LP64)

(christos)

Explicitly print a fs image filename on "fs image is too big" errors.

This will help future analysis of build errors caused by bloated
embedded ramdisk images and possible but unexpected other binaries.

Proposed on port-mips@ in discusson about ews4800mips daily build errors:
https://mail-index.netbsd.org/port-mips/2020/06/14/msg000950.html
and "Yes please!" from martin@.

(tsutsui)

Fix !DIAGNOSTIC compile.

(ad)

Use static constant rather than stack memset buffer for zero fpregs.

(riastradh)

Add some more id_aa64pfr0_el1 bits.

(riastradh)

Use proper "page" alignments for R5k Page Invalidate(S) op.  PR/55139

According to NEC "User's Manual VR5000, VR1000 64-BIT MICROPROCESSOR
INSTRUCTION" (U12754EJ1V0UMJ1), R5000 Page Invalidate (S) op does
"a page invalidate by doing a burst of 128 line invalidates to
the secondary cache at the page specified by the effective address
generated by the CACHE instruction, which must be page aligned."

This description looks a bit confusing, but "page" used here
implies fixed 32 byte cacheline * 128 lines == 4096 bytes,
not our variable "PAGE_SIZE" used in VM paging ops.  Note
the current default PAGE_SIZE for MIPS3 has been changed to 8192.

While here, also define and use proper macro for the "page" and CACHEOP
arg for the R5k Page_Invalidate_S op, as the manual also describes
the cache op field 10111 as "Page Invalidate" for the secondary cache.

No visible regression on Cobalt Qube 2700 (Rm5230) through
whole installation using netbsd-9 based Cobalt RestoreCD/USB.

(tsutsui)

Revert thinko in previous.

(simonb)

Include room for the trailing NUL in the way name string.

(simonb)

inet6: Allow addresses to be marked AUTOCONF from userland

(roy)

Use 32 byte cacheline ops (not 16 byte ones) for R5000 picache.  PR/55138

Commented "I think this is bad copy&paste" from skrll@.
No visible regression on Cobalt Qube 2700 (Rm5230) through
whole installation using netbsd-9 based Cobalt RestoreCD/USB.

(tsutsui)

Fix inconsistent mips_o32, _mips_o32, and __mips_o32 macro.  PR/54216

Not sure what the original intention was, but no responce for a year,
and no visible regression on Cobalt Qube 2700 (Rm5230) through
whole installation using netbsd-9 based Cobalt RestoreCD/USB.

(tsutsui)

Support Octeon Cavium cnMIPS I, II and III cores that have various
non-standard cache configurations (in terms of following MIPS spec
for defining cache configurations).

Move (most) Octeon support into a single place.

(simonb)

Move some "case 0" statements to be first in their switch statements.

(simonb)

Adjust previous - move consolidated debug printfs _after_ core specific
overrides, not before them.

(simonb)

Make core specific overrides a bit more readable / scalable (switch
instead of if / else / ...).
Move debug printfs to after core specific overrides in case any config
is updated.

(simonb)

KNF police - long lines and comments.

(simonb)

Define Octeon Cavium cache layouts for various cnMIPS cores.

(simonb)

Retire MIPS_CP0FL_USERLOCAL and MIPS_CP0FL_HWRENA and the flawed
logic that tried to deal with a MIPS processor that supports the ULR
CP0 register.  Probe correctly and save probed info somewhere we can
actually use it.  Avoids problems where libc expects ULR set to a
value but the CPU definition in the CPU table didn't have the right
combination of magic flags and thus never set ULR in the first place.

(simonb)

Put a message in build error logs to notify what this target builds.

(tsutsui)

Disable COMPAT_13 and UFS_EXTATTR.  Saves ~52kbytes.

(tsutsui)

Sort options to match GENERIC.

(tsutsui)

Disable npf instead of obsolete ipfilter.  Saves ~82kbytes.

Maybe other installation kernels that include GENEIRIC-like ones
and use "no pseudo-device" options should be checked:
https://mail-index.netbsd.org/source-changes/2018/08/01/msg097235.html

(tsutsui)

replace EX_NOWAIT with EX_WAITOK in device attach methods.
remove checks for failures that can no longer occur.

(chs)

Generate format after setting the prec value

Otherwise we end up with setting invalid printf format: "%.-1f" (from
"if (snprintf(p, sz, "%%.%df", prec) >= (int)sz)") for prec equal to -1,
which is not canonical (rejected by gcc and clang when used explicitly).

Detected by ASan (MKSANITIZER) for "jot 8".

(kamil)

Support optional square brackets around the host name.

