Matthew Dillon [Tue, 11 Sep 2012 23:07:44 +0000 (16:07 -0700)]
Merge branches 'hammer2' and 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly into hammer2
Matthew Dillon [Tue, 11 Sep 2012 21:39:17 +0000 (14:39 -0700)]
hammer - Add ssh-remote directive
* Adds a feature that allows you to set up a command="..." prefix for a
ssh key in your ~/.ssh/authorized_keys file that only runs the hammer
utility and only with specific commands and filesystem paths.
For example:
command="/sbin/hammer ssh-remote mirror-read,mirror-write /path/" ssh-rsa...
Currently requires a trailing '/' if you want to restrict the path to
a subdirectory. Multiple commands can be listed but the filesystem path
restriction is only currently tested for mirror-read and mirror-write.
* This allows ssh to be used for mirroring without having to give shell
access to the remote.
Matthew Dillon [Tue, 11 Sep 2012 16:54:02 +0000 (09:54 -0700)]
hammer - Add scoreboard file option
* Add -e <scoreboardfile> option for mirror-stream, so one can see the
progress of mirror-streams running in the background.
Nuno Antunes [Tue, 11 Sep 2012 10:49:51 +0000 (11:49 +0100)]
netisr: rename cpu_portfn() to netisr_portfn().
No functional change.
Searched and replaced with:
find sys/ -type f -exec sed -i "" 's/cpu_portfn/netisr_portfn/g' '{}' \;
Sascha Wildner [Tue, 11 Sep 2012 12:52:03 +0000 (14:52 +0200)]
hier.7: Oops, use Xr
Sascha Wildner [Tue, 11 Sep 2012 12:43:43 +0000 (14:43 +0200)]
hier.7: Add some words about /usr/share/terminfo.
Sascha Wildner [Tue, 11 Sep 2012 11:25:50 +0000 (13:25 +0200)]
Fix buildworld.
François Tigeot [Tue, 11 Sep 2012 09:30:37 +0000 (11:30 +0200)]
ixgbe: Document a performance tuning sysctl
* It is needed to consistently increase single stream TCP send
performance
* Without it, single tcp connections rarely reach the maximum
recorded speed of 9.4Gb/s
Sascha Wildner [Tue, 11 Sep 2012 08:39:21 +0000 (10:39 +0200)]
kernel/inet: Remove some unused variables.
Sascha Wildner [Tue, 11 Sep 2012 08:06:14 +0000 (10:06 +0200)]
mfi(4): Break some overly long lines.
Sepherosa Ziehau [Tue, 11 Sep 2012 01:40:48 +0000 (09:40 +0800)]
acpi/resource: We still need to make sure that IRQ is valid at least
Sascha Wildner [Mon, 10 Sep 2012 21:37:54 +0000 (23:37 +0200)]
kernel: Use NULL instead of 0 for pointers, part 1/x.
Found-with: Coccinelle (http://coccinelle.lip6.fr/)
Matthew Dillon [Mon, 10 Sep 2012 21:03:30 +0000 (14:03 -0700)]
hammer - Fix core dump during remote termination of mirror-stream
* Fix issue where remote-end hammer can core trying to fprintf() an error
message if a mirror-stream connection is lost unexpectedly.
Sascha Wildner [Mon, 10 Sep 2012 17:00:54 +0000 (19:00 +0200)]
kernel/acpi: Add missing include (for acpi_sci_irqno()).
François Tigeot [Sun, 9 Sep 2012 08:37:34 +0000 (10:37 +0200)]
ixgbe: Replace the TX lockmgr lock by a serializer
François Tigeot [Mon, 10 Sep 2012 12:11:41 +0000 (14:11 +0200)]
ixgbe: Do all RX/TX processing in ithreads
* There is no need to launch taskqueues if there is more data to
process, the next interrupt thread will handle it.
* This will be no later than 125µs in the general case anyway.
* This change can even increase performance by removing lock contention
between ithreads and taskqueues trying to run at the same time.
