d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
38,420 Commits 72 Branches 135 Tags 3.2 GiB
23fc6cddce
Commit Graph

15464 Commits

Author SHA1 Message Date
Tor Egge
23fc6cddce Close a race window where a tcp socket is closed while tcp_pcblist is 
copying out tcp socket info, causing a NULL pointer to be dereferenced.
 1999-06-16 19:05:17 +00:00
Dmitrij Tejblum
9a56b5f763 Make syscall tracing work: include "opt_ktrace.h" and make it compile. 1999-06-16 18:55:35 +00:00
Dmitrij Tejblum
71ddfdbbd5 Make sure syscall arguments properly aligned in ktrace records. 
Make syscall return value a register_t.

Based on a patch from Hidetoshi Shimokawa.
Mostly reviewed by:	Hidetoshi Shimokawa and Bruce Evans.
 1999-06-16 18:37:01 +00:00
Matt Jacob
233d9cfce2 Use vput instead of vrele. 
Reviewed by:	Matthew Dillon <dillon@apollo.backplane.com>
Submitted by:	Ville-Pertti Keinonen <will@iki.fi>
Obtained from:	Matthew Dillon <dillon@apollo.backplane.com>
 1999-06-16 18:35:58 +00:00
Nick Hibma
f06b2abff3 Remove option USBVERBOSE 
Add the options for debugging used in dev/usb/*.c

Submitted by:	Kazu
 1999-06-16 17:34:36 +00:00
Bill Paul
05222a2c7d Grrr! The PNIC II's multicast hash table is only 128 bits wide, not 512 
like the original PNIC and the MX98715A (from which the PNIC II is derived).
This requires special handling. Save the card type, and in mx_calchash(),
if we see that the card is a PNIC, return only the low 7 bits of the
hash instead of the low 9 bits.
 1999-06-16 16:27:30 +00:00
David Greenman
cd3fe8d008 Changed trypbuf to a getpbuf to work around a problem where redundant writes 
would occur when clustering them - caused by running out of buffers
and taking a degenerate code path as a result. It appears that waiting
instead for buffers to become available is okay.

Submitted by:	Matthew Dillon <dillon@apollo.backplane.com>
Discovered by: Craig A Soules <soules+@andrew.cmu.edu>
 1999-06-16 15:54:30 +00:00
Kazutaka YOKOTA
5c539f036d Print verbose messages when there is error. 
Oked by: des
 1999-06-16 14:04:45 +00:00
Bill Fumerola
73568d2bf1 Fix non-benign typo. 
Submitted by:	Daniel Baker <dbaker@cuckoo.com>
 1999-06-16 12:26:40 +00:00
Tor Egge
a31bad69da Clean up bitrot in interrupt tracing code. 1999-06-16 03:53:52 +00:00
Tor Egge
01cf8ad024 If we still haven't got a sufficient number of free buffers after the 
call to flushdirtybuffers() then sleep in waitfreebuffers().
PR:		11697
Reviewed by:	David Greenman, Matt Dillon
 1999-06-16 03:19:04 +00:00
Kirk McKusick
e4ab40bcb6 Get rid of the global variable rushjob and replace it with a function in 
kern/vfs_subr.c named speedup_syncer() which handles the speedup request.
Change the various clients of rushjob to use the new function.
 1999-06-15 23:37:29 +00:00
Matt Jacob
e672bf9cd6 If we retry this operation from the top of this routine, we need to 
make sure we've freed any allocated resources (to avoid a memory leak)
and and do the right thing with respect to the nfs node hash lock we'd
acquired.
 1999-06-15 23:24:14 +00:00
Julian Elischer
28808d20e9 Add a couple of command definitions. 
Also add offsets into the IDE parameter block so that it is humanly
possible to match the structure to the manufacturer's documentation.
(basically this is just changes to comments)
 1999-06-15 23:13:09 +00:00
Kirk McKusick
5e1a20fd26 COM_LLCONSOLE should over-ride COM_DEBUGGER request. 1999-06-15 22:15:10 +00:00
Dag-Erling Smørgrav
9b953cf681 Kill option FAILSAFE. 
PR:		i386/12187
Approved by:	bde
 1999-06-15 13:14:56 +00:00
Roger Hardiman
69be7dc19e Add VIA Apollo Master VT82c570 chipset. (I have 2 of these) 
Also, Apollo IDE controller has device ID of 1571 and 0571
 1999-06-15 12:01:26 +00:00
Mark Newton
009b52bf57 Ensoniq AudioPCI sound cards haven't worked since the newbus integration 
because they haven't been able to attach.  Now fixed.

