schwarze [Mon, 26 Feb 2018 16:23:43 +0000 (16:23 +0000)]
In ssl.h rev. 1.143 2018/02/22 17:27:07, jsing@ provided
SSL_CTX_get_ciphers(3). Merge the documentation from OpenSSL.
schwarze [Mon, 26 Feb 2018 16:02:14 +0000 (16:02 +0000)]
In ssl.h rev. 1.142 2018/02/22 17:25:18, jsing@ provided
SSL_SESSION_up_ref(3).
Merge the documentation from OpenSSL, tweaked by me.
schwarze [Mon, 26 Feb 2018 15:38:39 +0000 (15:38 +0000)]
In x509.h rev. 1.41 2018/02/22 17:22:02, jsing@ provided
X509_chain_up_ref(3). Merge the documentation from OpenSSL,
tweaked by me.
sthen [Mon, 26 Feb 2018 15:29:20 +0000 (15:29 +0000)]
sync
espie [Mon, 26 Feb 2018 14:58:36 +0000 (14:58 +0000)]
actually have safe do something, stupid typo
espie [Mon, 26 Feb 2018 14:09:27 +0000 (14:09 +0000)]
revert, unneeded
espie [Mon, 26 Feb 2018 14:07:39 +0000 (14:07 +0000)]
forgot the non-interactive case
espie [Mon, 26 Feb 2018 13:53:31 +0000 (13:53 +0000)]
spacing
mpi [Mon, 26 Feb 2018 13:43:51 +0000 (13:43 +0000)]
Serialize access to UID data with a rwlock.
Caller must call uid_release() after making whatever change they needed
on the result of uid_find().
Adapted from guenther@'s proctree diff, ok dlg@, visa@
mpi [Mon, 26 Feb 2018 13:33:25 +0000 (13:33 +0000)]
Fix a TOCTOU race that causes signals to be delivered more than once.
The race is only triggerable if one of the threads of a multi-threaded
program is in the middle of a NOLOCK syscall when a signal is received.
The problem is that `ps_sigact' is shared between threads so its access
should be serialized. In the case of SA_RESETHAND, the handler is reset
when a signal is delivered, so delivering the signal twice would put the
process in an "impossible" state where some threads were stopped and some
were waiting for the others to die.
Serialize signal checking & processing with the KERNEL_LOCK() for now,
and introduce postsig_done() gypped from FreeBSD, to make sure the lock
is held when resetting the handler.
Bug report from espie@, ok visa@
kettenis [Mon, 26 Feb 2018 13:11:48 +0000 (13:11 +0000)]
Fix typo in previous commit.
mpi [Mon, 26 Feb 2018 13:06:49 +0000 (13:06 +0000)]
Support dumping isochronous frames via bpf(4) using the USBPcap protocol.
ok deraadt@, ratchov@
espie [Mon, 26 Feb 2018 13:04:30 +0000 (13:04 +0000)]
- move sane into State.pm
- use it for every parameter to say/print
- extend it for actual printable characters, plus a hack for \n
- add confirmf for passing formats to the confirm
- adjust PkgCheck accordingly
- tweak the files test in PkgCheck to also hunt unknown files under
/var/db/pkg/somepackage
okay sthen@
dtucker [Mon, 26 Feb 2018 12:14:53 +0000 (12:14 +0000)]
Remove unneeded (local) include. ok markus@
nicm [Mon, 26 Feb 2018 08:09:56 +0000 (08:09 +0000)]
C-g for modes too, from Mike Hamrick.
ccardenas [Mon, 26 Feb 2018 06:47:06 +0000 (06:47 +0000)]
regen
ccardenas [Mon, 26 Feb 2018 06:46:10 +0000 (06:46 +0000)]
Add devices found on thinkpad e475 with an AMD A10-9600P
(Bristol Ridge, aka tweaked Carrizo) not to be confused with
Carrizo-L (16h apu).
ok jsg@, mlarkin@
ccardenas [Mon, 26 Feb 2018 06:36:53 +0000 (06:36 +0000)]
vmctl: clarify console error message
Clear up ambiguous error message when attaching to a console fails as
the vm id is not guaranteed to be populated.
ok mlarkin@
dtucker [Mon, 26 Feb 2018 03:56:44 +0000 (03:56 +0000)]
Add $OpenBSD$ markers to xmss files to help keep synced with portable.
ok djm@.
