beck [Wed, 5 Jul 2023 15:13:28 +0000 (15:13 +0000)]
Make softdep mounts a no-op
Softdep is a significant impediment to progressing in the vfs layer
so we plan to get it out of the way. It is too clever for us to
continue maintaining as it is.
ok kettenis@ kn@ tobhe@ and most of the g2k23 room except bluhm@
stsp [Wed, 5 Jul 2023 15:07:28 +0000 (15:07 +0000)]
Add iwm/iwx background scan task to the queue which it will be deleted from.
The corresponding task_add and task_del calls were operating on different
queues by mistake. Background scan tasks should now get cancelled properly
during driver state transitions.
ok mvs@
tb [Wed, 5 Jul 2023 14:41:18 +0000 (14:41 +0000)]
One more ECDSAerror goes.
tb [Wed, 5 Jul 2023 14:39:05 +0000 (14:39 +0000)]
ECDHerror() and ECDSAerror will go away
Move some trivial ones to ECerror().
discussed with jsing
bcook [Wed, 5 Jul 2023 13:06:06 +0000 (13:06 +0000)]
Adjust the public declaration of OpenSSLDie to use a portable
method of indicating that the function does not return.
ok tb@
tb [Wed, 5 Jul 2023 13:01:44 +0000 (13:01 +0000)]
Drop an incorrect part from a comment
kn [Wed, 5 Jul 2023 12:58:55 +0000 (12:58 +0000)]
Drop kernel lock before panic to avoid WITNESS report during fault
holding a spinlock, eg. malloc's malloc_mutex in "Data modified on freelist ..."
triggers "acquiring blockable sleep lock with spinlock or critical section held"
since kpageflttrap() grabs the kernel lock before fault() to serialise multiple
threds/faults avoid interleaved console text.
But fault() immediately sets the per-CPU panic string, so the kernel lock does
not really help here.
Use 'show panic' to recover from garbled console text if need be, as usual.
The i386 equivalent does not use the kernel lock, either.
OK bluhm kettenis
tb [Wed, 5 Jul 2023 12:56:52 +0000 (12:56 +0000)]
Missing . in comment
tb [Wed, 5 Jul 2023 12:54:46 +0000 (12:54 +0000)]
Fix #includes
tb [Wed, 5 Jul 2023 12:51:11 +0000 (12:51 +0000)]
Remove local prototypes for public API (?!)
tb [Wed, 5 Jul 2023 12:49:42 +0000 (12:49 +0000)]
Improve BN_bn2bin() error check for readability
tb [Wed, 5 Jul 2023 12:31:14 +0000 (12:31 +0000)]
Merge ECDH code that will stay into ecdh.c
tb [Wed, 5 Jul 2023 12:27:36 +0000 (12:27 +0000)]
Move ECDSA_size() to a more sensible place in this file
tb [Wed, 5 Jul 2023 12:18:21 +0000 (12:18 +0000)]
Merge ECDSA code that will stay into ecdsa.c
discussed with jsing
sashan [Wed, 5 Jul 2023 12:00:01 +0000 (12:00 +0000)]
Document DIOCXEND in pf(4) manpage.
With many improvements from jmc@ and kn@
OK jmc@, kn@ deraadt@
tb [Wed, 5 Jul 2023 11:37:45 +0000 (11:37 +0000)]
Rename ecs_local.h into ecdsa_local.h
tb [Wed, 5 Jul 2023 11:28:50 +0000 (11:28 +0000)]
Make variables in prototypes match function declarations
kettenis [Wed, 5 Jul 2023 11:07:36 +0000 (11:07 +0000)]
Implement support for the GPIOs on the JH7110. Makes it possible to reboot
the VisionFive 2 from OpenBSD.