Submitted by:	Christian Weisgerber <naddy@mips.rhein-neckar.de>
 1999-06-15 00:00:10 +00:00
John Birrell
bfc6f8cf11 Add the System V message queue and semaphore options for better 3rd party 
software support out-of-the-box (like i386 does).
 1999-06-14 22:59:55 +00:00
Nik Clayton
501bebb5be Include <sys/time.h> for correctness. BDE has a better version of this, 
but it's more complex, and in his words

    Commit your version, since it is the only one that is clearly permitted
    (if not best), and I'll untangle it later.

PR:		docs/11589
Reviewed by:	Bruce "he kicks ass" Evans
 1999-06-14 19:03:46 +00:00
Hidetoshi Shimokawa
7872cbaeea Fix typo confusing `fa' and 'fb' registers. 
Now, MATLAB works fine under OSF/1 emulation.
 1999-06-14 11:42:30 +00:00
Nick Hibma
85dd1453b1 1) remove the soft reset after a command failed. 
2) remove printing of errno and errstr
 1999-06-13 22:09:14 +00:00
Poul-Henning Kamp
7a7404d275 Eliminate the bogus procfs private almost struct dirent structure. 
Spotted by: Lars Hamren
Reviewed by: bde
 1999-06-13 20:53:16 +00:00
Nick Hibma
7b69218647 1) Add non-blocking I/O on read 
2) Add checks to make sure we do not dereference the softc when not found
because the device has been removed.
 1999-06-13 20:49:12 +00:00
Nick Hibma
680514601e 1) Add the USB controllers (will now print pretty strings when no usb 
support is compiled in)
2) Add probing for generic USB host controllers as well so we get them all
3) make the returned strings look alike in the whole file
 1999-06-13 20:48:29 +00:00
Nick Hibma
b24ecce39c Shorten the strings a bit (remove 'Host') 1999-06-13 20:46:10 +00:00
Alan Cox
dabd891d6e Use pmap_kenter instead of pmap_enter to map the message buffer. 1999-06-13 19:20:25 +00:00
Roger Hardiman
3d68f363a4 Removed checking for unit>NBKTR. 
No longer need when using device_get_softc.
(also fixed bug where loadable bktr module could only access 1st device)
 1999-06-13 16:05:00 +00:00
Roger Hardiman
d56282d89f Add bktr_load for the loadable Bt848/Bt878 driver. 1999-06-13 15:45:49 +00:00
Alan Cox
1c85e3df24 vm_map_protect: 
The wrong vm_map_entry is used to determine if writes must not be
	allowed due to COW.
 1999-06-12 23:10:38 +00:00
Roger Hardiman
05da1d9122 Add bktr (Bt848/Bt878 driver loadable module) to the Makefile i386 only section 1999-06-12 15:11:33 +00:00
Roger Hardiman
6c96acd35c Change perl5 to perl (same change as made in the fxp module) 1999-06-12 15:08:50 +00:00
Roger Hardiman
b72df81b21 Makefile to build a loadable bktr Bt848/Bt878 device driver 
This allows the bt848 grabber to be loaded in the boot loader or
after bootup without the need for a custom kernel
 1999-06-12 15:00:19 +00:00
Roger Hardiman
5a2f31f0fb Updated to use the new 4.x newbus API 
Also removed the BSDI support (for now)
This allows the driver to be loaded/unloaded as a KLD
and loaded in the boot loader phase whithout making a custom kernel.
 1999-06-12 14:54:56 +00:00
Ruslan Ermilov
adbdafc6b2 Don't accept divert/tee/pipe rules without corresponding option. 
PR:		10324
Reviewed by:	luigi
 1999-06-11 11:27:35 +00:00
Dmitrij Tejblum
23405ee478 Replace my previous fix of saving the FP state with a much simpler one: when 
we swap out fpcurproc, save its FP state.