Sepherosa Ziehau [Mon, 10 Sep 2012 09:45:15 +0000 (17:45 +0800)]
ifpoll: Dispatch netmsgs to netisr, which is MPSAFE now
Sepherosa Ziehau [Mon, 10 Sep 2012 09:03:41 +0000 (17:03 +0800)]
acpi/resource: Only skip SCI trigger/polarity configuration
The code before this commit could cause interrupt storm on certain
systems, on which certain IRQs are configured into different mode
but the configured IRQs are actually never used.
Reported-by: swildner@
Chris Turner [Mon, 10 Sep 2012 04:30:21 +0000 (04:30 +0000)]
periodic.conf(5): document new 'daily_clean_hammer_pfslist' variable.
On further reflection - as hammer cleanup is not limited to pfs's -
should this this variable be called 'fslist'? but then - does that
make the usage for pfs's less obvious. Hmm. A philosophical question.
Leave the name for now, and take a sip from a glass 100% full of
some percentage of air and liquid.
suggested-by: swildner@
Chris Turner [Mon, 10 Sep 2012 00:55:21 +0000 (00:55 +0000)]
periodic/daily: add option to specify list of pfs's to daily/160.clean-hammer
Add a new variable, daily_clean_hammer_pfslist, which can be used to
specify a list of pfs's to cleanup. This is useful e.g. on systems
with offline hammer slave pfs's which should be cleaned but would
not be done so with the default 'hammer cleanup' command.
Defaults to previous online-only cleanup behavior if this variable is
not set.
Sascha Wildner [Sun, 9 Sep 2012 21:16:28 +0000 (23:16 +0200)]
iscontrol(8): Remove unused header file.
François Tigeot [Sun, 9 Sep 2012 12:30:05 +0000 (14:30 +0200)]
ixgbe: Remove the Adaptative Interrupt Moderation code
* It didn't work correctly, the number of ixgbe interrupts per second
could become excessive
* Use a fixed rate for all interrupts, be they legacy, MSI or
MSI-X vectors. 8000 intr/s is a good default value.
* This change increases single stream tcp performance up to almost 50%
in the sending direction
François Tigeot [Thu, 6 Sep 2012 11:58:55 +0000 (13:58 +0200)]
ixgbe: Move sysctl creation to a separate function
Sascha Wildner [Sun, 9 Sep 2012 12:54:20 +0000 (14:54 +0200)]
kernel/bce: Fix an impossible && that should really be a ||.
Reviewed-by: sephe
Sascha Wildner [Sun, 9 Sep 2012 13:00:26 +0000 (15:00 +0200)]
Install apm(4)'s manual page on x86_64, too.
acpi(4) emulates apm(4)'s ioctls so it is relevant on x86_64 too.
Also, mention these things.
François Tigeot [Thu, 6 Sep 2012 12:43:17 +0000 (14:43 +0200)]
ixgbe: Enable existing MSI-X code
Sascha Wildner [Fri, 7 Sep 2012 21:24:45 +0000 (23:24 +0200)]
kernel: Remove two bogus break statements.
Sascha Wildner [Fri, 7 Sep 2012 20:18:45 +0000 (22:18 +0200)]
Remove pcidevs_data.h via 'make upgrade'.
Sascha Wildner [Fri, 7 Sep 2012 20:02:08 +0000 (22:02 +0200)]
BUS_SETUP_INTR.9: Add missing comma and bump .Dd
Sascha Wildner [Fri, 7 Sep 2012 19:51:17 +0000 (21:51 +0200)]
kernel/hptmv: Fix the completion of a write.
What is intended here is to perform BUS_DMASYNC_POSTWRITE when a write
is completed, but the 'else if' had the wrong check.
François Tigeot [Fri, 7 Sep 2012 19:18:12 +0000 (21:18 +0200)]
Rename notes/ directory to doc/
Antonio Huete Jimenez [Fri, 3 Sep 2010 08:39:27 +0000 (10:39 +0200)]
doc - Fix previous commit.
Antonio Huete Jimenez [Fri, 3 Sep 2010 08:32:38 +0000 (10:32 +0200)]
doc - Add some more porting notes regarding CVS ids.
Sascha Wildner [Thu, 25 Dec 2008 21:57:13 +0000 (22:57 +0100)]
Add some notes I had lying in my tree.
