KNF

(ad)

strlcpy() the command line only if it's present.

(bouyer)

if_skreg.h: remove duplicate or triplicate identical #defines

(jakllsch)

Omit needless comment.

We've already committed part of the write, so ERESTART is definitely
not appropriate at this point.

(riastradh)

Simplify loops by putting interrupt test at end.

(riastradh)

No need for a private pool cache.  kmem serves just fine.

(riastradh)

Make variable unused outside kern_entropy.c static.

(riastradh)

Simplify /dev/random without reference to entropy_depletion.

(riastradh)

Use sysctl_teardown() upon detach of mskc(4).

(jakllsch)

Try to get the starting bus number from _CRS before falling back to _BBN.
There are apparently cases where the first bus in _CRS does not match the
value of _BBN, and the consensus is that _CRS should take precedence.

(jmcneill)

expose acpi_pcidev_pciroot_bus

(jmcneill)

msk(4): Simplify setting the first List Element owner-bit last.

We don't actually need to do this, as the BMU won't proceed to the new
descriptor until we advance the ring-end pointer on the chip, but perhaps
it has value as paranoia.

(jakllsch)

Whitespace nit.

(simonb)

Clear PSL_TRACEDCHILD on ptrace(PT_DETACH)

This avoids potential race of detaching and attaching to a prestarted
process and receiving SIGTRAP instead of SIGSTOP.

(kamil)

Search the descriptors of the appropriate interface to ensure correct
match.

PR kern/55240 cdce(4) error "no data interface" when ECM USB IF are
preceded by ACM USB IF

(skrll)

Add a new function, sleepq_transfer(), that moves an lwp from one
sleepq to another.

(thorpej)

Eliminate curcpu_available() hack.

The entropy subsystem is no longer used before curcpu() and curlwp
are available on x86.

(riastradh)

Make curcpu_available() always true.

This should work now that x86 runs cpu_init_rng just after curcpu()
and curlwp are initialized, and no other architecture needs it to
work that early.

(riastradh)

Move cpu_rng_init a little later, just after cpu_init_msrs, on x86.

This way curcpu() and curlwp are available, so that we no longer need
any annoying conditionalization in kern_entropy.c.

(riastradh)

Factor randomization out of slotspace_rand.

slotspace_rand becomes deterministic; the randomization moves into
the callers instead.  Why?

There are two callers of slotspace_rand:

- x86/pmap.c pmap_bootstrap
- amd64/amd64.c init_slotspace

When the randomization was introduced, it used an x86-only
`cpu_earlyrng' abstraction that would hash rdseed/rdrand and rdtsc
output together.  Except init_slotspace ran before cpu_probe, so
cpu_feature was not yet filled out, so during init_slotspace, the
only randomization was rdtsc.

In the course of the recent entropy overhaul, I replaced cpu_earlyrng
by entropy_extract, and moved cpu_init_rng much earlier -- but still
after cpu_probe -- in order to reduce the number of abstractions
lying around and the number of copies of rdrand/rdseed logic.  In so
doing I added some annoying complication (see curcpu_available) to
kern_entropy.c to make it work early enough for init_slotspace, and
dropped the rdtsc.

For pmap_bootstrap that didn't substantively change anything.  But
for init_slotspace, it removed the only randomization.  To mitigate
this, this commit pulls the randomization out of slotspace_rand into
pmap_bootstrap and init_slotspace, so that

(a) init_slotspace can use rdtsc and a little private entropy pool in
    order to restore the prior (weak) randomization it had, and

(b) pmap_bootstrap, which runs a little bit later, can continue to
    use entropy_extract normally and get rdrand/rdseed too.

A subsequent commit will move cpu_init_rng just a wee bit later,
after cpu_init_msrs, so the kern_entropy.c complications can go away.
Perhaps someone else more wizardly with x86 can find a way to make
init_slotspace run a little later too, after cpu_probe and after
cpu_init_msrs and after cpu_rng_init, but I am not that wizardly.

(riastradh)

Forgot to commit this file as part of elf.c::rev1.21 mm.c::rev1.27.

