1
0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-25 11:37:56 +00:00
Commit Graph

3072 Commits

Author SHA1 Message Date
Peter Wemm
37b087a645 Clean up some low level bootstrap code:
- stop using the evil 'struct trapframe' argument for mi_startup()
  (formerly main()).  There are much better ways of doing it.
- do not use prepare_usermode() - setregs() in execve() will do it
  all for us as long as the p_md.md_regs pointer is set.  (which is
  now done in machdep.c rather than init_main.c.  The Alpha port did it
  this way all along and is much cleaner).
- collect all the magic %cr0 etc register settings into one place and
  have the AP's call that instead of using magic numbers (!!) that keep
  changing over and over again.
- Make it safe to call kthread_create() earlier, including during the
  device probe sequence.  It doesn't need the callback mechanism that
  NetBSD's version uses.
- kthreads created this way are root-less as they exist before the root
  filesystem is mounted.  init(1) is set up so that it aquires the root
  pointers prior to running.  If other kthreads want filesystem acccess
  we can make this code more generic.
- set all threads start times once we have decided what time it is.
- init uses a trampoline rather than the evil prepare_usermode() hack.
- kern_descrip.c has a couple of tweaks to deal with forking when there
  is no rootdir or cwd etc.
- adjust the early SYSINIT() sequence so that a few prereqisites are in
  place. eg: make sure the run queue is initialized before doing forks.

With this, the USB code can easily create a kthread to do the device
tree discovery.  (I have tested it, it works nicely).

There are still some open issues before this is truely useful.
- tsleep() does not like working before the clock is running.  It
  sort-of tries to spin wait, but it can do more useful things now.
- stopping a kthread in kld code at unload time is "interesting" but
  we have a solution for that.

The Alpha code needs no changes for this.  It already uses pretty much the
same strategies, but a little cleaner.
2000-08-11 09:05:12 +00:00
Tor Egge
4428d39d63 Don't skip IOAPIC id conflict detection when only one pci bus is present.
PR:		20312
Reviewed by:	Steve Roome <steve@sse0691.bri.hp.com>
2000-08-10 17:33:24 +00:00
Tor Egge
3c2498c0d3 Don't set flags on the mount structure before all permission checks have
been done.

Don't allow multiple mount operations with MNT_UPDATE at the same
time on the same mount point.  When the first mount operation
completed, MNT_UPDATE was cleared in the mount structure, causing
the second to complete as if it was a no-update mount operation
with the following bad side effects:

        - mount structure inserted multiple times onto the mountlist
        - vp->v_mountedhere incorrectly set, causing next namei
          operation walking into the mountpoint to crash with
          a locking against myself panic.

Plug a vnode leak in case vinvalbuf fails.
2000-08-09 01:57:11 +00:00
Robert Watson
e6a9ab52db o Introduce vn_extattr_{get,set}, wrapper routines for VOP_GETEXTATTR
and VOP_SETEXTATTR to simplify calling from in-kernel consumers,
  such as capability code.  Both accept a vnode (optionally locked,
  with ioflg to indicate that), attribute name, and a buffer + buffer
  length in UIO_SYSSPACE.  Both authorize the call as a kernel request,
  with cred set to NULL for the actual VOP_ calls.

Obtained from:	TrustedBSD Project
2000-08-08 17:15:32 +00:00
Jonathan Lemon
a114459191 Make the kqueue socket read filter honor the SO_RCVLOWAT value.
Spotted by:  "Steve M." <stevem@redlinenetworks.com>
2000-08-07 17:52:08 +00:00
Jonathan Lemon
ad91b6a280 Fix bug with timeout; previously, when attempting to poll the kqueue by
passing a zero-valued timeout, the code would always sleep for one tick.
Change code to avoid calling tsleep if we have no intention of sleeping.

Bring in bugfix from sys_select.c, r1.60 which also applies here.