ok jsing@
jsg [Wed, 5 Jul 2023 10:41:57 +0000 (10:41 +0000)]
drm/radeon: fix race condition UAF in radeon_gem_set_domain_ioctl
From Min Li
1af1cd7be370b08694d8752c97325fe51fdab6aa in linux-6.1.y/6.1.36
982b173a6c6d9472730c3116051977e05d17c8c5 in mainline linux
jsg [Wed, 5 Jul 2023 10:40:14 +0000 (10:40 +0000)]
drm/amd/display: fix the system hang while disable PSR
From Tom Chung
9bcac453890bf2c0ab5a7cefb407c0a9d6cbc4cb in linux-6.1.y/6.1.36
ea2062dd1f0384ae1b136d333ee4ced15bedae38 in mainline linux
jsg [Wed, 5 Jul 2023 10:38:11 +0000 (10:38 +0000)]
drm/amd/display: Add wrapper to call planes and stream update
From Rodrigo Siqueira
e538342002cbe468224f71b7ae116586e55c1134 in linux-6.1.y/6.1.36
81f743a08f3b214638aa389e252ae5e6c3592e7c in mainline linux
jsg [Wed, 5 Jul 2023 10:36:28 +0000 (10:36 +0000)]
drm/amd/display: Use dc_update_planes_and_stream
From Rodrigo Siqueira
8d855bc67630fa2b17855d85de61b9cd4300e3ad in linux-6.1.y/6.1.36
f7511289821ffccc07579406d6ab520aa11049f5 in mainline linux
kettenis [Wed, 5 Jul 2023 09:25:55 +0000 (09:25 +0000)]
They may have fixed the cache coherency problems in the JH7110, but it is
still criplled as the SD/MMC controllers only do 32-bit DMA.
ok jsing@
tb [Wed, 5 Jul 2023 08:39:40 +0000 (08:39 +0000)]
Drop useless ossl_ prefixes
discussed with jsing
tobhe [Wed, 5 Jul 2023 08:26:56 +0000 (08:26 +0000)]
Suspend from suspend_taskq in apmioctl.
ok kettenis@
jsg [Wed, 5 Jul 2023 08:15:34 +0000 (08:15 +0000)]
remove unused armv7 max_processes global
ok miod@
krw [Wed, 5 Jul 2023 08:05:04 +0000 (08:05 +0000)]
Simplify and clarify the description of auto allocation.
Tweaks and ok jmc@
krw [Wed, 5 Jul 2023 07:28:10 +0000 (07:28 +0000)]
Restore (R)esize functionality to sparc64 disklabel(8).
Empty or FS_UNUSED partitions don't intrude into the OpenBSD area
of a disk and therefore should not disable partition resizing.
Encountered by claudio@
jsg [Wed, 5 Jul 2023 01:21:51 +0000 (01:21 +0000)]
fw_update is in section 8 not 1
mvs [Tue, 4 Jul 2023 22:28:24 +0000 (22:28 +0000)]
Introduce SBL_WAIT and SBL_NOINTR sbwait() flags.
This refactoring is another step to make standalone socket buffers
locking. sblock() uses M_WAITOK and M_NOWAIT flags passed as the third
argument together with the SB_NOINTR flag on the `sb_flags' to control
sleep behaviour. To perform uninterruptible acquisition, SB_NOINTR flag
should be set before sblock() call. `sb_flags' modification requires to
hold solock() around sblock()/sbunlock() that makes standalone call
impossible.
Also `sb_flags' modifications outside sblock()/sbunlock() makes
uninterruptible acquisition code huge enough. This time only sorflush()
does this (and forgets to restore SB_NOINTR flag, so shutdown(SHUT_RDWR)
call permanently modifies socket locking behaviour) and this looks not
the big problem. But with the standalone socket buffer locking it will
be many such places, so this huge construction is unwanted.
Introduce new SBL_NOINTR flag passed as third sblock() argument. The
sblock() acquisition will be uninterruptible when existing SB_NOINTR
flag is set on `sb_flags' or SBL_NOINTR was passed.
