Commit Graph

243364 Commits

Author SHA1 Message Date
ozaki-r
f0423d34e6 Use if_get_byindex instead of if_byindex for MP-safe 2016-06-16 03:03:33 +00:00
ozaki-r
4ce95a7f59 Bump kernel version for mbuf (pkthdr#rcvif) change 2016-06-16 02:54:41 +00:00
ozaki-r
e1135cd9b9 Use curlwp_bind and curlwp_bindx instead of open-coding LP_BOUND 2016-06-16 02:38:40 +00:00
ozaki-r
8e415f7b54 Introduce curlwp_bind and curlwp_bindx
The API prevents the current LWP from migrating between CPUs during
the critical section (between curlwp_bind and curlwp_bindx). One use
case of it is psref(9) that has a contract that forbids such migrations.

Proposed at http://mail-index.netbsd.org/tech-kern/2016/06/13/msg020710.html
(check it out if you want to know why the function names are chosen)
2016-06-16 02:34:33 +00:00
dholland
7cf14915db Notes on MIPS models and architecture levels and their properties and
handling in NetBSD.

This is very preliminary so far, because I'm trying to get citations
from documentation for everything.

Please add stuff.
2016-06-16 01:33:35 +00:00
sevan
17a9d10433 Add a unit test for PR bin/51250 called se_output. se_output checks the output of cat
when invoked with '-se', to ensure that a '$' is printed on blank lines.
2016-06-16 01:04:58 +00:00
sevan
9edf879ce4 When invoked with -se, print a '$' on blank lines
Obtained from OpenBSD r1.13 of src/bin/cat/cat.c

Closes PR bin/51250
Reviewed by Christos@
2016-06-16 00:52:37 +00:00
sevan
f3ad9a4c7a Revert previous change made by me in r1.55 2016-06-16 00:45:37 +00:00
agc
0ad97abd01 Update netpgpverify to 20160617
Sync with changes in pkgsrc

+ don't assume a string is NUL-terminated - use fwrite(3) with a specific size
2016-06-15 20:34:28 +00:00
agc
f0cd146e8f Update netpgpverify to 20160616:
+ Bring over change from pkgsrc to add version.asc signature verification
to complement the noversion.asc cleartext signatures

+ Update version to 20160616
2016-06-15 16:51:46 +00:00
joerg
d9ced641b5 Mark obuf_printf as printf-like. 2016-06-15 15:47:50 +00:00
christos
874963f630 Harmonize the error messages about dynamic relocations. 2016-06-15 14:47:23 +00:00
macallan
1f03e35c95 g/c some leftovers from when we had static assignments of edge or level
triggers
2016-06-15 14:32:03 +00:00
riastradh
7c30fabe0e Remove dumplfs from the list now that it seems strict-aliasing-safe. 2016-06-15 14:11:15 +00:00
riastradh
08df494e80 Kill another -fno-strict-aliasing.
Generated code is same with/without.
2016-06-15 14:08:24 +00:00
riastradh
17e0569b61 <stdbool.h> for true and false. 2016-06-15 14:07:54 +00:00
riastradh
ebd94f1f29 Omit needless casts. 2016-06-15 13:57:39 +00:00
riastradh
7554812ac2 No more obvious strict aliasing violations here. 2016-06-15 13:57:26 +00:00
riastradh
356b70489c Kill another strict aliasing violation. 2016-06-15 13:47:26 +00:00
riastradh
d947e13b5f This hack no longer exists since the switch to ipsec-tools in 2005.
Another strict-aliasing problem killed.
2016-06-15 13:29:30 +00:00
jmcneill
75e20526bb If either "regulator-boot-on" or "regulator-always-on" properties are true,
explicitly enable the regulator at attach time. In addition, if the
"startup-delay-us" property is present, delay after enabling the regulator.
2016-06-15 13:13:40 +00:00
christos
4df2af8f4b Remove SIGPOLL, see https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=92362
for an explanation why.
2016-06-15 12:10:18 +00:00
christos
3ee1ef9906 Turn off GNU_RELRO for now. 2016-06-15 12:08:47 +00:00
ryo
2bdf21974d fix the tx stall problem. ("watchdog timeout")
TX descriptor should be written out once before writing READY flag.
2016-06-15 07:26:11 +00:00
ozaki-r
c7e18ccbde Protect if_byindex by pserialize 2016-06-15 06:01:21 +00:00
mrg
e6e96483bf replace the previous hack with something that i believe is actually
correct and, more importantly ;), works properly.

thanks for riastradh for hints about which bit was actually broken.
2016-06-15 05:01:58 +00:00
mlelstv
d48455317c Work around an issue with MSG_WAITALL that would block for too long by
replacing it with a receive loop.
2016-06-15 04:33:52 +00:00
mlelstv
bbe94f43d2 Remove throttling code, instead signal scsipi layer to reduce the openings
and retry the command. Start with a small openings number and let scsipi
request to grow it up to the current send window.

