dlg [Wed, 16 Feb 2022 01:45:31 +0000 (01:45 +0000)]
fill in support for rx prio handling.
dlg [Wed, 16 Feb 2022 01:25:45 +0000 (01:25 +0000)]
rewrite vxlan to better fit the current kernel infrastructure.
the big change is removing the integration with and reliance on
bridge(4) for learning vxlan endpoints. we have the etherbridge
layer now (which is used by veb, nvgre, bpe, etc) so vxlan can
operate independently of bridge(4) (or any other driver) while still
dynamically learning about other endpoints.
vxlan now uses the udp socket upcall mechanism to receive packets.
this means it actually creates and binds udp sockets to use rather
adding code in the udp layer for stealing packets from the udp
layer.
i think it's also important to note that this adds loop prevention
to the code. this stops a vxlan interface being used to transmit a
packet that was encapsulated in itself.
i want to clear this out of my tree where it's been sitting for
nearly a year. noone seems too concerned with the change either
way.
ok claudio@
rob [Tue, 15 Feb 2022 23:16:00 +0000 (23:16 +0000)]
Reintroduce ps state flag 'c' indicating chrooted process (via PS_BITS).
Ok deraat@
deraadt [Tue, 15 Feb 2022 21:17:12 +0000 (21:17 +0000)]
Since acpitoshiba brightness button processing no longer plays games
with AML parsing outside the acpi thread, the locking-release dance
around wsdisplay_{suspend,resume} can be removed
ok kettenis
kettenis [Tue, 15 Feb 2022 21:13:39 +0000 (21:13 +0000)]
Defer the actual setting of the display brightness to the acpi thread.
ok deraadt@
deraadt [Tue, 15 Feb 2022 16:54:48 +0000 (16:54 +0000)]
when the MI suspend code encounters problems, we need a way to
reset the MD state before bailing out. New MD function sleep_abort()
does that.
nicm [Tue, 15 Feb 2022 13:11:29 +0000 (13:11 +0000)]
Add an option (default off) to control the passthrough escape sequence.
Like set-clipboard and allow-rename it is safer to forbid this by
default.
nicm [Tue, 15 Feb 2022 13:03:02 +0000 (13:03 +0000)]
Do not create a buffer from an OSC 52 response if we have not sent a
query.
claudio [Tue, 15 Feb 2022 12:24:55 +0000 (12:24 +0000)]
com at acpi like com at isa needs com_active() for suspend/resume support.
Fixes my com0 issues after zzz and ZZZ.
OK kettenis@ miod@
jsg [Tue, 15 Feb 2022 12:04:44 +0000 (12:04 +0000)]
unifdef __OpenBSD__
kn [Tue, 15 Feb 2022 11:54:19 +0000 (11:54 +0000)]
Backout previous "Unwire with map lock held" (commitid: SsVz7dLGFgR21kFe)
The (known) lock order reversals which now occur more reliably and much
earlier on WITNESS boots with this diff knock out syzcaller reports since
syzcaller stops at the first "crash report":
https://syzkaller.appspot.com/bug?id=
81b39e970cd2eb21b97d1b31746c693e300fd2dd
jsg [Tue, 15 Feb 2022 10:36:59 +0000 (10:36 +0000)]
unifdef OpenBSD
ok miod@
jsg [Tue, 15 Feb 2022 10:30:23 +0000 (10:30 +0000)]
consistently use __linux__
visa [Tue, 15 Feb 2022 08:43:50 +0000 (08:43 +0000)]
Use knote_modify_fn() and knote_process_fn() in bpf.