The M_WAITOK and M_NOWAIT flags belongs to malloc(9). It has no M_NOINTR
flag and there is no reason to introduce it. So for consistency reasons
introduce new SBL_WAIT and use it together with SBL_NOINTR instead of
M_WAITOK and M_NOINTR respectively.
ok bluhm
anton [Tue, 4 Jul 2023 19:01:46 +0000 (19:01 +0000)]
favor hw.ncpuonline; ok bluhm@
cheloha [Tue, 4 Jul 2023 17:29:32 +0000 (17:29 +0000)]
amd64: cpu_info: rearrange members, keep ci_mds_tmp 32-byte aligned
ci_mds_tmp needs to be 32-byte aligned, otherwise we trip a CTASSERT
in amd64/cpu.c and break kernel compilation. However, ci_mds_tmp's
32-byte alignment is at risk: the size of schedstate_percpu is about
to change.
Move ci_curproc and ci_schedstate up just after ci_mds_buf. This puts
ci_mds_tmp at page offset 64 with no structs ahead of it in cpu_info.
With this arrangement it should remain 32-byte aligned without much
effort.
With input from guenther@.
ok guenther@
tb [Tue, 4 Jul 2023 15:09:31 +0000 (15:09 +0000)]
Avoid outputting invalid signatures
The caller can provide an r which will be added to the ECDSA_SIG unchecked.
This can happen via ECDSA_{,do_}sign_ex() or ECDSA_sign_setup() or else via
a custom sign_sig() handler. Therefore add a check that it is in the bounds
required.
Since k was long thrown away, there's no way to check kinv, so it needs to
be trusted. Misdesigned APIs that will output garbage everywhere...
ok jsing
tb [Tue, 4 Jul 2023 14:59:32 +0000 (14:59 +0000)]
Clean up ECDSA verification
Use variable names that correspond more closely to the standard. Use an
additional variable for s^-1 for readability. Annotate the code with
the corresponding steps from FIPS 186-5.
ok jsing
tb [Tue, 4 Jul 2023 14:57:05 +0000 (14:57 +0000)]
ECDSA signing: annotate code with steps corresponding to FIPS 185-6.
ok jsing
espie [Tue, 4 Jul 2023 14:41:26 +0000 (14:41 +0000)]
use local assignment to $ENV{TZ} like we do for signal handlers
patrick [Tue, 4 Jul 2023 14:32:21 +0000 (14:32 +0000)]
Instead of waiting for the inbound item to show up in the interrupt
handler, allocate it ourselves. The firmware doesn't seem to mind
that it's us doing the allocation. This fixes a splassert, because
the code that retrieves the item reaches through the shared memory
driver, which acquires the hardware mutex, which in turn mallocs.
ok kettenis@
sashan [Tue, 4 Jul 2023 14:23:38 +0000 (14:23 +0000)]
This diff limits the number of transactions/tickets
pf_open_trans() can issue for each clone of /dev/pf
to 512. The pf_open_trans() is currently being used
by DIOCGETRULES ioctl(2). The limit avoids processes
to consume all kernel memory by asking DIOCGETRULES
for more tickets. If DIOCGETRULES hits the limit, then
the application will see EBUSY error.
This diff was fine tuned with feedback from cluadio@,
deraadt@ and kn@.
OK kn@
espie [Tue, 4 Jul 2023 14:08:30 +0000 (14:08 +0000)]
stop saying the format is "simple" there are so many annotations.
"straightforward" is a better description at this point.
espie [Tue, 4 Jul 2023 14:03:16 +0000 (14:03 +0000)]
add some more fluff to documentation
espie [Tue, 4 Jul 2023 14:00:18 +0000 (14:00 +0000)]
explain why NO_TS_IN_PLIST exists
espie [Tue, 4 Jul 2023 13:58:01 +0000 (13:58 +0000)]
fix some very minor syntactic nits
jan [Tue, 4 Jul 2023 13:37:47 +0000 (13:37 +0000)]
Check for interface type ethernet before call ether_brport_isset()
Pointed out by bluhm.
ok bluhm@
kettenis [Tue, 4 Jul 2023 13:04:08 +0000 (13:04 +0000)]
enable dwqe(4)
kettenis [Tue, 4 Jul 2023 12:58:42 +0000 (12:58 +0000)]
Add StarFive JH7110 support.
