‚²112859 141 248 223 226 147 222 220 135 139 181 437 320 554 258 1796 154 229 125 408 185 311 427 195 323 280 333 272 249 209 216 350 221 992 133 463 330 176 134 126 277 474 338 111 135 132 180 699 201 179 135 127 112 443 223 151 103 112 200 203 170 257 134 290 119 211 122 265 157 160 169 158 967 118 130 269 176 150 144 192 319 161 157 150 125 132 163 120 185 345 293 280 275 161 154 136 120 136 254 185 463 428 190 303 359 317 116 110 408 1036 232 142 164 159 186 168 179 147 152 145 164 316 149 111 256 633 152 147 181 214 539 415 154 197 371 160 1876 443 387 280 180 574 620 K 10 svn:author V 3 alc K 8 svn:date V 27 2003-03-31T06:25:42.000000Z K 7 svn:log V 155 Recent changes to uipc_cow.c have eliminated the need for some sf_buf- related variables to be global. Make them either local to sf_buf_init() or static. END K 10 svn:author V 5 chris K 8 svn:date V 27 2003-03-31T08:08:59.000000Z K 7 svn:log V 128 Document the new mac_portacl(4) policy. Sponsored by: DARPA, Network Associates Laboratories Obtained from: TrustedBSD Project END K 10 svn:author V 5 chris K 8 svn:date V 27 2003-03-31T08:10:10.000000Z K 7 svn:log V 131 Document the new mac_portacl(4) policy. Sponsored by: DARPA, Network Associates Laboratories Obtained from: TrustedBSD Project END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T11:16:53.000000Z K 7 svn:log V 56 Strip the .comment section out from the BOOTMFS kernel. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T11:46:49.000000Z K 7 svn:log V 130 Revert revision 1.639 -- the "nodev" mount(8) option restriction does not apply to ${CHROOTDIR} file system since revision 1.712. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T12:29:31.000000Z K 7 svn:log V 128 Unmount the file system and detach an underlying memory disk even if the script fails somewhere in the middle. Prodded by: phk END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T12:35:26.000000Z K 7 svn:log V 44 NODEVFS cleanup: don't bother with MAKEDEV. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T12:41:27.000000Z K 7 svn:log V 48 Trace command execution. Grammar and spelling. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T12:46:18.000000Z K 7 svn:log V 90 Mention that 20021024 entry doesn't affect disks formatted in dangerously-dedicated mode. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T12:58:33.000000Z K 7 svn:log V 345 Slightly improve buildworld times by excluding crunchide(1) and kgzip(8) from the list of cross-tools during the normal, non-"make release" buildworld. Also, don't gratuitously build them, btxld(8) and elf2aout(1) for native architecture builds, since they have no known boostrapping issues along the supported upgrade path. Prodded by: peter END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-03-31T13:10:51.000000Z K 7 svn:log V 228 Enable cpp(1) warnings in system headers. GCC is oriented on glibc which is externally maintained, so GCC ships with these warnings turned off by default. This is also consistent with the src/contrib/gcc/c-lex.c,v 1.2 change. END K 10 svn:author V 3 des K 8 svn:date V 27 2003-03-31T13:45:36.000000Z K 7 svn:log V 461 If an ssh1 client initiated challenge-response authentication but did not respond to challenge, and later successfully authenticated itself using another method, the kbdint context would never be released, leaving the PAM child process behind even after the connection ended. Fix this by automatically releasing the kbdint context if a packet of type SSH_CMSG_AUTH_TIS is follwed by anything but a packet of type SSH_CMSG_AUTH_TIS_RESPONSE. MFC after: 1 week END K 10 svn:author V 3 des K 8 svn:date V 27 2003-03-31T13:48:18.000000Z K 7 svn:log V 165 - when using a child process instead of a thread, change the child's name to reflect its role - try to handle expired passwords a little better MFC after: 1 week END K 10 svn:author V 3 njl K 8 svn:date V 27 2003-03-31T17:29:43.000000Z K 7 svn:log V 1702 Clean up locking and resource management for pci/if_* - Remove locking of the softc in the attach method, instead depending on bus_setup_intr being at the end of attach (delaying interrupt enable until after ether_ifattach is called) - Call *_detach directly in the error case of attach, depending on checking in detach to only free resources that were allocated. This puts all resource freeing in one place, avoiding thinkos that lead to memory leaks. - Add bus_child_present check to calls to *_stop in the detach method to be sure hw is present before touching its registers. - Remove bzero softc calls since device_t should do this for us. - dc: move interrupt allocation back where it was before. It was unnecessary to move it. This reverts part of 1.88 - rl: move irq allocation before