Commit Graph

31678 Commits

Author SHA1 Message Date
John Baldwin
3aa68913bc Make this compile again when UMASS_DEBUG isn't defined. 2002-04-11 21:09:41 +00:00
John Baldwin
b106d2f56a - Set the base priority of an ithread that has no handlers when we set its
normal priority.
- Lock sched_lock while we dink with the priorities.
- Remove a few extra blank lines.
2002-04-11 21:03:35 +00:00
John Baldwin
5a882ddd66 Commented out locking that would be used in the ps command if locks were
used in ddb.
2002-04-11 21:01:34 +00:00
John Baldwin
8e4357a4cb Use proc lock to protect p_ucred pointer while we deference it to read a
few values.
2002-04-11 21:00:38 +00:00
John Baldwin
0af24d5151 Use td_ucred in a few spots. 2002-04-11 21:00:05 +00:00
Alan Cox
ab9ab5702e Regen 2002-04-11 17:35:53 +00:00
Alan Cox
a0805f6f7a Remove the requirement that Giant be held around osigreturn(). All platform-
specific implementations are MPSAFE.
2002-04-11 17:34:38 +00:00
Alan Cox
0d36f49e63 Add a comment that osigreturn() is MPSAFE. 2002-04-11 17:13:08 +00:00
John Baldwin
2c823989e1 Fix a warning due to the code assuming sizeof(int) == sizeof(void *) which
is not true on the alpha.  I think that other parts of this code also make
this implicit assumption as well.
2002-04-11 15:31:01 +00:00
Bernd Walter
0dc76ffdd4 LCA based systems can't handle more than 16 devices on pci bus 0.
Reviewed by:	gallatin
Approved by:	gallatin
2002-04-11 13:24:20 +00:00
Søren Schmidt
5b93eb04a9 Add yet another chip ID for a Promise TX2 chip. 2002-04-11 11:04:23 +00:00
Nick Hibma
b2f485ca0d Delay umass_cam_rescan by 200 ms to make sure attach is finished by the
time we tell CAM to rescan the bus. Together with the previous patch
this should avoid the problem where the devices would wedge because they
got spoken to over two different pipes.

Tested by:	Tomas Pluskal <plusik@pohoda.cz>
2002-04-11 10:34:15 +00:00
Peter Wemm
31c41f1b1e Bandaid for a buffer overrun in the module searching code. When breaking
up the module_path string, we would walk one past the end of the buffer.
This hurting ia64 originally, but it was probably also happening on i386
occasionally as well.  The effects were usually harmless, it would add
bogus "binary" search directories to the places it actually looked for
files.
2002-04-11 10:00:44 +00:00
Peter Wemm
2ab188f529 Finally fix loader completely for IA64. efifs_stat() wasn't setting
the S_IFREG bit for regular files.  This caused the path search code to
skip it when it finally did find the kernel (after the common/module.c
buffer overrun bug was fixed)
2002-04-11 09:50:11 +00:00
KATO Takenori
f11aa8f5ca MFi386: revision 1.517. 2002-04-11 09:45:41 +00:00
Paul Saab
2aa6b97298 Fix the I/O performance issues with the ciss driver where CAM was limiting
the number of transactions the device could handle to 1.

Obtained from:	msmith
2002-04-11 09:28:09 +00:00
Pierre Beyssac
2b9592079f Code cleanup, no functional change.
Patch adapted from PR.

PR:		i386/36015
Submitted by:	thomas@cuivre.fr.eu.org
MFC after:	1 week
2002-04-11 09:21:10 +00:00
Søren Schmidt
f8b318ce58 Only calculate Promise magic if a device is there for info. 2002-04-11 08:52:32 +00:00
Warner Losh
1bd2b5b4cd Integrate the hostap stuff that Thomas Skibo <skibo@packbell.net>
wrote.  This code was for 4.5-release, so I've ported it to -current
and made a few minor tweaks.  The biggest non-style tweak was to not
make access point the default.

More changes will be needed to get this actually working, but I wanted
to get a relatively pure baseline.  This doesn't seem to break what
works now.
2002-04-11 07:21:16 +00:00
Mike Barcroft
17dfcdfaed Remove bogus include of <machine/ansi.h>. 2002-04-11 06:53:40 +00:00
Bill Paul
9a29439fb9 Nortel Networks sells a RealTek 8139-based NIC that's basically
the same thing as the SMC 1211, but with their own vendor ID.
Update the device list to support this NIC. (Discovered these
cards lying around the lab at work.)
2002-04-11 06:12:51 +00:00
Warner Losh
b0db732e54 Add two more IEEE80211 defines for status. 2002-04-11 05:43:10 +00:00
Warner Losh
ea185f8108 better init 2002-04-11 03:31:45 +00:00
SUZUKI Shinsuke
f361efa0be initialize local variable explicitly
Reviewed by:	ume
Obtained from:	Fujitsu guys
MFC after:	1 week
2002-04-11 02:14:21 +00:00
Mike Silbersack
898568d8ab Remove some ISN generation code which has been unused since the
syncache went in.

MFC after:	3 days
2002-04-10 22:12:01 +00:00
David Malone
a983fdfe4c Move do_cpuid into the correct place in this file and make
the indentation more like the other multi-line assembley in
this file.

Someone who understands gcc constraints could update the
constraints for do_cpuid.
2002-04-10 21:18:46 +00:00
David Malone
4a771f59c2 Swap a bzero for an M_ZERO. Borris approved this ages ago, but
the hard drive with the patch on it went south before I committed
it.

