3a39a57 | Ken Smith | 08 August 2012, 18:39:29 UTC | Prep for 9.1-RC1 builds. Approved by: re (implicit) | 08 August 2012, 18:39:29 UTC |
e03e410 | Ken Smith | 06 August 2012, 00:15:49 UTC | Update __FreeBSD_version to reflect this being the 9.1-RELEASE release branch. Approved by: re (implicit) | 06 August 2012, 00:15:49 UTC |
2c36bc5 | Maksim Yevmenkin | 02 August 2012, 20:49:11 UTC | MFC r238622 Allow to specify no source-address-selection policy Approved by: re (kib) | 02 August 2012, 20:49:11 UTC |
c527d37 | Maksim Yevmenkin | 02 August 2012, 20:48:22 UTC | MFC r238599 Return zero from get_addrselectpolicy() when no source-address-selection policy is installed. Approved by: re (kib) | 02 August 2012, 20:48:22 UTC |
6c19383 | Andrey Chernov | 02 August 2012, 15:13:12 UTC | MFC r238795 Try to avoid all files dependence on the modification time of the large and often modified directory created symbolic links points to - it cause unnecessary full rebuilds each time make runs when directory is changed. So do it only if symbolic link does not exists, which usually means that objdir is clean anyway. Approved by: re (kib) | 02 August 2012, 15:13:12 UTC |
2c32108 | John Baldwin | 30 July 2012, 20:40:14 UTC | MFC 238164: Add another PS/2 keyboard PNP ID. This ID is listed as "Reserved by Microsoft" in the standard PNP ID table, but has been seen in the wild on at least one laptop. Approved by: re (kib) | 30 July 2012, 20:40:14 UTC |
32d7b1d | Jim Harris | 30 July 2012, 15:30:42 UTC | MFC r238755: Add rmb() to tsc_read_##x to enforce serialization of rdtsc captures. Intel Architecture Manual specifies that rdtsc instruction is not serialized, so without this change, TSC synchronization test would periodically fail, resulting in use of HPET timecounter instead of TSC-low. This caused severe performance degradation (40-50%) when running high IO/s workloads due to HPET MMIO reads and GEOM stat collection. Tests on Xeon E5-2600 (Sandy Bridge) 8C systems were seeing TSC synchronization fail approximately 20% of the time. Sponsored by: Intel Approved by: re (kib) | 30 July 2012, 15:30:42 UTC |
880f81e | Jilles Tjoelker | 30 July 2012, 12:25:20 UTC | MFC r237939: ctype_l(3): Note that not all these functions are specified by POSIX.1-2008. The function isascii_l() is not in POSIX even though isascii() is, probably because isascii() is marked as obsolete. The other functions, like digittoint_l() and ishexnumber_l(), are FreeBSD-specific just like their non-_l versions. Approved by: re (kib) | 30 July 2012, 12:25:20 UTC |
618e2c7 | Marius Strobl | 30 July 2012, 11:11:05 UTC | Pull the tier-2 card and change the sparc64 ZFS loader to no longer probe all diskN aliases for providers (which more or less corresponds to how the x86 version behaves) but instead probe only those listed in the boot-device OFW environment variable. This has the following advantages: - avoids otherwise unavoidable OFW warnings about failures to open disks for which aliases exist but no actual hardware is connected - avoids issues due to different diskN naming schemes - aligns us with Solaris Approved by: re (kib) | 30 July 2012, 11:11:05 UTC |
ed91c4f | Glen Barber | 29 July 2012, 18:52:01 UTC | MFC r238707: Document the following in rc.conf.5: - rtsold_enable - rtsold_flags - rtsol_flags Approved by: re (kib) | 29 July 2012, 18:52:01 UTC |
d7e4c53 | Bjoern A. Zeeb | 28 July 2012, 23:11:09 UTC | MFC r238871: Hardcode the loopback rx/tx checkum options for IPv6 to on without checking. This allows the FreeBSD 9.1 release process to move forward. Work around the problem that loopback connections to local addresses not on loopback interfaces and not on interfaces w/ IPv6 checksum offloading enabled would not work. A proper fix to allow us to disable the "checksum offload" on loopback for testing, measurements, ... as we allow for IPv4 needs to put in place later. PR: kern/170070 Approved by: re (kib) | 28 July 2012, 23:11:09 UTC |
