openbsd
8 years agoMention jumbo frames.
visa [Thu, 4 Aug 2016 13:22:20 +0000 (13:22 +0000)]
Mention jumbo frames.

8 years agoAdd support for jumbo frames.
visa [Thu, 4 Aug 2016 13:10:31 +0000 (13:10 +0000)]
Add support for jumbo frames.

8 years agoAdd support for pre-registering interrupts. This allows device drivers to
kettenis [Thu, 4 Aug 2016 12:17:36 +0000 (12:17 +0000)]
Add support for pre-registering interrupts.  This allows device drivers to
establish interrupts before their interrupt controller attaches, solving
dependency problems in various device trees.

Also add support for handing interrupt handlers over to parent interrupt
controllers.

ok jsg@ patrick@ (on an earlier diff)

8 years agoFix an assertion failure that happened when trying to add a page
schwarze [Thu, 4 Aug 2016 09:33:53 +0000 (09:33 +0000)]
Fix an assertion failure that happened when trying to add a page
with makewhatis -d to a completely empty database.
Reported by Mark Patruck <mark at wrapped dot cx>, thanks!

8 years agoadjust database file name and version number to what was committed
schwarze [Thu, 4 Aug 2016 09:18:53 +0000 (09:18 +0000)]
adjust database file name and version number to what was committed

8 years agotpm tweaks;
jmc [Thu, 4 Aug 2016 05:39:28 +0000 (05:39 +0000)]
tpm tweaks;

8 years agoApply a warning fix from lukem@NetBSD circa 1997 (more '()'), and
krw [Wed, 3 Aug 2016 23:37:25 +0000 (23:37 +0000)]
Apply a warning fix from lukem@NetBSD circa 1997 (more '()'), and
one from he@NetBSD circa 1998 (more '{}').

Makes gcc happier.

8 years agoApply fix from David Binderman via NetBSD. "Use all 3 of b, p, n
krw [Wed, 3 Aug 2016 23:22:48 +0000 (23:22 +0000)]
Apply fix from David Binderman via NetBSD. "Use all 3 of b, p, n
to determine if route changed".

Spotted by the Echelon team with AppChecker static analyzer.

NetBSD fix pointed out by millert@

8 years agoApply fix from mrouted 3.8 via NetBSD. Our mbone support may
krw [Wed, 3 Aug 2016 23:13:54 +0000 (23:13 +0000)]
Apply fix from mrouted 3.8 via NetBSD. Our mbone support may
be slightly dated.

Spotted by the Echelon team with AppChecker static analyzer.

8 years agomove acpimcfg to keep list sorted
jcs [Wed, 3 Aug 2016 22:50:47 +0000 (22:50 +0000)]
move acpimcfg to keep list sorted

8 years agoreference dwiic here too
jcs [Wed, 3 Aug 2016 22:47:43 +0000 (22:47 +0000)]
reference dwiic here too

8 years agoadd a man page for tpm(4)
jcs [Wed, 3 Aug 2016 22:46:44 +0000 (22:46 +0000)]
add a man page for tpm(4)

8 years agoFix typo, spotted by Holger Mikolon <holger@mikolon.com>
vgross [Wed, 3 Aug 2016 20:45:36 +0000 (20:45 +0000)]
Fix typo, spotted by Holger Mikolon <holger@mikolon.com>

8 years agonot having a century is a known layout by now
jcs [Wed, 3 Aug 2016 17:33:50 +0000 (17:33 +0000)]
not having a century is a known layout by now

8 years agodon't complain about invalid checksum if the bios isn't writing it
jcs [Wed, 3 Aug 2016 17:29:18 +0000 (17:29 +0000)]
don't complain about invalid checksum if the bios isn't writing it
properly, the user isn't going to do anything about it

8 years agoadd a minimal TPM 1.2 driver just to issue a "save state" command
jcs [Wed, 3 Aug 2016 17:23:38 +0000 (17:23 +0000)]
add a minimal TPM 1.2 driver just to issue a "save state" command
before suspending, in order to fix suspend/resume on some newer
machines.

more cleanup and TPM 2.0 support coming later.

earlier version ok pirofti
ok deraadt

8 years agoUse atomic operations to manipulate event masking bits
mikeb [Wed, 3 Aug 2016 17:14:41 +0000 (17:14 +0000)]
Use atomic operations to manipulate event masking bits

8 years agostrnames[i][0] == 'k' && strnames[i][0] == 'f' should be
krw [Wed, 3 Aug 2016 16:32:08 +0000 (16:32 +0000)]
strnames[i][0] == 'k' && strnames[i][0] == 'f' should be
strnames[i][0] == 'k' && strnames[i][1] == 'f'.