Approved by:	bp
2002-04-10 21:08:33 +00:00
Alan Cox
eac84263c4 o In osigreturn(), restore all of the registers in one place.
o Recent changes to osigreturn() and sigreturn() have made them MPSAFE.  Add
   a comment to this effect.

Submitted by:	bde (bullet #1)
Reviewed by:	jhb (bullet #2)
2002-04-10 20:08:07 +00:00
Doug Rabson
989db74d57 Add IA32 option for emulation of i386 binaries on the ia64 platform. 2002-04-10 19:35:50 +00:00
Doug Rabson
5eb29d9b90 Initial support for executing IA-32 binaries. This will not compile
without a few patches for the rest of the kernel to allow the image
activator to override exec_copyout_strings and setregs.

None of the syscall argument translation has been done. Possibly, this
translation layer can be shared with any platform that wants to support
running ILP32 binaries on an LP64 host (e.g. sparc32 binaries?)
2002-04-10 19:34:51 +00:00
Mike Silbersack
f2697d4d75 Totally nuke IPPORT_USERRESERVED, it is no longer used anywhere, update
remaining comments to reflect new ephemeral port range.

Reminded by:	Maxim Konovalov <maxim@macomnet.ru>
MFC after:	3 days
2002-04-10 19:30:58 +00:00
Doug Rabson
a989d30598 Save and restore the IA-32 state in cpu_switch(). Probably should only do
this if the thread has been executing IA-32 code.
2002-04-10 19:27:50 +00:00
Doug Rabson
b0685df774 Add suhword() and fuhword() for accessing 32-bit values ("half words") in
userland. All these functions should be renamed to be explicit about the
size of value being read or written.
2002-04-10 19:26:49 +00:00
Doug Rabson
4beaa04bf1 Add exception and syscall support for executing IA-32 binaries. 2002-04-10 19:24:59 +00:00
Doug Rabson
4fc38782f0 Add ucode values for SIGFPE etc. Copied from i386/include/signal.h. 2002-04-10 19:24:08 +00:00
Doug Rabson
02939fd82c Add fields for saving/restoring the IA-32 state. 2002-04-10 19:23:21 +00:00
Doug Rabson
01426578dc Add definitions for IA-32 exceptions, interrupts and intercepts. 2002-04-10 19:22:45 +00:00
Brian Somers
c0285beff4 Don't suggest that the digi firmware modules are normally loaded
dynamically, as this will only happen if you kldload digi after the
machine has booted or explicitly mention them in loader.conf.
2002-04-10 17:27:52 +00:00
Doug Rabson
ba4d213488 Call ast() from the syscall exit path as well as for full exception
restores.
2002-04-10 16:35:44 +00:00
Doug Rabson
794cc3d8ae Initialise PCPU_GET(current_pmap) in pmap_bootstrap - cpu_switch needs
to be sure that it is always correct and this was not true for the first
call to cpu_switch. When thread0 resumed later, it ended up calling
pmap_install with a null pmap, which is bad.
2002-04-10 16:34:52 +00:00
Mike Barcroft
49285705cc Remove the hack for segsz_t from <sys/types.h>; use the normal
_BSD_FOO_T_ method for defining segsz_t.
2002-04-10 15:58:13 +00:00
Mike Barcroft
7f0f1cfd57 Add manifest constants: _LITTLE_ENDIAN, _BIG_ENDIAN, _PDP_ENDIAN, and
_BYTE_ORDER.  These are far more useful than their non-underscored
equivalents as these can be used in restricted namespace environments.
Mark the non-underscored variants as deprecated.
2002-04-10 14:39:14 +00:00
Brian Somers
ef94eb4588 Yet another adjustment of digi_delay(). The delay is now consistent
irrespective of whether tsleep() or DELAY() is used.
2002-04-10 14:32:55 +00:00
Nick Hibma
380ee390fd Postpone I/O to the device until attach has finished. This fixes
attachment after boot in some cases where the device gets confused if
spoken to over two different pipes simultaneously (resulting in TIMEOUT
errors on requesting INQUIRY data).

This solution is far from perfect, as after umass_attach has returned
the USB stack fetches the string descriptors in order for creating the
event that is sent to userland. It just makes the chances of collisions
less likely.

Tested by: 	Tomas Pluskal <plusik@pohoda.cz>
2002-04-10 13:22:12 +00:00
Brian Somers
6e5b29b08e Change more uses of tsleep() to digi_delay() and make the loop counts more
sane.
2002-04-10 12:03:36 +00:00
Mike Barcroft
13c3fcc238 Unconditionalize the definition of INET_ADDRSTRLEN and
INET6_ADDRSTRLEN.  Doing this helps expose bogus redefinitions in 3rd
party software.
2002-04-10 11:59:02 +00:00
Søren Schmidt
56397b300c Fix the FreeBSD native ATA RAID code a bit. 2002-04-10 11:18:07 +00:00
Mike Barcroft
1fd0288692 Stop exporting ntohl(), ntohs(), htonl(), and htons() to userland in
<sys/param.h>.
2002-04-10 10:36:23 +00:00
Nick Hibma
2334e40985 intline == 0 is not a valid intline on 386. See pci_cfgintr() in
sys/i386/pci/pci_cfgreg.c.

This should resolve some cases where adding USB support to the kernel
produced an interrupt storm.
2002-04-10 07:35:28 +00:00
Peter Wemm
d637e9891d Add missing 'struct ifreq ifr;' that was forgotten in the last commit. 2002-04-10 06:07:16 +00:00