Alexander Polakov [Mon, 18 Jul 2011 10:58:24 +0000 (14:58 +0400)]
more printf -> kprintf + intrs + callouts
Alexander Polakov [Wed, 13 Jul 2011 15:35:55 +0000 (19:35 +0400)]
malloc -> kmalloc, printf->ksprintf + snd_verbose
Alexander Polakov [Wed, 13 Jul 2011 14:28:16 +0000 (18:28 +0400)]
hdac.c: fix header paths
Alexander Polakov [Wed, 13 Jul 2011 14:17:48 +0000 (18:17 +0400)]
Import new HDA code from FreeBSD (unmodified).
Sascha Wildner [Tue, 12 Jul 2011 19:02:07 +0000 (21:02 +0200)]
gdbinit: Some little simplification.
Sascha Wildner [Tue, 12 Jul 2011 18:39:17 +0000 (20:39 +0200)]
gdbinit: Add a function to show the unread portion of the kernel msgbuf.
kgdb only shows this at the start via a hard coded function. This
function is handy if you want to see it again and are too lazy to
scroll up or when it's out of the scrollback.
Sascha Wildner [Mon, 11 Jul 2011 20:18:18 +0000 (22:18 +0200)]
kernel/printcpuinfo(): Sync feature bit descriptions with FreeBSD.
Sascha Wildner [Mon, 11 Jul 2011 17:44:46 +0000 (19:44 +0200)]
vga(4): Remove a bogus check.
These are the remains of a larger chunk of code that I removed in
4e193d20. It was used to determine whether a MGA, CGA, EGA or VGA
was present in the box. I left the check in at the time because I
thought, why not?
However, ASpeed's VGA BIOS (at least as found on some Tyan S8230*
series motherboards) seems to have some flaws and ends up returning
0xf instead of 0x9. This caused vga(4) initialization to fail early
in the boot and no kernel messages would go to the screen.
Since the only thing that interests us is whether a VGA is present
or not, remove this check. There are other tests later on that will
detect what we want (such as checking the CRTC's address).
Reported-by: Francois Tigeot <ftigeot@wolfpond.org>
Sepherosa Ziehau [Mon, 11 Jul 2011 07:01:53 +0000 (15:01 +0800)]
msk(4): Pass through mii_mediachg() return value
Obtained-from: FreeBSD 192720
Sepherosa Ziehau [Mon, 11 Jul 2011 06:50:07 +0000 (14:50 +0800)]
msk(4): Align RX buffer on 8bytes for hardware w/ ram buffer
Add another hardware bug workaround for Yukon II controllers that
have hardware ram buffer. The silicon bug seem to be triggered by
pause frames if receive buffer is not aligned on FIFO word(8 bytes).
To workaround the issue, make sure to align Rx buffers on 8 bytes.
Unfortunately this workaround requires yet another Rx fixup for
strict alignment architecture machines to align IP header.
For newer hardwares that lacks ram buffer may not have this bug so
check number of available ram buffer size to see the existence of
ram buffer.
Obtained-frim: FreeBSD 183346
Sepherosa Ziehau [Mon, 11 Jul 2011 05:50:37 +0000 (13:50 +0800)]
msk(4): Workaround GMAC hardware hang of Yukon II on the receipt of pause frame
This bug seems to happen on certain hardware model/revision (e.g. 88E8053)
but it's not identified which hardwares are affected. To workaround it,
inrease GMAC FIFO threshold by one FIFO word to flush received pause frames.
Obtained-from: FreeBSD 176652
Aggelos Economopoulos [Sun, 10 Jul 2011 21:44:04 +0000 (23:44 +0200)]
ktrdump: handle %c in ktr format strings
Aggelos Economopoulos [Sun, 10 Jul 2011 18:11:23 +0000 (20:11 +0200)]
Mention the ktr format string limitations in ktr(9)
Aggelos Economopoulos [Sun, 10 Jul 2011 17:49:49 +0000 (19:49 +0200)]
ktrdump: fixup size modifiers for integers
The 'j' modifier is actually {u,}intmax_t, not {s,}size_t. Correct that
and add support for 'z' while at it.