Spotted by the Echelon team with AppChecker static analyzer.

millert@ naddy@ and ncurses 6.0 all agree.

8 years agoA couple of "a->blah == a->blah" -> "a->blah == b->blah".
krw [Wed, 3 Aug 2016 16:27:25 +0000 (16:27 +0000)]
A couple of "a->blah == a->blah" -> "a->blah == b->blah".

Spotted by the Echelon team with AppChecker static analyzer.

ok sashan@

8 years agoRemove the periodic timer and do rescheduling during Rx completion
mikeb [Wed, 3 Aug 2016 15:08:06 +0000 (15:08 +0000)]
Remove the periodic timer and do rescheduling during Rx completion

This change adds a check into the Rx ring completion routine that
schedules an interrupt task to be executed immediately after if
consumer index has already advanced itself.  The benefit of doing
this compared to an additional loop after replenishing the ring
(as done in FreeBSD for example) is that first of all this goes
through the loop in the taskqueue thread with a yeild check to
prevent CPU hogging and second is that it triggers Tx completion
as well since interrupt handler runs both.

8 years agoUse an atomic operation to clear pending event bits
mikeb [Wed, 3 Aug 2016 14:55:57 +0000 (14:55 +0000)]
Use an atomic operation to clear pending event bits

Pending event bits are located in a shared memory and are potentially
accessed by multiple CPUs running dom0 and the guest VM.  It appears
that a failure to synchronize changes to this shared memory leads to
race conditions resulting in the guest missing out on notifications.

8 years agoFold umass_atapi_attach() and umass_scsi_setup() into umass_scsi_attach() to
krw [Wed, 3 Aug 2016 13:44:49 +0000 (13:44 +0000)]
Fold umass_atapi_attach() and umass_scsi_setup() into umass_scsi_attach() to
shorten and simplify code.

No intentional functional change.

8 years agoSimplify the way we handle TLB flushes. Since ARMv7 effectively has a
kettenis [Wed, 3 Aug 2016 11:52:43 +0000 (11:52 +0000)]
Simplify the way we handle TLB flushes.  Since ARMv7 effectively has a
unified TLB there is not much point in optimizing TLB flushing for pages
that have never been executable.  The only difference is a flush of the
branch predictor and even that isn't necessary anymore on all but the oldest
Cortex cores.

ok patrick@

8 years agoWe only replace the first %%, not multiple (use %2 for second).
nicm [Wed, 3 Aug 2016 09:08:40 +0000 (09:08 +0000)]
We only replace the first %%, not multiple (use %2 for second).

8 years agoFix minimum size when pane status line is enabled, reported by Y Petremann.
nicm [Wed, 3 Aug 2016 09:07:02 +0000 (09:07 +0000)]
Fix minimum size when pane status line is enabled, reported by Y Petremann.

8 years agoThe ARMv7 architecture deprecates the separate Instruction and Data TLB
kettenis [Wed, 3 Aug 2016 07:05:05 +0000 (07:05 +0000)]
The ARMv7 architecture deprecates the separate Instruction and Data TLB
maintanenance instruction and recommends to only use the instructions that
operate on the unified TLB.  Those instructions will flush both TLBs on
implementations that still have separate Instruction and Data TLBs.  Switch
the TLB maintenance primitives over to use those.  This allows us to reduce
the number of primitives on armv7.  We still keep separate "ID" and "D"
variants as we still have to flush the branch predictor when changing the
mappings of pages that contain instructions.

ok jsg@

8 years agoshorten the openssl dhparam text;
jmc [Wed, 3 Aug 2016 06:43:21 +0000 (06:43 +0000)]
shorten the openssl dhparam text;
guenther helped rewrite the -dsaparam parts.

