Commit Graph

1399 Commits

Author SHA1 Message Date
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
Jake Burkholder
f02f55a8e3 Correct copyright language. 2001-08-10 04:51:19 +00:00
Jake Burkholder
416d9c00fa Add code to program the tick register and to setup its interrupt handler. 2001-08-10 04:49:33 +00:00
Jake Burkholder
228fa56391 Add early code to support interrupts. 2001-08-10 04:48:48 +00:00
Jake Burkholder
f69f6debc3 Fake up the frame pointers on a process's initial stack so they can be
restored correctly from the trapframe.

Submitted by:	tmm
2001-08-10 04:47:14 +00:00
Jake Burkholder
22693a0fef Handle all types of mmu misses from user mode.
Pass a context argument to tlb functions.
2001-08-10 04:44:33 +00:00
Jake Burkholder
a3698a6d7f Use the macro for getting the trap type from the trapframe.
Only set sticks (and acquire sched_lock) on entry from user mode.
Add handlers for all kinds of mmu misses, and for interrupts from
user mode.
Acquire Giant before calling into the vm system so this runs with
invariants.
Try to get the restrictions for page faults on user memory from
kernel mode right.
Only set pcb_onfault and return to the alternate return code if
this is actually a fault on user memory from kernel mode.
2001-08-10 04:43:31 +00:00
Jake Burkholder
c0eedfbccb Store 8 bytes instead of 4 in suword. Use a temporary stack that's known
to be locked in the tlb for calling openfirmware.

Submitted by:	tmm
2001-08-10 04:39:23 +00:00
Jake Burkholder
e822e0eefb Pass a context to tlb_store_slot, use a member(Sync) after setting the
secondary context register.
2001-08-10 04:37:03 +00:00
Jake Burkholder
9a6cacf3dd 1. Start the clock running early for testing.
2. Use the upcoming "tick" interface.
3. Save a call frame as well as a trap frame on proc0's initial stack.
4. Setup a pointer to the per-cpu interrupt queue.
5. Install the per-cpu pointer in interrupt and alternate globals as well.
6. Flush out setregs so exec works.

Submitted by:	tmm (3, 5, 6)
2001-08-10 04:36:00 +00:00
Jake Burkholder
346d41dd6f Set the pil to something sane on startup. 2001-08-10 04:32:13 +00:00
Jake Burkholder
4a7517d17f Add definitions needed by new assembler code. 2001-08-10 04:31:30 +00:00
Jake Burkholder
cc11501e3b 1. Add code to handle traps and interrupts from user mode.
2. Add spill and fill handlers for spills to the user stack on entry
   to the kernel.
3. Add code to handle instruction mmu misses from user mode.
4. Add code to handle level interrupts from kernel mode and vectored
   interrupt traps from either.
5. Save the pil in the trapframe on entry from kernel mode and restore
   it on return.

Submitted by:	tmm (1, 2)
2001-08-10 04:30:36 +00:00
Jake Burkholder
8dec6a681e Add code to handle stack traces that go all the way back to userland.
Use a better algorithm for finding out if an address is in the kernel.

Submitted by:	tmm
2001-08-10 04:23:41 +00:00
Jake Burkholder
8a7de3c054 Add trap types for interrupts. Ad definitions to get the interrupt level
from the trap type.
2001-08-10 04:22:24 +00:00
Jake Burkholder
fa01c0ff70 1. Add code to demap pages from the tlb for user contexts.
2. Add a context argument to most functions, instead of extracting it from
   from the tte.

