krw [Sun, 20 Jul 2008 01:53:43 +0000 (01:53 +0000)]
Add DIOCGPDINFO support. 'disklabel -d svnd0' now works.
martynas [Sat, 19 Jul 2008 23:35:44 +0000 (23:35 +0000)]
link to apache web server, not the project group. found while
catching up w/ the old email. submitted by Pierre Riteau on tech@
a year ago.
"no objection" jmc@. ok, "that's clearly correct" okan@
marco [Sat, 19 Jul 2008 22:41:58 +0000 (22:41 +0000)]
Full rewrite of metadata handling. This fixes power failures and crashes
that caused illegal checksums. The new metadata code is more or less ready
to deal with other vendor's metadata formats.
While here clean up the name space.
Fix thib's pool mess by adding removing bad flags in interrupt context.
tested on macppc, amd64, i386, sparc64 & hppa
sparc64 has issues with crypto however those do not seem to be softraid
specific.
help from okan@ ckuethe@ Will Backman and others
claudio [Sat, 19 Jul 2008 21:31:39 +0000 (21:31 +0000)]
Use errx instead of fprintf, exit combo.
Diff from Gleydson Soares gsoares (at) gmail (dot) com
OK henning@
miod [Sat, 19 Jul 2008 19:25:18 +0000 (19:25 +0000)]
In dec_1000a_intr_map(), use the SRM hint if available. This allows cards
plugged to the bottom 4 PCI slots of AlphaServer 1000A (attaching to pci1
behind a ppb) to get interrupts.
No regressions on AlphaServer 800 (which do not have these extra slots).
miod [Sat, 19 Jul 2008 18:13:04 +0000 (18:13 +0000)]
Add code to walk the eisa configuration, from NetBSD. This configuration
is not used yet, but this seems to ``warm up'' the eisa chips so that
accesses to the eisa bus later do not cause machine checks.
miod [Sat, 19 Jul 2008 17:10:03 +0000 (17:10 +0000)]
In _io_map(), make sure the whole range fits in one of the io windows,
instead of only the starting address. From NetBSD.
jmc [Sat, 19 Jul 2008 16:35:50 +0000 (16:35 +0000)]
minor wording tweaks;
brad [Sat, 19 Jul 2008 15:55:07 +0000 (15:55 +0000)]
Correct missed capitalization of MPLS.
kettenis [Sat, 19 Jul 2008 12:36:06 +0000 (12:36 +0000)]
Add acx(4) at cardbus(4).
reyk [Sat, 19 Jul 2008 12:10:07 +0000 (12:10 +0000)]
print the host parent id if available
reyk [Sat, 19 Jul 2008 11:38:54 +0000 (11:38 +0000)]
no need for using a TAILQ queue for the host children list, use a
singly-linked SLIST instead. the only noticeable change is the
reversed order to notify the children but it does not really matter
here. also only walk through the children host list if the host
itself is a potential parent.
mglocker [Sat, 19 Jul 2008 11:30:55 +0000 (11:30 +0000)]
If a resolution which has been requested over VIDEO_TRY_FMT or
VIDEO_S_FMT doesn't exactly match the devices available resolutions,
return the next best matching resolution which we have.
Makes some V4L2 apps happy when running them with the default resolution
(no options).
reyk [Sat, 19 Jul 2008 10:52:32 +0000 (10:52 +0000)]
If the new 'parent' keyword is specified for a host in a table,
inherit the state from another host with the specified Id; no
additional check will be for the inheriting host. This helps in
scenarios with lots of IP aliases that all point to the same service
on the same host (like web hosting with many SSL domains).
discussed with pyr, tested in different setups
reyk [Sat, 19 Jul 2008 10:35:31 +0000 (10:35 +0000)]
Use the IPv6 documentation prefix 2001:db8::/32 from RFC 3849 instead
of the deprecated 3ffe::/16 6bone test address space. Deprecated or
public IPv6 allocations should not be used for documentation; we have
the same approach for IPv4 by using RFC 1918 private addresses.
ok jmc@ todd@
kettenis [Sat, 19 Jul 2008 10:02:09 +0000 (10:02 +0000)]
Unbreak the tree. Pointed out by James Hartley.
mglocker [Sat, 19 Jul 2008 06:39:00 +0000 (06:39 +0000)]
In interrupt handler change printf -> DRPINTF.
art [Fri, 18 Jul 2008 23:43:31 +0000 (23:43 +0000)]
Add a macro that clears the want_resched flag that need_resched sets.
