Commit Graph

32199 Commits

Author SHA1 Message Date
jeff
b152d5fbb5 Remove the temporary alignment check in free().
Implement the following checks on freed memory in the bucket path:
	- Slab membership
	- Alignment
	- Duplicate free

This previously was only done if we skipped the buckets.  This code will slow
down INVARIANTS a bit, but it is smp safe.  The checks were moved out of the
normal path and into hooks supplied in uma_dbg.
2002-05-02 02:08:48 +00:00
obrien
d779b4edb0 Add some NOTES on the Comtrol Rocketport and the Digiboard drivers. 2002-05-01 23:03:31 +00:00
marcel
047c7fe5d2 Save the MCA info specific to the AP as part of the AP launch. 2002-05-01 22:19:03 +00:00
marcel
c9a9ce82cc Make ia64_mca_save_state MP safe. Protect access to the info block,
updating the sysctl tree and clearing the SAL state by a spin lock.
2002-05-01 22:10:03 +00:00
mjacob
ccea943059 If we get a DATA UNDERRUN error from QLogic FC cards, but the RQCS_RU bit
is not set in the scsi completion status, or if the residual is clearly
nonsense, then this was a command that suffered the loss of one or more
FC frames in the middle of the exchange.

Set HBA_BOTCH and hope it will get retried. It's the only thing we can do.

MFC after:	1 day
2002-05-01 21:58:36 +00:00
peter
5b777bef20 Try using genwakecode.sh instead of the perl version. 2002-05-01 21:54:07 +00:00
peter
ab5403ca0f Add a shell script to do what genwakecode.pl does.
Obtained from:	ade
2002-05-01 21:52:34 +00:00
alfred
798c53d495 Redo the sigio locking.
Turn the sigio sx into a mutex.

Sigio lock is really only needed to protect interrupts from dereferencing
the sigio pointer in an object when the sigio itself is being destroyed.

In order to do this in the most unintrusive manner change pgsigio's
sigio * argument into a **, that way we can lock internally to the
function.
2002-05-01 20:44:46 +00:00
imp
97feabed08 We don't need no stinkin' echos here.
Instead, don't run kldxref if you don't have one on your system.
2002-05-01 19:24:26 +00:00
semenu
595375107e Add an epic_stats_update() function (called once per second). Even though
we don't collect any stats in it, we mii_tick() in it! This fix the bug
when autonegotiating fullduplex modes.

Also, pause activity before setting TXCON in epic_miibus_statchg(). Though
i've never seen problmes from not doing that, the documentation says we
need to do it.

MFC after:	1 week
2002-05-01 19:23:04 +00:00
semenu
7c4d721144 MII_TICK handlers need not restart aneg on these PHYs, they behave pretty
well as is, so - just fetch current status upon MII_TICK.

Also do IFM_INST verification at the top of *_service() then doing it
separately for every case in switch.

acphy: do not read MII_ACPHY_DIAG twice, there is nothing latching.
qsphy: always fetch actual link status from MII_QSPHY_PCTL.

MFC after:	1 week
2002-05-01 19:15:54 +00:00
jhb
59e363cb4e Axe unused SESS_UNLOCK_NOSWITCH() and PGRP_UNLOCK_NOSWITCH() macros. The
MTX_NOSWITCH flag was deprecated a while ago.
2002-05-01 18:11:16 +00:00
alfred
21257e117d Fix some edge cases where bad string handling could occur.
Submitted by: ps
2002-05-01 08:29:41 +00:00
alfred
f34e021666 cleanup:
fix line wraps, add some comments, fix macro definitions, fix for(;;) loops.
2002-05-01 08:08:24 +00:00
peter
dd509e31ee Connect up kern_envp *before* we use it for getenv() and console probing.
It is a bit late after that when we have no consoles. :-]

Also, fix a comment nit and print a warning about missing metadata.
2002-05-01 06:52:08 +00:00
peter
dfb63f63b0 Add "CFLAGS+= -I${MAKESRCPATH}" when running under the kernel build,
otherwise #include "aicasm.h" etc do not work with gcc-3 and later.
2002-05-01 06:50:59 +00:00
cjc
6b0c9026c6 Enlighten those who read the FINE POINTS of the documentation a bit
more on how ipfw(8) deals with tiny fragments. While we're at it, add
a quick log message to even let people know we dropped a packet. (Note
that the second FINE POINT is somewhat redundant given the first, but
since the code is there, leave the docs for it.)

