ƒ²I205000 185 468 468 212 190 544 187 189 175 368 228 732 160 141 360 439 165 145 124 116 182 263 503 156 292 631 261 149 256 421 183 429 389 142 219 254 229 181 118 153 164 185 316 236 139 114 166 120 458 140 220 264 177 178 176 114 310 878 878 178 293 206 399 166 273 209 251 649 136 154 121 290 470 244 229 187 493 209 609 209 225 168 177 292 539 160 361 165 393 180 344 364 1125 428 144 175 113 224 193 210 1230 155 233 161 114 445 131 130 138 191 564 133 124 124 202 205 750 448 115 249 334 281 311 181 354 141 116 193 1174 482 158 306 197 249 128 216 144 445 192 623 261 285 147 123 183 145 363 802 409 143 213 680 166 198 162 533 193 364 534 144 301 628 578 237 364 258 196 136 265 227 490 125 156 470 1137 122 115 102 102 167 319 140 156 157 898 898 K 10 svn:author V 7 fabient K 8 svn:date V 27 2010-03-11T07:38:27.467246Z K 7 svn:log V 371 MFC r204878: Change the way shutdown is handled for log file. pmc_flush_logfile is now non-blocking and just ask the kernel to shutdown the file. From that point, no more data is accepted by the log thread and when the last buffer is flushed the file is closed. This will remove a deadlock between pmcstat asking for flush while it cannot flush the pipe itself. END K 10 svn:author V 7 fabient K 8 svn:date V 27 2010-03-11T07:39:36.809206Z K 7 svn:log V 371 MFC r204878: Change the way shutdown is handled for log file. pmc_flush_logfile is now non-blocking and just ask the kernel to shutdown the file. From that point, no more data is accepted by the log thread and when the last buffer is flushed the file is closed. This will remove a deadlock between pmcstat asking for flush while it cannot flush the pipe itself. END K 10 svn:author V 6 marius K 8 svn:date V 27 2010-03-11T07:46:17.175468Z K 7 svn:log V 116 Fix a typo in r204974 so that FTYPE_LNG matches the initialisers of the opmask array. Pointed out by: Peter Jeremy END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T08:03:56.961735Z K 7 svn:log V 94 Revert r204992 and just wrap it all in ifdef INVARIANTS to fix the debug and non-debug cases. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T08:30:05.114186Z K 7 svn:log V 446 I would've expected the Octeon to have a double-counting timer as the code suggests, but at least on my CAM-0100 this seems to not be the case. Pass 0 to mips_timer_init_params for double_count instead of 1 -- now timings on my system match wall clock. The tick code really should be updated for 64-bit counter reads on the CPUs that support them -- I wonder if the t_delta wrongness is due to inaccuracy from the hand-rolled counter wrapping. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T08:33:39.226619Z K 7 svn:log V 91 Wrap the proc wakeup special case for ddb in ifdef DDB. Submitted by: Giovanni Trematerra END K 10 svn:author V 3 avg K 8 svn:date V 27 2010-03-11T08:55:03.241324Z K 7 svn:log V 97 MFC r203776: acpi cpu: probe+attach before all other enumerated children X-MFCto7 after: 1 week END K 10 svn:author V 3 avg K 8 svn:date V 27 2010-03-11T08:58:13.381392Z K 7 svn:log V 83 MFC r203785: acpi: drop the second bus_generic_attach pass X-MFCto7 after: 1 week END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-11T11:09:58.211627Z K 7 svn:log V 276 Make script(1) a little less broken. Close the file descriptor to the TTY. There is no reason why the parent process should keep track of the descriptor. This ensures that the application inside properly drains the TTY during exit(2). Reported by: alfred MFC after: 2 weeks END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-11T11:28:29.627875Z K 7 svn:log V 136 Improve the change made in the previous commit. doshell() never returns, so there is no need to see whether we are the parent process. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2010-03-11T11:33:04.574563Z K 7 svn:log V 635 Update nfsrv_getsocksndseq() for changes in TCP internals since FreeBSD 6.x: - so_pcb is now guaranteed to be non-NULL and valid if a valid socket reference is held. - Need to check INP_TIMEWAIT and INP_DROPPED before assuming inp_ppcb is a tcpcb, as it might be a tcptw or NULL otherwise. - tp can never be NULL by the end of the function, so only check TCPS_ESTABLISHED before extracting tcpcb fields. The NFS server arguably incorporates too many assumptions about TCP internals, but fixing that is left for nother day. MFC after: 1 week Reviewed by: bz Reviewed and tested by: rmacklem Sponsored by: Juniper Networks END K 10 svn:author V 5 joerg K 8 svn:date V 27 2010-03-11T12:18:52.294481Z K 7 svn:log V 66 The number after the command is *not* optional. MFC after: 1 day END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T13:16:06.217269Z K 7 svn:log V 49 Fix a comment nit. Submitted by: Alexander Best END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T14:17:37.737242Z K 7 svn:log V 267 Print out the family and model from the cpu_id. This is especially useful given the advent of the extended family and extended model fields. The values are printed in hex to match their common usage in documentation. Submitted by: Alexander Best MFC after: 1 week END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-11T14:49:06.717829Z K 7 svn:log V 338 Provide groundwork for 32-bit binary compatibility on non-x86 platforms, for upcoming 64-bit PowerPC and MIPS support. This renames the COMPAT_IA32 option to COMPAT_FREEBSD32, removes some IA32-specific code from MI parts of the kernel and enhances the freebsd32 compatibility code to support big-endian platforms. Reviewed by: kib, jhb END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-11T14:54:54.891191Z K 7 svn:log V 65 Accidentally committed test code. Remove it. Big pointy hat: me END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-11T14:56:59.458351Z K 7 svn:log V 45 Regen after big endian compatibility import. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T15:13:55.768516Z K 7 svn:log V 32 Style fixes. Submitted by: bde END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T15:25:47.714439Z K 7 svn:log V 24 Small whitespace fixes. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-11T15:33:12.346315Z K 7 svn:log V 82 The first transport is away! Merge back in the final version of COMPAT_FREEBSD32. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T16:58:15.837786Z K 7 svn:log V 170 Fix a bug in the previous change: remove function-local definition of tcp_key and udp_key that shadows the global definition. PR: threads/144558 Submitted by: Sam Robb END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-11T17:03:32.636966Z K 7 svn:log V 410 - Use an initializer macro to initialize fields in 'fake' FILE objects used by *sprintf(), etc. - Explicitly initialize _fl_mutex to PTHREAD_MUTEX_INITIALIZER for all FILE objects. This is currently a nop on FreeBSD, but is import for other platforms (or in the future) where PTHREAD_MUTEX_INITIALIZER is not simply zero. PR: threads/141198 Reported by: Jeremy Huddleston @ Apple MFC after: 2 weeks END K 10 svn:author V 8 bschmidt K 8 svn:date V 27 2010-03-11T17:11:07.648038Z K 7 svn:log V 59 MFC r204213: Fix some typos. Approved by: rpaulo (mentor) END K 10 svn:author V 8 bschmidt K 8 svn:date V 27 2010-03-11T17:15:40.059645Z K 7 svn:log V 194 MFC r203934: Fix for the Intel WiFi Link 1000. The EEPROM image is in the OTPROM block before the last block, not in the last block itself. Approved by: rpaulo (mentor) Obtained from: OpenBSD END K 10 svn:author V 6 qingli K 8 svn:date V 27 2010-03-11T17:56:46.725233Z K 7 svn:log V 535 The if_tap interface is of IFT_ETHERNET type, but it does not set or update the if_link_state variable. As such RT_LINK_IS_UP() fails for the if_tap interface. Also, the RT_LINK_IS_UP() needs to bypass all loopback interfaces because loopback interfaces are considered up logically as long as the system is running. This patch fixes the above issues by setting and updating the if_link_state variable when the tap interface is opened or closed respectively. Similary approach is already done in the if_tun device. MFC after: 3 days END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T20:32:28.900324Z K 7 svn:log V 163 I don't know that a critical section is required here, but if it is we really don't want to get interrupted between reading the counter and calculating the ticks. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T20:41:21.344471Z K 7 svn:log V 53 Reapply r185998 which was overwritten at some point. END K 10 svn:author V 3 raj K 8 svn:date V 27 2010-03-11T21:04:29.986066Z K 7 svn:log V 163 Let detailed info about CPU features print on Marvell Sheeva CPU as well. Provide missing entry in the cpu_classes[]. Reported by: Maks Verver MFC after: 1 week END K 10 svn:author V 3 raj K 8 svn:date V 27 2010-03-11T21:16:54.345085Z K 7 svn:log V 328 Fix ARM cache handling yet more. 1) vm_machdep.c: remove the dangling allocations so they do not un-necessarily turn off the cache upon consecutive access. 2) busdma_machdep.c: remove the same amount than shadow mapped. Reported by: Maks Verver Submitted by: Mark Tinguely Reviewed by: Grzegorz Bernacki MFC after: 3 days END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:42:09.652707Z K 7 svn:log V 87 Use wMaxPacketSize for the uftdi input buffer size. Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:45:31.701866Z K 7 svn:log V 332 - make the usb_temp_setup() and usb_temp_unsetup() functions public so that other modules can generate USB descriptors. - extend the vendor specific request function by one length pointer argument, because not all descriptors store the length in the first byte. For example HID descriptors. Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:46:33.236938Z K 7 svn:log V 292 It appears that some UVISOR devices do not handle when the clear stall command is issued at the beginning of the initial IN/OUT data transfers. Reason unknown, probably firmware fault. Now the stall is only cleared on data transfer errors. PR: usb/144199 Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:47:25.977878Z K 7 svn:log V 46 Add new uvisor(4) device ID. PR: usb/144201 END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:48:10.182665Z K 7 svn:log V 122 isochronous endpoint descriptors should have two more bytes which are zero by default. Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:49:00.781335Z K 7 svn:log V 157 For USS820 driver we need to manually reset TX FIFO at each SETUP transaction because the chip doesn't do this by itself. Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:49:43.480571Z K 7 svn:log V 132 Make sure there is a way to reset the endpoint FIFO on transfer errors for ISOCHRONOUS transfers Submitted by: Hans Petter Selasky END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:50:36.951635Z K 7 svn:log V 85 Implement USB kernel driver detach from userland. Submitted by: Hans Petter Selasky END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T21:54:09.425829Z K 7 svn:log V 21 Report media status. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:54:23.412615Z K 7 svn:log V 57 add new vendor ID for APACER Submitted by: Paul B Mahol END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:55:25.424802Z K 7 svn:log V 68 Add new device ID for the SMC 2514HUB Submitted by: Alexander Best END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T21:57:01.716283Z K 7 svn:log V 89 extend search for Apple Function Key. PR: usb/144414 Submitted by: Hans Petter Selasky END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:01:48.617970Z K 7 svn:log V 218 Inline a function only referenced once and remove a comment about not calling that function from some place in the code. Make the ifinit function set the link status. This fixes rgmx with the recent routing changes. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T22:05:12.567042Z K 7 svn:log V 139 - Integrate latest driver code from OpenBSD - Drain our tasks from the ieee80211 taskqueue - Add more IDs Submitted by: Akinori Furukoshi END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2010-03-11T22:09:21.368373Z K 7 svn:log V 43 Add device ID for the NATURAL4000 keyboard END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:11:24.282235Z K 7 svn:log V 17 Merge from head. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:13:16.985468Z K 7 svn:log V 69 Remove a silly comment, remove gratuitous braces. Suggested by: imp END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:14:07.314385Z K 7 svn:log V 23 Remove use of typestr. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:22:06.400816Z K 7 svn:log V 360 o) Eliminate use of sc->typestr, which is always NULL. o) Inline octeon_rgmx_mark_ready into octeon_rgmx_init. o) Add a media status handler that reports link and media status. o) Set link state when if_init is called. o) Remove some printfs related to driver state changes. o) Remove some gratuitous comments. Reviewed by: imp Sponsored by: Packet Forensics END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:25:53.552864Z K 7 svn:log V 43 Don't force single user on Octeon anymore. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:29:45.654475Z K 7 svn:log V 122 Add bpf and random to Octeon configurations, since they're needed to run dhclient and ssh respectively. Reviewed by: imp END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-11T22:42:33.902407Z K 7 svn:log V 169 implement listing of a subset of pipes/queues/schedulers. The filtering of the output is done in the kernel instead of userland to reduce the amount of data transfered. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T22:44:00.854384Z K 7 svn:log V 80 Add my copyright. This file seems to consist of my cpufunc.h and pefo's pio.h. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-11T23:05:13.922793Z K 7 svn:log V 81 o) Use standard interrupt disable/restore functions. o) Spell __func__ properly. END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-11T23:27:16.571516Z K 7 svn:log V 82 alias page lock to page queue lock for architectures with pmaps without page lock END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-11T23:32:10.671856Z K 7 svn:log V 20 further narrow diff END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T00:25:04.491626Z K 7 svn:log V 215 - we can't assert that the page queue mutex is not owned if we're aliasing the page lock to it - we need to acquire the page queue lock before calling pmap_enter_object on architectures that don't use page lock END K 10 svn:author V 7 delphij K 8 svn:date V 27 2010-03-12T00:51:13.010732Z K 7 svn:log V 781 MFC r203760: Improve time precision for grdc(6): Traditionally, grdc would obtain time through time(3) which in turn gets only the second part of clock (CLOCK_SECOND), and sleep for 1 second after each screen refresh. This approach would have two problems. First, we are not guaranteed to be waken up at the beginning of a whole second, which will typically exhibit as a "lag" on second number. Second, because we sleep for whole second, and the refresh process would take some time, the error would accumulate from time to time, making the lag variable. Make grdc(6) to use time(3) to get time only at the beginning, and sample time in CLOCK_REALTIME_FAST granularity after refreshing, and use the nanosecond part to caculate how much time we want to sleep. PR: bin/120813 END K 10 svn:author V 7 delphij K 8 svn:date V 27 2010-03-12T00:51:25.473692Z K 7 svn:log V 781 MFC r203760: Improve time precision for grdc(6): Traditionally, grdc would obtain time through time(3) which in turn gets only the second part of clock (CLOCK_SECOND), and sleep for 1 second after each screen refresh. This approach would have two problems. First, we are not guaranteed to be waken up at the beginning of a whole second, which will typically exhibit as a "lag" on second number. Second, because we sleep for whole second, and the refresh process would take some time, the error would accumulate from time to time, making the lag variable. Make grdc(6) to use time(3) to get time only at the beginning, and sample time in CLOCK_REALTIME_FAST granularity after refreshing, and use the nanosecond part to caculate how much time we want to sleep. PR: bin/120813 END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T01:20:10.856668Z K 7 svn:log V 81 o) Send outgoing packets to BPF on transmit. o) Handle setting promiscuous mode. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T01:51:39.196310Z K 7 svn:log V 195 Change octeon_rgmx_rx_promiscuous into octeon_rgmx_config_cam which does general CAM configuration stuff (in theory) and checks interface flags for its own use rather than expecting a caller to. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T02:55:10.243512Z K 7 svn:log V 105 In the event there are no mappings in segment 16 (e.g. for real-mode OF), make sure we don't go too far. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T02:56:45.207280Z K 7 svn:log V 301 o) Send packets being queued for transmission up to BPF if there's a listener. o) Properly configure the CAM to handle IFF_PROMISC and note where IFF_ALLMULTI handling would go if we didn't already force the NIC to receive all multicast traffic. Reviewed by: imp Sponsored by: Packet Forensics END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T02:56:49.402547Z K 7 svn:log V 66 Reduce redundancy in exec_setregs() arguments. Suggested by: kib END K 10 svn:author V 3 jhb K 8 svn:date V 27 2010-03-12T03:08:47.976278Z K 7 svn:log V 180 Fix the previous attempt to fix kernel builds of HEAD on 7.x. Use the __gnu_inline__ attribute for PMAP_INLINE when using the 7.x compiler to match what 7.x uses for PMAP_INLINE. END K 10 svn:author V 4 neel K 8 svn:date V 27 2010-03-12T03:49:17.285167Z K 7 svn:log V 115 Make the ddb command "show tlb" SMP friendly. It now accepts an argument to dump out the tlb of a particular cpu. END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T04:44:20.079143Z K 7 svn:log V 156 - cleanup whitespace - keep statistics per-cpu for per-cpu flowtables to avoid cache line contention - add sysctls to accumulate stats and report aggregate END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T05:03:26.701515Z K 7 svn:log V 554 - restructure flowtable to support ipv6 - add a name argument to flowtable_alloc for printing with ddb commands - extend ddb commands to print destination address or 4-tuples - don't parse ports in ulp header if FL_HASH_ALL is not passed - add kern_flowtable_insert to enable more generic use of flowtable (e.g. system calls for adding entries) - don't hash loopback addresses - cleanup whitespace - keep statistics per-cpu for per-cpu flowtables to avoid cache line contention - add sysctls to accumulate stats and report aggregate MFC after: 7 days END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T05:08:05.680582Z K 7 svn:log V 42 create branch for integrating ZFS updates END K 10 svn:author V 5 joerg K 8 svn:date V 27 2010-03-12T05:16:24.005887Z K 7 svn:log V 60 (r205011) The "number" argument is everything but optional. END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T06:31:19.919198Z K 7 svn:log V 27 fix stats reporting sysctl END K 10 svn:author V 2 jh K 8 svn:date V 27 2010-03-12T06:56:51.842309Z K 7 svn:log V 198 MFC r204447: In _gettemp(), check that the length of the path doesn't exceed MAXPATHLEN. Otherwise the path name (or part of it) may not fit to carrybuf causing a buffer overflow. PR: bin/140228 END K 10 svn:author V 5 edwin K 8 svn:date V 27 2010-03-12T06:57:53.184104Z K 7 svn:log V 375 - With the introduction of -A, -B and -3, not all combinations of arguments makes sense anymore. For example, what would a combination of -3 (show three months) and -y (show the whole year) do? We will abort on these cases. - Move the debug option -d to -H (from highlight), while -d is now used for setting the day of "today" so that -y and friends can be tested. END K 10 svn:author V 4 neel K 8 svn:date V 27 2010-03-12T07:08:20.739878Z K 7 svn:log V 150 - Enable kernel stack guard page. - Unmap the unused kernel stack page that we cannot use because it is not aligned on a (PAGE_SIZE * 2) boundary. END K 10 svn:author V 6 brooks K 8 