875f343 | Michael Tuexen | 28 July 2012, 22:42:52 UTC | MFC r238790: Fix the sctp_sockstore union such that userland programs don't depend on INET and/or INET6 to be defined and in-tune with how the kernel was compiled. Approved by: re (kib) | 28 July 2012, 22:42:52 UTC |
4083739 | Hiroki Sato | 28 July 2012, 21:43:29 UTC | MFC r235285: Skip nd6 line with no warning message when the system does not support INET6. Spotted by: flo Approved by: re (kib) | 28 July 2012, 21:43:29 UTC |
3388d28 | Dimitry Andric | 28 July 2012, 11:28:00 UTC | MFC r238721: When WITHOUT_CLANG is being used, also clean out the clang 3.1 headers in OptionalObsoleteFiles.inc. PR: misc/169902 Submitted by: Thomas Eberhardt <sneakywumpus@googlemail.com> Approved by: re (kib) | 28 July 2012, 11:28:00 UTC |
3d0350c | Mikolaj Golub | 27 July 2012, 19:56:36 UTC | MFC r238753: Align the header with output. Approved by: re (kib) | 27 July 2012, 19:56:36 UTC |
ddd441d | Baptiste Daroussin | 27 July 2012, 09:35:06 UTC | MFC: r238461 Make pkg bootstrap program ask for confirmation before proceeding. The previous behaviour was to silently download and install the pkg package, without ever telling user about what it was doing and why. Approved by: re, des (mentor) | 27 July 2012, 09:35:06 UTC |
cf8ac0e | Ruslan Ermilov | 26 July 2012, 19:18:26 UTC | Synced mdoc(7) with head. Approved by: re (kib) | 26 July 2012, 19:18:26 UTC |
37d1246 | Christian Brueffer | 26 July 2012, 09:13:47 UTC | MFC: r238393 Add and utilize defines for the ATA device register. Approved by: re (kib) | 26 July 2012, 09:13:47 UTC |
94bc18f | Bjoern A. Zeeb | 25 July 2012, 19:18:28 UTC | MFC r238700: Fix a bug introduced in r221129 that leads to a panic when using bundled SAs. For now allow same address family bundles. PR: kern/164400 Approved by: re (kib) | 25 July 2012, 19:18:28 UTC |
c119c3f | Warren Block | 25 July 2012, 13:05:11 UTC | MFC r238705: Correct ugen.4 to show that it has been integrated into usb(4). Also fix some punctuation errors. Approved by: re (kib) | 25 July 2012, 13:05:11 UTC |
7d23bad | Christian Brueffer | 25 July 2012, 10:55:14 UTC | MFC: r238574 Fix a small memory leak in mpssas_get_sata_identify(). The change has been submitted upstream as well. Approved by: re (kib) | 25 July 2012, 10:55:14 UTC |
8caf6cc | Doug Barton | 24 July 2012, 22:32:03 UTC | MFV r238744: Heavy DNSSEC Validation Load Can Cause a "Bad Cache" Assertion Failure in BIND9 High numbers of queries with DNSSEC validation enabled can cause an assertion failure in named, caused by using a "bad cache" data structure before it has been initialized. CVE: CVE-2012-3817 Posting date: 24 July, 2012 Approved by: re (kib) | 24 July 2012, 22:32:03 UTC |
79045e5 | Alexander Motin | 23 July 2012, 15:19:21 UTC | MFC r238666: Fix typo in bzero() length argument during sense fetching. For me it at least fixed CD burning in PIO mode. Approved by: re (kib) | 23 July 2012, 15:19:21 UTC |
2904730 | David Xu | 23 July 2012, 09:33:31 UTC | Merge r238637,r238640,r238641,r238642: ------------------------------------------------------------------------ r238637 | davidxu | 2012-07-20 09:56:14 +0800 (Fri, 20 Jul 2012) | 6 lines Don't forget to release a thread reference count, replace _thr_ref_add() with _thr_find_thread(), so reference count is no longer needed. ------------------------------------------------------------------------ r238640 | davidxu | 2012-07-20 11:00:41 +0800 (Fri, 20 Jul 2012) | 2 lines Eliminate duplicated code. ------------------------------------------------------------------------ r238641 | davidxu | 2012-07-20 11:16:52 +0800 (Fri, 20 Jul 2012) | 2 lines Eliminate duplicated code. ------------------------------------------------------------------------ r238642 | davidxu | 2012-07-20 11:22:17 +0800 (Fri, 20 Jul 2012) | 2 lines Don't assign same value. Approved by: re (kib) | 23 July 2012, 09:33:31 UTC |