ether_ifattach. Problems might have been caused by allocating the irq after enabling interrupts on the card. - rl: call rl_stop before ether_ifdetach - sf: call sf_stop before ether_ifdetach - sis: add missed free of sis_tag - sis: check errors from tag creation - sis: move dmamem_alloc and dmamap_load to happen at same time as tag creation - sk: remove duplicate initialization of sk_dev - ste: add missed bus_generic_detach - ti: call ti_stop before ether_ifdetach - ti: add missed error setting in ti_rdata alloc failure - vr: add missed error setting in I/O, memory mapping cases - xl: add missed error setting in I/O, memory mapping cases - xl: remove multi-level goto on attach failure - xl: move dmamem_alloc and dmamap_load to happen at same time as tag creation - Calls to free(9) are unconditional because it is valid to call free with a null pointer. Reviewed by: imp, mdodd END K 10 svn:author V 4 bmah K 8 svn:date V 27 2003-03-31T17:38:18.000000Z K 7 svn:log V 61 New release notes/errata: SA-03:07. Submitted by: gshapiro END K 10 svn:author V 4 bmah K 8 svn:date V 27 2003-03-31T18:12:56.000000Z K 7 svn:log V 135 Release errata can also contain information uncovered late in the release cycle (after the release notes are closed). Note this fact. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-03-31T18:21:52.000000Z K 7 svn:log V 32 Match "serial" as well as "se". END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-03-31T18:35:37.000000Z K 7 svn:log V 315 Remove some debugging in the new OAM[*] and add a debug flag for other parts of it. [*] I've been asked what "OAM" means: It's an acronym used in the telecom industry, "Operations And Maintenance", and there it covers anything from a single unlabeled led on the frontpanel the the full nightmare of CMIP for SS7. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-03-31T18:38:31.000000Z K 7 svn:log V 93 Use new GEOM OAM. Kernels have supported this for a number of days, so people should be OK. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2003-03-31T19:24:37.000000Z K 7 svn:log V 218 If we fail to find our PCI ID in attach (this should never happen), then just return ENXIO directly instead of calling tl_detach() since that would panic since the softc mutex isn't initialized until after this check. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-03-31T19:56:55.000000Z K 7 svn:log V 333 - Allow the physical memory size that will be actually used by the kernel to be overridden by setting hw.physmem. - Fix a vm_map_find arg, we don't want to find space. - Add tracing and statistics for off colored pages. - Detect "stupid" pmap_kenters (same virtual and physical as existing mapping), and do nothing in that case. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2003-03-31T20:22:00.000000Z K 7 svn:log V 102 Add missing ()'s so that these drivers all compile again. Noticed by: jake Tested on: i386 (compile) END K 10 svn:author V 3 wes K 8 svn:date V 27 2003-03-31T21:09:57.000000Z K 7 svn:log V 230 Add a facility allowing processes to inform the VM subsystem they are critical and should not be killed when pageout is looking for more memory pages in all the wrong places. Reviewed by: arch@ Sponsored by: St. Bernard Software END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T21:40:47.000000Z K 7 svn:log V 186 - Use sigexit() instead of twiddling the signal mask, catch, ignore, and action bits to allow SIGILL to work as expected. This brings this file in line with other architectures. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:02:38.000000Z K 7 svn:log V 239 - Change trapsignal() to accept a thread and not a proc. - Change all consumers to pass in a thread. Right now this does not cause any functional changes but it will be important later when signals can be delivered to specific threads. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:12:09.000000Z K 7 svn:log V 178 - Mark signals which may be delivered to any thread in the process with SA_PROC. Signals without this flag should be directed to a particular thread if this is possible. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:30:08.000000Z K 7 svn:log V 155 - Add a signal for thread synchronization. Add an XXX so that maybe someone more knowledgeable on standards defined namespaces may ifdef this out. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:41:32.000000Z K 7 svn:log V 115 - Fix two calls to trapsignal() that were still passing in 'struct proc'. These were missed in my last commit. END K 10 svn:author V 6 julian K 8 svn:date V 27 2003-03-31T22:41:47.000000Z K 7 svn:log V 120 Do NOT return from an non-interruptable cv_wait, falsely claiming to have timed out. I don't know what I was thinking.. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:49:17.000000Z K 7 svn:log V 256 - Move p->p_sigmask to td->td_sigmask. Signal masks will be per thread with a follow on commit to kern_sig.c - signotify() now operates on a thread since unmasked pending signals are stored in the thread. - PS_NEEDSIGCHK moves to TDF_NEEDSIGCHK. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:51:19.000000Z K 7 svn:log V 127 - Move the NEEDSIGCHK and OLDMASK flags from proc to thread. - Move the signal mask to the thread. - Adjust a few comments. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:57:01.000000Z K 7 svn:log V 898 - The siglist in the proc holds signals that were blocked by all threads when they were delivered. In signotify() check to see if we have unblocked any of those signals and post them to the thread. - Use td_sigmask instead of p_sigmask in all cases. - In sigpending return both signals pending on the thread and proc. - Define a function, sigtd(), that finds the appropriate thread to deliver the signal to if psignal() has been called instead of tdsignal(). - Define a function, tdsignal(), that delivers a signal to a specific thread or if that thread has the signal blocked it may deliver it to the process where it will wait for a thread to unblock it. - Since we are delivering signals to a specific thread we do not need to abort the sleep of all threads. - Rename the old tdsignal() to tdsigwakeup(). - Save and restore the old signal mask to and from the thread. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T22:57:55.000000Z K 7 svn:log V 40 - Catch up with kernel signal changes. END K 10 svn:author V 6 thomas K 8 svn:date V 27 2003-03-31T23:26:10.000000Z K 7 svn:log V 367 Revert change 1.201 (removing mapping of VAPPEND to VWRITE). Instead, use the generic vaccess() operation to determine whether an operation is permitted. This avoids embedding knowledge on vnode permission bits such as VAPPEND in the NFS client. PR: kern/46515 vaccess() patch submitted by: "Peter Edwards" Approved by: tjr, roberto (mentor) END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T23:30:41.000000Z K 7 svn:log V 236 - Define sigwait, sigtimedwait, and sigwaitinfo in terms of kern_sigtimedwait() which is capable of supporting all of their semantics. - These should be POSIX compliant but more careful review is needed before we announce this. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T23:31:50.000000Z K 7 svn:log V 83 - According to mike@FreeBSD.org SIGTHR should be hiden by #ifdef __BSD_VISIBLE END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T23:33:45.000000Z K 7 svn:log V 41 - Regen for the sig*wait* system calls. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-03-31T23:36:40.000000Z K 7 svn:log V 33 - Add a placeholder for sigwait END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:16:32.000000Z K 7 svn:log V 183 - In npxgetregs() use the td argument to save the fpu state from and not curthread. Nothing currently depends on this behavior. - Clean up an extra newline. Obtained from: bde END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:18:55.000000Z K 7 svn:log V 380 - Define a new md function 'casuptr'. This atomically compares and sets a pointer that is in user space. It will be used as the basic primitive for a kernel supported user space lock implementation. - Implement this function in x86's support.s - Provide stubs that return -1 in all other architectures. Implementations will follow along shortly. Reviewed by: jake END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:30:30.000000Z K 7 svn:log V 244 - Add two files to support the thr threading interface. - sys/thr.h contains the user space visible api that is intended only for use in threading library packages. - kern/kern_thr.c contains thr system calls and other thr specific code. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:30:54.000000Z K 7 svn:log V 18 - Add kern_thr.c END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:31:37.000000Z K 7 svn:log V 42 - Add the four thr related system calls. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:34:29.000000Z K 7 svn:log V 39 - Regen for thr related system calls. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T00:35:12.000000Z K 7 svn:log V 87 - We now have to include umtx.h and ucontext.h in the system call related headers. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:10:42.000000Z K 7 svn:log V 605 - Add an api for doing smp safe locks in userland. - umtx_lock() is defined as an inline in umtx.h. It tries to do an uncontested acquire of a lock which falls back to the _umtx_lock() system-call if that fails. - umtx_unlock() is also an inline which falls back to _umtx_unlock() if the uncontested unlock fails. - Locks are keyed off of the thr_id_t of the currently running thread which is currently just the pointer to the 'struct thread' in kernel. - _umtx_lock() uses the proc pointer to synchronize access to blocked thread queues which are stored in the first blocked thread. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:11:22.000000Z K 7 svn:log V 107 - Add an entry and a head for the queue of threads blocked on a umtx. - Add a prototype for thr_exit1(). END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:12:24.000000Z K 7 svn:log V 86 - Include umtx.h in files generated by makesyscalls.sh - Add system calls for umtx. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:13:29.000000Z K 7 svn:log V 42 - Add the kern_umtx.c file to the build. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:15:56.000000Z K 7 svn:log V 34 - Add thr and umtx system calls. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:22:18.000000Z K 7 svn:log V 19 - Regen for umtx. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T01:26:20.000000Z K 7 svn:log V 349 - Borrow the KSE single threading code for exec and exit. We use the check if (p->p_numthreads > 1) and not a flag because action is only necessary if there are other threads. The rest of the system has no need to identify thr threaded processes. - In kern_thread.c use thr_exit1() instead of thread_exit() if P_THREADED is not set. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T02:32:07.000000Z K 7 svn:log V 129 - Mark the various thr syscalls as MP safe. Previously there was a bug if this was not done since thr_exit() unwinds giant. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T02:32:53.000000Z K 7 svn:log V 58 - thr_exit() should no longer be called with Giant held. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T02:34:21.000000Z K 7 svn:log V 10 - Regen. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T02:37:04.000000Z K 7 svn:log V 19 Implement casuptr. END K 10 svn:author V 3 tjr K 8 svn:date V 27 2003-04-01T02:42:02.000000Z K 7 svn:log V 107 Give the M_WAITOK flag explicitly to the MALLOC call to silence a runtime warning ("Bad malloc flags: 0"). END K 10 svn:author V 3 tjr K 8 svn:date V 27 2003-04-01T02:47:09.000000Z K 7 svn:log V 110 Specify the M_WAITOK flag explicitly in the MALLOC call to silence a runtime warning ("Bad malloc flags: 0"). END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T03:05:46.000000Z K 7 svn:log V 77 Rename pcb_fp to pcb_sp, so as to not be confused with floating point state. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T03:46:29.000000Z K 7 svn:log V 163 - Add libthr but don't hook it up to the regular build yet. This is an adaptation of libc_r for the thr system call interface. This is beta quality code. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T03:51:08.000000Z K 7 svn:log V 41 - Commit the forgotten libthr/sys bits. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T04:02:45.000000Z K 7 svn:log V 196 - Rename pcb_fpstate to pcb_ufp (user floating point), and change it to a simple array of 64 ints. - Use a critical section when saving floating point state in cpu_fork instead of sched_lock. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T04:19:29.000000Z K 7 svn:log V 26 Implement cpu_set_upcall. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T04:29:03.000000Z K 7 svn:log V 117 - Don't allow tf_wstate to be set in set_regs. - Clear FPRS_FEF in set_fpregs so the new registers will be reloaded. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T04:49:12.000000Z K 7 svn:log V 29 - Spell SIGSETOR correctly. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T04:58:50.000000Z K 7 svn:log V 171 - Add a flags field to struct pcb. Use this to keep track of wether or not the pcb has floating point registers saved in it. - Implement get_mcontext and set_mcontext. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T07:07:38.000000Z K 7 svn:log V 64 - Adjust the makefiles so we have a per architecture makefile. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T07:33:17.000000Z K 7 svn:log V 68 Remove the old config interface now that the new OAM is functional. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T07:33:56.000000Z K 7 svn:log V 77 Remove the old config interface, the new OAM is sufficiently functional now. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T08:10:21.000000Z K 7 svn:log V 66 Fix KASSERT syntax errors. Please compile LINT before commiting. END K 10 svn:author V 5 mdodd K 8 svn:date V 27 2003-04-01T08:21:44.000000Z K 7 svn:log V 872 Implement support for RFC 3514 (The Security Flag in the IPv4 Header). (See: ftp://ftp.rfc-editor.org/in-notes/rfc3514.txt) This fulfills the host requirements for userland support by way of the setsockopt() IP_EVIL_INTENT message. There are three sysctl tunables provided to govern system behavior. net.inet.ip.rfc3514: Enables support for rfc3514. As this is an Informational RFC and support is not yet widespread this option is disabled by default. net.inet.ip.hear_no_evil If set the host will discard all received evil packets. net.inet.ip.speak_no_evil If set the host will discard all transmitted evil packets. The IP statistics counter 'ips_evil' (available via 'netstat') provides information on the number of 'evil' packets recieved. For reference, the '-E' option to 'ping' has been provided to demonstrate and test the implementation. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T08:57:28.000000Z K 7 svn:log V 26 Fix KASSERT syntax error. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T09:02:58.000000Z K 7 svn:log V 38 Don't include needlessly. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T09:07:36.000000Z K 7 svn:log V 175 - p will be unused in cursig() if INVARIANTS is not defined. Access it through td->td_proc to avoid the unused variable. Spotted by: Maxim Konovalov END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T09:23:24.000000Z K 7 svn:log V 83 - smb_proc_intr is now spelled smb_td_intr. Noticed by: phk Pointy hat to: jeffr END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T09:24:12.000000Z K 7 svn:log V 57 - smb_td_intr takes a thread as an argument not a proc. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T11:11:38.000000Z K 7 svn:log V 53 This can't obviously be cleaner than system headers. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T11:29:29.000000Z K 7 svn:log V 100 Pass NO_WARNS to the build-tool stage too; -Wsystem-headers isn't known to the 4.x system compiler. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T11:39:04.000000Z K 7 svn:log V 227 Parallelize (on the top-level SUBDIR list) the "all" stage of buildworld. This gives 5-11% percent gain in real buildworld times on various UP and SMP systems here. I used 4 * hw.ncpu as an argument to "make -j" in my tests. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T12:34:47.000000Z K 7 svn:log V 69 Add #include so we don't rely on doing it. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T12:35:17.000000Z K 7 svn:log V 65 Include rather than trusting to do so. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T12:37:54.000000Z K 7 svn:log V 59 Mark bits that do not require an object directory as such. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T12:49:40.000000Z K 7 svn:log V 33 Introduce bioq_flush() function. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T12:50:40.000000Z K 7 svn:log V 41 libtelnet depends on OpenSSL. PR: 50507 END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T13:19:14.000000Z K 7 svn:log V 71 Start to split the GEOM/diskdriver specific bits into geom/geom_disk.h END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T13:33:28.000000Z K 7 svn:log V 28 Don't include . END K 10 svn:author V 3 yar K 8 svn:date V 27 2003-04-01T14:44:53.000000Z K 7 svn:log V 93 Don't die of SIGSEGV on a missing fstype field in a saved disklabel file. MFC after: 1 week END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T15:06:26.000000Z K 7 svn:log V 252 Use bioq_flush() to drain a bio queue with a specific error code. Retain the mistake of not updating the devstat API for now. Spell bioq_disksort() consistently with the remaining bioq_*(). #include where this is more appropriate. END K 10 svn:author V 3 yar K 8 svn:date V 27 2003-04-01T15:23:55.000000Z K 7 svn:log V 200 Give clearer diagnostics on some cases of invalid combinations of mdconfig(8) command-line arguments. Make mdconfig(8) accept "-a -f file -o options" equally with "-a -f file" (assuming "-t vnode".) END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2003-04-01T16:31:12.000000Z K 7 svn:log V 182 Add a crutch so that direct-mapped DMA works on alpha. This is a step toward preventing this driver from crashing an alpha at boot. Tested by: Oliver Lehmann END K 10 svn:author V 3 jhb K 8 svn:date V 27 2003-04-01T17:45:27.000000Z K 7 svn:log V 182 Complete removal of 320.rdist by removing its entry from periodic.conf and removing the related 220.backup-distfile script and associatd periodic.conf entry. Discussed with: obrien END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T18:55:04.000000Z K 7 svn:log V 69 I think the divorce successed, so stop #including END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T18:55:46.000000Z K 7 svn:log V 62 Remove the #define for bioqdisksort(), it's no longer needed. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T18:57:14.000000Z K 7 svn:log V 44 Include not END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-01T19:00:38.000000Z K 7 svn:log V 28 #include END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-01T19:59:30.000000Z K 7 svn:log V 45 Removed redundant lines. Noticed by: obrien END K 10 svn:author V 4 ceri K 8 svn:date V 27 2003-04-01T20:25:46.000000Z K 7 svn:log V 160 [1] - Document EHOSTUNREACH as a possible error [2] - Remove a contraction PR: docs/50401 Submitted by: [1] Slaven Rezic MFC after: 1 week END K 10 svn:author V 3 jhb K 8 svn:date V 27 2003-04-01T20:32:01.000000Z K 7 svn:log V 93 Remove 220.backup-distfile since it has been deleted. Reported by: mdodd Pointy hat to: jhb END K 10 svn:author V 3 hsu K 8 svn:date V 27 2003-04-01T21:16:46.000000Z K 7 svn:log V 370 Observe conservation of packets when entering Fast Recovery while doing Limited Transmit. Only artificially inflate the congestion window by 1 segment instead of the usual 3 to take into account the 2 already sent by Limited Transmit. Approved in principle by: Mark Allman , Hari Balakrishnan , Sally Floyd END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T22:39:31.000000Z K 7 svn:log V 334 - Restore old mutex code from libc_r. It is more standards compliant. This was changed because originally we were blocking on the umtx and allowing the kernel to do the queueing. It was decided that the lib should queue and start the threads in the order it decides and the umtx code would just be used like spinlocks. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T22:40:29.000000Z K 7 svn:log V 97 - Reenable setschedparam and the prioceiling code now that the mutex code has been restored. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-01T22:41:41.000000Z K 7 svn:log V 209 - Don't drop and reacquire giant in thread_suspend(). Change callers to do this manually. This will facilitate the unrolling of giant. - Don't allow giant to recurse anymore. This should never happen. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T23:18:13.000000Z K 7 svn:log V 265 - Set the version number in the mcontext in get_mcontext and check it in set_mcontext. - Don't make assumptions about the alignment of the mcontext inside of the ucontext; we have to save the floating point registers to the pcb and then copy to the mcontext. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T23:25:18.000000Z K 7 svn:log V 223 - Fix UC_COPY_SIZE. Adding up the size of structure fields doesn't take alignment into account. - Return EJUSTRETURN from set_context on success to avoid clobbering the first 2 out registers with td_retval on sparc64. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T23:28:50.000000Z K 7 svn:log V 23 Implement makecontext. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-01T23:52:24.000000Z K 7 svn:log V 17 Add MD makefile. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-02T03:05:39.000000Z K 7 svn:log V 314 - Define curthread as _get_curthread() and remove all direct calls to _get_curthread(). This is similar to the kernel's curthread. Doing this saves stack overhead and is more convenient to the programmer. - Pass the pointer to the newly created thread to _thread_init(). - Remove _get_curthread_slow(). END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-02T06:46:43.000000Z K 7 svn:log V 942 - Add in support for KSEs with 0 slice values on the run queue. If we try to select a KSE with a slice of 0 we will update its slice and insert it onto the next queue. - Pass the KSE instead of the ksegrp into sched_slice(). This more accurately reflects the behavior of the code. Slices are granted to kses. - Add a function kseq_nice_min() which finds the smallest nice value assigned to the kseg of any KSE on the queue. - Rewrite the logic in sched_slice(). Add a large comment describing the new slice selection scheme. To summarize, slices are assigned based on the nice value. Priorities are still calculated based on the nice and interactivity of a process. Slice sizes of 0 may be granted for KSEs whos nice is 20 or futher away from the lowest nice on the run queue. Other nice values are scaled across the range [min, min+20]. This fixes ULEs bad behavior with positively niced processes. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-02T08:02:27.000000Z K 7 svn:log V 138 - Make casuptr return the old value of the location we're trying to update, and change the umtx code to expect this. Reviewed by: jeff END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-02T08:03:42.000000Z K 7 svn:log V 49 Implement cpu_thread_setup. Fix cpu_set_upcall. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-02T08:15:07.000000Z K 7 svn:log V 71 Implement _get_curthread and _set_curthread. This is especially easy. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-02T08:21:24.000000Z K 7 svn:log V 66 - I meant divide by two and not shift by two in SCHED_PRI_NHALF. END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-02T08:22:33.000000Z K 7 svn:log V 93 - Make the interactivity calculator decay faster. - Make the pcpu estimator update faster. END K 10 svn:author V 5 maxim K 8 svn:date V 27 2003-04-02T09:20:08.000000Z K 7 svn:log V 74 o Fix error messages formatting, style. Prodded by: bde Reviewed by: bde END K 10 svn:author V 5 mdodd K 8 svn:date V 27 2003-04-02T10:28:47.000000Z K 7 svn:log V 85 Sync constant define with NetBSD. Requested by: Tom Spindler END K 10 svn:author V 4 nyan K 8 svn:date V 27 2003-04-02T10:49:48.000000Z K 7 svn:log V 54 Merged from sys/dev/syscons/syscons.c revision 1.400. END K 10 svn:author V 4 nyan K 8 svn:date V 27 2003-04-02T10:51:06.000000Z K 7 svn:log V 59 MFi386: revisions 1.556 and 1.557. Backout revision 1.312. END K 10 svn:author V 3 suz K 8 svn:date V 27 2003-04-02T11:35:51.000000Z K 7 svn:log V 53 fixed invalid pointer reference Obtained from: KAME END K 10 svn:author V 3 sos K 8 svn:date V 27 2003-04-02T11:36:43.000000Z K 7 svn:log V 72 Instead of not grapping a request binfinish(ENOMEM) it if malloc fails. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-02T13:09:50.000000Z K 7 svn:log V 223 Properly handle races between open/close and orphan. KASSERT the race between close and strategy, it is an error in the upper echelons if this happens, Add XXX: comment explaining why the ioctl/orphan race is not closed. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-02T13:10:40.000000Z K 7 svn:log V 57 Only orphan things if the open/close actually succeeded. END K 10 svn:author V 5 maxim K 8 svn:date V 27 2003-04-02T14:52:21.000000Z K 7 svn:log V 17 IP_EVIL -> IP_EF END K 10 svn:author V 3 hsu K 8 svn:date V 27 2003-04-02T15:24:50.000000Z K 7 svn:log V 163 Need to hold the same SMP lock for (knote) list traversal as for list manipulation. This lock also protects read-modify-write operations on the pipe_state field. END K 10 svn:author V 3 mux K 8 svn:date V 27 2003-04-02T16:47:16.000000Z K 7 svn:log V 540 Convert the fxp(4) driver to the busdma API. This patch is rather big because I had to significantly redesign the driver to make the busdma conversion possible. Most notably, hardware and software structures were carefully splitted to get rid of all the structs overlapping evilness. Special thanks to phk and Richard Puga for providing me with fxp(4) hardware to do this work. Thanks to marcel for testing this on ia64, and to Fred Clift for testing this on alpha. Tested on: i386, ia64, alpha END K 10 svn:author V 5 mdodd K 8 svn:date V 27 2003-04-02T18:02:58.000000Z K 7 svn:log V 58 - Use the correct constant define. - Add a missing break. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-02T20:13:59.000000Z K 7 svn:log V 56 Make disktab(5) MI (repo-copied from etc.i386/disktab). END K 10 svn:author V 5 mdodd K 8 svn:date V 27 2003-04-02T20:14:44.000000Z K 7 svn:log V 87 Back out support for RFC3514. RFC3514 poses an unacceptale risk to compliant systems. END K 10 svn:author V 2 ru K 8 svn:date V 27 2003-04-02T20:16:02.000000Z K 7 svn:log V 122 Drop the silly notion that fd* entries have both `a' and `b' partitions; this does nothing except upsetting disklabel(8). END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2003-04-02T20:22:29.000000Z K 7 svn:log V 442 If stat() on the terminal specified in utmp fails due to ENOENT, don't print a warning, and set the idletime variable for the entry to -1; then pick up the -1 later in sprint() and lprint() and ignore those idle times by printing just whitespace. When third party applications, such as kdm, insert utmp entries, they sometimes use strings like ":0", which can't be stat()'d and currently result in warnings that are not helpful to the user. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-02T20:41:18.000000Z K 7 svn:log V 322 Change events to have an array of "void *" references, and give the event posting functions varargs to fill these. Attribute g_call_me() to appropriate g_geom's where necessary. Add a flag argument to g_call_me() methods which will be used to signal cancellation of events in the future. This commit should be a no-op. END K 10 svn:author V 3 phk K 8 svn:date V 27 2003-04-02T21:10:04.000000Z K 7 svn:log V 62 Add handling for cancelled events in the g_call_me() methods. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2003-04-02T21:54:51.000000Z K 7 svn:log V 104 Lock the process before sending it a SIGIO. Not doing so is a panic(2) implementation with INVARIANTS. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2003-04-02T22:13:56.000000Z K 7 svn:log V 274 When kill(-1) returns ESRCH, it could be because the current process doesn't have a process group, which can occur if you're working with a custom init that doesn't set up a full tty context. Rather than refusing to reboot, ignore ESRCH from the kill attempt in reboot(8). END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-02T22:53:52.000000Z K 7 svn:log V 67 - Don't overrun the ldt buffer. Submitted by: gordan@freebsd.org END K 10 svn:author V 5 peter K 8 svn:date V 27 2003-04-02T23:53:30.000000Z K 7 svn:log V 1780 Commit a partial lazy thread switch mechanism for i386. it isn't as lazy as it could be and can do with some more cleanup. Currently its under options LAZY_SWITCH. What this does is avoid %cr3 reloads for short context switches that do not involve another user process. ie: we can take an interrupt, switch to a kthread and return to the user without explicitly flushing the tlb. However, this isn't as exciting as it could be, the interrupt overhead is still high and too much blocks on Giant still. There are some debug sysctls, for stats and for an on/off switch. The main problem with doing this has been "what if the process that you're running on exits while we're borrowing its address space?" - in this case we use an IPI to give it a kick when we're about to reclaim the pmap. Its not compiled in unless you add the LAZY_SWITCH option. I want to fix a few more things and get some more feedback before turning it on by default. This is NOT a replacement for Bosko's lazy interrupt stuff. This was more meant for the kthread case, while his was for interrupts. Mine helps a little for interrupts, but his helps a lot more. The stats are enabled with options SWTCH_OPTIM_STATS - this has been a pseudo-option for years, I just added a bunch of stuff to it. One non-trivial change was to select a new thread before calling cpu_switch() in the first place. This allows us to catch the silly case of doing a cpu_switch() to the current process. This happens uncomfortably often. This simplifies a bit of the asm code in cpu_switch (no longer have to call choosethread() in the middle). This has been implemented on i386 and (thanks to jake) sparc64. The others will come soon. This is actually seperate to the lazy switch stuff. Glanced at by: jake, jhb END K 10 svn:author V 4 jeff K 8 svn:date V 27 2003-04-03T00:29:28.000000Z K 7 svn:log V 349 - Keep seperate statistics and run queues for different scheduling classes. - Treat each class specially in kseq_{choose,add,rem}. Let the rest of the code be less aware of scheduling classes. - Skip the interactivity calculation for non TIMESHARE ksegrps. - Move slice and runq selection into kseq_add(). Uninline it now that it's big. END K 10 svn:author V 4 jake K 8 svn:date V 27 2003-04-03T03:34:50.000000Z K 7 svn:log V 293 - Pass a ucontext_t to _set_curthread. If non-NULL the new thread is set as curthread in the new context, so that it will be set automatically when the thread is switched to. This fixes a race where we'd run for a little while with curthread unset in _thread_start. Reviewed by: jeff END K 10 svn:author V 6 murray K 8 svn:date V 27 2003-04-03T04:38:14.000000Z K 7 svn:log V 184 Remove dokern.sh from the boot floppies once again. There is no associated module so the removal goes here, rather than in drivers.conf(5). Submitted by: bmah Approved by: re@ cabal END K 10 svn:author V 6 murray K 8 svn:date V 27 2003-04-03T04:39:44.000000Z K 7 svn:log V 85 MFR48: Remove awi driver from boot kernel -- we are short of space. Approved by: re END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2003-04-03T05:13:27.000000Z K 7 svn:log V 476 MFp4 @27667: WARNS=5 cleanup on i386. Remove the unused FILE\ *tf from print_mesg args, and the bogus passing in of an uninitialised FILE* for it. Call a timeval 'now' instead of 'clock' due to shadowing. Remove a nested localtime declaration. Make the delete invite argument match the ID type, u_int32_t. Use const for pointers to const items. Cast to long where printing as such. Include netinet/in.h for htonl/htons. Reviewed by: imp END K 10 svn:author V 3 imp K 8 svn:date V 27 2003-04-03T05:28:09.000000Z K 7 svn:log V 527 Kill aout to elf upgrade path. From an old comment in the makefile: # If -DWANT_AOUT is specified, a `make world' with OBJFORMAT=elf will # update the legacy support for aout. This includes all libraries, ld.so # and boot objects. This part of build should be regarded as # deprecated and you should _not_ expect to be able to do this past the # release of 4.0. You have exactly one major release to move entirely # to elf. Now that we're past 5.0, these aren't needed at all anymore and it is well past its freshness date. END