Sascha Wildner [Thu, 26 Apr 2007 17:35:03 +0000 (17:35 +0000)]
Use .Xr for manpage references.
Sascha Wildner [Thu, 26 Apr 2007 15:47:43 +0000 (15:47 +0000)]
Add a reference to manpath.config(5).
Sascha Wildner [Thu, 26 Apr 2007 15:45:59 +0000 (15:45 +0000)]
Use a smaller indentation because it looks better.
Sepherosa Ziehau [Thu, 26 Apr 2007 12:59:14 +0000 (12:59 +0000)]
- Factor out ieee80211_reset_state() from ieee80211_newstate(), so
802.11 layer can clean up its internal state machine without state
changing or going through driver code.
- Merge sta_disassoc() and sta_disauth() into sta_disconnect(), which
sends disassoc to associated STA and deauth to authenticating STA.
- Call ieee80211_reset_state(), if operational mode is going to be
changed, so we will do a proper state machine clean up. In the
original way, if we are acting as AP and switch directly to act as
STA, we will send out wrong management frames.
- Call ieee80211_reset_state(), if PHY is going to be changed. This
prevents panic (due to a proper assertion) under following scenario:
o Current PHY of the AP is 11b.
o An 11b STA joins the BSS created by the AP. Since AP's current
PHY is 11b, this STA will not be counted as long slot time STA,
i.e. long slot time STA count left untouched.
o Switch AP's PHY to 11g.
AP will panick (long slot time STA count is zero, while there is a
long slot time STA associated, in ieee80211_node_leave_11g()). The
assertion fails, because 802.11 state machine is doing its clean up
as if AP were using 11g PHY, but it actually used 11b PHY.
Sepherosa Ziehau [Thu, 26 Apr 2007 11:58:10 +0000 (11:58 +0000)]
- Don't disable interrupt in bge_intr(), just acknowledge it. This avoids
potential lose of interrupt when interrupt gets re-enabled, since writing
to BGE_MBX_IRQ0_LO also serves as interrupt acknowledgment.
- Move interrupt acknowledgment before reading status block. This lets us
read the correct/complete status block.
- Reset ifnet.if_timer, only if there is nothing pending to be TXed.
Obtained-from: FreeBSD (bde@freebsd.org)
Sascha Wildner [Thu, 26 Apr 2007 08:35:52 +0000 (08:35 +0000)]
Add a manual page explaining the format of /etc/manpath.config.
Sascha Wildner [Thu, 26 Apr 2007 08:33:24 +0000 (08:33 +0000)]
Sort alphabetically.
Sascha Wildner [Thu, 26 Apr 2007 08:25:41 +0000 (08:25 +0000)]
Revert my previous commit to nrelease/Makefile and add /usr/local/share/man
to manpath.config.
Requested-by: dillon
Sascha Wildner [Thu, 26 Apr 2007 08:17:54 +0000 (08:17 +0000)]
Enable manpath(1) to map more than one man dir to a bin dir.
Matthew Dillon [Thu, 26 Apr 2007 02:11:00 +0000 (02:11 +0000)]
Move syslink_desc to sys/syslink_rpc.h so kernel code does not need
to #include sys/syslink.h. Add a kernel config option 'SYSLINK' to build
with kern_syslink.c, so it can be worked on (read: broken) without
interfering with other developer's kernel builds. Add a shims file for
the syslink() system call for kernels not built with kern_syslink.c. The
shims file can be used generally for this purpose.
Matthew Dillon [Wed, 25 Apr 2007 21:55:33 +0000 (21:55 +0000)]
IPV6 type 0 route headers are a flawed design, insecure by default, and
open hosts and networks up to DOS attacks by allowing normal router IP
filtering to be bypassed. Disable them by default.
Submitted-by: Hasso Tepper <hasso@estpak.ee>
Obtained-from: OpenBSD
Sascha Wildner [Wed, 25 Apr 2007 11:45:28 +0000 (11:45 +0000)]
style(9) cleanup: Remove parameter names from prototypes.