af7fde6 | Gleb Smirnoff | 23 July 2012, 09:19:14 UTC | Merge from head r238572, r238573: ------------------------------------------------------------------------ r238572 | glebius | 2012-07-18 12:41:00 +0400 (ср, 18 июл 2012) | 3 lines When traversing global in_ifaddr list in the IFP_TO_IA() macro, we need to obtain IN_IFADDR_RLOCK(). ------------------------------------------------------------------------ r238573 | glebius | 2012-07-18 12:58:30 +0400 (ср, 18 июл 2012) | 5 lines Plug a reference leak: before doing 'goto again' we need to unref ia->ia_ifa if there is any. Submitted by: Andrey Zonov <andrey zonov.org> Approved by: re (kib) | 23 July 2012, 09:19:14 UTC |
2cece21 | Marius Strobl | 22 July 2012, 14:32:49 UTC | MFC: r238621 Revert the use of BUS_DMA_ALLOCNOW when creating the DMA tag for user data introduced in r236061 (MFC'ed to stable/9 in r237186). Using that flag doesn't make that much sense on this case as the DMA maps using it are also created during sym_pci_attach(). Moreover, due to the maxsegsz parameter used, doing so may exhaust the bounce pages pool on architectures requiring bounce pages. [1] While at it, use a slightly more appropriate maxsegsz parameter. PR: 169526 Submitted by: Mike Watters [1] Approved by: re (kib) | 22 July 2012, 14:32:49 UTC |
18f50d0 | Isabell Long | 22 July 2012, 11:00:02 UTC | MFC r238063: - Make ipfw's sched rules case insensitive, for user-friendliness. - Add a note to the ipfw(8) man page about the rules no longer being case sensitive. - Fix some typos in the man page. PR: docs/164772 Reviewed by: bz Approved by: gavin Approved by: re (kib) | 22 July 2012, 11:00:02 UTC |
e39333e | Konstantin Belousov | 22 July 2012, 00:44:22 UTC | MFC r238617: Fix several reads beyond the mapped first page of the binary in the ELF parser. Specifically, do not allow note reader and interpreter path comparision in the brandelf code to read past end of the page. This may happen if specially crafter ELF image is activated. Approved by: re (hrs) | 22 July 2012, 00:44:22 UTC |
553d3b4 | Glen Barber | 21 July 2012, 01:07:05 UTC | MFC r231257, r238606: r231257 (kevlo): Mention run(4) r238606: Xref mwl(4) and rum(4) in hostapd(8). Approved by: re (kib) | 21 July 2012, 01:07:05 UTC |
a7b260a | Michael Tuexen | 20 July 2012, 09:49:50 UTC | MFC r238550: Fix a refcount bug when freeing an association. While there: Change code to be consistent. Discussed with: rrs Approved by: re (kib) | 20 July 2012, 09:49:50 UTC |
7212d7c | Gabor Pali | 20 July 2012, 06:50:57 UTC | MFC r238527: - Add support for displaying process stack memory regions. Approved by: re (kib), rwatson | 20 July 2012, 06:50:57 UTC |
fa14e35 | Michael Tuexen | 19 July 2012, 09:32:59 UTC | MFC r238501, r 238514: Changes which improve compilation if neither INET nor INET6 is defined. Allow netstat to be build if INET is not defined in the kernel. Thanks to Garrett Cooper for reporting the issue. This also fixes a bug where netstat doesn't work correctly for SCTP on INET only (no INET6) kernels. Approved by: re@ | 19 July 2012, 09:32:59 UTC |
dbabe21 | Gleb Smirnoff | 18 July 2012, 16:13:03 UTC | Merge r230119, r238498 from head: ------------------------------------------------------------------------ r230119 | csjp | 2012-01-15 02:51:34 +0400 (вс, 15 янв 2012) | 9 lines Revert to the old behavior of allocating table/table entries using M_NOWAIT. Currently, the code allows for sleeping in the ioctl path to guarantee allocation. However code also handles ENOMEM gracefully, so propagate this error back to user-space, rather than sleeping while holding the global pf mutex. Reviewed by: glebius Discussed with: bz ------------------------------------------------------------------------ r238498 | glebius | 2012-07-15 23:10:00 +0400 (вс, 15 июл 2012) | 2 lines Use M_NOWAIT while holding the pf giant lock. Approved by: re (kib) | 18 July 2012, 16:13:03 UTC |
702c4a5 | Christian Brueffer | 18 July 2012, 10:56:48 UTC | MFC: r238363 - fix description of SECURITY_FREEZE_LOCK - add missing underscores to SECURITY_DISABLE_PASSWORD Approved by: re (hrs) | 18 July 2012, 10:56:48 UTC |