Adjust ccb and pdu counts to avoid ressource shortages. These are still
very ad-hoc numbers, but seem to be sufficient for a Gigabit link.

Use separate condvar for PDU pool and add counter to help debugging.

Revert setting PDU disposition to UNUSED before freeing. free_pdu
uses this as a flag to detect already returned PDUs.

Add reference counter for open commands to defer unmapping a session
that would lead to crashes in scsipi.

Move session cleanup to cleanup thread.

Use get_sernum to retrieve current serial number where possible and
make it check for immediate commands itself.

Adjust debug output.
2016-06-15 04:30:30 +00:00
mlelstv
ae0ee83b73 drop unused debug helper routine 2016-06-15 03:51:55 +00:00
mlelstv
e27d88e376 Fix field name in response pdu. 2016-06-15 03:40:59 +00:00
agc
02fd82b2cb netpgpverify: bring over changes in 20160615 from pkgsrc
+ perform check for start of ascii-armoured signature in a more efficient
way
2016-06-15 03:37:50 +00:00
mrg
6bfcf96505 build convert_endian() with "no-strict-aliases" optimiser attribute
to avoid some unknown miscompilation in endian_convert() that causes
ssh to exit on some output (for me, "cc -v".)

note in HACKS.  we should investigate this further if possible as
this seems to indicate a strict aliasing violation.  there certainly
are 32 and 64 bit object arrays being accessed with 8 bit accessors,
but i don't have time currently.
2016-06-15 02:12:14 +00:00
agc
ffd13a8c6d add file used for testing gpg --emit-no-version case 2016-06-14 20:55:00 +00:00
agc
4dc60d040f add test for signatures produced by gpg --no-emit-version 2016-06-14 20:47:43 +00:00
agc
dd98b26d9b Sync with pkgsrc sources as of version 20160614
+ pick up renaming changes to internal routines
+ fix for issue verifying signatures created by gpg --no-emit-version
+ add test for same
2016-06-14 20:47:08 +00:00
sjg
fee87d5733 meta_oodate: set needOODATE if oodate due to missing .meta file. 2016-06-14 18:16:06 +00:00
knakahara
d4315a0e8d fix: rxq->rxq_lock is uninitialized if NET_MPSAFE is not defined. 2016-06-14 17:09:20 +00:00
christos
e7090289a4 Add and document MKRELRO and set x86 to partial by default. 2016-06-14 13:26:45 +00:00
christos
0e6265fc35 Add support for GNU RELRO headers from Matthias Weckbecker. 2016-06-14 13:06:41 +00:00
skrll
59c4afa2c5 Simplify the code.
WM_[RT]X_* macros are removed and replaced with mutex_* calls directly.

OK from nakahara@ and msaitoh@
2016-06-14 09:07:22 +00:00
isaki
527eb93c1e In XM6 TypeG (and original XM6), its version number is
represented by BCD, not decimal.
Reported by GIMONS (the author of XM6 typeG).
2016-06-14 07:51:10 +00:00
pgoyette
4f13ec197c Re-arrange PAX_ASLR options to keep the _DEBUG next to the main option. 2016-06-14 03:34:01 +00:00
christos
d83ec840b3 fix PIC code. 2016-06-14 03:05:24 +00:00
ryoon
64125f9b55 Fix buld with clang, add warning option exceptions 2016-06-13 20:03:07 +00:00
abhinav
762c2184ca Make the function static in the definition (prototype was already declared static).
Missed in my last commit.
2016-06-13 14:58:57 +00:00
christos
954c4c5c26 Simplify, no functional change. 2016-06-13 14:23:26 +00:00
knakahara
a6f4292e65 eliminate unnecessary splnet 2016-06-13 08:37:15 +00:00
knakahara
e4ff09f05d MP-ify fastforward to support GATEWAY kernel option.
I add "ipflow_lock" mutex in ip_flow.c and "ip6flow_lock" mutex in ip6_flow.c
to protect all data in each file. Of course, this is not MP-scalable. However,
it is sufficient as tentative workaround. We should make it scalable somehow
in the future.

ok by ozaki-r@n.o.
2016-06-13 08:34:23 +00:00
knakahara
14ea9af5f7 make ipflow_reap() static function. 2016-06-13 08:29:55 +00:00
knakahara
f2808ade1a remove unnecessary splnet before pool_{get,put} 2016-06-13 08:04:44 +00:00