Create a new thread state to describe threads that would be ready to run
except for the fact tha they are presently swapped out. Also add a process
flag to indicate that the process has started the struggle to swap
back in. This will be needed for the case where multiple threads
start the swapin action top a collision. Also add code to stop
a process fropm being swapped out if one of the threads in this
process is actually off running on another CPU.. that might hurt...
Mike Barcroft [Mon, 29 Jul 2002 17:41:23 +0000 (17:41 +0000)]
Create a new header <machine/_stdint.h> for storing MD parts of
<stdint.h>. Previously, parts were defined in <machine/ansi.h> and
<machine/limits.h>. This resulted in two problems:
(1) Defining macros in <machine/ansi.h> gets in the way of that
header only defining types.
(2) Defining C99 limits in <machine/limits.h> adds pollution to
<limits.h>.
Warner Losh [Mon, 29 Jul 2002 15:36:22 +0000 (15:36 +0000)]
If we are in hostap mode, do not go into promisc mode. This causes
problems with the firmware and will result in a) poor performance and
b) the inability to associate certain types of cards (most notibly
cisco).
Idea obtained from OpenBSD, but I implemented it by clearing the
IFF_PROMISC flag rather than the refusing to honor it downstream.
Tim J. Robbins [Mon, 29 Jul 2002 14:50:54 +0000 (14:50 +0000)]
When translating and the -c option is specified, handle the case where the
second string argument is more than one character in length in the way
required by SUSv3 (and the way GNU textutils and SVR4 do it).
Ruslan Ermilov [Mon, 29 Jul 2002 08:51:04 +0000 (08:51 +0000)]
Ignore -C, -p, and -S options of install(1) when used with the -d
option. Warn about COPY being phased out. Restore the old method
of always comparing before installing: INSTALL="install -C".
Jeff Roberson [Mon, 29 Jul 2002 06:26:55 +0000 (06:26 +0000)]
- Backout the patch made in revision 1.75 of vfs_mount.c. The vputs here
were hiding the real problem of the missing unlock in sync_inactive.
- Add the missing unlock in sync_inactive.
Alan Cox [Mon, 29 Jul 2002 05:42:44 +0000 (05:42 +0000)]
o Pass VM_ALLOC_WIRED to vm_page_grab() rather than calling vm_page_wire()
in pmap_new_thread(), pmap_pinit(), and vm_proc_new().
o Lock page queue accesses by vm_page_free() in pmap_object_init_pt().
Mike Silbersack [Mon, 29 Jul 2002 04:32:35 +0000 (04:32 +0000)]
Revert rev 1.131.
1.131 is slightly broken, and I would commit the fix to that here, but it
has been reported that any deviation from the original code is causing
problems with some 82557 chips, causing them to lock hard.
Until those issues have been figured out, going back to the original
code is the best plan.
Luigi Rizzo [Mon, 29 Jul 2002 02:48:09 +0000 (02:48 +0000)]
whitespace commit: fix indentation in fxp_intr_body.
This is also done in preparation of a subsequent fix
for the handling of RNR conditions in polling mode.
Don Lewis [Sun, 28 Jul 2002 21:06:14 +0000 (21:06 +0000)]
Make a temporary copy of the output data in the generic sysctl handlers
so that the data is less likely to be inconsistent if SYSCTL_OUT() blocks.
If the data is large, wire the output buffer instead.
This is somewhat less than optimal, since the handler could skip the copy
if it knew that the data was static.
If the data is dynamic, we are still not guaranteed to get a consistent
copy since another processor could change the data while the copy is in
progress because the data is not locked. This problem could be solved if
the generic handlers had the ability to grab the proper lock before the
copy and release it afterwards.
This may duplicate work done in other sysctl handlers in the kernel which
also copy the data, possibly while a lock is held, before calling they call
a generic handler to output the data. These handlers should probably call
SYSCTL_OUT() directly.
Don Lewis [Sun, 28 Jul 2002 19:59:31 +0000 (19:59 +0000)]
Wire the sysctl output buffer before grabbing any locks to prevent
SYSCTL_OUT() from blocking while locks are held. This should
only be done when it would be inconvenient to make a temporary copy of
the data and defer calling SYSCTL_OUT() until after the locks are
released.
Fix a bug introduced in previous commit. Due to the interaction of the
direct physical mappings with virtual page colour, we need to flush the
data cache when a page changes colour. I missed one case which broke
pipes.
David Malone [Sun, 28 Jul 2002 16:05:07 +0000 (16:05 +0000)]
Improve WARNS situation for kdump:
1) Define _KERNEL while including sys/time.h to get some function prototypes.
2) Add prototypes and ANSIify definitions.
3) Constness changes.
4) Remove register keyword.
5) Actually return a sensible value from main.
6) Make fread_tail take a void * instead of a char *.
7) Avoid a signedness warning by casting to a size_t. Should be safe
enough 'cos we also check for nonnegativity.
8) Be extra chummy with sigset_t rather than passing a struct to printf
and pretending it is an int.
Mike Barcroft [Sun, 28 Jul 2002 15:59:51 +0000 (15:59 +0000)]
Revert the previous delta; uintfptr_t needs to be available to
userland for libc/gmon to compile, so the typedef in <machine/types.h>
isn't good enough. This is really ugly since we end up with the
actual value which uintfptr_t is typedef'd from, in multiple places.
This is bug for bug compatible with the other FreeBSD architectures.