8e22200 | Christian Brueffer | 18 July 2012, 10:42:30 UTC | MFC: r238486 Fix typo in a message. Approved by: re (hrs) | 18 July 2012, 10:42:30 UTC |
5e205e7 | Michael Tuexen | 18 July 2012, 09:16:59 UTC | MFC r238475: #ifdef INET and INET6 consistently. This also fixes a bug, where it was done wrong. Approved by: re@ | 18 July 2012, 09:16:59 UTC |
e19f9d6 | Mateusz Guzik | 18 July 2012, 04:52:37 UTC | MFC r238220, r238272: Unbreak handling of descriptors opened with O_EXEC by fexecve(2). While here return EBADF for descriptors opened for writing (previously it was ETXTBSY). Add fgetvp_exec function which performs appropriate checks. == Follow-up commit to r238220: Pass only FEXEC (instead of FREAD|FEXEC) in fgetvp_exec. _fget has to check for !FWRITE anyway and may as well know about FREAD. Make _fget code a bit more readable by converting permission checking from if() to switch(). Assert that correct permission flags are passed. PR: kern/169651 In collaboration with: kib Approved by: re (hrs), trasz (mentor) | 18 July 2012, 04:52:37 UTC |
59b3442 | Alexander V. Chernikov | 17 July 2012, 19:29:32 UTC | MFC r237479, r238265 Finally fix lookup (account remaining '\0') and deletion (provide valid key length for radix lookup). Submitted by: Ihor Kaharlichenko<madkinder at gmail.com> (prev version) Approved by: re(hrs), kib(mentor) Sponsored by: Shtorm ISP | 17 July 2012, 19:29:32 UTC |
a5c3c08 | Konstantin Belousov | 17 July 2012, 19:00:46 UTC | MFC r237203,r237223 (by phk): Fix writes to the video memory of the VGA to always use half-word sized writes. Approved by: re (hrs), phk | 17 July 2012, 19:00:46 UTC |
85c2dd5 | Dimitry Andric | 17 July 2012, 18:04:34 UTC | MFC r238429: Pull in r159895 from upstream clang trunk: When marking virtual functions as used for a class' vtable, mark all functions which will appear in the vtable as used, not just those ones which were declared within the class itself. Fixes an issue reported as comment#3 in PR12763 -- we sometimes assert in codegen if we try to emit a reference to a function declaration which we've not marked as referenced. This also matches gcc's observed behavior. This should fix clang assertions when building certain components of the LibreOffice port. Approved by: re (kib) | 17 July 2012, 18:04:34 UTC |
81729ca | Michael Tuexen | 17 July 2012, 11:33:24 UTC | MFC r238458: Provide the correct notification type (SCTP_SEND_FAILED_EVENT) for unsent messages. Approved by: re@ | 17 July 2012, 11:33:24 UTC |
d7f1b4f | Michael Tuexen | 17 July 2012, 11:18:39 UTC | MFC r238454: Bugfix: Send up a COMM UP notification for active 1-to-1 style sockets also in the case where the assoc comes up due to a remotely started handshake (collision case). Approved by: re@ | 17 July 2012, 11:18:39 UTC |
5cbdb67 | Andrey V. Elsukov | 16 July 2012, 08:16:41 UTC | MFC r238281: We don't need to check the result of sending signal when R flag is specified. Submitted by: Ilya A. Arkhipov Approved by: re (kib) | 16 July 2012, 08:16:41 UTC |
3d9e0c1 | David Xu | 16 July 2012, 02:10:26 UTC | MFC r238328: Executing CPUID with EAX set to 1 to actually get feature flags. PR: 169730 Approved by: re (kib) | 16 July 2012, 02:10:26 UTC |
742a885 | Andriy Gapon | 15 July 2012, 20:38:02 UTC | MFC r238168,238169: further dtrace disassembler fixes for long nop decoding - segment-override prefixes are not invalid in long mode - fix nopw (0f 1f) behavior with respect to modifiers Reported by: emaste Tested by: emaste Obtained from: Illumos commit 13442:4adbe6de60c8 Approved by: re (kib) | 15 July 2012, 20:38:02 UTC |
