ƒ¯K146574 208 256 157 138 175 123 198 242 237 160 138 152 339 173 199 238 126 134 224 145 466 139 125 362 331 199 211 159 176 278 303 128 161 481 159 190 218 239 189 239 239 114 166 243 206 122 152 429 193 127 160 118 365 200 407 134 429 295 209 484 151 309 300 235 125 118 168 148 325 313 213 129 110 149 254 164 121 232 202 114 111 118 428 166 125 147 160 121 252 538 156 338 148 194 197 139 201 249 134 245 783 143 323 158 293 120 701 218 241 287 229 368 141 351 283 1637 252 210 253 163 198 144 154 308 154 602 178 124 292 206 159 182 232 116 884 123 267 186 151 148 165 137 186 121 129 366 527 136 169 325 183 129 164 219 181 193 151 268 742 189 361 268 155 1027 365 241 270 833 263 143 412 157 175 218 4043 402 112 511 156 188 154 159 225 188 162 346 323 205 236 199 163 269 402 204 286 153 237 178 152 124 153 237 153 237 178 342 144 135 364 K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T15:41:34.000000Z K 7 svn:log V 161 Factor out ProcExec() into its own file and rename it to Proc_Exec() for consistency with the rest of make. Obtained from: DragonFlyBSD (except for the rename) END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-24T15:42:42.000000Z K 7 svn:log V 60 3Com 3C2000-T reported as working. Submitted by: anonymous END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-24T15:44:02.000000Z K 7 svn:log V 41 MFC: 3Com 3C2000-T reported as working. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T15:45:11.000000Z K 7 svn:log V 81 Fix a 64-bit warning by casting an int64_t to intmax_t and printing it with %jd. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T15:46:50.000000Z K 7 svn:log V 29 Bump the warning level to 6. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T15:48:10.000000Z K 7 svn:log V 103 Make shell.o dependend on the Makefile just in case someone changes the default shell in the Makefile. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T15:58:35.000000Z K 7 svn:log V 147 Get rid of the third argument to Var_Value() the pointer it pointed to has always been set to NULL for some time now. Obtained from: DragonFlyBSD END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-24T16:05:51.000000Z K 7 svn:log V 142 The caller of Var_Value() should not change the variable value. Make this clear by constifying the return value. Obtained from: DragonFlyBSD END K 10 svn:author V 6 damien K 8 svn:date V 27 2005-05-24T16:48:07.000000Z K 7 svn:log V 65 Add new ral(4) and ural(4) drivers. Approved by: silby (mentor) END K 10 svn:author V 2 ps K 8 svn:date V 27 2005-05-24T17:52:14.000000Z K 7 svn:log V 47 Copyout to userland if kern_sigaction succeeds END K 10 svn:author V 2 ps K 8 svn:date V 27 2005-05-24T17:53:27.000000Z K 7 svn:log V 61 MFC: rev 1.32 copyout to userland if kern_sigaction succeeds END K 10 svn:author V 3 mux K 8 svn:date V 27 2005-05-24T18:31:34.000000Z K 7 svn:log V 246 - Move the REISERFS option in the MD NOTES file for i386, it doesn't exist on other architectures yet. - While I'm here, fix the formatting of the options line. The keyword "options" should be followed by a space and then a tab, not 2 tabs. END K 10 svn:author V 8 schweikh K 8 svn:date V 27 2005-05-24T18:54:07.000000Z K 7 svn:log V 76 Update (reduce) the list of committers with MAINTAINER= lines in Makefiles. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T20:42:08.000000Z K 7 svn:log V 103 Fix one more misuse of u_long when uint32_t is actually meant. Submitted by: oliver MFC after: 3 days END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:12:49.000000Z K 7 svn:log V 142 Don't set the tuner type to TUNER_MT2032 if BKTR_OVERRIDE_TUNER is defined or if hw.bt848.tuner has been set, so that we can force the tuner. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:24:40.000000Z K 7 svn:log V 31 Use -march=armv5te for Xscale. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:25:32.000000Z K 7 svn:log V 39 arm defaults to unsigned char as well. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:42:31.000000Z K 7 svn:log V 128 Make sure we clean the RAS start address once we're done. This fixes the random segfaults which occurs at high interrupts rate. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:43:16.000000Z K 7 svn:log V 50 Asm version of bswap16(). Obtained from: NetBSD END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-05-24T21:43:49.000000Z K 7 svn:log V 370 Since the BAT mapping was removed a long time back, it can't be assumed that modules are contiguous in memory (they're not) so don't blindly __syncicache start/end. In fact, don't bother syncing the icache for modules since the kernel will do it after fixing up relocations. This fixes the trap when loading modules at boot time. Reported by: orlando at break dot net END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:44:34.000000Z K 7 svn:log V 44 Use asm versions of in_cksum() and friends. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:45:26.000000Z K 7 svn:log V 30 Use a more sane value for HZ. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T21:47:10.000000Z K 7 svn:log V 266 Write back affected pages in pmap_qremove() as well. This removes the need to change the DACR when switching to a kernel thread, thus making userland thread => kernel thread => same userland thread switch cheaper by totally avoiding data cache and TLB invalidation. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T22:10:35.000000Z K 7 svn:log V 235 - Try to avoid calling malloc() in bus_dmamap_create() and bus_dmamem_alloc() for the dmamap by using static dmamaps. - Don't do anything for BUS_DMASYNC_PREREAD and BUS_DMASYNC_POSTWRITE in bus_dmamap_sync(), it's not needed anymore. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T22:21:44.000000Z K 7 svn:log V 103 Don't set the default of kern.fallback_elf_brand to FreeBSD for arm, as binutils now do the job for us END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T23:05:26.000000Z K 7 svn:log V 115 Correctly setup the UND stack in cpu_set_upcall(), and the trapframe in cpu_thread_setup(), as done in cpu_fork(). END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T23:06:02.000000Z K 7 svn:log V 64 We need to decrease p->p_lock after vm_fault() has been called. END K 10 svn:author V 3 mux K 8 svn:date V 27 2005-05-24T23:06:10.000000Z K 7 svn:log V 84 Only build mount_reiserfs(8) on i386, it exists only there for now. Spotted by: ru END K 10 svn:author V 4 csjp K 8 svn:date V 27 2005-05-24T23:42:09.000000Z K 7 svn:log V 184 If sysctlbyname fails for kernel related reasons, tag the errno string to the end of the error message. I think we used errx() there when we really wanted an err(). MFC after: 1 week END K 10 svn:author V 3 jmg K 8 svn:date V 27 2005-05-24T23:42:50.000000Z K 7 svn:log V 210 make stat return an zero'd struct, and be a FIFO again... This is only to fix libc_r since it requires stat to close fd's, and so commented in the code... PR: threads/75795 Reviewed by: ps MFC after: 1 week END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T23:55:09.000000Z K 7 svn:log V 33 Remove kcopy(), we don't use it. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-24T23:57:22.000000Z K 7 svn:log V 66 MFp4: Setup arm9 to write back by default. Obtained from: NetBSD END K 10 svn:author V 6 marcel K 8 svn:date V 27 2005-05-25T07:25:12.000000Z K 7 svn:log V 385 For ISA DMA maps, the chipsets scatter/gather feature is used. As such, the segments pointer in the DMA tag will always be NULL. In bus_dmamap_load(), temporarily point the segments pointer in the DMA tag to a local variable so that we don't dereference a NULL pointer. Reset the segments pointer to NULL after calling the callback function with it. PR: alpha/30486 MFC after: 1 week END K 10 svn:author V 2 ru K 8 svn:date V 27 2005-05-25T07:31:05.000000Z K 7 svn:log V 68 Fix order (I introduced some time ago) broken in previous revision. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T08:46:31.000000Z K 7 svn:log V 96 Document why there is a dependency of certain object files from the Makefile. Requested by: ru END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T12:00:03.000000Z K 7 svn:log V 123 Apply a vendor patch: silence gcc on arm (alignent when casting pointer). Submitted by: Olivier Houchard END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T12:00:03.000000Z K 7 svn:log V 144 This commit was generated by cvs2svn to compensate for changes in r146609, which included commits to RCS files with non-trunk default branches. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T12:03:26.000000Z K 7 svn:log V 95 Apply vendor patch: include the header files from the canonical place. Explained to me by: ru END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T12:03:26.000000Z K 7 svn:log V 144 This commit was generated by cvs2svn to compensate for changes in r146611, which included commits to RCS files with non-trunk default branches. END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T12:04:44.000000Z K 7 svn:log V 144 No need to specify the include subdirectory. During buildworld the correct files are included from a temp. include directory. Explained by: ru END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-05-25T12:32:06.000000Z K 7 svn:log V 21 MFi386: Add ReiserFS END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-05-25T12:57:02.000000Z K 7 svn:log V 70 MFC 1.65(phk): Do not attach MBR on top of an MBR. Approved by: phk END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-25T13:10:04.000000Z K 7 svn:log V 150 - Call root_mount_rel() when provider IS created, not earlier. This should close the race observed by Daniel Eriksson. - Remove redundant wakeup(). END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T13:33:58.000000Z K 7 svn:log V 111 Add a missing comma which prevents compilation with debugging enabled. Spotted by: Donatas END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-25T13:44:55.000000Z K 7 svn:log V 27 Increase the refresh rate. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-25T13:46:32.000000Z K 7 svn:log V 57 Remove bits specific to CPUs we won't support (< armv4). END K 10 svn:author V 6 peadar K 8 svn:date V 27 2005-05-25T13:52:03.000000Z K 7 svn:log V 333 Separate out address-detaching part of if_detach into if_purgeaddrs, so if_tap doesn't need to rely on locally-rolled code to do same. The observable symptom of if_tap's bzero'ing the address details was a crash in "ifconfig tap0" after an if_tap device was closed. Reported By: Matti Saarinen (mjsaarin at cc dot helsinki dot fi) END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-25T15:35:31.000000Z K 7 svn:log V 96 Document support for some Dell PERC adapters. Submitted by: Muthu_T@Dell.com MFC after: 3 days END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-25T15:41:28.000000Z K 7 svn:log V 30 MFC: soundcard -> sound card END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-25T15:44:10.000000Z K 7 svn:log V 63 MFC: rev. 1.17 + 1.18 Document support for the 5714 and 5789. END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-25T15:54:17.000000Z K 7 svn:log V 26 After provider creation!! END K 10 svn:author V 5 harti K 8 svn:date V 27 2005-05-25T16:06:14.000000Z K 7 svn:log V 270 Under certain conditions the condition parser would go one past end of the string. Until now this caused no harm, because the buffer code used to tack two NULs onto buffers. With the new, soon to come, parsing code this isn't the case anymore in all cases, so fix this. END K 10 svn:author V 6 peadar K 8 svn:date V 27 2005-05-25T16:27:43.000000Z K 7 svn:log V 104 MFC 1.15: lockmgr(...,LK_DRAIN,...) requires a balancing LK_RELEASE. Remove an unused variable. END K 10 svn:author V 8 sheldonh K 8 svn:date V 27 2005-05-25T16:30:43.000000Z K 7 svn:log V 309 Release maintainership. More ambitious minds have plans for the ntp docs. Last I heard, Harlan Stenn was considering using FreeBSD's pages as a starting point for the ISC NTP distribution's own pages. If that happens, everyone wins and these can go away, to be replaced by imported versions in contrib/ntp. END K 10 svn:author V 8 sheldonh K 8 svn:date V 27 2005-05-25T16:32:14.000000Z K 7 svn:log V 37 MFC rev 1.11: release maintainership END K 10 svn:author V 8 schweikh K 8 svn:date V 27 2005-05-25T17:37:57.000000Z K 7 svn:log V 331 Looking just at Makefiles was slightly too narrow to catch all "inofficial" declarations of maintainership. Grep all plain files, and insert the actual command the list was generated with, so future updates avoid that pitfall. Removed sheldonh@ who just released his maintainership of ntp/doc after I informed him of this effort. END K 10 svn:author V 2 ps K 8 svn:date V 27 2005-05-25T17:55:27.000000Z K 7 svn:log V 203 This is conform with the terminology in M.Mathis and J.Mahdavi, "Forward Acknowledgement: Refining TCP Congestion Control" SIGCOMM'96, August 1996. Submitted by: Noritoshi Demizu, Raja Mukerji END K 10 svn:author V 8 netchild K 8 svn:date V 27 2005-05-25T21:03:13.000000Z K 7 svn:log V 111 Add a note how to use nextboot(8) to test a kernel only once. Approved by: mentor (joerg) Discussed with: imp END K 10 svn:author V 4 csjp K 8 svn:date V 27 2005-05-26T00:31:38.000000Z K 7 svn:log V 390 MFC revision 1.144 date: 2005/05/19 21:09:44; author: csjp; state: Exp; lines: +2 -0 Add cases for ENOBUFS and ENETDOWN. We need to make sure that we catch all transient errors. This