Right now when mi_switch picks up the same proc, we didn't clear the
flag which would mean that every time we service an AST we would attempt
a context switch. For some architectures, amd64 being probably the
most extreme, that meant attempting to context switch for every
trap and interrupt.
Now we clear_resched explicitly after every context switch, even if it
didn't do anything. Which also allows us to remove some more code
in cpu_switchto (not done yet).
miod@ ok
jmc [Fri, 18 Jul 2008 22:51:01 +0000 (22:51 +0000)]
no need for .Pp before or after .Sh;
jmc [Fri, 18 Jul 2008 22:44:21 +0000 (22:44 +0000)]
+.Xr crash 8 ,
from Stephan A. Rickauer
kettenis [Fri, 18 Jul 2008 22:27:50 +0000 (22:27 +0000)]
Add ohci(4) and ehci(4) at cardbus(4).
kettenis [Fri, 18 Jul 2008 22:05:29 +0000 (22:05 +0000)]
Determine the free address space by looking at the "available" property of
the PCI host bridge if we're not running on an UltraBook. Fix allocation of
bus number such that it works on machines that have OpenBoot 4.x.
mglocker [Fri, 18 Jul 2008 21:45:24 +0000 (21:45 +0000)]
For negotation first get the devices default values (GET_DEF) instead
asking for (GET_CUR) directly. This gets us better negotation values.
miod [Fri, 18 Jul 2008 21:39:52 +0000 (21:39 +0000)]
Protect dma operations with splvm instead of splhigh, that's faster than
splbio and won't delay clock interrupts.
miod [Fri, 18 Jul 2008 21:39:14 +0000 (21:39 +0000)]
Make IPL_VM level 5, which is guaranteed to be above bio tty and net, instead
of making it dynamic and the smallest value above the former three. Idea
from NetBSD.
mglocker [Fri, 18 Jul 2008 18:49:11 +0000 (18:49 +0000)]
Implement VIDIOC_S_FMT and VIDIOC_TRY_FMT. Now V4L2 applications can
set their desired image size, and therefore users can manipulate the
image size, too via the application.
Also tested by brad@
kurt [Fri, 18 Jul 2008 16:58:06 +0000 (16:58 +0000)]
Add support for the kernel to recognize, load and execute position
independent executables using the uvm_map_pie() function to
randomize the load address. okay miod@, kettenis@, drahn@
kurt [Fri, 18 Jul 2008 16:40:17 +0000 (16:40 +0000)]
Add new uvm function called uvm_map_pie() which takes align as a
parameter and returns an aligned random load address for position
independent executables to use. This also adds three new vmparam.h
defines to specify the maximum address, minimum address and minimum
allowed alignment for uvm_map_pie() to use. The PIE address range
for i386 was carefully selected to work well within the i386 W^X
framework.
With much help and feedback from weingart@.
okay weingart@, miod@, kettenis@, drahn@
mglocker [Fri, 18 Jul 2008 14:45:50 +0000 (14:45 +0000)]
Add OmniVision OV7670 cam to the supported devices which comes in the
Dell XPS M1330 Notebooks.