Submitted-by: Hasso Tepper <hasso@estpak.ee>
Sascha Wildner [Wed, 25 Apr 2007 09:35:31 +0000 (09:35 +0000)]
Use .Nm
Sascha Wildner [Wed, 25 Apr 2007 09:22:06 +0000 (09:22 +0000)]
Create a symbolic link /usr/local/share/man -> /usr/local/man because
apparently stuff using autotools prefers to install its manual pages
in this location.
Observed-and-suggested-by: Hasso Tepper <hasso@estpak.ee>
Sascha Wildner [Wed, 25 Apr 2007 08:13:36 +0000 (08:13 +0000)]
Use .Fn instead of .Xr.
Sepherosa Ziehau [Tue, 24 Apr 2007 11:06:47 +0000 (11:06 +0000)]
- Cleanup header file inclusion.
- Nuke no longer needed functions.
- Nuke no longer needed fields in bge_softc.
Sepherosa Ziehau [Mon, 23 Apr 2007 15:14:37 +0000 (15:14 +0000)]
Correct hardware csum offload support for bge(4)
Obtained-from: FreeBSD (w/ modification)
Sepherosa Ziehau [Mon, 23 Apr 2007 13:15:57 +0000 (13:15 +0000)]
Mention wlan_ratectl_{onoe,sample}
Reminded-by: swildner@
Sepherosa Ziehau [Mon, 23 Apr 2007 12:35:03 +0000 (12:35 +0000)]
Nuke unnecessary setting of ural(4)'s security registers. ural(4) only
supports software encrypt/decrypt. The nuked code itself is quite
problematic, as pointed out by Sam Leffler -- wk->wk_keyix should be
replaced by the loop count.
Noticed-by: Hans Petter Selasky <hselasky@c2i.net>
Simon Schubert [Sun, 22 Apr 2007 23:03:48 +0000 (23:03 +0000)]
Add support for generic prosigns by enabling ligatures.
A ligature can be created by putting the letters in angle brackets, like `<KA>'.
Simon Schubert [Sun, 22 Apr 2007 22:04:20 +0000 (22:04 +0000)]
Add the possibility to dump the generated audio to a file instead of playing it.
This can be used to create morse mp3s from input texts, like this:
morse -o | lame -r -s 44.1 -x -m m --resample 11 - morse.mp3
Add a comment on the timing calculation.
Sascha Wildner [Sun, 22 Apr 2007 20:40:20 +0000 (20:40 +0000)]
Properly use .Cd
Sascha Wildner [Sun, 22 Apr 2007 16:29:50 +0000 (16:29 +0000)]
Fix function name.
Sepherosa Ziehau [Sun, 22 Apr 2007 11:44:56 +0000 (11:44 +0000)]
Turn off Sample TX rate control algorithm support for 2661d BBP. Due to
delayed/batched TX status report, Sample does not play well with this kind
of BBP. Use Onoe TX rate control algorithm for 2661d BBP.
YONETANI Tomokazu [Sun, 22 Apr 2007 10:54:43 +0000 (10:54 +0000)]
Put a DELAY in the AT-style keyboard detection loop to wait
certain amount of time without relying on ISA bus timing.
Suggested-by: Matt
DragonFly-Bug: http://bugs.dragonflybsd.org/issue573
YONETANI Tomokazu [Sun, 22 Apr 2007 10:43:00 +0000 (10:43 +0000)]
When enabling A20, put upper limit on amount of time we wait for the
keyboard controller to get ready (65K x ISA access time, visually
around 1 second). If we have wait more than that amount it's likely
that the hardware is a legacy-free one and simply doesn't have keyboard
controller and doesn't require enabling A20 at all.
Taken-from: FreeBSD (with modifications suggested by Matt)
relavant commit logs(and discussion):
http://docs.FreeBSD.org/cgi/mid.cgi?
200604110439.k3B4dTOD072774
http://docs.FreeBSD.org/cgi/mid.cgi?
200604112053.k3BKrnC1053262
http://docs.FreeBSD.org/cgi/mid.cgi?
