ƒ¯x75201 189 383 160 135 290 135 245 200 229 135 187 381 188 126 124 143 307 310 1405 322 223 168 124 126 136 115 277 244 444 128 308 182 161 264 137 717 257 122 265 406 205 182 153 283 269 199 259 240 450 154 164 175 184 226 257 170 186 216 199 256 176 308 140 437 760 240 164 134 268 129 223 163 144 488 505 137 252 246 147 208 140 168 168 174 199 200 117 156 129 191 122 147 237 185 173 633 194 154 265 174 170 254 125 1079 148 230 287 165 194 419 202 167 166 200 661 263 209 161 157 210 229 422 356 123 445 106 229 122 166 319 311 337 376 117 296 301 234 340 215 173 182 108 172 143 627 163 102 187 102 170 150 101 171 319 164 129 125 683 401 218 471 304 608 707 143 270 159 167 345 237 186 271 168 1233 257 132 124 7584 165 261 232 130 349 160 137 119 230 133 155 148 K 10 svn:author V 3 ken K 8 svn:date V 27 2001-04-04T22:50:57.000000Z K 7 svn:log V 290 Don't allow immediate values of 0 for operations that can take either an immediate value or the accumulator. 0 is the chip's internal representation for the accumulator, and so 0 is an invalid immediate value when the accumulator can also be specified as an argument. Submitted by: gibbs END K 10 svn:author V 3 phk K 8 svn:date V 27 2001-04-04T23:07:22.000000Z K 7 svn:log V 68 Fix a potential ROOT-exploit in NTPD. PR: 26358 Reviewed by: dima END K 10 svn:author V 3 phk K 8 svn:date V 27 2001-04-04T23:09:10.000000Z K 7 svn:log V 43 MFC: Fix a potential ROOT-exploit in NTPD. END K 10 svn:author V 3 gad K 8 svn:date V 27 2001-04-04T23:27:35.000000Z K 7 svn:log V 197 Fix bpf devices so select() recognizes that they are always writable. PR: 9355 Submitted by: Bruce Evans Reviewed by: Garrett Rooney (see pr :-) END K 10 svn:author V 5 assar K 8 svn:date V 27 2001-04-05T00:37:06.000000Z K 7 svn:log V 41 MFC: 1.19: disable RSA Approved by: jkh END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-05T01:24:05.000000Z K 7 svn:log V 153 Document the extattr_(get|set|delete)_fd calls, and add a note warning people that these calls are likely to change in the future. Reviewed by: rwatson END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-05T01:25:42.000000Z K 7 svn:log V 108 Make it clear that commands within labels must be indented to be recognized. PR: 26293 Reviewed by: brian END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-05T01:36:39.000000Z K 7 svn:log V 135 New release note: Fix for ntpd(8) vulnerability. Correction: fetch(1) recently gained a -U option, not ftp(1) (pointed out by des). END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-05T01:36:54.000000Z K 7 svn:log V 40 Install the acl_create_entry.3 man page END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-05T01:38:36.000000Z K 7 svn:log V 94 MFC: ntpd(8) vulnerability fix, correct fetch(1)/ftp(1) typo. Approved by: jkh (implicitly) END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-05T01:48:07.000000Z K 7 svn:log V 285 - Add descriptions and cross-references for the ACL editing library functions. - Place the acl_dup() description in alphabetical order. - Move the POSIX.1e descriptions under the ENVIRONMENT section to the STANDARDS section. Reviewed by: rwatson Obtained from: TrustedBSD Project END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-05T02:23:48.000000Z K 7 svn:log V 94 When we change the interface MTU, run through the routing table and tweak all route MTUs too. END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-05T02:25:44.000000Z K 7 svn:log V 32 MFC: Keep route MTUs up to date END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T02:46:21.000000Z K 7 svn:log V 29 This builds fine with -Wall. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T02:46:40.000000Z K 7 svn:log V 48 Default the UTC question to "yes" on the Alpha. END K 10 svn:author V 8 bmilekic K 8 svn:date V 27 2001-04-05T03:55:27.000000Z K 7 svn:log V 209 Various style fixes. Also place the macros under #ifdef _KERNEL. Equally hide the internal structures such as the freelist structs which include condition variables. Reviewed by: bde Mostly suggested by: bde END K 10 svn:author V 8 bmilekic K 8 svn:date V 27 2001-04-05T04:20:48.000000Z K 7 svn:log V 212 Change a couple of M_WAITOKs used in M_PREPEND() to M_TRYWAITs, which is what they should be. As the returned mbuf is already checked for failure of M_PREPEND even in the wait case, nothing more to be done here. END K 10 svn:author V 7 rwatson K 8 svn:date V 27 2001-04-05T06:12:38.000000Z K 7 svn:log V 1307 o Rather than arbitrarily construct a credential in the nfs_statfs() VFS operation, make use of the calling process's credential. This solution may not be ideal (there are a number of other possible proposals, including making use of the proc0 credential, adding a credential argument to the VFSOP, and switching from a hard-coded ucred to a hard-coded nfscred), it is simple and appears to work. The arguments against using simply crget() are fairly strong: it is the only place in the code (other than a nearly identical invocation in ncp) where crget() is invoked, other than in the process credential creation code; as ucred becomes extensible, this use of crget() without appropriate context results in less and less meaningful credential data. The implementation here will probably be tweaked as a result of experimentation and further exploration of the requirements. In the mean-time, it allows progress to be made in ucred expansion for new security models without causing a crash every time df is used on an NFS mounted file system. This code has been interop tested against FreeBSD and Solaris NFS servers. While using the process credentials should not introduce interop problems, please let me know if any turn out to exist. Reviewed by: freebsd-arch END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-05T06:56:10.000000Z K 7 svn:log V 226 Don't leak resources: Don't leak iospace when irq allocation fails. (call wi_free()) Call bus_release_resource() with the correct "rid" obtained from bus_alloc_resource() that's saved in the softc instead of a hardcoded 0. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T07:35:59.000000Z K 7 svn:log V 131 mdoc(7) police rev 1.11: convert descriptions and cross-references for the ACL editing library functions to the plain tagged list. END K 10 svn:author V 3 bde K 8 svn:date V 27 2001-04-05T07:37:55.000000Z K 7 svn:log V 76 Fixed some printf format errors (don't assume that ntohl() returns u_long). END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T07:42:56.000000Z K 7 svn:log V 33 Record -lposix1e merge with -lc. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T07:47:06.000000Z K 7 svn:log V 35 Unbreak world. Submitted by: jhay END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T07:51:48.000000Z K 7 svn:log V 45 mdoc(7) police: add missing LIBRARY section. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T07:59:19.000000Z K 7 svn:log V 24 Record -lposix1e death. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-05T08:04:40.000000Z K 7 svn:log V 185 MFC: - rev 1.17 (off-by-nine error when building a list of include files) - rev 1.18 (fix to rev 1.17) - rev 1.20 (search for include files in lexicographical order) Approved by: jkh END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-05T09:25:37.000000Z K 7 svn:log V 148 WI_TIMEOUT is 65536 65536 * 10msec == 10 minutes 55 seconds == hung machine Instead wait for the busy bits for a max of ~2 seconds (200 * 10msec) END K 10 svn:author V 3 mpp K 8 svn:date V 27 2001-04-05T09:37:55.000000Z K 7 svn:log V 351 We don't use the "pseudo-device" configuration option any longer, so update the example to use the correct definition. Add an example for documenting kernel compile options, along with a small example of how to reference them in the main text of the man page (I.e. the .Dv macro). Inspired-by: a brief exchange I saw in in the commit messages mail END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-05T09:47:07.000000Z K 7 svn:log V 33 test should be for == 0, not < 0 END K 10 svn:author V 3 dfr K 8 svn:date V 27 2001-04-05T10:28:52.000000Z K 7 svn:log V 215 Don't call prom_open() multiple times. This confuses some versions of SRM and makes it impossible to boot from floppy and CD on some AlphaServer platforms. Detective work by: Michael Richards END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T11:13:07.000000Z K 7 svn:log V 90 Correct the sysctl names to match the tuneables. Proberly flush the tag queue on detach. END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T11:15:24.000000Z K 7 svn:log V 69 Add new flag ATPR_F_QUIET to atapi_request. Cleanup error handling. END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T11:17:33.000000Z K 7 svn:log V 171 On burners that return faulty ready on fixate, wait for the expected fixate time before returning. Dont print error messages to the console on READ_TOC on a blank media. END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-05T12:11:02.000000Z K 7 svn:log V 42 Remove the recently-depricated LIBPOSIX1E END K 10 svn:author V 3 joe K 8 svn:date V 27 2001-04-05T13:59:51.000000Z K 7 svn:log V 624 Fix a few bugs in BSDPAN: * Fix a bug which prevented the second invocation of overloaded subs governed by SelfLoader from functioning. * Fix a bug with XS modules. MakeMaker determines where the xsubpp preprocessor is located by adding "ExtUtils" string to the Perl's system path. At the same time, BSDPAN has to fool MakeMaker into thinking that the Perl's system path is elsewhere. Now we `reverse-adjust' the notion of the Perl's system path for a moment, so xsubpp utility is found. This should fix the breakage with some p5- ports. Reported by: vanilla Submitted by: Anton Berezin END K 10 svn:author V 6 nsayer K 8 svn:date V 27 2001-04-05T14:09:15.000000Z K 7 svn:log V 161 Reactivate SRA. Make handling of SIGINT and SIGQUIT follow SIGTSTP in TerminalNewMode(). This allows people to break out of SRA authentication if they wish to. END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T15:45:53.000000Z K 7 svn:log V 30 More error handling cleanups. END K 10 svn:author V 3 imp K 8 svn:date V 27 2001-04-05T17:00:46.000000Z K 7 svn:log V 172 Soften the dire warnings about this code. Things are kinda working now and it does compile :-). There are still some issues, but it is a good time to soften the warning. END K 10 svn:author V 7 iedowse K 8 svn:date V 27 2001-04-05T17:18:36.000000Z K 7 svn:log V 309 Add IPv6 support to showmount(8). This replaces IPv4-specific code with calls to the new protocol-independent clnt_*_create functions provided by ti-rpc. Martin submitted a more complex patch to achieve this, but it turns out that clnt_create() does everything we need. Reviewed by: Martin Blapp END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T17:21:54.000000Z K 7 svn:log V 112 MFC: Add tunables to control DMA/WC/TAGS/ATAPI_DMA and associated sysctl's to read settings. Approved by: jkh END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T17:23:43.000000Z K 7 svn:log V 90 MFC: remove opsolete WC/TAGS/ATAPI_DMA options, they are now tuneables Approved by: jkh END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-05T17:24:48.000000Z K 7 svn:log V 61 MFC: Update to reflect the new tunables instead of options. END K 10 svn:author V 6 nsayer K 8 svn:date V 27 2001-04-05T17:27:08.000000Z K 7 svn:log V 187 1. Revert the removal of SRA 2. MFC sys_bsd.c v1.5 - fix the bug that started the fuss. Make SIGINT and SIGQUIT follow the state of SIGTSTP, thus making SRA abortable. Approved by: jkh END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-05T18:12:50.000000Z K 7 svn:log V 175 New release notes: ata(4) configure options are now loader tunables, isp(4) target mode support (submitted by mjacob). Update hardware lists from isp(4) manpage, fix typos. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-05T18:14:25.000000Z K 7 svn:log V 105 MFC: ata(4) loader tunables. Update hardware lists from isp(4) manpage. Approved by: jkh (implicitly) END K 10 svn:author V 3 ume K 8 svn:date V 27 2001-04-05T19:45:02.000000Z K 7 svn:log V 166 - correct logic of per-address input packet counts for lo0 - reject packets to fe80::xxxx%lo0 (xxxx != 1) Submitted by: JINMEI Tatuya END K 10 svn:author V 3 ume K 8 svn:date V 27 2001-04-05T19:56:54.000000Z K 7 svn:log V 147 IPv6 support for skeyaccess(3). You can specify IPv6 address using `internet' keyword into /etc/skey.access. Not Objected to by: -audit and -net END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T20:45:25.000000Z K 7 svn:log V 354 Rev 1.10 actually only tested booting from floppies. Thus that is all that was fixed. There are other problems why the current sources for the Alpha `cdboot' do not produce a working loader. Because we use the 4.0-RELEASE `cdboot' binary, it will not get these fixes at the current time. Thus CD booting on certain AlphaServer boxes is still broken. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T20:52:33.000000Z K 7 svn:log V 59 Remove -Wall. It breaks compiling with non-gcc compilers. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T21:34:47.000000Z K 7 svn:log V 69 Bump the default / size up another 10MB to 120MB. Requested by: jhb END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-05T22:36:41.000000Z K 7 svn:log V 80 Don't need to preload the `vn' KLD anymore as we use `md' for release building. END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-05T22:53:35.000000Z K 7 svn:log V 90 It's possible to kldload netgraph (as well as the other ng_* bits), so do it if required. END K 10 svn:author V 3 gad K 8 svn:date V 27 2001-04-06T02:14:39.000000Z K 7 svn:log V 133 Make sure the euid is restored in an obscure error situation. Submitted by: J Wunsch in freebsd-audit END K 10 svn:author V 5 green K 8 svn:date V 27 2001-04-06T05:52:53.000000Z K 7 svn:log V 162 Label the OpenSSH section, then add a small explanation for the existence of the non-typo "csshd" service. Prodded by: markm Reviewed by: markm, dwhite, rwatson END K 10 svn:author V 5 billf K 8 svn:date V 27 2001-04-06T06:52:25.000000Z K 7 svn:log V 76 pipe/queue are the only consumers of flow_id, so only set it in those cases END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-06T07:20:27.000000Z K 7 svn:log V 94 Axe the per-cpu variable witness_spin_check as it was replaced by the per-cpu spinlocks list. END K 10 svn:author V 3 bde K 8 svn:date V 27 2001-04-06T11:34:03.000000Z K 7 svn:log V 123 MFC (rev.1.14: don't follow the null pointer went the server sends an unparseable timestamp). PR: 25243 Approved by: jkh END K 10 svn:author V 3 bde K 8 svn:date V 27 2001-04-06T11:37:50.000000Z K 7 svn:log V 106 MFC (rev.1.14: don't follow the null pointer went the server sends an unparseable timestamp). PR: 25243 END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-06T14:15:38.000000Z K 7 svn:log V 160 - Correct off-by-one error and buffer underflow from previous fix - int -> unsigned char fixes Submitted by: ache, dillon, Mark Andrews, et.al. (on -security) END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-06T14:34:15.000000Z K 7 svn:log V 81 Do not log every potential exploit attempt since a denial-of-service may result. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-06T14:40:26.000000Z K 7 svn:log V 216 Execute this script in a "C" locale environment. Currently, cs_CZ.ISO_8859-2 locale's collation sequence is broken, and this caused grep(1) to skip some include files. Reported by: Michal Mertl END K 10 svn:author V 7 darrenr K 8 svn:date V 27 2001-04-06T15:52:28.000000Z K 7 svn:log V 44 fix security hole created by fragment cache END K 10 svn:author V 6 nsayer K 8 svn:date V 27 2001-04-06T15:56:10.000000Z K 7 svn:log V 341 Clean up telnet's argument processing a bit. autologin and encryption is now the default, so ignore the arguments that turn it on. Add a new -y argument to turn off encryption in case someone wants to do that. Sync these changes with the man page (including removing the now obsolete statement about availability only in the US and Canada). END K 10 svn:author V 7 wollman K 8 svn:date V 27 2001-04-06T16:43:21.000000Z K 7 svn:log V 663 Latest timezone database form Arthur Olson. This update includes changes to the following locations: Antarctica Australia (additional historical comments) Bangladesh (new spelling of Dhaka) Brazil (multiple changes; America/Porto_Acre renamed America/Rio_Branco) CNMI Canada Chile Dominican Republic East Timor Falkland Islands Fiji France (additional history) Guam Israel (additional historical comments) Latvia Mexico Moldova (Europe/Tiraspol removed) Netherlands (additional history) Paraguay Philippines (additional history) Tonga United States (additional historical comments) Obtained from: Arthur Olson; END K 10 svn:author V 7 wollman K 8 svn:date V 27 2001-04-06T16:43:21.000000Z K 7 svn:log V 143 This commit was generated by cvs2svn to compensate for changes in r75264, 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 2001-04-06T16:43:22.000000Z K 7 svn:log V 68 This commit was manufactured by cvs2svn to create tag 'tzdata2001b'. END K 10 svn:author V 7 wollman K 8 svn:date V 27 2001-04-06T16:46:52.000000Z K 7 svn:log V 38 Merge tzdata2001b from vendor branch. END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-06T17:51:27.000000Z K 7 svn:log V 172 MFC revs 1.3 and 1.4: - Correct off-by-one error and buffer underflow from previous fix - int -> unsigned char fixes - Prevent potential syslog DoS Approved by: jkh END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-06T18:21:00.000000Z K 7 svn:log V 34 MFC: Fix potential remote exploit END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-06T19:06:19.000000Z K 7 svn:log V 127 Correct a bug in rev 1.589. The `make obj' that was removed for sysinstall was also effectively a `make obj' for src/release. END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-06T19:18:35.000000Z K 7 svn:log V 71 Add ATA66 and ATA100 mode support for Acer chipsets. MFC candidate :) END K 10 svn:author V 3 sos K 8 svn:date V 27 2001-04-06T19:23:47.000000Z K 7 svn:log V 52 MFC: Add ATA66 & ATA100 support for Acer chipsets. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-06T21:37:52.000000Z K 7 svn:log V 395 - Split out the functionality of displaying the contents of a single lock list into a public witness_list_locks() function. Call this function twice in witness_list() instead of using an evil goto. - Adjust the 'show locks' command to take an optional parameter which specifies the pid of a process to list the locks of. By default the locks held by the current process are displayed. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-06T21:41:52.000000Z K 7 svn:log V 412 Add a new ddb command 'show pcpu' which lists some of the per-cpu data. Specifically, the cpuid, curproc, curpcb, npxproc, and idleproc members. Also, if witness is compiled into the kernel, then a list of all the spin locks held by this CPU is displayed. By default the information for the current CPU is displayed, but a decimal cpu id may be specified as a parameter to obtain information on a specific CPU. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-06T21:48:19.000000Z K 7 svn:log V 42 replace hardcoded 1518 with ETHER_MAX_LEN END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-06T22:21:57.000000Z K 7 svn:log V 156 use correct contants (from net/ethernet.h) ETHER_TYPE_LEN instead of sizeof(u_int16_t) when looking at an ethernet header ETHERTYPE_IP instead of 0x800 END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-07T01:48:55.000000Z K 7 svn:log V 150 Blow away all the help files on the MFSROOT for the Alpha. Someone did something that really blew up the size of the crunchgen'ed sysinstall+friends. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-07T01:53:14.000000Z K 7 svn:log V 52 Strip comments from the MFSROOT's /stand/etc files. END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-07T02:53:35.000000Z K 7 svn:log V 113 The sess column went away last December with v1.26 of keyword.c Remove it from ``jfmt''. Forgotten by: mckusick END K 10 svn:author V 7 darrenr K 8 svn:date V 27 2001-04-07T03:40:32.000000Z K 7 svn:log V 44 fix security hole created by fragment cache END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-07T06:12:10.000000Z K 7 svn:log V 73 Reorder the Alpha cut-out list to match the order in the GENERIC kernel. END K 10 svn:author V 5 billf K 8 svn:date V 27 2001-04-07T08:19:20.000000Z K 7 svn:log V 74 Typo ("no not" -> "do not") Submitted by: Dave McKay END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-07T11:09:28.000000Z K 7 svn:log V 83 MFC: (rev 1.21) Execute this script in a "C" locale environment. Approved by: jkh END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-07T11:13:46.000000Z K 7 svn:log V 107 Introduce SCRIPTS for bsd.prog.mk. See bsd.README for details. Idea stolen from NetBSD. Reviewed by: bde END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-07T11:16:41.000000Z K 7 svn:log V 105 Update documentation in line with what the code does PR: 25435 Submitted by: dd Forgotten by: mckusick END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-07T11:21:35.000000Z K 7 svn:log V 26 beforeinstall -> SCRIPTS. END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-07T11:22:26.000000Z K 7 svn:log V 62 Remove tsess for now. This (and sess) may come back shortly. END K 10 svn:author V 5 green K 8 svn:date V 27 2001-04-07T11:46:51.000000Z K 7 svn:log V 35 MFC: add "csshd" Approved by: jkh END K 10 svn:author V 3 phk K 8 svn:date V 27 2001-04-07T13:24:59.000000Z K 7 svn:log V 99 Don't make fsck go quite _that_ fast: In Preen mode we only checked one partition per disk device. END K 10 svn:author V 5 orion K 8 svn:date V 27 2001-04-07T14:12:53.000000Z K 7 svn:log V 28 Quieten when re-triggering. END K 10 svn:author V 5 orion K 8 svn:date V 27 2001-04-07T14:16:31.000000Z K 7 svn:log V 53 MFC sample rate fixes for release. Approved by: jkh END K 10 svn:author V 3 des K 8 svn:date V 27 2001-04-07T15:26:31.000000Z K 7 svn:log V 144 If the server's reply to the SIZE command is unparseable, reset us->size to -1 so the caller will know it's invalid. This is an MFC candidate. END K 10 svn:author V 5 peter K 8 svn:date V 27 2001-04-07T16:48:51.000000Z K 7 svn:log V 91 Add missing KMODDEPS=snd_pcm for two drivers in RELENG_4 PR: kern/26394 Approved by: jkh END K 10 svn:author V 3 des K 8 svn:date V 27 2001-04-07T17:30:48.000000Z K 7 svn:log V 81 MFC: set us->size to -1 when server's response is unparseable. Approved by: jkh END K 10 svn:author V 3 des K 8 svn:date V 27 2001-04-07T19:51:12.000000Z K 7 svn:log V 540 Let pseudofs into the warmth of the FreeBSD CVS repo. It's not finished yet (I still have to find a way to implement process- dependent nodes without consuming too much memory, and the permission system needs tightening up), but it's becoming hard to work on without a repo (I've accidentally almost nuked it once already), and it works (except for the lack of process-dependent nodes, that is). I was supposed to commit this a week ago, but timed out waiting for jkh to reply to some questions I had. Pass him a spoonful of bad karma :) END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-07T20:47:29.000000Z K 7 svn:log V 98 Add id for the IO Data ET/T PR: 23877 Submitted by: Makoto MATSUSHITA END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-07T20:51:24.000000Z K 7 svn:log V 58 Use getopt instead of a home grown one Submitted by: DES END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2001-04-07T23:48:46.000000Z K 7 svn:log V 167 fix cd9660 to work on files larger than ISO_DEFAULT_BLOCK_SIZE and unbreak cdboot on alphas (which has been broken since just after 4.0-RELEASE) submitted by: jlemon END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2001-04-07T23:52:31.000000Z K 7 svn:log V 77 build cdboot from sources now that the cd9660 fs support works MFC candidate END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2001-04-08T00:01:54.000000Z K 7 svn:log V 73 no longer needed now that we are able to build cdboot from sources again END K 10 svn:author V 5 greid K 8 svn:date V 27 2001-04-08T00:07:53.000000Z K 7 svn:log V 159 Add yet another chip revision of the ES1371 which requires initialisation delays PR: 26415 Submitted by: Jose M. Alcaide Reviewed by: cg END K 10 svn:author V 6 jedgar K 8 svn:date V 27 2001-04-08T00:15:00.000000Z K 7 svn:log V 30 MFC: globbing limits for ftpd END K 10 svn:author V 4 grog K 8 svn:date V 27 2001-04-08T02:18:14.000000Z K 7 svn:log V 985 struct drive: Workaround size mismatch in struct drive on alpha platform only. This struct is shared between kernel and userland and contains a dev_t, which now refers to a completely different structure in userland and kernel: int the kernel, userland dev_t is called udev_t, and dev_t is struct specinfo *. In practical terms, kernel dev_t is a pointer and userland dev_t is an int, which on the alpha are different sizes. As stated, this is a workaround. The correct fix would be in two parts: 1. Don't define structures in the kernel with the same name and a different content as a userland structure. This is a kernel-wide issue: kernel dev_t needs another name. 2. Don't export unnecessary kernel structures to userland. This can be done within Vinum. It is currently on the tuit queue. Sleuth work by: Normand Leclerc Tested by: Normand Leclerc Blanket approved by: jkh END K 10 svn:author V 4 jake K 8 svn:date V 27 2001-04-08T04:15:26.000000Z K 7 svn:log V 55 Fix a precedence bug. ! has higher precedence than &. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-08T04:28:39.000000Z K 7 svn:log V 136 New release notes: Fix IPFilter fragment cache, libposix1e integration into libc. Note default root partition for alpha is now 120MB. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-08T04:31:30.000000Z K 7 svn:log V 193 MFC: Fix IPFilter fragment cache. Semi-MFC: ata(4) supports ATA66/ATA100 on Acer chipsets (not noted in -CURRENT because this note covered by existing items). Approved by: jkh (implicitly) END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-08T06:58:48.000000Z K 7 svn:log V 74 Install links to extattr_(get|set|delete)_fd.2 (from extattr_get_file.2). END K 10 svn:author V 4 nate K 8 svn:date V 27 2001-04-08T16:43:05.000000Z K 7 svn:log V 100 - Applied PHK's NTPDX remote-exploit fix. { Tested on my 2.2.8 box using the sample exploit code } END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2001-04-08T16:43:59.000000Z K 7 svn:log V 321 insert a magical second memory barrier prior to calling draina() in badaddr_read(). This fixes 'machine check in pal mode' halts on ev5 2100As. MFC candidate -- after spending 6 hours tracking this down, I checked and discovered that it has been in NetBSD for over a year, so it should be safe for MFC into 4.3-RELEASE END K 10 svn:author V 4 nate K 8 svn:date V 27 2001-04-08T16:51:41.000000Z K 7 svn:log V 108 - Merge in better fix for remote buffer expoit from RELENG_3. Prodded by: Chris Faulhaber END K 10 svn:author V 5 orion K 8 svn:date V 27 2001-04-08T18:24:49.000000Z K 7 svn:log V 73 MFC sample rate fixes and 4 speaker audio for release. Approved by: jkh END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-08T18:41:22.000000Z K 7 svn:log V 75 wd0 -> ad0 PR: 26343 Submitted by: Sergey A. Osokin END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-08T19:11:25.000000Z K 7 svn:log V 108 Mention that locks are inherited across an exec. PR: 24802 Submitted by: Kenneth Ingham END K 10 svn:author V 7 iedowse K 8 svn:date V 27 2001-04-08T19:21:50.000000Z K 7 svn:log V 564 The maximum RPC message size was set at 8k for UDP. This is lower than the default buffer size in the old RPC code (8800 bytes), and it could not be overriden by the application. This caused problems with CFS (/usr/port/security/cfs). Change this default back to UDPMSGSIZE (8800 bytes), but more importantly, allow applications to use larger message sizes for all protocols if desired. Choose an arbitrary maximum message size of 256k instead of using the default as the maximum (which is silly). Reported by: ache Reviewed by: alfred, Martin Blapp END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-08T19:24:41.000000Z K 7 svn:log V 167 MFC:insert a magical second memory barrier prior to calling draina() in badaddr_read(). This fixes 'machine check in pal mode' halts on ev5 2100As. Approved by: jkh END K 10 svn:author V 5 markm K 8 svn:date V 27 2001-04-08T20:11:51.000000Z K 7 svn:log V 114 From the author: Fix a bug when the return values from the overridden sub were lost during the first invocation. END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-08T20:14:14.000000Z K 7 svn:log V 70 if the feeder chain returned no data, do not try to acquire the data. END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-08T20:17:03.000000Z K 7 svn:log V 66 fix feeder initialisation methods to return correct result codes. END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-08T20:20:52.000000Z K 7 svn:log V 118 minor tweaks in speed and format setting routines. don't stop exploring the feeders if a feeder fails to initialise. END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-08T20:26:22.000000Z K 7 svn:log V 137 add a software sample rate conversion feeder. this uses linear interpolation for reasonable quality whilst not using too much cpu time. END K 10 svn:author V 5 joerg K 8 svn:date V 27 2001-04-08T20:29:09.000000Z K 7 svn:log V 327 Move the decision whether we want to request authentication from our peer out from sppp_lcp_open() to sppp_lcp_up(). For one, this makes things look more symmetrical to sppp_lcp_close(), and somehow it also just occurred to me that an Up event following the open caused the value of the authentication option to be clobbered. END K 10 svn:author V 3 jkh K 8 svn:date V 27 2001-04-08T20:42:52.000000Z K 7 svn:log V 263 According with RFC 2330 ("Framework for IP Performance Metrics", from Paxson et al, Status: Informational, May 1998), we should use "bits per second" and "k" as 1000 not 1024 for throughput measures. Submitted by: Eduardo Souza Machado da Silva END K 10 svn:author V 3 jkh K 8 svn:date V 27 2001-04-08T20:43:22.000000Z K 7 svn:log V 31 MFC: correct metrics reporting END K 10 svn:author V 3 joe K 8 svn:date V 27 2001-04-08T21:35:17.000000Z K 7 svn:log V 352 Don't complain about deprecated environment variables, like MUSIC_CD, being present in the environment if the user has CDROM defined, or has specified a device on the command line. This avoids users of ports like 'workman' that use these variables getting gratuitous warnings from cdcontrol. Suggested by: John Sellens END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-08T21:50:41.000000Z K 7 svn:log V 11 Style fix. END K 10 svn:author V 5 greid K 8 svn:date V 27 2001-04-08T23:02:06.000000Z K 7 svn:log V 134 Reinitialise the DSP and mixer after a resume from suspend PR: 22372 Submitted by: Hiroyuki Aizu Reviewed by: cg END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-08T23:02:12.000000Z K 7 svn:log V 27 Fix typo in error message. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-08T23:09:21.000000Z K 7 svn:log V 71 Add FreeBSD IDs to some files and make the others consistent in style. END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T05:41:41.000000Z K 7 svn:log V 223 Correctly initialize free_ccbq so that if we fail to attach (as is possible for some systems where the device is there, but the BIOS hasn't allocated memory resources for it), we don't panic. Submitted by: Gerard Roudier END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T05:43:49.000000Z K 7 svn:log V 215 MFC: Correctly initialize free_ccbq so that if we fail to attach (as is possible for some systems where the device is there, but the BIOS hasn't allocated memory resources for it), we don't panic. Approved by: jkh END K 10 svn:author V 3 imp K 8 svn:date V 27 2001-04-09T06:33:36.000000Z K 7 svn:log V 244 Two minor fixes: o Change the number of init tries from 5 to a #define. o Allow up to 5s rather than 2s for commands to complete. This is still much less than 51 minutes, but makes my intel card init with more reliability than before. END K 10 svn:author V 2 bp K 8 svn:date V 27 2001-04-09T09:39:29.000000Z K 7 svn:log V 284 Add function prototypes and base module for kernel side iconv library. Add simple "xlat" converter which performs 8to8 table based conversion. Unicode converter will be added in the near future. Reviewed by: silence on arch@ Files placement reviewed by: bde Obtained from: smbfs END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T10:28:33.000000Z K 7 svn:log V 21 Remove a stale file. END K 10 svn:author V 5 greid K 8 svn:date V 27 2001-04-09T11:10:34.000000Z K 7 svn:log V 201 Change the behaviour of mixer slightly such that not explicitly specifying a mixer device will set/get the master volume levels PR: 25156 Submitted by: Tony Finch Reviewed by: cg, nik END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-09T12:04:44.000000Z K 7 svn:log V 209 enable the rate conversion feeder. the main benefit this gives for now is that via686 audio devices on motherboards with ac97 codecs that do not support vra will be able to use sample rates other than 48khz. END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-09T12:46:19.000000Z K 7 svn:log V 139 ``|'' should be more binding than ``!'' so that this isn't broken: if ! echo bla | wc -c ; then echo broken fi Obtained from: NetBSD END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-09T13:02:20.000000Z K 7 svn:log V 248 Merge in latest CSRG revisions: - .Fn and .Fc now print a final semicolon (`;') after a function declaration in the SYNOPSIS - .