Tested by kettenis@
kevlo [Fri, 18 Jul 2008 14:24:18 +0000 (14:24 +0000)]
remove unnessasary cast.
ok millert@
thib [Fri, 18 Jul 2008 13:38:40 +0000 (13:38 +0000)]
o Use mbufs, for the RX ring, instead of malloc()'ing an MCLBYTES sized buffer.
o On non-strict alignment archs, dont copy the mbuf, every time, unload it, and send
it up the stack and just get a new one for the rx ring. We still do the copy on
strict alignment archs though...
o create a function to handle mbuf allocation for the rx ring, vr_mbuf_alloc(),
use it to allocate the mbufs and shuffle the bus dma setup around.
ideas/code from vic(4) and sis(4);
ok reyk@, brad@, dlg@
tested by many, been in snapshots for a while.
martynas [Fri, 18 Jul 2008 13:08:58 +0000 (13:08 +0000)]
fix cases, found by lint:
- actually raise inexacts for expm1, and support
- __exp__E always returns, because inexact check always succeeds.
add NOTREACHED
- some cases in lgamma FALLTHROUGH, as intended
ok millert@
reyk [Fri, 18 Jul 2008 12:35:27 +0000 (12:35 +0000)]
merge host_v6 with relayd's version to use getaddrinfo instead of
inet_pton. host_v4, host_v6, and host_dns could be merged into one
function using getaddrinfo but i keep it in multiple functions to keep
it in sync with the other daemons using this common code.
reyk [Fri, 18 Jul 2008 12:30:06 +0000 (12:30 +0000)]
merge log_host with relayd's version using getnameinfo and rename it
to print_host.
reyk [Fri, 18 Jul 2008 12:26:52 +0000 (12:26 +0000)]
terminate the input buffer on failure in print_host()
reyk [Fri, 18 Jul 2008 12:00:20 +0000 (12:00 +0000)]
the Belkin
F5D7010 version # is a RTL8185 cardbus device. this
chipset revision is not supported yet, so add it commented out.
reyk [Fri, 18 Jul 2008 11:45:24 +0000 (11:45 +0000)]
sync
reyk [Fri, 18 Jul 2008 11:44:22 +0000 (11:44 +0000)]
Add Belkin Wireless G Notebook Card
F5D7010 (rtw(4) RTL8185-based)
marco [Fri, 18 Jul 2008 03:54:18 +0000 (03:54 +0000)]
POLL causes concurrency issues with event driven batteries. Remove for now
until I have time to implement proper fix.
Noted by several folks.
martynas [Fri, 18 Jul 2008 00:25:22 +0000 (00:25 +0000)]
oops. move infnan to md.vax. pointed out by theo
kettenis [Thu, 17 Jul 2008 23:08:08 +0000 (23:08 +0000)]
Zero mbreg_t structure *before* initializing any members, not when we're
halfway through. Makes the ISP12160 correctly load its firmware again on
sparc64.
ok krw@
brad [Thu, 17 Jul 2008 19:59:44 +0000 (19:59 +0000)]
Add a flag and allow the driver to identify the BCM5788 chipsets. They're
slightly different from the other chipsets using the BCM5705 ASIC.
ok jsg@
reyk [Thu, 17 Jul 2008 16:41:06 +0000 (16:41 +0000)]
final reorder diff to use TAILQ_INSERT_TAIL instead of
TAILQ_INSERT_HEAD. now tables and relays are also matching the order
in the config file.
ok pyr@
reyk [Thu, 17 Jul 2008 16:28:07 +0000 (16:28 +0000)]
add the hosts in order to get host ids that match the order in the config file. ok pyr@
reyk [Thu, 17 Jul 2008 16:12:04 +0000 (16:12 +0000)]
give sane ids to hosts, tables, redirections, relays, etc. - start
counting at 1 and do not assign an id before inheriting a real table.
makes more sense in the relayctl output.
ok pyr@
martynas [Thu, 17 Jul 2008 15:36:28 +0000 (15:36 +0000)]
properly raise inexact; ok millert@
reyk [Thu, 17 Jul 2008 15:10:14 +0000 (15:10 +0000)]
use getaddrinfo/getnameinfo to parse ipv6 addresses instead of
inet_pton/inet_ntop to allow specifing and printing the IPv6 scope
identifier. synced host_v6() with ntpd's version to use getaddrinfo()
instead of inet_pton() - host_v4, host_v6, and host_dns could all use
getaddrinfo in a single function by specifing different flags but this
would diverge from the other daemons using this common interface so we
keep this little overhead.
discussed with henning@
ok pyr@
jsing [Thu, 17 Jul 2008 13:50:49 +0000 (13:50 +0000)]
Remove clause 3 from NetBSD license.
ok deraadt@
jsg [Thu, 17 Jul 2008 13:16:29 +0000 (13:16 +0000)]
Increase the size of the diskinfo struct, which has grown
4 bytes in the new firmware.
