]> CyberLeo.Net >> Repos - FreeBSD/FreeBSD.git/log
FreeBSD/FreeBSD.git
15 years agoRemove ATH_SUPPORT_TDMA and use IEEE80211_SUPPORT_TDMA instead. It
Sam Leffler [Mon, 30 Mar 2009 19:23:49 +0000 (19:23 +0000)]
Remove ATH_SUPPORT_TDMA and use IEEE80211_SUPPORT_TDMA instead.  It
doesn't make much sense to configure driver support w/o net80211.
Note this means ath now depends on opt_wlan.h.

15 years agoWe don't know what these pages are going to be used for, they should be
Robert Noland [Mon, 30 Mar 2009 18:01:42 +0000 (18:01 +0000)]
We don't know what these pages are going to be used for, they should be
un-cached.  This got lost somewhere with all the bus_dma fixups.

MFC after: 3 days

15 years agoAdd a simple manage for the refcount(9) API.
John Baldwin [Mon, 30 Mar 2009 18:01:12 +0000 (18:01 +0000)]
Add a simple manage for the refcount(9) API.

Requested by: rwatson

15 years agoLoad the right microcode for RS780.
Robert Noland [Mon, 30 Mar 2009 17:49:21 +0000 (17:49 +0000)]
Load the right microcode for RS780.

MFC after: 3 days

15 years agoTwo fixes:
Warner Losh [Mon, 30 Mar 2009 17:25:56 +0000 (17:25 +0000)]
Two fixes:
(1) Delete all children when detaching to keep from adding a phy each
    driver reload.
(2) All AX88x90 chips have the RST issue.

15 years agoThis is a major reworking of the AX88x90 support.
Warner Losh [Mon, 30 Mar 2009 16:15:06 +0000 (16:15 +0000)]
This is a major reworking of the AX88x90 support.

o Introduce new chip_type AX88790.  There's a few places we need to know the
  exact chip for workaronds.
o Explain the AX88190 workaround for the ISR bits being stuck, and don't
  apply them to the AX88790.  The datasheet says the bits are fixed, and
  experience confirms.
o Fix mii bit-bang read code to read and discard the 'floating' bit.
o Remove empty ed_pccard_ax88x90_mii_reset routine
o Report error from mii_phy_probe
o Don't use ed_probe_Novel_generic for ax88x90 chips.  It puts them into
  an odd state sometimes.  Instead, use a more stream-lined version that
  avoids the trouble spots.  This was copied and tweaked from the original.
o Move chip reset into its own routine.
o Minor code optimiation on getting MAC address
o Add code for coping with AX88790 cards that are in power down state and
  need to be kicked before the PHY registers for the internal phy read right.
o Remove ugly cap of PHYs at 17.
o For AX88790, we need to set a special bit for accessig phy 16 (the internal
  phy) and clear it for all others according to a chip erratum.
o streamline the bit-bang code for AX88x90: the delays aren't needed according
  to the datasheet timing diagrams and also the Linux driver
o Fix minor bit definition for direction bit.
o Generally: Some comments reformatted
o Only try the toshiba probe on cards labelled as toshiba

# From another Akihabara card (this one from a few years ago from a
# friend in Japan).  Fix the Corega FEther II PCC-TXD.  This one is
# still on sale new, as of a few weeks ago.  should fix all other AX88x90
# based cards, but I have some testing left to finish on my collection...

15 years agoNew PHY driver for the internal PHY found in the AX88790. There's a
Warner Losh [Mon, 30 Mar 2009 16:01:09 +0000 (16:01 +0000)]
New PHY driver for the internal PHY found in the AX88790.  There's a
number of quirks for this device, and this implements just the basics.
The 2.5s powerdown recommended in the datasheet will be next...

15 years agoFix memory leak in semunload().
Christian Brueffer [Mon, 30 Mar 2009 15:01:29 +0000 (15:01 +0000)]
Fix memory leak in semunload().

PR: 133064
Submitted by: Mateusz Guzik <mjguzik@gmail.com>
MFC after: 1 week

15 years agowhen it failed to inform rx/tx dma setup it should clean up allocated
Weongyo Jeong [Mon, 30 Mar 2009 11:51:05 +0000 (11:51 +0000)]
when it failed to inform rx/tx dma setup it should clean up allocated
dma-related buffers and tags.

MFC after: 3 days

15 years agocorrects a error message.
Weongyo Jeong [Mon, 30 Mar 2009 11:23:14 +0000 (11:23 +0000)]
corrects a error message.

15 years agohandles more exceptional cases when the driver failed to attach.
Weongyo Jeong [Mon, 30 Mar 2009 08:48:33 +0000 (08:48 +0000)]
handles more exceptional cases when the driver failed to attach.