Alex Hornung [Sun, 10 Jul 2011 09:35:30 +0000 (10:35 +0100)]
vfs_conf - remove dead struct
Sascha Wildner [Sat, 9 Jul 2011 19:50:35 +0000 (21:50 +0200)]
vga(4): Fix the kernel build with VGA_NO_BIOS set.
Reported-by: Francois Tigeot <ftigeot@wolfpond.org>
Matthew Dillon [Sat, 9 Jul 2011 18:37:05 +0000 (11:37 -0700)]
kernel - Fix improper persist state in tcp
* Fix another case where the persist timer could wind up getting set
while in TCPS_SYN_RECEIVED.
Matthew Dillon [Sat, 9 Jul 2011 18:34:24 +0000 (11:34 -0700)]
HAMMER VFS - Fix inconsistent atime updates on read()
* HAMMER's optimization of the read() path to improve concurrency broke
atime updates in the cache case. There was a big XXX comment in the
code about it.
* Fix the issue in a way that avoids acquiring the fs_token in the cache
case. If the ATIME flag is already set all we need to do is update
ip->ino_data.atime. If it is not set we have to get the fs_token and
go through the normal transaction.
Reported-by: marino
Matthew Dillon [Sat, 9 Jul 2011 06:45:22 +0000 (23:45 -0700)]
kernel - Fix improper persist state in tcp
* When TCP is doing a simultanious connect and sitting in the
TCPS_SYN_RECEIVED state a degenerate 0 window (due to +SYN -ACK packets)
case occurs where the code improperly tries to switch over from the
retry timer to the persist timer, which panics.
The invalid persist state was caught by code placed earlier in the
year.
* Adjust the code to document the case and stay with the retry timer.
Sascha Wildner [Sat, 9 Jul 2011 03:40:39 +0000 (05:40 +0200)]
kernel: Fix compiling SMBFS and NETSMB into the x86_64 kernel.
Matthew Dillon [Fri, 8 Jul 2011 23:05:42 +0000 (16:05 -0700)]
vkernel - enhance the pidfile option and fix memimg file scanning
* Enhance the pidfile option to leave the file descriptor open and hold
an active lock on the pidfile while the vkernel is running.
This allows scripts to test whether the pidfile is real or stale.
* When scanning for memimg files only stop if we are unable to create
a new file. Skip files owned by other users instead of giving up.
* Adjust the vkernel manual page to document the pidfile feature. Also
document that -I can take a vknetd socket path.
Sascha Wildner [Wed, 6 Jul 2011 20:00:11 +0000 (22:00 +0200)]
less(1): Raise WARNS to 2.
Sascha Wildner [Wed, 6 Jul 2011 19:58:18 +0000 (21:58 +0200)]
kernel: Move the est(4) source code to dev/powermng/est.
Just like the other related drivers (such as powernow(4)).
Sascha Wildner [Wed, 6 Jul 2011 19:11:36 +0000 (21:11 +0200)]
Install the est(4) manual page for both i386 and x86_64.
Sascha Wildner [Wed, 6 Jul 2011 19:05:26 +0000 (21:05 +0200)]
kernel: FWIW, allow the CPU_ENABLE_EST option for x86_64 kernels too.
Sascha Wildner [Wed, 6 Jul 2011 18:57:19 +0000 (20:57 +0200)]
kernel: Remove obsolete comment.
Thomas Nikolajsen [Wed, 6 Jul 2011 09:03:21 +0000 (11:03 +0200)]
kernel: Fix buildkernel for kernel w/ 'options IPSEC', but w/o IPSEC_ESP
Thomas Nikolajsen [Wed, 6 Jul 2011 09:37:10 +0000 (11:37 +0200)]
VKERNEL64: Sync to VKERNEL; i.e. add IPSEC, TMPFS & NULLFS
Thomas Nikolajsen [Wed, 6 Jul 2011 09:10:54 +0000 (11:10 +0200)]
vkernel64: Sync to vkernel modulo pc64
I.e. add conf/files files present in both platform/vkernel and platform/pc64.