(maxv)

On debugger attach to a prestarted process don't report SIGTRAP

Introduce PSL_TRACEDCHILD that indicates tracking of birth of a process.
A freshly forked process checks whether it is traced and if so, reports
SIGTRAP + TRAP_CHLD event to a debugger as a result of tracking forks-like
events. There is a time window when a debugger can attach to a newly
created process and receive SIGTRAP + TRAP_CHLD instead of SIGSTOP.

Fixes races in t_ptrace_wait* tests when a test hangs or misbehaves,
especially the ones reported in tracer_sysctl_lookup_without_duplicates.

(kamil)

If no random seed file exists on boot, create one.

rndctl -S triggers entropy consolidation, so whatever we gathered
during kernel startup -- interrupt timings, autoconf timings, &c. --
will be incorporated into the seed and into subsequent data read from
/dev/urandom, just like if rndctl -L had run at this boot, and the
seed will carry them into the next boot too.

But it still avoids frequently consolidating entropy on any regular
schedule, in order to continue to mitigate iterative-guessing
attacks.

(riastradh)

Omit needless verbiage in error message.

(riastradh)

Go back using cpu_info_primary, all events are bound to vCPU 0 by default.
Register the event handler on cpu_info_primary.
While there. update the event counter on interrupts.
Finally this should be MP-safe.

(bouyer)

This should be mpsafe, register the event handler as such.

(bouyer)

Change event_set_handler() to take the target CPU parameter. If ci is NULL,
  event_set_handler() will choose the CPU and bind the event.
  If ci is not NULL the caller is responsible for binding the event.
Use a IPI xcall to register the handlers if needed.
pull in a hack from x86 to force pirq handlers to be mpsafe if registered at
a level != IPL_VM. This is for the com at isa interrupt handler, which
registers at IPL_HIGH and has to way to tell it's mpsafe (taking
KERNEL_LOCK at IPL_HIGH causes deadlocks on MP systems).

(bouyer)

Localify.

(maxv)

Trigger entropy consolidation before saving seed.

This way, whenever /etc/security runs infrequently (daily), or the
operator manually issues rndctl -S, we ensure that all samples taken
during the entire boot are hashed together in the seed for the next
boot.

This should be infrequent enough that it's unlikely to enable the
iterative-guessing attacks that we try to mitigate by not frequently
consolidating entropy.

(riastradh)

Rework rndctl seed load sequence again.

Go back to the book's order, now that writing to /dev/random
guarantees to consolidate entropy -- this way the _next_ boot is no
less secure than the current boot, in the event that entropy sources
like interrupt timings provided any security that we just don't know
how to measure honestly.

Make sure to open the old seed to overwrite and the new seed to write
anew first so that we can determine whether the medium is read-only
before accepting the file's entropy estimate.

(riastradh)

Touch up rndctl(8) a bit.

(riastradh)

Print `entropy: ready' only when we first have full entropy.

Now that we consolidate entropy in rndctl -L and equivalent, not just
when the operator chooses, epoch != -1 no longer necessarily means
full entropy -- it just means `time to (re)seed, whether justified by
entropy accounting or by explicit consolidation'.

There is a bug on x86 systems with RDRAND/RDSEED that prevents this
message from appearing at all: it happens so early that consinit has
not run yet, so it just goes into oblivion.  Need to fix that some
other way!

(riastradh)

Consolidate entropy on RNDADDDATA and writes to /dev/random.

The man page for some time has advertised:

  Writing to either /dev/random or /dev/urandom influences subsequent
  output of both devices, guaranteed to take effect at next open.

So let's make that true again.

It is a conscious choice _not_ to consolidate entropy frequently.
For example, if you have a _slow_ HWRNG, which provides 32 bits of
entropy every few seconds, and you reveal a hash that to the
adversary before any more comes in, the adversary can in principle
just keep guessing the intermediate state by a brute force search
over ~2^32 possibilities.

To mitigate this, the kernel generally tries to avoid consolidating
entropy from the per-CPU pools until doing so would bring us from
zero entropy to full entropy.

However, there are various _possible_ sources of entropy which are
just hard to give honest estimates for that are valid on ~all
machines -- like interrupt timings.  The time at which we read a seed
in, which usually happens via /etc/rc.d/random_seed early in
userland, is a reasonable time to gather this up.  An operator or
system engineer who knows another opportune moment can always issue
`sysctl -w kern.entropy.consolidate=1'.