%I implemented - .At outputs ``AT&T UNIX'' if called without arguments - minor cleanup Obtained from: CSRG archives END K 10 svn:author V 8 gallatin K 8 svn:date V 27 2001-04-09T14:21:24.000000Z K 7 svn:log V 117 Document clear_error command on sable/lynx. Thanks to Timothy D. First for figuring this one out END K 10 svn:author V 7 sobomax K 8 svn:date V 27 2001-04-09T16:36:48.000000Z K 7 svn:log V 77 Whitespace only: Convert several tabs into spaces, to make style consistent. END K 10 svn:author V 3 jkh K 8 svn:date V 27 2001-04-09T16:49:50.000000Z K 7 svn:log V 90 Clarify that we're talking about KBytes per second vs kBits per second. Reminded by: bde END K 10 svn:author V 3 jkh K 8 svn:date V 27 2001-04-09T16:50:09.000000Z K 7 svn:log V 16 MFC: xfer stats END K 10 svn:author V 7 sobomax K 8 svn:date V 27 2001-04-09T16:55:32.000000Z K 7 svn:log V 76 Whitespace only: convert several tabs into spaces to make style consistent. END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-09T17:08:20.000000Z K 7 svn:log V 52 MFC: merge latest CSRG revisions. Approved by: jkh END K 10 svn:author V 7 sobomax K 8 svn:date V 27 2001-04-09T17:24:29.000000Z K 7 svn:log V 530 Allow user to omit font size specification when loading a font. In addition the following fixes had been made: - check the size of the font being loaded and compare it with possible sizes to minimise possibility of loading something that is not a fontfile at all and turning console screen into garbage; - prevent buffer overflow (and coredump as a result ) when loading valid uuencoded file with size that exceeds allocated buffer; - correct and improve several error messages. Approved by: -audit, -hackers (silently) END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:19:20.000000Z K 7 svn:log V 67 Add the Omni 56K Plus modem Submitted by: kazarov@izmiran.rssi.ru END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:19:41.000000Z K 7 svn:log V 7 Regen. END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:22:20.000000Z K 7 svn:log V 91 Update the description for the EPSON PID 0x010a. It seems to be reused in the 8700 series. END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:26:18.000000Z K 7 svn:log V 7 Regen. END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:44:11.000000Z K 7 svn:log V 74 Add the Abocom URE 450 ethernet adapter. Submitted by: dima@bog.msu.su END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:45:02.000000Z K 7 svn:log V 54 Again an ID that has been reused. Update description. END K 10 svn:author V 7 n_hibma K 8 svn:date V 27 2001-04-09T18:45:32.000000Z K 7 svn:log V 6 Rege. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-09T19:35:53.000000Z K 7 svn:log V 76 Add a comment out console line for AlphaServer 8200 and 8400 ("TurboLaser") END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T21:29:44.000000Z K 7 svn:log V 223 Add Marvell PHY support for 10/100/1000 LIVENGOOD_CU Intel NIC. Parag Patel did all of the grunt work, so he gets the credit. Register definitions and actions inferred from a Linux driver, so Intel also gets some 'credit'. END K 10 svn:author V 7 cvs2svn K 8 svn:date V 27 2001-04-09T21:29:45.000000Z K 7 svn:log V 68 This commit was manufactured by cvs2svn to create branch 'RELENG_4'. END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-09T21:33:47.000000Z K 7 svn:log V 38 comment out a boot-time debug message END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T21:34:52.000000Z K 7 svn:log V 30 We now depend on miibus_if.h. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-09T21:43:45.000000Z K 7 svn:log V 590 - One can now specify the decimal pid of a process to trace as a parameter. Since pid's are not in the kernel address space, this doesn't conflict with the funcionality of specifying an arbitrary frame pointer to the trace command. - If the first function of a backtrace maps to fork_trampoline, then this is a newly fork'd process that has not been executed yet, so just print out the first frame and then return for that case. - Lower the default count from 65535 to 1024. ddb doesn't trace into userland, and if the stack gets hosed and starts looping it's less annoying. END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T21:47:11.000000Z K 7 svn:log V 305 Pick up changes from Parag Patel and Kachun Lee, and self: 1. The offsets for some registers change in LIVENGOOD. Gratuitously. 2. Define LIVENGOOD and LIVENGOOD_CU part numbers. Add some more specific LIVENGOOD defaults. 3. Add definitions for PHY support for the copper LIVENGOOD part (10/100/1000). END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T21:48:50.000000Z K 7 svn:log V 122 Add in MII support for LICENGOOD copper part (10/100/1000). Add in some more flags for verbose as well as debug printing. END K 10 svn:author V 6 mjacob K 8 svn:date V 27 2001-04-09T21:54:15.000000Z K 7 svn:log V 375 Several things: 1. Pick up MII/PHY support for Livengood copper part (10/100/1000) from Parag Patel. It was a fairly complete but not quite platform independent job. 2. Finish silly offset differences that LIVENGOOD vs. WISEMAN registers have (so the !)$*!)$*!$ fiber LIVENGOOD now works too). 3. Ansify the source. So- we now suppor tthe PRO1000F and PRO1000T adapters. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-09T22:29:10.000000Z K 7 svn:log V 210 MFC: Fix pkg_version(1) version comparison, pkg_version -t, make test target for testing. Makefile 1.7->1.8 pkg_version.1 1.14->1.15 pkg_version.pl 1.17->1.18 test-pkg_version.sh 1.1 Approved by: jkh END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-09T22:34:05.000000Z K 7 svn:log V 515 Maintain a reference count on the witness struct. When the reference count drops to 0 in witness_destroy, set the w_name and w_file pointers to point to the string "(dead)" and the w_line field to 0. This way, if a mutex of a given name is used only in a module, then as long as all mutexes in the module are destroyed when the module is unloaded, witness will not maintain stale references to the mutex's name in the module's data section causing a panic later on when the w_name or w_file field's are examined. END K 10 svn:author V 6 dillon K 8 svn:date V 27 2001-04-09T23:35:19.000000Z K 7 svn:log V 611 Late edition: rpcinfo, mount, and umount are using different mechanisms to do portmap lookups. Some use tcp connections to portmap, others use udp. In anycase, using udp for RPC requests causes the old 'reply from the wrong IP address' problem to rear its ugly head so this commit is a quick hack to add the -h option to portmap. The semantics are the same as -h in nfsd (including backwards compatibility if you don't use -h). portmap no longer exists in current so this is 4.x only. I will check current's new rpc mechanism to see if it has the same problem and fix it there if it does. Approved: jkh END K 10 svn:author V 2 bp K 8 svn:date V 27 2001-04-10T00:56:19.000000Z K 7 svn:log V 52 Avoid endless recursion on panic. Reviewed by: jhb END K 10 svn:author V 2 dd K 8 svn:date V 27 2001-04-10T01:03:29.000000Z K 7 svn:log V 178 Make the list in the DIAGNOSTICS section "-tag" instead of "-diag": the former makes it more obvious as to there the error message starts and the explanation begins. PR: 26431 END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T03:47:40.000000Z K 7 svn:log V 62 Fix a comment within a comment warning due to a missing "*/". END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T03:55:19.000000Z K 7 svn:log V 70 Include so that read(2) and write(2) don't cause warnings. END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T04:11:50.000000Z K 7 svn:log V 247 To be consistent, use the __weak_reference macro from instead of #pragma weak to create weak definitions. This macro is improperly named, though, since a weak definition is not the same thing as a weak reference. Suggested by: bde END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T04:19:21.000000Z K 7 svn:log V 139 To be consistent, use the __weak_reference macro from instead of #pragma weak to create weak definitions. Suggested by: bde END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T04:22:24.000000Z K 7 svn:log V 89 Added a missing set of braces to a conditional that encompasses more than one statement. END K 10 svn:author V 8 deischen K 8 svn:date V 27 2001-04-10T04:25:49.000000Z K 7 svn:log V 173 Clean up a bit. Use the correct TAILQ link when walking the thread lists to free thread resources after a fork (in the child). Also remember to free the dead thread list. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-10T04:48:58.000000Z K 7 svn:log V 75 MFC: glob(3) limits, ftpd(8) usage of same, ntpd(8) vulnerability closed. END K 10 svn:author V 6 alfred K 8 svn:date V 27 2001-04-10T05:29:26.000000Z K 7 svn:log V 1136 Add more diagnostic output for failure. s/1518/ETHER_MAX_LEN Some style changes, add some braces, mostly residual from having a lot of debug hooks added while working on this driver. Bring in a plethora of changes from NetBSD: revision 1.58 date: 2001/03/08 11:07:08; author: ichiro; state: Exp; lines: +17 -1 it wait until busy flag disappears. it was able to prevent some cards with late initializing faling in wi_reset(). revision 1.41 date: 2000/10/13 19:15:08; author: jonathan; state: Exp; lines: +4 -2 Fix wi_intr() to avoid touching card registers during insert/remove events, when sharing an interrupt with other devices: check sc->sc_enabled, and drop the interrupt if its' off. revision 1.30 date: 2000/08/18 04:11:48; author: jhawk; state: Exp; lines: +4 -4 Copy wi_{dst,src}_addr from struct wi_frame into faked-up ether_header instead of addr1 and addr2. THis means that tcpdump -e will show the correct MAC address for communications with access points instead of showing the BSSID. In the future there should be 802.11 support for bpf/libpcap/tcpdump, but that is aways down the road. END K 10 svn:author V 2 bp K 8 svn:date V 27 2001-04-10T07:59:06.000000Z K 7 svn:log V 165 Import kernel part of SMB/CIFS requester. Add smbfs(CIFS) filesystem. Userland part will be in the ports tree for a while. Obtained from: smbfs-1.3.7-dev package. END K 10 svn:author V 2 bp K 8 svn:date V 27 2001-04-10T08:16:06.000000Z K 7 svn:log V 41 Add netsmb and smbfs include directories END K 10 svn:author V 5 brian K 8 svn:date V 27 2001-04-10T08:31:13.000000Z K 7 svn:log V 30 kldload ng_pppoe as necessary END K 10 svn:author V 8 mckusick K 8 svn:date V 27 2001-04-10T08:38:59.000000Z K 7 svn:log V 7485 Directory layout preference improvements from Grigoriy Orlov . His description of the problem and solution follow. My own tests show speedups on typical filesystem intensive workloads of 5% to 12% which is very impressive considering the small amount of code change involved. ------ One day I noticed that some file operations run much faster on small file systems then on big ones. I've looked at the ffs algorithms, thought about them, and redesigned the dirpref algorithm. First I want to describe the results of my tests. These results are old and I have improved the algorithm after these tests were done. Nevertheless they show how big the perfomance speedup may be. I have done two file/directory intensive tests on a two OpenBSD systems with old and new dirpref algorithm. The first test is "tar -xzf ports.tar.gz", the second is "rm -rf ports". The ports.tar.gz file is the ports collection from the OpenBSD 2.8 release. It contains 6596 directories and 13868 files. The test systems are: 1. Celeron-450, 128Mb, two IDE drives, the system at wd0, file system for test is at wd1. Size of test file system is 8 Gb, number of cg=991, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=35 2. PIII-600, 128Mb, two IBM DTLA-307045 IDE drives at i815e, the system at wd0, file system for test is at wd1. Size of test file system is 40 Gb, number of cg=5324, size of cg is 8m, block size = 8k, fragment size = 1k OpenBSD-current from Dec 2000 with BUFCACHEPERCENT=50 You can get more info about the test systems and methods at: http://www.ptci.ru/gluk/dirpref/old/dirpref.html Test Results tar -xzf ports.tar.gz rm -rf ports mode old dirpref new dirpref speedup old dirprefnew dirpref speedup First system normal 667 472 1.41 477 331 1.44 async 285 144 1.98 130 14 9.29 sync 768 616 1.25 477 334 1.43 softdep 413 252 1.64 241 38 6.34 Second system normal 329 81 4.06 263.5 93.5 2.81 async 302 25.7 11.75 112 2.26 49.56 sync 281 57.0 4.93 263 90.5 2.9 softdep 341 40.6 8.4 284 4.76 59.66 "old dirpref" and "new dirpref" columns give a test time in seconds. speedup - speed increasement in times, ie. old dirpref / new dirpref. ------ Algorithm description The old dirpref algorithm is described in comments: /* * Find a cylinder to place a directory. * * The policy implemented by this algorithm is to select from * among those cylinder groups with above the average number of * free inodes, the one with the smallest number of directories. */ A new directory is allocated in a different cylinder groups than its parent directory resulting in a directory tree that is spreaded across all the cylinder groups. This spreading out results in a non-optimal access to the directories and files. When we have a small filesystem it is not a problem but when the filesystem is big then perfomance degradation becomes very apparent. What I mean by a big file system ? 