Add accessor functions to get the number of gets and puts on pools and
pool caches.

(simonb)

Flush input to EOL correctly in games/fish.

PR 54885 from Mouse, with a somewhat different patch.

(dholland)

acpi_timer: use ACPI-Fast if WAET timer one read flag is set

The Windows ACPI Emulated Devices Table (WAET) has a hint to inform the OS
that a single read of the PM timer is reliable. Honour this flag.

(jmcneill)

don't put CPUs in separate zones, turns out there really isn't much that
keeps the airflow apart around them

(macallan)

aarch64: Remove empty source file and references to it.

(jmcneill)

Regen: Add some Compaq display controllers.

(thorpej)

Add some Compaq display controllers.

(thorpej)

build fix

(jmcneill)

s/structres/structures + s/whenver/whenever

(andvar)

Fix all remaining typos, mainly in comments but also in few definitions and log messages, reported by me in PR kern/54889.
Also fixed some additional typos in comments, found on review of same files or typos.

(andvar)

Build fix: vtophys takes vaddr_t, not a ptr

(jmcneill)

Report our interrupt trigger type.

(thorpej)

Noralize autoconfiguration messages, and report interrupt trigger type.

(thorpej)

Report which trigger type we're using for our interrupt.

(thorpej)

aarch64: Restrict KERNEL_SUFFIXES to img only

(jmcneill)

amd64: liveimage: use RC_CONF_EXTRA instead of rc.conf.d

(jmcneill)

Create temporary files in . rather than /tmp so that they get cleaned up
automatically by atf.

(gson)

Don't blindly establish our interrupt handler as IST_LEVEL.  If the INTDEF
register has the INTHIGH bit set, the controller is going to keep the
line low when *not* asserting an interrupt, and since EISA level-tiggered
interrupts are active-low, this would result in a forever-interrupt-storm.

So, if INTHIGH is set in INTDEF, establish our interrupt handler as
IST_EDGE, which will program the EISA PIC to detect the interrupt on
the rising edge of the IRQ line.

(thorpej)

Correct the comment about the INTHIGH bit in the INTDEF register.  It
does indeed mean "IRQ signal is active-high", but "else edge" is not
correct; level-triggered EISA interrupt are active-low, and edge-triggered
EISA interrupts are rising-edge, so INTHIGH would in fact mean "edge".

(thorpej)

evbarm: liveimage: use RC_CONF_EXTRA instead of rc.conf.d

(jmcneill)

bootimage: allow appending additional rc.conf options

If RC_CONF_EXTRA is defined, append the contents of this file to the
target file-system's rc.conf

(jmcneill)

Instead of getting an estimate of the number of kernel buffers,
allocating memory for them, requesting all the metadata contents of
these buffers (and repeating in the unlikely case of the number of
buffers increasing too much since the estimate) and then straight away
throwing all the contents out just to count how many buffers there were,
just get the initial estimate from the kernel and subtract the slop.

Reduces system CPU usage of "systat vm" by approx 80% for any system
with a reasonable number of buffers.

(simonb)

Expose KERN_BUFSLOP in <sys/sysctl.h>.

(simonb)

Pad out the slop for kern.buf based on the passed in element size,
rather than a size of an unrelated struct.

(simonb)

smbios: Add character device for accessing SMBIOS tables

The /dev/smbios character device gives an aperture into physical memory
that allows read-only access to the SMBIOS header and tables.

(jmcneill)

acpi: /dev/acpi: fix bounds check when reading tables

(jmcneill)

efiboot: Allow "fs" as alias for "initrd", for compat with x86

(jmcneill)

For sh3, increment PC when PT_CONTINUE from trigger_trap(), as already
done for aarch64, arm, and powerpc. Otherwise, child is trapped to the
PTRACE_BREAKPOINT_ASM (== trapa) instruction indefinitely.

Fix tests/lib/libc/sys/t_ptrace_wait*:core_dump_procinfo.