This adds support for IPSEC
Thomas Nikolajsen [Wed, 6 Jul 2011 09:23:30 +0000 (11:23 +0200)]
padlock(4): Fix build
Thomas Nikolajsen [Wed, 6 Jul 2011 09:21:56 +0000 (11:21 +0200)]
bce(4): Fix build
Alex Hornung [Mon, 4 Jul 2011 18:18:08 +0000 (19:18 +0100)]
crypto(4,9) - update with new algorithms
Sepherosa Ziehau [Mon, 4 Jul 2011 07:32:43 +0000 (15:32 +0800)]
unix(4): Mention SOCK_SEQPACKET
Reviewed-by: swildner@
Alex Hornung [Sun, 3 Jul 2011 15:33:25 +0000 (16:33 +0100)]
dm - change remove_all traversal direction
* Make remove_all iterate backwards, so that devices that depend on
other dm volumes created earlier are removed first; otherwise
remove_all stops at the first dm volume on which another one depends.
Sascha Wildner [Sun, 3 Jul 2011 15:08:54 +0000 (17:08 +0200)]
kernel: Little style(9) issue.
Sepherosa Ziehau [Sun, 3 Jul 2011 08:10:07 +0000 (16:10 +0800)]
bce(4): Add more definition for the RX frame header
Obtained-from: FreeBSD 218423
Sepherosa Ziehau [Sun, 3 Jul 2011 08:08:53 +0000 (16:08 +0800)]
bce(4): Update firmware to 6.0.x
Obtained-from: FreeBSD 218529
Alex Hornung [Sun, 3 Jul 2011 00:16:43 +0000 (01:16 +0100)]
devfs - replace subname* with related* functions
* Completely remove the subname* cruft that removed devices based on
whether they started with a given string. This was used to remove,
for example, all slices and partitions whenever a disk was removed.
The issue with it was that it was solely based on name, so if one
created for example two (unrelated) dm volumes, called "foo1" and
"foo1234", then removing "foo1" would leave "foo1234" dangling.
* Instead we introduce the concept of devices (cdevs) having parents:
now the slices have the disk as their parent, and the partitions have
the slice as their parents. The related* functions in devfs
recursively deal with these relationships given a parent node.
* This fix will be MFC'ed in a week or two if no issues pop up.
Sascha Wildner [Sun, 3 Jul 2011 02:38:59 +0000 (04:38 +0200)]
Sort the file list in /usr/share/examples/etc.
Sascha Wildner [Sun, 3 Jul 2011 02:18:51 +0000 (04:18 +0200)]
Update the README in /usr/share/examples/etc.
Sascha Wildner [Sun, 3 Jul 2011 01:57:52 +0000 (03:57 +0200)]
Remove obsolete file.
Peter Avalos [Sat, 2 Jul 2011 22:50:40 +0000 (12:50 -1000)]
Merge branch 'vendor/LESS'
Peter Avalos [Sat, 2 Jul 2011 22:49:01 +0000 (12:49 -1000)]
Import less-444.
Fix ungetc bug; remove vestiges of obsolete -l option.
Sascha Wildner [Sat, 2 Jul 2011 20:49:10 +0000 (22:49 +0200)]
make upgrade: Remove various old files that were forgotten over the years.