guenther [Mon, 26 Feb 2018 03:12:41 +0000 (03:12 +0000)]
Delete pmap_update_2pg(), the #include of <machine/segments.h>, trailing
whitespace, and the leading underbars of '__inline'
ok mlarkin@
dtucker [Mon, 26 Feb 2018 03:03:05 +0000 (03:03 +0000)]
Add newline at end of file to prevent compiler warnings.
schwarze [Sun, 25 Feb 2018 22:52:48 +0000 (22:52 +0000)]
In x509_vfy.h rev. 1.24 2018/02/22 17:19:31, jsing@ provided
X509_STORE_set_ex_data(3) and X509_STORE_get_ex_data(3).
Document them.
schwarze [Sun, 25 Feb 2018 22:24:18 +0000 (22:24 +0000)]
In x509_vfy.h rev. 1.24 2018/02/22 17:19:31, jsing provided
X509_STORE_get0_objects(3). Merge the documenteation from OpenSSL,
heavily tweaked by me. While here, document six additional public
functions closely related to X509_STORE_set1_param(3) that OpenSSL
lacks documentation for. No Copyright-worthy amount of text remains
in X509_STORE_set1_param.3, so switch to my Copyright and license.
kettenis [Sun, 25 Feb 2018 20:43:33 +0000 (20:43 +0000)]
Add RK805 support.
kettenis [Sun, 25 Feb 2018 20:42:42 +0000 (20:42 +0000)]
Add RK3328 support.
kettenis [Sun, 25 Feb 2018 20:42:13 +0000 (20:42 +0000)]
More RK3328 clocks.
schwarze [Sun, 25 Feb 2018 20:26:51 +0000 (20:26 +0000)]
In x509_vfy.h rev. 1.23 2018/02/22 17:17:09, jsing@ provided
X509_OBJECT_get0_X509(3) and X509_OBJECT_get0_X509_CRL(3).
Since they are undocumented in OpenSSL, write a new manual
page from scratch, also documenting five closely related
public functions that have already been available before.
schwarze [Sun, 25 Feb 2018 17:46:38 +0000 (17:46 +0000)]
In x509_vfy.h rev. 1.22 2018/02/22 17:15:09, jsing@ provided
X509_STORE_up_ref(3). X509_STORE_new(3) and X509_STORE_free(3)
have already been available earlier. Import the documentation from
OpenSSL, adding some precision.
krw [Sun, 25 Feb 2018 17:24:44 +0000 (17:24 +0000)]
Bunch of whitespace fixes.
ok otto@
schwarze [Sun, 25 Feb 2018 16:26:15 +0000 (16:26 +0000)]
In x509_vfy.h rev. 1.21 2018/02/22 17:11:30, jsing@ provided
X509_STORE_CTX_get0_store(3). It is undocumented in OpenSSL,
so write some documentation from scratch.
schwarze [Sun, 25 Feb 2018 16:04:07 +0000 (16:04 +0000)]
In x509_vfy.h rev. 1.21 2018/02/22 17:11:30, jsing@ provided
X509_STORE_CTX_get0_chain(3). Adapt the documentation.
It is absurd that OpenSSL documents the two almost identical functions
X509_STORE_CTX_get0_chain(3) and X509_STORE_CTX_get1_chain(3) in
two different manual pages, with quite different wordings, and without
even referencing each other. It is very obvious that they have
lost their way in their own mire of functions.
espie [Sun, 25 Feb 2018 14:47:21 +0000 (14:47 +0000)]
much better checks
- add an explicit dir thingy so that we don't look for /var/db/pkg/+CONTENTS
- actually prepend dir name to info name, oops
- also check for files/dirs that do not belong to root:wheel or have too
lax permissions.
espie [Sun, 25 Feb 2018 14:20:39 +0000 (14:20 +0000)]
add explicit permission checks.
pkg_check is most often run as root. In case of a garbled file systems
it won't notice when permissions are completely wrong.
espie [Sun, 25 Feb 2018 14:19:26 +0000 (14:19 +0000)]
sprinkle quite a few more calls to safe
kettenis [Sun, 25 Feb 2018 13:26:44 +0000 (13:26 +0000)]
RK3328 support.
kettenis [Sun, 25 Feb 2018 13:25:57 +0000 (13:25 +0000)]
Initial RK3328 clocks.
stsp [Sun, 25 Feb 2018 12:40:51 +0000 (12:40 +0000)]
Don't pass our configured SSID to iwm(4) firmware during a background scan.
