Commit graph

2072 commits

Author SHA1 Message Date
Eivind Eklund
2ae353f9a7 Rename one of the two devfs_link's to devfs_makelink. 1998-12-10 19:57:01 +00:00
Jordan K. Hubbard
337c96916f poll(2) sets POLLNVAL for descriptors passed in that are less than
0.  This makes it difficult to do efficient manipulation of the
struct pollfd since you can't leave a slot empty.

PR:		8599
Submitted-by:	Marc Slemko <marcs@znep.com>
1998-12-10 01:53:26 +00:00
Robert V. Baron
efb73e5aca In ktrwrite, use uio_procp = curproc vs 0 1998-12-10 01:47:41 +00:00
Eivind Eklund
d51523c6c9 Get rid of CTLTYPE_OPAQUE in a SYSCTL_OPAQUE - it is added my the
SYSCTL_OPAQUE macro.
1998-12-09 02:26:45 +00:00
Bruce Evans
e1501bb6eb Backed out the FIOASYNC fix in rev.1.108. fcntl(fd, F_SETFL, flags)
depends on the bug.  It does an FIOASYNC ioctl to sync the setting
of the O_ASYNC "file" flag with drivers even if the setting hasn't
changed.

PR:		9003
1998-12-08 10:22:07 +00:00
Archie Cobbs
f1d19042b0 The "easy" fixes for compiling the kernel -Wunused: remove unreferenced static
and local variables, goto labels, and functions declared but not defined.
1998-12-07 21:58:50 +00:00
Eivind Eklund
4979978b8d Fix grouping of statements. This remove a potential panic in the soft
updates code.  While I'm here, remove an unintended trigraph.

Reviewed by:	Kirk McKusick <kirk@freebsd.org>
1998-12-07 17:23:45 +00:00
Andrey A. Chernov
8e3c23be56 Move stime declaration to main block, otherwise can left uninitialized
in rare cases.
Found by: Eivind Eklund <eivind@yes.no>
1998-12-07 07:59:20 +00:00
Archie Cobbs
2326715f79 Avoid compiler warning (printf arg type mismatch) when compiling #ifdef DEBUG 1998-12-06 00:03:30 +00:00
Kirk McKusick
fcdfed00d5 Even the most recently allocated buffer may not have its b_blkno
field properly filled in, so we must do a VOP_BMAP on that buffer
as well if it is not resolved.
Submitted by: Luoqi Chen <luoqi@watermarkgroup.com>
1998-12-05 06:12:14 +00:00
Archie Cobbs
2127f26023 Examine all occurrences of sprintf(), strcat(), and str[n]cpy()
for possible buffer overflow problems. Replaced most sprintf()'s
with snprintf(); for others cases, added terminating NUL bytes where
appropriate, replaced constants like "16" with sizeof(), etc.

These changes include several bug fixes, but most changes are for
maintainability's sake. Any instance where it wasn't "immediately
obvious" that a buffer overflow could not occur was made safer.

Reviewed by:	Bruce Evans <bde@zeta.org.au>
Reviewed by:	Matthew Dillon <dillon@apollo.backplane.com>
Reviewed by:	Mike Spengler <mks@networkcs.com>
1998-12-04 22:54:57 +00:00
David Greenman
911e8dbc2a Fixed broken code in sendfile(2) when using file offsets. 1998-12-03 12:35:47 +00:00
Archie Cobbs
8245f3f5b4 Add snprintf(3) and vsnprintf(3) capability to the kernel.
Reviewed by:	bde
1998-12-03 04:45:57 +00:00
KATO Takenori
9ad861edee - For some old Cyrix CPUs, %cr2 is clobbered by interrupts. This
problem is worked around by using an interrupt gate for the page
   fault handler.  This code was originally made for NetBSD/pc98 by
   Naofumi Honda <honda@kururu.math.sci.hokudai.ac.jp> and has already
   been in PC98 tree.  Because of this bug, trap_fatal cannot show
   correct page fault address if %cr2 is obtained in this function.
   Therefore, trap_fatal uses the value from trap() function.
-  The trap handler always enables interruption when buggy application
   or kernel code has disabled interrupts and then trapped.  This code
   was prepared by Bruce Evans <bde@FreeBSD.org>.