Alex Hornung [Sat, 2 Jul 2011 14:54:55 +0000 (15:54 +0100)]
dm_target_crypt - Add twofish & serpent support
* Adds both CBC and XTS mode support for both Twofish and Serpent
Alex Hornung [Sat, 2 Jul 2011 14:17:32 +0000 (15:17 +0100)]
opencrypto - Add serpent & twofish XTS modes
* also minor cleanup
Alex Hornung [Sat, 2 Jul 2011 12:26:42 +0000 (13:26 +0100)]
tests - minor cosmetic fix
Alex Hornung [Sat, 2 Jul 2011 12:18:37 +0000 (13:18 +0100)]
tests - add serpent via cryptodev tests
Alex Hornung [Sat, 2 Jul 2011 12:14:38 +0000 (13:14 +0100)]
opencrypto - add serpent support to cryptosoft/dev
Alex Hornung [Sat, 2 Jul 2011 12:07:01 +0000 (13:07 +0100)]
crypto - Add Serpent implementation
Sepherosa Ziehau [Sat, 2 Jul 2011 11:58:41 +0000 (19:58 +0800)]
bce(4): Correct ifmedia related bits
- Changing media in bce_stop() is not necessary at all
- Pass through error code returned by mii_mediachg()
Obtained-from: FreeBSD 210260 210261
Sepherosa Ziehau [Sat, 2 Jul 2011 11:37:24 +0000 (19:37 +0800)]
bce(4): Detect context initialization failure
Obtained-from: FreeBSD 210259
Sepherosa Ziehau [Sat, 2 Jul 2011 11:28:55 +0000 (19:28 +0800)]
bce(4): Avoid NULL pointer access, if no flash info could be found
Obtained-from: FreeBSD 210257
Sepherosa Ziehau [Sat, 2 Jul 2011 10:58:44 +0000 (18:58 +0800)]
bce(4): Track bootcode status
Obtained-from: FreeBSD 206268 (partial)
Sepherosa Ziehau [Sat, 2 Jul 2011 10:24:24 +0000 (18:24 +0800)]
bce(4): Print detailed device information only if bootverbose is enabled
Sepherosa Ziehau [Sat, 2 Jul 2011 10:17:33 +0000 (18:17 +0800)]
bce(4): - Added a workaround for NC-SI management firmware
That firmware would allow frames to be accepted while the driver
is resetting the hardware. This failure is generally observed
when broadcast frames are received during driver load and will
generate "Unable to write CTX memory" errors.
Obtained-from: FreeBSD 202717
Sepherosa Ziehau [Wed, 29 Jun 2011 13:41:56 +0000 (21:41 +0800)]
bce(4): Create dedicated shared memory access routines.
Add code to read bootcode firwmare version.
Obtained-from: FreeBSD 194781
Alex Hornung [Sat, 2 Jul 2011 10:58:07 +0000 (11:58 +0100)]
tests - add twofish via cryptodev tests
Alex Hornung [Sat, 2 Jul 2011 09:53:33 +0000 (10:53 +0100)]
opencrypto - add twofish support to cryptosoft/dev
Alex Hornung [Sat, 2 Jul 2011 09:33:16 +0000 (10:33 +0100)]
crypto - Add Twofish implementation
Alex Hornung [Mon, 27 Jun 2011 20:17:24 +0000 (21:17 +0100)]
opencrypto - Add AES GCM/GMAC support
Obtained-from: OpenBSD
Alex Hornung [Mon, 27 Jun 2011 18:07:03 +0000 (19:07 +0100)]
opencrypto - (minor) sync with OpenBSD
Sepherosa Ziehau [Fri, 1 Jul 2011 05:08:44 +0000 (13:08 +0800)]
tcp: Remove PR_IMPLOPCL; it had been invented to support T/TCP
Since we have dropped T/TCP support, keep this flag in place could
confuse TCP processing in protocol threads.
Antonio Huete Jimenez [Thu, 30 Jun 2011 19:55:53 +0000 (21:55 +0200)]
hammer(8) - Fix snapshot count in info directive.
- We were wrongly using the HAMMER filesystem fd instead of a per-PFS one.
- Adapt the snapshot count in the case of error
Sascha Wildner [Thu, 30 Jun 2011 03:59:13 +0000 (05:59 +0200)]
kernel: Remove BCE_NVRAM_WRITE_SUPPORT option.
Antonio Huete Jimenez [Wed, 29 Jun 2011 22:34:13 +0000 (00:34 +0200)]
jail - Automatically mount devfs filesystem in every jail if specified.
Obtained-from: FreeBSD
Antonio Huete Jimenez [Wed, 29 Jun 2011 21:41:41 +0000 (23:41 +0200)]
rc.conf - Fix a wrong reference for jail_procfs_enable
Sascha Wildner [Wed, 29 Jun 2011 20:33:09 +0000 (22:33 +0200)]
kernel: Fix building with BCE_DEBUG.