svn:date V 27 2010-03-12T07:26:37.551818Z K 7 svn:log V 133 Regen: * Hart: rev 671 of pcidevs.txt; 22-01-2008 (D-M-Y). * Boemler: vendors.txt (2010-03126) PR: kern/133733 MFC after: 1 week END K 10 svn:author V 3 mav K 8 svn:date V 27 2010-03-12T07:49:10.079033Z K 7 svn:log V 95 Mask disk_idx to avoid panic because of extra bits set. PR: kern/102211 Submitted by: yoichi END K 10 svn:author V 3 rrs K 8 svn:date V 27 2010-03-12T08:10:30.304480Z K 7 svn:log V 400 With the recent change of the sctp checksum to support offload, no delayed checksum was added to the ip6 output code. This causes cards that do not support SCTP checksum offload to have SCTP packets that are IPv6 NOT have the sctp checksum performed. Thus you could not communicate with a peer. This adds the missing bits to make the checksum happen for these cards. PR: 144529 MFC after: 2 weeks END K 10 svn:author V 3 uqs K 8 svn:date V 27 2010-03-12T10:01:06.651237Z K 7 svn:log V 116 Fix several typos in macros or macro misusage. Found by: make manlint Reviewed by: ru Approved by: philip (mentor) END K 10 svn:author V 6 qingli K 8 svn:date V 27 2010-03-12T10:24:58.992261Z K 7 svn:log V 513 The flow-table module retrieves the destination and source address as well as the transport protocol port information from the outbound packets. The routing code is generic and compares every byte in the given sockaddr object. Therefore the temporary sockaddr objects must be cleared due to padding bytes. In addition, the port information must be stripped or the route search will either fail or return the incorrect route entry. Unit testing is done using OpenVPN over the if_tun interface. MFC after: 7 days END K 10 svn:author V 3 uqs K 8 svn:date V 27 2010-03-12T11:05:37.037668Z K 7 svn:log V 116 Fix typo in macro name and macro usage. Found by: make manlint Reviewed by: ru Approved by: harti, philip (mentor) END K 10 svn:author V 3 pjd K 8 svn:date V 27 2010-03-12T12:07:21.169891Z K 7 svn:log V 132 Remove bogus assertion. Reported by: Johan Ström Obtained from: OpenSolaris, Bug ID 6827260 MFC after: 1 week END K 10 svn:author V 3 pjd K 8 svn:date V 27 2010-03-12T13:36:58.290548Z K 7 svn:log V 76 Force commit to correct Bug ID: Obtained from: OpenSolaris, Bug ID 6920880 END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-12T13:53:52.652657Z K 7 svn:log V 83 svn merge -r 204832:205079 svn+ssh://svn.freebsd.org/base/stable/8/release/picobsd END K 10 svn:author V 7 glebius K 8 svn:date V 27 2010-03-12T14:51:42.302761Z K 7 svn:log V 195 Fix 'netstat -f netgraph', which I had broken in r163463 ling time ago in 2006. This linked list is actually needed for userland. PR: kern/140446 Submitted by: Adrian Steinmann END K 10 svn:author V 7 glebius K 8 svn:date V 27 2010-03-12T15:04:59.837449Z K 7 svn:log V 442 Now fix functionality of 'netstat -f netgraph' that hasn't worked starting from netgraph import in 1999. netstat(8) used pointer to node as node address, oops. That didn't work, we need the node ID in brackets to successfully address a node. We can't look into ng_node, due to inability to include netgraph/netgraph.h in userland code. So let the node make a hint for a userland, storing the node ID in its private data. MFC after: 2 weeks END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-12T15:21:13.973629Z K 7 svn:log V 66 partial merge of ipfw3 changes in preparation for MFC to RELENG_8 END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T15:57:56.065340Z K 7 svn:log V 260 Don't pretend we are limited to a 32-bit address space any more, and move KVA from its previous cramped 512 MB slot to an expansive 1024 TB domain in high memory. Next on the list is actually worrying about SLB replacement so we can use more than 16 GB of it. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T16:08:52.487928Z K 7 svn:log V 65 Bump kmem parameters on powerpc64 now that KVA is vastly larger. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T17:15:44.884607Z K 7 svn:log V 292 Move USER_ADDR into high memory, too. With the exception of possible OF mappings, all memory below 2^32 is now used for direct mapped quantities only, paving the way for a large-page direct map in this region, and removing assumptions about where physical memory happens to have been mapped. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-12T17:23:26.463512Z K 7 svn:log V 80 Missed some s/ps_strings/imgp->ps_strings that broke the amd64 and ia64 builds. END K 10 svn:author V 4 fanf K 8 svn:date V 27 2010-03-12T17:55:29.295036Z K 7 svn:log V 250 Update to upstream version 2.3 Add -V (display version) and -S (list controlling symbols per #if with nesting information) options, and improve unifdefall debug output. Done committing 0 revisions to SVN Obtained from: http://dotat.at/prog/unifdef END K 10 svn:author V 7 yongari K 8 svn:date V 27 2010-03-12T18:18:04.829560Z K 7 svn:log V 267 Reorder interrupt handler a bit such that producer/consumer index of status block is read first before acknowledging the interrupts. Otherwise bge(4) may get stale status block as acknowledging an interrupt may yield another status block update. Reviewed by: marius END K 10 svn:author V 7 yongari K 8 svn:date V 27 2010-03-12T18:41:41.028203Z K 7 svn:log V 1027 Implement Rx checksum offloading for Yukon EC, Yukon Ultra, Yukon FE and Yukon Ultra2. These controllers provide very simple checksum computation mechanism and it requires additional pseudo header checksum computation in upper stack. Even though I couldn't see much performance difference with/without Rx checksum offloading it may help notebook based controllers. Actually controller can compute two checksum value by giving different starting position of checksum computation on received frame. However, for long time, Marvell's checksum offloading engine have been known to have several silicon bugs so don't blindly trust computed partial checksum value. Instead, compute partial checksum twice by giving the same checksum computation position and compare the result. If the value is different it's clear indication of hardware bug. This configuration lose IP checksum offloading capability but I think it's better to take safe route. Note, Rx checksum offloading for Yukon XL was still disabled due to known silicon bug. END K 10 svn:author V 4 jkim K 8 svn:date V 27 2010-03-12T19:14:58.375448Z K 7 svn:log V 334 Tidy up callout for select(2) and read timeout. - Add a missing callout_drain(9) before the descriptor deallocation.