Aren/NetBSD
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
166,967 Commits 606 Branches 0 Tags 3.1 GiB
a686e4e754
Commit Graph

166967 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
christos
a686e4e754 back all changes out until I fix it properly. 2008-02-15 21:29:50 +00:00
christos
4b3df4502b Improve dh's fix: 
- make sure i is initialized
	- don't subtract 1 in comparison, add 1 to the other side
 2008-02-15 20:08:11 +00:00
matt
4498b1fe25 Fix inconsistent definitions 2008-02-15 17:26:06 +00:00
elad
d6faea9863 Allow no signatures file to be specified for the "load" command, and use 
the default one in such cases.

Update command usage and documentation to reflect the above.
 2008-02-15 15:33:56 +00:00
ad
fb00b83874 Give bbusy() an interlock argument. If the we need to wait for the buffer, 
the interlock is dropped and reacquired when awoken. This allows for
busying buffers attached to a list that is not locked by bufcache_lock.
 2008-02-15 13:46:04 +00:00
ad
b2fa822a33 The buffer LOCKED flag need not be under the protection of bufcache_lock, 
BUSY is enough.
 2008-02-15 13:30:56 +00:00
gson
f3b879c064 Fix unaligned write to PCI configuration space. Should fix PR kern/37824. 2008-02-15 13:28:25 +00:00
ad
b0dbb911e1 Regen. 2008-02-15 13:10:23 +00:00
ad
653195d46d Mark mount/unmount MPSAFE. PR kern/38034. 2008-02-15 13:09:57 +00:00
ad
74d14c2433 Add dummy hardclock_ticks. 2008-02-15 13:07:02 +00:00
ad
13df0c89b3 vflush: yield at least every 100ms. PR kern/38034. 2008-02-15 13:06:02 +00:00
tsutsui
00bd6f7500 Remove obsolete platform support options. 2008-02-15 12:28:09 +00:00
skrll
4cfa0739ca Cast via (uintptr_t) in RAS_ADDR to deal with hppa RAS_DECL weirdness. 2008-02-15 09:50:40 +00:00
he
75194e6a79 Typo correction: s/than/that/ in one instance. 2008-02-15 09:24:52 +00:00
dholland
cb84a8637d fix typo in comment 2008-02-15 09:18:56 +00:00
dholland
a9c8da72a1 Avoid SIGSEGV caused by signed/unsigned mismatch. (hi xtos) 2008-02-15 08:55:31 +00:00
uwe
16d9a389fe sh4_switch_resume: Now that r2 is free use it to keep SH4_PTEH to 
avoid re-fetching it uncached while running on P2.
 2008-02-15 04:37:37 +00:00
uwe
cbd01ac88b Prettify sh4_switch_resume - <TAB> between operands only makes the 
code hard to read as any intdended "neatly lined up" effect is
destroyed by long label names anyway.

G/c unused SH4_ITLB_AA.

No functional change.
 2008-02-15 04:11:08 +00:00
uwe
1e93965e99 Use dt instruction for the loop, free r2 we used as a counter. 2008-02-15 04:03:17 +00:00
uwe
b5a1661b3f sh4_switch_resume: don't jump back to P1 b/c we are going to return 
immediately after that anyway and our caller is in P1.