OK dlg@
jsg [Tue, 15 Feb 2022 08:17:50 +0000 (08:17 +0000)]
unifdef CDIOCCLOSE __OpenBSD__
espie [Tue, 15 Feb 2022 08:02:03 +0000 (08:02 +0000)]
make use of pkg_* infra simplifications
okay afresh1@
espie [Tue, 15 Feb 2022 07:35:35 +0000 (07:35 +0000)]
simplify code based on afresh1@'s remark: indeed I always need to
have a state once I'm past the initial locking
djm [Tue, 15 Feb 2022 05:13:36 +0000 (05:13 +0000)]
document the unbound/host-bound options to PubkeyAuthentication;
spotted by HARUYAMA Seigo
dlg [Tue, 15 Feb 2022 04:19:52 +0000 (04:19 +0000)]
only tweak ifp if_flags while holding NET_LOCK.
tun_dev_open and tun_dev_close were being optmistic.
dlg [Tue, 15 Feb 2022 04:16:10 +0000 (04:16 +0000)]
make tun_link_state take the ifnet pointer instead of tun_softc.
it only works on struct ifnet data, so passing ifp makes it clearer
what's actually being manipulated. also fix tun_dev_open so
tun_link_state is called before if_put instead of immediately after.
jsg [Tue, 15 Feb 2022 03:53:58 +0000 (03:53 +0000)]
unifdef TIOCHPCL, 4.3BSD compat ioctl
ok deraadt@ guenther@
jsg [Tue, 15 Feb 2022 03:31:17 +0000 (03:31 +0000)]
remove unused and uneeded bits in a byte define
posix requires a byte to be 8 bits
deraadt [Tue, 15 Feb 2022 02:38:16 +0000 (02:38 +0000)]
MI disable_lid_wakeups() is not needed, x86 systems can do this
in sleep_resume(), which seems sensible for other future systems also
deraadt [Tue, 15 Feb 2022 02:29:23 +0000 (02:29 +0000)]
The DVACT_POWERDOWN step was accidentally moved in relation to some MD
steps, so shuffle the MD steps to reinforce the order.
jsg [Tue, 15 Feb 2022 00:27:11 +0000 (00:27 +0000)]
fix mask in hppa inst_trap_return()
inst_trap_return() was checking for rfir by masking with 0xfc001fc0
which made it impossible to match rfir (0xca0)
rfi: return from interruption
00 rv rv rv 60 0
6 5 5 3 8 5
rfir: return from interruption and restore
00 rv rv rv 65 0
6 5 5 3 8 5
from "PA-RISC 1.1 Architecture and Instruction Set Reference Manual"
where rv indicates reserved bits
change the mask to only mask out reserved bits and check for rfi
in addition to rfir
ok miod@
kettenis [Mon, 14 Feb 2022 23:42:37 +0000 (23:42 +0000)]
Make sure we call acpi_resume_cpu() with the right ACPI state. This was
accidentally broken with the reorganization of the suspend/resume code that
was committed a few days ago.
ok deraadt@
jsg [Mon, 14 Feb 2022 23:20:46 +0000 (23:20 +0000)]
don't deref pointer before test it is safe
ok miod@
jsg [Mon, 14 Feb 2022 23:02:15 +0000 (23:02 +0000)]
fix cmalo_media_change() test
ok claudio@ deraadt@
krw [Mon, 14 Feb 2022 22:38:24 +0000 (22:38 +0000)]
Use fdisk -b to create the desired 1MB MSDOS boot partition
rather than relying on /usr/mdec/mbr.
Tested by tobhe@
kn [Mon, 14 Feb 2022 21:11:02 +0000 (21:11 +0000)]
Unwire with map lock held
This is an updated version of uvm_map.c r1.283 "Unwire with map lock held".
The previous version introduced a use-after-free by not unlocking vm_map
locks in uvm_map_teardown(), resulting in dangling references on the
reaper's lock list (thanks visa!).
Lock and unlock the map in around uvm_map_teardown() instead.
This code path holds the last reference, hence the lock isn't strictly
needed except for satisfying upcoming locking assertions.
Tested on amd64, arm64, i386, macppc, octeon, sparc64.
This time also with WITNESS enabled (except on sparc64 which builds but does
not boot with WITNESS; this is a known issue).