Peter Avalos [Sun, 6 Apr 2008 19:08:30 +0000 (19:08 +0000)]
Lose LK_EXCLUSIVE in the lockinit example and a typo fix.
Sascha Wildner [Sat, 29 Dec 2007 18:35:59 +0000 (18:35 +0000)]
Add some documentation on converting sleep mutexes and fix a couple of typos.
Submitted-by: Aggelos Economopoulos <aoiko@cc.ece.ntua.gr>
Sascha Wildner [Wed, 31 Oct 2007 04:52:52 +0000 (04:52 +0000)]
Add a section about renamed kernel functions and clean up a bit.
Matthew Dillon [Sun, 3 Dec 2006 20:49:59 +0000 (20:49 +0000)]
Add an internal document describing (in a very incomplete way at the moment) how to port a driver from FreeBSD.
François Tigeot [Fri, 7 Sep 2012 09:43:17 +0000 (11:43 +0200)]
Document bus_setup_intr_descr(9)
Sascha Wildner [Fri, 7 Sep 2012 15:57:04 +0000 (17:57 +0200)]
VFS.9: Add a reference to VFS_SET(9).
Reported-by: Raimundo Santos <raitech@gmail.com>
Sascha Wildner [Fri, 7 Sep 2012 12:57:48 +0000 (14:57 +0200)]
kernel/ipx: Add a missing 'goto set_head;'.
Matthew Dillon [Fri, 7 Sep 2012 06:21:01 +0000 (23:21 -0700)]
Merge branches 'hammer2' and 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly into hammer2
François Tigeot [Sat, 1 Sep 2012 17:25:04 +0000 (19:25 +0200)]
ixgbe: Explicitely enable PCIe bus mastering
Sascha Wildner [Thu, 6 Sep 2012 11:11:03 +0000 (13:11 +0200)]
kernel: Use NULL for pointers in some places.
Sascha Wildner [Wed, 5 Sep 2012 21:32:42 +0000 (23:32 +0200)]
kernel: Remove pcidevs_data.h. It serves no purpose.
Sascha Wildner [Tue, 4 Sep 2012 17:21:16 +0000 (19:21 +0200)]
kernel/ixgbe: Add missing braces.
Sepherosa Ziehau [Mon, 3 Sep 2012 09:46:58 +0000 (17:46 +0800)]
tcp: Implement asynchronized pru_rcvd
This mainly avoids extra scheduling cost on the reception path due to
lwkt_domsg(). lwkt_sendmsg() is now used to carry out TCP pru_rcvd.
Since TCP's pru_rcvd could be batched, one pru_rcvd netmsg is embedded
into struct socket to avoid pru_rcvd netmsg allocation for each pru_rcvd,
and this netmsg will be used by lwkt_sendmsg(). Whether this embedded
pcu_rcvd netmsg should be sent or not is determined by its MSG_DONE bit.
Since user thread and netisr thread could be on different CPUs, the
embedded pru_rcvd netmsg's MSG_DONE bit is protected by a spinlock.
To cope with the following race that could drop window updates,
tcp_usr_rcvd() replies asynchronized rcvd netmsg before tcp_output():
netisr thread user thread
tcp_usr_rcvd() sorcvtcp()
{ {
tcp_output() :
: :
: sbunlinkmbuf()
: if (rcvd & MSG_DONE) (2)
: lwkt_sendmsg(rvcd)
: :
lwkt_replymsg(rcvd) (1)
}
At (2) window update is dropped, since rcvd netmsg is not replied yet at (1)
The result:
On i7-2600 (4C/8T, 3.4GHz):
32 parallel netperf -H 127.0.0.1 -t TCP_STREAM -P0 -l 30 (4 runs, unit: Mbps)
old 30253.88 30242.58 30162.55 30101.51
new 33962.74 33798.70 33499.92 33482.35
This gives ~12% performance improvement.
Nuno Antunes [Mon, 3 Sep 2012 04:30:05 +0000 (05:30 +0100)]
msgport.9: Fix a path.