fixes situations where transient error conditions such as network interfaces losing carrier signals or the system running out of mbufs would result in the permanent removal of forwarding syslog messages. END K 10 svn:author V 7 thompsa K 8 svn:date V 27 2005-05-26T03:21:16.000000Z K 7 svn:log V 55 Add myself to the calendar. Approved by: bms (mentor) END K 10 svn:author V 5 adamw K 8 svn:date V 27 2005-05-26T04:21:53.000000Z K 7 svn:log V 214 MFC rmuser.sh 1.9: Fix a particularly egregious grammar error. It's not that I mind the implication that a user's name owns a punctuation mark, but I think life is simpler this way. Approved by: ssouhlal and wes END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-26T06:50:00.000000Z K 7 svn:log V 203 Plug mbuf leak, that I have introduced in 1.85. Also restore important comment from if_ethersubr.c:1.178. While here adjust formatting, to make code more readable. Reported by: Alexey Kamyshev, rwatson END K 10 svn:author V 4 ache K 8 svn:date V 27 2005-05-26T06:57:57.000000Z K 7 svn:log V 141 Fix: printed output flags (onocr) and (onlret) same as oxtabs PR: 81256 Submitted by: Arseny Nasokin END K 10 svn:author V 3 dfr K 8 svn:date V 27 2005-05-26T09:32:41.000000Z K 7 svn:log V 33 MFC: r1.4 - align stack for SSE. END K 10 svn:author V 3 dfr K 8 svn:date V 27 2005-05-26T09:33:28.000000Z K 7 svn:log V 26 MFC: Align stack for SSE. END K 10 svn:author V 4 ache K 8 svn:date V 27 2005-05-26T09:42:18.000000Z K 7 svn:log V 75 MFC: 1.20 printed output flags (onocr) and (onlret) PR: 81256 END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-26T10:48:35.000000Z K 7 svn:log V 51 New release note: Read-only support for ReiserFS. END K 10 svn:author V 7 roberto K 8 svn:date V 27 2005-05-26T10:57:03.000000Z K 7 svn:log V 228 - Add further functionality to check for invalid characters - Remove keyword 'continue' for more indepth error reporting on each line - WARNS 6 Clean Submitted by: Liam J. Foy MFC after: 1 week END K 10 svn:author V 4 bmah K 8 svn:date V 27 2005-05-26T14:38:16.000000Z K 7 svn:log V 219 New release notes: autoboot_delay -1 (+MFC), NgATM 1.2, texinfo 4.8, pkg_version -I. Modified release notes: bsnmp 1.9, new manpages (devfs.conf(5), devfs.rules(5), pthread_atfork(3)). Other MFCs noted: sed(1) -l. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2005-05-26T14:41:20.000000Z K 7 svn:log V 119 Move sed(1) -l release note to the right section. Don't know what I was thinking (was I thinking?) when I wrote this. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2005-05-26T14:43:26.000000Z K 7 svn:log V 36 MFC: autoboot_delay -1, sed(1) -l. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2005-05-26T14:44:41.000000Z K 7 svn:log V 17 MFC: sed(1) -l. END K 10 svn:author V 6 archie K 8 svn:date V 27 2005-05-26T14:59:38.000000Z K 7 svn:log V 54 MFC: (rev. 1.10) Correct L2TP header offset handling. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-26T15:01:13.000000Z K 7 svn:log V 158 Don't call vm_page_dirty() in pmap_nuke_pv(), it's not the place to do so, and it leads to funny things, such as pmap_remove_all() marking the page as dirty. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-26T15:02:47.000000Z K 7 svn:log V 69 Don't enable interrupts in the dispatcher, there's no need to do so. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-26T16:05:22.000000Z K 7 svn:log V 26 s/_KLD_MODULE/KLD_MODULE/ END K 10 svn:author V 8 keramida K 8 svn:date V 27 2005-05-26T17:54:16.000000Z K 7 svn:log V 134 Interlink systat(1), iostat(8) and vmstat(8) through their SEE ALSO sections, so that users of one can learn about the others easily. END K 10 svn:author V 7 stefanf K 8 svn:date V 27 2005-05-26T19:51:59.000000Z K 7 svn:log V 105 MFC 1.3: The header glue.h should provide just a declaration for the variable __sglue, not a definition. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T20:06:57.000000Z K 7 svn:log V 19 decode mkdir args. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T20:47:29.000000Z K 7 svn:log V 16 MFC: 1.46, 1.47 END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T20:50:46.000000Z K 7 svn:log V 23 MFC: procctl doc xref. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T21:26:00.000000Z K 7 svn:log V 332 MFC: When doing the initial open of the proc via procfs, complain if we are unable to open the /proc/pid/mem file. Otherwise doing a truss on a nonexistant pid makes us return success even though no such process exists. I've been working on this somewhat so I'm moving the parts I'm touching to be as style(9) compliant as I can. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-26T22:14:37.000000Z K 7 svn:log V 71 We are past 4.4BSD - use our new-found stat flags for pipes and fifos. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-26T22:30:12.000000Z K 7 svn:log V 30 Document a couple of gotchas. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T22:49:06.000000Z K 7 svn:log V 52 decode utimes, lutimes, futimes, chflags, lchflags. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-26T22:49:08.000000Z K 7 svn:log V 65 Remove an errno reset that became unnecessary. Noticed by: juli END K 10 svn:author V 6 alfred K 8 svn:date V 27 2005-05-26T22:51:04.000000Z K 7 svn:log V 26 MFC: 1.48 utimes/chflags. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-26T23:01:30.000000Z K 7 svn:log V 156 Baby, we are not in Kansas anymore. Nor are we in 1996 or FreeBSD 2.1. Note that these papers are mostly quite old, and add a pointer to more recent docs. END K 10 svn:author V 9 tackerman K 8 svn:date V 27 2005-05-26T23:32:02.000000Z K 7 svn:log V 439 Changes to update driver with latest Intel driver version 2.1.7 - Changed from using explicit devices id to using descriptive labels. - Added support for 82573 and 82546 Quad adapters. - Corrected support for 82547EI and 82541ER (mac_type was not assigned) - Removed #ifdef DBG_STATS and extraneous code. if_em_hw.c/if_em_hw.h - Added support for 82573 and 82546 Quad adapters. - Brought forward Intel's most current mac and phy changes. END K 10 svn:author V 9 tackerman K 8 svn:date V 27 2005-05-26T23:33:24.000000Z K 7 svn:log V 58 Latest README to correspond to latest Intel version 2.1.7 END K 10 svn:author V 3 mux K 8 svn:date V 27 2005-05-27T00:05:16.000000Z K 7 svn:log V 245 Use clnt_create_timed() instead of clnt_create(). The former has an additional argument that allows us to specify a timeout, like we do for the subsequent clnt_call() calls. Submitted by: Jeremie Le Hen MFC after: 3 weeks END K 10 svn:author V 8 keramida K 8 svn:date V 27 2005-05-27T00:21:12.000000Z K 7 svn:log V 51 Add cross-references to iostat, systat and vmstat. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-27T00:39:11.000000Z K 7 svn:log V 99 Make the example for using xterm as a login manager to match the X of the noughties, not eighties. END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-27T00:40:06.000000Z K 7 svn:log V 101 MFC rev 1.50: Fix one more misuse of u_long when uint32_t is actually meant. Submitted by: oliver END K 10 svn:author V 6 cognet K 8 svn:date V 27 2005-05-27T00:45:39.000000Z K 7 svn:log V 44 Remove pmap_deactivate(), we do not use it. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-27T01:09:42.000000Z K 7 svn:log V 105 Since this is already off the vendor branch: Our kernel is now in /boot/kernel/kernel, not plain /kernel END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-05-27T04:27:22.000000Z K 7 svn:log V 152 Remove sleep queue hack, it is no longer needed with current sleep queue. Actually, it causes process to hang when it is being debugged. PR: gnu/77818 END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-27T04:52:21.000000Z K 7 svn:log V 37 Document support for the 82573 chip. END K 10 svn:author V 6 eivind K 8 svn:date V 27 2005-05-27T06:07:21.000000Z K 7 svn:log V 149 Fixup of last commit: Use the name X instead of XFree86 for the server binary, thus being compatible with both XFree86 and X.org. Noticed by: danfe END K 10 svn:author V 7 stefanf K 8 svn:date V 27 2005-05-27T10:00:22.000000Z K 7 svn:log V 686 Fix long (and long long) to long double, unsigned to long double and unsigned long (and unsigned long long) to long double conversions. - Add a parameter that specifies the position of the sign bit to the _QP_TTOQ macro, previously it always looked at bit 31. Pass a negative number to disable sign inspection for unsigned types. This fixes _Qp_xtoq(), _Qp_uitoq() and _Qp_uxtoq(). - In the functions __fpu_itof() and __fpu_xtof(), look at the sign bit to decide whether we're doing a conversion from an unsigned type. If so, don't negate the mantissa if the integer exceeds the biggest signed number. PR: 55773 Patch by: Stephen Paskaluk (based upon) MFC after: 2 weeks END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T11:58:01.000000Z K 7 svn:log V 47 Merge kernel unit allocator. Approved by: phk END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:02:35.000000Z K 7 svn:log V 226 MFC: o Use subr_unit allocator. This simplifies code much: - Remove get_free_unit(). - Remove SLIST of nodes. - Remove global mutex. o Increase NGD_MAX to 999. o Move ngd_mod_event() up to netgraph methods. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:03:11.000000Z K 7 svn:log V 62 MFC 1.40: Use subr_unit allocator instead of own functions. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:04:13.000000Z K 7 svn:log V 196 MFC: 1.30: Remove extra ifnet pointer from private data. It can be accessed via arpcom. 1.29: Remove ng_connect_t where it is unused. 1.28: Use subr_unit allocator instead of own functions. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:05:24.000000Z K 7 svn:log V 24 Add vnode_if.h to SRCS. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-27T12:25:42.000000Z K 7 svn:log V 604 Back out ipx.h:1.18, which introduced a Linux API compatibility field in the ipx_net data structure. Doing so introduced a stronger alignment requirement for the address structure, which in turn propagated into other dependent data structures, which turns out not to be suported by the available IPX source code. As a result, a number of user space applications, such as IPX routing components, failed to operate correctly. RELENG_5_3 candidate? PRs: 74059, 80266 Pointy hat to: bms Fix by: bde Tested by: Keith White MFC after: 1 week Suffering: great END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:31:29.000000Z K 7 svn:log V 121 MFC: Since there is no way to queue a function call to node, create ng_queue_fn() - a queue version of ng_send_fn(). END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:31:49.000000Z K 7 svn:log V 144 MFC: When used as divert socket we need to decouple stack when node is entered from socket side. Use ng_queue_fn() instead of ng_send_fn(). END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T12:32:29.000000Z K 7 svn:log V 190 MFC: In div_output() explicitly set m->m_nextpkt to NULL. If divert socket is not userland, but ng_ksocket, then m->m_nextpkt may be non-NULL. In this case we would panic in sbappend. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T14:22:30.000000Z K 7 svn:log V 132 Make NETGRAPH_DEBUG a kernel option, so that it can be turned off without hacking source. In collaboration with: ru, julian END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T14:48:35.000000Z K 7 svn:log V 271 MFC: Use uma(9) for allocating netgraph items: - ng_getqblk() simply runs uma_zalloc(). - ng_free_item() simply frees. - ngq_mtx is pushed down under NETGRAPH_DEBUG. - NGQF_FREE is removed. Increase default maxalloc to 512. Reviewed by: julian END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-27T14:51:32.000000Z K 7 svn:log V 45 MFC: Make netgraph ISR and callout MPSAFE. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-05-27T15:29:01.000000Z K 7 svn:log V 258 Fix for 64-bit platforms. random() returns values between 0 and RAND_MAX, and RAND_MAX != LONG_MAX on 64-bit platforms. PR: amd64/81279 Submitted by: Vivek Khera vivek at khera dot org Submitted by: Adriaan de Groot groot at kde dot org MFC after: 1 week END K 10 svn:author V 7 davidxu K 8 svn:date V 27 2005-05-27T15:57:27.000000Z K 7 svn:log V 186 Remove thread_upcall_check, it was used to avoid race bug in earlier day's sleep queue code, today the bug no longer exists. please see 04/25/2004 freebsd-threads@ mailing list archive. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-27T17:16:43.000000Z K 7 svn:log V 1539 In the current world order, each socket has two mutexes: a mutex that protects socket and receive socket buffer state, and a second mutex to protect send socket buffer state. In some places, the mutex shared between the socket and receive socket buffer will be acquired twice, once by each layer, resulting in some inconsistency, but providing the abstraction benefit of being able to more easily separate the two mutexes in the future if desired. When transitioning a socket to the SS_ISDISCONNECTING or SS_ISDISCONNECTED states, grab the socket/receive socket buffer lock once rather than grabbing it as the socket lock, modifying socket state, then grabbing a second time as the receive lock in order to modify the socket buffer state to indicate no further data can be read. This change is believed to close a race between the change in socket state and the change in socket buffer state, which for a remotely initiated close on a UNIX domain socket, resulted in soreceive() returning ENOTCONN rather than an EOF condition. A similar race still exists in the case of send, however, and is harder to fix as the socket and send socket buffer mutexes are not the same, and we would like to avoid holding combinations of socket mutexes over sb_upcall until we've finished clarifying the locking protocol for upcalls. This change has the side affect of reducing the number of mutex operations to initiate disconnect or perform disconnect on a socket by two. PR: 78824 Rerported by: Marc Olzheim MFC after: 2 weeks END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-27T17:52:56.000000Z K 7 svn:log V 155 dd a '-n' option to ministat, which causes it to display only summary statistics, not graph and statistical test output. Useful for automated processing. END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-27T19:02:12.000000Z K 7 svn:log V 117 disable defining NI_WITHSCOPEID. It was obsoleted, and was exist only for backward compatibility since 5.2-RELEASE. END K 10 svn:author V 4 emax K 8 svn:date V 27 2005-05-27T19:11:33.000000Z K 7 svn:log V 159 Do not tread 128-bit UUID as int128. Provide separate macros to get/put 128-bit UUID libsdp(3). Fix 128-bit UUID printing in sdpcontrol(8). MFC after: 3 days END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-27T19:15:46.000000Z K 7 svn:log V 71 Use consistent style in functions I want to modify in the near future. END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-27T19:21:08.000000Z K 7 svn:log V 105 Sync locking in freebsd4_getfsstat() with getfsstat(). Giant is probably also needed in kern_fhstatfs(). END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-27T19:23:48.000000Z K 7 svn:log V 52 Remove (now) unused argument 'td' from cvtstatfs(). END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-27T19:25:39.000000Z K 7 svn:log V 62 Remove (now) unused argument 'td' from bsd_to_linux_statfs(). END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-05-27T19:26:11.000000Z K 7 svn:log V 215 Print out the commands from /boot.config after parsing them so that they output is sent to the correct console(s). PR: kern/66425 Submitted by: Gavin Atkinson gavin at ury dot york dot ac dot uk MFC after: 1 week END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-05-27T19:28:04.000000Z K 7 svn:log V 62 Fix a warning by adding a missing 'const'. MFC after: 1 week END K 10 svn:author V 3 jhb K 8 svn:date V 27 2005-05-27T19:31:00.000000Z K 7 svn:log V 509 - Add support to the loader for multiple consoles. - Teach the i386 and pc98 loaders to honor multiple console requests from their respective boot2 binaries so that the same console(s) are used in both boot2 and the loader. - Since the kernel doesn't support multiple consoles, whichever console is listed first is treated as the "primary" console and is passed to the kernel in the boot_howto flags. PR: kern/66425 Submitted by: Gavin Atkinson gavin at ury dot york dot ac dot uk MFC after: 1 week END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-27T19:49:16.000000Z K 7 svn:log V 86 MFC 1.38: fixed a bug that incorrect IPsec request level may be returned for proto AH END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-27T20:44:57.000000Z K 7 svn:log V 32 reduce cast. MFC after: 1 week END K 10 svn:author V 4 emax K 8 svn:date V 27 2005-05-28T00:48:42.000000Z K 7 svn:log V 198 Move AVM USB Bluetooth-Adapter BlueFritz! from "broken" devices list (where I incorrectly put it initially) to "ignored" devices list (where it should be). Pointy hat goes to me. MFC after: 3 days END K 10 svn:author V 6 brooks K 8 svn:date V 27 2005-05-28T06:11:38.000000Z K 7 svn:log V 110 Update refrenced URL for SNMP list of ifTypes to refer to iana.org instead of a dead location on ftp.isi.edu. END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-28T07:17:51.000000Z K 7 svn:log V 62 MFC: rev. 1.31 Document support for some Dell PERC adapters. END K 10 svn:author V 8 tanimura K 8 svn:date V 27 2005-05-28T07:46:44.000000Z K 7 svn:log V 85 Let OSPFv3 go through ipfw. Some more additional checks would be desirable, though. END K 10 svn:author V 8 tanimura K 8 svn:date V 27 2005-05-28T09:32:43.000000Z K 7 svn:log V 134 Add 6300ESB, which should be treated as ICH4. PR: kern/81573 Submitted by: OOTOMO Hiroyuki MFC after: 1 week END K 10 svn:author V 8 tanimura K 8 svn:date V 27 2005-05-28T09:41:40.000000Z K 7 svn:log V 19 MFC 1.52: 6300ESB. END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-28T11:54:38.000000Z K 7 svn:log V 791 MFC: NI_WITHSCOPEID cleanup contrib/tcp_wrappers/hosts_access.c: 1.5 contrib/tcp_wrappers/socket.c: 1.8 contrib/telnet/telnetd/telnetd.c: 1.28 lib/libc/net/rcmd.c: 1.41 lib/libc/net/res_send.c: 1.50 lib/libpam/modules/pam_unix/pam_unix.c: 1.50 lib/libutil/logwtmp.c: 1.20 lib/libutil/realhostname.c: 1.18 libexec/rlogind/rlogind.c: 1.44 libexec/tftpd/tftpd.c: 1.36 release/picobsd/tinyware/login/pico-login.c: 1.6 sbin/ifconfig/af_inet6.c: 1.2 sbin/route/route.c: 1.78 usr.bin/netstat/route.c: 1.76 usr.sbin/inetd/inetd.c: 1.129 usr.sbin/lpr/common_source/net.c: 1.9 usr.sbin/lpr/lpd/lpd.c: 1.41 usr.sbin/mountd/mountd.c: 1.81 usr.sbin/ppp/ncpaddr.c: 1.16 usr.sbin/rip6query/rip6query.c: 1.7 usr.sbin/syslogd/syslogd.c: 1.143 END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-05-28T12:41:55.000000Z K 7 svn:log V 30 Enable ReiserFS note on pc98. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T12:58:54.000000Z K 7 svn:log V 170 Acquire Giant explicitly in fhopen(), fhstat(), and kern_fhstatfs(), so that we can start to eliminate the presence of non-MPSAFE system call entries in syscalls.master. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T12:59:33.000000Z K 7 svn:log V 90 Mark fhopen(), fhstat(), and fhstatfs() as MSTD, since they now acquire Giant themselves. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:07:43.000000Z K 7 svn:log V 55 Also mark the COMPAT4 version of fhstatfs() as MPSAFE. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:08:26.000000Z K 7 svn:log V 52 Regenerate system call tables from syscalls.master. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:09:41.000000Z K 7 svn:log V 69 Mark kenv(2) as MPSAFE, since it appears to be properly locked down. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:09:56.000000Z K 7 svn:log V 41 Regenerate from updated syscalls.master. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:11:35.000000Z K 7 svn:log V 90 Acquire Giant explicitly in quotactl() so that the syscalls.master entry can become MSTD. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:12:04.000000Z K 7 svn:log V 25 Mark quotactl() as MSTD. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:13:01.000000Z K 7 svn:log V 33 Regenerate from syscalls.master. END K 10 svn:author V 2 bz K 8 svn:date V 27 2005-05-28T13:15:44.000000Z K 7 svn:log V 274 Fix use of uninitialized variable len in ngd_send. Note: len gets intialized to 0 for sap == NULL case only to make compiler on amd64 happy. This has nothing todo with the former uninitialized use of len in sap != NULL case. Reviewed by: glebius Approved by: pjd (mentor) END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:23:42.000000Z K 7 svn:log V 430 Mark the following compatability system calls as MCOMPAT or MCOMPAT4 based on the their simply wrapping MPSAFE implementations of existing MPSAFE system calls: getfsstat() lseek() stat() lstat() truncate() ftruncate() statfs() fstatfs() Note that ogetdirentries() is not marked MPSAFE because it does not share the MPSAFE implementation used for getdirentries(), and requires separate locking to be implemented. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T13:24:05.000000Z K 7 svn:log V 40 Regenerate for updated syscalls.master. END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-05-28T13:40:27.000000Z K 7 svn:log V 76 Change the spkr_set_pitch() function to a macro to fix low level profiling. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T14:34:41.000000Z K 7 svn:log V 228 Explicitly acquire Giant around the ntp_gettime() and assert it in the sysctl path. While this code is close to MPSAFE, it may require some additional locking. Mark ntp_gettime1() as GIANT_REQUIRED for now. Suggested by: phk END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T14:35:05.000000Z K 7 svn:log V 87 Mark ntp_gettime() as MSTD, since its system call path will acquire Giant if required. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-28T14:35:43.000000Z K 7 svn:log V 33 Regenerate from syscalls.master. END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-28T16:23:29.000000Z K 7 svn:log V 72 Document 'jid' keyword for ps(1) and '-j' option for pgrep(1)/pkill(1). END K 10 svn:author V 3 alc K 8 svn:date V 27 2005-05-28T17:13:36.000000Z K 7 svn:log V 126 pmap_enter() no longer requires Giant. Therefore, stop acquiring and releasing it in pmap_enter_quick(). MFC after: 3 weeks END K 10 svn:author V 3 alc K 8 svn:date V 27 2005-05-28T17:56:18.000000Z K 7 svn:log V 89 Update some comments to reflect the change from spl-based to lock-based synchronization. END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-28T18:25:52.000000Z K 7 svn:log V 97 Fix mismerge. Unbreak kernel build. Please tinderbox. Put on pointy hat. Phone call from: maxim END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-28T21:56:41.000000Z K 7 svn:log V 59 integrate changes from libpcap-0.9.1-096 Reviewed by: bms END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-28T22:29:44.000000Z K 7 svn:log V 175 Prevent loading modules with are compiled into the kernel. PR: kern/48759 Submitted by: Pawe³ Ma³achowski Patch from: demon MFC after: 2 weeks END K 10 svn:author V 3 gad K 8 svn:date V 27 2005-05-28T22:42:41.000000Z K 7 svn:log V 649 Change the way options are parsed on the `#!'-line of a shell-script. Instead of having the kernel parse that line and add an entry to the argument list for each 'separate word' it finds, have it add only one entry which holds all the words found on that line. The old behavior is useful in some situations, but it does not match the way any other operating system will parse that line. This has been discussed in the thread "Bug in #! processing - One More Time" on the freebsd-arch mailing list (starting back on Feb 24, 2005). The first few messages in that thread provide the background in much detail. PR: 16393 Reviewed by: freebsd-arch END K 10 svn:author V 3 gad K 8 svn:date V 27 2005-05-28T22:45:31.000000Z K 7 svn:log V 97 Provide info on the incompatible change in v1.33 of sys/kern/imgact_shell.c Discussed with: imp END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-28T23:20:05.000000Z K 7 svn:log V 268 Fix panic when module is compiled in and it is loaded from loader.conf. Only panic is fixed, module will be still listed in kldstat(8) output. Not sure what is correct fix, because adding unloading code in case of failure to linker_init_kernel_modules() doesn't work. END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-05-29T04:42:30.000000Z K 7 svn:log V 174 Remove bus_{mem,p}io.h and related code for a micro-optimization on i386 and amd64. The optimization is a trivial on recent machines. Reviewed by: -arch (imp, marcel, dfr) END K 10 svn:author V 3 tjr K 8 svn:date V 27 2005-05-29T08:18:48.000000Z K 7 svn:log V 63 Move information about exit status into a DIAGNOSTICS section. END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-05-29T08:43:44.000000Z K 7 svn:log V 930 Add VESA mode support for syscons, which enables the support of 15, 16, 24, and 32 bit modes. To use that, syscons(4) must be built with the compile time option 'options SC_PIXEL_MODE', and VESA support (a.k.a. vesa.ko) must be either loaded, or be compiled into the kernel. Do not return EINVAL when the mouse state is changed to what it already is, which seems to cause problems when you have two mice attached, and applications are not likely obtain useful information through the EINVAL caused by showing the mouse pointer twice. Teach vidcontrol(8) about mode names like MODE_, where is the video mode number from the vidcontrol -i mode output. Also, revert the video mode if something fails. Obtained from: DragonFlyBSD Discussed at: current@ with patch attached [1] PR: kern/71142 [2] Submitted by: Xuefeng DENG [1], Cyrille Lefevre [2] END K 10 svn:author V 6 grehan K 8 svn:date V 27 2005-05-29T08:51:21.000000Z K 7 svn:log V 269 The end values passed to rman_manage_region() for PCI i/o and mem spaces were 1 too large. This resulted in the rman list not being sorted correctly, and USB ports not being discovered on older TiBooks. Detective work by: Andreas Tobler END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-29T10:17:29.000000Z K 7 svn:log V 143 MFC: rev. 1.39 - Add a HARDWARE section based on the previous 'Supported Cards' subsection and the hardware notes. - Xref all snd_* drivers END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-29T10:20:13.000000Z K 7 svn:log V 172 MFC: - Autogenerate device listings for sound(4) aka pcm(4) - Rename the Audio Devices section into Sound Devices, this is more in line with out sound and snd_* drivers END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-29T10:21:38.000000Z K 7 svn:log V 740 MFC: The ai_addrlen of a struct addrinfo used to be a size_t, per RFC 2553. In XNS5.2, and subsequently