8 years agosmall refactor of cipher.c: make ciphercontext opaque to callers
djm [Wed, 3 Aug 2016 05:41:57 +0000 (05:41 +0000)]
small refactor of cipher.c: make ciphercontext opaque to callers
feedback and ok markus@

8 years agoFix bug introduced in rev 1.467 which causes "buffer_get_bignum_ret:
dtucker [Wed, 3 Aug 2016 04:23:55 +0000 (04:23 +0000)]
Fix bug introduced in rev 1.467 which causes "buffer_get_bignum_ret:
incomplete message" errors when built with WITH_SSH1 and run such that no
Protocol 1 ephemeral host key is generated (eg "Protocol 2", no SSH1 host
key supplied).  Reported by rainer.laatsch at t-online.de, ok deraadt@

8 years agoFix the begin of word bug in s-command, similar to sed.
martijn [Tue, 2 Aug 2016 20:17:06 +0000 (20:17 +0000)]
Fix the begin of word bug in s-command, similar to sed.

nit fix and OK schwarze@

8 years agoMove to libevent; ok florian@
jca [Tue, 2 Aug 2016 17:00:09 +0000 (17:00 +0000)]
Move to libevent; ok florian@

8 years agoEnsure that "lim" is initialized and reset at each loop run.
jca [Tue, 2 Aug 2016 16:17:54 +0000 (16:17 +0000)]
Ensure that "lim" is initialized and reset at each loop run.

Avoids possible invalid memory accesses.  ok florian@

8 years agoIgnore the -bindir option, meaningless on OpenBSD.
jca [Tue, 2 Aug 2016 16:09:55 +0000 (16:09 +0000)]
Ignore the -bindir option, meaningless on OpenBSD.

Lack of -bindir handling is not a problem right now because -bindir is
passed to cc(1) which ignores it, and because read(2) on /usr/local/bin
returns 0 right now, so ld(1) copes with it.  But there are plans to
make read(2) fail with EISDIR.

Input from and bulk builds thanks to ajacoutot@, ok ajacoutot@ espie@

8 years agoAllow specifying an alternate socket path.
jca [Tue, 2 Aug 2016 16:05:32 +0000 (16:05 +0000)]
Allow specifying an alternate socket path.

This allows one to run multiple ripd instances, for example to serve
multiple rdomains.  Diff from Nima GHOTBI, ok claudio@ florian@ benno@

8 years agocapitalize a question like all the others
jcs [Tue, 2 Aug 2016 16:00:10 +0000 (16:00 +0000)]
capitalize a question like all the others

ok krw

8 years agoRevert previous since it adds new symbols.
jsing [Tue, 2 Aug 2016 07:47:11 +0000 (07:47 +0000)]
Revert previous since it adds new symbols.

Requested by deraadt@

8 years agowhitespace and 80 columns, no binary change
benno [Mon, 1 Aug 2016 21:25:53 +0000 (21:25 +0000)]
whitespace and 80 columns, no binary change

8 years agosync http.h with relayd
benno [Mon, 1 Aug 2016 21:15:30 +0000 (21:15 +0000)]
sync http.h with relayd
ok reyk@

8 years agorfc 7230 3.3.3 says: response with a status code of 1xx
benno [Mon, 1 Aug 2016 21:14:45 +0000 (21:14 +0000)]
rfc 7230 3.3.3 says: response with a status code of 1xx
(Informational) or 204 (No Content) MUST not have a Content-Length.
Add a check for that.
ok reyk@

8 years agoLooks like patrick@ committed an older version of the diff. Fix this, and
kettenis [Mon, 1 Aug 2016 21:08:20 +0000 (21:08 +0000)]
Looks like patrick@ committed an older version of the diff.  Fix this, and
rename a few variables and functions to be a bit more consistent with other
armv7 code.

8 years agoRepair makewhatis -t, sorry for breaking it in the previous commit.
schwarze [Mon, 1 Aug 2016 20:46:33 +0000 (20:46 +0000)]
Repair makewhatis -t, sorry for breaking it in the previous commit.
Committing this quickly because it hurts package builders.
Regression reported by naddy@.