(rin)

#include <sys/param.h> for COHERENCY_UNIT

(skrll)

#include <sys/param.h> for COHERENCY_UNIT and some KNF while I'm here.

(skrll)

For evbppc, use C version of memcpy(3), memcmp(3), and memmove(3)
consistently for *.{po,pico,go} (for RUMP), in order to avoid
alignment faults for 403.

(rin)

For evbppc, use C version of bcopy(3), memcpy(3), memcmp(3), and
memmove(3) consistently for debug library (*.go) in order to avoid
alignment faults for 403.

(rin)

The number of components used must be at least 2.  An odd number of components
is not allowed.

PR bin/45162

(oster)

evbarm: Add build.sh live-image support for armv7/aarch64

(jmcneill)

smbios: Support SMBIOS 2.x tables.

(jmcneill)

efiboot: Add support for SMBIOS 2.x tables.

(jmcneill)

Various disk queue "peek" routines were only ever used in the simulator
version of RAIDFrame.  Remove them from here.

(oster)

lint: rename cstk_t to control_statement

Rename the variables as well.  Their previous name 'ci' was not easy to
understand, the 'i' may have meant 'stack item'.  The new name 'cs'
simply means 'control statement'.

No functional change.

(rillig)

lint: remove a few unnecessary abbreviations

No functional change.

(rillig)

lint: make offset and alignment positive numbers

There's no use case for negative alignment or offsets in structures.

No functional change.

(rillig)

lint: write canonical form of keywords

For keywords that have a single spelling variant (such as __packed),
write this form in the source, to make it searchable.  This also avoids
a few calls to malloc.

Previously, some keywords had leading underscores and some hadn't, this
was inconsistent.

No functional change.

(rillig)

tests/lint: test lexer for the GCC extension __thread

(rillig)

lint: move single semicolon to external_declaration

This aligns more closely with the grammar from GCC's parser.  The global
cleanup from the grammar rule 'external_declaration:
top_level_declaration' is not performed anymore, which doesn't matter
since there is nothing to clean up after a single semicolon.

No functional change.

(rillig)

Use wildcard matches more consistently, and attach pci, eisa, isa, and
usb to their interface attributes, which makes for easier trimming-down
with "no" directives.

(thorpej)

gcc hates strncpy()

(martin)

Emulate the IEEE flavors of the FIX extension instructions:
- FTOIS, FTOIT
- ITOFS, ITOFT
- SRQTS, SQRTT

Left out for now are the VAX FP flavors:
- ITOFF
- SQRTF, SQRTG

(thorpej)

IEEE SQRTS / SQRTT is fix_float opclass 11, not fix_float opclass 10
(which is VAX SQRTF / SQRTG).

(thorpej)

- Expose alpha_fp_complete_at() to callers outside fp_complete.c.
- Use ufetch_32() instead of copyin() to fetch the instruction.

(thorpej)

All IO is async in the RAIDframe kernel driver, so desc->async_flag
isn't needed.  Cleanup the flag from rf_DoAccess() and its caller as
well.

(oster)

The 'pss_issued' pool is unused, so remove it.

(oster)

Extensive mechanical changes to the pools used in RAIDframe.

Alloclist remains not per-RAID, so initialize that pool
separately/differently than the rest.

The remainder of pools in RF_Pools_s are now per-RAID pools.  Mostly
mechanical changes to functions to allocate/destroy per-RAID pools.
Needed to make raidPtr available in certain cases to be able to find
the per-RAID pools.

Extend rf_pool_init() to now populate a per-RAID wchan value that is
unique to each pool for a given RAID device.

TODO: Complete the analysis of the minimum number of items that are
required for each pool to allow IO to progress (i.e. so that a request
for pool resources can always be satisfied), and dynamically scale
minimum pool sizes based on RAID configuration.

(oster)

getiobuf() can return NULL if there are no IO buffers available.
RAIDframe can't deal with that, so create a dedicated pool of buffers
to use for IO.  PR_WAITOK is fine here, as we pre-allocate more than
we need to guarantee IO can make progress.  Tuning of pool still to
come.