Prompted by a suggestion from nia@ to consolidate entropy at the
first transition to userland.  I chose not to do that because it
would likely cause warning fatigue on systems that are perfectly fine
with a random seed -- doing it this way instead lets rndctl -L
trigger the consolidation automatically.  A subsequent commit will
reorder the operations in rndctl again to make it work out better.

(riastradh)

Tickets #890 - #895

(martin)

Pull up following revision(s) (requested by jmcneill in ticket #895):

sys/arch/arm/cortex/gic_v2m.h: revision 1.3
sys/arch/arm/cortex/gic_v2m.c: revision 1.9

Do not store a pointer to the passed in struct pci_attach_args

(martin)

Pull up following revision(s) (requested by kim in ticket #894):

usr.bin/finger/finger.1: revision 1.22
usr.bin/finger/util.c: revision 1.30

Add lastlogx support

(martin)

Pull up following revision(s) (requested by kim in ticket #893):

usr.sbin/lastlogin/lastlogin.c: revision 1.16
usr.sbin/lastlogin/lastlogin.c: revision 1.17
usr.sbin/lastlogin/lastlogin.c: revision 1.18
usr.sbin/lastlogin/lastlogin.c: revision 1.19
usr.sbin/lastlogin/lastlogin.c: revision 1.20
usr.sbin/lastlogin/lastlogin.8: revision 1.13
usr.sbin/lastlogin/lastlogin.8: revision 1.14

Size output columns dynamically by default to fit contents.

If the passwd entry is not found for a lastlogx entry, cons up a fake
struct passwd where pw_name is the numeric uid in parentheses. This was
already implemented for lastlog entries in revision 1.13.

If -n is specified more than once, also print the user numerically
(ie, uid instead of username) for lastlog entries. This was already
implemented for lastlogx entries in revision 1.13.

Reorder the lastlogx host name numeric condition so it better matches
the user name numeric condition.

Use memcpy for copying out lastlog and lastlogx text
Lastlog and lastlogx text fields are not NUL-terminated when original data
is truncated.

Output records in original order

Append to the list of output records instead of pushing on it, so we
don't reverse the order of records (when not sorting).

New sentence, new line.

Sync usage with man page.

(martin)

Pass full pathname to df, and print warning message on failure.

No need to extract dirname; `df -P /var/db/entropy-file' and `df -l
/var/db/entropy-file' work just fine.

(riastradh)

Fix LOCKDEBUG compilation on i386.

(maxv)

Update the comments.

(maxv)

Clarify.

(maxv)

Explain more.

(maxv)

Pull up following revision(s) (requested by macallan in ticket #892):

sys/dev/sdmmc/if_bwfm_sdio.c: revision 1.15

add entry for BCM43362, found on Cubietruck
ok jmcneill@

(martin)

Pull up following revision(s) (requested by jmcneill in ticket #891):

sys/arch/arm/sunxi/sun4i_a10_ccu.c: revision 1.12

Add A20 CLK_OUT_A and CLK_OUT_B clocks

(martin)

Ticket #1546

(martin)

Pull up following revision(s) (requested by aymeric in ticket #1546):

bin/rcp/rcp.c: revision 1.50

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

Reported by Daniel Goujot, Georges-Axel Jaloyan, Ryan Lahfa, and David Naccache.

(martin)

Pull up following revision(s) (requested by aymeric in ticket #890):

bin/rcp/rcp.c: revision 1.50

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

Reported by Daniel Goujot, Georges-Axel Jaloyan, Ryan Lahfa, and David Naccache.

(martin)

If we encounter relocations from a section that the bootloader dropped,
AND if the section is a note, then skip the relocations.

Considering a note that the bootloader dropped, there are two possible
sides for the relocations: (1) the relocations from the note towards the
rest of the binary, and (2) the relocations from the rest of the binary
towards the note.

We skip (1), which is correct, because the notes do not play any role at
run time. If we encounter (2) however then there is a bug in the kernel,
so add a sanity check against that.

This fixes KASLR since the latest Xen changes (which introduced .note.Xen).