OK mpi visa
deraadt [Mon, 14 Feb 2022 19:25:03 +0000 (19:25 +0000)]
acpi_indicator() can take the softc*, instead of void*
rob [Mon, 14 Feb 2022 18:09:08 +0000 (18:09 +0000)]
Revert change to ps for displaying chrooted process.
Ok deraadt
krw [Mon, 14 Feb 2022 16:10:45 +0000 (16:10 +0000)]
Whitespace nit.
kettenis [Mon, 14 Feb 2022 14:57:00 +0000 (14:57 +0000)]
Use iic_is_compatible(9) and use the more generic compatible string to
match the device.
ok visa@
kettenis [Mon, 14 Feb 2022 14:55:53 +0000 (14:55 +0000)]
Pass the entire compatible property for devices attaching to the iic(4) bus.
ok visa@
job [Mon, 14 Feb 2022 14:47:49 +0000 (14:47 +0000)]
Draw a bit more attention to transport switches between RRDP, RSYNC, and cache
OK claudio@ tb@
job [Mon, 14 Feb 2022 14:46:16 +0000 (14:46 +0000)]
Name the maximum number of deltas permitted in a RRDP Notification file
OK claudio@ tb@
aoyama [Mon, 14 Feb 2022 13:03:52 +0000 (13:03 +0000)]
Delete incomplete implementation of bus_space_{alloc,free} in bus.h.
This prevents errors on luna88k in src/usr.sbin/fw_update/pattern.c
which includes dev/pci/pcivar.h, which includes machine/bus.h.
Suggested by miod@
jsg [Mon, 14 Feb 2022 12:54:43 +0000 (12:54 +0000)]
OF_getproplen() returns -1 on error, so don't store result in size_t
spotted by and ok tb@, ok tobhe@
claudio [Mon, 14 Feb 2022 11:26:05 +0000 (11:26 +0000)]
Introduce a signal context that is used to pass signal related information
from cursig() to postsig() or the caller itself. This will simplify locking.
Also alter sigactsfree() a bit and move it into process_zap() so ps_sigacts
is always a valid pointer.
OK semarie@
espie [Mon, 14 Feb 2022 10:34:36 +0000 (10:34 +0000)]
make db locking self-contained
nicm [Mon, 14 Feb 2022 09:10:48 +0000 (09:10 +0000)]
Do not return error with -q, GitHub issue 3065.
kevlo [Mon, 14 Feb 2022 08:14:19 +0000 (08:14 +0000)]
Add the EM_RISCV value to match <sys/exec_elf.h>
ok miod@
visa [Mon, 14 Feb 2022 06:39:19 +0000 (06:39 +0000)]
pty test is not expected to fail any longer.
jsg [Mon, 14 Feb 2022 05:30:53 +0000 (05:30 +0000)]
reprint device name in attach error paths if needed
jsg [Mon, 14 Feb 2022 05:20:03 +0000 (05:20 +0000)]
check return value of drm_attach_pci()
adapted from a diff by Ted Bullock
jsg [Mon, 14 Feb 2022 04:55:55 +0000 (04:55 +0000)]
remove unused code
jsg [Mon, 14 Feb 2022 04:41:13 +0000 (04:41 +0000)]
no need for ifndef __sparc64__
from Ted Bullock who tested on sparc64 with xvr-100
dlg [Mon, 14 Feb 2022 04:33:18 +0000 (04:33 +0000)]
update sbchecklowmem() to better detect actual mbuf memory usage.
previously sbchecklowmem() (and sonewconn()) would look at the mbuf
and mbuf cluster pools to see if they were approaching their hard
limits. based on how many mbufs/clusters were allocated against the
limits, socket operations would start to fail with ENOBUFS until
utilisation went down.
mbufs and clusters have changed a lot since then though. there are
now many mbuf cluster pools, not just one for 2k clusters. because
of this the mbuf layer now limits the amount of memory all the mbuf
pools can allocate backend pages from rather than limit the individual
pools. this means sbchecklowmem() ends up looking at the default
pool hard limit, which is UINT_MAX, which in turn means means
sbchecklowmem() probably never applies backpressure. this is made
worse on multiprocessor systems where per cpu caches of mbuf and
cluster pool items are enabled because the number of in use pool
items is distorted by the cpu caches.