Modify error handling slightly; passing in an invalid fd will now result
in EBADF returned in the eventlist, while an attempt to change a knote
which does not exist will result in ENOENT being returned.  Previously
such attempts would fail silently without notification.

Pointed out by: nicolas.leonard@animaths.com
	        Rick Reed (rr@yahoo-inc.com)
2000-08-07 16:45:42 +00:00
Paul Saab
c206a8609e Change the behavior of isa_nmi to log an error message instead of
panicing and return a status so that we can decide whether to drop
into DDB or panic.  If the status from isa_nmi is true, panic the
kernel based on machdep.panic_on_nmi, otherwise if DDB is
enabled, drop to DDB based on machdep.ddb_on_nmi.

Reviewed by:	peter, phk
2000-08-06 14:17:21 +00:00
Tor Egge
e666f57c3e Be more verbose when changing APIC ID on an IO APIC.
Don't allow cpu entries in the MP table to contain APIC IDs out of range.

Don't write outside array boundaries if an IO APIC entry in the MP table
contains an APIC ID out of range.

Assign APIC IDs for all IO APICs according to section 3.6.6 in the
Intel MP spec:

  - If the current APIC ID on an IO APIC doesn't conflict with other
    IO APICs or CPUs, that APIC ID should be used.  The copy of the MP
    table must be updated if the corresponding APIC ID in the MP table
    is different.

  - If the current APIC ID was in conflict with other units, the
    corresponding APIC ID specified in the MP table is checked for conflict.

  - If a conflict is still found then fall back to using a new unique ID.
    The copy of the MP table must be updated.

  - IDs out of range is considered to be in conflict.

During these operations, the IO_TO_ID array cannot be used, since any
conflict would have caused information loss.  The array is then corrected,
since all APIC ID conflicts should have been resolved.

PR:	20312, 18919
2000-08-06 00:04:03 +00:00
Jeffrey Hsu
51b86781c0 Modify to use fixed STAILQ_LAST().
Reviewed by:	dfr
2000-08-03 16:37:46 +00:00
Peter Wemm
2c7f8b4ebd Fix self referential dependencies. eg: uhub was packaged along with
usb, all in usb.ko.  uhub depends on usb.  The bug was that the preload
processing only adds a module to the list once it's internal dependencies
are resolved... Since it was not "seeing" the internal usb module it
believed that uhub had a missing dependency.
2000-08-02 21:08:53 +00:00
Peter Wemm
af4b2d2d1c Fix the SYSINIT() bubble sort. This was fixed in kern_linker.c already. 2000-08-02 21:05:21 +00:00
Jonathan Lemon
1dfd47607b Back out rev 1.12; its not clear that this is the right thing to do,
and in any event, it wasn't done correctly in the first place.
2000-08-01 04:27:50 +00:00
Luoqi Chen
3fb50adb4c Handle write page faults (both write only or read-modify-write) as MI vm
write-only faults.  This would allow write-only mmapped regions to function
correctly.
2000-07-31 14:47:14 +00:00
Alfred Perlstein
9ad48853de mbstat should be a read-only sysctl.
Submitted by: Bosko Milekic <bmilekic@dsuper.net>
2000-07-31 09:24:32 +00:00
Paul Saab
030f7b3faa Remove unnecessary call to splnet when setting an accept filter
since we are already at splnet.
2000-07-31 08:23:43 +00:00
Peter Wemm
3a285cc807 Regen. (Fix SYS_exit) 2000-07-29 10:07:38 +00:00
Peter Wemm
4e0f152bbe Sigh. Fix SYS_exit problems. I misunderstood the significance of these
trailing options.
2000-07-29 10:05:25 +00:00
Paul Saab
0e461cb7e2 Remove this file incase of further confusion. 2000-07-29 04:09:07 +00:00
Peter Wemm
69065e880a Regenerate with makesyscalls.sh 2000-07-29 00:21:50 +00:00
Peter Wemm
ac2b067b9a Change the 'exit()' system call to 'sys_exit()'. This avoids overlapping
gcc's internal exit() prototypes and the (futile) hackery that we did to
try and avoid warnings.  main() was renamed for similar reasons.
Remove an exit related hack from makesyscalls.sh.
2000-07-29 00:16:28 +00:00
Peter Wemm
5dec52bada Fix the #ifdef VFS_AIO to not compile a whole bunch of unused stuff in the
!VFS_AIO case.  Lots of things have hooks into here (kqueue, exit(),
 sockets, etc), I elected to keep the external interfaces the same
 rather than spread more #ifdefs around the kernel.
2000-07-28 23:10:10 +00:00
Peter Wemm
f7ce4efc8a Fix a const related warning. 2000-07-28 22:41:56 +00:00
Peter Wemm
93e8459a02 Fix some style nits.
Fix(?) some compile warnings regarding const handling.
2000-07-28 22:40:04 +00:00
Peter Wemm
c828c7b784 Fix warnings - make kevent args in comment match those in syscalls.master.
Deal with consts.
2000-07-28 22:32:25 +00:00
Peter Wemm
b31ae1adc5 Fix a warning that has been annoying me for some time:
"kern/sys_generic.c:358: warning: cast discards qualifiers from pointer
   target type"
The idea for using the uintptr_t intermediate cast for de-constifying
a pointer was hinted at by bde some time ago.
2000-07-28 22:17:42 +00:00
Robert Watson
fc3345a4a7 o Modify extattr_{set,get}() syscalls so that partial reads and writes
with an error condition such as EINTR, EWOULDBLOCK, and ERESTART,
  are reported to the application, not silently conceal.  This
  behavior was copied from the {read,write}v() syscalls, and is
  appropriate there but not here.
o Correct a bug in extattr_delete() wherein the LOCKLEAF flag is
  passed to the wrong argument in namei(), resulting in some
  unexpected errors during name resolution, and passing in an unlocked
  vnode.

