0c9ce6ed1f
New file with interrupt-handling code. Add intr_fasttrap().
1993-10-22 19:27:40 +00:00
f1dda6c852
Add intr.c.
1993-10-22 19:17:32 +00:00
1b8b52c8c5
mapping the mem or kmem devices is *very* dangerous with the current
...
vm system. Don't allow it.
1993-10-21 23:55:02 +00:00
80a3179079
vanquish the tree of the last traces of /dev/acct
1993-10-21 00:29:03 +00:00
430d001c1a
get rid of ACCOUNTING option. it's mandatory, and there's no define any more.
1993-10-20 23:04:52 +00:00
27bc255b86
kill accounting.
1993-10-18 23:22:56 +00:00
d7498e64da
fix the 'size allocated vs. MAXBSIZE' check for buffers. Instead
...
of making vfs_bio *and* machdep.c deal with it, just deal with it
here, where we actually allocate the physical memory for pages.
1993-10-18 23:14:13 +00:00
9eafd05204
Add I486_CPU.
1993-10-18 16:25:42 +00:00
f0c6138247
Made "start" the entry point (for consistency with other "ld" lines).
1993-10-18 14:41:31 +00:00
f8b564f223
Fixed typo.
1993-10-18 14:35:45 +00:00
14ba800a2f
Provided entry point (_boot1) to ld.
...
Use DESTDIR when installing.
1993-10-18 13:34:26 +00:00
14e6ea89aa
Put definition of __LDPGSZ and `struct relocation_info' where they belong:
...
in <machine/exec.h>.
1993-10-18 09:46:56 +00:00
593a6c40ef
Explicit entry point specified in kernel link command.
...
(so far, we've been lucky only because vers.o doesn't contain text)
1993-10-17 20:56:09 +00:00
107d011b19
Add KBR_EXTENDED.
1993-10-17 14:04:15 +00:00
7620c96e5d
Make atdevbase an int.
1993-10-17 14:03:54 +00:00
6067dc9ea8
Minor cleanup.
1993-10-17 05:34:23 +00:00
7e8063835d
Use `swap generic'.
1993-10-16 07:48:04 +00:00
2ce6de16d5
Add isa_establish(), for all it's worth (which is to say, nothing).
1993-10-16 05:25:21 +00:00
c73d85d341
Panic if DMA with odd length attempted on a 16-bit channel.
1993-10-16 03:55:59 +00:00
4bcf529b86
oops; forgot to move these definitions over, before...
1993-10-15 13:47:22 +00:00
fba9bbc5d9
add back the bufpages limit
1993-10-15 11:55:39 +00:00
4f94812b29
New file with DMA code from isa.c, modified to check for some errors.
1993-10-14 05:22:57 +00:00
5448e175cf
Only include isa/dma.c if a driver which needs it is configured.
1993-10-14 05:17:24 +00:00
0eabc3bc13
sysarch() belongs in sys_machdep.c
1993-10-13 08:25:57 +00:00
a1c3bf517a
use the new location for libkern, auto-depend on it, and include
...
the right one, for profiling or non-profiling.
1993-10-13 06:43:17 +00:00
d466119aa4
Convert space to tabs on the trapframe and sigframe lines
1993-10-10 07:23:34 +00:00
4119b752e0
Add symbols KERNSIZE and PDRSHIFT so that they can be used along with
...
KERNBASE to caclulate several constants in locore.s used for setting
up the vm system.
1993-10-10 07:07:52 +00:00
4cbf1951c1
Put back assembler versions of setrq() and remrq(). GCC 2 is just too lame.
1993-10-09 22:49:29 +00:00
c8431a9508
Add locore2.c as standard.
1993-10-09 22:17:36 +00:00
99e736549a
Remove `device-driver' flags.
1993-10-09 08:55:24 +00:00
7b64280a28
patch from rod grimes:
...
Fix definitions to be in terms of a minimum number of constants by
using the SHIFT value as the basis for the other values, for example
instead of:
#define NBPG 4096
#define PGSHIFT 12
use:
#define PGSHIFT 12
#define NBPG (1 << PGSHIFT)
1993-10-09 01:35:50 +00:00
f51618137b
genassym needs a dependency...
1993-10-09 00:31:21 +00:00
fa29c74cd0
fix the mmap of kmem somewhat, but it's still not perfect...
...
(or, at least, it crashes the machines in some situations where
i don't think it should...)
1993-10-06 23:37:09 +00:00
d4435d2bd4
get rid of excess target line for machdep.o
1993-10-06 21:19:11 +00:00
004534e441
Add FASTLINKS.