(maxv)

Do not store a pointer to the passed in struct pci_attach_args

(jmcneill)

Cast physical address to uintptr_t, fix PAE build.
Pointed out by John D. Baker

(bouyer)

Add lastlogx support

(kim)

Add -i to SYNOPSIS.

(wiz)

Fix macro usage.

(wiz)

Fix macro usage. Fix typo. New sentence, new line.

(wiz)

Sync usage with man page.

(wiz)

New sentence, new line.

(wiz)

Whitespace fix

(sborrill)

Ticket #1731

(sborrill)

Pull up the following revisions(s) (requested by aymeric in ticket #1731):
bin/rcp/rcp.c: revision 1.50

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

(sborrill)

Ticket #1731

(sborrill)

Pull up the following revisions(s) (requested by aymeric in ticket #1731):
bin/rcp/rcp.c: revision 1.50

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

(sborrill)

Ticket #1731

(sborrill)

Pull up the following revisions(s) (requested by aymeric in ticket #1731):
bin/rcp/rcp.c: revision 1.50

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

(sborrill)

add entry for BCM43362, found on Cubietruck
ok jmcneill@

(macallan)

Add PRID definition for newer SiByte SB1 cores (rev 0x11).
Add a constant for SiByte/BCRM cacheable coherent TLB cache attribute.

(simonb)

Add A20 CLK_OUT_A and CLK_OUT_B clocks

(jmcneill)

Operation zfs_zget_cleaner() cannot fail, comment and add assertions.

(hannken)

Revert Rev. 1.31 as it is no longer possible for the handle to be NULL.

(hannken)

Revert Rev. 1.63 and add a comment why we have to zil_commit() here:

Operation zfs_znode.c::zfs_zget_cleaner() depends on this
zil_commit() as a barrier to guarantee the znode cannot
get freed before its log entries are resolved.

(hannken)

Fix two mistakes in entropy accounting.

1. When consolidating entropy from per-CPU pools, drop the amount
  pending to zero; otherwise the entropy consolidation thread might
  keep consolidating every second.

This uncovered a feedback loop with kern.entropy.depletion=1 and
on-demand entropy sources, which is that depleting the pool and then
requesting more from it causes the on-demand entropy sources to
trigger reseed, which causes cprng_fast/strong to request more which
depletes the pool again which causes on-demand entropy sources to
trigger reseed, and so on.

To work around this:

2. Set a rate limit on reseeding (advancing the entropy epoch) when
  kern.entropy.depletion=1; otherwise reseeding gets into a feedback
  loop when there are on-demand entropy sources like RDRAND/RDSEED.

(By default, kern.entropy.depletion=0, so this mainly only affects
systems where you're simulating what happens when /dev/random blocks
for testing.)

(riastradh)

The event channel is not necesserely bound to the primary CPU.
Use the CPU returned in the interrupt handler instead.

(bouyer)

Document the HDAUDIO_ENABLE_HDMI and HDAUDIO_ENABLE_DISPLAYPORT options.

(rkujawa)

Drop xbdi_lock() before calling intr_disestablish(), it needs to take
cpu_lock.

(bouyer)

Attempts to make MP-safe:
xpmap_ptom_unmap() doens't need to be called at splvm; we own the pa.
Use atomic ops to change pmap_pa_end
Make sure the ptom/mtop are up to date before giving the pages back to the
pool.

(bouyer)

Make MP-safe: make sure the xpq_queue* are flushed before making the
pages visible to UVM.

(bouyer)

We can't take v_interlock with a spin lock held. Release xbdi_lock
before scheduling the I/O, and take again once queued.
It looks safe to drop the lock here.

(bouyer)

xpq_queue_* use per-cpu queue; splvm() is enough to protect them.
remove the XXX SMP comments.

(bouyer)

Output records in original order

Append to the list of output records instead of pushing on it, so we
don't reverse the order of records (when not sorting).

(kim)

x86/x86/ipi.c should not be built for XENPV, even if dom0ops is defined.

(bouyer)

Use memcpy for copying out lastlog and lastlogx text

Lastlog and lastlogx text fields are not NUL-terminated when original data
is truncated.

(kim)

Tweak logic to decide whether a medium is safe for an rndseed.

- Teach rndctl to load the seed, but treat it as zero entropy, if the
  medium is read-only or if the update fails.