Obtained from:	TrustedBSD Project
2000-07-28 19:52:38 +00:00
Jonathan Lemon
ab2adc20f2 Have kevent() automatically restart if interrupted by a signal. If this
is not desired, then the user can register an EV_SIGNAL filter to
explicitly catch a signal event.

Change requested by: jayanth, ps, peter
		     "Why is kevent non-restartable after a signal?"
2000-07-27 23:06:14 +00:00
Brian Feldman
3c89e357f0 Distinguish between whether ktraceing was enabled before an IO
operation or after it.  If the ktrace operation was enabled while the
process was blocked doing IO, the race would allow it to pass down
invalid (uninitialized) data and panic later down the call stack.
2000-07-27 03:45:18 +00:00
Robert Watson
3ce7b7aa84 o Lock vnode before calling extattr_* VOP's, and modify vnode spec to
allow for that.
o Remember to call NDFREE() if exiting as a result of a failed
  vn_start_write() when snapshotting.

Reviewed by:	mckusick
Obtained from:	TrustedBSD Project
2000-07-26 20:29:20 +00:00
Kirk McKusick
54e53ebda7 Now that buffer locks can be recursive, we need to delete the panics
that complain about them.

Obtained from:	Brian Fundakowski Feldman <green@FreeBSD.org>
2000-07-25 18:28:46 +00:00
Kirk McKusick
aec3bbe11c Do not need vrele(nd.ni_vp) as that is done by NDFREE(&nd, 0);
Submitted by:	Peter Holm <pho@freebsd.org>
2000-07-25 05:38:54 +00:00
Robert Watson
e2e45aa8a0 o Add missing function return types from capability syscall call stubs,
fix compiler warning.

Submitted by:	jake
2000-07-25 03:37:36 +00:00
Kirk McKusick
9b97113391 This patch corrects the first round of panics and hangs reported
with the new snapshot code.

Update addaliasu to correctly implement the semantics of the old
checkalias function. When a device vnode first comes into existence,
check to see if an anonymous vnode for the same device was created
at boot time by bdevvp(). If so, adopt the bdevvp vnode rather than
creating a new vnode for the device. This corrects a problem which
caused the kernel to panic when taking a snapshot of the root
filesystem.