this switches sbchecklowmem to looking at the page allocations made
by all the pools instead. the big benefit of this is that the page
allocations are much more representative of the overall mbuf memory
usage in the system. the downside is is that the backend page
allocation accounting does not see idle memory held by pools. pools
cannot release partially free pages to the page backend (obviously),
and pools cache idle items to avoid thrashing on the backend page
allocator. this means the page allocation level is higher than the
memory used by actual in-flight mbufs.
however, this can also be a benefit. the backend page allocation is a
kind of smoothed out "trend" line. mbuf utilisation over short periods
can be extremely bursty because of things like rx ring dequeue and fill
cycles, or large socket sends. if you're trying to grow socket
buffers while these things are happening, luck becomes an important
factor in whether it will work or not. because pools cache idle items,
the backend page utilisation better represents the overall trend
of activity in the system and will give more consistent behaviour here.
this diff is deliberately simple. we're basically going from "no
limits" to "some sort of limit" for sockets again, so keeping the
code simple means it should be easy to understand and tweak in the
future.
ok djm@ visa@ claudio@
jsg [Mon, 14 Feb 2022 03:51:42 +0000 (03:51 +0000)]
change a NULL pa_memex test after use to an earlier KASSERT
from Ted Bullock
guenther [Mon, 14 Feb 2022 03:38:59 +0000 (03:38 +0000)]
Mark all the rpc 'ops' vectors, for auth, client, service, and xdr,
as const, moving them from .data to .data.rel.ro. The other BSDs
did this a long time ago; NetBSD did a chunk in 1998, which is long
enough I didn't bother to get exact dates for others.
ok deraadt@ millert@
jsg [Mon, 14 Feb 2022 02:57:21 +0000 (02:57 +0000)]
test the correct pointer for acpi_intr_establish() result
ok deraadt@ gnezdo@
jsg [Mon, 14 Feb 2022 00:53:40 +0000 (00:53 +0000)]
OF_getproplen() returns -1 on error, so don't store result in size_t
ok kettenis@
bluhm [Sun, 13 Feb 2022 23:11:10 +0000 (23:11 +0000)]
The length value in bpf_movein() is casted to from size_t to u_int
and then rounded before checking. Put the same check before the
calculations to avoid overflow.
Reported-by: syzbot+6f29d23eca959c5a9705@syzkaller.appspotmail.com
OK claudio@
jmc [Sun, 13 Feb 2022 21:27:51 +0000 (21:27 +0000)]
- for -a, get the brackets right
- for -l and -m remove the brackets
- note that -m whilst charging now displays estimated recharge time
- A/C -> AC
from jan stary;
tweaked a little by myself
otto [Sun, 13 Feb 2022 20:02:30 +0000 (20:02 +0000)]
Handle the case in freechunks where not a single allocation has happened.
ok deraadt@
sthen [Sun, 13 Feb 2022 19:37:41 +0000 (19:37 +0000)]
sync
mlarkin [Sun, 13 Feb 2022 19:15:09 +0000 (19:15 +0000)]
remove stray tab whitespace. no code change.
tobhe [Sun, 13 Feb 2022 16:44:50 +0000 (16:44 +0000)]
Fix return value check of OF_getproplen(). If "reset-gpios" is not found
in the device tree -1 is returned, causing a panic in the following malloc
call.
ok kn@ patrick@ kettenis@
deraadt [Sun, 13 Feb 2022 15:56:55 +0000 (15:56 +0000)]
Move some MI pieces out of suspend_mp/resume_mp
ok kettenis
kettenis [Sun, 13 Feb 2022 15:54:07 +0000 (15:54 +0000)]
The Apple M1 SoC has two mechanism for doing IPIs. The first method uses
the interrupt controller, the second method used implementation-specific
CPU registers. The M1 Pro/Max SoCs use a modified interrupt controller
that no longer provides the first method. So switch to the second method
in preparation for adding M1 Pro/Max support to OpenBSD. As a bonus this
mechanism seems to be faster (which may be why Apple calls these "fast"
IPIs).