200604260605.k3Q65GZZ095546
Simon Schubert [Sun, 22 Apr 2007 10:22:32 +0000 (10:22 +0000)]
Pull morse(6) into the new millenium and teach it to use sound(4).
Feedback-by: swildner@
Sepherosa Ziehau [Sun, 22 Apr 2007 09:14:46 +0000 (09:14 +0000)]
- Define macro for the number of channels. Use it instead of the magic
number.
- Define maximum and default channel TX power.
- The EEPROM regions, which contain 5GHz and 2GHz channels' TX power, are
not contiguous. Add rt2661_read_txpower_config() to read them seperately.
- Lower channel's TX power is in the lower byte of the 16bits EEPROM value,
while upper channel's TX power is in the upper byte of the 16bits EEPROM
value.
- Channel's TX power should never exceed maximum channel TX power, if it is
set it to default channel TX power.
- Rename rt2661_read_eeprom() to rt2661_read_config(), since we already have
rt2661_eeprom_read().
EEPROM layout, EEPROM values' layout and various limit values are obtained
from Ralink RT61 Linux driver.
Sascha Wildner [Sun, 22 Apr 2007 07:57:31 +0000 (07:57 +0000)]
Fix typo and use section numbers for .Xr.
Sepherosa Ziehau [Sun, 22 Apr 2007 05:18:38 +0000 (05:18 +0000)]
TXRX_CSR5 stores allowable ACK rates instead of basic rates, so rename
rt2661_set_basicrates() to rt2661_set_ackrates(). Also turn on 6Mbits/s,
12Mbits/s and 24Mbits/s in ACK rates, if we are not forced to be in 11b
mode and to start/join an ERP BSS. Now rt2661 part will send ACK at
24Mbits/s instead of 6Mbits/s, if the data to be acknowledged are sent at
rate above/equal 24Mbits/s, which is standard conforming and improves
performance a bit.
# This problem is identified by running following command on any NIC
# that can capture control frames and is aware of ERP:
# tcpdump -ni iface -y ieee802_11_radio not wlan type mgt
# Pay attention to the rate, at which ACK gets sent ;]
Sepherosa Ziehau [Sun, 22 Apr 2007 04:35:12 +0000 (04:35 +0000)]
Const-fy rate set argument of ieee80211_iserp_rateset().
Sepherosa Ziehau [Sun, 22 Apr 2007 04:16:26 +0000 (04:16 +0000)]
Busdma(9)-fy, mainly obtained from FreeBSD:
if_bge.c rev 1.{45,76,81,85,101,105,106,109,151}
if_bgereg.h rev 1.{19,40}
Local changes:
- Factor out bge_dma_block_{alloc,free}() to avoid code duplication.
- Move jumbo buffer pool allocation/free into bge_dma_{alloc,free}().
- In bge_newbuf_std(), use bus_dmamap_load_mbuf() instead of bus_dmamap_load().
- Nuke vm header files inclusion.
- In bge_start():
o Bail out if OACTIVE is turned on or if bge(4) is not running yet.
o Move ifq_dequeue() before bge_encap(), since m_defrag() in bge_encap()
will trash the mbuf parameter passed in.
o Make sure there are enough TX descriptors, before dequeuing mbuf and
calling bge_encap().
- Don't call ifnet.if_start() on very interrupt, but at the end of bge_txeof().
- In bge_txeof(), if there are not enough TX descriptors, then don't clear
OACTIVE.
- Minor style changes.
Matthew Dillon [Sun, 22 Apr 2007 04:08:59 +0000 (04:08 +0000)]
sbappendcontrol() was improperly setting sb_lastmbuf, creating a possible
mbuf leak and/or sb_cc count mismatch. The passed mbuf 'm' is a chain,
not a single mbuf, and must be iterated through to locate the last mbuf.
Optimize sbappend(). M_EOR is always set in the last mbuf and the sockbuf's
sb_lastmbuf and sb_lastrecord fields can be used instead of iterating
through all the mbufs in the sockbuf. Only UNIX domain stream sockets used
this function. Normal TCP connections use the already-optimized
sbappendstream() call.