[1] - Prefer callout_init_mtx(9) over callout_init(9) and let the callout subsystem handle the mutex for callout function. PR: kern/144453 Submitted by: Alexander Sack (asack at niksun dot com)[1] MFC after: 1 week END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T19:26:45.144799Z K 7 svn:log V 50 re-update copyright to 2010 pointed out by danfe@ END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T19:29:27.471248Z K 7 svn:log V 81 don't block on stream locking, too many parallel prefetches can kill performance END K 10 svn:author V 4 jkim K 8 svn:date V 27 2010-03-12T19:42:42.915852Z K 7 svn:log V 20 Fix a style(9) nit. END K 10 svn:author V 7 rnoland K 8 svn:date V 27 2010-03-12T19:43:39.737865Z K 7 svn:log V 127 Add pci ids for Intel Ironlake chipsets. These behave just like g45 for agp. Tested by: Torfinn Ingolfsen MFC after: 3 days END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-12T19:58:51.487660Z K 7 svn:log V 99 flowtable_get_hashkey is only used by a DDB function - move under #ifdef DDB pointed out by jkim@ END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T21:06:35.582256Z K 7 svn:log V 112 Use octeon_fpa_alloc_phys in a situation in which we don't need a usable pointer, rather than octeon_fpa_alloc. END K 10 svn:author V 7 delphij K 8 svn:date V 27 2010-03-12T21:14:56.801326Z K 7 svn:log V 1132 Two optimizations to MI strlen(3) inspired by David S. Miller's blog posting [1]. - Use word-sized test for unaligned pointer before working the hard way. Memory page boundary is always integral multiple of a word alignment boundary. Therefore, if we can access memory referenced by pointer p, then (p & ~word mask) must be also accessible. - Better utilization of multi-issue processor's ability of concurrency. The previous implementation utilized a formular that must be executed sequentially. However, the ~, & and - operations can actually be caculated at the same time when the operand were different and unrelated. The original Hacker's Delight formular also offered consistent performance regardless whether the input would contain characters with their highest-bit set, as it catches real nul characters only. These two optimizations has shown further improvements over the previous implementation on microbenchmarks on i386 and amd64 CPU including Pentium 4, Core Duo 2 and i7. [1] http://vger.kernel.org/~davem/cgi-bin/blog.cgi/2010/03/08#strlen_1 MFC after: 1 month END K 10 svn:author V 7 delphij K 8 svn:date V 27 2010-03-12T21:15:35.515504Z K 7 svn:log V 59 Follow up commit to reindent the code. MFC after: 1 month END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T21:20:36.164572Z K 7 svn:log V 135 Remove FAU initialize/enable/disable stubs, which are unused and as far as I can tell from the Simple Executive, entirely unnecessary. END K 10 svn:author V 7 rnoland K 8 svn:date V 27 2010-03-12T21:34:23.232267Z K 7 svn:log V 65 Add support of Intel Pineview chips, aka IGD. MFC after: 3 days END K 10 svn:author V 5 ticso K 8 svn:date V 27 2010-03-12T22:39:35.438693Z K 7 svn:log V 20 fix type in comment END K 10 svn:author V 3 rrs K 8 svn:date V 27 2010-03-12T22:58:52.391792Z K 7 svn:log V 352 The proper fix for the delayed SCTP checksum is to have the delayed function take an argument as to the offset to the SCTP header. This allows it to work for V4 and V6. This of course means changing all callers of the function to either pass the header len, if they have it, or create it (ip_hl << 2 or sizeof(ip6_hdr)). PR: 144529 MFC after: 2 weeks END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-12T23:23:46.160335Z K 7 svn:log V 36 sh: Add tests for " and $ inside `. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T23:46:26.591116Z K 7 svn:log V 33 Actually remove fau.c reference. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-12T23:57:27.528803Z K 7 svn:log V 41 Remove some unused functions and macros. END K 10 svn:author V 7 delphij K 8 svn:date V 27 2010-03-13T00:15:06.084547Z K 7 svn:log V 95 Correct bugs fixed but not merged from my own tree. Pointy hat to: delphij MFC after: 1 month END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T01:09:14.650754Z K 7 svn:log V 466 o) Do not unconditionally update the last known link status in the softc. If report_link isn't set, when octeon_rgmx_config_speed is called the first time it will tell the driver (essentially) that we have already marked the interface up. Likewise, don't change media speed and duplex if only the link status is at issue. o) Remove manual changing of link state and let octeon_rgmx_config_speed do the heavy lifting. Sponsored by: Packet Forensics END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T01:18:33.964679Z K 7 svn:log V 36 Report half duplex status properly. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T02:09:15.219715Z K 7 svn:log V 27 Fix spelling of noreorder. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T02:10:20.276342Z K 7 svn:log V 27 Fix spelling of noreorder. END K 10 svn:author V 3 imp K 8 svn:date V 27 2010-03-13T03:06:48.776292Z K 7 svn:log V 109 Make this conform to the other top-level Makefile subdir listings with one file per line. Suggested by: bde END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T04:00:36.988093Z K 7 svn:log V 107 Add an implementation, only compile-tested, of makecontext with ABI-correct argument encoding from NetBSD. END K 10 svn:author V 8 jmallett K 8 svn:date V 27 2010-03-13T04:55:47.630848Z K 7 svn:log V 652 o) Use octeon_fpa_alloc_phys in a situation in which we don't need a usable pointer, rather than octeon_fpa_alloc. o) Report half duplex status properly. o) Do not unconditionally update the last known link status in the softc. If report_link isn't set, when octeon_rgmx_config_speed is called the first time it will tell the driver (essentially) that we have already marked the interface up. Likewise, don't change media speed and duplex if only the link status is at issue. [1] o) Remove manual changing of link state and let octeon_rgmx_config_speed do the heavy lifting. [1] Reviewed by: [1] imp Sponsored by: Packet Forensics END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-13T09:21:00.118476Z K 7 svn:log V 356 Remove COMPAT_43TTY from stock kernel configuration files. COMPAT_43TTY enables the sgtty interface. Even though its exposure has only been removed in FreeBSD 8.0, it wasn't used by anything in the base system in FreeBSD 5.x (possibly even 4.x?). On those releases, if your ports/packages are less than two years old, they will prefer termios over sgtty. END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-13T11:06:47.431309Z K 7 svn:log V 24 MFH the Newcons branch. END K 10 svn:author V 6 brucec K 8 svn:date V 27 2010-03-13T11:08:57.544114Z K 7 svn:log V 153 Free the memory allocated via strdup. PR: bin/113881 Submitted by: Alexander Drozdov dzal_mail mtu-net.ru Approved by: rrs (mentor) MFC after: 1 week END K 10 svn:author V 6 brucec K 8 svn:date V 27 