(oster)

fix typo in unused obsolete set name.  noted by tnn.

(mrg)

rtld expects an absolute path in execname provided via AUXV in order to
handle $ORIGIN properly and checks for that. Since we are calling its guts
directly, provide one.

(christos)

revert changes to fflush.c (1.18) and fvwrite.c (1.25) until we investigate
collateral damage. Breaks h_intr.c test. Requested by RVP.

(christos)

revert changes to fflush.c (1.18) and fvwrite.c (1.25) until we investigate
collateral damage. Breaks h_intr.c test. Requested by RVP.

(christos)

It's unlikely that a process will be using restartable atomic sequences
on Alpha, and furthermore it's unlikely that any given context switch
will be returning to one even if the process has them.  So, re-arrange
the RAS processing in cpu_switchto() so that the most likely code paths
are predicted by the branch predictor.  On an EV4-class processor, this
will save ~4-6 cycles on just about every context switch.

(thorpej)

#include <sys/param.h>

(skrll)

Minor shuffle to bring <machine/param.h> and <machine/limits.h> in before
any <sys/*.h> headers and for the COHERENCY_UNIT and CACHE_LINE_SIZE
defaults to be provided after the <machine/*.h> includes, but before the
<sys/*.h> includes.

COHERENCY_UNIT and CACHE_LINE_SIZE are used by a few <sys/*.h> filss.

I checked a handful of kernel builds produce the same binary before and
after this change.  I'll check more.

(skrll)

Various minor cleanups and bug fixes to the FP software completion code:
- Use __CTASSERT() instead of rolling our own compile-time assertion
  using cpp.
- Use __BIT() &c instead of rolling our own.
- Improve some comments.
- Define a default FP_C and FPCR value that is self-consistent, and
  initialize it properly at process creation time.
- Fix signal information when the trap shadow cannot be resolved.
- Use defined constants rather than magic numbers for the exception
  summary bits.
- Add a machdep sysctl to enable FP software-completion debugging.

(thorpej)

module_hold() returns void, not int

XXX module_hold() and module_rele() should probably not be exposed, since
XXX they really aren't MP-safe.

(pgoyette)

Use __BITS() to define MDLWP_FP_C.

(thorpej)

Expose SMBIOS tables using sysctl machdep.dmi.*, same as x86.

(jmcneill)

x86's platform.c no longer has any x86 specific code in it, so move it to
dev/smbios_platform.c to let other ports use it

(jmcneill)

Separate MI smbios interface from MD specific code.

(jmcneill)

Remove commented line that is a duplicate of a real line.

(oster)

lint: move assignments to $$ at the end of the action

They are closely related to return statements.

While here, add some more remarks from reviewing the grammar.

No functional change.

(rillig)

lint: rename grammar rules for specifier-qualifier-list

No functional change.

(rillig)

lint: remove unnecessary assertion for type qualifier

If there is any new type qualifier that lint should support (such as
_Atomic from C11), the obvious place is to look in the grammar rule
type_qualifier, so there is no need to have an assertion.

No functional change.

(rillig)

lint: rename ignore_up_to_rparen to match the grammar rule

No functional change.

(rillig)

libm: build fma(3), fmaf(3), fmal(3) on alpha

Alpha has no hardware-assisted fused multiply-add, so the fallback
C implementation must be used. A basic smoke test shows the
implementation works for
"x * x - 1.0f" vs "fmaf(x, x, -1.0f)" with "volatile float x = 1.0008f;"

(tnn)

Bump the version for ALTQ/CBQ changes

Welcome to 9.99.87

(ozaki-r)

altq: fix CBQ with RED and/or pf

(ozaki-r)

altq, cbq: apply tweaks for PPP to only PPP interfaces

(ozaki-r)

altqstat, cbq: follow uint64_t changes

PR kern/56319

(ozaki-r)

altq, cbq: change long to uint64_t for ILP32 archs

PR kern/56319

(ozaki-r)

altqstat: follow changes of types of variables

PR kern/56319

(ozaki-r)

altqstat: show raw values of avgidle, maxidle and minidle

The values are sometimes small and cooked values of them can be less
informative such as 0 or -1.