Sascha Wildner [Wed, 29 Jun 2011 16:34:53 +0000 (18:34 +0200)]
df(1): Properly indent.
Antonio Huete Jimenez [Wed, 29 Jun 2011 16:07:04 +0000 (18:07 +0200)]
df - Properly handle paths passed when -t is used.
Show only the related mount points for the specified filesystem(s) when
option -t is used.
Obtained-from: FreeBSD
Sascha Wildner [Wed, 29 Jun 2011 09:49:02 +0000 (11:49 +0200)]
etc/rc.d/lvm: Set permissions to 644 like elsewhere.
Sepherosa Ziehau [Tue, 28 Jun 2011 14:29:44 +0000 (22:29 +0800)]
bce(4): Update to 4.6.x firmware
Obtained-from: FreeBSD 189325
Sepherosa Ziehau [Tue, 28 Jun 2011 08:13:28 +0000 (16:13 +0800)]
bce(4): Add several HP OEM parts' PCI IDs.
Obtained-from: FreeBSD 187133 187317
Sepherosa Ziehau [Tue, 28 Jun 2011 07:58:20 +0000 (15:58 +0800)]
bce(4): Don't count InFramesL2FilterDiscards into ifnet.if_ierrors
This value does not represent a real packet error but simply indicate
that an unexpected unicast or multicast error was received by the NIC,
which was not counted in the past as well.
Obtained-from: FreeBSD 186169
Sascha Wildner [Tue, 28 Jun 2011 05:58:19 +0000 (07:58 +0200)]
Sync zoneinfo database with tzdata2011h from elsie.nci.nih.gov
antarctica: 8.8 -> 8.9
asia: 8.64 -> 8.65
europe: 8.32 -> 8.33
iso3166.tab 8.6 -> 8.9
southamerica: 8.49 -> 8.50
zone.tab: 8.43 -> 8.45
* antarctica: Remove unused RussAQ rules.
* asia: Add comment on retaining "old" Russia rules for the benefit
of Armenia.
* europe: Russia abandons DST in 2011 (thanks to Alexander Krivenyshev).
"Old" Russia rules retained for now for Belarus.
* iso3166.tab: Change AN (Netherlands Antilles) to CW (Curacao).
* southamerica: Use links for places with new iso3166 codes.
* zone.tab: Change AN (Netherlands Antilles) to CW (Curacao).
Sepherosa Ziehau [Tue, 28 Jun 2011 05:45:01 +0000 (13:45 +0800)]
em(4)/emx(4): Update to Intel's 7.1.7
- Restore stock version of e1000_read_mac_addr_generic(); if this function
is called after hardware reset, it just works as expected.
- For em(4), add 82567/82577/82578/82579/82583, PCH and PCH2 support.
- For em(4) and emx(4)
o Reorganize hardware initialization sequence
o Fix ITR and EIAC settings for 82574
o Fix hardware control support for chips with AMT
Alex Hornung [Mon, 27 Jun 2011 16:50:42 +0000 (17:50 +0100)]
dm_target_crypt - add plain64 ivgen
* The plain64 ivgen works just the same as the plain ivgen, but treats
the sector as 64 bits.
Alex Hornung [Mon, 27 Jun 2011 16:48:24 +0000 (17:48 +0100)]
dm - fix typo in previous commit
Alex Hornung [Mon, 27 Jun 2011 16:33:52 +0000 (17:33 +0100)]
tree.h - Add RB_AUGMENT support
Submitted-by: Brills Peng
Alex Hornung [Mon, 27 Jun 2011 16:29:12 +0000 (17:29 +0100)]
committer.7 - Add note about MFC hook
Alex Hornung [Mon, 27 Jun 2011 16:22:35 +0000 (17:22 +0100)]
dm - check error on vn_open
Submitted-by: Adam Hoka
Sepherosa Ziehau [Sun, 26 Jun 2011 13:30:41 +0000 (21:30 +0800)]
bce(4): Add BCM5709 and BCM5716 support
Thanks David Christensen <davidch@broadcom.com> for sending me a
sample 5709 card for testing.