* sys/kern/netisr.c -> sys/net/netisr.c
Pointed-out-by: Romick <yellowrabbit2010@gmail.com>
Sascha Wildner [Mon, 3 Sep 2012 04:04:38 +0000 (06:04 +0200)]
acpi: Remove some unused files.
Sascha Wildner [Mon, 3 Sep 2012 03:04:55 +0000 (05:04 +0200)]
Remove another unused header (<machine/ultrasound.h>).
Sascha Wildner [Mon, 3 Sep 2012 02:28:19 +0000 (04:28 +0200)]
Remove an unused old ral(4) firmware header.
Sascha Wildner [Mon, 3 Sep 2012 02:18:05 +0000 (04:18 +0200)]
Remove an unused and empty header (<machine/mtpr.h>).
Sascha Wildner [Sun, 2 Sep 2012 16:08:50 +0000 (18:08 +0200)]
libprop: Add two missing MLINKS.
Sascha Wildner [Sun, 2 Sep 2012 15:55:30 +0000 (17:55 +0200)]
tbridge: Sync NAME, SYNOPSIS and MLINKS.
Sascha Wildner [Sun, 2 Sep 2012 15:47:26 +0000 (17:47 +0200)]
libm: Add missing MLINK.
Sascha Wildner [Sun, 2 Sep 2012 11:31:33 +0000 (13:31 +0200)]
Bring in a VFS_SET(9) manual page (from FreeBSD).
Sascha Wildner [Sat, 1 Sep 2012 22:26:04 +0000 (00:26 +0200)]
link.2: Use .Fn and add .Nm linkat
Sascha Wildner [Fri, 31 Aug 2012 22:06:05 +0000 (00:06 +0200)]
amr(4): Remove some unused code (amr_timeout()).
Taken-from: FreeBSD
Sascha Wildner [Fri, 31 Aug 2012 21:00:35 +0000 (23:00 +0200)]
Add CARP support to the GENERIC kernels.
http://leaf.dragonflybsd.org/mailarchive/users/2012-08/msg00086.html
Suggested-by: Charles Rapenne <charles.rapenne@gmail.com>
Approved-by: sephe
Alex Hornung [Thu, 30 Aug 2012 08:17:12 +0000 (08:17 +0000)]
usched_bsd4 - topology sched: go for safe defaults
* When we don't know how to handle the level of the current CPU, go for
safe defaults disabling both cache-coherent and SMT scheduling.
* This seems to be the case on single-core machines where the level is
the PACKAGE_LEVEL.
Reported-by: Max Herrgard, Ferruccio Zamuner, David Shao
Dragonfly-bug: http://bugs.dragonflybsd.org/issue2408
Dragonfly-bug: http://bugs.dragonflybsd.org/issue2413
Sepherosa Ziehau [Thu, 30 Aug 2012 04:13:08 +0000 (12:13 +0800)]
sorecvtcp: Remove unapplied code
Sepherosa Ziehau [Thu, 30 Aug 2012 03:26:13 +0000 (11:26 +0800)]
socket: Replicate soreceive() to sorecvtcp() for cleanup and optimization
Sepherosa Ziehau [Wed, 29 Aug 2012 09:51:35 +0000 (17:51 +0800)]
jme: RX interrupt to ~6000Hz, TX interrupt to ~4000Hz
Sepherosa Ziehau [Wed, 29 Aug 2012 09:35:08 +0000 (17:35 +0800)]
jme: Record number of times that RX ring becomes empty
Sepherosa Ziehau [Wed, 29 Aug 2012 09:10:43 +0000 (17:10 +0800)]
jme: Allow MSI-X be evenly spreaded across CPUs
Sepherosa Ziehau [Wed, 29 Aug 2012 08:40:29 +0000 (16:40 +0800)]
igb: Set default RX descriptor count to 512
Sepherosa Ziehau [Wed, 29 Aug 2012 02:25:47 +0000 (10:25 +0800)]
acpi/sci: Test ACPI MADT interrupt override entry's preferred mode first
Sascha Wildner [Tue, 28 Aug 2012 22:16:01 +0000 (00:16 +0200)]
wpa_supplicant(8): Use libpcap functions.
Sascha Wildner [Tue, 28 Aug 2012 22:08:24 +0000 (00:08 +0200)]
kernel/acpi: Add smart battery support.