The brackets are required when using numeric IPv6 addresses as they
contain colons as part of their syntax.  We do not enforce that the
thing in the brackets is a numeric IPv6 address - this matches scp
syntax and behavior.

(uwe)

genfs_putpages(): when building a cluster make use of pages in the in the
existing uvm_page_array.

(ad)

If a vnode is marked with VI_EXECMAP then in all likelyhood it has pages.

(ad)

NetBSD 9.99.67 - struct lwp changed on x86

(ad)

Print a rate limited warning if the TSC timecounter goes backwards from the
viewpoint of any single LWP.

(ad)

Partial scan of components

(sevan)

Fix encblkno8 legacy support.  Add a test vector while here.

What a crock!

This is deliberately _not_ neatly abstracted because the whole
configurable `iv method' mechanism is a mistake and should never be
used for anything new.

(riastradh)

Constify.

(riastradh)

Eliminate uio indirection for cgd crypto.

We don't actually use it, and we only ever used it kludgily in the
CBC encryption direction in the past anyway.

(riastradh)

tmux-3.1b out.

(wiz)

Remove -46 from the -c usage.  While here, sync usage with man page
(s/devfile/device).

(uwe)

New xz released

(sevan)

usr.bin/make: make Str_Match faster for repeated asterisks

Conceptually related to https://en.wikipedia.org/wiki/ReDoS.

(rillig)

g/c vm.idlezero

(ad)

g/c vm_page_zero_enable

(ad)

uvm_pagerealloc(): resurrect the insertion case.

(ad)

uvm_pagerealloc(): resurrect the insertion case.

(ad)

Create a compatibility symlink panel0 -> lcdpanel0

PR port-cobalt/55009

(thorpej)

When reading in the MD MAKEDEV.conf, perform block / char major
substitutions that may be present in that file.

PR port-cobalt/55009

(thorpej)

Use more shrinked x_foo binaries.

(tsutsui)

Add comments over fpu_kern_enter/leave.

(riastradh)

Zero the fpu registers on fpu_kern_leave.

Avoid Spectre-class attacks on any values left in them.

(riastradh)

Draft opencrypto support for Allwinner Crypto Engine.

XXX Can't handle nonzero crd_skip yet.

(riastradh)

Rework sun8i crypto.

- Preallocate tasks and DMA maps together for now, for 4k transfers.
- Confine setup of the task descriptor to a single function, without
  bus_dmamap_t as an input; just use the preallocated DMA maps.
- Take the DMA map part out of sun8i_crypto_buf.
  => Not much left here, just a dmamem segment and kva mapping.

This should make it easier to use with opencrypto.

(riastradh)

Truncate hw.sun8icryptoN.rng queries to 4096 bytes.

...rather than fail entirely.

(riastradh)

Move cgd selftest from module init to cgdattach.

This defers it until considerably later at boot, after cpu_attach has
run, which will be needed in order to make AES-NI work.

(riastradh)

Specify which cgd self-test failed and dump the mismatch.

(riastradh)

Shrink AES-XTS state by 280 bytes.

(riastradh)

Convert malloc -> kmem.

(riastradh)

Remove obsolete comment -- AES block size is always 128.

(riastradh)

Tidy up includes.

(riastradh)

Print cgd self-test noise only with verbose boot.

(riastradh)

Fold `cipher prep' into `cipher' in cgd.

Simplify some logic along the way and u_int*_t -> uint*_t.

(riastradh)

Nix trailing whitespace.

(riastradh)

Nix trailing whitespace.

(riastradh)

Add -46 to usage.

(wiz)

Fix incompatible function pointer casts

(kamil)

Comment out "options PMAPCOUNTERS".  It adds much overhead and is mostly of
interest to people hacking on the VM system or pmap.  Proposed on port-arm@.

(ad)

Correct a comment (or at least comment on what we do instead of half of
what we do).

(simonb)

Move MIPSNN_CFG3_ULRI so that it doesn't appear in some random position
among the other config3 register definitions.

(simonb)

Use the correct config3 field name (ULRI) for UserLocal register is
implemented bit.

(simonb)

BAU

(sevan)

Note some hard-coded capabilties that can be probed.

XXX: Fix this and CPU table in mips/mips_machdep.c one day...

(simonb)

Support IPv6.

Use getaddrinfo(3).  Add -4 and -6 command line options.  Obey USE_INET6.

(uwe)

Add a comment to say that instruction encoding 0x7c03e83b is "rdhwr $3,$29".