(ozaki-r)

altq, cbq: treat time in nanoseconds

Time granularity in microseconds is sometimes not enough to provide
accurate bandwidth control; actual bandwidth on a capped class
can exceed its limit considerably in some cases.

Treating time in nanoseconds requires the following changes:
- Use timespec instead of timeval
- Use nanotime(9) instead of microtime(9)
- Change the type of some variables, especially *idle, to long

(ozaki-r)

altqstat: adopt ps_per_byte

PR kern/56319

(ozaki-r)

libaltq, cbq: convert ns_per_byte to ps_per_byte

PR kern/56319

(ozaki-r)

need <sys/param.h> for COHERENCY_UNIT

Minor KNF along the way.

(skrll)

need <sys/param.h> for COHERENCY_UNIT

(skrll)

altq, cbq: convert ns_per_byte to ps_per_byte

Also the type of variables of it is changed to u_long from int.

This change provides fine-grain resolution of bandwidth.  For example
750 Mbps was treated as 800 Mbps internally because bandwidth was
represented as nanoseconds per byte.  Converting the representation
to picoseconds per byte enables to treat 750 Mbps as-is.

PR kern/56319

(ozaki-r)

Add AUDIO_SETINFO_gain_balance test.
The test checks whether AUDIO_SETINFO can change the gain and the balance
at the same time (if MD driver has the capability).  See PR kern/56308.

(isaki)

AUDIO_SETINFO: fix a bug that the gain and the balance could not be set
at the same time.  Fix PR kern/56308.

(isaki)

lint: use consistent naming scheme for functions that build nodes

No functional change.

(rillig)

lint: split 'build' into build_binary and build_unary

No functional change.

(rillig)

Extract just the random bits to feed to /dev/urandom.

This makes no difference in the randomness of the pool, but it improves
on the estimation (if any) of how many random bits were obtained.
Also make the ftp -q time out a bit longer since I got some time outs.

(rhialto)

Don't override /etc/rc.conf if it sets ec2_init.

(rhialto)

lint: make char unsigned on aarch64

This fixes tests msg_074 and msg_076, which previously warned for
'\xff': conversion of 'int' to 'char' is out of range [119]

The commit from 2014-08-10 that first defined char as signed had the
remark "Enough for a distribution build".  At that time, there was no
unit test for lint1 that would have detected the signedness of char.

(rillig)

PR 56303: do not borrow from the default swap allocation if we are in
tiny ram conditions and will need to enable swap early.

(martin)

RAM size will not change during installation, do the sysctl() dance
only once.

(martin)

need <sys/param.h> for COHERENCY_UNIT

(christos)

no need for extern kmutex_t itimer_mutex or #include <sys/mutex.h> here.

(skrll)

Trailing whitespace

(skrll)

Define the f_float in-memory format.  Line of the fields in s_float and
t_float.

(thorpej)

Define the FIX float function codes (ITOFF, ITOFS, ITOFT).

(thorpej)

Emulate the CIX extension instructions (CTPOP, CTTZ, CTLZ).

(thorpej)

NFC - if/else blocks start with the same mutex_exit, just move it up.

(andvar)

- In unaligned_fixup(), use defined constants for opcodes, not magic numbers.

- In handle_opdec(), use ufetch_int() to fetch the instruction and
  _u{fetch,store}_{8,16}() when emulating BWX instructions rather
  than copyin() / copyout() (they're faster).

- Add event counters for BWX instruction emulation.