in POSIX-2001 and RFC 3493, it was changed to a socklen_t. And, the n_net of a struct netent used to be an unsigned long integer. In XNS5, and subsequently in POSIX-2001, it was changed to an uint32_t. To accomodate for this while preserving ABI compatibility with the old interface, we need to prepend or append 32 bits of padding, depending on the (LP64) architecture's endianness. include/netdb.h: 1.36 (partly) lib/libc/net/getaddrinfo.c: 1.69 lib/libc/net/getnetbydns.c: 1.30 lib/libc/net/getnetbyht.c: 1.15 lib/libc/net/getnetbynis.c: 1.18 lib/libc/net/getnetnamadr.c: 1.20 (partly) lib/libc/net/netdb_private.h: 1.9 END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-29T10:29:21.000000Z K 7 svn:log V 165 MFC: rev. 1.281 Sort the sound section by driver name. The exception to this is sound(4), which is the master driver and will hopefully be renamed to snd(4) soon. END K 10 svn:author V 8 brueffer K 8 svn:date V 27 2005-05-29T10:34:38.000000Z K 7 svn:log V 46 MFC: rev. 1.264 by simon Add snd_audiocs(4). END K 10 svn:author V 3 pjd K 8 svn:date V 27 2005-05-29T10:36:37.000000Z K 7 svn:log V 319 MFC: named 1.22 We need to use 'applyset' command for devfs, 'apply hide' is not enough, because new devfs entries can show up later and one can access such entires from inside named's chroot. In rc.d scripts we can use devfs_domount() function with devfsrules_hide_all policy and unhide 'null' and 'random' manually. END K 10 svn:author V 4 nyan K 8 svn:date V 27 2005-05-29T11:53:14.000000Z K 7 svn:log V 64 Sync with syscons update (Add new member to struct sc_rndr_sw). END K 10 svn:author V 7 glebius K 8 svn:date V 27 2005-05-29T12:20:41.000000Z K 7 svn:log V 79 Fix check for leading zero, so that it does not block two zeroes in hook name. END K 10 svn:author V 6 marius K 8 svn:date V 27 2005-05-29T12:47:39.000000Z K 7 svn:log V 122 Add vr_init_t member to sc_rndr_sw_t instances in order to unbreak compilation after sys/dev/syscons/syscons.h rev. 1.83. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T13:38:07.000000Z K 7 svn:log V 3945 Kernel malloc layers malloc_type allocation over one of two underlying allocators: a set of power-of-two UMA zones for small allocations, and the VM page allocator for large allocations. In order to maintain unified statistics for specific malloc types, kernel malloc maintains a separate per-type statistics pool, which can be monitored using vmstat -m. Prior to this commit, each pool of per-type statistics was protected using a per-type mutex associated with the malloc type. This change modifies kernel malloc to maintain per-CPU statistics pools for each malloc type, and protects writing those statistics using critical sections. It also moves to unsynchronized reads of per-CPU statistics when generating coalesced statistics. To do this, several changes are implemented: - In the previous world order, the statistics memory was allocated by the owner of the malloc type structure, allocated statically using MALLOC_DEFINE(). This embedded the definition of the malloc_type structure into all kernel modules. Move to a model in which a pointer within struct malloc_type points at a UMA-allocated malloc_type_internal data structure owned and maintained by kern_malloc.c, and not part of the exported ABI/API to the rest of the kernel. For the purposes of easing a possible MFC, re-use an existing pointer in 'struct malloc_type', and maintain the current malloc_type structure size, as well as layout with respect to the fields reused outside of the malloc subsystem (such as ks_shortdesc). There are several unused fields as a result of no longer requiring the mutex in malloc_type. - Struct malloc_type_internal contains an array of malloc_type_stats, of size MAXCPU. The structure defined above avoids hard-coding a kernel compile-time value of MAXCPU into kernel modules that interact with malloc. - When accessing per-cpu statistics for a malloc type, surround read - modify - update requests with critical_enter()/critical_exit() in order to avoid races during write. The per-CPU fields are written only from the CPU that owns them. - Per-CPU stats now maintained "allocated" and "freed" counters for number of allocations/frees and bytes allocated/freed, since there is no longer a coherent global notion of the totals. When coalescing malloc stats, accept a slight race between reading stats across CPUs, and avoid showing the user a negative allocation count for the type in the event of a race. The global high watermark is no longer maintained for a malloc type, as there is no global notion of the number of allocations. - While tearing up the sysctl() path, also switch to using sbufs. The current "export as text" sysctl format is retained with the same syntax. We may want to change this in the future to export more per-CPU information, such as how allocations and frees are balanced across CPUs. This change results in a substantial speedup of kernel malloc and free paths on SMP, as critical sections (where usable) out-perform mutexes due to avoiding atomic/bus-locked operations. There is also a minor improvement on UP due to the slightly lower cost of critical sections there. The cost of the change to this approach is the loss of a continuous notion of total allocations that can be exploited to track per-type high watermarks, as well as increased complexity when monitoring statistics. Due to carefully avoiding changing the ABI, as well as hardening the ABI against future changes, it is not necessary to recompile kernel modules for this change. However, MFC'ing this change to RELENG_5 will require also MFC'ing optimizations for soft critical sections, which may modify exposed kernel ABIs. The internal malloc API is changed, and modifications to vmstat in order to restore "vmstat -m" on core dumps will follow shortly. Several improvements from: bde Statistics approach discussed with: ups Tested by: scottl, others END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T13:40:00.000000Z K 7 svn:log V 305 Modify vmstat(8)'s domem() routine, which is responsible for extracting malloc(9) statistics from kernel memory or a kernel coredump, to catch up with recent changes to adopt per-CPU malloc(9) statistics. The new routines walk the per-CPU statistics pools and coalesce them for presentation to the user. END K 10 svn:author V 6 avatar K 8 svn:date V 27 2005-05-29T14:56:51.000000Z K 7 svn:log V 17 Typo correction. END K 10 svn:author V 5 simon K 8 svn:date V 27 2005-05-29T15:33:35.000000Z K 7 svn:log V 416 Make sure the dest_realm buffer is always NUL terminated. This code is only used when rlogin is compiled with Kerberos 4 support, and since the Kerberos code in rlogin has been removed from FreeBSD 5.0 and newer, this is not an MFC. It is not believed that this bug could be exploited as a security vulnerability. PR: bin/80661 Submitted by: Przemyslaw Frasunek Approved by: cperciva END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T15:47:31.000000Z K 7 svn:log V 59 Move FreeBSD Id outside of copyright. Initialize variable. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T15:49:53.000000Z K 7 svn:log V 91 Reduce compiler warning: variable might be used uninitialized, by giving an initial value. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T15:52:48.000000Z K 7 svn:log V 57 Move variable initialization to reduce compiler warning. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T15:57:00.000000Z K 7 svn:log V 62 rscid -> __FBSDID. Mark parameter as __unused when necessary. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T16:01:12.000000Z K 7 svn:log V 127 Even if variable is never used uninitialized by the semantic, reduce compiler warning by giving an initial value in all cases. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T16:04:46.000000Z K 7 svn:log V 91 Reduce compiler warning: variable might be used uninitialized, by giving an initial value. END K 10 svn:author V 8 charnier K 8 svn:date V 27 2005-05-29T16:07:10.000000Z K 7 svn:log V 65 Give variable an initial value. Use errx() instead of fprintf(). END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-29T16:08:39.000000Z K 7 svn:log V 253 MFC: teach IPv6 to async DNS resolver in ypserv: - query AAAA RR on ipnodes.byname query. - query .ip6.arpa. on ipnodes.byaddr query. usr.sbin/ypserv/yp_dnslookup.c: 1.26 usr.sbin/ypserv/yp_extern.h: 1.17 usr.sbin/ypserv/yp_server.c: 1.38, 1.39 END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T16:10:33.000000Z K 7 svn:log V 226 Add place-holder audit.h that defines only au_event_t, which is needed in order to modify the system call table to include event identifiers. The full audit.h will be merged at a later date. Obtained from: TrustedBSD Project END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T16:11:34.000000Z K 7 svn:log V 108 For consistency with more system include files, add a trailing '_' to the define guards in audit_kevents.h. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T16:17:00.000000Z K 7 svn:log V 139 Do install BSM include files (such as they are) when installing system includes. Submitted by: wsalamon Obtained from: TrustedBSD Project END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T16:18:29.000000Z K 7 svn:log V 102 Add /usr/include/bsm to mtree creation set. Submitted by: wsalamon Obtained from: TrustedBSD Project END K 10 svn:author V 7 delphij K 8 svn:date V 27 2005-05-29T16:28:07.000000Z K 7 svn:log V 67 Constify savedir pointer, and remove an accompaning strdup() call. END K 10 svn:author V 3 ume K 8 svn:date V 27 2005-05-29T17:06:09.000000Z K 7 svn:log V 176 MFC 1.36: add targets for ipnodes.byname and ipnodes.byaddr which carries IPv6 addresses. these targets are off by default, and these are built when there is /var/yp/ipnodes. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2005-05-29T17:27:52.000000Z K 7 svn:log V 305 Due to a last minute change in the #ifdefing in malloc.h before committing, a nested include of param.h is required so that MAXCPU is visible to all consumers of sys/malloc.h. In an earlier version of the patch, the malloc_type_internal structure was only conditionally visible. Pointed out by: delphij END K 10 svn:author V 4 bmah K 8 svn:date V 27 2005-05-29T17:32:27.000000Z K 7 svn:log V 110 Remove dangling reference to the no-longer-existant-in-this-branch Migration Guide. Pointed out by: keramida END K 10 svn:author V 8 schweikh K 8 svn:date V 27 2005-05-29T17:43:24.000000Z K 7 svn:log V 188 Chop a '>' in a feature name (RSVD2>) that snuck in; this now balances the <> flags displayed at boot, e.g. without this Features2=0x41d,MON,DS_CPL,CNTX-ID> MFC after: 1 week END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T17:46:52.000000Z K 7 svn:log V 61 Virgin import of libpcap v0.9.1 (alpha 096) from tcpdump.org END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T17:46:52.000000Z K 7 svn:log V 144 This commit was generated by cvs2svn to compensate for changes in r146768, which included commits to RCS files with non-trunk default branches. END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2005-05-29T17:46:53.000000Z K 7 svn:log V 82 This commit was manufactured by cvs2svn to create tag 'libpcap-vendor-v0_9_1-096'. END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:09:04.000000Z K 7 svn:log V 60 resolve merge conflicts Reviewed by: bms (earlier version) END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:12:46.000000Z K 7 svn:log V 32 update for import of v0.9.1-096 END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:17:16.000000Z K 7 svn:log V 61 Virgin import of tcpdump v3.9.1 (alpha 096) from tcpdump.org END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:17:16.000000Z K 7 svn:log V 144 This commit was generated by cvs2svn to compensate for changes in r146773, which included commits to RCS files with non-trunk default branches. END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:24:21.000000Z K 7 svn:log V 61 Virgin import of tcpdump v3.9.1 (alpha 096) from tcpdump.org END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T18:24:21.000000Z K 7 svn:log V 144 This commit was generated by cvs2svn to compensate for changes in r146775, which included commits to RCS files with non-trunk default branches. END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2005-05-29T18:24:22.000000Z K 7 svn:log V 82 This commit was manufactured by cvs2svn to create tag 'tcpdump-vendor-v3_9_1-096'. END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T19:09:28.000000Z K 7 svn:log V 249 resolve merge conflicts and update for proper build; including: o print-fr.c returned to code on vendor branch o remove pmap_prot.h include from print-sunrprc.c o remove gcc/i386-specific ntoh* write-arounds from tcpdump-stdinc.h Reviewed by: bms END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T19:11:21.000000Z K 7 svn:log V 52 update after import of v3.9.1-096 Reviewed by: bms END K 10 svn:author V 3 sam K 8 svn:date V 27 2005-05-29T19:15:34.000000Z K 7 svn:log V 43 bump version for libpcap (+tcpdump) import END K 10 svn:author V 3 phk K 8 svn:date V 27 2005-05-29T19:29:19.000000Z K 7 svn:log V 271 Hand it to spaff, he knows how to say it: Using encryption on the Internet is the equivalent of arranging an armoured car to deliver credit card information from someone living in a cardboard box to someone living on a park bench. -- Gene Spafford, Purdue University. END