To allow this to work on older firmwares add a "short read allowed"
flag to arc_msgbuf(), which will allow short reads by getting
the read length from the header of the response the firmware sends.
Only set the flag on the arc_msgbuf() call that deals with the diskinfo
struct for now.
Discussed with and ok dlg@
djm [Thu, 17 Jul 2008 08:51:07 +0000 (08:51 +0000)]
strip trailing '.' from hostname when HostbasedUsesNameFromPacketOnly=yes
report and patch from res AT qoxp.net (bz#1200); ok markus@
djm [Thu, 17 Jul 2008 08:48:00 +0000 (08:48 +0000)]
strnvis preauth banner; pointed out by mpf@ ok markus@
sthen [Thu, 17 Jul 2008 08:14:24 +0000 (08:14 +0000)]
Pasto in error message for setspppkey()
ok mbalmer@
drahn [Wed, 16 Jul 2008 20:33:42 +0000 (20:33 +0000)]
Fix several of the 'large offset' PLT entries. With help from kurt.
ok kurt@
miod [Wed, 16 Jul 2008 20:03:20 +0000 (20:03 +0000)]
Add a new parameter to pckbc_cnattach(): flags to put in the pckbc bowels.
For now, only one such flag is defined, PCKBC_CANT_TRANSLATE. It hints
pckbc that the device does not behave correctly to the ``set translation''
commands.
Set this flag if we are running on a Tadpole Ultrabook machine, which needs it.
This makes the built-in keyboard work correctly on this laptop (with the
help of the software translation pckbd diff).
tested & ok kettenis@
miod [Wed, 16 Jul 2008 20:00:45 +0000 (20:00 +0000)]
If pckbc_xt_translation() fails, do not try to force the keyboard in table
1; instead, try table 2 first, then 1: some older keyboards, connected to
a castrated 8042, do not know how to talk in table 1 (or worse, will report
they do while they don't).
If automatic translation is not possible, remember which table the keyboard
ended configured in, and convert the scancodes from table 2 to table 1 if
necessary.
ok kettenis@
miod [Wed, 16 Jul 2008 16:41:03 +0000 (16:41 +0000)]
One more pckbd.
miod [Wed, 16 Jul 2008 16:40:05 +0000 (16:40 +0000)]
Send unused string literal ``mapme'' to nirvana.
miod [Wed, 16 Jul 2008 16:34:34 +0000 (16:34 +0000)]
This file is long unused.
miod [Wed, 16 Jul 2008 16:33:18 +0000 (16:33 +0000)]
It is no longer necessary to regen gsckbdmap after changing this file.
miod [Wed, 16 Jul 2008 16:32:06 +0000 (16:32 +0000)]
Implement pckbc_xt_translation() in gsckbc. Thus, we can attach pckbd instead of
gsckbd; the former will cause a proper translation page to be selected by
the keyboard.
Because of this, we no longer depend on the page the keyboard is left in
by the PDC (page 2 for all machines but the PrecisionBook, which is in
page 3), and there is no longer any need to use separate keyboard maps.
miod [Wed, 16 Jul 2008 15:49:22 +0000 (15:49 +0000)]
Now that uvm_pglistalloc() does not lose on large memory gaps, do not
restrict the memory allocation range in _dmamem_alloc().
martynas [Wed, 16 Jul 2008 15:25:51 +0000 (15:25 +0000)]
some archaic 'n = a(b), c(d)' use that confused lint
ok millert@
martynas [Wed, 16 Jul 2008 15:15:26 +0000 (15:15 +0000)]
install infnan.3 only for vax. pointed out by millert@
ok millert@
martynas [Wed, 16 Jul 2008 15:14:33 +0000 (15:14 +0000)]
'no subject' check is never triggered. check if it's actually empty
ok millert@
martynas [Wed, 16 Jul 2008 15:11:16 +0000 (15:11 +0000)]
add -E flag and 'skipempty' option to skip sending messages with
empty bodies. useful for sending mails from crontabs.