MFC after: 3 days

15 years agoImplement support for RTLD_NODELETE flag for dlopen() and -z nodelete
Konstantin Belousov [Mon, 30 Mar 2009 08:47:28 +0000 (08:47 +0000)]
Implement support for RTLD_NODELETE flag for dlopen() and -z nodelete
static linker option. Do it by incrementing reference count on the loaded
object and its dependencies.

Reviewed by: davidxu, kan

15 years agofix a bug of uses after free.
Weongyo Jeong [Mon, 30 Mar 2009 08:39:42 +0000 (08:39 +0000)]
fix a bug of uses after free.

Pointed by: dchagin
MFC after: 3 days

15 years agoFurther rate limit the root wait status, it will be printed once per
Andrew Thompson [Mon, 30 Mar 2009 05:57:55 +0000 (05:57 +0000)]
Further rate limit the root wait status, it will be printed once per
root_mount_rel() wakeup.

15 years agoAdd PHY entry for the ASIX 88x90 internal PHYs.
Warner Losh [Mon, 30 Mar 2009 01:47:32 +0000 (01:47 +0000)]
Add PHY entry for the ASIX 88x90 internal PHYs.

15 years agoSharpen the saw:
Marcel Moolenaar [Mon, 30 Mar 2009 01:03:58 +0000 (01:03 +0000)]
Sharpen the saw:
o  PC98 uses 32-bit block numbers. Limit the scheme to 2^32-1
   blocks when the media is larger. The 32-bit block numbers
   are implicit (16-bit cylinder * 8-bit head * 8-bit sector).

15 years agoSharpen the saw:
Marcel Moolenaar [Mon, 30 Mar 2009 00:53:46 +0000 (00:53 +0000)]
Sharpen the saw:
o  MBR uses 32-bit block numbers. Limit the scheme to 2^32-1
   blocks when the media is larger.

15 years agoSharpen the saw:
Marcel Moolenaar [Mon, 30 Mar 2009 00:48:42 +0000 (00:48 +0000)]
Sharpen the saw:
o  EBR uses 32-bit block numbers. Limit the scheme to 2^32-1
   blocks when the media is larger.
o  Calculate the number of entries based on the rounded media
   size, rather than the raw media size.

15 years agoSharpen the saw:
Marcel Moolenaar [Mon, 30 Mar 2009 00:33:43 +0000 (00:33 +0000)]
Sharpen the saw:
o  Don't create a GPT scheme underneath another scheme when
   the probe doesn't allow it.

15 years agoReplace v_dd vnode pointer with v_cache_dd pointer to struct namecache
Alexander Kabaev [Sun, 29 Mar 2009 21:25:40 +0000 (21:25 +0000)]
Replace v_dd vnode pointer with v_cache_dd pointer to struct namecache
in directory vnodes. Allow namecache dotdot entry to be created pointing
from child vnode to parent vnode if no existing links in opposite
direction exist. Use direct link from parent to child for dotdot lookups
otherwise.

This restores more efficient dotdot caching in NFS filesystems which
was lost when vnodes stoppped being type stable.

Reviewed by: kib

15 years agoo add ic_rt to track the rate table for the current channel; this enables
Sam Leffler [Sun, 29 Mar 2009 21:17:08 +0000 (21:17 +0000)]
o add ic_rt to track the rate table for the current channel; this enables
  calculation of packet transmit times to do things like check txop limits
o remove equivalent driver code and convert to use net80211 state

15 years agoFix typo.
Christian Brueffer [Sun, 29 Mar 2009 21:08:48 +0000 (21:08 +0000)]
Fix typo.

15 years agotoggling fast-frames requires clocking the state machine
Sam Leffler [Sun, 29 Mar 2009 21:00:27 +0000 (21:00 +0000)]
toggling fast-frames requires clocking the state machine

15 years agoEmulate the FIODGNAME ioctl in our 32-bit emulator.
Ed Schouten [Sun, 29 Mar 2009 20:09:51 +0000 (20:09 +0000)]
Emulate the FIODGNAME ioctl in our 32-bit emulator.

It's quite strange that nobody reported this issue before. It turns out
functions like ttyname(), ptsname() and fdevname() don't work in
compat32. This means it't not even possible to run applications like
script(1) inside a 32-bit FreeBSD jail.

Fix this by converting 32-bit fiodgname_arg structures to their 64-bit
equivalent.