ok patrick@
visa [Sun, 13 Feb 2022 13:05:51 +0000 (13:05 +0000)]
Use knote_modify() and knote_process() in obvious places.
visa [Sun, 13 Feb 2022 13:03:02 +0000 (13:03 +0000)]
Add helper functions for f_modify and f_process to condense code
These new functions, knote_modify() and knote_process(), implement
the logic that is common to most f_modify and f_process instances.
The code is inlined so as to not add yet another call frame on the
already towering stack of kqueue functions. Also, the _fn versions
allow direct calling of an event function when there is only one
filter type to handle.
visa [Sun, 13 Feb 2022 12:58:46 +0000 (12:58 +0000)]
Rename knote_modify() to knote_assign()
This avoids verb overlap with f_modify.
mbuhl [Sun, 13 Feb 2022 12:26:54 +0000 (12:26 +0000)]
SKEEYSEED -> SKEYSEED
kettenis [Sun, 13 Feb 2022 12:02:21 +0000 (12:02 +0000)]
Some PCIe devices on the M1 and M1 Pro/Max need to be explicitly powered on
(most notably the WiFi chip). This is done through a GPIO controlled by
the SMC. Add support for this and make sure we wait for the 100ms required
by the PCIe standard when we use that GPIO to turn on power. This makes sure
these devices are available even if U-Boot doesn't turn them on for us.
ok patrick@
kettenis [Sun, 13 Feb 2022 11:58:53 +0000 (11:58 +0000)]
Add basic GPIO support.
ok patrick@
ajacoutot [Sun, 13 Feb 2022 00:02:16 +0000 (00:02 +0000)]
Add /usr/local/share/examples/login.conf.d to the default hierarchy.
ok sthen@ espie@ deraadt@
eric [Sat, 12 Feb 2022 18:22:04 +0000 (18:22 +0000)]
use new libtls signer api
ok tb@
deraadt [Sat, 12 Feb 2022 16:26:57 +0000 (16:26 +0000)]
wsdisplay checks are no longer needed here
deraadt [Sat, 12 Feb 2022 16:25:42 +0000 (16:25 +0000)]
wsdisplay checks are no longer needed here
deraadt [Sat, 12 Feb 2022 16:22:40 +0000 (16:22 +0000)]
add MP stubs
deraadt [Sat, 12 Feb 2022 16:22:03 +0000 (16:22 +0000)]
for non-SUSPEND kernel, put the ioctl pieces entirely inside #ifdef
visa [Sat, 12 Feb 2022 14:07:26 +0000 (14:07 +0000)]
Reduce code duplication in pipe event filters
Use the f_event callback for checking event state within the pipe
event filters. This enables the same f_modify and f_process functions
to handle the different filter types.
OK anton@
jsing [Sat, 12 Feb 2022 12:42:19 +0000 (12:42 +0000)]
Add test coverage for OBJ_txt2obj()/OBJ_obj2txt() with no_name == 0.
espie [Sat, 12 Feb 2022 09:46:19 +0000 (09:46 +0000)]
switch to using the common code in BaseState
(BaseState is a "light weight" base class for State, that does the
common stuff like printing out stuff pkg-style, or exec'ing commands
and reporting exit/signal status, but without needing to create an
object--necessarily--or without the extra option handling code.
jsing [Sat, 12 Feb 2022 03:07:24 +0000 (03:07 +0000)]
Avoid potential single byte overread in asn1_parse2().
A fix for this was previously commited in r1.32, however while this added
a bounds check the logic means we still fall through and perform the
overread. Fix the logic such that we only log the error if the bounds check
fails. While here, flip the test around such that we check for validity then
print (which is more readable and matches earlier code).
ok inoguchi@ tb@
jsing [Sat, 12 Feb 2022 03:02:47 +0000 (03:02 +0000)]
Revise for OBJ_obj2txt() OID arc limits.
jsing [Sat, 12 Feb 2022 03:01:59 +0000 (03:01 +0000)]
Limit OID text conversion to 64 bits per arc.
