Commit Graph

8096 Commits

Author SHA1 Message Date
Warner Losh
d3b6477a19 It turns out that while Toshiba laptops don't want to route interrupts
multiple times, others do.  The last strategy, which was to assume
that already routed interrupts were good and just return them doesn't
work for some laptops.  So, instead, we have a new strategy: we notice
that we have an interrupt that's already routed.  We go ahead and try
to route it, none the less.  We will assume that it is correctly
routed, even if the route fails.  We still assume that other failures
in the bios32 call are because the interrupt is NOT routed.

Note: some laptops do not support the bios32 interface to PCI BIOS and
we need to call it via the INT 2A interface.  That is another windmill
to till at later.

Also correct a minor typo and minor whitespace nits.

Strong MFC candidate.
2001-08-28 16:35:01 +00:00
Warner Losh
d626906b0c MFS: IRQ ordering, PRVERB and more whining in pcibios_get_version on failure.
Check return value from bios32.

[[ Yes, I was bad and committed this to stable first.  I should have done
   the commit in the other order. ]]
2001-08-27 20:44:38 +00:00
Peter Wemm
9d170964e4 There is nothing more embarresing than having three goes at correcting
typos in the same paragraph.  s/in in/in/

Submitted by:	iedowse
2001-08-27 05:18:12 +00:00
Peter Wemm
76cb0cadf1 Enable hardwiring of things like tunables from embedded enironments
that do not start from loader(8).
2001-08-27 05:11:53 +00:00
Peter Wemm
d675a0adcc I missed a typo in the last commit: s/whach/which/
Submitted by:	bde
2001-08-26 21:55:54 +00:00
Peter Wemm
359ebc389c Argh! Revert accidental commit. 2001-08-25 07:42:27 +00:00
Peter Wemm
547a9e66fd vm_page_zero_idle() is no longer MD. 2001-08-25 04:54:25 +00:00
Julian Elischer
15dac10b05 Add another comment.
check for 'teh's this time..
2001-08-25 02:44:38 +00:00
Peter Wemm
268bdb43f9 Optionize UPAGES for the i386. As part of this I split some of the low
level implementation stuff out of machine/globaldata.h to avoid exposing
UPAGES to lots more places.  The end result is that we can double
the kernel stack size with 'options UPAGES=4' etc.

This is mainly being done for the benefit of a MFC to RELENG_4 at some
point.  -current doesn't really need this so much since each interrupt
runs on its own kstack.
2001-08-25 02:20:02 +00:00
Peter Wemm
61af199168 s/teh/the/ 2001-08-25 02:06:59 +00:00
Julian Elischer
6a9344d45a Add an explanatory note that would have saved me an hour or two
of confusion had it been there when I started reading the code..
2001-08-25 01:16:47 +00:00
John Baldwin
e5dfa86a70 Axe a commented, unused #define related to the old giant lock. 2001-08-24 16:52:31 +00:00
John Baldwin
134b29765c INTREN() and INTRDIS() have been MP safe for a while. Don't say that they
are "soon to be" MP safe in the comments.
2001-08-24 16:51:47 +00:00
John Baldwin
a0f831f573 Remove references to the old giant kernel lock in various comments. 2001-08-24 16:46:15 +00:00
Peter Wemm
c1c2006ca7 Export the actual KERNBASE to the symbol table. We can use nlist() to get
this without having to second guess it in userland.
2001-08-24 08:29:54 +00:00
Peter Wemm
5897d41109 Move cpu_fxsr definition to C code (so debug info is generated) and where
it is easily #ifdef'ed so that we dont miss unintentional references to it.
2001-08-24 08:27:24 +00:00
John Baldwin
1432aa0c5e Add a new kernel option RESTARTABLE_PANICS. If this option is present,
then one can restart from a panic by resetting the panicstr variable to
NULL.  This commit conditionalizes the previously committed functionality
on this variable.  It also removes the __dead2 attribute from the panic()
function so that when one continues from a panic() the behavior will
be predictable.
2001-08-23 20:32:21 +00:00
Peter Wemm
66875a2aff Fix a comment error that was fixed in the pc98 version. hw.maxmem is
really hw.physmem.
2001-08-23 01:49:12 +00:00
Peter Wemm
9f9360003f Dont add UPAGES to the %cs segment limit. There is nothing there except
page tables.
2001-08-23 01:25:21 +00:00
Peter Wemm
3b703181e3 Dont compile in SSE fxsave/fxrstor instructions if CPU_ENABLE_SSE isn't
active.
2001-08-23 01:03:56 +00:00
Mitsuru IWASAKI
e4e9098988 Move CR4.PGE enabling code after paging is enabled via CR0.PG based on
the description (2.5. CONTROL REGISTERS) of Intel developer's manual at:
ftp://download.intel.com/design/PentiumII/manuals/24319202.pdf

Reviewed by:	peter, bde, tlambert2@mindspring.com
Pointed-out by:	"Shin'ya Kumabuchi" <kumabu@t3.rim.or.jp>
MFC after:	1 week
2001-08-22 15:41:43 +00:00
Matthew Dillon
219d632c15 Move most of the kernel submap initialization code, including the
timeout callwheel and buffer cache, out of the platform specific areas
and into the machine independant area.  i386 and alpha adjusted here.
Other cpus can be fixed piecemeal.