Reported by: kris
Tested by: kris

15 years agoremove bogus nwid use; that's a compat shim for netbsd
Sam Leffler [Sun, 29 Mar 2009 18:14:45 +0000 (18:14 +0000)]
remove bogus nwid use; that's a compat shim for netbsd

15 years agofix wired-wireless failover example and remove incorrect
Sam Leffler [Sun, 29 Mar 2009 18:05:04 +0000 (18:05 +0000)]
fix wired-wireless failover example and remove incorrect
comment about WPA not working

15 years agoEliminate ic_myaddr so changing the mac address of a device works correctly:
Sam Leffler [Sun, 29 Mar 2009 17:59:14 +0000 (17:59 +0000)]
Eliminate ic_myaddr so changing the mac address of a device works correctly:
o remove ic_myaddr from ieee80211com
o change ieee80211_ifattach to take the mac address of the physical device
  and use that to setup the lladdr.
o replace all references to ic_myaddr in drivers by IF_LLADDR
o related cleanups (e.g. kill dead code)

PR: kern/133178
Reviewed by: thompsa, rpaulo

15 years agoQuery DNS only once per an address family.
Hajimu UMEMOTO [Sun, 29 Mar 2009 17:55:11 +0000 (17:55 +0000)]
Query DNS only once per an address family.

Obtained from: KAME
MFC after: 2 weeks

15 years agoGet rid of VSTAT and replace it with VSTAT_PERMS, which is somewhat
Edward Tomasz Napierala [Sun, 29 Mar 2009 17:45:48 +0000 (17:45 +0000)]
Get rid of VSTAT and replace it with VSTAT_PERMS, which is somewhat
better defined.

Approved by: rwatson (mentor)

15 years agoUnbreak OF_interpret() and its standard implementation after r186347.
Marius Strobl [Sun, 29 Mar 2009 15:10:49 +0000 (15:10 +0000)]
Unbreak OF_interpret() and its standard implementation after r186347.

15 years agoExtend comment in copyright notice as requested by author.
Michael Reifenberger [Sun, 29 Mar 2009 13:35:20 +0000 (13:35 +0000)]
Extend comment in copyright notice as requested by author.

Submitted by: G.Otsuji

15 years agoFixup relative pointers after channel realloc. It fixes crash on systems
Alexander Motin [Sun, 29 Mar 2009 07:10:52 +0000 (07:10 +0000)]
Fixup relative pointers after channel realloc. It fixes crash on systems
with several HDA codecs per controller.

While I am there, remove some unneeded dereferences.

Submitted by: Tor Egge <Tor.Egge@cvsup.no.freebsd.org>

15 years agoAdd SDT DTrace probes for VFS vnode operations in the vfs:vop
Robert Watson [Sun, 29 Mar 2009 03:30:15 +0000 (03:30 +0000)]
Add SDT DTrace probes for VFS vnode operations in the vfs:vop
provider namespace.  These are inserted dynamically into the
VOP_..._AP() functions created from vnode_if.src.  Each VOP has
entry and return probes, as arg0 the primary vnode, arg1 the
vnode operation argument structure pointer, providing access to
IN and OUT arguments, and for return probes, arg2 the return
value.

MFC after: 1 month
Sponsored by: Google, Inc.

15 years agoMark the declaration of bus_space_map 'static' as the implementation is.
Bjoern A. Zeeb [Sat, 28 Mar 2009 23:24:34 +0000 (23:24 +0000)]
Mark the declaration of bus_space_map 'static' as the implementation is.
Follow one of the two most common indent schemes in this file.
This unbreaks a few mips kernel builds.

15 years agoFor kernel builds reduce the impact of svnversion, just scanning
Bjoern A. Zeeb [Sat, 28 Mar 2009 23:17:18 +0000 (23:17 +0000)]
For kernel builds reduce the impact of svnversion, just scanning
src/sys and not the entire src/ tree.

An earlier solution by peter had been comitted in r183528 and backed out
in r183566 due to problems with newvers.sh also called from other places
during world build. With the extra test this survived a make universe.

15 years ago- Add files that should have been added in r190507.
Ulf Lilleengen [Sat, 28 Mar 2009 21:06:59 +0000 (21:06 +0000)]
- Add files that should have been added in r190507.

15 years agoUpdate to include Toshiba LANCT00A card being supported, as well as a
Warner Losh [Sat, 28 Mar 2009 20:59:25 +0000 (20:59 +0000)]
Update to include Toshiba LANCT00A card being supported, as well as a
caution that while the Mitsubishi M8895 has a DP83902, its ASIC is
undocumented and doens't work with either the NE2000 or the WD80x3
drivers.

15 years agoenable setting the mac address of 802.11 devices
Sam Leffler [Sat, 28 Mar 2009 17:36:56 +0000 (17:36 +0000)]
enable setting the mac address of 802.11 devices

15 years agoImport the gvinum work that have been done during and after Summer of Code 2007.
Ulf Lilleengen [Sat, 28 Mar 2009 17:20:08 +0000 (17:20 +0000)]
Import the gvinum work that have been done during and after Summer of Code 2007.
The work have been under testing and fixing since then, and it is mature enough
to be put into HEAD for further testing.