3f71778 | Gleb Smirnoff | 15 July 2012, 19:52:23 UTC | Merge r237929, r237930 from head: ------------------------------------------------------------------------ r237929 | glebius | 2012-07-01 19:30:43 +0400 (вс, 01 июл 2012) | 6 lines In g_mirror_regular_request() upon successful delivery treat BIO_DELETE requests same way as BIO_WRITE removing them from queue. This fixes panic with BIO_DELETE operations on geom_mirror. Reviewed by: pjd ------------------------------------------------------------------------ r237930 | glebius | 2012-07-01 19:43:52 +0400 (вс, 01 июл 2012) | 20 lines Make geom_mirror more friendly to SSDs. To properly support TRIM, we need to pass BIO_DELETE requests down to providers that support it. Also, we need to announce our support for BIO_DELETE to upper consumer. This requires: - In g_mirror_start() return true for "GEOM::candelete" request. - In g_mirror_init_disk() probe below provider for "GEOM::candelete" attribute, and mark disk with a flag if it does support BIO_DELETE. - In g_mirror_register_request() distribute BIO_DELETE requests only to those disks, that do support it. Note that we announce "GEOM::candelete" as true unconditionally of whether we have TRIM-capable media down below or not. This is made intentionally, because upper consumer (usually UFS) requests the attribite only once at mount time. And if user ever migrates his mirror from HDDs to SSDs, then he/she would get TRIM working without remounting filesystem. Reviewed by: pjd Approved by: re (kib) | 15 July 2012, 19:52:23 UTC |
9539623 | Dag-Erling Smorgrav | 15 July 2012, 12:13:15 UTC | MFH r237270: no need to specify passive mode, it is the default. Approved by: re | 15 July 2012, 12:13:15 UTC |
d1462cb | Dag-Erling Smorgrav | 15 July 2012, 12:00:47 UTC | MFH r237269: switch default password hash to sha512. Approved by: re | 15 July 2012, 12:00:47 UTC |
2fa4336 | Isabell Long | 15 July 2012, 11:39:56 UTC | MFC r237573: Add more locale-specific functions to the relevant man pages and Makefiles: - libc/stdtime/strftime.3 - libc/stdtime/strptime.3 - libc/stdlib/strfmon.3 Approved by: gabor (mentor) Approved by: re (kib) | 15 July 2012, 11:39:56 UTC |
d07669c | Dag-Erling Smorgrav | 15 July 2012, 11:39:35 UTC | MFH r236751: document sha256 / sha512 support MFH r236892: remove mention of auth.conf from programs that don't use it MFH r236963: remove dead code relating to auth.conf MFH r236965 r236966 r236967 r237005 r237006 r237011: retire auth.conf Approved by: re | 15 July 2012, 11:39:35 UTC |
20e6541 | Bjoern A. Zeeb | 15 July 2012, 11:27:15 UTC | MFC r238248: Implement handling of "atomic fragements" as outlined in draft-gont-6man-ipv6-atomic-fragments to mitigate one class of possible fragmentation-based attacks. Approved by: re (kib) | 15 July 2012, 11:27:15 UTC |
087d978 | Christian Brueffer | 15 July 2012, 11:26:24 UTC | MFC: r234577 Remove duplicate -l description. Approved by: re (kib) | 15 July 2012, 11:26:24 UTC |
0a17640 | Bjoern A. Zeeb | 15 July 2012, 11:13:09 UTC | MFC r238222: As mentioned in the commit message of r237571 (copied from a prototype patch of mine) also check if the 2nd in6_setscope() failed and return the error in that case. Approved by: re (kib) | 15 July 2012, 11:13:09 UTC |
2450d9c | Christian Brueffer | 15 July 2012, 10:55:43 UTC | MFC: r237921 Similar to all.log, mention that /var/log/console.log has to be created and chmod'ed to make logging work. Approved by: re (kib) | 15 July 2012, 10:55:43 UTC |
171f846 | Navdeep Parhar | 15 July 2012, 05:24:19 UTC | MFC r238313: Fix a bug in code that calculates the number of the first interrupt vector for a port. This affected the gigabit ports of T422 cards (the ones with 2x10G ports and 2x1G ports). Approved by: re (kib) | 15 July 2012, 05:24:19 UTC |
9e40c87 | Ken Smith | 15 July 2012, 03:43:56 UTC | 9.1-BETA1 builds done, shift back to PRERELEASE for the rest of the 9.1 release cycle. Approved by: re (implicit) | 15 July 2012, 03:43:56 UTC |
4453249 | Ken Smith | 10 July 2012, 19:10:39 UTC | Ready for 9.1-BETA1 builds. Approved by: re (implicit) | 10 July 2012, 19:10:39 UTC |
89054a8 | Ken Smith | 10 July 2012, 19:08:56 UTC | Adjustments for 9.0 -> 9.1. Approved by: re (implicit) | 10 July 2012, 19:08:56 UTC |