I don't know which laptops have them, as far as I can tell, none of mine
do. But as some seem to have it, it's worth supporting.
https://en.wikipedia.org/wiki/Smart_Battery_System
This is the patch from the tracker with some later updates FreeBSD did.
Taken-from: FreeBSD
Dragonfly-bug: <https://bugs.dragonflybsd.org/issues/1229>
Submitted-by: Dmitry Komissaroff <aunoor@gmail.com>
Sascha Wildner [Tue, 28 Aug 2012 16:32:09 +0000 (18:32 +0200)]
mpt(4): We don't have MOD_QUIESCE.
Sepherosa Ziehau [Tue, 28 Aug 2012 07:10:58 +0000 (15:10 +0800)]
tcp/ncr: Avoid using magic number
Sepherosa Ziehau [Tue, 28 Aug 2012 07:00:28 +0000 (15:00 +0800)]
tcp: RFC3517bis is now officially RFC6675
Sepherosa Ziehau [Tue, 28 Aug 2012 06:30:47 +0000 (14:30 +0800)]
acpi/pcib: Guard against invalid GSI provided by PRT
Sepherosa Ziehau [Tue, 28 Aug 2012 06:12:10 +0000 (14:12 +0800)]
acpi/resource: Use legacy_intr_find to detect invalid IRQ configure
Sepherosa Ziehau [Tue, 28 Aug 2012 06:02:10 +0000 (14:02 +0800)]
pci: Guard against wrong user supplied IRQ assignment
Sepherosa Ziehau [Tue, 28 Aug 2012 05:43:35 +0000 (13:43 +0800)]
acpi/fadt: Make sure that SCI IRQ is valid
Sepherosa Ziehau [Tue, 28 Aug 2012 04:46:20 +0000 (12:46 +0800)]
pci/pir: Make sure that the IRQ is allowed to be configured
If the IRQ is already configured into non-level/low mode, we should not
change the trigger and polarity
Sepherosa Ziehau [Tue, 28 Aug 2012 04:23:26 +0000 (12:23 +0800)]
MachIntr: Add two methods to find IRQ
- Find IRQ conforming to the specified trigger and polarity, if it was
configured.
- Find IRQ by GSI, the located IRQ must conform to the specified trigger
and polarity if it was configured.
Sepherosa Ziehau [Mon, 27 Aug 2012 02:43:48 +0000 (10:43 +0800)]
acpi/intr: Fix comment, source index is GSI _not_ IRQ
Sascha Wildner [Mon, 27 Aug 2012 18:00:46 +0000 (20:00 +0200)]
kernel: Turn the delay before mounting root into a tunable.
Submitted-by: Francis Gudin <fgudin@nerim.net>
Dragonfly-bug: <http://bugs.dragonflybsd.org/issues/2373>
Sascha Wildner [Mon, 27 Aug 2012 17:57:49 +0000 (19:57 +0200)]
/boot/defaults/loader.conf: Adjust a comment.
François Tigeot [Sat, 25 Aug 2012 15:25:26 +0000 (17:25 +0200)]
ixgbe: add tso_pullup function
* This routine rearranges mbuf chains to get more continuous bytes,
potentially increasing tcp send performance
* Single TCP streams are now able to push slightly more than 4Gb/s
under the right circumstances
Taken-from: Sepherosa Ziehau's work on the igb(4) driver
François Tigeot [Sat, 25 Aug 2012 14:35:41 +0000 (16:35 +0200)]
ixgbe: enable existing FreeBSD IPv4 TSO code
* This is not perfect but increases sending speeds up to about 3.5 Gb/s
by TCP connection
* Total throughput has been measured up to 9.22 Gb/s in the sending
direction
François Tigeot [Mon, 20 Aug 2012 19:59:03 +0000 (21:59 +0200)]
kernel: IFCAP_TSO is really IPv4-specific
* Rename IFCAP_TSO to IFCAP_TSO4 to make things crystal clear, keeping
the old name for compatibility
* Add IFCAP_TSO6 for the future
Matthew Dillon [Sun, 26 Aug 2012 17:09:19 +0000 (10:09 -0700)]
kernel - reduce kern.maxvnodes default on 32 bit systems
* Reduce maximum kern.maxvnodes on standard-configured 32 bit systems
from ~100K vnodes to ~80K vnodes to reduce kmem usage.