Suggested by:	bde
 1999-06-10 20:40:59 +00:00
Jonathan Lemon
5a86dd8c67 Change variable used for calculating ending address of physical memory 
from 'int' to 'vm_offset_t'.

Spotted by:	 Richard Cownie <tich@ma.ikos.com>
 1999-06-10 02:48:51 +00:00
Andrey A. Chernov
97a83933e5 Add detection code for NVidia Riva 128,TNT,TNT2 
(TNT2 from me)

PR: 12094
Submitted by: Ilya Naumov camel@avias.com
 1999-06-09 11:46:43 +00:00
Dmitrij Tejblum
2105375397 Use kmem_alloc_nofault() rather than kmem_alloc_pageable() to allocate 
kernel virtual address space for UPAGES.
 1999-06-08 17:14:22 +00:00
Dmitrij Tejblum
a839bdc8af Add a function kmem_alloc_nofault() - same as kmem_alloc_pageable(), but 
create a nofault entry. It will be used to allocate kmem for upages.

(I am not too happy with all this, but it's better than nothing).
 1999-06-08 17:03:28 +00:00
Dmitrij Tejblum
be960acd20 Keep fpcurproc locked in memory, so that we always can save the FP state 
correctly.

This should fix the "pmap_changebit didn't" panic that some people see.

Reviewed by:	dfr
 1999-06-08 16:42:19 +00:00
Mike Smith
79fc0bf4a0 From the submitter: 
- this causes POSIX locking to use the thread group leader
   (p->p_leader) as the locking thread for all advisory locks.
   In non-kernel-threaded code p->p_leader == p, so this will have
   no effect.

   This results in (more) correct POSIX threaded flock-ing semantics.

   It also prevents the leader from exiting before any of the children.
   (so that p->p_leader will never be stale) in exit1().

   We have been running this patch for over a month now in our lab
   under load and at customer sites.

Submitted by:	John Plevyak <jplevyak@inktomi.com>
 1999-06-07 20:37:29 +00:00
Archie Cobbs
05292ba234 ksprintn() may be called with base=2, so redefine MAXNBUF accordingly. 
Other brucification tweaks.

Obtained from:	bde@freebsd.org
 1999-06-07 18:26:26 +00:00
Steve Price
2df9ad0bd7 The FD options are now in /sys/conf/options so need to duplicate them 
here and keep people from building new kernels. :)
 1999-06-06 22:45:04 +00:00
Bill Paul
32af3bf215 Small optimization: use wi_seek() in wi_read_record() and wi_write_record() 
instead of frobbing the buffer access path registers directly. Saves a
few lines of duplicated code.
 1999-06-06 16:44:04 +00:00
KATO Takenori
ebc253718c Sync with sys/i386/conf/majors.i386 revision 1.76. 1999-06-06 12:50:46 +00:00
KATO Takenori
621ba6f2bb Sync with sys/boot/i386/libi386/bootinfo.c revision 1.19. 1999-06-06 12:46:40 +00:00
Poul-Henning Kamp
e1f523d464 Reserve cdev major 126 for sync support. 1999-06-06 10:41:35 +00:00
Poul-Henning Kamp
2f55ead758 typo in previous commit 1999-06-06 09:28:01 +00:00
Poul-Henning Kamp
cf4b937128 Introduce IFF_SMART bit. 
This means that the driver will add/delete routes when it knows it is
up/down, rather than have the generic code belive it is up if configured.

This is probably most useful for serial lines, although many PHY chips
could probably tell us if we're connected to the cable/hub as well.
 1999-06-06 09:17:51 +00:00