8 years agoAllow a bel character in wall. This was missed when migrating away from
martijn [Mon, 1 Aug 2016 20:30:25 +0000 (20:30 +0000)]
Allow a bel character in wall. This was missed when migrating away from
vis(3). This brings the allowed characters on par with write(1).

Noticed by consus <at> gmx <dot> com
OK millert@

8 years agoRemove vi's "directory" option and TMPDIR support.
bentley [Mon, 1 Aug 2016 18:27:35 +0000 (18:27 +0000)]
Remove vi's "directory" option and TMPDIR support.

ok jung@

8 years agothe default smtpd setup does not require newaliases be run;
jmc [Mon, 1 Aug 2016 18:20:19 +0000 (18:20 +0000)]
the default smtpd setup does not require newaliases be run;
noted by consus

ok gilles

8 years agoBump TLS_API for addition of ALPN support.
jsing [Mon, 1 Aug 2016 17:40:23 +0000 (17:40 +0000)]
Bump TLS_API for addition of ALPN support.

8 years agoAdd ALPN support to libtls.
jsing [Mon, 1 Aug 2016 17:32:19 +0000 (17:32 +0000)]
Add ALPN support to libtls.

ok beck@ doug@

8 years agozap unused var
jca [Mon, 1 Aug 2016 17:04:18 +0000 (17:04 +0000)]
zap unused var

8 years agoMake sure variables are initialized before using them
stefan [Mon, 1 Aug 2016 16:40:09 +0000 (16:40 +0000)]
Make sure variables are initialized before using them

In elf.c, do not increment `s' before it is initialized. At the time
of the increment, `s' is otherwise unused anyway.

In elflink.c, initialize sec_contents and l_sec_contents to make
sure that the free(sec_contents) and free(l_sec_contents) are called
on valid pointers.

ok jca@ guenther@

8 years agoDo not pass VM name directly to setproctitle format string.
stefan [Mon, 1 Aug 2016 16:32:10 +0000 (16:32 +0000)]
Do not pass VM name directly to setproctitle format string.

Use a secure idiom instead.
ok mlarkin@

8 years agobring the light of ansi to a few more files
tedu [Mon, 1 Aug 2016 15:58:22 +0000 (15:58 +0000)]
bring the light of ansi to a few more files

8 years agoDon't forget to destroy the taskqueue on interrupt disestablish
mikeb [Mon, 1 Aug 2016 14:37:39 +0000 (14:37 +0000)]
Don't forget to destroy the taskqueue on interrupt disestablish

8 years agoImplement an FDT-aware interrupt establish API. This means the drivers
patrick [Mon, 1 Aug 2016 14:17:00 +0000 (14:17 +0000)]
Implement an FDT-aware interrupt establish API.  This means the drivers
don't need to know where to attach to.  Instead the API will take care
of finding the correct interrupt establish for a given device node and
will call it with the correct data.

Adapted from the OFW GPIO framework.

ok kettenis@

8 years agoMark shared producer and consumer indices volatile
mikeb [Mon, 1 Aug 2016 13:48:33 +0000 (13:48 +0000)]
Mark shared producer and consumer indices volatile

8 years agodocument the new file format
schwarze [Mon, 1 Aug 2016 12:24:37 +0000 (12:24 +0000)]
document the new file format

8 years agoSet SDEV_UMASS on all umass devices, even when they use ATAPI or
krw [Mon, 1 Aug 2016 10:56:31 +0000 (10:56 +0000)]
Set SDEV_UMASS on all umass devices, even when they use ATAPI or
UFI command sets. Probe all the LUNs a device reports instead of
assuming ATAPI and UFI devices always have 1 LUN.

Motivated by the device David Vasek reported as claiming to be
ATAPI but wanting to use >1 LUN. Committing quickly to smoke out
the inevitable device that works in the exact opposite way. All
hail USB.

ok kettenis@

8 years agoRemove the dependency on SQLite without loss of functionality.
schwarze [Mon, 1 Aug 2016 10:32:39 +0000 (10:32 +0000)]
Remove the dependency on SQLite without loss of functionality.
Drop the obsolete names_check() now that we deleted MLINKS.
Run "doas makewhatis" after compiling and installing this.