The current implementation uses an unsigned long, then switches to BN once
the arc exceeds its size. However, the complexity of BN_bn2dec() is
quadratic in the length of number being converted. This means that OIDs
with excessively large arcs take a lot of computation to convert to text.
While the X.660 specification states that arcs are unbounded, in reality
they are not overly large numbers - 640K^W64 bits ought to be enough for
any arc. Remove BN entirely, switch from unsigned long to uin64_t and fail
if an arc exceeds this size.
Identified via oss-fuzz timeouts - should fix #41028 and #44372.
ok tb@
tb [Fri, 11 Feb 2022 17:41:55 +0000 (17:41 +0000)]
Fix length check of IP addresses for name constraints
An IP address in a name constraint is actually an IP address concatenated
with a netmask, so it is twice as long as usual.
This fixes a third bug introduced in r1.3 and reported by Volker Schlecht
ok jsing
tb [Fri, 11 Feb 2022 17:39:36 +0000 (17:39 +0000)]
Add missing error check for a2i_GENERAL_NAME()
Fixes a segfault reported by Volker Schlecht.
ok jsing
tb [Fri, 11 Feb 2022 17:37:54 +0000 (17:37 +0000)]
Fix a double free in v2i_NAME_CONSTRAINTS()
a2i_GENERAL_NAME() modifies and returns the out argument that was
passed in unless out == NULL, in which case it returns something
freshly allocated. Thus, in v2i_GENERAL_NAME_ex() we must only free
ret if out == NULL so v2i_NAME_CONSTRAINTS() can free correctly.
Issue reported by Volker Schlecht
ok jsing
sthen [Fri, 11 Feb 2022 17:19:59 +0000 (17:19 +0000)]
adjust bsd.port.mk(5) manual for change to login.conf.d file handling
jmc [Fri, 11 Feb 2022 17:09:39 +0000 (17:09 +0000)]
space needed between arg and punctuation;
jsing [Fri, 11 Feb 2022 16:39:50 +0000 (16:39 +0000)]
Test OBJ_obj2nid() with an undefined (NID_undef) object.
jsing [Fri, 11 Feb 2022 16:39:16 +0000 (16:39 +0000)]
Make OBJ_obj2nid() work correctly with NID_undef.
Currently OBJ_obj2nid() with NID_undef returns NID_ccitt - this is due to
doing a lookup on an empty value and having NID_undef conflict with an
uninitialised NID value.
Somewhat based on OpenSSL
0fb99904809.
ok tb@
jsing [Fri, 11 Feb 2022 16:34:23 +0000 (16:34 +0000)]
Add initial regress for objects.
cheloha [Fri, 11 Feb 2022 16:09:21 +0000 (16:09 +0000)]
tr(1): rename lookup tables
The lookup tables in tr(1) are called "string1" and "string2". This is
misleading:
- They aren't C strings, they are lookup tables.
- The names "string1" and "string2" don't hint at what their contents
actually mean. The meaning of a given table changes with tr(1)'s
operating mode. There are five different modes.
It would be considerably easier to see what is happening at a glance
if the tables were named for their corresponding byte transformation.
So instead of two tables named "string1" and "string2" we'll have
three tables named "delete", "squeeze", and "translate". In addition,
the "string" passed to the setup() function will be called a "table".
With this patch the code in main() is way easier to understand.
Hopefully this makes subsequent patches easier to review.
Thread: https://marc.info/?l=openbsd-tech&m=
164355980600901&w=2
No complaints on tech@ after over a week.
espie [Fri, 11 Feb 2022 15:49:25 +0000 (15:49 +0000)]
be more consistent in packing-list usage
remove a bunch of stuff that's redundant with update-plist(1) in the
update-plist target description, but instead also points at SUBST_VARS
for the default handling of substituted variables by update-plist.
espie [Fri, 11 Feb 2022 15:36:40 +0000 (15:36 +0000)]
use packing-list consistently
millert [Fri, 11 Feb 2022 15:11:35 +0000 (15:11 +0000)]
Fix typo in comment for f_favail.