2010-03-13T11:17:39.882341Z K 7 svn:log V 238 Change the 'amt' parameter in format_k2 from int to unsigned long long to match the values passed in and prevent the SIZE field being corrupted when more than 2TB is allocated. PR: bin/129706 Approved by: rrs (mentor) MFC after: 1 week END K 10 svn:author V 7 rnoland K 8 svn:date V 27 2010-03-13T11:51:18.519509Z K 7 svn:log V 184 Add support for Intel Pineview chips, aka IGD Slightly modified version of the submitted patch. PR: 143427 Submitted by: Mamoru Sumida MFC after: 3 days END K 10 svn:author V 2 jh K 8 svn:date V 27 2010-03-13T12:02:44.300569Z K 7 svn:log V 219 Use an unique directory name instead of hardcoded /tmp/.diskless. A malicious user could create a file named /tmp/.diskless and cause the script to misbehave. PR: conf/141258 Reported by: Jon Passki MFC after: 1 week END K 10 svn:author V 6 jkoshy K 8 svn:date V 27 2010-03-13T15:10:07.050939Z K 7 svn:log V 86 Correct a typo. Submitted by: Rahul Saxena END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-13T16:15:15.427516Z K 7 svn:log V 258 Build kernel modules as relocatable objects. They are much more easy to work with in a NUMA environment. This commit effectively breaks module support. The breakage is fixed when the loader and kernel know how to load and relocate the object files on ia64. END K 10 svn:author V 5 ticso K 8 svn:date V 27 2010-03-13T16:37:17.722602Z K 7 svn:log V 47 MFC 204462,204463,204476: fix multicast hashes END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-13T18:06:17.640380Z K 7 svn:log V 25 Small style(9) cleanups. END K 10 svn:author V 7 rnoland K 8 svn:date V 27 2010-03-13T18:14:51.158457Z K 7 svn:log V 97 Fix build after typo. Reported by: Sergey V. Dyatko MFC after: 3 days END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-13T18:34:19.294033Z K 7 svn:log V 1077 Go out on a limb and build the kernel proper as a relocatable as well. This has some bigger consequences, but again it's much more easy to work with relocatables in a NUMA environment. In particular, the effects of this change are: 1. We can't load kernels any more -- not a big deal for Altix, because we couldn't load kernels in the first place. We needed to load the kernel at a fixed physical address (= 64M) and on Altix there's no physical memory at that address. 2. The kernel object size increased by ~39%. The debug kernel object file increased by about ~99%. Lastly, the symbols object file saw about a 150% increase. Whether this is acceptable or not is left for later to decide -- after we can compare the memory footprint. 3. The loader can't check for ET_EXEC to determine if the object is the kernel proper or not. Consequently, we have no entry point. It feels acceptable to expect the entry point to be called "__start" or something along those lines. The kernel proper is the object file that contains that symbol. END K 10 svn:author V 5 simon K 8 svn:date V 27 2010-03-13T19:22:41.115162Z K 7 svn:log V 387 Merge OpenSSL 0.9.8m into head. This also "reverts" some FreeBSD local changes so we should now be back to using entirely stock OpenSSL. The local changes were simple $FreeBSD$ lines additions, which were required in the CVS days, and the patch for FreeBSD-SA-09:15.ssl which has been superseded with OpenSSL 0.9.8m's RFC5746 'TLS renegotiation extension' support. MFC after: 3 weeks END K 10 svn:author V 5 simon K 8 svn:date V 27 2010-03-13T19:30:29.970443Z K 7 svn:log V 64 Regenerate manual pages for OpenSSL 0.9.8m. MFC after: 3 weeks END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-13T20:43:11.168166Z K 7 svn:log V 210 sh: Fix longjmp clobber warnings in parser.c. Make parsebackq a function instead of an emulated nested function. This puts the setjmp usage in a smaller function where it is easier to avoid bad optimizations. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-13T21:28:14.477961Z K 7 svn:log V 97 Provide a mechanism for handling kernel SLB spills that does not fill me with terror to imagine. END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-13T21:41:52.215076Z K 7 svn:log V 154 Don't bottleneck on acquiring the stream locks - this avoids a massive drop off in throughput with large numbers of simultaneous reads MFC after: 7 days END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-13T21:52:21.962350Z K 7 svn:log V 34 fix compilation under ZIO_USE_UMA END K 10 svn:author V 5 kmacy K 8 svn:date V 27 2010-03-13T21:53:48.938169Z K 7 svn:log V 121 make UMA the default allocator for ZFS buffers - this avoids a great deal of contention in kmem_alloc MFC after: 7 days END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-13T22:30:37.126964Z K 7 svn:log V 44 Missed a file. Reported by: Andreas Tobler END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-13T22:30:52.993983Z K 7 svn:log V 349 sh: Add test for assignment errors (e.g. trying to change a readonly var). We currently ignore readonly status for assignments before regular builtins and external programs (these assignments are not persistent anyway), so just check that the readonly variable really is not changed. The test depends on the command builtin changes for 'command :'. END K 10 svn:author V 5 simon K 8 svn:date V 27 2010-03-13T22:37:07.053551Z K 7 svn:log V 98 Readd $FreeBSD$ to the OpenSSL config file as that's useful for mergemaster. Suggested by: dougb END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-13T22:53:17.790864Z K 7 svn:log V 527 sh: Do not abort on a redirection error if there is no command word. Although simple commands without a command word (only assignments and/or redirections) are much like special builtins, POSIX and most shells seem to agree that redirection errors should not abort the shell in this case. Of course, the assignments persist and assignment errors are fatal. To get the old behaviour portably, use the ':' special builtin. To get the new behaviour portably, given that there are no assignments, use the 'true' regular builtin. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-14T00:32:18.739765Z K 7 svn:log V 160 Don't spill existing SLB entries. We can't handle this yet, and it results in randomly demapping bits of the kernel. Which is bad. Reported by: Andreas Tobler END K 10 svn:author V 7 weongyo K 8 svn:date V 27 2010-03-14T01:57:32.110999Z K 7 svn:log V 188 fixes a broken software beacon miss handler. There is a race to check vap->iv_bmiss_count == 0 in ieee80211_swbmiss because iv_swbmiss_task is enqueued by taskqueue. Reviewed by: rpaulo END K 10 svn:author V 7 weongyo K 8 svn:date V 27 2010-03-14T02:13:11.104333Z K 7 svn:log V 51 enables S/W beacon miss handler. Reported by: imp END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-14T02:45:45.361080Z K 7 svn:log V 28 Pull-in libstand as well... END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-14T02:50:24.803814Z K 7 svn:log V 88 Include load_elf64_obj.c on ia64. This adds the fundamental support for ET_REL objects. END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-14T02:54:36.688327Z K 7 svn:log V 50 Add link_elf_obj.c for support of ET_REL objects. END K 10 svn:author V 5 dougb K 8 svn:date V 27 2010-03-14T05:22:46.993850Z K 7 svn:log V 268 Make it more clear in the docs that -a is not compatible with -iFU, and enforce this in the code. Apparently a lot of users mistakenly combine -a with these flags and are then mystified that no changes were made. While I'm here, fix a trailing space in mergemaster.8 END K 10 svn:author V 2 ed K 8 svn:date V 27 2010-03-14T10:18:58.141477Z K 7 svn:log V 710 Trim down libcompat by removing . Erwin ran an exp-run with libcompat and removed. It turns out the regexp library is almost entirely unused. In fact, it looks like it is sometimes used by accident. Because these function names clash with libc's , some application use both and libcompat, which means they link against the wrong regex library. This commit removes the regexp library and reimplements re_comp() and re_exec() using . It seems the grammar of the regular expressions accepted by these functions is similar to POSIX EREs. After this commit, 1 low-profile port will be broken, but the maintainer already has a patch for it sitting in his mailbox. END K 10 svn:author V 5 edwin K 8 svn:date V 27 2010-03-14T10:24:03.416572Z K 7 svn:log V 314 Create regression tests for ncal: - A couple of tests to check if the layout of the generated calenders is correct. - A couple of tests to see if impossible combinations for -3, -A, -m, -y etc properly abort. - A couple of test to confirm that the order of -A, -B, -3 etc give the right number of months. END K 10 svn:author V 3 kib K 8 svn:date V 27 2010-03-14T13:04:09.353811Z K 7 svn:log V 51 Make it compile on LP64 arches. MFC after: 3 days END K 10 svn:author V 3 kib K 8 svn:date V 27 2010-03-14T13:05:48.860426Z K 7 svn:log V 120 Adjust style, fix typo, do not print errno because it is meaningless there. This test fails for now. MFC after: 3 days END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-14T13:07:40.689917Z K 7 svn:log V 584 MFC r204410: Include terminated threads in ps's process cpu time field. When a kinfo_proc is filled, first fill_kinfo_proc_only() fills in ki_runtime using p->p_rux.rux_runtime (all cpu time used by the process including terminated threads). If information for a specific thread is requested, fill_kinfo_thread() then overwrites this with the thread's td->td_runtime (good). If not, fill_kinfo_aggregate() overwrote it with the sum of all threads' td->td_runtime which does not include terminated threads. This affects ps(1)'s TIME field, not its %CPU field nor anything in top(1). END K 10 svn:author V 3 kib K 8 svn:date V 27 2010-03-14T13:29:09.408684Z K 7 svn:log V 74 Remove dot at the end of errx() message. Noted by: bde MFC after: 3 days END K 10 svn:author V 5 gavin K 8 svn:date V 27 2010-03-14T13:32:40.717044Z K 7 svn:log V 103 Merge r183385 from head (by cognet) Advertise bit 26 as sse2. PR: kern/130724 Approved by: cognet END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-14T13:51:12.609670Z K 7 svn:log V 67 sh: Add test for redirection error on subshell (should not abort). END K 10 svn:author V 6 jilles K 8 svn:date V 27 2010-03-14T14:24:35.341099Z K 7 svn:log V 437 sh: Do not abort on a redirection error on a compound command. Redirection errors on subshells already did not abort the shell because the redirection is executed in the subshell. Other shells seem to agree that these redirection errors should not abort the shell. Also ensure that the redirections will be cleaned up properly in cases like command eval '{ shift x; } 2>/dev/null' Example: { echo bad; } MFC after: 1 week END K 10 svn:author V 5 gavin K 8 svn:date V 27 2010-03-14T16:03:36.099832Z K 7 svn:log V 269 Merge r202461 from head: Implement an "-x" option to cp(1), for compatibility with Linux and feature parity with du(1) and similar: When set, cp(1) will not traverse mount points. PR: bin/88056 Initial patch by: Graham J Lee END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2010-03-14T18:59:11.983286Z K 7 svn:log V 437 Abstract out initialization of most aspects of struct inpcbinfo from their calling contexts in {IP divert, raw IP sockets, TCP, UDP} and create new helper functions: in_pcbinfo_init() and in_pcbinfo_destroy() to do this work in a central spot. As inpcbinfo becomes more complex due to ongoing work to add connection groups, this will reduce code duplication. MFC after: 1 month Reviewed by: bz Sponsored by: Juniper Networks END K 10 svn:author V 3 imp K 8 svn:date V 27 2010-03-14T19:04:42.653414Z K 7 svn:log V 52 Go ahead and add USB support to the generic config. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-14T21:14:25.430401Z K 7 svn:log V 200 Use a large-page-mapped direct mapped region on powerpc64 if possible. This is a fairly sweeping change and has doubtless broken many things, but this should be the last major change to the MMU code. END K 10 svn:author V 7 attilio K 8 svn:date V 27 2010-03-14T22:38:18.856107Z K 7 svn:log V 531 Checkin a facility for specifying a passthrough FIB from userland. arcconf tool by Adaptec already seems to use for identifying the Serial Number of the devices. Some simple things (like FIB setup and bound checks) are retrieved from the Adaptec's driver, but this implementation is quite different because it does use the normal buffer dmat area for loading segments and not a special one (like the Adaptec's one does). Sponsored by: Sandvine Incorporated Discussed with: emaste, scottl Reviewed by: emaste, scottl MFC: 2 weeks END K 10 svn:author V 7 yongari K 8 svn:date V 27 2010-03-14T23:23:57.042985Z K 7 svn:log V 481 It seems PCI_OUR_REG_[1-5] registers are not mapped on PCI configuration space on Yukon Ultra(88E8056) such that accesses to these registers were NOPs which in turn make msk(4) instable on this controller. Use indirect access method to access PCI_OUR_REG_[1-5] registers. This should fix a long standing instability bug which prevented msk(4) working on Yukon Ultra. Special thanks to koitsu who gave me remote access to his system. PR: kern/114631, kern/116853 MFC after: 1 week END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-14T23:52:30.372142Z K 7 svn:log V 141 Add ia64_elf_obj to file_formats[] and unify the SKI and EFI support by moving the definition from the separate conf.c to the common exec.c. END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-15T00:27:40.544941Z