Properly propogate M_EOR in sbappendcontrol() and sbappendaddr().
Do various code and comment cleanups.
Matthew Dillon [Sun, 22 Apr 2007 01:25:04 +0000 (01:25 +0000)]
Sync netstat up to the sockbuf changes.
Matthew Dillon [Sun, 22 Apr 2007 01:13:17 +0000 (01:13 +0000)]
Give the sockbuf structure its own header file and supporting source file.
Move all sockbuf-specific functions from kern/uipc_socket2.c into the new
kern/uipc_sockbuf.c and move all the sockbuf-specific structures from
sys/socketvar.h to sys/sockbuf.h.
Change the sockbuf structure to only contain those fields required to
properly management a chain of mbufs. Create a signalsockbuf structure
to hold the remaining fields (e.g. selinfo, mbmax, etc).
Change the so_rcv and so_snd structures in the struct socket from
a sockbuf to a signalsockbuf.
Remove the recently added sorecv_direct structure which was being used
to provide a direct mbuf path to consumers for socket I/O. Use the newly
revamped sockbuf base structure instead. This gives mbuf consumers
direct access to the sockbuf API functions for use outside of a struct
socket. This will also allow new API functions to be added to the sockbuf
interface to ease the job of parsing data out of chained mbufs.
Matthew Dillon [Sun, 22 Apr 2007 00:59:27 +0000 (00:59 +0000)]
Just throw all the main arguments for syslink() into syslink_info and
pass the structure. Do not pass the descriptor separately, do not pass
a pointer to the structure size (just pass the size directly). The
search routines just return one structure at a time so a return size
field is not needed.
Start revamping syslink() to make it more mbuf-centric. This work is
very much still in progress.
Sascha Wildner [Sat, 21 Apr 2007 22:42:22 +0000 (22:42 +0000)]
Don't use .Xr for pathnames.
Peter Avalos [Sat, 21 Apr 2007 20:18:21 +0000 (20:18 +0000)]
Merge from vendor branch TNFTP:
Sync with NetBSD. In addition to some bug fixes, this brings in -s srcaddr
functionality -- using srcaddr as the local IP address for the connection.
Peter Avalos [Sat, 21 Apr 2007 20:18:21 +0000 (20:18 +0000)]
Sync with NetBSD. In addition to some bug fixes, this brings in -s srcaddr
functionality -- using srcaddr as the local IP address for the connection.
Peter Avalos [Sat, 21 Apr 2007 17:44:12 +0000 (17:44 +0000)]
Fix a memory leak in the uname/gname lookup cache.
Obtained-from: FreeBSD
Sascha Wildner [Sat, 21 Apr 2007 15:34:26 +0000 (15:34 +0000)]
Remove argument to .Os.
Sascha Wildner [Sat, 21 Apr 2007 15:25:18 +0000 (15:25 +0000)]
Bump .Dd for the -L option and add -L to the SYNOPSIS.
Matthew Dillon [Sat, 21 Apr 2007 02:26:48 +0000 (02:26 +0000)]
Convert all pr_usrreqs structure initializations to the .name = data format.
Matthew Dillon [Fri, 20 Apr 2007 22:20:12 +0000 (22:20 +0000)]
Make 'last mounted on' reporting by fsck work again. Add a new option
to fsck (-L) which dumps the last mounted on field.
Submitted-by: Dave Hayes <dave@jetcafe.org>
Simon Schubert [Fri, 20 Apr 2007 20:36:34 +0000 (20:36 +0000)]
Create the right symlinks for libpthread.so.0.
Submitted-by: Hasso Tepper <hasso@estpak.ee>
Sascha Wildner [Fri, 20 Apr 2007 11:00:56 +0000 (11:00 +0000)]
Fix link names with $DESTDIR.
Sascha Wildner [Fri, 20 Apr 2007 08:04:20 +0000 (08:04 +0000)]
1) Remove redundant -print's in EXAMPLES.