(simonb)

add entry for ASIX AX99100 PCIe 4port serial card

(ryo)

regen

(ryo)

add ASIX AX99100 Multi I/O (Serial,Parallel,I2C,SPI,LocalBus,GPIO) Controller

(ryo)

nbmake bootstrap: silent configure if MAKEVERBOSE==0

Be consistent with the silencing of configure in tools/
and suppress the output in build.sh configure of nbmake

(lukem)

nbmake bootstrap: be quieter if MAKEVERBOSE==0

More accurately simulate <bsd.own.mk> and don't even print
the "compile" lines with MAKEVERBOSE=0

(lukem)

nbmake bootstrap: consistency fix in MAKEVERBOSE<2 support

(lukem)

nbmake bootstrap: if MAKEVERBOSE < 2, output similar to <bsd.own.mk>

(lukem)

tools: configure --silent if MAKEVERBOSE == 0

(lukem)

tools: if MAKEVERBOSE < 2, quieten automake builds

(lukem)

usr.bin/make: remove redundant parentheses around return

(rillig)

usr.bin/make: consistently use ++ for incrementing pointers

(rillig)

usr.bin/make: fix typo in comment

(rillig)

COMPAT_90 doesn't necessarily imply COMPAT_50. So include compat in6_var.h in
either case.

Fixes evbarm build that starts with COMPAT_60.

(mlelstv)

Cleanup whitespace

(martin)

Initialise the mutex before we use it.

(jdc)

SCTP: Use ifp->if_mtu rather than ND_IFINFO(ifp)->linkmtu

(roy)

ndp: Sync usage and SYNPOSIS with reality

(roy)

Use more markup.

(wiz)

pmf-powerdown == false  -->  pmf-no-powerdown == true

(thorpej)

Add nd6.h compat header

(roy)

Update for proplib(3) API changes.

(thorpej)

external/gpl3/gcc: Suppress -Werror=maybe-uninitialized

Seems like false positive since the ASM_GENERATE_INTERNAL_LABEL macro
stores the value into prev_label, so it is alright for prev_label to
be uninitialized.

Error was reported when build.sh was run with MKSANITIZER=yes flag.

Reviewed by: kamil@

(fox)

Revert accidental part of prior

(roy)

Remove unused OPENBSD_RAW_PART. (copied from arc?)

(tsutsui)

9.99.66 welcomes the removal of in-kernel RA handling

(roy)

Note removal of RA handling.

(roy)

Remove in-kernel handling of Router Advertisements

This is much better handled by a user-land tool.
Proposed on tech-net here:
https://mail-index.netbsd.org/tech-net/2020/04/22/msg007766.html

Note that the ioctl SIOCGIFINFO_IN6 no longer sets flags. That now
needs to be done using the pre-existing SIOCSIFINFO_FLAGS ioctl.

Compat is fully provided where it makes sense, but trying to turn on
RA handling will obviously throw an error as it no longer exists.

Note that if you use IPv6 temporary addresses, this now needs to be
turned on in dhcpcd.conf(5) rather than in sysctl.conf(5).

(roy)

ipfilter: Prepare for life without in kernel RA

(roy)

-  Remove extra kpreempt_disable() -> kpreempt_enable() because
  ixgbe_handle_msf() was changed from softint to workqueue.
- Set schedule_wqs_ok before enabling interrupt to prevent the race.
- Fix comment.

(msaitoh)

more proplib API catchup

(macallan)

Update for proplib(3) API changes.

(thorpej)

i2c-indirect-config == false  -->  i2c-no-indirect-config == true

(thorpej)

lib/libpam: Fix the possible -Werror=stringop-truncation

Replace strncpy(3) with the safer strlcpy(3) and adjust the code.

Error was reported when build.sh was run with MKLIBCSANITIZER=yes flag.

Reviewed by: kamil@, christos@

(fox)

Change previous to only apply when building with clang.

(thorpej)

QOI

(sevan)

Counter tweaks:

- Don't need to count anonpages+filepages any more; clean+unknown+dirty for
  each kind of page can be summed to get the totals.

- Track the number of free pages with a counter so that it's one less thing
  for the allocator to do, which opens up further options there.

- Remove cpu_count_sync_one(). It has no users and doesn't save a whole lot.
  For the cheap option, give cpu_count_sync() a boolean parameter indicating
  that a cached value is okay, and rate limit the updates for cached values
  to hz.

(ad)

Unbreak clang builds by removing questionable linker warning sections
trggered all over the place.

(joerg)

Counter tweaks:

- Don't need to count anonpages+filepages any more; clean+unknown+dirty for
  each kind of page can be summed to get the totals.

- Track the number of free pages with a counter so that it's one less thing
  for the allocator to do, which opens up further options there.

- Remove cpu_count_sync_one().  It has no users and doesn't save a whole lot.
  For the cheap option, give cpu_count_sync() a boolean parameter indicating
  that a cached value is okay, and rate limit the updates for cached values
  to hz.

(ad)