K 7 svn:log V 263 Fix two small bugs. The PowerPC 970 does not support non-coherent memory access, and reflects this by autonomously writing LPTE_M into PTE entries. As such, we should not panic if LPTE_M changes by itself. While here, fix a harmless typo in moea64_sync_icache(). END K 10 svn:author V 10 nwhitehorn K 8 svn:date V 27 2010-03-15T00:29:15.407213Z K 7 svn:log V 157 Set memory access as guarded on large page direct maps that edge into physical memory holes, and optimize moea64_kextract(), thereby avoiding a LOR at boot. END K 10 svn:author V 3 phk K 8 svn:date V 27 2010-03-15T08:58:35.138650Z K 7 svn:log V 103 Comment a fine point, so it does not get lost when people borrow code from FreeBSD for other purposes. END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T09:40:23.252455Z K 7 svn:log V 42 missing changes to let the kernel compile END K 10 svn:author V 7 attilio K 8 svn:date V 27 2010-03-15T14:20:16.742338Z K 7 svn:log V 168 Make the code more readable and compiling on 64-bits arch different than amd64. Sponsored by: Sandvine Incorporated Submitted by: emaste MFC: 2 weeks X-MFC: r205160 END K 10 svn:author V 5 bruno K 8 svn:date V 27 2010-03-15T15:33:32.988464Z K 7 svn:log V 132 MFC r204519: Deliver siginfo when signal is generated by thr_kill(2) (SI_USER with properly filled si_uid and si_pid). PR: 141956 END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T15:43:35.833481Z K 7 svn:log V 395 Implement "lookup dscp N" which does a lookup of the DSCP (top 6 bits of ip->ip_tos) in a table. This can be useful to direct traffic to different pipes/queues according to the DSCP of the packet, as follows: ipfw add 100 queue tablearg lookup dscp 3 // table 3 maps dscp->queue This change is a no-op (but harmless) until the two-line kernel side is committed, which will happen shortly. END K 10 svn:author V 5 joerg K 8 svn:date V 27 2010-03-15T16:06:52.658774Z K 7 svn:log V 31 then -> than MFC after: 1 day END K 10 svn:author V 2 jh K 8 svn:date V 27 2010-03-15T16:39:52.227482Z K 7 svn:log V 65 MFC r204872: Pass the correct size to memset(). PR: bin/128094 END K 10 svn:author V 6 marcel K 8 svn:date V 27 2010-03-15T16:53:09.292440Z K 7 svn:log V 374 Have cpu_throw() loop on blocked_lock as well. This bug has existed a long time and has gone unnoticed just as long, because I kept using sched_4bsd (due to sched_ule not working with preemption), but GENERIC had sched_ule by default -- including SMP. While here, remove unused inclusion of , remove totally bogus inclusion of . END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T17:14:27.254910Z K 7 svn:log V 1041 + implement (two lines) the kernel side of 'lookup dscp N' to use the dscp as a search key in table lookups; + (re)implement a sysctl variable to control the expire frequency of pipes and queues when they become empty; + add 'queue number' as optional part of the flow_id. This can be enabled with the command queue X config mask queue ... and makes it possible to support priority-based schedulers, where packets should be grouped according to the priority and not some fields in the 5-tuple. This is implemented as follows: - redefine a field in the ipfw_flow_id (in sys/netinet/ip_fw.h) but without changing the size or shape of the structure, so there are no ABI changes. On passing, also document how other fields are used, and remove some useless assignments in ip_fw2.c - implement small changes in the userland code to set/read the field; - revise the functions in ip_dummynet.c to manipulate masks so they also handle the additional field; There are no ABI changes in this commit. END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T17:52:55.850242Z K 7 svn:log V 28 wrong name for this file... END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T17:54:01.451745Z K 7 svn:log V 21 and wrong place too! END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T17:54:53.811842Z K 7 svn:log V 9 MFStable END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T17:55:46.560438Z K 7 svn:log V 9 MFStable END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T18:09:21.586488Z K 7 svn:log V 73 small fixes to estimate the buffer size when requesting all pipes/flows. END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T18:20:51.609754Z K 7 svn:log V 224 print correctly commands of the form ipfw add 100 allow ip from { 1.2.3.4 or 5.6.7.8 } (note that the above example could be better written as ipfw add 100 allow dst-ip 1.2.3.4,5.6.7.8 Submitted by: Riccardo Panicucci END K 10 svn:author V 4 jkim K 8 svn:date V 27 2010-03-15T18:22:19.980022Z K 7 svn:log V 47 MFC: r204235 Fix FBIO_ADPINFO ioctl on amd64. END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T18:26:09.866614Z K 7 svn:log V 62 accept lower case m as a synonym for Mega (bit/s or bytes/s). END K 10 svn:author V 5 luigi K 8 svn:date V 27 2010-03-15T18:27:51.351114Z K 7 svn:log V 63 almost completed merge of ipfw3 from HEAD into my working tree END K 10 svn:author V 6 marius K 8 svn:date V 27 2010-03-15T18:32:57.711032Z K 7 svn:log V 802 MFC: r204974, r205002 - The OPSZ macro actually only does the right thing for int32 and int64 operands but not for double and extended double ones. Instead of trying to fix the macro just nuke it and unroll the loops in the correct way though as extended double operands turn out to be the only special case. - For FxTO{s,d,q} the source operand is int64 so rs2 has to be re-decoded after setting type accordingly as it's generally decoded using the low 2 bits as the type, which are 0 for these three instructions. - Similarly, in case of F{s,d,q}TOx the target is int64 so rd has to be re-decoded using not only the operand mask appropriate for int64 but also the correct register number encoding. - Use const where appropriate. - Wrap long lines. Submitted by: Peter Jeremy (partly) END K 10 svn:author V 6 marius K 8 svn:date V 27 2010-03-15T18:33:23.721859Z K 7 svn:log V 802 MFC: r204974, r205002 - The OPSZ macro actually only does the right thing for int32 and int64 operands but not for double and extended double ones. Instead of trying to fix the macro just nuke it and unroll the loops in the correct way though as extended double operands turn out to be the only special case. - For FxTO{s,d,q} the source operand is int64 so rs2 has to be re-decoded after setting type accordingly as it's generally decoded using the low 2 bits as the type, which are 0 for these three instructions. - Similarly, in case of F{s,d,q}TOx the target is int64 so rd has to be re-decoded using not only the operand mask appropriate for int64 but also the correct register number encoding. - Use const where appropriate. - Wrap long lines. Submitted by: Peter Jeremy (partly) END