1993-10-06 12:15:33 +00:00
501c1b99ad
Add cy.
1993-10-06 12:14:31 +00:00
8cc28d314c
Merge changes from trunk.
1993-10-06 12:09:35 +00:00
e9b92335b0
Add DRQUNK and IOBASEUNK.
1993-10-06 12:09:25 +00:00
81a9642670
Added entry for Cyclades Cyclom-{4/8/16}Y multiport async serial board
...
driver.
1993-10-06 09:31:30 +00:00
e256404adc
Cyclades Cyclom-{4/8/16}Y multiport async serial board driver.
1993-10-06 09:30:16 +00:00
13a48b8263
Goes with the new cy driver.
1993-10-06 09:27:27 +00:00
a92b57431b
make nbuf = bufpages, now that vfs_bio can do the rational thing.
1993-10-04 04:57:09 +00:00
174c25bd21
add mmmmap (mem device mmap routine). possible problems:
...
(1) the 'physmem' constant has a value which i don't think is
correct
(2) not sure what happens if you mmap a region of kmem which is
not in the kernel's address space...
1993-10-02 00:16:24 +00:00
ee70c53e3f
Display a less cryptic message when the ring buffer overflows.
1993-10-01 23:12:04 +00:00
f969ea9aa8
Fix probe routine to work for multiple controllers.
1993-10-01 23:07:27 +00:00
018d1230b2
Read ethernet address correctly on NE1000 boards.
1993-10-01 22:59:57 +00:00
4d4bfb2287
*** empty log message ***
1993-10-01 04:00:45 +00:00
892a4681ae
* Revision 2.5 93/09/30 17:44:14 davidg
...
* patch from vak@zebub.msk.su (Serge V.Vakulenko) to work around
* a hardware bug in cheap WD clone boards where the PROM checksum
* byte is always zero
*
* Revision 2.4 93/09/29 21:24:30 davidg
* Added software NIC reset in NE probe to work around a problem
* with some NE boards where the 8390 doesn't reset properly on
* power-up. Remove initialization of IMR/ISR in the NE probe
* because this is inherent in the reset.
*
* Revision 2.3 93/09/29 15:10:16 davidg
* credit Charles Hannum
*
* Revision 2.2 93/09/29 13:23:25 davidg
* added no multi-buffer override for 3c503
*
* Revision 2.1 93/09/29 12:32:12 davidg
* changed multi-buffer count for 16bit 3c503's from 5 to 2 after
* noticing that the transmitter becomes idle because of so many
* packets to load.
*
* Revision 2.0 93/09/29 00:00:19 davidg
* many changes, rewrites, additions, etc. Now supports the
* NE1000, NE2000, WD8003, WD8013, 3C503, 16bit 3C503, and
* a variety of similar clones. 16bit 3c503 now does multi
* transmit buffers. Nearly every part of the driver has
* changed in some way since rev 1.30.
1993-10-01 02:01:43 +00:00
e57ef97529
Add LKM.
1993-09-30 20:15:25 +00:00
83f32ad82b
Add needs-flag (or needs-count for tty devices) to all driver in the
...
switch tables. Remove ancient ethernet drivers.
1993-09-30 20:14:55 +00:00
845512b39a
TEST: Enable com ports.
...
files.i386.newconf: Add needs-count to com.c for now.
1993-09-30 17:30:32 +00:00
81c776a8e4
Enable mms.
1993-09-29 15:24:41 +00:00
c27a6e4acf
Fix TIMEZONE specs; add DST.
1993-09-29 06:57:10 +00:00
5085e2ca53
files.i386: Defunct.
...
files.i386.newconf: Remove lpa.
TEST: Enable lms and lpt.
*: Change all lpa's to lpt's. Change fd[0-3] to fd* and wd[0-3] to wd*.
Remove all explicit irq's.
1993-09-29 05:37:29 +00:00
ff24138e60
get rid of an "#if 0", and allocate buffers here.
1993-09-29 04:51:12 +00:00
4a21fe3648
ifdef out the ttyfree()'s, so that fill_eproc doesn't panic when
...
a process's session still holds a ref to a tty which has been deallocated
and reused.
1993-09-29 02:36:21 +00:00
b85a24ec4e
get rid of stray printfs (turn them into "lprintf"s)
1993-09-28 21:26:59 +00:00
cd216b5938
Make Ctrl-Space input a NUL and Ctrl-/ input a ^/ (undo in Emacs).
1993-09-28 04:55:17 +00:00
ef83446787
Exports the pccons ioctls.
1993-09-28 03:26:17 +00:00
ceb2a05027
Added keyboard typematic rate adjustment ioctl. Rudimentary support for
...
entering DDB from X11 by means of the F12 key (when XSERVER_DDB is
defined).