1. A big filesystem is a filesystem which occupy 20-30 or more percent of total drive space, i.e. first and last cylinder are physically located relatively far from each other. 2. It has a relatively large number of cylinder groups, for example more cylinder groups than 50% of the buffers in the buffer cache. The first results in long access times, while the second results in many buffers being used by metadata operations. Such operations use cylinder group blocks and on-disk inode blocks. The cylinder group block (fs->fs_cblkno) contains struct cg, inode and block bit maps. It is 2k in size for the default filesystem parameters. If new and parent directories are located in different cylinder groups then the system performs more input/output operations and uses more buffers. On filesystems with many cylinder groups, lots of cache buffers are used for metadata operations. My solution for this problem is very simple. I allocate many directories in one cylinder group. I also do some things, so that the new allocation method does not cause excessive fragmentation and all directory inodes will not be located at a location far from its file's inodes and data. The algorithm is: /* * Find a cylinder group to place a directory. * * The policy implemented by this algorithm is to allocate a * directory inode in the same cylinder group as its parent * directory, but also to reserve space for its files inodes * and data. Restrict the number of directories which may be * allocated one after another in the same cylinder group * without intervening allocation of files. * * If we allocate a first level directory then force allocation * in another cylinder group. */ My early versions of dirpref give me a good results for a wide range of file operations and different filesystem capacities except one case: those applications that create their entire directory structure first and only later fill this structure with files. My solution for such and similar cases is to limit a number of directories which may be created one after another in the same cylinder group without intervening file creations. For this purpose, I allocate an array of counters at mount time. This array is linked to the superblock fs->fs_contigdirs[cg]. Each time a directory is created the counter increases and each time a file is created the counter decreases. A 60Gb filesystem with 8mb/cg requires 10kb of memory for the counters array. The maxcontigdirs is a maximum number of directories which may be created without an intervening file creation. I found in my tests that the best performance occurs when I restrict the number of directories in one cylinder group such that all its files may be located in the same cylinder group. There may be some deterioration in performance if all the file inodes are in the same cylinder group as its containing directory, but their data partially resides in a different cylinder group. The maxcontigdirs value is calculated to try to prevent this condition. Since there is no way to know how many files and directories will be allocated later I added two optimization parameters in superblock/tunefs. They are: int32_t fs_avgfilesize; /* expected average file size */ int32_t fs_avgfpdir; /* expected # of files per directory */ These parameters have reasonable defaults but may be tweeked for special uses of a filesystem. They are only necessary in rare cases like better tuning a filesystem being used to store a squid cache. I have been using this algorithm for about 3 months. I have done a lot of testing on filesystems with different capacities, average filesize, average number of files per directory, and so on. I think this algorithm has no negative impact on filesystem perfomance. It works better than the default one in all cases. The new dirpref will greatly improve untarring/removing/coping of big directories, decrease load on cvs servers and much more. The new dirpref doesn't speedup a compilation process, but also doesn't slow it down. Obtained from: Grigoriy Orlov END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-10T10:35:29.000000Z K 7 svn:log V 70 Correct some cut-n-paste errors. Also embellish the UP1100 a little. END K 10 svn:author V 3 nik K 8 svn:date V 27 2001-04-10T10:36:44.000000Z K 7 svn:log V 168 Add information about the new options to newfs and tunefs which set the expected average file size and number of files per directory. Could do with some fleshing out. END K 10 svn:author V 6 obrien K 8 svn:date V 27 2001-04-10T10:38:35.000000Z K 7 svn:log V 136 MFC revs 1.79 (cut-n-paste + embellish), 1.78 (sable/lynx `clear_error') Requested by: wilko Approved by: jkh (vicariously thru wilko) END K 10 svn:author V 2 cg K 8 svn:date V 27 2001-04-10T13:52:26.000000Z K 7 svn:log V 39 lock the mutex, not the softc pointer. END K 10 svn:author V 5 greid K 8 svn:date V 27 2001-04-10T14:28:21.000000Z K 7 svn:log V 254 Add another card to the list of Neomagic 256AV's which don't have AC97 codecs. Also, add some additional code to check for future cards without this feature - attempting to initialise them as AC97 cards will hang the machine. PR: 26427 Reviewed by: cg END K 10 svn:author V 5 greid K 8 svn:date V 27 2001-04-10T15:56:45.000000Z K 7 svn:log V 66 Add another ID to the badcards (non-ac97) table Approved by: jkh END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-10T16:15:01.000000Z K 7 svn:log V 46 vnconfig(8) -> mdconfig(8). Reviewed by: phk END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-10T16:29:41.000000Z K 7 svn:log V 28 vnconfig(8) -> mdconfig(8). END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-10T17:04:17.000000Z K 7 svn:log V 136 MFC: VLAN bugfixes, Vinum bugfixes, pkg_version comparison fix and -t option, portmap -h, GNATS update. Approved by: jkh (implicitly) END K 10 svn:author V 2 ru K 8 svn:date V 27 2001-04-10T17:04:31.000000Z K 7 svn:log V 42 psroff(1) has never been a part of Groff. END K 10 svn:author V 4 bmah K 8 svn:date V 27 2001-04-10T17:06:50.000000Z K 7 svn:log V 62 New release notes: FFS dirpref speedup, GNATS 3.113, BSDPAN. END K 10 svn:author V 3 jhb K 8 svn:date V 27 2001-04-10T19:01:30.000000Z K 7 svn:log V 56 NOBLOCKRANDOM doesn't exist anymore as a kernel option. END