Submitted by:	Bruce Evans <bde@FreeBSD.org>
		Naofumi Honda <honda@kururu.math.sci.hokudai.ac.jp>
1998-12-02 08:15:17 +00:00
Eivind Eklund
0bfe299097 Check return value of malloc() in expand_name.
Reviewed by:	sef
1998-12-02 01:53:48 +00:00
Poul-Henning Kamp
510eb5b9db Make the previous behaviour the default, add a sysctl which you
can set if your hw/sw produces the "calcru negative..." message.

Setting the alternate method (sysctl -w kern.timecounter.method=1)
makes the the get{nano|micro}*() functions call the real thing at
resulting in a measurable but minor overhead.

I decided to NOT have the "calcru" change the method automatically
because you should be aware of this problem if you have it.

The problems currently seen, related to usleep and a few other corners
are fixed for both methods.
1998-11-29 20:31:02 +00:00
David Greenman
f2b678d4cb Compare p_cpulimit with RLIM_INFINITY before comparing it with the process
runtime. p_runtime is unsigned while p_cpulimit is not, so this avoids the
nasty side effect of the process getting killed when the runtime comes up
"negative" due to other bugs.
1998-11-27 11:44:22 +00:00
Tor Egge
18830dba83 Don't forget to update the pmap associated with aio daemons when adding
new page directory entries for a growing kernel virtual address space.
1998-11-27 01:14:21 +00:00
Tor Egge
f0ac792bec Attempt to handle interrupts delivered to all IO APICs by using the first
IO APIC with a sufficient number of pins.
1998-11-26 23:14:23 +00:00
Eivind Eklund
9fcdafaefc Staticize. 1998-11-26 18:50:24 +00:00
Bruce Evans
e7414d1bfa Fixed the previous fix - stathz doesn't give the statclock frequency
when it is 0.

Submitted by:	mostly by Hidetoshi Shimokawa <simokawa@sat.t.u-tokyo.ac.jp>
1998-11-26 16:49:55 +00:00
Bruce Evans
44f05f9c92 Oops, yet again back out some local changes that shouldn't have been
in the previous commit.
1998-11-26 14:05:58 +00:00
Bruce Evans
e4d0e37b33 Fixed scaling of p_pctcpu. It was wrong by a factor of stathz/hz.
Until recently, this was half compensated for in at least ps and top
by multiplying by 100/stathz to get a better wrong factor of 100/hz.
1998-11-26 14:00:08 +00:00
Poul-Henning Kamp
c2906d55d0 Make timecounters more resistant to badly behaved SW/HW which locks
out interrupts for too long.  If you still see the "calcru: negative
time..." message you can increase NTIMECOUNTER (see LINT).

Sideeffect is that a timecounter is required to not wrap around in
less than (1 + delta) seconds instead of the (1/hz + delta) required
until now.

Many thanks to: msmith, wpaul, wosch & bde
1998-11-23 09:59:02 +00:00
Søren Schmidt
8843cc3508 Add a kludge to prevent panicing when using VM86 and hitting here
with a NULL curproc.

Originally by: Tor Egge (IIRC)
1998-11-23 09:34:19 +00:00
Bruce Evans
f6233520ce Fixed a missing include. `SYSININT(...);' garbage in gave null garbage out. 1998-11-23 09:33:35 +00:00
Don Lewis
9d2b090975 We can't call fsetown() from sonewconn() because sonewconn() is be called
from an interrupt context and fsetown() wants to peek at curproc, call
malloc(..., M_WAITOK), and fiddle with various unprotected data structures.
The fix is to move the code that duplicates the F_SETOWN/FIOSETOWN state
of the original socket to the new socket from sonewconn() to accept1(),
since accept1() runs in the correct context.  Deferring this until the
process calls accept() is harmless since the process can't do anything
useful with SIGIO on the new socket until it has the descriptor for that
socket.

One could make the case for not bothering to duplicate the
F_SETOWN/FIOSETOWN state and requiring the process to explicitly make the
fcntl() or ioctl() call on the new socket, but this would be incompatible
with the previous implementation and might break programs which rely on
the old semantics.