1993-09-28 03:25:01 +00:00
e7ecd24ac3
Do not provide quad constants if compiling in a strict ansi or a strict
...
posix environment.
1993-09-24 18:22:31 +00:00
06c199bb44
Renamed from clock.h.
1993-09-24 09:04:56 +00:00
054e0b73bc
clock.c: Finish conversion. Make rtcput() work if VRT bit not set. Speed up
...
time conversion in delay().
fd.c: Minor changes to autoconfig. Use dv_parent rather than storing another
pointer to the fdc_softc in each fd_softc. Change disk label stuff for our
kernel.
icu.s: Make splx() and splnone() return previous cpl for now.
isa.c: Minor autoconfig changes.
isa.h: rtcin() is now nvram() and is declare in nvram.h. Make isaphysmem a
caddr_t.
pccons.c: XXXX: Maybe we should insist they open /dev/io now.
timerreg.h: #define TIMER_NPORTS.
wd.c: Delete unused wdnoreloc() and wddospart(), and use remaining two bits in
unit number. struct disk --> struct wd_softc. struct board --> struct
wdc_softc. Still needs much work.
1993-09-24 08:49:21 +00:00
3540378534
GENERIC, TEST: Remove all irqs.
...
std.i386: Re-add clock0 and timer0.
1993-09-24 08:45:03 +00:00
2dc346f0d1
add accounting
1993-09-23 21:31:56 +00:00
4212e64961
Include sys/signal.h instead of machine/signal.h
1993-09-21 20:51:12 +00:00
b54c7ce913
Be sure to get sys/signal.h
1993-09-21 13:43:15 +00:00
a9f5756dd2
In resettodr() handle rtcget() failure.
1993-09-21 13:09:10 +00:00
39804a17e5
Remove unneeded saving of segment regs for USER_LDT.
1993-09-21 13:08:34 +00:00
d0b8b43cf2
Add USER_LDT option.
1993-09-19 19:11:24 +00:00
42f1476a71
Changes for per-process/user settable LDT and full sigcontexts
1993-09-19 19:00:44 +00:00
90add535c9
Add tISP offset
1993-09-19 18:59:35 +00:00
a3e9fd6947
Add per-process LDT slot.
1993-09-19 18:59:17 +00:00
044ac34c5a
Define signal frame here.
1993-09-19 18:58:39 +00:00
66f540572a
Defines architecure dependant syscalls.
1993-09-19 18:58:14 +00:00
5413da0ff5
Machine dependant signal.h. Defines sigcontext and sig_atomic_t for i386 port.
1993-09-19 18:57:48 +00:00
0145d113ae
get rid of the #ifndef KERNEL around the include of sys/cdefs.h,
...
a la 4.4, because it seriously breaks things otherwise, given
inclusion of endian.h in types.h
1993-09-17 01:44:08 +00:00
a46b224043
Add constants for quad support.
1993-09-16 06:01:11 +00:00
111aecd0a9
Add an #ifndef _POSIX_SOURCE...
1993-09-16 05:59:43 +00:00
cb098b5eb0
Changes to use a trap frame for syscalls.
1993-09-16 03:24:13 +00:00
606544f20d
Add _QUAD_HIGHWORD and _QUAD_LOWWORD for Torek's quad stuff.
1993-09-16 02:25:45 +00:00
a3b397dd18
compile boot blocks with -O6. this makes them "fit". this is sick.
1993-09-15 22:17:54 +00:00
ff02097496
No need to lospecifiy load address at 1M, now the default for any kernel.
1993-09-14 20:01:19 +00:00
fd156a97fa
New i386 code.
1993-09-14 17:28:06 +00:00
5823e758f1
Remove SYMTAB_SPACE from config so that boot symbols are used.
1993-09-13 18:19:26 +00:00
7d2eb9e6b6
Default LOAD_ADDRESS at 0xfe100000 (1M)
1993-09-13 15:28:23 +00:00
b119f3ec4f
Add MACHINE_NONCONTIG option
1993-09-13 14:22:42 +00:00
c3f643cd8a
Remove NEED_* defines, as they are now satisfied by libkern.
1993-09-11 00:16:40 +00:00
0d2e8f0fcb
Link with libkern.
1993-09-11 00:14:13 +00:00
cab4d78107
Removed functions moved to libkern: strcpy, strlen, ffs, etc.
1993-09-11 00:12:56 +00:00
c87683c012
Make various macros work if outb() is being inlined.