MFC after:	1 day
2002-05-01 06:29:16 +00:00
peter
7b6f84310a Remove two unused headers (<machine/frame.h> and <machine/psl.h>).
psl.h is 100% bogus to be referenced here, especially from alpha MD code.
2002-05-01 06:24:51 +00:00
obrien
81dea2aac7 "pointers are not permitted as case values", so force the macros to ints. 2002-05-01 04:18:36 +00:00
obrien
590c999a50 makeobjops.awk is stricter on syntax than the perl version. 2002-05-01 03:52:40 +00:00
obrien
7ef86b3115 Use makeobjops.awk rather than makeobjops.pl.
(with big thanks to Oliver Fromme <olli@fromme.com>)
2002-05-01 03:28:14 +00:00
obrien
77bbdad49c Awk version of makeobjops.PL.
Note the invocation ordering is slightly different:
	awk -f makeobjops.awk foo.m -ch

Submitted by:	Oliver Fromme <olli@fromme.com>
2002-05-01 03:19:19 +00:00
obrien
d2ad9809fd Remove the trailing ^M's. makeobjops.awk does not like them. 2002-05-01 03:07:40 +00:00
peter
84ae7c9225 Cosmetic tweaks. Try and keep the style more consistent, catch some stray
whitespace and update a comment.
2002-05-01 02:51:50 +00:00
peter
336f78b18a Zap some stale unused headers, including one machine/psl.h (which is
a stub on alpha).  Compile tested on alpha and x86.
2002-05-01 02:17:33 +00:00
peter
aac61d6bcd Zap KMODDEPS entries so that people do not wonder why it isn't working on
-current.

Apolgies to: anhold
2002-05-01 01:34:48 +00:00
peter
bf12d371f8 Catch any stray KMODDEPS entries to make sure they do not keep turning up. 2002-05-01 01:32:28 +00:00
peter
55a74432bb kern_tc.c doesn't use <machine/psl.h>, and having this #include breaks
other platforms.
2002-05-01 01:31:26 +00:00
obrien
0c1a773004 Remove this Perl script. There have been zero bug reports against
vnode_if.awk.
2002-05-01 00:40:44 +00:00
mdodd
ed57320b79 Document the location (in the source tree) of the "Porter's Handbook". 2002-04-30 23:55:16 +00:00
mdodd
2e0d89a50d Bump __FreeBSD_version for mtx_init() change.
Document same.

Forgotten by:	 jhb
2002-04-30 23:54:03 +00:00
scottl
6f6413a223 Fix a bug where the aacp device would only talk to bus 0 on the
controller.
2002-04-30 22:50:26 +00:00
scottl
3845bd469f Note that the aacp device requires CAM 2002-04-30 22:47:26 +00:00
jeff
968fe15c4d Convert longs to u_longs in stats. This will hold off wrap arounds for a
while longer.
2002-04-30 22:39:32 +00:00
alc
0e84366ae7 o Convert the vm_page buckets mutex to a spin lock. (This resolves
an issue on the Alpha platform found by jeff@.)
 o Simplify vm_page_lookup().

Reviewed by:	jhb
2002-04-30 21:24:47 +00:00
phk
5ae616a516 Brucifixion ? Yes, out that door, row on the left, one patch each.
Many thanks to:	bde
2002-04-30 20:42:06 +00:00
phk
38a0945624 Brucifixion ? Yes, out that door, row on the left, one patch each. 2002-04-30 19:48:45 +00:00
phk
26ffc19d1e Don't export timecounter structures under debug. with sysctl, they
contain no truly interesting data anymore.
2002-04-30 19:34:31 +00:00
dillon
8468513da0 These are Alexander Kabaev's VFSops fixes (see the thread 'Found: module
loading breakage').  The patch fixes serious issues with the VFS
operations vector array which results in a crash when a filesystem module
adding a new VOP is loaded into the kernel.  Basically what was happening
before was that the old operations vector was being freed and a new one
allocated.  The original MALLOC code tended to reuse the same address
for the case and so the bug did not rear its ugly head until the new memory
subsystem was emplaced.