Earlier version tested by jmc@ and jturner@;
"commit it all" deraadt@   "commit and dodge" krw@

8 years agoshorten the openssl dgst text; in particular, do not try to
jmc [Mon, 1 Aug 2016 07:23:29 +0000 (07:23 +0000)]
shorten the openssl dgst text; in particular, do not try to
list all the available digests;

8 years agoupdate currency exchange rates;
jmc [Mon, 1 Aug 2016 07:22:00 +0000 (07:22 +0000)]
update currency exchange rates;

8 years agodocument default for DISTDIR
espie [Mon, 1 Aug 2016 00:48:22 +0000 (00:48 +0000)]
document default for DISTDIR

8 years agoSome base system pages, for example perl(1), contain non-ASCII
schwarze [Sun, 31 Jul 2016 23:35:26 +0000 (23:35 +0000)]
Some base system pages, for example perl(1), contain non-ASCII
characters in their source code, so switch on charset autodetection
in the same way as in man(1) itself.
Issue reported by Pavan Maddamsetti at gmail dot com on bugs@.

8 years agoRemove devmap stuff which is unused on armv7.
kettenis [Sun, 31 Jul 2016 22:27:07 +0000 (22:27 +0000)]
Remove devmap stuff which is unused on armv7.

ok patrick@

8 years agoAccording to te armv7 ARM TLB entries that caused a Permission fault might
kettenis [Sun, 31 Jul 2016 22:04:44 +0000 (22:04 +0000)]
According to te armv7 ARM TLB entries that caused a Permission fault might
be held in the TLB.  On top of that valid page table entries might be
speculatively loaded into the TLB.  As a result we need to flush TLB entries
even when the page in question has not been referenced.

Fixes pmap_fault_fixup messages on Cortex-A53, and presumably also on
Cortex-A7.

ok patrick@, guenther@

8 years agoClose /dev/hotplug on exec(). Otherwise a restart of the daemon may
natano [Sun, 31 Jul 2016 20:13:12 +0000 (20:13 +0000)]
Close /dev/hotplug on exec(). Otherwise a restart of the daemon may
fail, because the device is occupied by a child process.

from Alexey Vatchenko
ok jca

8 years agobump for LibreSSL 2.5.x
bcook [Sun, 31 Jul 2016 19:30:32 +0000 (19:30 +0000)]
bump for LibreSSL 2.5.x

8 years agoRename the route socket to a more useful name.
jca [Sun, 31 Jul 2016 19:23:24 +0000 (19:23 +0000)]
Rename the route socket to a more useful name.

Avoid a bunch of "warning: declaration of 's' shadows a global
declaration" while here.

8 years agoTurn two global variables into variables local to main.
jca [Sun, 31 Jul 2016 19:17:41 +0000 (19:17 +0000)]
Turn two global variables into variables local to main.

8 years agoDuplicated decls
jca [Sun, 31 Jul 2016 19:12:26 +0000 (19:12 +0000)]
Duplicated decls

8 years agoAsk for DHO_BOOTFILE_NAME and DHO_TFTP_SERVER by default. May prove
krw [Sun, 31 Jul 2016 18:55:16 +0000 (18:55 +0000)]
Ask for DHO_BOOTFILE_NAME and DHO_TFTP_SERVER by default. May prove
useful to autoinstall in the future and in the meantime can provide
information on what the dhcp server has done with the desired server
name and file name info.

Original diff from Patrik Lundin via tech@

8 years agoUse m_devget(9).
kettenis [Sun, 31 Jul 2016 12:51:49 +0000 (12:51 +0000)]
Use m_devget(9).

ok claudio@

8 years agoUse ansi style function declarations. No binary change.
jsg [Sun, 31 Jul 2016 09:18:01 +0000 (09:18 +0000)]
Use ansi style function declarations.  No binary change.
From Robert Tate.