Based on an initial diff from jsing@
ok jsing@, patrick@
kettenis [Tue, 4 Jul 2023 12:48:42 +0000 (12:48 +0000)]
Handle rx error.
ok jsing@, patrick@
jasper [Tue, 4 Jul 2023 12:27:44 +0000 (12:27 +0000)]
consistenly name instruction defines
jasper [Tue, 4 Jul 2023 12:20:10 +0000 (12:20 +0000)]
Unbreak kprobes. Since LLVM emits endbr64 instructions for IBT, dt could no
longer identify function boundaries and as such no kprobes were found anymore.
Adjust the parser accordingly.
ok mpi@
sashan [Tue, 4 Jul 2023 11:34:19 +0000 (11:34 +0000)]
The recent change to DIOCGETRULE allows applications which
periodically read rules from pf(4) to consume all kernel
memory. The bug has been discovered and root caused by florian@.
In this particular case it was snmpd(8) what ate all kernel
memory.
This commit introduces DIOCXEND to pf(4) so applications such
as snmpd(8) and systat(1) to close ticket/transaction when
they are done with fetching the rules. This change also
updates snmpd(8) and systat(1) to use newly introduced
DIOCXEND ioctl(2).
OK claudio@, deraadt@, kn@
bluhm [Tue, 4 Jul 2023 11:27:13 +0000 (11:27 +0000)]
Run Perl tests from our regress tree. It needs a Perl obj directory
with Makefile. To create the test environment, use the
Makefile.bsd-wrapper in /usr/src/gnu/usr.bin/perl and build Perl
there if it does not exist. This allows to test Perl easily and
in a consistent manner. Do not link tests to regress tree due to
this reach around and the combination of build and test.
jsg [Tue, 4 Jul 2023 11:14:00 +0000 (11:14 +0000)]
remove prototypes for removed functions
tb [Tue, 4 Jul 2023 10:53:42 +0000 (10:53 +0000)]
Extract private key and group order in s computation
This pushes a few variables no longer needed in ossl_ecdsa_sign_sig() into
ecdsa_compute_s() separating API logic and pure computation a bit more.
ok beck
bluhm [Tue, 4 Jul 2023 10:48:19 +0000 (10:48 +0000)]
Remove redundant code when calculating checksum.
OK jmatthew@
kettenis [Tue, 4 Jul 2023 10:34:03 +0000 (10:34 +0000)]
Add a bunch of clocks for GMAC0 and GMAC1 on the JH7110 SoC.
Based on an initial diff from jsing@
ok jsing@
tb [Tue, 4 Jul 2023 10:31:57 +0000 (10:31 +0000)]
Use key for the EC_KEY everywhere
tb [Tue, 4 Jul 2023 10:26:47 +0000 (10:26 +0000)]
Some more consistency in variable names
tb [Tue, 4 Jul 2023 10:23:34 +0000 (10:23 +0000)]
Normalize ECDSA_SIG to be sig everywhere
jmatthew [Tue, 4 Jul 2023 10:22:39 +0000 (10:22 +0000)]
Make the hardware counters available on BCM5705 and newer available as
kstats. Some of the hardware counters are already used in
bge_stats_update_regs() to update interface counters and work around
hardware bugs, and all counters reset on read, so to keep things simple
the work is split between that function and bge_kstat_read(), which
reads the rest of the counters.
tested by bluhm@ on BCM5720 (with counters) and BCM5704 (without), and
by me on BCM5720 A0 (with counters and hardware bugs)
ok bluhm@ dlg@
tb [Tue, 4 Jul 2023 10:14:37 +0000 (10:14 +0000)]
Normalize on digest and digest_len rather than dgst dlen dgstlen, etc.
tb [Tue, 4 Jul 2023 10:06:36 +0000 (10:06 +0000)]
Rework ecdsa_prepare_digest()
Make it take an EC_KEY instead of a group order in preparation for further
cleanup. Rename m into e to match the standard better. Also buy some vowels
for jsing.
ok beck jsing
jsg [Tue, 4 Jul 2023 09:47:51 +0000 (09:47 +0000)]
Remove mbuf low watermark vars. Unused since uipc_mbuf.c rev 1.244.
ok kn@ bluhm@
kettenis [Tue, 4 Jul 2023 09:00:24 +0000 (09:00 +0000)]
Add some extra parenthesis for consistency.
stolen from jsing@
tb [Tue, 4 Jul 2023 08:47:01 +0000 (08:47 +0000)]
Plug leak in the ssltest
Removing -tls1 moved some tests from the legacy stack to the TLSv1.3 stack.