Change the calling convention of vn_write_suspend_wait() to be the
same as vn_start_write().

Split out softdep_flushworklist() from softdep_flushfiles() so that
it can be used to clear the work queue when suspending filesystem
operations.

Access to buffers becomes recursive so that snapshots can recursively
traverse their indirect blocks using ffs_copyonwrite() when checking
for the need for copy on write when flushing one of their own indirect
blocks. This eliminates a deadlock between the syncer daemon and a
process taking a snapshot.

Ensure that softdep_process_worklist() can never block because of a
snapshot being taken. This eliminates a problem with buffer starvation.

Cleanup change in ffs_sync() which did not synchronously wait when
MNT_WAIT was specified. The result was an unclean filesystem panic
when doing forcible unmount with heavy filesystem I/O in progress.

Return a zero'ed block when reading a block that was not in use at
the time that a snapshot was taken. Normally, these blocks should
never be read. However, the readahead code will occationally read
them which can cause unexpected behavior.

Clean up the debugging code that ensures that no blocks be written
on a filesystem while it is suspended. Snapshots must explicitly
label the blocks that they are writing during the suspension so that
they do not cause a `write on suspended filesystem' panic.

Reorganize ffs_copyonwrite() to eliminate a deadlock and also to
prevent a race condition that would permit the same block to be
copied twice. This change eliminates an unexpected soft updates
inconsistency in fsck caused by the double allocation.

Use bqrelse rather than brelse for buffers that will be needed
soon again by the snapshot code. This improves snapshot performance.
2000-07-24 05:28:33 +00:00
Brian Feldman
55af4c7d94 Using an atomic operation here won't help if nobody else uses them (for
this).  Use the simple_lock() on v_interlock like elsewhere.
2000-07-23 22:19:49 +00:00
Brian Feldman
25ead03462 Solve the problem where it is possible to get the kernel stuck in
a loop down in pmap_init_pt().  A subtraction causes the number of
pages to become negative, that was assigned to an unsigned variable,
and there is a lot of iteration.  The bug is due to the ELF image
activator not properly checking for its files being the correct size
as specified by the ELF header.

The solution is to check that the header doesn't ask for part of a
file when that part of the file doesn't exist.  Make sure to set
VEXEC at the proper times to make the executables immutable (remove
race conditions).  Also, the ELF format specifiies header entries
that allow embedding of other executables (hence how ld-elf.so.1
gets loaded, but not the same as loading shared libraries), so those
executables need to be set VEXEC, too, so they're immutable.

Reviewed by:	peter
2000-07-23 06:49:46 +00:00
Alfred Perlstein
f408896444 only allow accept filter modifications on listening sockets
Submitted by: ps
2000-07-20 12:17:17 +00:00
Alfred Perlstein
85f5e7f098 disallow unload until we do proper refcounting 2000-07-20 12:12:41 +00:00
Jonathan Lemon
2ba03123c5 Fix a bug which would cause some knotes to get lost when two kqueues
were being used in a process at the same time.

Test case provided by:  Chris Peiffer <peifferc@CS.Stanford.EDU>
2000-07-18 21:41:47 +00:00
Jonathan Lemon
a8e65b915e Simplify kqueue API slightly.
Discussed on:	-arch
2000-07-18 19:31:52 +00:00
Peter Wemm
f03c9f90d1 Patch up some bogons in the resource_find() vs resource_find_hard()
interfaces.  The original resource_find() returned a pointer to an internal
resource table entry.  resource_find_hard() dereferences the actual
passed in value (oops!) - effectively trashing random memory due to
the pointer being passed in with a random initial value.

Submitted by:  bde
2000-07-18 06:08:27 +00:00
Andrzej Bialecki
bd3cdc3105 These patches implement dynamic sysctls. It's possible now to add
and remove sysctl oids at will during runtime - they don't rely on
linker sets. Also, the node oids can be referenced by more than
one kernel user, which means that it's possible to create partially
overlapping trees.