A lot have changed in this time, and here are the most important:
- Gvinum now uses one single workerthread instead of one thread for each
  volume and each plex. The reason for this is that the previous scheme was
  very complex, and was the cause of many of the bugs discovered in gvinum.
  Instead, gvinum now uses one worker thread with an event queue, quite
  similar to what used in gmirror.
- The rebuild/grow/initialize/parity check routines no longer runs in
  separate threads, but are run as regular I/O requests with special flags.
  This made it easier to support mounted growing and parity rebuild.
- Support for growing striped and raid5-plexes, meaning that one can extend the
  volumes for these plex types in addition to the concat type. Also works while
  the volume is mounted.
- Implementation of many of the missing commands from the old vinum:
  attach/detach, start (was partially implemented), stop (was partially
  implemented), concat, mirror, stripe, raid5 (shortcuts for creating volumes
  with one plex of these organizations).
- The parity check and rebuild no longer goes between userland/kernel, meaning
  that the gvinum command will not stay and wait forever for the rebuild to
  finish. You can instead watch the status with the list command.
- Many problems with gvinum have been reported since 5.x, and some has been hard
  to fix due to the complicated architecture. Hopefully, it should be more
  stable and better handle edge cases that previously made gvinum crash.
- Failed drives no longer disappears entirely, but now leave behind a dummy
  drive that makes sure the original state is not forgotten in case the system
  is rebooted between drive failures/swaps.
- Update manpage to reflect new commands and extend it with some examples.

Sponsored by:   Google Summer of Code 2007
Mentored by:    le
Tested by:      Rick C. Petty <rick-freebsd2008 -at- kiwi-computer.com>

15 years agoDo not dereference NULL pointer. refobj is NULL for the objects that are
Konstantin Belousov [Sat, 28 Mar 2009 15:54:08 +0000 (15:54 +0000)]
Do not dereference NULL pointer. refobj is NULL for the objects that are
preloaded.

Reported and tested by: ed

15 years agoAdd support for Phenom (Family 10h) to cpufreq.
Michael Reifenberger [Sat, 28 Mar 2009 08:54:47 +0000 (08:54 +0000)]
Add support for Phenom (Family 10h) to cpufreq.
Its a newer version provided by the author than in the PR.

PR: kern/128575
Submitted by: Gen Otsuji annona2 [at] gmail.com

15 years agoAdd some casts to silence compiler warning about signedness.
Xin LI [Sat, 28 Mar 2009 07:44:08 +0000 (07:44 +0000)]
Add some casts to silence compiler warning about signedness.

15 years agoo Don't access VPD even if hardware advertised the capability.
Pyun YongHyeon [Sat, 28 Mar 2009 07:39:35 +0000 (07:39 +0000)]
o Don't access VPD even if hardware advertised the capability.
  It seems that some revision of controller hang while accessing
  the VPD. Because VPD access routine are unused, nuke it.
o Let TWSI reload EEPROM if VPD capability is detected. Reloading
  EEPROM will also set ethernet address so age(4) now reads AGE_PAR0
  and AGE_PAR1 register to get ethernet address. This removes a lot
  of hack and enhance readability a lot.
o Double PHY reset timeout as it takes more time to take PHY out of
  power-saving state.
o Explicitly check power-saving state by checking undocumented PHY
  registers. If link is not up, poke undocumented registers to take
  PHY out of power-saving state. This is the same way what Linux
  does. On resume, make sure to wake up PHY.
o Don't rely on auto-clearing feature of master reset bit, just wait
  1ms and check idle status of MAC.
o Add PCI device revision information in bootverbose mode.
This should fix occasional controller hang in device attach phase.

Reported by: barbara < barbara.xxx1975 at libero DOT it >
Tested by: barbara < barbara.xxx1975 at libero DOT it >

15 years agoMinor changes from Berkeley DB 1.86 and further improvements from OpenBSD.
Xin LI [Sat, 28 Mar 2009 07:31:02 +0000 (07:31 +0000)]
Minor changes from Berkeley DB 1.86 and further improvements from OpenBSD.
This does not include the new hash routines since they will cause problems
when reading old hash files.

Since mpool(3) has been changed, provide a compatibility shim for older
binaries.

Obtained from: OpenBSD

15 years agoAllow O_SYNC and O_NOFOLLOW flags in dbopen().
Xin LI [Sat, 28 Mar 2009 07:26:00 +0000 (07:26 +0000)]
Allow O_SYNC and O_NOFOLLOW flags in dbopen().

Obtained from: OpenBSD

15 years agoPlug memory leaks and a potential NULL dereference.
Xin LI [Sat, 28 Mar 2009 07:20:39 +0000 (07:20 +0000)]
Plug memory leaks and a potential NULL dereference.