Submitted by:	tmm (1).
2001-08-10 04:21:44 +00:00
Jake Burkholder
0cb2d36969 Add fields that point to per-cpu interrupt data. 2001-08-10 04:18:48 +00:00
Jake Burkholder
db6c479810 Add a field to trapframe for saving the pil. 2001-08-10 04:18:13 +00:00
Jake Burkholder
c13c9cbf4f Add asis for interrupt registers. 2001-08-10 04:17:29 +00:00
David E. O'Brien
688985be2b The author isn't a [UC] Regents. Correct the copyright language. 2001-08-09 02:32:05 +00:00
David E. O'Brien
b2d4cd8f6e Fix VCS ID spamage. 2001-08-09 02:29:32 +00:00
David E. O'Brien
d410360643 Restore the proper copyright on this and remove the gratuitous changes from
sys/alpha/include/elf.h.
2001-08-09 02:24:05 +00:00
David E. O'Brien
73a4930297 The author isn't a [UC] Regents. Correct the copyright language. 2001-08-09 02:09:34 +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
Jake Burkholder
b3a1f1765a Handle dmmu protection faults as well as misses. Enable tracking of
the modify and reference tte bits.  Implementing allocating of tsb pages.
Make tsb_stte_lookup do the right thing with the kernel pmap.
2001-08-06 02:34:20 +00:00
Jake Burkholder
e15d0e4cf9 Add page fault and high level tsb miss handlers. 2001-08-06 02:32:38 +00:00
Jake Burkholder
a5f2d2a0eb Handle switching switching mmu contexts and mapping the new primary tsb.
Rework some register usage and code placement.  Comment.
2001-08-06 02:31:30 +00:00
Jake Burkholder
04a8a9ba95 Save the primary mmu context around calls to the prom, and install
nucleus context.  The prom runs at trap level 0, so there's no
implicit nucleus context and we have to force it.
2001-08-06 02:29:13 +00:00
Jake Burkholder
cfe8998cd6 Remove some debug code. 2001-08-06 02:27:59 +00:00
Jake Burkholder
9592342320 Handle managed and unmanaged mapping better. Allocate an vm object for
the tsb pages.
2001-08-06 02:27:34 +00:00
Jake Burkholder
2a9921d8bd Add trap handlers for dmmu faults from user mode, and for faults from
accessing user address space in kernel mode.
2001-08-06 02:26:04 +00:00
Jake Burkholder
2de67fb944 The kernel runs at a much lower address now. 2001-08-06 02:24:52 +00:00
Jake Burkholder
2e74c0876f Fix macros for dealing with tte contexts.
Add tte bits for initializing tsbs and for specifying managed mappings.
2001-08-06 02:24:18 +00:00
Jake Burkholder
d15f935586 Oops. Last commit to tsb.h should have gone here.
Fix macros for eadling with tte contexts and add macros for sfsr fields.
2001-08-06 02:21:53 +00:00
Jake Burkholder
3af9da91bc Fix macros for setting and extracting the context field in ttes and
add macros for the fields in sfsr.
2001-08-06 02:20:36 +00:00
Jake Burkholder
8d94222282 Add a vm_object and page count to struct pmap for allocating tsb pages. 2001-08-06 02:19:52 +00:00
Thomas Moestl
491842fc97 Sigh. Add two files needed for the sparc64 fp contect switching code
that were forgotten in the last commit.

Pointy hat to: 	tmm
2001-08-05 03:47:02 +00:00
Thomas Moestl
e64b70af32 Add floating point context switching code for sparc64.
Reviewed by:	jake
2001-08-04 18:55:15 +00:00
Jake Burkholder
7cabcd9841 Move some code related to managing pv entries from the pmap module to
the pv module.  It works now that vtophys for sttes works.
2001-08-03 01:27:15 +00:00
Jake Burkholder
d741ca6bdb Fix a bug translating virtual translation table entry addresses to physical
addresses.  It helps to use the physical address that the virtual address
actually maps to (doh!).  Comment out some code that crashes.

Found independently by:	tmm
2001-08-03 01:21:24 +00:00
Jake Burkholder
3db6135894 Define proc0paddr. Call init_param() as early as possible. 2001-08-03 01:11:49 +00:00
Jake Burkholder
1994658f04 Add an Elfhashelt type for sparc64. 2001-08-03 01:00:41 +00:00
Jake Burkholder
89bf8575ee Flesh out the sparc64 port considerably. This contains:
- mostly complete kernel pmap support, and tested but currently turned
  off userland pmap support
- low level assembly language trap, context switching and support code
- fully implemented atomic.h and supporting cpufunc.h
- some support for kernel debugging with ddb
- various header tweaks and filling out of machine dependent structures
2001-07-31 06:05:05 +00:00
Jake Burkholder
98bb5304e1 Add skeleton machine dependent headers and c files for a port of freebsd
to a new architecture.  This is the base of the sparc64 port, but contains
limited machine dependent code, and can be used a base for ports.  Included
are:
- standard machine dependent headers, tweaked for a 64 bit, big endian
  architecture, including empty versions of all the machine dependent
  structures
- a machine independent atomic.h, which can be used until a port has
  support for interrupts and the operations really need to be atomic
- stub versions of all the machine dependent functions, which panic
  when called and print out the name of the function that needs to
  be implemented.  functions which are normally in assembly files are
  not included, but this should reduce the number of different undefined
  references on the first few compiles from hundreds to 5 or 6
Given minimal startup code and console support it should be trivial to
make this compile and run the first few sysinits on almost any architecture.

Requested by:   alfred, imp, jhb
2001-07-31 05:45:16 +00:00
Warner Losh
0e3cd5fe1b Obrien created this directory, but I didn't cvs add cvsignore 2001-07-01 23:37:03 +00:00
Warner Losh
1f94b9005c Don't need the .keep_me files. Obrien and I committed past each other.
Add 0-9 to the list of possible kernel names at matsushita-san's
suggestion.

Submitted by: Makoto MATSUSHITA-san <matusita@jp.FreeBSD.org>
2001-07-01 23:35:44 +00:00
David E. O'Brien
1453323996 Grrr, this was added in the wrong place.
Have I mentioned before how much I hate remove CVS when adding directories???
Ask me about it (got 3 hours...)
2001-06-30 15:21:34 +00:00
David E. O'Brien
5e6ded4212 Ensure sys/${MACHINE}/compile/FOO exists
Reviewed by: arch, imp, peter, and the USENIX terminal room secret kernel cabal
2001-06-30 15:16:29 +00:00
David E. O'Brien
791eca5f7b Ensure sys/${MACHINE}/compile/FOO exists
Reviewed by: arch, imp, peter and
  the USENIX terminal room secret kernel cabal
2001-06-30 07:12:34 +00:00