- Teach rndctl to accept `-i' flag instructing it to ignore the
  entropy estimate in the seed.

- Teach /etc/rc.d/random_seed to:
  (a) assume nonlocal file systems are unsafe, and use -i, but
  (b) assume / is safe, even if it is nonlocal.
  If the medium is nonwritable, leave it to rndctl to detect that.
  (Could use statvfs and check for ST_LOCAL in rndctl, I guess, but I
  already implemented it this way.)

Treating nonlocal / as safe is a compromise: it's up to the operator
to secure the network for (e.g.) nfs mounts, but that's true whether
we're talking entropy or not -- if the adversary has access to the
network that you've mounted / from, they can do a lot more damage
anyway; this reduces warning fatigue for diskless systems, e.g. test
racks.

(riastradh)

Update to reflect change to message.

(riastradh)

Don't reject seed file entropy estimates, until one is nonzero.

We try to avoid counting the seed file's entropy twice, e.g. once
from the boot loader and once from rndctl via /etc/rc.d/random_seed.

But previously, if you had a /var/db/entropy-file that was deemed to
have zero entropy, that would prevent rndctl -L from _ever_ setting a
nonzero entropy estimate, even if you (say) copy a seed file over
from another machine (over a non-eavesdroppable medium) and try to
load it in with rndctl -L, e.g. via `/etc/rc.d/random_seed start'.

Now we accept the first _nonzero_ entropy estimate from a seed file.

The operator can still always trick the kernel into believing there's
entropy in the system by writing data to /dev/random, if the operator
knows something the kernel doesn't; this only affects the _automated_
seed file loading.

(riastradh)

In sink(), upon error, avoid multiple replies to the source as this
would lead to a desynchronization of the protocol and further files or
directories to be ignored or corrupted.

Reported by Daniel Goujot, Georges-Axel Jaloyan, Ryan Lahfa, and David Naccache.

(aymeric)

KASSERT() that the per-cpu queues are run at IPL_VM after boot.

(bouyer)

No need to run hypercalls at IPL_VM.
XXX still not MP-safe

(bouyer)

Convert remaining splfoo/splx and tsleep/wakeup to mutex and condvar.
Mark kernel threads and interrupt handlers MPSAFE.

(bouyer)

Add getentropy() to libc - a simple wrapper to access the kernel CSPRNG.

Posted to tech-userlevel@ a week ago and reviewed by riastradh@.

GETENTROPY(3)   Library Functions Manual GETENTROPY(3)

NAME
    getentropy - fill a buffer with high quality random data

LIBRARY
    Standard C Library (libc, -lc)

SYNOPSIS
    #include <unistd.h>

    int
    getentropy(void *buf, size_t buflen);

DESCRIPTION
    The getentropy() function fills a buffer with high quality random data,
    suitable for seeding cryptographically secure psuedorandom number
    generators.

    getentropy() is only intended for seeding random number generators and is
    not intended for use by regular code which simply needs secure random
    data.  For this purpose, please use arc4random(3).

    The maximum value for buflen is 256 bytes.

IMPLEMENTATION NOTES
    getentropy() reads from the sysctl(7) variable kern.arandom.

RETURN VALUES
    The getentropy() function returns 0 on success, and -1 if an error
    occurred.

ERRORS
    getentropy() will succeed unless:

    [EFAULT] The buf argument points to an invalid memory address.

    [EIO] Too many bytes were requested.

SEE ALSO
    arc4random(3), rnd(4)

STANDARDS
    The getentropy() function is non-standard.

HISTORY
    The getentropy() function first appeared in OpenBSD 5.6, then in
    FreeBSD 12.0, and NetBSD 10.

(nia)

The name of the action used with blacklistd's control script is rem,
not remove.  From Jose Luis Duran in PR bin/55195.

(snj)

netsmb is no more

(snj)

If the passwd entry is not found for a lastlogx entry, cons up a fake
struct passwd where pw_name is the numeric uid in parentheses. This was
already implemented for lastlog entries in revision 1.13.

If -n is specified more than once, also print the user numerically
(ie, uid instead of username) for lastlog entries. This was already
implemented for lastlogx entries in revision 1.13.

Reorder the lastlogx host name numeric condition so it better matches
the user name numeric condition.