Obtained from: OpenBSD

15 years agoSimplify the logic when determining whether to zero out a db file to after
Xin LI [Sat, 28 Mar 2009 07:09:51 +0000 (07:09 +0000)]
Simplify the logic when determining whether to zero out a db file to after
open().  The previous logic only initializes the database when O_CREAT is
set, but as long as we can open and write the database, and the database
is empty, we should initialize it anyway.

Obtained from: OpenBSD

15 years ago - If (keysize+datasize)%(bsize=14)==0, insertion of a `big key' would cause
Xin LI [Sat, 28 Mar 2009 06:47:05 +0000 (06:47 +0000)]
 - If (keysize+datasize)%(bsize=14)==0, insertion of a `big key' would cause
   an invariant (actually, an ugly hack) to fail, and all Hell would break
   loose.

   When deleting a big key, the offset of an empty page should be bsize, not
   bsize-1; otherwise an insertion into the empty page will cause the new key to
   be elongated by 1 byte.

   Make the packing more dense in a couple of cases.

 - fix NULL dereference exposed on big bsize values;

Obtained from: NetBSD via OpenBSD

15 years agoAvoid NULL deference which causes DEBUG1 to crash, not to print useful
Xin LI [Sat, 28 Mar 2009 06:40:48 +0000 (06:40 +0000)]
Avoid NULL deference which causes DEBUG1 to crash, not to print useful
information.

Obtained from: OpenBSD

15 years ago - Avoid overwriting the cursor page when the cursor page becomes the
Xin LI [Sat, 28 Mar 2009 06:38:31 +0000 (06:38 +0000)]
 - Avoid overwriting the cursor page when the cursor page becomes the
LRU page.
 - Fix for sequential retrieval failure when using large key/data pairs.

Obtained from: OpenBSD

15 years agoFix a crash when iterating over a hash and removing its elements.
Xin LI [Sat, 28 Mar 2009 06:30:43 +0000 (06:30 +0000)]
Fix a crash when iterating over a hash and removing its elements.

Obtained from: OpenBSD

15 years agoOnly squeeze a short key/value pair onto a page with other complete key/value
Xin LI [Sat, 28 Mar 2009 06:25:33 +0000 (06:25 +0000)]
Only squeeze a short key/value pair onto a page with other complete key/value
pairs, not onto a page containing the end of a big pair.

Obtained from: NetBSD via OpenBSD

15 years agoDiff reduce against OpenBSD, no functional change.
Xin LI [Sat, 28 Mar 2009 06:23:10 +0000 (06:23 +0000)]
Diff reduce against OpenBSD, no functional change.

15 years agoWhen reading via memory, read in (amount + 1) / 2 (to properly round
Warner Losh [Sat, 28 Mar 2009 06:22:11 +0000 (06:22 +0000)]
When reading via memory, read in (amount + 1) / 2 (to properly round
up) rather than amount + 1 / 2, which is the same as amount, or 2x too
many words which leads to data corruption.

# This fixes the sbdrop panics I was seeing with the Toshiba LANCT00A.

15 years agoReturn meaningful errno in overflow case; print error message to stderr
Xin LI [Sat, 28 Mar 2009 06:12:39 +0000 (06:12 +0000)]
Return meaningful errno in overflow case; print error message to stderr
in one more case.

Obtained from: NetBSD via OpenBSD

15 years agoUse pread(2) and pwrite(2) instead of lseek(2) + read(2) / write(2).
Xin LI [Sat, 28 Mar 2009 06:05:53 +0000 (06:05 +0000)]
Use pread(2) and pwrite(2) instead of lseek(2) + read(2) / write(2).

Obtained from: NetBSD via OpenBSD

15 years agodb/btree/bt_open.c: check return value of snprintf() and return value
Xin LI [Sat, 28 Mar 2009 05:57:27 +0000 (05:57 +0000)]
db/btree/bt_open.c: check return value of snprintf() and return value
if the result is truncated.

db/hash/hash_page.c: use the same way to create temporary file as
bt_open.c; check snprintf() return value.

Obtained from: OpenBSD

15 years agoSeveral signed/unsigned warning fixes.
Xin LI [Sat, 28 Mar 2009 05:45:29 +0000 (05:45 +0000)]
Several signed/unsigned warning fixes.

15 years agoIt turns out that the Toshiba LANCT00A PC Card is really like the
Warner Losh [Sat, 28 Mar 2009 04:56:56 +0000 (04:56 +0000)]
It turns out that the Toshiba LANCT00A PC Card is really like the
Toshiba PCETC ISA card, and even has the same board type code in the
card ID (0x14).  So, for this card, call ed_probe_WD80x3_generic after
setting things up apropriately.  This makes the card attach and kinda
work (I'm seeing panics in sbdrop).  Since history has shown that the
WD80x3 probe routine is dangerous, only do it for this card.  Also,
disable the memory range check to make sure it is an valid ISA memory.
I think that it is bogus, but I'm not 100% sure, for these cards.