1993-09-09 15:16:14 +00:00
1c880d62bf
As of this revision, all known bugs have been fixed. Some of the fixes include:
...
1) fixed 3c503 lock-up if the thinwire cable was disconnected at boot time
2) 8013EBT boards now work (quite well!) in 16bit/16k mode
3) ED_NO_DOUBLE_BUFFERING flag now works
4) slightly higer performance (about 3%) with 16bit WD/SMC boards
5) support for WD8013WC (10BaseT) boards
Additionally, the probe code has been reorganized to be much cleaner. This
revision of the driver is 1.25. The release notes have been updated as well.
1993-09-09 09:40:56 +00:00
aabddf622f
STDC -> __STDC__
1993-09-08 16:55:17 +00:00
6409dcaef1
Trivial stylistic change.
1993-09-07 07:35:41 +00:00
0230178143
Clean up deleted files.
1993-09-07 02:25:20 +00:00
a6b1318c68
Adding the (non-optional) option MACHINE_NONCONTIG
1993-09-07 00:56:05 +00:00
515949e42e
Add a bunch more shit that was missing.
1993-09-07 00:36:27 +00:00
5be4167214
Rearrange SCSI probing to actually work.
1993-09-07 00:25:50 +00:00
0dc9a11469
Rearrange ethernet probing in the vain hope that maybe, just maybe, the ne0
...
probe's effect on ed0 will be reversed when the latter is reset.
1993-09-06 23:35:19 +00:00
8c7f0d2e48
Comment out we0 and ec0, and add equivalent ed0 and ed1.
1993-09-06 23:31:33 +00:00
5e532f3f9f
Add a generous dose of comments.
1993-09-06 23:29:33 +00:00
07e57f53a2
Put back printf() for entry point. It certainly fits.
1993-09-06 22:09:45 +00:00
a75978c2e0
Change `NFS' to `NFSSERVER,NFSCLIENT'.
1993-09-06 21:15:56 +00:00
229a34c6b3
Correct I/O ports for mice.
1993-09-06 18:37:56 +00:00
0535b6089e
Avoid clash of struct ns_addr with #define ns_addr.
1993-09-06 18:30:39 +00:00
84c634662f
Move the `#include "bpfilter.h"' to a more logical and easy to find place.
1993-09-06 18:27:22 +00:00
bec04754dc
Update for new select(), and fix another `struct arpcom.ac_enaddr'.
1993-09-06 18:24:35 +00:00
91b8683f3e
Make Ctrl-Space and Ctrl-2 send a NUL, even without XSERVER defined.
...
(This has annoyed me for a while, and people are complaining about it.)
1993-09-06 05:52:33 +00:00
9cd4c25c2a
Second wd0 should be wd1.
1993-09-05 22:34:12 +00:00
59d53ecfbf
Check some boundary conditions is cursor movement and scrolling.
1993-09-05 17:56:58 +00:00
7c8271a011
I give up. Clearly, SYMTAB_SPACE is required at the moment to build a
...
kernel that doesn't go foom in locore.s
1993-09-05 16:42:54 +00:00
b477be52eb
Yet more of the ptrace() reorg; now ptrace_setregs() and ptrace_getregs()
...
are present, along with PT_GETREGS and PT_SETREGS ptrace commands.
1993-09-05 03:53:52 +00:00
9840b0b7af
update for new exec; a few name changes and (*gasp*) addition of comments.
1993-09-05 01:31:39 +00:00
59a008c712
change maximum stack size to 8M.
1993-09-04 22:21:28 +00:00
3aa99f85dc
better ptrace() support from Sean Eric Fagan <sef@kithrup.com>
1993-09-04 05:32:18 +00:00
37e8d5291e
get rid of bogosities of the form "vm_maxsaddr + MAXSSIZ", and replace
...
them with USRSTACK. the former was a hack by wfj to accomodate his screwy
execve() implementation. the latter is a (minor) reversion to net/2.
1993-09-04 01:29:22 +00:00
2403a33987
decouple MAXDSIZ and MAXSSIZ. MAXSSIZ now is 2M; shouldn't need more,
...
and saves time in exec.
1993-09-04 00:38:47 +00:00
4475f55d8f
Include systm.h to get prototypes (and possibly inlines) of *max functions.
1993-09-04 00:02:48 +00:00
c10d5aa752
get rid of bogus #define of resettodr(); now setting time actually works!
1993-09-02 02:10:35 +00:00
73c9dd61b0
Add ed2, as per GENERICAHA.
1993-09-01 05:36:33 +00:00
3db4b24245
Add ed2 at 0x300, irq 10, 0xcc000, which people can configure by hardware.