Apparently this can cause a firmware crash during a TX command on 7265 devices.
Why this happens is unclear.
Problem reported and workaround tested by trondd on bugs@
I have verified that hidden SSID APs still work, though we won't be able to
seamlessly roam between them anymore. Seems like a fair trade-off for now.
stsp [Sun, 25 Feb 2018 12:40:06 +0000 (12:40 +0000)]
My previous commit to iwn(4) broke the scan loop.
The problem happened if we didn't find an AP to connect to after one scan
iteration. The net80211 stack then performs a SCAN -> SCAN transition to
kick off another scan, but the driver treated this transition as a no-op
and remained in SCAN state doing nothing.
To fix this, introduce a flag which keeps track of whether a firmware
scan command is in progress, and start another scan during a SCAN->SCAN
transition if no scan is in progress. Matches what iwm(4) does.
Note that previously (i.e. in 6.2), iwn(4) would always try to start
a new scan regardless of what the firmware was currently doing.
Problem noticed by myself and also by deraadt@
test & ok tb@
tb [Sun, 25 Feb 2018 12:14:59 +0000 (12:14 +0000)]
sync
schwarze [Sun, 25 Feb 2018 10:53:16 +0000 (10:53 +0000)]
In x509.h rev. 1.37 2018/02/22 17:01:44, jsing@ provided
X509_CRL_set1_lastUpdate(3) and X509_CRL_set1_nextUpdate(3) and
in rev. 1.39 2018/02/22 17:06:42 X509_set1_notBefore(3) and
X509_set1_notAfter(3). Document them, again from scratch
because what OpenSSL provides is slightly confusing.
schwarze [Sun, 25 Feb 2018 10:07:34 +0000 (10:07 +0000)]
Correct the description of ASN1_item_free(3) with respect to
optional reference counting. ETOOMUCHMAGIC. By the way, this
public hellhole is still undocumented in OpenSSL.
schwarze [Sun, 25 Feb 2018 09:49:08 +0000 (09:49 +0000)]
In x509.h rev. 1.35 2018/02/22 16:53:42, jsing@ provided
X509_CRL_up_ref(3). Since it is undocumented in OpenSSL,
write some documentation from scratch. While here, also
correct the description of X509_CRL_free(3) and mention
X509_CRL_dup(3), too.
jmc [Sun, 25 Feb 2018 07:22:22 +0000 (07:22 +0000)]
Pp useless before/after S{h,s};
jsg [Sun, 25 Feb 2018 01:53:29 +0000 (01:53 +0000)]
set NCPU so the install script will use bsd.mp when hw.ncpufound > 1
ok kettenis@
dlg [Sun, 25 Feb 2018 01:52:25 +0000 (01:52 +0000)]
default the nvgre vnetid to the first valid value according to the rfc
guenther [Sun, 25 Feb 2018 01:45:01 +0000 (01:45 +0000)]
Typo: auxilliary -> auxiliary
dlg [Sun, 25 Feb 2018 01:16:12 +0000 (01:16 +0000)]
give egre an example
nvgre had an example of joining nvgre networks together over the
internet with egre. this takes it away and reworks it for an egre
example.
while here, get rid of the hostnames in the shell prompts so things
are more consistent, which jmc and i discussed.
schwarze [Sat, 24 Feb 2018 23:42:40 +0000 (23:42 +0000)]
In x509.h rev. 1.34 2018/02/22 16:50:30, jsing@ provided
X509_REQ_get_signature_nid(3), in rev. 1.36 2018/02/22 16:58:45
X509_CRL_get_signature_nid(3), and in rev. 1.40 2018/02/22 17:09:28
X509_get0_tbs_sigalg(3). Merge the documentation from OpenSSL.
kettenis [Sat, 24 Feb 2018 22:28:08 +0000 (22:28 +0000)]
RK3328 support.
schwarze [Sat, 24 Feb 2018 21:39:29 +0000 (21:39 +0000)]
In x509.h rev. 1.33 2018/02/22 16:47:50, jsing@ provided
X509_REVOKED_get0_serialNumber(3) and X509_REVOKED_get0_revocationDate(3).