Sascha Wildner [Sun, 26 Aug 2012 11:15:35 +0000 (13:15 +0200)]
growfs(8)/ffsinfo(8): Use __func__
Sascha Wildner [Sun, 26 Aug 2012 11:14:13 +0000 (13:14 +0200)]
growfs(8): Fix building with -DFS_DEBUG.
Sascha Wildner [Sun, 26 Aug 2012 10:35:14 +0000 (12:35 +0200)]
make_dev.9: Fix typo.
Sascha Wildner [Sun, 26 Aug 2012 09:55:37 +0000 (11:55 +0200)]
kernel/ffs: Compare v_rdev directly when updating mounts.
Nowadays the vnode's v_rdev is always associated, so it can be used
directly to compare the devices.
Pointed-out-by: alexh
Sascha Wildner [Sat, 25 Aug 2012 20:48:12 +0000 (22:48 +0200)]
kernel/linprocfs: Implement /proc/devices.
As close to Linux' format as I could get, and enough to make LSI's
MegaCli utility (for RAID configuration) work with mfi(4).
Sascha Wildner [Sun, 26 Aug 2012 10:12:46 +0000 (12:12 +0200)]
kernel/devfs: Fill out cdev_t's si_major field (and adjust one use case).
Sascha Wildner [Sun, 26 Aug 2012 10:31:13 +0000 (12:31 +0200)]
kernel/devfs: Pass more arguments to the callback for devfs_scan_callback.
is_alias signals whether it is an alias or a real device. If true, name
is the alias name. If false, it is the device name.
Reviewed-by: alexh
Sascha Wildner [Thu, 23 Aug 2012 22:40:22 +0000 (00:40 +0200)]
kernel/devfs: Rename devfs_nodetype members (conflicts with {,lin}procfs.h).
François Tigeot [Sat, 25 Aug 2012 12:38:08 +0000 (14:38 +0200)]
ixgbe: enable VLAN code
Sascha Wildner [Sat, 25 Aug 2012 06:04:39 +0000 (08:04 +0200)]
kernel/devfs: Assert that the callback isn't NULL in devfs_scan_callback().
Sascha Wildner [Sat, 25 Aug 2012 05:12:50 +0000 (07:12 +0200)]
Add a device_get_name(9) manual page (from FreeBSD).
Matthew Dillon [Sat, 25 Aug 2012 00:09:58 +0000 (17:09 -0700)]
kernel - Do not allow destroyed namecache entries to be re-resolved
* Do not allow a destroyed namecache entry to be re-resolved, as it
might resolve as a completely different file, or even resolve as
a file when it was originally a directory and so forth.
* Fixes inconsistencies in the current-dir fields in proc->p_fd.
* Note that most VFS's can't re-resolve a disconnected directory
anyway but HAMMER1 actually can, so this was causing a problem
in tests with HAMMER1 (the same tests that originally crashed
tmpfs were also crashing HAMMER1, but for a different reason).
Matthew Dillon [Sat, 25 Aug 2012 00:05:40 +0000 (17:05 -0700)]
Merge branches 'hammer2' and 'master' of ssh://crater.dragonflybsd.org/repository/git/dragonfly into hammer2
Matthew Dillon [Fri, 24 Aug 2012 23:30:20 +0000 (16:30 -0700)]
tmpfs - Fix numerous races and adjust to use cache_unlink()
* Fix numerous issues when chdir'd into a directory that is then
rmdir'd.
* Rewrite the link count handling code for directories, also removing
two now unecessary recursions.
* Do not allow new files to be created in a directory which has been
rmdir'd.
* Use the new cache_unlink facility.
Matthew Dillon [Fri, 24 Aug 2012 23:29:49 +0000 (16:29 -0700)]
kernel - Adjust devfs, hammer, and puffs to use the new cache_unlink()
* Adjust devfs, hammer, and puffs to use the new cache_unlink() mechanic.