7eaca5c | Ken Smith | 10 July 2012, 19:05:27 UTC | Shift to "FreeBSD 9.1" for what gets printed in manual pages. Approved by: re (implicit) | 10 July 2012, 19:05:27 UTC |
4647484 | Navdeep Parhar | 09 July 2012, 17:25:56 UTC | Re-enable IFCAP_TSO6 in cxgb(4) and cxgbe(4) in stable/9. The kernel changes needed for all this to work have now been MFC'd to 9 by bz@. This is a direct commit to stable/9 that removes earlier changes made to drivers in this branch only. Approved by: re (kib) | 09 July 2012, 17:25:56 UTC |
f72cfb7 | Michael Tuexen | 09 July 2012, 13:21:49 UTC | MFC r238121: Use CSUM_SCTP_IPV6 for IPv6. Approved by: re@ | 09 July 2012, 13:21:49 UTC |
e8d543d | Alexander V. Chernikov | 09 July 2012, 11:53:06 UTC | Merge r237162, r237163, r237164, r237226. Fix improper L4 header handling for IPv6 packets passed via DLT_RAW. Set netflow v9 observation domain value to fib number instead of node id. This fixes multi-fib netflow v9 export. Use time_uptime instead of getnanotime for accouting integer number of seconds. Simplify IP pointer recovery in case of mbuf reallocation. Approved by: ae(mentor) Approved by: re | 09 July 2012, 11:53:06 UTC |
e29d834 | Konstantin Belousov | 09 July 2012, 10:33:53 UTC | MFC r237988: Issue proper diagnostic on the short writes, also consider the case of write reporting 0 bytes as short write. Approved by: re (hrs) | 09 July 2012, 10:33:53 UTC |
b808b49 | Pedro Giffuni | 09 July 2012, 09:11:07 UTC | MFC r238173: Fix issue resizing bin/sh This partially reverts some changes from r237448 that are causing breakage when resizing under bin/sh . Reverted changes from NetBSD are: Mar 10 20:46:15 2009 - editline.3 read.c make el_gets set the count to -1 on error to distinguish between EOF and error. Feb 19 15:20:22 2009 - read.c sig.c sig.h reset and redraw on sigcont. From Anon Ymous. Feb 15 21:24:13 2009 don't restart on EINTR, instead return NULL immediately. From Anon Ymous PR: 169603 Reported by: Peter Jeremy, David Shao MFC after: 3 days Approved by: re | 09 July 2012, 09:11:07 UTC |
f9def65 | Robert N. M. Watson | 09 July 2012, 08:42:54 UTC | Merge r234032 from head to stable/9: When allocation of labels on files is implicitly disabled due to MAC policy configuration, avoid leaking resources following failed calls to get and set MAC labels by file descriptor. Reported by: Mateusz Guzik <mjguzik at gmail.com> + clang scan-build Approved by: re (kib) | 09 July 2012, 08:42:54 UTC |
82806d8 | Robert N. M. Watson | 09 July 2012, 08:27:04 UTC | Merge r229521 from head to stable/9: Add missing -w to kgdb(1)'s usage(). Sponsored by: Adara Networks, Inc. Approved by: re (kib) | 09 July 2012, 08:27:04 UTC |
aa167d9 | Robert N. M. Watson | 08 July 2012, 21:21:45 UTC | Merge r231712 from head to stable/9: When initialising the CP0 status register during boot on 64-bit MIPS, set all three of the kernel, supervisor, and user-mode 64-bit mode flags. While FreeBSD does not currently use the supervisor ring (and hence this is effectively a NOP on most systems), doing this avoids triggering an exception on 64-bit MIPS CPUs that don't support 32-bit compatibility mode, and therefore don't allow clearing the SX bit. Reviewed by: gonzo Sponsored by: DARPA, SRI International Approved by: re (kib) | 08 July 2012, 21:21:45 UTC |
ff36e2a | Robert N. M. Watson | 08 July 2012, 20:41:54 UTC | Merge r234034 from head to stable/9: Merge a local fix to OpenBSM's libauditd to avoid a directory descriptor leak when iterating over possible audit trail directories. This fix will be merged upstream in an identical form, but hasn't yet appeared in an OpenBSM release. Submitted by: guido Obtained from: TrustedBSD Project Approved by: re (kib) | 08 July 2012, 20:41:54 UTC |
d0c6415 | Jack F Vogel | 08 July 2012, 20:39:00 UTC | MFC of 238149 Approved by:re | 08 July 2012, 20:39:00 UTC |
1a7c0fe | Jack F Vogel | 08 July 2012, 20:35:56 UTC | MFC of the e1000 drivers: 236406,238148,238151,238181, and 238214 Approved by:re | 08 July 2012, 20:35:56 UTC |