(thorpej)

Various ena(4) bug fixes, from KUSABA Takeshi <t-kusaba@iij.ad.jp>:

[PATCH 01/32] include device_xname for evcnt group.
[PATCH 02/32] fix improper NULL check.
[PATCH 03/32] add tx drop counter
[PATCH 04/32] remove unimplemented m_getjcl
[PATCH 05/32] make ENA_MEM_{ALLOC,FREE}_COHERENT symmetric.
[PATCH 06/32] disestablish the correct interrupt.
[PATCH 07/32] fix null check target.
[PATCH 08/32] use if_initialize() and if_register() instead of
[PATCH 09/32] free all pci-related resource.
[PATCH 10/32] no need to call if_free(), but to call if_detach
[PATCH 11/32] add some locking assertions.
[PATCH 12/32] use bus_size_t for bus_dma instead of uint32_t.
[PATCH 13/32] no need to pass interlock, ena_timer_service does not
[PATCH 14/32] enable ena(4) to down I/F
[PATCH 15/32] destroy I/O queues before disestablishing msix
[PATCH 16/32] count input/output packet stats.
[PATCH 17/32] fix missing #ifdef LRO, NetBSD does not support LRO
[PATCH 18/32] MP-ify TX, allocate mbuf queue to each TX ring.
[PATCH 19/32] down the interface first when to detach, to prevent
[PATCH 20/32] default link speed should be unknown.
[PATCH 21/32] protect ena_adapter members only by "global_mtx".
[PATCH 22/32] lower global_mtx interrupt level
[PATCH 23/32] reorder function declaration
[PATCH 24/32] add locking notes and some marking.
[PATCH 25/32] process RX in workqueue context, as same as FreeBSD
[PATCH 26/32] count rx_drops correctly.
[PATCH 27/32] no need to lock when attach/detach, down/up
[PATCH 28/32] fix memory leak.
[PATCH 29/32] add "stopping" flag to ena_ring.
[PATCH 30/32] make the flags atomic.
[PATCH 31/32] do not schedule timer when device is down.
[PATCH 32/32] no need to start timer if I/F is down.

(jmcneill)

Release mutexes in few more places on failure path. Reviewed them in chfs code after fixing PR kern/56242.
ok riastradh

(andvar)

Mention caution changing the defaults.

(christos)

Re-arrange the kernel pmap fast-path in pmap_extract() so that
when DEBUG is not enabled, the compiler can emit a tail-call to
vtophys_internal().

(thorpej)

There is already a fast-path in pmap_extract() for the kernel pmap, so
don't bother doing a conditional for kernel vs. user-space here.

KASSERT() that pmap_extract() succeeds; it is a programming error if
it does not, and it's not a great idea to insert a garbage address
into the SGMAP page table.

(thorpej)

there is no need to keep pvos for unmanaged mappings on a hidden p/v list,
since "unmanaged" means that we don't want to find such pvos on any p/v list.
instead, just don't put such pvos on any p/v list at all and remove
the two hidden p/v lists for unmanaged mappings.  code mostly from martin,
to implement rin's suggestion of unifying the two hidden lists.

(chs)

Match the declaration in <sys/kauth.h>

(christos)

Remove definitions for CACHE_LINE_SIZE and COHERENCY_UNIT which are the same
as the default.

(christos)

There's no need to adjust `iov' in the error path.
Returning the amount written is all that's needed.
from RVP

(christos)

binutils 2.37 has been released.

(rin)

Abolish all the silly indirection macros for initializing vnode ops tables.

These are things of the form #define foofs_op genfs_op, or #define
foofs_op genfs_eopnotsupp, or similar. They serve no purpose besides
obfuscation, and have gotten cutpasted all over everywhere.

Part 3; cvs randomly didn't commit all the files the first time, still
hunting down the files it skipped.

(dholland)

Abolish all the silly indirection macros for initializing vnode ops tables.

These are things of the form #define foofs_op genfs_op, or #define
foofs_op genfs_eopnotsupp, or similar. They serve no purpose besides
obfuscation, and have gotten cutpasted all over everywhere.

Part 2; cvs randomly didn't commit these changes before, and then hid
them from me until I touched the files to force it to rethink. Dunno
what happened.

There's probably more of these, going to have to scan the tree the
hard way.