I removed probing for the WD80x3 in 2005 when I added support for the
AX88x90 and DL100xx cards since none of my cards had ever matched it
and PAO3 removed it and none of the cards in their database died.

It is possible there are other quirks about this card too, since no
other open source OS supports it, or even claims to support it.  But
it was a fun half hour hack...

15 years agoWhen allocating memory, zero out them if we don't intend to overwrite them
Xin LI [Sat, 28 Mar 2009 04:00:46 +0000 (04:00 +0000)]
When allocating memory, zero out them if we don't intend to overwrite them
all; before freeing memory, zero out them before we release it as free
heap.  This will eliminate some potential information leak issue.

While there, remove the PURIFY option.  There is a slight difference between
the new behavior and the old -DPURIFY behavior, with the latter initializes
memory with 0xff's.  The difference between old and new approach does not
generate observable difference.

Obtained from: OpenBSD (partly).

15 years agoAdd new V* constants, neccessary for granular permission checks
Edward Tomasz Napierala [Fri, 27 Mar 2009 21:47:56 +0000 (21:47 +0000)]
Add new V* constants, neccessary for granular permission checks
in NFSv4 ACLs.  While here, get rid of VALLPERM; it wasn't used anyway.

Approved by: rwatson (mentor)

15 years agoAdd explicit test so that non-existence of /usr/games/fortune
Ruslan Ermilov [Fri, 27 Mar 2009 21:13:14 +0000 (21:13 +0000)]
Add explicit test so that non-existence of /usr/games/fortune
doesn't affect command exit status.

15 years agoSort NECINFORTIA correctly.
Warner Losh [Fri, 27 Mar 2009 20:40:17 +0000 (20:40 +0000)]
Sort NECINFORTIA correctly.

15 years agoAdd bromax axnet based cards to the mix. This was harvested from the linux
Warner Losh [Fri, 27 Mar 2009 19:15:28 +0000 (19:15 +0000)]
Add bromax axnet based cards to the mix.  This was harvested from the linux
driver.  Not sure who sold it/rebadged it.

Add stub entries for Mitsubishi B8895 and Toshiba LANCT00A to the
driver with a comment that they don't work /* NG */.[*] These are
DP83902A based cards, which should work, but don't seem to.  Likely
they are from the days before the ne2000 roamed the earth and use a
non-standard hookup (see if_ed_isa or if_ed_cbus for some examples).
Unless I happen to stumble into the right one, these may never work,
but I'm tired of omitting them from commits.

[*] The Japanese adopted OK from English, but also use NG for its
opposite.

15 years agoDon't strip TTY device name to the last '/'.
Ed Schouten [Fri, 27 Mar 2009 19:13:36 +0000 (19:13 +0000)]
Don't strip TTY device name to the last '/'.

We've seen this bug in other applications before: we have some
applications that use strrchr(tty, '/') on the TTY device name. This
isn't valid when using pts(4), because the device name will be stripped
to "0" instead of "pts/0".

This fixes issues with login(1) ignoring /etc/ttys and missing utmp
records.

Reported by: Barney Cordoba <barney_cordoba yahoo com>
Reviewed by: rwatson

15 years agoHarvest a previously unknown Bromax design based on the AX88x90 from
Warner Losh [Fri, 27 Mar 2009 19:08:15 +0000 (19:08 +0000)]
Harvest a previously unknown Bromax design based on the AX88x90 from
the Linux axnet driver.

15 years agoRevert 190445 change to this file restoring:
Doug Ambrisko [Fri, 27 Mar 2009 17:00:49 +0000 (17:00 +0000)]
Revert 190445 change to this file restoring:
typedef l_long          l_off_t;
Change l_mmap_argv's to l_ulong for pgoff.  This restores prior behaviour
to consumers of l_off_t but allows mmap to mmap a 32bit position which a
Linux application requires to access SMBIOS data via /dev/mem.

Reviewed by: dchagin
Prompted by: rdivacky

15 years agoThe buffer start and end phys addresses should be 0x0 for an OHCI zero length
Andrew Thompson [Fri, 27 Mar 2009 16:56:01 +0000 (16:56 +0000)]
The buffer start and end phys addresses should be 0x0 for an OHCI zero length
packet, this fixes LS/FS devices on the Gateworks 2348 XScale board.

Reviewed by: HPS

15 years agoCorrect typo.
Konstantin Belousov [Fri, 27 Mar 2009 15:46:02 +0000 (15:46 +0000)]
Correct typo.

Noted by: kensmith

15 years agoWhitespace/spelling fixes in advance of upcoming functional changes.
Jamie Gritton [Fri, 27 Mar 2009 13:13:59 +0000 (13:13 +0000)]
Whitespace/spelling fixes in advance of upcoming functional changes.