6e5f02d | Mikolaj Golub | 08 July 2012, 19:29:08 UTC | Fix RTTVAR scale in net.inet.tcp.hostcache.list sysctl. Reviewed by: andre Approved by: re (kib) | 08 July 2012, 19:29:08 UTC |
05d7858 | Michael Tuexen | 08 July 2012, 16:14:42 UTC | MFC r237049: Pass flowid explicitly through the stack instead of taking it from the mbuf chain at different places. While there: Fix several bugs related to VRFs. Approved by: re@ | 08 July 2012, 16:14:42 UTC |
6481434 | Michael Tuexen | 08 July 2012, 16:11:16 UTC | MFC r236956: Unify the sending of ABORT, SHUTDOWN-COMPLETE and ERROR chunks. While there: Fix also some minor bugs and prepare for SCTP/DTLS. Approved by: re@ | 08 July 2012, 16:11:16 UTC |
fd3ea57 | Michael Tuexen | 08 July 2012, 16:00:45 UTC | MFC r236332: Seperate SCTP checksum offloading for IPv4 and IPv6. While there: remove some trainling whitespaces. Approved by: re@ | 08 July 2012, 16:00:45 UTC |
b4c6f13 | Michael Tuexen | 08 July 2012, 15:37:58 UTC | MFC r236087: Get rid of SCTP specific code to avoid CRC32C computations on loopback. Just use offloading. Approved by: re | 08 July 2012, 15:37:58 UTC |
6796175 | Bjoern A. Zeeb | 08 July 2012, 14:21:36 UTC | MFC r236170,236176: It turns out that too many drivers are not only parsing the L2/3/4 headers for TSO but also for generic checksum offloading. Ideally we would only have one common function shared amongst all drivers, and perhaps when updating them for IPv6 we should introduce that. Eventually we should provide the meta information along with mbufs to avoid (re-)parsing entirely. To not break IPv6 (checksums and offload) and to be able to MFC the changes without risking to hurt 3rd party drivers, duplicate the v4 framework, as other OSes have done as well. Introduce interface capability flags for TX/RX checksum offload with IPv6, to allow independent toggling (where possible). Add CSUM_*_IPV6 flags for UDP/TCP over IPv6, and reserve further for SCTP, and IPv6 fragmentation. Define CSUM_DELAY_DATA_IPV6 as we do for legacy IP and add an alias for CSUM_DATA_VALID_IPV6. This pretty much brings IPv6 handling in line with IPv4. TSO is still handled in a different way and not via if_hwassist. Update ifconfig to allow (un)setting of the new capability flags. Update loopback to announce the new capabilities and if_hwassist flags. Individual driver updates will have to follow, as will SCTP. Removed the IFCAP_ prefix when printing the IPv6 checksum capabilities. Approved by: re | 08 July 2012, 14:21:36 UTC |
8f8201f | Bjoern A. Zeeb | 08 July 2012, 12:39:02 UTC | MFC r236093: Trim the extra $FreeBSD$ from the comment below the license. We use the __FBSDID() macro on the file now instead. Approved by: re | 08 July 2012, 12:39:02 UTC |
3c8a973 | Bjoern A. Zeeb | 08 July 2012, 12:36:44 UTC | MFC r235950: Factor out the tcp_hc_getmtu() call. As the comments say it applies to both v4 and v6, so only write it once making it easier to read the protocol family specifc code. Approved by: re | 08 July 2012, 12:36:44 UTC |
a595069 | Bjoern A. Zeeb | 08 July 2012, 12:34:12 UTC | MFC r235953: No need to hold the (expensive) rt lock over (expensive) logging. Approved by: re | 08 July 2012, 12:34:12 UTC |
3aee8ef | Bjoern A. Zeeb | 08 July 2012, 12:30:25 UTC | MFC r235986: Use M_ZERO with malloc rather than calling bzero() ourselves. Change if () panic() checks to KASSERT()s as they are only catching invariants in code flow but not dependent on network input/output. Move initial assigments indirecting pointers after the lock has been aquired. Passing layer boundries, reset M_PROTOFLAGS. Remove a NULL assignment before free. Approved by: re | 08 July 2012, 12:30:25 UTC |
dddac2b | Bjoern A. Zeeb | 08 July 2012, 12:26:32 UTC | MFC r235985: Properly protect the inp read access when handling the control code. In the past this was expensive but given the rlock it's not so much anymore. Approved by: re | 08 July 2012, 12:26:32 UTC |