(dholland)

For CIA/Pyxis and Tsunami/Typhoon/Titan, save the firmware's idea of
the DMA window configuration, and restore it at shutdown time.  Make
sure that all assumptions we've made on the firmware's configuration
of DMA windows is correct.

(thorpej)

Fix ptyfs link and symlink ops.

They were set (via macro obfuscation just eliminated) to
genfs_abortop, which is the wrong thing. These code paths are
currently inaccessible because of EROFS enforcement in namei, but if
either were reached odd behavior would likely ensue. Use
genfs_eopnotsupp like the rest of the ops that don't make sense here.

(dholland)

Abolish all the silly indirection macros for initializing vnode ops tables.

These are things of the form #define foofs_op genfs_op, or #define
foofs_op genfs_eopnotsupp, or similar. They serve no purpose besides
obfuscation, and have gotten cutpasted all over everywhere.

(dholland)

Use macros for the canned parts of device and fifo vnode op tables.

Add GENFS_SPECOP_ENTRIES and GENFS_FIFOOP_ENTRIES macros that contain
the portion of the vnode ops table declaration that is
(conservatively) the same in every fs. Use these in every fs that
supports devices and/or fifos with separate ops tables.

Note that ptyfs works differently (it has one type of vnode with
open-coded dispatch to the specfs code, which I haven't changed in
this commit) and rump/librump/rumpvfs/rumpfs.c has an indirect dynamic
dispatch that already does more or less the same thing, which I also
haven't changed.

Also note that this anticipates a few bits in the next changeset here
and there, and adds missing but unreachable calls in some cases (e.g.
most fses weren't defining whiteout on devices and fifos, but it isn't
reachable there), and it changes parsepath on devices and fifos to
genfs_badop from genfs_parsepath (but it's not reachable there
either).

It appears that devices in kernfs were missing kqfilter, so it's
possible that if you try to use kqueue on /kern/rootdev that it'll
explode.

And finally note that the ops declaration tables aren't
order-dependent. (Other than vop_default_desc has to come first.)
Otherwise this wouldn't work.

(dholland)

According to section 8.1.2.2 of the Tsunami/Typhoon hardware reference
manual (DS-0025A-TE), the SGMAP TLB is arranged as 168 locations of 4
consecutive quadwords.  It seems that on some revisions of the Pchip,
SGMAP translation is not perfectly reliable unless we align the DMA
segments to the TLBs natural boundaries (observed on the API CS20).

N.B. the Titan (as observed on a Compaq DS25) does not seem to have this
problem, but we'll play it safe and run this way on both variants.

PR port-alpha/40604.

(thorpej)

aiomixer: Honor terminal colors. Looks better in cool-retro-term.

(nia)

Fix confusion arising from whether FOLLOW or NOFOLLOW is 0.

In vn_open, don't set and then throw away FOLLOW, and clarify the
comment about requesting FOLLOW/NOFOLLOW behavior.

Related to PR 56316.

(dholland)

PR kern/56316: vn_open doesn't follow symlinks

remove setting of O_NOFOLLOW. restores ZFS being able
to open devices via symlinks again.

(kardel)

skip symbol tables that were unloaded again to avoid EFAULT when reading
ksyms.

also restore TAILQ_FOREACH idiom.

(mlelstv)

Mark t_execsnoop xfail (PR 53417)

(dholland)

Allow for the SGMAP implementation to specify a minimum alignment for
SGMAP DMA segments.  If not specified, PAGE_SIZE will be used, as before.

(thorpej)

Oops, pass the correct bus_dma_tag_t to the "hi" sgmap init function.

(thorpej)

- Don't bother issuing a warning if we program the WBASE / WSM register
  to different values than the firmware.
- Reduce the differences in how we initialize the DMA tags vis a vis the
  other Alpha implementations.
- Use Window 2 to provide a 1G @ 3G PCI SGMAP window on systems with more
  than 1G of RAM, rather than falling back on the ISA DMA window which is
  small and could get starved by PCI devices.
- Make sure we set TBASE to 0 for direct-mapped windows.