Juli Mallett [Sun, 28 Jul 2002 13:46:09 +0000 (13:46 +0000)]
Search the include path for the argument to -f, to make lives easier for
those of us who want to figure out how old Jim Mock is, but only want to
type 'calendar -f calendar.freebsd'. This is done in a way that should
be totally backwards compatible with no noticable differences, at all.
Add a -v option that prevents switching virtual terminals while this
terminal is locked. This permits the user to easily lock the entire
console from a single terminal.
Implement this (quoted from the updated man page): If the first token
of a rule specification is a single dash (``-''), rules are read from
the standard input and the rest of the specification is ignored.
Use more standard section names in an attempt to unconfuse new users
(and make it easier to find stuff for the old ones). I originally put
everything under one "RULE SUBSYSTEM" so it didn't look like devfs(8)
is dedicated to the rule subsystem, but since nobody has any idea what
else might go here, it's not worth confusing people just to save a
little time for someone that might add something rule-unrelated to
devfs(8) in a few years.
Juli Mallett [Sun, 28 Jul 2002 03:57:08 +0000 (03:57 +0000)]
As of revision 1.38 of make/parse.c, our make(1) will warn too. Note that
this isn't just for the sake of testing behaviour, and that things really
do break if this regression occurs.
Unimplement panic(8) by making sure that we don't recurse into a
ruleset. If we do, that means there's a ruleset loop (10 includes 20
include 30 includes 10), which will quickly cause a double fault due
to stack overflow (since "include" is implemented by recursion).
(Previously, we only checked that X didn't include X.)
Juli Mallett [Sun, 28 Jul 2002 03:52:41 +0000 (03:52 +0000)]
Print a warning when we are given two scripts for one target. This is neither
as wide-reaching nor intensive as NetBSD's similar, but the warning uses the
same text.
Stock -current has more than 300 files in /etc, so 255 inodes for the
/etc filesystem isn't enough; consequently, add "-i 4096" to the newfs
command for /etc. This results in 1022 inodes, which should be enough
for the forseeable future (although I don't know why we would ever
have more than 1000 files in a default /etc).
Alan Cox [Sat, 27 Jul 2002 23:20:32 +0000 (23:20 +0000)]
o Require that the page queues lock is held on entry to vm_pageout_clean()
and vm_pageout_flush().
o Acquire the page queues lock before calling vm_pageout_clean()
or vm_pageout_flush().
David Malone [Sat, 27 Jul 2002 23:06:52 +0000 (23:06 +0000)]
If a socket is disconnected for some reason (like a TCP connection
not responding) then drop any data on the outgoing queue in
soisdisconnected because there is no way to get it to its destination
any longer.
The only objection to this patch I got on -net was from Terry, who
wasn't sure that the condition in question could arise, so I provided
some example code.
David Malone [Sat, 27 Jul 2002 22:53:44 +0000 (22:53 +0000)]
Make test check the tv_nsec part of a struct stat when comparing
the mtimes of a file. (This is probably only useful if you have
vfs.timestamp_precision set to something nonzero).
John Baldwin [Sat, 27 Jul 2002 22:15:42 +0000 (22:15 +0000)]
- Fixup whitespace after previous commit.
- To minimize whitespace changes, remove a test that didn't define
_CPUCFLAGS if both NO_CPU_CFLAGS and NO_CPU_COPTFLAGS were defined
since it is redundant (we don't use _CPUCFLAGS if those are defined).
John Baldwin [Sat, 27 Jul 2002 22:04:05 +0000 (22:04 +0000)]
If there is not a CPUTYPE defined by default, then allow for _CPUCFLAGS
to tune for more advanced processors while still supporting the minimum
processor in an architecture. We can do this with the '-mtune=' option
to gcc for alpha, sparc64, and powerpc and with the mis-named '-mcpu='
option for i386.
This defaults to tuning i386 builds for i686 machines though not using
any instructions that aren't found on an 80386. For alpha it defaults
to tuning for an EV5.
Approved by: peter
Peril sensitive sunglasses borrowed from: peter
Implement a direct mapped address region, like alpha and ia64. This
basically maps all of physical memory 1:1 to a range of virtual addresses
outside of normal kva. The advantage of doing this instead of accessing
phsyical addresses directly is that memory accesses will go through the
data cache, and will participate in the normal cache coherency algorithm
for invalidating lines in our own and in other cpus' data caches. So
we don't have to flush the cache manually or send IPIs to do so on other
cpus. Also, since the mappings never change, we don't have to flush them
from the tlb manually.
This makes pmap_copy_page and pmap_zero_page MP safe, allowing the idle
zero proc to run outside of giant.
Robert Watson [Sat, 27 Jul 2002 19:56:27 +0000 (19:56 +0000)]
Reserve VCACHEDLABEL vnode flag for use by the TrustedBSD MAC
implementation. This flag will indicate that the security label
in the vnode is currently valid, and therefore doesn't need to
be refreshed before an access control decision can be made. Most
file systems (or stdvops) will set this flag after they load the
MAC label from disk the first time to prevent redundant disk I/O;
some file synthetic file systems (procfs, for example) may not.
Obtained from: TrustedBSD Project
Sponsored by: DARPA, NAI Labs
John Baldwin [Sat, 27 Jul 2002 19:43:52 +0000 (19:43 +0000)]
If we are building docs with this release and PREFETCHDISTFILES is defined
then download the distfiles for the ports needed to build the docs before
we enter the chroot environment. This is useful since often times releases
get in a funk trying to download distfiles in the chroot.