from rivo nurges <rix at estpak dot ee>, with missing 'skipempty'
documentation and usage update from me
ok millert@
martynas [Wed, 16 Jul 2008 15:04:09 +0000 (15:04 +0000)]
ansify infnan example. sure millert@
reyk [Wed, 16 Jul 2008 15:02:19 +0000 (15:02 +0000)]
relay_connect() may fail, close the session in the bindany callback
if it does.
martynas [Wed, 16 Jul 2008 15:01:19 +0000 (15:01 +0000)]
infnan always returned wrong values, because compiler assumed it
returns integer value, so add missing prototype
ok millert@
martynas [Wed, 16 Jul 2008 14:57:27 +0000 (14:57 +0000)]
better description for TMPDIR. from millert@
martynas [Wed, 16 Jul 2008 14:56:11 +0000 (14:56 +0000)]
man page tweaks:
- for argument of -f flag, name -> file
- document TMPDIR, better description from millert@
ok millert@
martynas [Wed, 16 Jul 2008 14:53:41 +0000 (14:53 +0000)]
- use strncmp/strncasecmp instead of comparing by character
- simplify istrlcpy, no need to check for isupper
- line[0] is redundant, because strcasecmp will take care of it
ok millert@
reyk [Wed, 16 Jul 2008 14:49:44 +0000 (14:49 +0000)]
use getsockname() to find out the local address of a connection before
doing a nat lookup. this fixes nat lookups when the relay is
listening to a wildcard IPv4/IPv6 address (like 0.0.0.0 or ::).
martynas [Wed, 16 Jul 2008 14:49:09 +0000 (14:49 +0000)]
- err with the pathbuf, if we know it
- use tmpdir instead of /tmp
ok millert@
reyk [Wed, 16 Jul 2008 14:38:33 +0000 (14:38 +0000)]
fix nat lookup to use the correct pf_addr offset. now it also works
with ipv6.
djm [Wed, 16 Jul 2008 11:52:19 +0000 (11:52 +0000)]
this loop index should be automatic, not static
djm [Wed, 16 Jul 2008 11:51:14 +0000 (11:51 +0000)]
rename variable first_gc -> last_gc (since it is actually the last
in the list).
millert [Wed, 16 Jul 2008 10:33:44 +0000 (10:33 +0000)]
Dynamically allocate fd_sets so we are not limited to FD_SETSIZE
connections. With help from djm@. OK djm@
canacar [Wed, 16 Jul 2008 10:23:39 +0000 (10:23 +0000)]
Cleanup #define maze leftover from pftop compatibility.
Pointed out by mcbride@, makes sense henning@
henning [Wed, 16 Jul 2008 09:00:44 +0000 (09:00 +0000)]
link udp pcbs to pf states, same as done for tcp already
ok markus, also tested david sthen
brad [Wed, 16 Jul 2008 03:56:25 +0000 (03:56 +0000)]
Enable the read DMA engine's PCI read request burst length long burst
mode (4KB) for PCIe chips. This resolves the poor TX performance for
the PCIe chips. The result being a bit under double the TX performance
on a Gig connection (roughly 495 Mb/s -> 940 Mb/s).
Tested by reyk@, sthen@, brad@ and a few end users.
kettenis [Tue, 15 Jul 2008 22:49:01 +0000 (22:49 +0000)]
Add a timecounter based on the $sys_tick register, and use it on machines
that have it. Initial diff from art@.
kettenis [Tue, 15 Jul 2008 22:46:58 +0000 (22:46 +0000)]
Use -Wa,-Av9b instead of -Wa,-Av9a such that I can use %sys_tick in an
upcoming change.