2) Remove ports specific example.
[2] Spotted-by: Trevor Kendall <trevorjkendall@gmail.com>
Sascha Wildner [Fri, 20 Apr 2007 07:04:09 +0000 (07:04 +0000)]
Add a DEVICES section and vkd(4)/vke(4) MLINKS.
Sascha Wildner [Fri, 20 Apr 2007 06:24:42 +0000 (06:24 +0000)]
Make hw.vkeX.tap_unit sysctl read only.
Matthew Dillon [Fri, 20 Apr 2007 05:42:25 +0000 (05:42 +0000)]
Clean up the so_pru_soreceive() API a bit to make it easier to read
mbuf chains without having to use a fake UIO.
Matthew Dillon [Thu, 19 Apr 2007 19:06:01 +0000 (19:06 +0000)]
We have a few generation sets for Red-Black trees that implement RLOOKUP
(ranged-lookup) for fixed numeric fields. Add another one implementing
a generic RLOOKUP where the comparison function is passed to the generator.
Simon Schubert [Thu, 19 Apr 2007 12:52:29 +0000 (12:52 +0000)]
Error out when parsing invalid IPv6 addresses.
Submitted-by: Tatsuya BIZENN <bizenn@visha.org>
Obtained-from: NetBSD (supposedly)
DragonFly-Bug: http://bugs.dragonflybsd.org/issue591
Matthew Dillon [Thu, 19 Apr 2007 03:16:33 +0000 (03:16 +0000)]
Allocations of size greater then the radix were not returning the correct
starting block. For example, the second allocation of 32 blocks was returning
block 1 instead of block 32.
Also remove a redundant check in the code that was wasting cycles.
Submitted-by: Jason Smethers <jason@smethers.net>
Sascha Wildner [Wed, 18 Apr 2007 18:39:11 +0000 (18:39 +0000)]
Remove 'register'.
Sascha Wildner [Wed, 18 Apr 2007 18:32:12 +0000 (18:32 +0000)]
Ansify parameter declarations and fix minor style issues.
Simon Schubert [Tue, 17 Apr 2007 23:20:54 +0000 (23:20 +0000)]
Build libc_r before other libs, particularly before libpthread.
Fixes world breakage reported by tuxillo, aggelos and chlamydia.
Simon Schubert [Tue, 17 Apr 2007 22:25:45 +0000 (22:25 +0000)]
Point P_tmpdir to /tmp instead of /var/tmp.
Temp files are much better of placed in /tmp than in the not-so-volatile
/var/tmp. Additionally, people might get space issues or might have tuned
their /tmp for speed, but not /var/tmp.
Libiberty, among other consumers is using P_tmpdir to place the temp files,
leading compiler temp files being created in /var/tmp.
Matthew Dillon [Tue, 17 Apr 2007 17:28:04 +0000 (17:28 +0000)]
Don't allow snd_nxt to be set to a value less then snd_una when restoring
snd_nxt in the TCP limited transmit code. This can occur if the code
blocks and acks are processed prior to restoration.
This bug fix is currently unverified. The possible trigger is running
netstat -an on a machine very heavily loaded with 6000+ network connections.
Reported-by: Peter Avalos <pavalos@theshell.com>
Simon Schubert [Tue, 17 Apr 2007 12:34:07 +0000 (12:34 +0000)]
Add a proxy libpthread.
In short, it works like this: ld uses libpthread.so when linking. There
the SONAME is set to libpthread.so.0. This however is a symlink (or maybe
varsymlink) to one of the threading libs. Voila, threading libs can be
switched at execution time and we won't get breakage because of two different
threading libs being used in the same binary.
Idea-by: joerg and me.
Simon Schubert [Tue, 17 Apr 2007 12:18:37 +0000 (12:18 +0000)]
Change the compilers to alias -pthread to -lpthread.
Now threaded programs do not get linked to libc_r anymore, but to the
proxy lib libpthread. As soon as all programs and libraries are relinked
to use libpthread exclusively (and not libc_r.so anymore), threading libs
can be switched by changing the /usr/lib/libpthread.so.0 symlink.