Reviewed by:    freebsd-smp, jake
2001-08-22 04:07:27 +00:00
Peter Wemm
3d6fde76ed Introduce two new sysctl's.. vm.kvm_size and vm.kvm_free. These are
purely informational and can give some advance indications of tuning
problems.  These are i386 only for now as it seems that the i386 is
the only one suffering kvm pressure.
2001-08-22 00:50:46 +00:00
John Baldwin
3a9e0f5bd1 Push down Giant some in trap_pfault() so we don't grab Giant around
trap_fatal() to make restarting from panic's slightly easier.  Before if
one did 'w 0 0' in ddb, the longjmp in ddb inside of trap_fatal() would
result in Giant being held (or recursed one level deeper) which led to
problems later on.  You can now drop to teh debugger, do 'w 0 0', and
continue w/o a problem.
2001-08-21 23:15:25 +00:00
Warner Losh
0b9427de88 The general conesnsus on irc was that pci bios for config registers
and such was just a bad idea and one that users should be forced to
enable if they want it.  This patch introduces a hw.pci.enable_pcibios
tunable for those people.  This does not impact the pcibios interrupt
routing at all.

Approved by: peter, msmith
2001-08-21 07:53:37 +00:00
Matthew Dillon
6db5812aef Fix bug in physmem_est calculation - the kernel_map size was not being
converted into pages.

Fix bug in maxbcache calculation, nbuf must be tested against maxbcache
rather then physmem_est.

Obtained from:	bde
2001-08-21 07:20:06 +00:00
Peter Wemm
573be82757 Detect a certain type of PCIBIOS brain damage. For some reason,
some bios vendors took it apon themselves to "censor" the
host->pci bridges from PCIBIOS callers, even when the caller
explicitly asks for them.  This includes certain Compaq machines
(eg: DL360) and some laptops.

If we detect this, shut down pcibios and revert to using IO
port bashing.

Under -current, apcica does a better job anyway.
2001-08-21 03:10:55 +00:00
Peter Wemm
e8ebc08f80 Make COMPAT_43 optional again. XXX we need COMPAT_FBSD3 etc for this
stuff.
2001-08-21 02:32:59 +00:00
Matthew Dillon
2f9e4e8025 Limit the amount of KVM reserved for the buffer cache and for swap-meta
information.  The default limits only effect machines with > 1GB of ram
and can be overriden with two new kernel conf variables VM_SWZONE_SIZE_MAX
and VM_BCACHE_SIZE_MAX, or with loader variables kern.maxswzone and
kern.maxbcache.  This has the effect of leaving more KVM available for
sizing NMBCLUSTERS and 'maxusers' and should avoid tripups where a sysad
adds memory to a machine and then sees the kernel panic on boot due to
running out of KVM.

Also change the default swap-meta auto-sizing calculation to allocate half
of what it was previously allocating.  The prior defaults were way too high.
Note that we cannot afford to run out of swap-meta structures so we still
stay somewhat conservative here.
2001-08-20 00:41:12 +00:00
Peter Wemm
4faa812afa There is nothing special that requires SSE to be only on 686 class cpus.
This enables 586-only SMP kernels to compile again.

Problem reported by:  Jacek Jedrzejczak <jacol@ids.gda.pl>
2001-08-18 04:04:45 +00:00
David E. O'Brien
589278dbae style(9) and make consistent across platforms 2001-08-16 09:29:35 +00:00
Andrey A. Chernov
2a54e09dff OFF_T -> OFF (more standard style) 2001-08-15 19:50:59 +00:00
Andrey A. Chernov
527b997c3d Add OFF_T_MAX/OFF_T_MIN 2001-08-15 19:12:46 +00:00
Bill Paul
3708f8da9c Teach bus_dmamem_free() about contigfree(). This is a bit of a hack,
but it's better than the buggy behavior we have now. If we contigmalloc()
buffers in bus_dmamem_alloc(), then we must configfree() them in
bus_dmamem_free(). Trying to free() them is wrong, and will cause
a panic (at least, it does on the alpha.)

I tripped over this when trying to kldunload my busdma-ified if_rl
driver.
2001-08-15 17:26:54 +00:00
John Baldwin
2be7b27a41 Whitespace fixes to make this mostly fit in 80 columns. 2001-08-15 16:57:12 +00:00
Bruce Evans
3d66f80905 Use interrupt gates instead of trap gates for breakpoint and trace
traps, so that ddb can keep control (almost) no matter how it is
entered.  This breaks time-critical interrupts while the system is
stopped in ddb, but I haven't noticed any significant problems except
that applications become confused about the time.  Lost time will be
adjusted for later.  Anyway, the half-baked disabling of interrupts in
Debugger() gives the same problems for the usual way of entering ddb.
2001-08-13 13:22:50 +00:00
Bruce Evans
d072acf902 Removed he BPTTRAP() macro and its use. It was intended for restoring
bug for bug compatibility to ddb trap handlers after fixing the debugger
trap gates to be interrupt gates, but the fix was never committed.  Now
I want the fix to apply to ddb.
2001-08-13 12:42:13 +00:00
Bruce Evans
807ef07ea1 Fixed minor numbers when there is more than one cy card.
PR:		19256
Submitted by:	initial version by yokota
MFC after:	1 week
2001-08-13 10:52:15 +00:00
Bruce Evans
0ce18d0354 Use ttymalloc() instead of a static array of `struct tty'. This will
be a regression until `pstat -t' actually understands the results of
ttymalloc().