(kim)

remove the per-channel spin lock and instead make sure that events
add/remove will only be done on the CPU the handler is bound to, with
interrupts disabled.
Should be similar to the native x86 interrupts add/remove.

(bouyer)

Bump date for previous change...  I frequently forget that bit.

(kre)

Allocate one more byte so that we are always NUL-terminated, and remove
the extra commented out NUL-terminations. As suggested in:

    http://mail-index.netbsd.org/source-changes-d/2020/04/01/msg012470.html

(christos)

Size output columns dynamically by default to fit contents.

(kim)

Bump INSTIMAGEMB, hopefully fix releng autobuilds

(bouyer)

kill is built-in to more than just csh(1).
While here, add missing Xr sh 1 (which was previously needed, moreso now)
and also include STOP and CONT in the list of common signals.

(kre)

Ticket #1730

(sborrill)

Fix bug in sha384 checksum computation introduced when converting the code
to avoid alignment issues.

(sborrill)

Avoid buffer overflow

Disable the offending code.

OK by kre@

(kamil)

Fix a bug in the futex_wake_highest_pri test case.  Still fails as
expected.

(thorpej)

Behave like gmake: empty string does not match % pattern

(christos)

lwp_unpark(): no need to acquire LWP refs or drop the proc lock.

On the hacky benchmarks I have, held over from the transition to 1:1
threading, this restores pthread_cond_signal() perf to radixtree/sleepq
levels, and semes much better than either with pthread_cond_broadcast() and
10 threads.  It would be interesting to see what might be achieved with a
lockless lookup, which is within grasp now thanks to pid_table being used
for lookup.

(ad)

PR port-amiga/55233 Process hangs indefinitely if not calling syscalls for a while

cpu_intr_p() is broken on amiga, fix it.

From code inspection it looks like amiga and other m68k ports check for ASTs
with interrupts enabled in some cases, which is racy.  Not fixed.

(ad)

switch hppa to new binutils (2.34)

(skrll)

Disable attempt to use ifunc on hppa.  It doesn't work and new binutils
throws a build error.

(skrll)

Mark __cerror as hidden to avoid using the PLT.  This is required for new
binutils where the PLT stubs now use %t1 (%r22) which is used to pass
the errno to __cerror.

(skrll)

Add a space before any non-nullified instruction.  NFCI.

(skrll)

fix KASSERT() for MHLEN case in m_defrag() - network stack usually does
m_adj(ETHER_ALIGN) so check that the mbuf chain data fits
M_LEADINGSPACE() + M_TRAILINGSPACE()

(jdolecek)

Pull up following revision(s) (requested by maya in ticket #888):

sys/kern/uipc_sem.c: revision 1.59

Release the collision if we find one.

Candidate fix for:
panic: lock error: Mutex: mutex_vector_enter,542: locking against myself: lock 0xffff8f611abd37e0 cpu 8 lwp 0xffff8f60a3c6a040
cpu8: Begin traceback...
vpanic() at netbsd:vpanic+0x178
snprintf() at netbsd:snprintf
lockdebug_abort() at netbsd:lockdebug_abort+0xe6
mutex_vector_enter() at netbsd:mutex_vector_enter+0x3c1
ksem_close_fop() at netbsd:ksem_close_fop+0x17
closef() at netbsd:closef+0x69
fd_free() at netbsd:fd_free+0x101
exit1() at netbsd:exit1+0x118
sys_exit() at netbsd:sys_exit+0x3d
syscall() at netbsd:syscall+0x299

Would be nice to have an automatic test for this.  Since semids are
only 24 bits, we only need to create a few thousand of them to have a
high probability of collision.  Maybe we should bump default semmax
while here...

(martin)

Gather the section filtering in a single function, and add a sanity check
when relocating, to make sure the section we're accessing is mappable.

Currently this check fails, because of the Xen section, which has RELAs but
is an unmappable unallocated note.

Also improve the prekern ASSERTs while here.

(maxv)

Ticket #1545

(martin)

Apply patch, requested by he in ticket #1545:

Fix bug revealing itself in sha384 checksum computation: one
important statement was overlooked when converting the code to
avoid alignment issues.