Use the following command to find all packages still linking to libc_r:
find /usr/pkg/bin /usr/pkg/sbin /usr/pkg/lib /usr/pkg/libexec -type f \
-exec sh -c '\
objdump -p $0 2>/dev/null | grep -q "NEEDED.*libc_r\.so" || exit 0; \
pkg_info -F -e $0' {} \; | sort | uniq
Matthew Dillon [Mon, 16 Apr 2007 17:40:16 +0000 (17:40 +0000)]
Probably the last change to the syslink() system call. Allow a generic
structure to be passed and returned and revamp the command structure.
Matthew Dillon [Mon, 16 Apr 2007 17:36:04 +0000 (17:36 +0000)]
Bring in the skeleton infrastructure and manual page for the new syslink
utility.
Sepherosa Ziehau [Sun, 15 Apr 2007 12:50:49 +0000 (12:50 +0000)]
Expose number of TX/RX descriptors through read-only sysctl hw.emX.{txd,rxd}.
Submitted-by: Hasso Tepper <hasso@estpak.ee>
Approved-by: dillon@
Gregory Neil Shapiro [Sat, 14 Apr 2007 17:43:48 +0000 (17:43 +0000)]
Change build infrastructure over to sendmail 8.14.1.
Note that mail filters (aka, milters) compiled against the libmilter
included in the base operating system should be recompiled.
Sepherosa Ziehau [Sat, 14 Apr 2007 05:14:40 +0000 (05:14 +0000)]
For bge_rxeof(), return immediately if no RX descs need to be processed, this
optimization avoids two unnecessary register accessing and increase performance
a lot. Same optimization is applied to bge_txeof(), but it will not have
real effect until bge(4) uses busdma(4) (i.e. will avoid extra sync)
Obtained-from: FreeBSD (oleg@freebsd.org)
Sepherosa Ziehau [Sat, 14 Apr 2007 04:35:10 +0000 (04:35 +0000)]
- Return error if reading eeprom times out.
Obtained-from: FreeBSD
- Strip extra blank lines in bge_attach().
Sepherosa Ziehau [Sat, 14 Apr 2007 04:22:14 +0000 (04:22 +0000)]
BGE_MBX_TX_HOST_PROD0_LO is write-only, avoid reading it.
Obtained-from: NetBSD via FreeBSD
Simon Schubert [Fri, 13 Apr 2007 12:26:56 +0000 (12:26 +0000)]
Switch binutils default to 2.17 and remove old gcc2 code.
Simon Schubert [Fri, 13 Apr 2007 12:26:42 +0000 (12:26 +0000)]
Hook binutils-2.17 into the build.
Simon Schubert [Fri, 13 Apr 2007 12:24:33 +0000 (12:24 +0000)]
Add build infrastructure for binutils-2.17.
Simon Schubert [Fri, 13 Apr 2007 12:12:27 +0000 (12:12 +0000)]
Convert assembly which accesses segment descriptors to use 16bit ops.
Simon Schubert [Fri, 13 Apr 2007 12:08:19 +0000 (12:08 +0000)]
Add our READMEs.
Simon Schubert [Fri, 13 Apr 2007 12:05:44 +0000 (12:05 +0000)]
Merge from vendor branch BINUTILS:
Import binutils-2.17.
Simon Schubert [Fri, 13 Apr 2007 12:05:44 +0000 (12:05 +0000)]
Import binutils-2.17.
YONETANI Tomokazu [Fri, 13 Apr 2007 02:51:34 +0000 (02:51 +0000)]
fix critical-section mismatch
Sascha Wildner [Thu, 12 Apr 2007 19:50:20 +0000 (19:50 +0000)]
Ansify parameter declarations.
In-collaboration-with: Alexey Slynko <slynko@tronet.ru>
Sascha Wildner [Thu, 12 Apr 2007 18:35:09 +0000 (18:35 +0000)]
Ansify parameter declarations.