On a HRR, the alpn callback would be called twice and allocate the global
twice, thereby leaking. So free it up front.
Joint suffering with bcook and beck
tb [Tue, 4 Jul 2023 07:38:31 +0000 (07:38 +0000)]
Factor the computation of ECDSA s into a function
ossl_ecdsa_sign_sig() is already complicated enough. The math bit is
entirely self contained and does not need to obfuscate control flow
and logic.
with feedback from and ok jsing
espie [Tue, 4 Jul 2023 07:04:11 +0000 (07:04 +0000)]
document the recent addition of "Building from scratch" and what it means.
dlg [Tue, 4 Jul 2023 03:59:21 +0000 (03:59 +0000)]
add support for unix domain sockets to ssh -W
ok djm@ dtucker@
jsg [Tue, 4 Jul 2023 03:56:07 +0000 (03:56 +0000)]
m_reclaim() was removed in uipc_mbuf.c rev 1.195
dlg [Tue, 4 Jul 2023 02:56:11 +0000 (02:56 +0000)]
a little knf, no functional change
jsg [Tue, 4 Jul 2023 02:01:55 +0000 (02:01 +0000)]
remove unused global var
ok sashan@
espie [Mon, 3 Jul 2023 19:12:08 +0000 (19:12 +0000)]
record when pkg_add is actually doing something.
details of rc code to be set in stone later (currently as an option)
anton [Mon, 3 Jul 2023 17:55:51 +0000 (17:55 +0000)]
favor REGRESS_EXPECTED_FAILURES
espie [Mon, 3 Jul 2023 17:01:59 +0000 (17:01 +0000)]
also make things 5.36 if not used
nicm [Mon, 3 Jul 2023 16:47:43 +0000 (16:47 +0000)]
Add support for marking lines with a shell prompt based on the OSC 133
extension, from Munif Tanjim in GitHub issue 3596.
tobhe [Mon, 3 Jul 2023 15:54:07 +0000 (15:54 +0000)]
Enable mouse.tp.mtbuttons for apldcms
ok bru@
kn [Mon, 3 Jul 2023 15:52:51 +0000 (15:52 +0000)]
use consistent queue(9) example for LIST removal; OK bluhm mvs
krw [Mon, 3 Jul 2023 15:27:07 +0000 (15:27 +0000)]
An empty string indicates there is no NXTNUM().
Fixes segfault in -e if you happen to delete everything after
a partition line's ":".
tb [Mon, 3 Jul 2023 14:51:09 +0000 (14:51 +0000)]
sign_sig: drop ckinv
The only reason ckinv exists is to be able to avoid a copy. This copy
leaks some timing info, that will be mitigated in a subsequent step.
It is an unused or at least uncommonly used codepath.
ok jsing
tb [Mon, 3 Jul 2023 13:53:54 +0000 (13:53 +0000)]
Rework the logic in ECDSA sign_sig()
If the caller supplied both kinv and r, we don't loop but rather throw
an undocumented error code that no one uses, which is intended to tell
the caller to run ECDSA_sign_setup() and try again.
Use a boolean that indicates this situation so that the logic becomes
a bit more transparent.
ok jsing
jasper [Mon, 3 Jul 2023 13:51:55 +0000 (13:51 +0000)]
add regress tests for the remainder of the function provided by our uuid.h
jasper [Mon, 3 Jul 2023 13:44:45 +0000 (13:44 +0000)]
add a few tests related to --cflags and --libs in combination with Requires.private.
skipping one (currently failing) test which exhibits the behaviour described
by humm@ljabl.com on bugs@
claudio [Mon, 3 Jul 2023 11:51:27 +0000 (11:51 +0000)]
Use new ibuf API. Use ibuf_data() and ibuf_size() instead of direct
struct access and use ibuf_set_n16() to update the length field.