This patch replaces the temporary workaround Dave O'Brien comitted in 1.58.

The patch is clean enough that I intend to MFC it to stable at some point.

Submitted by:	Alexander Kabaev <ak03@gte.com>
MFC after:	1 week
2002-04-30 18:44:32 +00:00
julian
2d4af411cc Need more includes. 2002-04-30 16:43:51 +00:00
julian
adee5febf6 Add the myson controllers to LINT
MFC after:	2 weeks
2002-04-30 16:08:16 +00:00
julian
b393629971 Add the extra argument to mtx_init() 2002-04-30 15:50:09 +00:00
benno
6ff9e30a59 Add sigcode.S 2002-04-30 11:13:16 +00:00
jeff
21868731b0 Add a new UMA debugging facility. This will overwrite freed memory with
0xdeadc0de and then check for it just before memory is handed off as part
of a new request.  This will catch any post free/pre alloc modification of
memory, as well as introduce errors for anything that tries to dereference
it as a pointer.

This code takes the form of special init, fini, ctor and dtor routines that
are specificly used by malloc.  It is in a seperate file because additional
debugging aids will want to live here as well.
2002-04-30 07:54:25 +00:00
arr
009ce05bfe - Take advantage of the M_ZERO flag that can now be passed to uma_zalloc.
- Remove atm_uma_ctor() as the M_ZERO will zero out the memory for us.
2002-04-30 07:41:59 +00:00
marcel
79ba8b4281 Add ar.lc and ar.ec to the trapframe. These are not saved for syscalls,
only for exceptions.

While adding this to exception_save and exception_restore, it was hard
to find a good place to put the instructions. The code sequence was
sufficiently arbitrarily ordered that the density was low (roughly 67%).
No explicit bundling was used.
Thus, I rewrote the functions to optimize for density (close to 80% now),
and added explicit bundles and nop instructions. The immediate operand
on the nop instruction has been incremented with each instance, to make
debugging a bit easier when looking at recurring patterns. Redundant
stops have been removed as much as possible. Future optimizations can
focus more on performance. A well-placed lfetch can make all the
difference here!

Also, the FRAME_Fxx defines in frame.h were mostly bogus. FRAME_F10 to
FRAME_F15 were copied from FRAME_F9 and still had the same index. We
don't use them yet, so nothing was broken.
2002-04-30 06:37:25 +00:00
scottl
4a9b6b4afd Fix udf_read(). Honor the uio_resid when determining the size of
the block to read and copy out.  This removes the hack in
udf_readatoffset() for only reading one block at a time.  WooHoo!
Remove a redundant test for fragmented fids in both udf_readdir()
and udf_lookup().  Add comment to both as to why the test is
written the way it is.  Add a few more safety checks for brelse().

Thanks to Timothy Shimmin <tes@boing.melbourne.sgi.com> for pointing
out these problems.
2002-04-30 05:05:05 +00:00
jeff
06a56984b5 Move the implementation of M_ZERO into UMA so that it can be passed to
uma_zalloc and friends.  Remove this functionality from the malloc wrapper.

Document this change in uma.h and adjust variable names in uma_core.
2002-04-30 04:26:34 +00:00
imp
07648d50d3 skibo tells me that that I didn't apply all of his last patch, and
sent me a replacement patch that fixes the problem.  The challenge
buffer was not large enough by a factor of 4 (due to my changing the
size from 128 to 32, but not u_int8_t to u_int32_t).

MFC after: 1 day
Submitted by: skibo@pacbell.net
2002-04-30 03:46:27 +00:00
alc
4d466c829c o Revert vm_fault1() to its original name vm_fault(), eliminating the wrapper
that took its place for the purposes of acquiring and releasing Giant.
2002-04-30 03:44:34 +00:00