e1b6afe136
todo.
2006-11-04 13:25:52 +00:00
d4cb7c851a
Fix matching of options with common prefixes (e.g. console and console_speed).
...
While here remove an useless assignment and bind a variable to a loop.
2006-11-04 11:37:12 +00:00
9477ac30bc
Add "@uid" keyword translation, to translate effective user-id of the
...
process.
2006-11-04 10:14:00 +00:00
fe9e2303fd
Change KAUTH_SYSTEM_RAWIO to KAUTH_DEVICE_RAWIO_SPEC (moving the raw i/o
...
requests to the device scope) and add KAUTH_DEVICE_RAWIO_PASSTHRU.
Expose iskmemdev() through sys/conf.h.
okay yamt@
2006-11-04 09:30:00 +00:00
5bfcdd68c1
- ltsleep(): for now, stay at splsched() when releasing sched_lock, or we
...
may allow wakeup() to occur before switching away. PR/32962.
- mi_switch(): don't inspect p->p_cred or send signals without holding the
kernel lock.
2006-11-03 20:46:00 +00:00
1cffc415cb
- issignal(): acquire the kernel lock before inspecting signal state
...
- lwp_exit2(): don't drop the kernel lock until after doing wakeup()
2006-11-03 19:46:03 +00:00
0ae566badc
fix deadlocks due to ksiginfo_pool spl problems.
2006-11-03 12:18:41 +00:00
1cd3142c15
make siginfo_pool and ksiginfo_pool static.
2006-11-03 11:41:07 +00:00
4d91d6d252
ltsleep: fix a race with wakeup().
2006-11-02 16:26:25 +00:00
d6bdd9bb62
printf() -> log().
2006-11-01 22:45:14 +00:00
3f78162b5c
implement kern.arandom properly, instead of lying about it and only filling
...
the first 4 bytes of the array with random data.
2006-11-01 22:27:43 +00:00
7aa8c7d98c
Pointing one element past an array is fine, pointing before it isn't.
2006-11-01 11:37:59 +00:00
1a7bc55dcc
remove some __unused from function parameters.
2006-11-01 10:17:58 +00:00
caa14e011b
kill signal "dolock" hacks.
...
related to PR/32962 and PR/34895. reviewed by matthew green.
2006-11-01 09:46:14 +00:00
86a24f7e0c
- remove an unnecessary #include.
...
- make some variables static.
2006-11-01 09:36:28 +00:00
60392d9383
kill sched_psignal.
...
related to PR/32962 and PR/34895. reviewed by matthew green.
2006-11-01 09:33:45 +00:00
dd706203e4
mi_switch: move rlimit and autonice handling out of sched_lock in order to
...
simplify locking.
related to PR/32962 and PR/34895. reviewed by matthew green.
2006-11-01 09:32:52 +00:00
a2be0ed655
Revert the changes I introduced trying to solve tmpfs' NFS export problem.
...
Requested by yamt@
2006-10-31 08:12:46 +00:00
52d8744717
Use integers, not pointers to integers, for KAUTH_REQ_NETWORK_SOCKET_OPEN.
...
Reminded by yamt@, thanks!
2006-10-30 16:53:48 +00:00
ff7c31739a
Use consistent logging messages.
2006-10-30 11:29:12 +00:00
ce42f55037
add the emulation in kinfo_proc2
2006-10-29 22:34:07 +00:00
d165928f40
check for negative length in PT_COREDUMP (from Neil)
2006-10-28 17:58:22 +00:00
066e9189c4
Adapt to changes suggested by yamt@ to get rid of __UNCONST() stuff.
...
While here, don't leak pathbuf on success.
2006-10-28 11:43:45 +00:00
8f3077c34d
avoid allocating from ksiginfo_pool when we know we aren't going to
...
use it. idea partly from yamt. assert SCHED_ASSERT_UNLOCKED() in
all the places we call ksiginfo_queue() without a ksiginfo.
fixes recent panics detected by LOCKDEBUG.
2006-10-28 08:09:31 +00:00
145a8382e1
Don't leak handle.
2006-10-27 22:17:09 +00:00
31a62606ea
Merge kernel and userland rmd160 and sha2 implementation.
...
XXX: We still install rmd160.h and sha2.h in /usr/include/crypto, unlike
the other hash functions which get installed in /usr/include for compatibility.
2006-10-27 21:20:48 +00:00
52bab3d8ab
Don't allocate MAXPATHLEN on the stack.
...
Prompted by and initial diff okay yamt@
2006-10-27 20:16:10 +00:00
17685f3fdc
vmem: use correct function names for ASSERT_SLEEPABLE.
2006-10-27 15:05:16 +00:00
1342c5a275
Use prop_dictionary_set_int32() to set the error number.