(martin)

Tickets #888 and #889

(martin)

Pull up following revision(s) (requested by jmcneill in ticket #889):

sys/external/bsd/compiler_rt/dist/lib/builtins/clear_cache.c: revision 1.4

Align addresses to cache lines in __clear_cache for aarch64.

This corrects an issue where if the start and end address fall in different
lines, and the end address is not cache line size aligned, the last line
will not be invalidated properly.

Patch from compiler-rt upstream: https://reviews.llvm.org/rCRT323315

(martin)

Skip traceme_raise* tests rather than failing upfront

This avoid reporting the tests as unexpected failure.

(kamil)

Make DOM0OPS build for PVH/PVHVM too

(bouyer)

disable I/O or mem decode before probing BAR size.
Bar size is probed writing 0xffffffff to the BAR and reading back; but while
doing this the decoding address is not guaranteed to be valid and could have
side effect.

Xen PVH enforces disabling decoding before writing to a BAR.

Proposed on tech-kern@, got positive comments

(bouyer)

A tiny little more.

(christos)

New sysctl kern.entropy.gather=1 to trigger entropy gathering.

Invokes all on-demand RNG sources.  This enables HWRNG driver
developers to use a dtrace probe on rnd_add_data to examine the data
coming out of the HWRNG:

dtrace -n 'fbt::rnd_add_data:entry /args[0]->name == "amdccp0"/ {
  ...examine buffer args[1] length args[2]...
}'

(riastradh)

Revert "Use cv_timedwaitclock_sig in futex."

Turned out to break things; we'll do this another way.

(riastradh)

Revert "Make sure futex waits never return ERESTART."

Part of redoing the timedwaitclock changes, which were buggy and
committed a little too fast.

(riastradh)

Align addresses to cache lines in __clear_cache for aarch64.

This corrects an issue where if the start and end address fall in different
lines, and the end address is not cache line size aligned, the last line
will not be invalidated properly.

Patch from compiler-rt upstream: https://reviews.llvm.org/rCRT323315

(jmcneill)

due to the locking wild west for if_ioctl can't assert IFNET_LOCKED()
in xennet_ioctl(), so just do splnet() like other drivers do, and hope for best

fixes failed KASSERT() e.g. when starting rpcbind(), which ends
up calling this via sys_setsockopt()->sosetopt()->...->in6_addmulti()->
if_mcast_op(), this path doesn't currently take IFNET_LOCK()

(jdolecek)

write down some observations about the locking wild west around if_mcast_op()

(jdolecek)

adjust comment - sosetopt() path doesn't take IFNET_LOCK()

(jdolecek)

add a NOMPSAFE comment for if_mcast_op(), it is called from context
which doesn't hold IFNET_LOCK() in some cases, and calls if_ioctl

this needs to be sorted out for NET_MPSAFE

(jdolecek)

build start_xenpvh only if XEN. Fixes NET4501 build issue reported by
John D. Baker

(bouyer)

bump version - struct ifnet change

(jdolecek)

remove struct ifnet if_mcastop, it's not used by anything

(jdolecek)

Fix KASAN, init_xen_early must be called after kasan_early_init.

(maxv)

Add a space before any non-nullified instruction.  NFCI.

(skrll)

Use the delay slot

(skrll)

Remove unnecesary #define/#undef _LOCORE

(skrll)

Move misc tests out of t_ptrace_wait.c to t_ptrace_misc_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Add OWASP

(ginsbach)

Move core tests out of t_ptrace_wait.c to t_ptrace_core_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move siginfo tests out of t_ptrace_wait.c to t_ptrace_siginfo_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move threads tests out of t_ptrace_wait.c to t_ptrace_threads_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move topology tests out of t_ptrace_wait.c to t_ptrace_topology_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move exec() tests out of t_ptrace_wait.c to t_ptrace_exec_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move LWP tests out of t_ptrace_wait.c to t_ptrace_lwp_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Remove the duplicate ATF_TP_ADD_TC() entries

Event mask tests are already defined in
ATF_TP_ADD_TCS_PTRACE_WAIT_EVENTMASK().

(kamil)

Move eventmask tests out of t_ptrace_wait.c to t_ptrace_eventmask_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Remove the duplicate ATF_TP_ADD_TC() entries

Byte transfer tests are already defined in
ATF_TP_ADD_TCS_PTRACE_WAIT_BYTETRANSFER().