(thorpej)

- Define the DAC enable bit that's present in WSBA3.
- Define symbolic constants for the valid WSM values.

(thorpej)

Use strlcpy() to copy the wedge device name (the code still assumes
the string is short enough and 0 terminated, but now gcc is happy).

(martin)

PR 56310: if we fail to create a wedge this either means there is
a bug here (and we requested something nonsensial), or there are pre-
existing "foreign" wedges which disturb our work.
So remove all wedges on this disk that we do not know about and retry
to add our new wedge.

(martin)

provide an equivalent alignment for __float128 for clang and i386 that
does not have it. Idea from mrg@

(christos)

Mark vlan_safe_ifpromisc_locked() as "__unused" to appease LLVM.

Maybe completely remove this short helper?

(hannken)

PR 56313: fix eroneous = that was meant to be ==

(martin)

PR 56310: avoid assert() failures (or crashes) when the runtime addition
of a wedge fails (for whatever reasons).

(martin)

puc(4): Add Intel Core 5G (mobile) KT.

(riastradh)

regen.

(mrg)

add current generation NVIDIA graphics cards (3050-3090 etc.)

(mrg)

Back in rev 1.21, mhitch@ fixed an issue with his 1.5GB RAM PWS 500au
by using a fall-back to the ISA DMA window if DMA was out of range for
the 1G @ 1G PCI DMA window.  Alas, the ISA DMA window is pretty small
(8M @ 8M), and it's possible to starve it with PCI devices that might
have, for example, large control data structures there.

So, instead, if the system has more than 1G of RAM, use Window 3
(previously unused) as a SGMAP window 1G @ 3G, and set that as the
fall-back if the direct-mapped window fails.

(thorpej)

Fix incorrect function name, some grammar and typos in comments. Remove trailing tab symbol.
No functional change intended.

(andvar)

Instrument success/faulure of phyisically contiguous uarea allocation.

(thorpej)

The Alpha AXP Architecture Reference Manual is explcit that the only
valid bits in the PSL are the IPL and USER bits, the latter of which
will always be clear when in the kernel, and that all other bits MBZ.
So, when reading the PSL to get the current IPL, don't bother masking
with ALPHA_PSL_IPL_MASK.

(thorpej)

On LCA45 systems, look at the memory controller's Bcache configuration
to initialize uvmexp.ncolors rather than hard-coding a value per model
(some models can have more than one configuration).

(thorpej)

Define the memory controller registers, and contents for the Cache
register.

(thorpej)

When fflush fails, adjust pointers and the io vectors. From RVP. Fixes
core-dump at cvs(1) exit(3).

(christos)

Add missing RCSID, remove non-breaking space characters in copyright, remove vim modeline (from yambo in IRC).
ok riastradh

(andvar)

Remove commented code, including empty if block. Replace tab with whitespace in usage declaration.

(andvar)

PR port-sh3/56311

Correct misinterpretation for the cause of the problem, and link to
the bug report for upstream.

(rin)

tests, altq: fix checks of altqd startup

Hopefully the fix stabilizes test results on qemu/anita.

(ozaki-r)

lint: reduce verboseness of merge_signedness

No functional change.

(rillig)

lint: rename merge_type_specifiers to merge_signedness

No functional change.

(rillig)

lint: replace switch statement in dcs_merge_declaration_specifiers

Grouping the rules by their abstract type took a lot of visual space.
Instead, move each of the rules from C11 6.7.2 into its own if
statement, so that the rules almost read like in the standard.

No functional change.

(rillig)

lint: extract dcs_merge_declaration_specifiers from end_type

No functional change.

(rillig)

lint: clean up mix of 'case' and 'if'

No functional change.

(rillig)

lint: remove warning about _Complex long double in traditional C

Complex types have been introduced in C99, so there is no danger of
anyone trying to use them in traditional C.

(rillig)

Make sure that mutex is released before conditional return statements. Fixes PR kern/56242
ok riastradh

(andvar)

lint: fix internal error for sizeof(typeof)

(rillig)