Á8182984 16 1170 271 168 168 211 241 276 241 287 754 263 480 371 626 626 2146 K 10 svn:author V 7 attilio K 8 svn:date V 27 2008-09-12T21:44:01.755287Z K 7 svn:log V 1072 - For any lock list we hold the head in order to reduce allocation from the free list and in this way avoid contention on the w_mtx. In order to make the code simple, we rely on the rule that when the head has not a child it also doesn't have other subsequent entries. Actually this assertion is broken because we can free all the head children and quit witness_unlock() with the head still allocated, with no children and subsequent entries present. Fix this by shifting the head if other entries are present and still freeing the object, but leaving always an head. - Fix witness_thread_has_locks() in order to report, correctly, if the lock list linked to a specific thread has children or not based on the above explained rule. - Fix a printout into DDB's "show alllocks" command in order to show, correctly, the process name that is really what we want. - Fix style(9) for a comment. Tested by: Giovanni Trematerra Reported by: Marko Kiiskila Sponsored by: Nokia END K 10 svn:author V 3 jhb K 8 svn:date V 27 2008-09-12T21:44:31.515969Z K 7 svn:log V 178 MFC: Remove the never-enabled attempt at using device cloning to provide per-file descriptor storage and replace it with cdevpriv (and enable it) instead. Approved by: re (kib) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2008-09-12T21:46:23.177765Z K 7 svn:log V 76 MFC: Fail detach if cpufreq_unregister() fails. Approved by: re (kensmith) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2008-09-12T21:46:59.373307Z K 7 svn:log V 76 MFC: Fail detach if cpufreq_unregister() fails. Approved by: re (kensmith) END K 10 svn:author V 3 jhb K 8 svn:date V 27 2008-09-12T21:50:42.501235Z K 7 svn:log V 118 MFC: Use kvm_getcptime() to fetch the equivalent of the kern.cp_time sysctl from a crash dump. Approved by: re (kib) END K 10 svn:author V 7 roberto K 8 svn:date V 27 2008-09-12T22:15:43.607684Z K 7 svn:log V 144 MFH r182944: Makefile.inc already defines OPENSSL if crypto is available/wanted. PR: bin/127296 Submitted by: olivier Approved by: re (gnn) END K 10 svn:author V 7 roberto K 8 svn:date V 27 2008-09-12T22:18:35.261183Z K 7 svn:log V 179 MFH r182857: Apply updated patch from bin/92839 to avoid two possible buffer overflows. PR: bin/92839 Submitted by: Helge Oldach Approved by: re (gnn) END K 10 svn:author V 7 roberto K 8 svn:date V 27 2008-09-12T22:20:24.149584Z K 7 svn:log V 144 MFH r182944: Makefile.inc already defines OPENSSL if crypto is available/wanted. PR: bin/127296 Submitted by: olivier Approved by: re (gnn) END K 10 svn:author V 7 roberto K 8 svn:date V 27 2008-09-12T22:22:31.093984Z K 7 svn:log V 190 MFH r182857: Apply updated patch from bin/92839 to avoid two possible buffer overflows. PR: bin/92839 Submitted by: Helge Oldach Approved by: re (gnn) END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2008-09-13T00:16:20.090012Z K 7 svn:log V 657 Extend global UNIX domain socket rwlock coverage to include incrementing and decrementing unp_rights, which may otherwise be corrupted. Be slightly more conversative in where we read unp_rights. This relates to one of the symptoms reported in the noted PR, but may not correct the actual high system time problem. The reporter has confirmed stability but not that the problem is eliminated. However, this is a useful fix to a clear locking bug. Note that this is not an MFC as the UNIX domain socket garbage collector has been replaced in 8.x. PR: 127024 Reported by: Anton Yuzhaninov Reviewed by: kib Approved by: re (kib) END K 10 svn:author V 3 mav K 8 svn:date V 27 2008-09-13T07:36:16.477005Z K 7 svn:log V 170 MFC r182749 Add one more SII3132 chip PCI ID found on Adaptec Serial ATA II RAID 1220SA. Submitted by: Dmitry S. Luhtionov Approved by: re (kib) END K 10 svn:author V 3 mav K 8 svn:date V 27 2008-09-13T09:17:02.537470Z K 7 svn:log V 387 We can't implicitly trust the hook on NGQF_FN/NGQF_FN2 processing in