Besides we were not switching to P1 anyway, b/c mova is pc-relative so
without additional arithmetic we were still jumping to a P2 address.
 2008-02-15 03:52:15 +00:00
uwe
10bd77c376 We don't need to/can't wire u-area if we use P1_STACK. 
Not actually tested the P1_STACK case, but I guess it was never tested
on SH4 anyway (it would write garbage to TLB).  It used to work on SH3
b/c "wiring" was done using curupte hack and that was just an
innocuous pointer update.
 2008-02-15 03:16:10 +00:00
uwe
2a753fff49 G/c vestiges of curpute hack that was obsoleted by the fast TLB miss handler. 
sh3_switch_setup() has nothing to do now so don't bother calling it.
 2008-02-15 03:02:43 +00:00
uwe
5130d8b2db G/c vestiges of curpute hack that was obsoleted by the fast TLB miss handler. 
sh3_switch_setup() has nothing to do now.
 2008-02-15 02:55:57 +00:00
christos
e654b454bc remove error/noerror const 2008-02-15 02:50:53 +00:00
christos
649f359b39 fix != expansion bug I introduced. 2008-02-15 02:38:07 +00:00
christos
4d9d6704c1 - symbolic names for error and no error 
- more size_t
 2008-02-15 02:37:51 +00:00
uwe
8b441bdff6 G/c vestiges of curpute hack that was obsoleted by the fast TLB miss handler. 2008-02-15 02:34:46 +00:00
uwe
cde91eb236 If we are returning from a nested interrupt we know we are returning 
to kernel so we can skip the PSL_MD check.
 2008-02-14 22:46:14 +00:00
christos
2b28370627 - use pid_t/size_t as appropriate instead of int. 
- use %ld to print pids.
- fix a bit of lint.
- WARNS=4
 2008-02-14 22:11:20 +00:00
ad
377f098ab0 Adjust mutex/rwlock definitions to match reality now that there is only 
one implementation of each. PR lib/38030.
 2008-02-14 21:40:51 +00:00
garbled
86a2448d10 Some powerpc cleanup. Remove unneeded/bad usage of extern oeacpufeat. 
Convert asm code to use %r register format.  Done by comparison to
disassembled output, double checked with diff of dissasembled output
before and after, and test booted on my 7044.
 2008-02-14 19:41:54 +00:00
tnn
c4bfcd4e81 Add a history section and a compatibility section. 2008-02-14 18:08:35 +00:00
ad
bd48672369 - Clarify how this works. 
- Note the locking rules.
 2008-02-14 17:39:50 +00:00
ad
d4c279feec - Remove pointless AUTHORS section. 
- Remove implementation details.
 2008-02-14 17:25:51 +00:00
he
663b0ebf3e Don't include <powerpc/oea/pmap.h>, use <powerpc/pmap.h> instead. 
Discussed with garbled@, fixes build problem for LKMs.
 2008-02-14 15:21:22 +00:00
ad
ac336fcd77 Remove 'contributed to' text from the TNF license on this file, so that 
it can not be complained about.
 2008-02-14 15:01:45 +00:00
ad
60c1b8843d Make schedstate_percpu::spc_lwplock an exernally allocated item. Remove 
the hacks in sparc/cpu.c to reinitialize it. This should be in its own
cache line but that's another change.
 2008-02-14 14:26:57 +00:00
ad
71eb65050c Make sparc use atomic_cas_ulong() for mutex and rwlock operations, and 
disable the custom mutex/rwlock code. PR port-sparc/37930. ok martin@
 2008-02-14 14:07:35 +00:00
yamt
8ae4ab89dd use time_uptime instead of getmicrotime() for ph_time. 2008-02-14 11:45:24 +00:00
matt
a54b0d709b Add multiple inclusion protection. 
add oeacpufeat.
 2008-02-14 05:35:36 +00:00
joerg
2fa7e92d02 Missing assym.h dependencies. 2008-02-14 00:35:56 +00:00
joerg
7eb4a6d1de Fix syntax. 2008-02-14 00:25:39 +00:00
joerg
34021e2617 Add missing assym.h dependencies. 2008-02-14 00:09:36 +00:00
joerg
cf13ecfa10 GC unused variable. 2008-02-14 00:04:51 +00:00
joerg
75501168d3 Add missing assym.h dependencies. 2008-02-13 23:37:17 +00:00
garbled
53ee4a269c Fix a botch in the comparison of two addresses, where I forgot to add in 
the offset of a range.  With this fix to the openpic initialization,
ofppc now boots single user on a 7046-B50.  Thanks to Jochen Kunz for
giving me remote access to a B50 to hack on it.
 2008-02-13 21:12:32 +00:00
garbled
dc6a7fd211 Add support to ofwboot for booting on the 7046-B50. Also likely to work 
on the 7043-150.  The B50 wants not only a note at the head of the
binary, like the 7044 did, but it also wants a PHDR that points at it.
Because the IBM firmware doesn't like the real note section that ld
generates (it puts it at the end of the file), we instead point PT_NOTE
at the text section.  This works, because at the top of the text section,
there is a note hacked in, which was required for the 7044.

Also, change the mode from virtual mode to real mode in the note, because
the B50 hangs if you load it in virtual.  Tested to work on a B50, 7044,
and Pegasos2.

Yay IBM.
 2008-02-13 20:11:38 +00:00
drochner
85b70008e0 make the mutex securing a flag which controls screen switched adaptive 
for sanity -- it is not yet in a code path which is called from
keyboard interrupt handlers
 2008-02-13 19:01:35 +00:00
drochner
e6c8051893 -if an RSDT entry points to a table with invalid checksum, print 
a warning and continue, rather than aborting -- some BIOSes disable
 tables this way (in my case the TCPA table if the TPM is disabled)
-no terminating newline in err(3) calls
 2008-02-13 18:59:18 +00:00
matt
88f1e662d1 #include <sys/cpu.h> for curlwp 2008-02-13 18:43:16 +00:00