Merge the documentation from OpenSSL, but put it into X509_REVOKED_new(3)
rather than into X509_CRL_get0_by_serial(3) because it fits better there.
kn [Sat, 24 Feb 2018 21:22:44 +0000 (21:22 +0000)]
Remove obsolete www/drupal6, www/horde and www/zope.
Prompted by sthen.
kn [Sat, 24 Feb 2018 21:05:57 +0000 (21:05 +0000)]
Document MODCARO_RUSTFLAGS.
ok sthen
krw [Sat, 24 Feb 2018 21:00:00 +0000 (21:00 +0000)]
If you can modify p_cpg in when 'm'odifying a partition, you should be
able to do the same when 'a'dding a partition. Only in 'X'pert mode
of course.
ok otto@
schwarze [Sat, 24 Feb 2018 20:57:49 +0000 (20:57 +0000)]
In x509.h rev. 1.33 2018/02/22 16:47:50, jsing@ provided
X509_REVOKED_get0_extensions(3) and in rev. 1.36 2018/02/22 16:58:45
X509_CRL_get0_extensions(3). Merge the documentation from OpenSSL
and fix a few minor typos while here.
cheloha [Sat, 24 Feb 2018 20:00:07 +0000 (20:00 +0000)]
Replace popen/setjmp/pclose with a manual pipe/fork/exec/wait.
We can limit the time we wait on wall(1) without the complexity
inherent to setjmp.
Actually wait (instead of waitpid) to pick up any straggler wall
processes from prior timewarn() calls.
With a tweak from millert@ to ensure we don't accidentally close
stdin before we exec wall.
ok millert@ tb@
schwarze [Sat, 24 Feb 2018 19:24:09 +0000 (19:24 +0000)]
In ssl.h rev. 1.141 2018/02/20 18:07:11, tb@ provided
SSL_SESSION_get_protocol_version(3).
Import the documentation form OpenSSL, tweaked by me, OK tb@.
schwarze [Sat, 24 Feb 2018 19:17:10 +0000 (19:17 +0000)]
In bio.h rev. 1.39 2018/02/22 16:38:43, jsing@ provided BIO_up_ref(3).
Merge the documentation from OpenSSL, tweaked by me. While here,
fix the in parts imprecise, in parts incorrect descriptions of
BIO_new(3), BIO_set(3), BIO_free(3), and BIO_free_all(3).
schwarze [Sat, 24 Feb 2018 17:11:20 +0000 (17:11 +0000)]
In evp.h rev. 1.58 2018/02/20 18:05:28, tb@ provided
EVP_PKEY_get0_EC_KEY(3). Merge the documentation from OpenSSL.
schwarze [Sat, 24 Feb 2018 15:18:47 +0000 (15:18 +0000)]
In dh.h rev. 1.23 2018/02/20 17:59:31, tb@ provided DH_bits(3).
Merge the documentation from OpenSSL, tweaked by me.
schwarze [Sat, 24 Feb 2018 15:03:35 +0000 (15:03 +0000)]
In bio.h rev. 1.35 2018/02/20 17:55:26, tb@ provided BIO_set_shutdown(3)
and BIO_get_shutdown(3). Write the documentation from scratch because
what OpenSSL provides doesn't explain the difference to BIO_set_close(3)
and is also worded in a rather confusing way.
sthen [Sat, 24 Feb 2018 14:28:58 +0000 (14:28 +0000)]
sync
kettenis [Sat, 24 Feb 2018 14:10:36 +0000 (14:10 +0000)]
bcmrng(4)
schwarze [Sat, 24 Feb 2018 13:51:50 +0000 (13:51 +0000)]
In dsa.h rev. 1.28 2018/02/20 17:48:35, tb@ provided
DSA_SIG_get0(3) and DSA_SIG_set0(3).
Merge the documentation from OpenSSL, tweaked by me.
kettenis [Sat, 24 Feb 2018 13:46:57 +0000 (13:46 +0000)]
Enable bcmrng(4).
kettenis [Sat, 24 Feb 2018 13:46:22 +0000 (13:46 +0000)]
Add bcmrng(4), a driver for the random number generator on the Raspberry Pi.
kettenis [Sat, 24 Feb 2018 13:43:33 +0000 (13:43 +0000)]
I wrote this in 2018. Pointed out by jmc@
schwarze [Sat, 24 Feb 2018 13:25:50 +0000 (13:25 +0000)]
In dsa.h rev. 1.27 2018/02/20 17:45:44, tb@ provided
DSA_clear_flags(3), DSA_test_flags(3), and DSA_set_flags(3),
and in rev. 1.29 2018/02/20 17:52:27 DSA_get0_engine(3).