1993-09-01 02:31:10 +00:00
5bce7d4176
Add prototype for pmap_bootstrap().
1993-08-30 18:09:57 +00:00
86d51ac6f8
pagers are not options, not pseudo-devices
1993-08-30 07:07:53 +00:00
9aed375ba3
tty XXstart() routines return void
1993-08-29 13:46:31 +00:00
aaf4f8dbc5
Changes for boot loading of symbols.
1993-08-29 12:48:10 +00:00
d48a7d8497
There is no longer any good reason to allocate a small number of ptys.
...
Increased to 32 so people will not bitch at me next time around.
1993-08-29 01:54:22 +00:00
468c5f5158
put PROCFS in pain...
1993-08-28 11:00:35 +00:00
966d7491cc
Changes: Boot omagic kernels (like Mach), pass cnvmem, extmem, and esym args, load symbol table into memory, enable kbd clock.
1993-08-28 01:18:43 +00:00
534b51feac
Declare find_isadev() function.
1993-08-28 00:13:55 +00:00
62164fe535
Add find_isadev() and have sysbeep() use common defines.
1993-08-28 00:13:00 +00:00
548e9c5a3f
Remove warnings.
1993-08-28 00:11:33 +00:00
db43439bb0
i386 Machine dependant changes to handle non-contiguous memory.
1993-08-28 00:02:45 +00:00
a54d809cd8
Need LOAD_ADDRESS for depend pass.
1993-08-27 23:58:20 +00:00
42484186e9
Load at 1M and turn-on MACHINE_NONCONTIG option.
1993-08-27 23:55:28 +00:00
4aca3365cc
i386 Machine dependant changes to handle non-contiguous memory.
1993-08-27 23:52:22 +00:00
538e35852c
ugh.
1993-08-26 09:52:16 +00:00
225917bfe6
0.9 "official" patch 001:
...
(1) added support for ed1 in both generic kernels at:
device ed1 at isa? port 0x250 net irq 9 iomem 0xd8000 vector edintr
(2) brought if_ed.c up to DG's 1.19 rev.
1993-08-26 00:26:50 +00:00
c5e6262e4e
Add ACCOUNTING and GATEWAY options.
1993-08-24 21:39:35 +00:00
4543c1ebfd
changes from 0.9-ALPHA2 to 0.9-BETA
1993-08-23 04:59:38 +00:00
c41b22b4f6
Removed CLK_TCK, it's value is not machine dependant.
...
It's supposed to be defined in <time.h>, not <limits.h> anyway.
1993-08-18 00:22:44 +00:00
35fdf4ad74
Declare ntohl, ntohs, htonl, and htons functions.
...
Use GCC's inline assembly iff we are using GCC.
Use bswap instruction if using gcc and defined(KERNEL) &&
((defined(I486_CPU) || defined(I586_CPU)) && !defined(I386_CPU))
1993-08-16 18:30:35 +00:00
af6c95dd23
sigh..
1993-08-14 07:31:28 +00:00
3604916e13
ppp from paul mackerras
1993-08-14 06:29:28 +00:00
38239524ee
Absolutely unimportant change to RCS identifier stuff.
1993-08-14 01:29:36 +00:00
4ed4b0d387
delete a reference to the old 'as' device driver.
1993-08-10 08:42:25 +00:00
342e987d36
Theo runs ppp
1993-08-10 04:01:03 +00:00
e5ea45d312
there are some devices out there which use... NO I/O PORTS!
1993-08-08 09:54:23 +00:00
cd56dd1638
merge in changes from netbsd-0-9-ALPHA2
1993-08-07 07:53:27 +00:00
d7080696e9
Put RCS ids back...
1993-08-04 00:42:51 +00:00
63c0a317bb
Add ALIGNBYTES for new fts().
1993-08-03 21:51:10 +00:00
f21b797b62
Bring to rtc clock code from Mach. Provides resettodr().
1993-08-03 12:33:17 +00:00
4e1ade1731
Further cleanup. Remove IDXSHIFT in favor of using PGSHIFT and an indexed
...
load. (Same code size but faster.)
1993-08-03 07:03:01 +00:00
a2a26e452b
Make coding style a little more consistent and cleanup some constants.
1993-08-03 06:34:35 +00:00
27ba85335c
Emit KERNBASE.
1993-08-03 06:33:55 +00:00
03302d8a3a
Fix bogus use of splx() in is_reset().
1993-08-03 05:06:49 +00:00
c12fc78e87
Nuke RCS log.