Approved by: bz (mentor)

15 years agoDocument missing requests.
Diomidis Spinellis [Fri, 27 Mar 2009 11:03:02 +0000 (11:03 +0000)]
Document missing requests.

15 years agoSharpen the saw:
Marcel Moolenaar [Fri, 27 Mar 2009 05:48:42 +0000 (05:48 +0000)]
Sharpen the saw:
o  BSD uses 32-bit block numbers. Limit the scheme to 2^32-1
   blocks when the media is larger.

15 years agoPut the "%d retries" debug message in the transmit path behind WPI_DEBUG_TX
Juli Mallett [Fri, 27 Mar 2009 05:44:53 +0000 (05:44 +0000)]
Put the "%d retries" debug message in the transmit path behind WPI_DEBUG_TX
rather than behind a seemingly accidental constant likely left over from one of
the related drivers which uses log levels rather than per-facility debugging
flags.  This should get rid of contextless messages on the console for people
who have not set (or cleared the default) debugging flags.

15 years agoSharpen the saw:
Marcel Moolenaar [Fri, 27 Mar 2009 05:35:12 +0000 (05:35 +0000)]
Sharpen the saw:
o  Don't create an APM scheme underneath another scheme when
   the probe doesn't allow it.
o  APM uses 32-bit block numbers. Limit the scheme to 2^32-1
   blocks when the media is larger.

15 years agoSkip the allocation of the root hold token if the mount already happened.
Andrew Thompson [Fri, 27 Mar 2009 03:52:08 +0000 (03:52 +0000)]
Skip the allocation of the root hold token if the mount already happened.

15 years agoMFi386: r190447
Yoshihiro Takahashi [Fri, 27 Mar 2009 03:45:42 +0000 (03:45 +0000)]
MFi386: r190447

  Convert gdt_segs and ldt_segs initialization to C99 style.

15 years agoo) Check that no overrun or CRC errors were encountered in receiving a
Juli Mallett [Fri, 27 Mar 2009 03:17:25 +0000 (03:17 +0000)]
o) Check that no overrun or CRC errors were encountered in receiving a
   packet.  Linux, OpenBSD and our iwn(4) all do this.  It also results in
   a huge performance improvement (and the rejection of a fair number of
   apparently-bad packets on receive) on my hardware.
o) Like the wpi(4) driver in OpenBSD, and like our iwn(4), also drop runt
   packets.
o) Don't bother doing IFQ_POLL and then IFQ_DRV_DEQUEUE, just do
   IFQ_DRV_DEQUEUE outright.  This is more similar to how OpenBSD and our
   iwn(4) work.

Reviewed by: sam

15 years agoadd superg+tdma include files for protocol defs needed to do ie parsing
Sam Leffler [Thu, 26 Mar 2009 21:29:38 +0000 (21:29 +0000)]
add superg+tdma include files for protocol defs needed to do ie parsing

15 years agoo move tdma+superg protocol defs out of public view
Sam Leffler [Thu, 26 Mar 2009 21:29:03 +0000 (21:29 +0000)]
o move tdma+superg protocol defs out of public view
o add #ifdef _KERNEL to superg+tdma include files so they can be used
  by user code to get the protocol defs

15 years agoMove the code to update cpu_cx_count out of acpi_cpu_generic_cx_probe() and
John Baldwin [Thu, 26 Mar 2009 21:10:35 +0000 (21:10 +0000)]
Move the code to update cpu_cx_count out of acpi_cpu_generic_cx_probe() and
into acpi_cpu_startup() which is where all the other code to update this
global variable lives.  This fixes a bug where cpu_cx_count was not updated
correctly if acpi_cpu_generic_cx_probe() returned early.

PR: kern/108581
Debugged by: Bruce Cran
Reviewed by: avg, njl, sepotvin
MFC after: 3 days

15 years agoSigh, not my day. Check-in the update version that didn't have
Doug Ambrisko [Thu, 26 Mar 2009 20:23:21 +0000 (20:23 +0000)]
Sigh, not my day.  Check-in the update version that didn't have
the linux_compat mistakes.

15 years agoAdd in forgotten module ipmi_linux.
Doug Ambrisko [Thu, 26 Mar 2009 19:15:31 +0000 (19:15 +0000)]
Add in forgotten module ipmi_linux.