8 years agoregen
jsg [Sun, 31 Jul 2016 07:37:04 +0000 (07:37 +0000)]
regen

8 years agoAdd Xeon E3-1200 v5 host bridge and some more Sunrise Point H PCH ids.
jsg [Sun, 31 Jul 2016 07:36:16 +0000 (07:36 +0000)]
Add Xeon E3-1200 v5 host bridge and some more Sunrise Point H PCH ids.

8 years agoRecognise Cortex A35 and Cortex A73.
jsg [Sun, 31 Jul 2016 06:24:38 +0000 (06:24 +0000)]
Recognise Cortex A35 and Cortex A73.

8 years agosync
jsg [Sun, 31 Jul 2016 04:24:06 +0000 (04:24 +0000)]
sync

8 years agoinstall acpials.4
jsg [Sun, 31 Jul 2016 04:23:14 +0000 (04:23 +0000)]
install acpials.4

8 years agoInstead of testing MIDR values for every model of Cortex processor check
jsg [Sun, 31 Jul 2016 03:49:51 +0000 (03:49 +0000)]
Instead of testing MIDR values for every model of Cortex processor check
MMFR0 for an ARMv7 VMSA MMU that can handle short descriptors when
setting ARMv7 function pointers.  ARMv8 in AArch32 mode is documented to
set the same bits.

ok patrick@

8 years agoRevert unneeded version bump for perl modules
afresh1 [Sat, 30 Jul 2016 23:55:35 +0000 (23:55 +0000)]
Revert unneeded version bump for perl modules

OK sthen@

8 years agouse the style from the man page examples for getaddrinfo, which makes a
halex [Sat, 30 Jul 2016 22:04:04 +0000 (22:04 +0000)]
use the style from the man page examples for getaddrinfo, which makes a
bit more sense

ok jung@ deraadt@

8 years agosync
sthen [Sat, 30 Jul 2016 21:40:55 +0000 (21:40 +0000)]
sync

8 years agoCheck for wraparound before the "commit" phase of uvm_map() and uvm_mapanon(),
kettenis [Sat, 30 Jul 2016 16:43:44 +0000 (16:43 +0000)]
Check for wraparound before the "commit" phase of uvm_map() and uvm_mapanon(),
to prevent hitting assertions and/or corrupting data structures during that
phase.

ok deraadt@, tedu@

8 years agoAdd a few checks for potential integer overflow and underflow related to the
kettenis [Sat, 30 Jul 2016 16:37:54 +0000 (16:37 +0000)]
Add a few checks for potential integer overflow and underflow related to the
size of an address range.

ok deraadt@, tedu@

8 years agoadd acpials(4), an acpi driver for ambient light sensors
jcs [Sat, 30 Jul 2016 16:25:04 +0000 (16:25 +0000)]
add acpials(4), an acpi driver for ambient light sensors

hw.sensors.acpials0.illuminance0=11.00 lx (ambient light sensor)

8 years agoadd an authors section for some drivers i authored
jcs [Sat, 30 Jul 2016 15:44:45 +0000 (15:44 +0000)]
add an authors section for some drivers i authored

8 years agohook up regression tests for jot(1)
tb [Sat, 30 Jul 2016 13:57:21 +0000 (13:57 +0000)]
hook up regression tests for jot(1)

8 years agoAdd regression tests for jot(1).
tb [Sat, 30 Jul 2016 13:55:54 +0000 (13:55 +0000)]
Add regression tests for jot(1).

Adapted from FreeBSD's test suite by attila () stalphonsos ! com.

Minor tweaks:
* make tests work properly with 'make obj' and setting VERBOSE
* add 10 tests checking some examples in our man page
* disable the dddd test for now since it loops forever

Besides dddd, the following five tests fail: dhhd dhhd2 wp3 wp4 wp5
Suggested fixes for these are on tech@

Many thanks to attila for doing the hard work!

8 years agotest suite for the new mandoc.db(5) format;
schwarze [Sat, 30 Jul 2016 10:56:13 +0000 (10:56 +0000)]
test suite for the new mandoc.db(5) format;
will be enabled after committing in src/usr.bin/mandoc

8 years agoRemove disabled jumbo frame checks. They will not be needed
visa [Sat, 30 Jul 2016 09:45:09 +0000 (09:45 +0000)]
Remove disabled jumbo frame checks. They will not be needed
for jumbo frame support.