1993-08-03 05:06:26 +00:00
015cae18dd
more /* in comment warning fixes. in this case it looks like something
...
was cut and pasted badly :)
1993-08-03 01:53:36 +00:00
4a385c1228
with -Wall would bomb, on incomplete type 'selinfo'. bpfilter.h include
...
moved down, and select.h included.
1993-08-03 01:52:57 +00:00
65456d9ca0
Add a panic() after the swtch() is cpu_exit() just in case, and to shut up
...
GCC.
1993-08-02 23:20:55 +00:00
da7f7e95b4
Put in a gratuitous cast to shut up GCC 2.
1993-08-02 23:19:18 +00:00
52efca41cd
Nuke remaining patchkit headers.
1993-08-02 18:17:24 +00:00
aee4b07b8a
Add RCS identifiers, remove some completely useless RCS logs and patchkit
...
headers, and a few other insignificant changes.
1993-08-02 17:48:44 +00:00
4862b84c92
Add RCS identifiers (this time on the correct side of the branch), and
...
incorporate recent changes in netbsd-0-9 branch.
1993-08-01 19:22:24 +00:00
804be109a6
Feep.
1993-08-01 00:32:27 +00:00
690cae8181
Add RCS indentifiers.
1993-07-31 15:17:49 +00:00
3643c134f1
Change "#endif FOO" to "#endif /* FOO */", to keep gcc -Wall, lint, etc.
...
from complaining about a trivial issue.
1993-07-29 21:42:19 +00:00
7b2afa7fe6
incorporate changes from 0-9-base to 0-9-ALPHA
1993-07-28 02:20:34 +00:00
b74ddcc84b
Allow one to fault in DDB and survive.
1993-07-22 13:04:45 +00:00
556176fc24
Call ddb_init instead of kdb_init.
1993-07-22 13:04:21 +00:00
2103705fc5
Added arpcom counters ipackets, opackets etc.
1993-07-21 04:25:07 +00:00
8141ce5d83
kill damned RCS Log
1993-07-19 16:58:20 +00:00
b9b6bd3397
Add ${DEBUG} to CFLAGS and -f to dbsym.
1993-07-19 16:52:16 +00:00
886f6afa56
Reading file marks or writing end of tape return 0 bytes, AND THEY
...
DON'T SET THE ERROR BIT ON THE BUFFER UNLESS THE AUTHOR OF PHYSIO
IS STUPID!
1993-07-19 16:44:38 +00:00
9bd082f70e
replace jolitz's vfs__bio with a better one from CMU via mw.
...
so, replace vfs__bio, and deal with attendant changes.
1993-07-19 16:38:35 +00:00
570ad76179
Change open and close routines for swapdev back to enodev. What a kluge.
1993-07-19 10:57:30 +00:00
fd13ccf3cc
Fix open and close entries for swap in bdevsw.
1993-07-19 10:35:12 +00:00
e916333aea
disable all AUTO_EOI's, at bde's suggestion--
...
they don't gain much, and on old chips it doesn't work properly
1993-07-19 08:27:06 +00:00
e6c1d2670f
Phooey. Need a better mechanism for dealing with multiple sets of local
...
changes.
1993-07-18 21:26:34 +00:00
1318be7191
CC should be cc. cc is gcc2...
1993-07-18 16:44:16 +00:00
d2e484320a
Add DUMMY_NOPS.
1993-07-18 15:11:31 +00:00
a700cbd173
my bucket of mud has changed a bit..
1993-07-18 13:29:27 +00:00
8fa2ba491b
rework the way effectual NOPs are enabled/disabled to make they delay
...
things in the default case.
1993-07-18 13:08:41 +00:00
2565e5e15a
Change to work with new config stuff for specifying load address.
1993-07-18 10:08:22 +00:00
7b721ac9fa
set mountroot to ufs_mountroot. swapgeneric is currently broken for other
...
reasons (the interface to SCSI disks isn't a clean match, and i'm
not entirely sure what a good way to do it is...)
1993-07-18 09:51:26 +00:00
46570ee334
Oops. Back out a change that shouldn't be committed...
1993-07-18 09:48:49 +00:00
28df966f59
Use new -T option to dbsym.
1993-07-18 09:47:40 +00:00
003ff4e190
Add a `XXX'...
1993-07-18 08:43:19 +00:00
297f54dd01
* copyin/out(), copyin/outstr(), s/fubyte() et al protection fixes,
...
including COW support, using code based on that from Wolfgang Solfrank and
Christoph Robitschko.
* user segment descriptors fixed to disallow access to user area.
* bde's boot >= 1MB fixes.
1993-07-18 08:22:59 +00:00
c558337807
Add some kitchen sinks.
1993-07-17 16:54:16 +00:00
f18d296b3c
Stylistic change to attach message to match others.