Sepherosa Ziehau [Thu, 12 Apr 2007 12:54:07 +0000 (12:54 +0000)]
When read BBP registers, avoid writing to BBPCSR until it is no longer busy.
After this bug fixing, TX/RX antenna setup can be safely put after BBP
initialization, which is a correct place for it, since BBP initialization
will overwrite RX antenna BBP register with default value. Before this bug
fixing, putting TX/RX antenna setup after BBP initailization always results
in strange TX/RX problems, which I experienced when I fiddled with my ASUS
WL-107G; and some OpenBSD folks had this problems too, before Damien reverted
related changes in OpenBSD.
2526 RF does not require special I/Q flipping during TX/RX antenna setup.
Obtained-from: Ralink RT2500 Linux driver
Gregory Neil Shapiro [Thu, 12 Apr 2007 03:30:24 +0000 (03:30 +0000)]
Add DragonFly instructions file to new version directory.
Note that I have changed to using a single directory for the major
version instead of a new directory for each minor version. This
is a better method of tracking versions IMHO.
Gregory Neil Shapiro [Thu, 12 Apr 2007 03:26:51 +0000 (03:26 +0000)]
Merge from vendor branch SENDMAIL:
Import sendmail 8.14.1
Gregory Neil Shapiro [Thu, 12 Apr 2007 03:26:51 +0000 (03:26 +0000)]
Import sendmail 8.14.1
Simon Schubert [Wed, 11 Apr 2007 22:49:22 +0000 (22:49 +0000)]
Disable ar support in libarchive due to buffer overflow bugs.
Submitted-by: joerg@
Matthew Dillon [Wed, 11 Apr 2007 21:04:09 +0000 (21:04 +0000)]
Kernel virtual memory must be mapped on a segment address boundary. Try
numerous addresses instead of a single fixed address to accomodate systems
that might have been compiled with a different MAXDSIZ.
Matthew Dillon [Wed, 11 Apr 2007 18:51:33 +0000 (18:51 +0000)]
Use SHUT_RD instead of a hardcoded value of 0 in calls to shutdown().
Submitted-by: "Nuno Antunes" <nuno.antunes@gmail.com>
Obtained-from: FreeBSD
Sascha Wildner [Wed, 11 Apr 2007 18:00:06 +0000 (18:00 +0000)]
Remove ldconfig_paths_aout which was removed from the system.
Matthew Dillon [Wed, 11 Apr 2007 17:46:39 +0000 (17:46 +0000)]
Remove ldconfig_paths_aout, it is no longer used.
Add /usr/pkg/lib to the default ldconfig_paths and remove /usr/X11R6/lib.
The path is now: /usr/lib/compat /usr/pkg/lib /usr/pkg/xorg/lib /usr/local/lib
pkgsrc packages use rpaths - absolute paths, and do not need ldd library
search paths. However, we do want these libraries to be easily accessible
to manually compiled programs and until GCC starts using rpaths by default,
maintaining a default list is a good idea.
Submitted-by: Hasso Tepper <hasso@estpak.ee>
Sascha Wildner [Wed, 11 Apr 2007 09:12:08 +0000 (09:12 +0000)]
<> -> ()
Sascha Wildner [Wed, 11 Apr 2007 09:07:33 +0000 (09:07 +0000)]
Misc. mdoc fixes.
Sascha Wildner [Wed, 11 Apr 2007 06:51:35 +0000 (06:51 +0000)]
Don't use \*[Px] because it leads to an unwanted font size change.
Found-in: FreeBSD
Sascha Wildner [Tue, 10 Apr 2007 21:09:00 +0000 (21:09 +0000)]
Swap "underflow" and "overflow" in table header.
Obtained-from: FreeBSD
Sascha Wildner [Mon, 9 Apr 2007 21:20:38 +0000 (21:20 +0000)]
Separate punctuation.
Sascha Wildner [Mon, 9 Apr 2007 20:47:01 +0000 (20:47 +0000)]
Oops, forgot to remove parentheses.
Sascha Wildner [Mon, 9 Apr 2007 20:40:46 +0000 (20:40 +0000)]
Use .Fn for functions, fix .Nm usage.