Found by: mlaier

15 years agoFix enough bits so that fast frames work again:
Sam Leffler [Thu, 26 Mar 2009 19:13:11 +0000 (19:13 +0000)]
Fix enough bits so that fast frames work again:
o include ath ie in beacon frames
o fix probe response check for including ath ie
o add ieee80211_add_athcap shorthand for ap-side ie additions

15 years agodynamic turbo mode is only implemented in sta mode; don't enable it
Sam Leffler [Thu, 26 Mar 2009 19:07:56 +0000 (19:07 +0000)]
dynamic turbo mode is only implemented in sta mode; don't enable it
for other operating modes based on the capability

15 years agoneed to adjust htinfo offset when csa is inserted
Sam Leffler [Thu, 26 Mar 2009 19:06:00 +0000 (19:06 +0000)]
need to adjust htinfo offset when csa is inserted

15 years agoadjust tdma ie offset when beacon frame contents changes
Sam Leffler [Thu, 26 Mar 2009 19:04:11 +0000 (19:04 +0000)]
adjust tdma ie offset when beacon frame contents changes

15 years agoConvert gdt_segs and ldt_segs initialization to C99 style.
Konstantin Belousov [Thu, 26 Mar 2009 18:07:13 +0000 (18:07 +0000)]
Convert gdt_segs and ldt_segs initialization to C99 style.

Reviewed by: jhb

15 years agoAllow the attach routine to fail gracefully and not panic the system.
Warner Losh [Thu, 26 Mar 2009 17:36:19 +0000 (17:36 +0000)]
Allow the attach routine to fail gracefully and not panic the system.

15 years agoAdd stuff to support upcoming BMC/IPMI flashing of newer Dell machine
Doug Ambrisko [Thu, 26 Mar 2009 17:14:22 +0000 (17:14 +0000)]
Add stuff to support upcoming BMC/IPMI flashing of newer Dell machine
via the Linux tool.
     -  Add Linux shim to ipmi(4)
     -  Create a partitions file to linprocfs to make Linux fdisk see
        disks.  This file is dynamic so we can see disks come and go.
     -  Convert msdosfs to vfat in mtab since Linux uses that for
        msdosfs.
     -  In the Linux mount path convert vfat passed in to msdosfs
        so Linux mount works on FreeBSD.  Note that tasting works
        so that if da0 is a msdos file system
                /compat/linux/bin/mount /dev/da0 /mnt
        works.
     -  fix a 64it bug for l_off_t.
Grabing sh, mount, fdisk, df from Linux, creating a symlink of mtab to
/compat/linux/etc/mtab and then some careful unpacking of the Linux bmc
update tool and hacking makes it work on newer Dell boxes.  Note, probably
if you can't figure out how to do this, then you probably shouldn't be
doing it :-)

15 years agoChange the priority from high to normal. This makes sure that
Marcel Moolenaar [Thu, 26 Mar 2009 16:42:24 +0000 (16:42 +0000)]
Change the priority from high to normal. This makes sure that
the BSD or GPT schemes can take precedence as appropriate.

15 years agothere is no dev_mkdb anymore
Luigi Rizzo [Thu, 26 Mar 2009 10:22:40 +0000 (10:22 +0000)]
there is no dev_mkdb anymore

15 years agobump date.
Weongyo Jeong [Thu, 26 Mar 2009 06:12:04 +0000 (06:12 +0000)]
bump date.

Pointed by: yongari

15 years agoadds malo(4).
Weongyo Jeong [Thu, 26 Mar 2009 04:17:35 +0000 (04:17 +0000)]
adds malo(4).

15 years agobwi(4) and mwl(4) doesn't exist at CURRENT.
Weongyo Jeong [Thu, 26 Mar 2009 04:16:40 +0000 (04:16 +0000)]
bwi(4) and mwl(4) doesn't exist at CURRENT.

15 years agoThe malo(4) device driver first appeared in FreeBSD 7.1 not 8.0.
Weongyo Jeong [Thu, 26 Mar 2009 02:59:10 +0000 (02:59 +0000)]
The malo(4) device driver first appeared in FreeBSD 7.1 not 8.0.

Reported by: John <lists_at_reiteration.net>
MFC after: 3 days

15 years agoFix up waiting on vblank again... This reverts a last minute change that
Robert Noland [Thu, 26 Mar 2009 02:10:18 +0000 (02:10 +0000)]
Fix up waiting on vblank again... This reverts a last minute change that
I made on the last patch, it seems to upset suspend/resume and shutdown.

MFC after: 3 days

15 years agoTweak comments.
Warner Losh [Wed, 25 Mar 2009 22:21:53 +0000 (22:21 +0000)]
Tweak comments.

15 years agoAdd RIOS PCCARD 3.
Warner Losh [Wed, 25 Mar 2009 22:21:38 +0000 (22:21 +0000)]
Add RIOS PCCARD 3.

15 years agoRIOS Systems co was naughty and used a 'random' id. Use it with the
Warner Losh [Wed, 25 Mar 2009 22:20:36 +0000 (22:20 +0000)]
RIOS Systems co was naughty and used a 'random' id.  Use it with the
PCCARD3 name to describe the RIOS PC Card III Ethernet that I have.

15 years agoAdd recently added cards.
Warner Losh [Wed, 25 Mar 2009 22:14:46 +0000 (22:14 +0000)]
Add recently added cards.