1993-07-17 16:41:27 +00:00
331148c31a
Make these files GCC 2 happy.
1993-07-17 16:20:24 +00:00
ad087b06a0
Finish moving struct definitions outside of function declarations.
1993-07-17 15:56:59 +00:00
3aed8ce5e5
Fixed two occarances of ldos which should have been lods.
...
(From Christoph Robitschko)
1993-07-17 12:28:47 +00:00
97c38bc76a
gcc-2.4.5's ffs() has been rendered safe for kernel compiling...
1993-07-17 12:22:00 +00:00
2ee2073a53
scsi_switch had wrong entries for these controllers...
1993-07-17 11:00:09 +00:00
45efbe01cd
#include cpufunc,h so inb() and outb() are inlined.
1993-07-16 15:44:22 +00:00
534b762f34
patch from <ache@astral.msk.su> to fix problems:
...
nestart() logic totally rewriten, neput() code moved out of loop.
Fixed error in rxend buffer size: assing 32768 to short.
nerecv() fixed for correct buffer sizes for NE1000.
Detected card type displayed at autoconf time.
1993-07-15 12:57:05 +00:00
ba408056a6
Use splnet, not splimp.
1993-07-13 21:53:06 +00:00
e086df5a24
Fix addressing bug in bdb junk.
1993-07-13 21:51:36 +00:00
499f536ddc
up default data size limit to 16M, like on freefall, so everything will
...
compile properly with gcc2.
1993-07-13 10:16:40 +00:00
bcfa1d8dd3
Fix handling of bad syscall numbers.
1993-07-12 13:53:36 +00:00
6cc173695a
moved bfdttach point to same place as other drivers, from greenman
1993-07-12 13:13:41 +00:00
e28cd102b5
Add some more swap.
1993-07-12 12:07:46 +00:00
395bc5f7a2
physmem is in clusters, not in bytes... that wasn't
...
taken into account in the buffer size calculations
1993-07-12 12:00:45 +00:00
8401efb773
add -fno-builtin to CFLAGS until gcc bug fixed
1993-07-12 11:58:27 +00:00
3be4221095
Change tty code to use clist interface, but with ring buffer implementation.
...
Also, fix a couple of bugs in tty.c and pccons.c, and some gross kluginess
in the hp300 stuff.
1993-07-12 11:36:53 +00:00
c3f293f92b
Make splimp = splnet | spltty, and get rid of SLIP kluge for splnet and
...
spltty.
1993-07-11 14:12:40 +00:00
0daaa4f16c
Make splimp = splnet | spltty, and get rid of SLIP kluge for splnet and
...
spltty.
1993-07-11 14:03:51 +00:00
14f33d91ea
Made boot up messages cleaner
1993-07-11 12:12:35 +00:00
84d1e74079
Fixes from bde, including support for loading @ any MB boundary (e.g. a
...
kernel linked for 0xfe100000 will load at the 1MB mark) and read-ahead
buffering to speed booting from floppies. Also works with aha174x
controllers in enhanced mode.
1993-07-11 12:02:17 +00:00
6d1a5a2ad5
New isolan driver. Multiple card and bpf support added.
1993-07-11 11:18:23 +00:00
4909db9a7e
do quoting right, and give boat-anchor enough symbol space
1993-07-11 10:29:55 +00:00
62ddceb6fc
Yah, yah...
1993-07-11 09:53:44 +00:00
7d3cf97d73
Oops. Back out clist changes, but leave the bug fix.
1993-07-11 09:45:07 +00:00
4d187f65ae
DDB makes these kernels too large. damn.
1993-07-11 09:03:13 +00:00
cc86d5be93
don't ignore errors from dbsym... it might say that, e.g. there's
...
not enough symbol space!
1993-07-11 08:42:22 +00:00
dd71ecc3aa
This is so utterly revolting I refuse to describe it, except to say that it
...
fixes the problem of console output hogging the CPU.
1993-07-11 07:54:47 +00:00
1695b91b15
need more space for symbols than DDB normally gives you...
1993-07-11 07:28:09 +00:00
3e2b1078a3
kill an extraneous # which was annoying cpp
1993-07-11 06:09:48 +00:00
083cf2243a
Fix some probs with stack tb's. Deal with syscalls.
1993-07-10 03:26:44 +00:00
6d65f0bf71
Cleanup warnings and add netbsd kernel name suffix.