8 years agoDrop a redundant parameter of octeon_eth_buf_free_work().
visa [Sat, 30 Jul 2016 09:29:14 +0000 (09:29 +0000)]
Drop a redundant parameter of octeon_eth_buf_free_work().

8 years agoStop pretending we can still boot without an FDT. Remove the code paths
kettenis [Sat, 30 Jul 2016 08:07:01 +0000 (08:07 +0000)]
Stop pretending we can still boot without an FDT.  Remove the code paths
handling ATAGS and cleanup the early boot code.

ok patrick@, tom@, jsg@

8 years agoRemove compat code.
ajacoutot [Sat, 30 Jul 2016 06:31:17 +0000 (06:31 +0000)]
Remove compat code.

8 years agoThis is effectively a ksh(1) script.
ajacoutot [Sat, 30 Jul 2016 06:25:21 +0000 (06:25 +0000)]
This is effectively a ksh(1) script.

8 years agoPrep for relro: make sure it's off for any non-PIE stand/ program
guenther [Sat, 30 Jul 2016 03:25:48 +0000 (03:25 +0000)]
Prep for relro: make sure it's off for any non-PIE stand/ program

ok millert@ kettenis@

8 years agoold keys no longer needed
deraadt [Sat, 30 Jul 2016 03:05:22 +0000 (03:05 +0000)]
old keys no longer needed

8 years agoAdd a periodic timer to workaround missing completion events
mikeb [Fri, 29 Jul 2016 22:25:28 +0000 (22:25 +0000)]
Add a periodic timer to workaround missing completion events

8 years agoDisable receive ring slot accounting
mikeb [Fri, 29 Jul 2016 22:01:57 +0000 (22:01 +0000)]
Disable receive ring slot accounting

In the Netfront case it appears that by having scheduler decide when
to run the workload we don't need to involve additional mechanisms
to artificially limit resource availability to achieve better
performance under heavy load.  On the contrary, by performing a
single pass through Rx and Tx completion rings and having scheduler
decide when to run the next attempt, we limit the amount of time
spent in the packet processing and achieve system responsiveness.

8 years agoMove xen interrupt handlers to dedicated task queues
mikeb [Fri, 29 Jul 2016 21:27:43 +0000 (21:27 +0000)]
Move xen interrupt handlers to dedicated task queues

Handling receive and transmit for multiple networking interfaces
in a "shared interrupt" within normal interrupt vector code path
introduces too much delay from the hypervisor POV which prevents
it from injecting further completion event interrupts for Rx and
Tx queues.

Additionally, Netfront backend driver includes a mechanism to
detect Rx ring stalls and "turn the carrier off" when the guest
is not replenishing the ring (e.g. due to missing completion
interrupts) that relies on guest waking up periodically and making
sure that the Rx ring completion handling is progressing.

Having tried both task queue + timeout and interrupts + timeout
approaches, it appears that using  the task queue is more flexible
and provides superior performance under heavy network load.

8 years agoLoop until we've read all available responses
mikeb [Fri, 29 Jul 2016 21:05:26 +0000 (21:05 +0000)]
Loop until we've read all available responses

8 years agoadd a check that the arguments to isavail don't overflow.
tedu [Fri, 29 Jul 2016 20:44:40 +0000 (20:44 +0000)]
add a check that the arguments to isavail don't overflow.
callers should probably check too, but checking here won't hurt.
possible panic reported by tim newsham.
ok kettenis

8 years agoFix regression introduced by patch to CVE-2016-0718: Tag names were cut off in some...
rpointel [Fri, 29 Jul 2016 18:57:39 +0000 (18:57 +0000)]
Fix regression introduced by patch to CVE-2016-0718: Tag names were cut off in some cases.

8 years agoall of the update functions take a size_t. correct type and casts.
tedu [Fri, 29 Jul 2016 18:35:45 +0000 (18:35 +0000)]
all of the update functions take a size_t. correct type and casts.
ok deraadt millert