Obtained-from: FreeBSD 170810, 176850, 179771, 179772, 182293, 185593
Sepherosa Ziehau [Mon, 27 Jun 2011 11:37:09 +0000 (19:37 +0800)]
miidevs.h: regen
Sepherosa Ziehau [Mon, 27 Jun 2011 11:36:48 +0000 (19:36 +0800)]
mii: Add PHY IDs for 5709C
Sepherosa Ziehau [Mon, 27 Jun 2011 07:54:19 +0000 (15:54 +0800)]
em(4)/emx(4): Update ig_hal to Intel's 7.1.7
Local modification:
- Keep old e1000_read_mac_addr_generic() so that 82574L and 82571EB
could be probed correctly (instead of bailing out at "Invalid MAC
address")
Newer device support is not added into em(4) and emx(4) yet.
Sascha Wildner [Sun, 26 Jun 2011 15:55:48 +0000 (17:55 +0200)]
gdbinit: Fix lstok and infotok cmds (t_flags and tr_stallpc were removed).
Alex Hornung [Sat, 25 Jun 2011 18:18:54 +0000 (19:18 +0100)]
tools - Add a git commit-msg hook for MFC'ing
Sascha Wildner [Fri, 24 Jun 2011 11:16:22 +0000 (13:16 +0200)]
kernel: Remove some Micro Channel and EISA remains.
Sascha Wildner [Fri, 24 Jun 2011 11:03:21 +0000 (13:03 +0200)]
kernel: Use NULL for DRIVER_MODULE()'s evh & arg (which are pointers).
This is just cosmetics for easier reading.
Sascha Wildner [Fri, 24 Jun 2011 10:10:02 +0000 (12:10 +0200)]
kernel: Remove unused file.
Alex Hornung [Sun, 19 Jun 2011 09:03:03 +0000 (10:03 +0100)]
padlock - fix rng code for x86_64
* I originally used the wrong calling convention (windows amd64), so
switch to the sysv convention.
* Has been tested on a Via Nano L2200
Alex Hornung [Sun, 19 Jun 2011 07:30:36 +0000 (08:30 +0100)]
padlock - Shuffle around, add rng x86_64 support
* Move the config to the platform dependent conf/files, as the assembly
is platform dependent, and so is padlock in general.
* Add a preliminary rng_harvest_x86_64.S, but as there is no padlock
documentation for x86_64, the implementation could be wrong.
Alex Hornung [Sun, 19 Jun 2011 07:12:27 +0000 (08:12 +0100)]
padlock - fix bug in rng
* The xstore-rng instruction increments the buffer pointer in edi
automatically, so we shouldn't increment it manually, too.
Alex Hornung [Sat, 18 Jun 2011 18:34:23 +0000 (19:34 +0100)]
padlock rng - rework
* Rework to use a custom loop so that it doesn't stall when there is no
more entropy left to harvest.
Sascha Wildner [Sat, 18 Jun 2011 08:57:01 +0000 (10:57 +0200)]
token.9: Don't list lwkt_relpooltoken() for which there are no docs (yet).
Sascha Wildner [Sat, 18 Jun 2011 08:52:34 +0000 (10:52 +0200)]
kernel: Unbreak building padlock(4) into the kernel.
Peter Avalos [Sat, 18 Jun 2011 01:34:39 +0000 (15:34 -1000)]
Patch the contrib sources for awk instead of having separate files.
Peter Avalos [Sat, 18 Jun 2011 01:30:31 +0000 (15:30 -1000)]
Update website for awk in README.DRAGONFLY.
Peter Avalos [Sat, 18 Jun 2011 01:29:44 +0000 (15:29 -1000)]
Merge branch 'vendor/AWK'
Peter Avalos [Sat, 18 Jun 2011 01:05:21 +0000 (15:05 -1000)]
Import awk-
20110506.
* added #ifdef for isblank (allows -ffoo as well as -f foo arguments).
* fixed a subtle (and i hope low-probability) overflow error in
fldbld, by adding space for one extra \0.
* changed srand() to return the previous seed