martynas [Tue, 15 Jul 2008 19:23:26 +0000 (19:23 +0000)]
- chraise can be replaced with toupper; no need to check for islower
- remove quite some code, use strcasestr instead of reimplementing
it each time
- use strncasecmp, instead of comparing through each character
"looks fine" millert@
martynas [Tue, 15 Jul 2008 19:13:25 +0000 (19:13 +0000)]
set cp to the value of PAGER, so that type1 doesn't try to use it
uninitialized (page=1), or use wrong previously-set value of crt
(page=0). ok millert@
kettenis [Tue, 15 Jul 2008 17:50:20 +0000 (17:50 +0000)]
Backout rev 1.162. This change made us muck with with pci config space at
address 0x1a and 0x1e, and that's not where the PCIe capability stuff
lives. Potentially it was mucking with an IO BAR (super dangerous).
But probably it was achieving nothing at all.
ok dlg@, marco@, brad@
oga [Tue, 15 Jul 2008 16:57:48 +0000 (16:57 +0000)]
Remove some debug code that i forgot to remove in a previous commit.
Pointed out by Chris Cappuccio, thanks!
jsg [Tue, 15 Jul 2008 13:21:17 +0000 (13:21 +0000)]
Adapt some code from the FreeBSD driver required to make
newer PCI Express adapters (ie 8168C*/8102*) work.
V2 Checksum offload format in RTL8102 devices not yet supported.
No objections from brad@. Thanks to everyone who tested.
thib [Tue, 15 Jul 2008 12:10:48 +0000 (12:10 +0000)]
theres no need too m_adj() the mbuf by 8bytes (u_int64_t) in
sis_newbuf(), so remove the call.
ok and lots of prodding dlg@, brad@
espie [Tue, 15 Jul 2008 11:10:51 +0000 (11:10 +0000)]
make the version number stuff mandatory.
Restrict specifications to the parts that actually make sense (in particular,
remove most wildcard support, it's no longer a good idea in true
version number handling), explain about v*, and be more specific about
default package specs (since the fuzzyness in the original description
predates several changes to bsd.port.mk).
Also add version numbers and wildcards where they make sense so the examples
are more complete.
espie [Tue, 15 Jul 2008 10:48:14 +0000 (10:48 +0000)]
deprecate package names without version numbers.
pvalchev [Tue, 15 Jul 2008 03:07:40 +0000 (03:07 +0000)]
sync
djm [Tue, 15 Jul 2008 02:23:14 +0000 (02:23 +0000)]
number of pipelined requests is not 64; prodded by Iain.Morgan AT nasa.gov
miod [Mon, 14 Jul 2008 20:01:37 +0000 (20:01 +0000)]
Fix struct sysioreg layout after the struct iommureg layout change;
ok kettenis@
mglocker [Mon, 14 Jul 2008 19:57:36 +0000 (19:57 +0000)]
Make VIDIOC_ENUM_FMT list all available formats for the attached device.
miod [Mon, 14 Jul 2008 16:23:07 +0000 (16:23 +0000)]
Regen
miod [Mon, 14 Jul 2008 16:22:23 +0000 (16:22 +0000)]
/dev/tuner description should point out to bktr(4), spotted by jmc.
miod [Mon, 14 Jul 2008 16:21:52 +0000 (16:21 +0000)]
Do not create /dev/joy* and /dev/{r,}mcd* nodes, these devices are not
available on amd64.
art [Mon, 14 Jul 2008 15:17:08 +0000 (15:17 +0000)]
Don't lock timeouts in db_show_callout.
All cpus are stopped and this cpu blocks all interrupts. It doesn't make
sense to grab locks that ddb can then jump past with longjmp.
Noticed by Pierre Riteau. I just forgot about the bug until reminded
today.
miod [Mon, 14 Jul 2008 14:00:00 +0000 (14:00 +0000)]
Use uvm_km_valloc_prefer_wait() instead of uvm_km_valloc_wait() in vmapbuf().
ok art@
miod [Mon, 14 Jul 2008 13:37:39 +0000 (13:37 +0000)]
Be sure to propagate PSL_O in psw on PCXU* processors to userland processes
and signal handlers.
ok kettenis@