Submitted by:	mostly by yokota
2001-08-13 07:00:21 +00:00
Mitsuru IWASAKI
dfc356335a Fix some trivial bugs.
- fix segment limit mis-calculation for GCODE_SEL, GDATA_SEL, GPRIV_SEL,
   LUCODE_SEL and LUDATA_SEL.
 - move `loader(8) metadata' related printf() after cninit().
 - use atop macro (address to pages) for segment limit calculation
   instead of i386_btop macro (bytes to pages).
 - fix style bugs for the declarations of ints.

Reviewed by:    bde, msmith (and arch & audit ML)
2001-08-12 08:37:39 +00:00
John Baldwin
0f345c4054 Add an optimization where we check hte PS_ASTPENDING and PS_NEEDRESCHED
flags with interrupts disabled to see if we should call ast() during
doreti.  This was mostly submitted by Bruce, but his original patch did
the looping in ast() in assembly rather than in the ast() function itself.
Once we've actually called into the ast() function, it's cheaper to just
loop inside the function rather than returning from the function,
performing the check, and then calling the function again.  However, we
can optimize the first check to avoid calling the function at all.
Other architectures may choose to implement this optimization if they
wish but it is not required for correct operation.

Submitted by:	bde
2001-08-10 23:00:07 +00:00
John Baldwin
688ebe120c - Close races with signals and other AST's being triggered while we are in
the process of exiting the kernel.  The ast() function now loops as long
  as the PS_ASTPENDING or PS_NEEDRESCHED flags are set.  It returns with
  preemption disabled so that any further AST's that arrive via an
  interrupt will be delayed until the low-level MD code returns to user
  mode.
- Use u_int's to store the tick counts for profiling purposes so that we
  do not need sched_lock just to read p_sticks.  This also closes a
  problem where the call to addupc_task() could screw up the arithmetic
  due to non-atomic reads of p_sticks.
- Axe need_proftick(), aston(), astoff(), astpending(), need_resched(),
  clear_resched(), and resched_wanted() in favor of direct bit operations
  on p_sflag.
- Fix up locking with sched_lock some.  In addupc_intr(), use sched_lock
  to ensure pr_addr and pr_ticks are updated atomically with setting
  PS_OWEUPC.  In ast() we clear pr_ticks atomically with clearing
  PS_OWEUPC.  We also do not grab the lock just to test a flag.
- Simplify the handling of Giant in ast() slightly.

Reviewed by:	bde (mostly)
2001-08-10 22:53:32 +00:00
Peter Wemm
74a40576eb Add the screen savers for test coverage. 2001-08-09 00:05:30 +00:00
Peter Wemm
2aca0c28d3 Zap 'ptrace(PT_READ_U, ...)' and 'ptrace(PT_WRITE_U, ...)' since they
are a really nasty interface that should have been killed long ago
when 'ptrace(PT_[SG]ETREGS' etc came along.  The entity that they
operate on (struct user) will not be around much longer since it
is part-per-process and part-per-thread in a post-KSE world.

gdb does not actually use this except for the obscure 'info udot'
command which does a hexdump of as much of the child's 'struct user'
as it can get.  It carries its own #defines so it doesn't break
compiles.
2001-08-08 05:25:15 +00:00
Nate Williams
d4b85e6acf -Finished cleanup of old 'ThinkPad' comments that are no longer useful.
Reminded by:	bde
2001-08-06 16:04:39 +00:00
Nate Williams
3633ea98d7 - Removed comment about ThinkPad keyboards from the PCVT line. Any ThinkPad
that needs this probably won't run -current, as it's at least 5 years old.
2001-08-05 21:56:11 +00:00
John Baldwin
c6da9f6a8c - Fix a comment.
- Whitespace fixes.

Submitted by:	bde
2001-08-04 20:43:48 +00:00
Jake Burkholder
7e5102989e Use a machine dependent type, Elf_Hashelt, for the elements of the elf
dynamic symbol table buckets and chains.  The sparc64 toolchain uses 32
bit .hash entries, unlike other 64 bits architectures (alpha), which use
64 bit entries.

Discussed with: dfr, jdp
2001-07-31 03:46:39 +00:00
Mark Murray
7d42654638 Diff-reduce this with GENERIC.
OK'ed by:	imp
2001-07-28 20:40:49 +00:00
Peter Wemm
0b27d7104f Make PMAP_SHPGPERPROC tunable. One shouldn't need to recompile a kernel
for this, since it is easy to run into with large systems with lots of
shared mmap space.

Obtained from:	yahoo
2001-07-27 01:08:59 +00:00