This bug was discovered by Andrew Gallatin <gallatin@cs.duke.edu>.
1998-11-23 00:45:39 +00:00
Bruce Evans
7dd89537f1 Fixed some missing cases in the check for ioctls that involve modification.
Many (mostly machine-dependent ones) are still missing.  NIST-PCTS found
this bug for all the ioctls used to implement the POSIX tc* functions
(TIOCCBRK, TIOCDRAIN, TIOCSPGRP, TIOCSBRK, TIOCSTART and TIOCSTOP), and
I found FIOASYNC, TIOCCONS, TIOCEXCL, TIOCHPCL, TIOCNXCL, TIOCSCTTY and
TIOCSDRAINWAIT by inspection.  TIOCSPGRP was ifdefed out for some reason.

Handle tcsetattr()'s historical speed conversions correctly and more
centrally:
- don't store speeds of 0 in the final termios struct.  Drivers can now
  depend on tp->t_ispeed and tp->t_ospeed giving the actual speed.
  Applications can now depend on tcgetattr() being POSIX.1 conformant.
- convert from a proposed input speed of 0 to the proposed output speed
  (except if that is 0, convert to the current output speed).  Drivers
  can now depend on the proposed input speed being nonzero.
- don't reject negative speeds.  Negative speeds can't happen now that
  speed_t is unsigned, and rejecting invalid speeds is a bug - tcsetattr()
  is supposed to succeed if it can "perform any of the requested actions",
  so it shouldn't fail in practice.
1998-11-22 09:04:09 +00:00
David Greenman
4f699173cb Closed a very narrow and rare race condition that involved net interrupts,
bio interrupts, and a truncated file that along with the precise alignment
of the planets could result in a page being freed multiple times or a
just-freed page being put onto the inactive queue.
1998-11-18 09:00:47 +00:00
Kirk McKusick
96334b8fc8 Because buffers may be tossed and recreated at will under the new VM
system, the mapping from logical to physical block number may be lost.
Hence we have to check for a reconstituted buffer and redo the call to
VOP_BMAP if the physical block number has been lost.
1998-11-17 00:31:12 +00:00
Kenneth D. Merry
d9e371b995 Now that the wd driver is fixed (Thanks Bruce!), re-enable the
devstat_end_transaction error message that gets printed whenever the
busy count is < 0.

This will help catch drivers that improperly implement devstat(9) support.
1998-11-15 23:57:22 +00:00
Doug Rabson
921433915d A couple of minor sanity checks.
Submitted by: Nick Hibma <nick.hibma@jrc.it>
1998-11-15 18:11:21 +00:00
David Greenman
efac52b4ab In sendfile(2), check against sb_lowat when filling the socket buffer,
rather than 0.
1998-11-15 16:55:09 +00:00
Bruce Evans
4ae860afde Fixed the type of vfs_modevent (a little late, to match the corresponding
change in sys/mount.h).
1998-11-15 15:18:30 +00:00
Bruce Evans
14fddaf97a Fixed a missing include. <sys/kernel.h> is needed by the new
MALLOC_DEFINE() and MALLOC_DEFINE() is needed by the recently
reenabled "reallocblks" code, but <sys/kernel.h> was only included
if CLUSTERDEBUG was defined.  This was too harmless.  gcc only
warns about garbage like `SYSINIT(blech);' at file scope ...
1998-11-15 14:11:06 +00:00
David Greenman
f2efb8e4c8 Fixed a couple of nits in sendfile(2): clear PG_ZERO before unbusying
the page, and use passed-in "p" rather than curproc in uio struct.
1998-11-14 23:36:17 +00:00
Garrett Wollman
14177d72cd My changes to the new device interface:
- Interface wth the new resource manager.
	- Allow for multiple drivers implementing a single devclass.
	- Remove ordering dependencies between header files.
	- Style cleanup.
	- Add DEVICE_SUSPEND and DEVICE_RESUME methods.
	- Move to a single-phase interrupt setup scheme.

Kernel builds on the Alpha are brken until Doug gets a chance to incorporate
these changes on that side.

Agreed to in principle by: dfr
1998-11-14 21:58:51 +00:00
Mike Smith
d02d6d040a Don't count non-local dirty buffers as outstanding when shutting down.
This avoids the fsck-on-reboot symptoms if you're shutting down with a
hung or unreachable NFS server mounted.  Also remove non-local
filesystems from the mount list to prevent the system hanging when it tries
to unmount them (for the same reason).