2006-10-26 05:08:01 +00:00
adf8d7aab2
Introduce KAUTH_REQ_NETWORK_SOCKET_OPEN, to check if opening a socket is
...
allowed. It takes three int * arguments indicating domain, type, and
protocol. Replace previous KAUTH_REQ_NETWORK_SOCKET_RAWSOCK with it (but
keep it still).
Places that used to explicitly check for privileged context now don't
need it anymore, so I replaced these with XXX comment indiacting it for
future reference.
Documented and updated examples as well.
2006-10-25 22:49:22 +00:00
9e6aff1165
Added missing whitespace in sysctl description string. PR kern/34915.
2006-10-25 20:55:11 +00:00
5ce26ddf80
enable namecache enter reverse for linux procfs lookups.
...
XXX: how expensive is that?
2006-10-25 18:56:38 +00:00
1b7ad03044
Defer the vnode test for directory until after the cache lookup. This makes
...
getcwd work with plain files too, if we allow plain files in the vnode->name
cache.
2006-10-25 18:55:49 +00:00
133d7c5ba4
Add a function (ksyms_init_explicit) to add a symbol table when we know the
...
exact address of the symtab and strtab ELF sections contents beforehand.
This will be used in i386's Multiboot code to add the initial kernel ksyms
without having to prepare a minimal ELF image that matches the format
expected by the reguluar ksyms_init function.
This part was reviewed by cube@ and successfully passes a full i386 release
build.
2006-10-25 13:46:36 +00:00
3c6ae35ab2
- Add a new disk ioctl (DIOCGDISKINFO) to get the disk-info dictionary
...
for the disk.
- Add a new function, disk_ioctl(), that does generic disk ioctl handling.
DIOCGDISKINFO is handled here now, and others will be added in the future.
- In the wd driver, fill in the dk_info member of struct disk and use the
new disk_ioctl() function.
2006-10-25 04:04:45 +00:00
2d8eaf9b86
Various cleanups.
...
Make some "#ifdef DIAGNOSTIC" blocks always compiled in; others
convert to KASSERT() where appropriate.
Add some sanity checks and comments while here.
2006-10-24 22:38:41 +00:00
84bd46b9f9
Add support to allow a file system to not permit being exported over NFS.
...
Approved by elad@ and wrstuden@
2006-10-24 21:53:10 +00:00
8ea60a7f23
Remove specificdata_setspecific_nowait() as result of discussion on tech-kern.
...
Add _lwp_getspecific_by_lwp() to get lwp specific data from other lwp's.
Protected by #ifdef _LWP_API_PRIVATE.
Approved by: Jason Thorpe <thorpej@netbsd.org>
2006-10-24 10:05:45 +00:00
5239dc5a67
qc_init: fix a problem introduced by rev.1.13.
...
namely, use quantum-size alignment rather than no alignment.
PR/34879 from Peter Postma.
2006-10-23 13:36:33 +00:00
c2da5d160e
PR/34873: Ryo Shimizu: sendmsg() can cause kernel panic
...
Patch applied, thanks!
2006-10-23 10:37:22 +00:00
a37172bb09
avoid a problem in kpsignal2() holding sched_lock while wanting to call
...
pool_get(). pre-allocate the ksiginfo_t before taking sched_lock and
use it if necessary, or freeing it if it is unused.
rename ksiginfo_{get,put}() to ksiginfo_{dequeue,queue}() (idea from chuq.)
this fixes PR#32962. thanks to christos, chuq and go for help/ideas.
2006-10-22 20:48:45 +00:00
8c71201d08
make ktruser enforce the maximum buffer length, and return an error.
2006-10-22 18:19:49 +00:00
96d8335980
kauth_cred_uucvt() -> kauth_uucred_to_cred(), introduce kauth_cred_to_uucred()
...
per tech-kern proposal
2006-10-22 13:07:15 +00:00
54c1c4a4dc
Remove todo that has been documented for a long time now.
2006-10-22 10:37:19 +00:00
982b4f66e8
remove a todo which has been done.
2006-10-22 10:19:25 +00:00
0d14f90ef6
qc_init: when initializing pools, use align==1 (ie. no alignment)
...
rather than align==0, which is converted to ALIGN(1) by pool_init.
2006-10-22 09:42:30 +00:00
7a4c53719e
avoid yet another GCC warning only exposed at -O3.
2006-10-22 00:48:14 +00:00
e2d2f0dc15
pool_get(): use PR_WAITOK, not M_WAITOK (which happens to be PR_NOWAIT ..)
...
XXX: not the first time this mistake has been made
2006-10-21 17:01:56 +00:00
d094ff8f40
Avoid printf() the selected timecounter as it is called in the path of hardclock().
...
The information is available via sysctl anyway.
2006-10-21 07:16:54 +00:00
a5d447aee5
Add an XXX to remind me why it's there when grepping. (securelevel ref)
2006-10-20 22:22:48 +00:00
yamt