OK tb@
tobhe [Mon, 3 Jul 2023 11:43:15 +0000 (11:43 +0000)]
Delete some more references to dead policy code.
Fixes -DNAMESPACE
ok tb@
tb [Mon, 3 Jul 2023 11:10:28 +0000 (11:10 +0000)]
sign_sig: test on assignment
tb [Mon, 3 Jul 2023 11:06:28 +0000 (11:06 +0000)]
sign_setup: split another check into two
tobhe [Mon, 3 Jul 2023 10:59:16 +0000 (10:59 +0000)]
typo
nicm [Mon, 3 Jul 2023 10:48:26 +0000 (10:48 +0000)]
Do not risk writing over the end of the buffer when it ends in #
(because strchr \0 will be non-NULL), reported by Robert Morris in
GitHub issue 3610.
claudio [Mon, 3 Jul 2023 10:34:31 +0000 (10:34 +0000)]
Use ibuf_data() and ibuf_size() instead of accessing struct ibuf.
OK tb@
claudio [Mon, 3 Jul 2023 10:34:08 +0000 (10:34 +0000)]
Use ibuf_data() and ibuf_size() instead of accessing struct ibuf.
OK tb@
tb [Mon, 3 Jul 2023 10:21:25 +0000 (10:21 +0000)]
Split range checks for ECDSA r and ECDSA s
requested by jsing
tb [Mon, 3 Jul 2023 10:19:52 +0000 (10:19 +0000)]
Switch a couple of test from ucmp to cmp
This is confusing, as both sides involved should be unsigned. The ec
code is undecided on whether the group order can be negative. It should
never be, so lets see what happen with this slightly stricter check.
discussed with jsing
tb [Mon, 3 Jul 2023 10:16:14 +0000 (10:16 +0000)]
ossl_ecdsa_verify_sig(): simplify range checks
The checks whether r and s lie in the interval [1, order) were a bit
uglier than necessary. Clean this up.
ok beck jsing
tb [Mon, 3 Jul 2023 10:10:58 +0000 (10:10 +0000)]
List variables in a somewhat more sensible order
tb [Mon, 3 Jul 2023 10:09:12 +0000 (10:09 +0000)]
In ossl_ecdsa_verify_sig() use BN_CTX more idiomatically
ok beck jsing
tb [Mon, 3 Jul 2023 10:06:00 +0000 (10:06 +0000)]
Split a bunch of unrelated checks
ok beck jsing
tb [Mon, 3 Jul 2023 10:04:05 +0000 (10:04 +0000)]
Make ossl_ecdsa_verify_sig() single exit
ok beck jsing
tb [Mon, 3 Jul 2023 09:59:20 +0000 (09:59 +0000)]
Switch ossl_ecdsa_verify() to timingsafe_memcmp()
Requested by jsing
tb [Mon, 3 Jul 2023 09:55:42 +0000 (09:55 +0000)]
Streamline ossl_ecdsa_verify()
Make it single exit and use API more idiomatically and some other
cosmetics.
ok beck jsing
claudio [Mon, 3 Jul 2023 09:51:38 +0000 (09:51 +0000)]
Use ibuf_data() instead of direct access to ibuf->buf,
use ibuf_size() instead of direct access to ibuf->wpos,
use ibuf_left() in places where the code checks if there is enough
space left in the ibuf.
OK tb@
claudio [Mon, 3 Jul 2023 09:40:47 +0000 (09:40 +0000)]
Use ibuf_data() instead of directly accessing ibuf->buf.
OK tb@
claudio [Mon, 3 Jul 2023 09:38:08 +0000 (09:38 +0000)]
Use ibuf_data() instead of accessing ibuf->buf directly.
OK tb@