Drew points out that there's a good argument for forcibly removing all
"non syncable" filesystems from the mount list (eg. NFS mounts, disks
that aren't responding, etc.) as this then allows you to sync and
cleanly unmount their parents.  No such change is included in this
patch.

Submitted by:	Andrew Gallatin <gallatin@cs.duke.edu>
1998-11-13 22:40:37 +00:00
Doug Rabson
92c6ae04da A couple of fixes to device_delete_child() to delete all children of
the device and to free the device which is being deleted (not its
parent - oops).

Submitted by: Nick Hibma <nick.hibma@jrc.it>
1998-11-13 09:39:37 +00:00
David Greenman
1c680b45a2 Restored the "reallocblks" code to its former glory. What this does is
basically do a on-the-fly defragmentation of the FFS filesystem, changing
file block allocations to make them contiguous. Thanks to Kirk McKusick
for providing hints on what needed to be done to get this working.
1998-11-13 01:01:44 +00:00
Peter Wemm
ba031106b1 kldsym(2) prototype implementation 1998-11-11 13:04:40 +00:00
Peter Wemm
de8a10f07a regenerate (+kldsym) 1998-11-11 12:57:05 +00:00
Peter Wemm
325e13dd19 A kldsym(2) syscall prototype for extracting information from the in-kernel
linker.  This is intended to replace kvm_mkdb etc.  The first version
only does name->value lookups, but it's open ended.  value->name lookups
would probably be a good thing to do too.

It's been suggested to try and connect the symbol tables to sysctl (which
is probably a more flexible way of doing it if it's done right), but that
is far more complex and difficult than I was ready to have a shot at.
1998-11-11 12:45:14 +00:00
Don Lewis
62d6ce3af2 I got another batch of suggestions for cosmetic changes from bde. 1998-11-11 10:56:07 +00:00
Don Lewis
831d27a9f5 Installed the second patch attached to kern/7899 with some changes suggested
by bde, a few other tweaks to get the patch to apply cleanly again and
some improvements to the comments.

This change closes some fairly minor security holes associated with
F_SETOWN, fixes a few bugs, and removes some limitations that F_SETOWN
had on tty devices.  For more details, see the description on the PR.

Because this patch increases the size of the proc and pgrp structures,
it is necessary to re-install the includes and recompile libkvm,
the vinum lkm, fstat, gcore, gdb, ipfilter, ps, top, and w.

PR:		kern/7899
Reviewed by:	bde, elvind
1998-11-11 10:04:13 +00:00
Peter Wemm
1c5bb3eaa1 add #include <sys/kernel.h> where it's needed by MALLOC_DEFINE() 1998-11-10 09:16:29 +00:00
Peter Wemm
c6105dce78 Register and unregister vnodeops the new way rather than leaking on
unload.
1998-11-10 09:12:40 +00:00
Peter Wemm
5ad3d2212f kzipboot uses kern/inflate.c outside the kernel by providing its own
minimal malloc/free implementation.  Stop passing M_GZIP to it.
1998-11-10 09:08:49 +00:00
Peter Wemm
4e61198e8f Make the vnode opv vector construction fully dynamic. Previously we
leaked memory on each unload and were limited to items referenced in
the kernel copy of vnode_if.c.  Now a kernel module is free to create
it's own VOP_FOO() routines and the rest of the system will happily
deal with it, including passthrough layers like union/umap/etc.

Have VFS_SET() call a common vfs_modevent() handler rather than
inline duplicating the common code all over the place.

Have VNODEOP_SET() have the vnodeops removed at unload time (assuming a
module) so that the vop_t ** vector is reclaimed.

Slightly adjust the vop_t ** vectors so that calling slot 0 is a panic
rather than a page fault.  This could happen if VOP_something() was called
without *any* handlers being present anywhere (including in vfs_default.c).
slot 1 becomes the default vector for the vnodeop table.

TODO: reclaim zones on unload (eg: nfs code)
1998-11-10 09:04:09 +00:00
Peter Wemm
edfbe15080 Arrange for unload-time linker set hooks to be called. While cut/pasting
some code, I changed the original to be consistant with the rest of the
file rather than duplicating the problems.
1998-11-10 08:49:28 +00:00