(kamil)

Move signal tests out of t_ptrace_wait.c to t_ptrace_signal_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move fork/vfork/posix_spawn tests out of t_ptrace_wait.c to t_ptrace_fork_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Bump (c) year

(kamil)

Move clone() tests out of t_ptrace_wait.c to t_ptrace_clone_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move byte transfer tests out of t_ptrace_wait.c to t_ptrace_bytetransfer_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move kill()-like tests out of t_ptrace_wait.c to t_ptrace_kill_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move PT_STEP tests out of t_ptrace_wait.c to t_ptrace_step_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move syscall tests out of t_ptrace_wait.c to t_ptrace_syscall_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

Move register tests out of t_ptrace_wait.c to t_ptrace_register_wait.h

The same tests are now included with the preprocessor in t_ptrace_wait.c.

No functional change intended.

(kamil)

constify

(jdolecek)

Prevent double definition of pt_entry_t from machine/param.h

(joerg)

New timedwaitclock_setup.

C99 initializers would have been nice, but part of the struct is
explicit parameters and part of the struct is implicit state, and
-Wmissing-field-initializers can't discriminate between them
(although for some reason it doesn't always fire!).

Instead, just do:

        struct timedwaitclock T;

        timedwaitclock_setup(&T, timeout, clockid, flags, epsilon);
        while (...) {
                error = timedwaitclock_begin(&T, &timo);
                if (error)
                        ...
                error = waitwhatever(timo);
                timedwaitclock_end(&T);
                ...
        }

(riastradh)

boardype should not be common.

(joerg)

note feature-sg and jumbo frames for xennet(4)/xvif(4), MSI for XenPV

(jdolecek)

add support for using MSI for XenPV Dom0

use PHYSDEVOP_map_pirq to get the pirq/gsi for MSI/MSI-X, switch also INTx
to use it instead of PHYSDEVOP_alloc_irq_vector

MSI confirmed working with single-vector MSI for wm(4), ahcisata(4), bge(4)

XXX added some provision for MSI-X, but it doesn't actually work (no interrupts
delivered), needs some further investigation; disable MSI-X for XENPV
via flag in x86/pci/pci_machdep.c

(jdolecek)

Improve markup.

(wiz)

Break line after macro arguments end. Use \- for minus.

(wiz)

Add a test case for PR kern/55230.  It is currently marked as expect-fail.

(thorpej)

Tickets #885 - #887

(martin)

Pull up following revision(s) (requested by christos in ticket #887):

crypto/external/bsd/netpgp/lib/verify/Makefile: revision 1.13
crypto/external/bsd/netpgp/lib/verify/verify.map: revision 1.1

Add a linker map to hide all the symbols the this library accidentally
exported. In particular the following symbols:

DSA_SIG_free
DSA_SIG_new
DSA_do_sign
DSA_do_verify
DSA_free
DSA_new
DSA_size
RSA_check_key
RSA_free
RSA_generate_key
RSA_new
RSA_private_decrypt
RSA_private_encrypt
RSA_public_decrypt
RSA_public_encrypt

conflict with libcrypto and break pkg_add which links against both
libraries.

(martin)

Release the collision if we find one.

Candidate fix for:

panic: lock error: Mutex: mutex_vector_enter,542: locking against myself: lock 0xffff8f611abd37e0 cpu 8 lwp 0xffff8f60a3c6a040
cpu8: Begin traceback...
vpanic() at netbsd:vpanic+0x178
snprintf() at netbsd:snprintf
lockdebug_abort() at netbsd:lockdebug_abort+0xe6
mutex_vector_enter() at netbsd:mutex_vector_enter+0x3c1
ksem_close_fop() at netbsd:ksem_close_fop+0x17
closef() at netbsd:closef+0x69
fd_free() at netbsd:fd_free+0x101
exit1() at netbsd:exit1+0x118
sys_exit() at netbsd:sys_exit+0x3d
syscall() at netbsd:syscall+0x299

Would be nice to have an automatic test for this.  Since semids are
only 24 bits, we only need to create a few thousand of them to have a
high probability of collision.  Maybe we should bump default semmax
while here...

(riastradh)