Merge the documentation from OpenSSL, tweaked by me.
kettenis [Sat, 24 Feb 2018 13:23:21 +0000 (13:23 +0000)]
fix pasto
jmc [Sat, 24 Feb 2018 13:14:09 +0000 (13:14 +0000)]
sort;
schwarze [Sat, 24 Feb 2018 13:04:47 +0000 (13:04 +0000)]
In rsa.h rev. 1.37 2018/02/20 17:42:32, tb@ provided
RSA_clear_flags(3), RSA_test_flags(3), and RSA_set_flags(3).
Merge the documentation, tweaked by me.
jsg [Sat, 24 Feb 2018 12:46:45 +0000 (12:46 +0000)]
Enable interrupts while running interrupt handlers like we do on
agintc(4) and ampintc(4).
ok kettenis@ patrick@
visa [Sat, 24 Feb 2018 11:42:30 +0000 (11:42 +0000)]
Declare ci_ipl volatile to prevent the compiler from optimizing
or reordering accesses to the variable. Assume that the assembler
preserves the correct sequence of instructions, which allows the
removal of the explicit noreorder/reorder toggles from the C code.
With ci_ipl being volatile, drop mips_sync() calls that follow
the accesses of the variable. The sync is redundant as a compiler
barrier. In addition, the MIPS64 CPU designs should not need the
sync for pipeline or write buffer control. According to miod@,
the use of the instruction is a carryover from code targeting
early MIPS designs that lack tight integration with the cache
and write buffer.
Discussed with and testing help from miod@.
Tested on CN5020, CN6120, CN7130, CN7360, Loongson 2F and 3A1000,
R4400, R8000, R10000 and R16000.
kettenis [Sat, 24 Feb 2018 11:17:00 +0000 (11:17 +0000)]
bcmtemp(4)
kettenis [Sat, 24 Feb 2018 11:07:40 +0000 (11:07 +0000)]
Enable bcmtemp(4).
kettenis [Sat, 24 Feb 2018 11:06:59 +0000 (11:06 +0000)]
Add bcmtemp(4), a driver for the temperature sensor on the Raspberry Pi.
phessler [Sat, 24 Feb 2018 10:39:35 +0000 (10:39 +0000)]
add an alias for vmctl show, matches finger memory and the rest of the
*ctl programs
OK mlarkin@ benno@ deraadt@
kettenis [Sat, 24 Feb 2018 09:45:10 +0000 (09:45 +0000)]
Make ncpusfound count the available processors even if they don't attach.
ok patrick@
dlg [Sat, 24 Feb 2018 08:53:36 +0000 (08:53 +0000)]
make the gre flowid output always 2 chars so payloads stay lined up.
dlg [Sat, 24 Feb 2018 07:20:04 +0000 (07:20 +0000)]
when comparing nvgre entries, compare a with b, not a with itself.
dlg [Sat, 24 Feb 2018 06:31:47 +0000 (06:31 +0000)]
return early on some unhandled ioctls rather than err().
kettenis [Fri, 23 Feb 2018 21:47:08 +0000 (21:47 +0000)]
Disable mailbox interrupts on all cores upon attach instead of enabling the
mailbox 0 interrupt on all cores. Seems to make GENERIC.MP work on the rpi3.
krw [Fri, 23 Feb 2018 21:39:58 +0000 (21:39 +0000)]
'partions' -> 'partitions', nuke some erroneous whitespace.
ok otto@
schwarze [Fri, 23 Feb 2018 21:34:37 +0000 (21:34 +0000)]
After opening a file with gzdopen(3), we have to call gzclose(3) or
we leak memory internally used by zlib to keep compression state.
Bug reported by Wolfgang Mueller <vehk at vehk dot de> who also
provided an incomplete patch, part of which i'm using in this commit.
schwarze [Fri, 23 Feb 2018 19:12:09 +0000 (19:12 +0000)]
In bio.h rev. 1.34 2018/02/20 17:15:27, jsing@ provided
BIO_get_new_index(3), in rev. 1.37 2018/02/20 18:13:31, tb@ provided
BIO_meth_get_read(3), BIO_meth_get_puts(3), BIO_meth_get_gets(3),
BIO_meth_get_ctrl(3), BIO_meth_get_create(3), and BIO_meth_get_destroy(3),
and in rev. 1.38 2018/02/20 18:17:17 BIO_meth_get_callback_ctrl(3)
and BIO_meth_set_callback_ctrl(3).