1993-07-09 15:24:10 +00:00
4cbf5e04e2
update for better FP routines, from AT&T & elsewhere
1993-07-09 09:39:13 +00:00
24ee0231e8
Diskless boot prom code from Jim McKim (mckim@lerc.nasa.gov)
1993-07-08 16:03:44 +00:00
770db91c65
add "diediedie", a function to make the system panic & dump core
...
in a nice, predictable fashion. useful for DDB, when you want
a core... (i.e. "call diediedie")
1993-07-08 09:55:54 +00:00
80dc38edb3
documentation error
1993-07-08 08:22:17 +00:00
e4c42383fb
Implement scroll lock using sleep() and wakeup() so we don't get stuck at
...
spltty and freeze everyone.
1993-07-08 07:27:29 +00:00
f485ebbacc
get rid if wfj's stack limit check-disabling bogosities. (his old
...
exec needed them. stoopid...)
1993-07-08 04:01:13 +00:00
1eacd36710
pccons.c now dynamically allocates it's "struct tty"
...
cons.c's "struct tty *cn_tty" wasn't used by any of the kernel, and goes away.
1993-07-07 11:00:23 +00:00
bab148ba4e
delete inclusion of rlist.h again. grr.
1993-07-07 06:52:39 +00:00
aa2992a474
make the rlist code all go away. replace it with resource map code,
...
as written by Wolfgang Solfrank.
1993-07-07 06:02:26 +00:00
9bff90a5c1
added a timeout debugging mode to wd.c so that people can see how slow
...
their controllers really are. Look for the WDCNDELAY_DEBUG definition.
1993-07-07 05:50:35 +00:00
d670635e8b
ok, 5 second command timeouts..
1993-07-07 01:05:42 +00:00
ceec40c98f
clean up code for timeout/untimeout/wakeup prototypes.
1993-07-06 06:06:26 +00:00
21e01d6cd2
change the way dump info is printed back to the original (net/2) way,
...
rather than the jolitz way...
1993-07-06 00:48:06 +00:00
86999439b2
number of changes:
...
1. controller reset code moved into one function, called from more places now.
2. more places now do timeouts. TIPCAT code turned on.
3. blew away the unused WDOPENLOCK test code.
4. delay is now done using WDCNDELAY repetitions of DELAY(25). Up to 2.5 second
pause accepted from controller (some controllers have some commands that are
actually this slow)
5. some of the old timeout code was really whacked.
1993-07-06 00:42:43 +00:00
2d53469109
get rid of extraneous inclusion of rlist.h
1993-07-06 00:31:52 +00:00
2cb34d6d65
added a missing {/} pair that barsoom seems to have taken out, but
...
which the standard 386bsd 0.1 driver did have. This is for a piece of
code in the bad block handling. found/fixed by hpeyerl.
1993-07-05 04:44:48 +00:00
3322c7b691
all timeout code now calls DELAY() so the timeout delays are the same
...
length on all processors. Where possible, the controller is reset if it
times out.
1993-07-05 03:20:57 +00:00
b274bb4126
add support for checking entry against max user address
1993-07-05 01:53:56 +00:00
748e864b38
No need for the DELAY() macro. in userland. It just polutes the namespace.
1993-07-05 01:12:46 +00:00
ed61b76822
increased the timeout inside wdcommand(). Some controllers impliment some
...
commands very slowly.
1993-07-05 00:08:27 +00:00
c286f8c4fd
move julian's scsi files back into files.i386
1993-07-04 20:08:15 +00:00
abf0a9b802
do the right thing, so things aren't built in /usr/src/sys
1993-07-04 14:02:53 +00:00
d6325033cd
make i386 setregs do the "normal thing". should probably do something
...
better with retval... (currenty does nothing)
1993-07-04 10:48:39 +00:00
12b46ebf29
Got rid of the SHOW_STI I had added to the BPTTRAP() macro - it breaks
...
(in the type 9 trap and into ddb sense) the kernel's breakpoint handling.
1993-07-04 05:49:17 +00:00
e50f336a76
kernel profiling says that i should make "over_net_stuff_for_spl0" a local
...
label! (i.e. i missed this one before...)
1993-07-03 13:13:34 +00:00
77219acd07
change interrupt vector names to begin with "X" rather than "V", so we're
...
like every other BSD system on the planet...
1993-07-03 12:32:25 +00:00
7e940b55e0
add support for David Greenman "ed" driver
1993-07-03 12:21:07 +00:00
d77e126ab2
add support for David Greenman's "ed" driver.
1993-07-03 12:19:36 +00:00
b8a46f69af
Unless INTR_DEBUG is set, don't report all of the "random" event types.
...
also, unless INTR_DEBUG is set, make the internal branch labels for
the interrupt code be local labels, so profiling works OK.
1993-07-03 09:43:11 +00:00
mycroft