Add sysctl contexts to help programmers manage multiple dynamic
oids in convenient way.

Please see the manpages for detailed discussion, and example module
for typical use.

This work is based on ideas and code snippets coming from many
people, among them:  Arun Sharma, Jonathan Lemon, Doug Rabson,
Brian Feldman, Kelly Yancey, Poul-Henning Kamp and others. I'd like
to specially thank Brian Feldman for detailed review and style
fixes.

PR:		kern/16928
Reviewed by:	dfr, green, phk
2000-07-15 10:26:04 +00:00
Alfred Perlstein
af0e6bcdf0 Make mbstat.m_mtypes seperate and viewable via sysctl, also
expand the size from short to ulong

Submitted by: Ian Dowse <iedowse@maths.tcd.ie>
PR: kern/19809
2000-07-15 06:02:48 +00:00
Paul Saab
88f675ba30 Change the way NMI's are handled. Before, if DDB was enabled and
a NMI occured, you could type continue in DDB and the kernel would
not attempt to detect what type of NMI was recieved.  Now we check
for the type of NMI first and then go to DDB if it is enabled.

This will solve the problem with having DDB enabled and getting an
NMI due to some possibly bad error and being able to continue the
operation of the kernel when you really want to panic and know
what happened.

Submitted by:	jhb
2000-07-14 11:49:44 +00:00
Robert Watson
e8483a05a6 o Commit two of two, introducing __cap_{get,set}_{fd,file} syscalls to
modify capability sets on files.

Obtained from:	TrustedBSD Project
2000-07-13 20:38:52 +00:00
Robert Watson
92eebb8a9b o Introduce syscall prototypes, stubs for __cap_{get,set}_{fd,file},
syscalls to manage capability sets on files.  First of two commits.

Obtained from:	TrustedBSD Project
2000-07-13 20:31:24 +00:00
John Baldwin
9c386f6b7d For infinite timeouts, set both the tv_sec and tv_usec fields to zero in
poll() and select().

Noticed by:	Wesley Morgan <morganw@chemicals.tacorp.com>
2000-07-13 02:12:25 +00:00
John Baldwin
4da144c091 Fix a very obscure bug in select() and poll() where the timeout would
never expire if poll() or select() was called before the system had been
in multiuser for 1 second.  This was caused by only checking to see if
tv_sec was zero rather than checking both tv_sec and tv_usec.
2000-07-12 22:46:40 +00:00
Jun-ichiro itojun Hagino
f38211642f remove m_pulldown statistics, which is highly experimental and does not
belong to *bsd-merged tree
2000-07-12 16:39:13 +00:00
Kirk McKusick
f2a2857bb3 Add snapshots to the fast filesystem. Most of the changes support
the gating of system calls that cause modifications to the underlying
filesystem. The gating can be enabled by any filesystem that needs
to consistently suspend operations by adding the vop_stdgetwritemount
to their set of vnops. Once gating is enabled, the function
vfs_write_suspend stops all new write operations to a filesystem,
allows any filesystem modifying system calls already in progress
to complete, then sync's the filesystem to disk and returns. The
function vfs_write_resume allows the suspended write operations to
begin again. Gating is not added by default for all filesystems as
for SMP systems it adds two extra locks to such critical kernel
paths as the write system call. Thus, gating should only be added
as needed.

Details on the use and current status of snapshots in FFS can be
found in /sys/ufs/ffs/README.snapshot so for brevity and timelyness
is not included here. Unless and until you create a snapshot file,
these changes should have no effect on your system (famous last words).
2000-07-11 22:07:57 +00:00
Boris Popov
2ff087318a Correct SYSINIT execution order in the case when KLD contains more
than one SYSINIT with the same 'subsystem' id and different 'order' id.

Reviewed by:	peter
2000-07-09 23:58:56 +00:00