From Alf Schlichting
florian [Fri, 11 Feb 2022 12:58:18 +0000 (12:58 +0000)]
It's fw_update's job to say what went wrong.
input jca
OK kmos sthen deraadt kn
sthen [Fri, 11 Feb 2022 12:42:24 +0000 (12:42 +0000)]
describe LOGIN_CONF, help from espie
kn [Fri, 11 Feb 2022 12:06:48 +0000 (12:06 +0000)]
Backout previous "Unwire with map lock held" (commitid: eQBvWUwShD91dN9Z)
WITNESS builds broke^W^Wkernels panic on boot as reported by anton and bluhm.
Booting bsd.mp in single-user mode inside VMM shows:
root on sd0a (
5f9e458ed30b39ab.a) swap on sd0b dump on sd0b
Enter pathname of shell or RETURN for sh:
witness: lock order reversal:
1st 0xfffffd801f8ce468 vmmaplk (&map->lock)
2nd 0xfffffd801b8162c0 inode (&ip->i_lock)
lock order "&ip->i_lock"(rrwlock) -> "&map->lock"(rwlock) first seen at:
#0 rw_enter_read+0x38
#1 uvmfault_lookup+0x8a
#2 uvm_fault_check+0x32
#3 uvm_fault+0xfb
#4 kpageflttrap+0x12c
#5 kerntrap+0x91
#6 alltraps_kern_meltdown+0x7b
#7 copyout+0x53
#8 ffs_read+0x1f6
#9 VOP_READ+0x41
#10 vn_rdwr+0xa1
#11 vmcmd_map_readvn+0xa0
#12 exec_process_vmcmds+0x88
#13 sys_execve+0x732
#14 start_init+0x26f
#15 proc_trampoline+0x1c
lock order data w1 -> w2 missing
# exit
kernel: protection fault trap, code=0
Stopped at witness_checkorder+0x312: movl 0x10(%r14),%ecx
gkoehler reported faults on poisened addresses on macppc dual G5.
kn [Fri, 11 Feb 2022 09:25:04 +0000 (09:25 +0000)]
Backout previous "Assert vm map locks" (commitid: sRNBfzX2dJrxFDmb)
WITNESS builds broke as reported by anton and bluhm:
root on sd0a (
5ec49b3ad23eb2d4.a) swap on sd0b dump on sd0b
kernel: protection fault trap, code=0
Stopped at witness_checkorder+0x4ec: movl 0x10(%r12),%ecx
https://syzkaller.appspot.com/bug?id=
be02b290a93c648986c35370a271aad4135a5044
https://syzkaller.appspot.com/text?tag=CrashLog&x=
136e9aa4700000
visa [Fri, 11 Feb 2022 07:28:29 +0000 (07:28 +0000)]
Replace manual !klist_empty()+knote() with KNOTE().
OK mpi@
visa [Fri, 11 Feb 2022 07:27:07 +0000 (07:27 +0000)]
Check klist emptiness instead of NULL pointer in KNOTE()
All callers of KNOTE() supply a non-NULL klist argument. Replace the
NULL pointer check with klist emptiness check as a small optimization.
OK mpi@
visa [Fri, 11 Feb 2022 07:25:50 +0000 (07:25 +0000)]
Inline klist_empty() for more economic machine code.
OK mpi@
deraadt [Fri, 11 Feb 2022 01:55:12 +0000 (01:55 +0000)]
the sleep_clocks() hook is not needed because the architectures which
need to do this can do it a few moments later in a different hook
afresh1 [Fri, 11 Feb 2022 00:46:58 +0000 (00:46 +0000)]
Lock the pkg db while running
At least when not running in the installer.
Suggestions from espie@
Works for me sthen@
fine deraadt@