ng_apply_item(). There are possible (and I have got one) use-after-free class panics because of it. If hook is specified, require it to be valid at the apply time. The only exceptions are the internal ng_con_part2(), ng_con_part3() and ng_rmhook_part2() functions which are specially made to work with invalid hooks. END K 10 svn:author V 2 bz K 8 svn:date V 27 2008-09-13T14:06:36.868967Z K 7 svn:log V 279 Turn on TCP_SIGNATURE for LINT builds. This should catch situations we ran into in the past where places hidden by TCP_SIGNATURE were missed. It is possible to turn it on now that FAST_IPSEC (now know as IPSEC) is enabled for LINT and the default and only IPsec implementation. END K 10 svn:author V 6 marius K 8 svn:date V 27 2008-09-13T15:17:21.013911Z K 7 svn:log V 530 MFC: r182688 Bypass isa_probe_children(9) and directly call bus_generic_attach(9) in order to avoid the invasive probes done by identify-routines of ISA drivers, which may access unassigned addresses or those of unrelated devices and thus in turn can trigger master/target aborts as revealed by r182108 and ahc(4). I think that this is also the cause of the hang previously seen on B100 blades during boot. Bypassing isa_probe_children(9) also avoids adding ISA hints, which just can be wrong for sparc64. Approved by: re (kib) END K 10 svn:author V 6 marius K 8 svn:date V 27 2008-09-13T15:17:23.733853Z K 7 svn:log V 530 MFC: r182688 Bypass isa_probe_children(9) and directly call bus_generic_attach(9) in order to avoid the invasive probes done by identify-routines of ISA drivers, which may access unassigned addresses or those of unrelated devices and thus in turn can trigger master/target aborts as revealed by r182108 and ahc(4). I think that this is also the cause of the hang previously seen on B100 blades during boot. Bypassing isa_probe_children(9) also avoids adding ISA hints, which just can be wrong for sparc64. Approved by: re (kib) END K 10 svn:author V 3 mav K 8 svn:date V 27 2008-09-13T16:56:03.269909Z K 7 svn:log V 2052 My massive snd_hda driver update. Because of using more clear and same time more functional codec parser new driver is able to handle more codecs, use them better then before and without most of previous quirks. All of tested codecs itself manage playback, record, input mixing and monitoring quite fine. In all of investigated trouble cases problem was found or in nonstandard codec usage or incorrect codec configuration made by BIOS. Most of that cases could be fixed using device hints, some of which are already included to the driver. New driver supports multiple codecs per HDA bus, multiple audio function groups per codec and multiple logical sound devices per audio function group. So don't worry when you get several PCM devices instead of one, it is normal. It is usual situation with powerful codecs to provide, for example, 3 PCM devices: one for 7.1 playback and main recording, one for headset and one for digital SPDIF I/O. New driver implements Universal Audio Architecture (UAA) much better then previous one. Most information about recommended codec usage now taken from the codec configuration registers initialized by BIOS. User may alter that configuration using device hints to reconfigure logical audio devices to his needs in a very broad range up to the limits of the codec functionality. New driver supports digital PCM playback and AC3 pass-through. I am not sure about completeness of this implementation, but I have several success stories including my own. Vchans subsystem does not support AC3 pass-through so it had to be disabled for that devices at this moment. New driver is ready for multichannel playback, but until our OSS is unable to use this it will just duplicate same stereo stream into all channel pairs. New driver supports suspend/resume. I am unable to really test this part myself, but I have got several success stories. Driver has very informative verbose boot messages. So if you have any questions or problems - enable and read them first. Discussed on: freebsd-multimedia@ Tested by: many END