Matthew Dillon [Mon, 9 Apr 2007 17:10:00 +0000 (17:10 +0000)]
Add subr_alist.c. This is a bitmap allocator that works very similarly to
subr_blist.c (swap allocator), but with added considerations.
1. All allocations must be in powers of 2.
2. All allocations will be aligned to the allocation size.
3. No allocation size limit (blist was limited to 32 blocks per allocation)
Like the blist allocator, the alist is arranged in a linear array suitable
for direct mapping onto a storage medium. A dataspace of 2^31-1 blocks may
be represented. Approximately 3 bits of kernel memory is used per block.
This allocator will be used by HAMMER and ANVIL (filesystem and filesystem
storage manager), and by syslink route nodes to chop out individual addresses
and subnets. We may also use this allocator to improve the allocation of
physical memory.
Sascha Wildner [Mon, 9 Apr 2007 09:45:12 +0000 (09:45 +0000)]
Remove redundant -Wall option.
Sascha Wildner [Mon, 9 Apr 2007 09:15:48 +0000 (09:15 +0000)]
* Raise WARNS to 6 and fix all warnings.
* Do some style(9) cleanup.
Sascha Wildner [Sun, 8 Apr 2007 14:56:37 +0000 (14:56 +0000)]
Remove obsolete header files.
Sepherosa Ziehau [Sun, 8 Apr 2007 12:03:18 +0000 (12:03 +0000)]
Fix comment: 1000baseTX -> 1000baseT
Sepherosa Ziehau [Sun, 8 Apr 2007 09:43:57 +0000 (09:43 +0000)]
Nuke unnecessary NULL node checks.
Sepherosa Ziehau [Sun, 8 Apr 2007 09:41:41 +0000 (09:41 +0000)]
Code cleanup:
- Nuke commented out spl*() and function declaration.
- Break long comment lines in rum_ioctl().
- Nuke unnecessary NULL node check in rum_start().
Resource alloc/free fixes and cleanup:
- Fix several resource leakages:
o In rum_start(), decrease node refcnt on error handling path.
o In rum_stop(), free rum_softc.stats_xfer.
o In rum_tx_free_list(), free possible pending TX mbuf.
- In rum_tx_data(), if usbd_transfer() fails, clear rum_tx_data.m and
rum_tx_data.ni, so later resource free routine will not double free
them.
- Change the redundant resource free code in rum_detach() into assertion
to make sure everything are clear.
- In rum_alloc_{tx,rx}_list(), don't call rum_free_{tx,rx}_list, if
error happens, since caller will do that in rum_stop().
Rest of changes are used to fix following problems:
1) Random system hanging if following script is excuted:
#!/bin/csh
while (1)
ifconfig rum0 up
ifconfig rum0 down
end
2) Random system hanging during 'ifconfig rum0 down', if rum(4) is
attached to uhci/ohci.
3) Promised system hanging if wpa_supplicant(8) is used and killed after
connection to AP is established, and rum(4) is attached to uhci/ohci.
4) Random serializer recursive assertion, if rum(4) is under heavy load
and brought down.
- In rum_stop(), turn on RUM_FLAG_STOPPED, so USB transanction will
not be restarted in rum_rxeof() and rum_stats_timeout() during device
halting.
- In rum_stop(), release ifnet.if_serializer before aborting RX/TX pipe.
This make sure serializer will not be recursive held,
e.g. abort RX pipe -> rum_rxeof().
- Don't turn off TSF sync, when 802.11 state transits to INIT. Since
o Chip will be reset immediately after INIT state transition, so
turning off TSF sync does not make much sense.
o If rum(4) is under heavy RX/TX load, turning off TSF sync will
stall various USB operations, thus hang the whole system.
- After above change, perform INIT state transition in rum_newstate(),
so that 802.11 state machine is promised to be halted during detach
and stop routine.
- Use critical section and RUM_FLAG_CONFIG to protect various USB operation
sequences, which are not intended to be interfered.