BIO_meth_get_write(3) will become public with the next minor bump.
Merge the documentation from OpenSSL, tweaked by me, OK tb@.
kettenis [Fri, 23 Feb 2018 19:08:56 +0000 (19:08 +0000)]
Get rid of the cpu_on_fn hook and call the psci(4) functions directly instead
like we already do in the code that flushes the BTB.
ok jsg@
schwarze [Fri, 23 Feb 2018 18:53:49 +0000 (18:53 +0000)]
clarify documentation of macro keys
schwarze [Fri, 23 Feb 2018 18:24:41 +0000 (18:24 +0000)]
Logically, the following are are type names - just like .Vt,
some of them with an optional variable name following:
- .Ft
- .Fa in the SYNOPSIS
- .Fn second and later arguments in the SYNOPSIS
So add these to the .Vt macro table in the mandoc.db(5) database.
During my LibreSSL work, i'm getting really tired of typing
$ man -k Vt,Ft,Fa=some_type_name
over and over again; now, this becomes just:
$ man -k Vt=some_type_name
markus [Fri, 23 Feb 2018 15:58:37 +0000 (15:58 +0000)]
Add experimental support for PQC XMSS keys (Extended Hash-Based Signatures)
The code is not compiled in by default (see WITH_XMSS in Makefile.inc)
Joint work with stefan-lukas_gazdag at genua.eu
See https://tools.ietf.org/html/draft-irtf-cfrg-xmss-hash-based-signatures-12
ok djm@
visa [Fri, 23 Feb 2018 15:41:08 +0000 (15:41 +0000)]
Drop redundant bzero() calls. ses_ghash is allocated with M_ZERO, so
it is unnecessary to zero the struct's fields right after allocation.
OK mikeb@
schwarze [Fri, 23 Feb 2018 14:50:21 +0000 (14:50 +0000)]
In dh.h rev. 1.22 2018/02/20 17:38:15, tb@ provided
DH_clear_flags(3), DH_test_flags(3), and DH_set_flags(3),
in rev. 1.24 2018/02/20 18:01:42 DH_set_length(3), and
in rev. 1.25 2018/02/22 16:41:04, jsing@ provided DH_get0_engine(3).
Merge the documentation from OpenSSL, tweaked by me.
schwarze [Fri, 23 Feb 2018 12:16:08 +0000 (12:16 +0000)]
In bn.h rev. 1.38 2018/02/20 17:13:14, jsing@ provided
BN_GENCB_new(3), BN_GENCB_free(3), and BN_GENCB_get_arg(3).
Merge the documentation from OpenSSL, verbatim.
jmc [Fri, 23 Feb 2018 09:50:30 +0000 (09:50 +0000)]
various tweaks; ok dlg
jmc [Fri, 23 Feb 2018 07:38:09 +0000 (07:38 +0000)]
some cleanup for BindInterface and ssh-keyscan;
dlg [Fri, 23 Feb 2018 07:05:39 +0000 (07:05 +0000)]
regen
dlg [Fri, 23 Feb 2018 07:04:57 +0000 (07:04 +0000)]
masanobu saitoh pointed out i had the wrong id for XXV710 SFP28 devs
akoshibe [Fri, 23 Feb 2018 05:17:39 +0000 (05:17 +0000)]
Add unsetrdomain() and option -rdomain to return an interface to routing
domain 0.
OK phessler, henning, deraadt, stsp, benno
djm [Fri, 23 Feb 2018 05:14:05 +0000 (05:14 +0000)]
Add ssh-keyscan -D option to make it print its results in SSHFP format
bz#2821, ok dtucker@
dtucker [Fri, 23 Feb 2018 04:18:46 +0000 (04:18 +0000)]
Add missing braces.
Caught by the tinderbox's -Werror=misleading-indentation, ok djm@
dlg [Fri, 23 Feb 2018 03:03:36 +0000 (03:03 +0000)]
lint fix
djm [Fri, 23 Feb 2018 03:03:00 +0000 (03:03 +0000)]
unbreak interop test after SSHv1 purge;
patch from Colin Watson via bz#2823