aboutsummaryrefslogtreecommitdiff
path: root/sys/kern
Commit message (Expand)AuthorAgeFilesLines
* Don't re-lock the sched lock if we didn't unlock it.Julian Elischer2002-08-231-1/+1
* - Fix a mistake in my last few commits. The PDROP flag stops msleep fromJeff Roberson2002-08-232-6/+2
* s/sus/sys/ in the a.out kernel case.Peter Wemm2002-08-222-2/+2
* slight cleanup of single-threading code for KSE processesJulian Elischer2002-08-224-6/+53
* Replace (ab)uses of "NULL" where "0" is really meant.Archie Cobbs2002-08-221-1/+1
* Instead of grabbing the userland a.out.h/link.h (or worse, fromPeter Wemm2002-08-221-3/+5
* Instead of nlist.h and link.h, use sys/nlist_aout.h and sys/link_elf.hPeter Wemm2002-08-222-4/+4
* Spell proprly properly:Robert Watson2002-08-221-1/+1
* Include <sys/systm.h> for the declarations of many things instead ofBruce Evans2002-08-221-0/+1
* o Remove the AIOCBLIST_ASYNCFREE flag and related code. It's never set.Alan Cox2002-08-221-19/+2
* - Closer inspection revealed a possible deadlock situation in vn_lock() thatJeff Roberson2002-08-221-16/+13
* - Make vn_lock() vget() and VOP_LOCK() all behave the same way WRTJeff Roberson2002-08-222-18/+9
* - Return two shared locks to exclusive locks. This was premature.Jeff Roberson2002-08-221-2/+11
* - Fix interlock handling in vn_lock(). Previously, vn_lock() could returnJeff Roberson2002-08-221-10/+12
* - Fix interlock handling in vn_lock(). Previously, vn_lock() could returnJeff Roberson2002-08-221-17/+9
* Don't use "NULL" when "0" is really meant.Archie Cobbs2002-08-212-3/+3
* Revert some suspension/sleep/signal code from KSE-IIIJulian Elischer2002-08-213-43/+49
* - Document two cases, one in vget and the other in vn_lock, where the stateJeff Roberson2002-08-212-0/+2
* - If vn_lock fails with the LK_INTERLOCK flag set, interlock will not beJeff Roberson2002-08-211-2/+3
* - Add two new debugging macros: ASSERT_VI_LOCKED and ASSERT_VI_UNLOCKEDJeff Roberson2002-08-212-8/+65
* - Hold the vnode lock across unlink() so that the v_vflag check is safe.Jeff Roberson2002-08-212-30/+40
* Close a race in process label changing opened due to dropping theRobert Watson2002-08-191-7/+10
* Regen.Robert Watson2002-08-192-3/+3
* mac_syscall is now implemented, switch to MSTD.Robert Watson2002-08-191-1/+1
* Pass active_cred and file_cred into the MAC framework explicitlyRobert Watson2002-08-196-32/+57
* Provide an implementation of mac_syscall() so that security modulesRobert Watson2002-08-191-0/+38
* Break out mac_check_pipe_op() into component check entry points:Robert Watson2002-08-192-8/+50
* Break out mac_check_vnode_op() into three seperate checks:Robert Watson2002-08-196-37/+74
* Assert process locks in proces-related access control checks.Robert Watson2002-08-191-0/+6
* Add a missing vnode assertion for the exec() check.Robert Watson2002-08-191-0/+2
* Keep a copy of the credential used to mount filesystems around soPoul-Henning Kamp2002-08-191-3/+7
* A side effect of some debugging: prototypify and deregister.Poul-Henning Kamp2002-08-181-45/+18
* Increase size of ifnet.if_flags from 16 bits (short) to 32 bits (int). To avoidMaxim Sobolev2002-08-181-5/+5
* In continuation of early fileop credential changes, modify fo_ioctl() toRobert Watson2002-08-176-20/+26
* Further improved the performance of sbreserve() by moving the calculationDavid Greenman2002-08-162-10/+52
* Wrap maintenance of varios nmac{objectname} counters in MAC_DEBUG so weRobert Watson2002-08-161-0/+46
* Correct white space nits that crept in during my recent merges ofRobert Watson2002-08-162-3/+3
* Make similar changes to fo_stat() and fo_poll() as made earlier toRobert Watson2002-08-169-56/+71
* Rewrote the space check algorithm in sbreserve() so that the extremelyDavid Greenman2002-08-162-2/+2
* In order to better support flexible and extensible access control,Robert Watson2002-08-1514-62/+91
* Rename mac_check_socket_receive() to mac_check_socket_deliver() so thatRobert Watson2002-08-151-10/+10
* Fix return case for negative namelen by jumping to normal exit processingRobert Watson2002-08-151-2/+4
* Make m_flags an int instead of a short, this is consistent with theBosko Milekic2002-08-151-1/+1
* Sync to trustedbsd_mac tree: default to sigsegv rather than copy-on-writeRobert Watson2002-08-151-1/+1
* return foo -> return (foo)Alfred Perlstein2002-08-151-22/+22
* Moved sf_buf_alloc and sf_buf_free function declarations to sys/socketvar.hDavid Greenman2002-08-132-4/+0
* Remove obsolete comment about sf_buf_* functions being static. They wereDavid Greenman2002-08-131-3/+0
* Remember to unlock the (optional) vnode in vfs_stdextattrctl(). FailingPoul-Henning Kamp2002-08-131-0/+2
* Add a #include for <sys/mount.h>Poul-Henning Kamp2002-08-131-0/+1
* Make SYSVSEM mpsafe. Each semaphore set gets its own lock, howeverAlfred Perlstein2002-08-131-87/+172