3b26bf7 | Bjoern A. Zeeb | 08 July 2012, 12:23:27 UTC | MFC r235981,236394 In case forwarding is turned on for a given address family, refuse to queue the packet for LRO and tell the driver to directly pass it on. This avoids re-assembly and later re-fragmentation problems when forwarding. It's not the best solution but the simplest and most effective for the moment. Make TCP LRO work properly with VIMAGE kernels rather than just panicing. There's no VIMAGE context set there yet as this is before if_ethersubr.c. Approved by: re | 08 July 2012, 12:23:27 UTC |
b9f9320 | Bjoern A. Zeeb | 08 July 2012, 12:17:56 UTC | MFC r235962: Factor out Hop-By-Hop option processing. It's still not heavily used, it reduces the footprint of ip6_input() and makes ip6_input() more readable. Approved by: re | 08 July 2012, 12:17:56 UTC |
26ec9f1 | Bjoern A. Zeeb | 08 July 2012, 12:12:34 UTC | MFC r235961: Add code to handle pre-checked TCP checksums as indicated by mbuf flags to save the entire computation for validation if not needed. In the IPv6 TCP output path only compute the pseudo-header checksum, set the checksum offset in the mbuf field along the appropriate flag as done in IPv4. In tcp_respond() just initialize the IPv6 payload length to 0 as ip6_output() will properly set it. Approved by: re | 08 July 2012, 12:12:34 UTC |
aeba9f8 | Bjoern A. Zeeb | 08 July 2012, 12:01:11 UTC | MFC r235960: Simple yet effective change enabling checksum "offload" on loopback for IPv6 to avoid expensive computations. Approved by: re | 08 July 2012, 12:01:11 UTC |
b4aa243 | Bjoern A. Zeeb | 08 July 2012, 11:57:11 UTC | MFC r235959: Defer checksum calulations on UDP6 output and respect the mbuf flags set by NICs having done checksum validation for us already, thus saving the computing time in the input path as well. Approved by: re | 08 July 2012, 11:57:11 UTC |
a60ccff | Bjoern A. Zeeb | 08 July 2012, 11:53:13 UTC | MFC r235958,236130: Add support for delayed checksum calculations in the IPv6 output path. We currently cannot offload to the card if we add extension headers (which incl. fragmentation). Fix two SCTP offload support copy&paste bugs: calculate checksums if fragmenting and no need to flag IPv4 header checksums in the IPv6 forwarding path. Correctly get the payload length in host byte order. While we already plan to support >64k payload here, the IPv6 header payload length obviously is only 16 bit and the calculations need to be right. Approved by: re | 08 July 2012, 11:53:13 UTC |
36e70b1 | Bjoern A. Zeeb | 08 July 2012, 11:42:21 UTC | MFC r235956: Hide the ip6aux functions. The only one referenced outside ip6_input.c is not compiled in yet (__notyet__) in route6.c (r235954, r238231). We do have accessor functions that should be used. Approved by: re | 08 July 2012, 11:42:21 UTC |
e798abb | Bjoern A. Zeeb | 08 July 2012, 11:33:49 UTC | MFC r235955: Simplify the code removing a return from an earlier else case, not differing from the default function return called now. Approved by: re | 08 July 2012, 11:33:49 UTC |
e65bedc | Bjoern A. Zeeb | 08 July 2012, 11:28:33 UTC | MFC r235954: We currently nowhere set IP6A_SWAP making the entire check useless with the current code. Keep around but do not compile in. Approved by: re | 08 July 2012, 11:28:33 UTC |
b84ac12 | Bjoern A. Zeeb | 08 July 2012, 11:24:22 UTC | MFC r235944: Significantly update tcp_lro for mostly two things: 1) introduce basic support for IPv6 without extension headers. 2) try hard to also get the incremental checksum updates right, especially also in the IPv4 case for the IP and TCP header. Move variables around for better locality, factor things out into functions, allow checksum updates to be compiled out, ... Leave a few comments on further things to look at in the future, though that is not the full list. Update drivers with appropriate #includes as needed for IPv6 data type in LRO. Approved by: re | 08 July 2012, 11:24:22 UTC |
61fada4 | Bjoern A. Zeeb | 08 July 2012, 10:46:55 UTC | MFC r235941: in_cksum.h required ip.h to be included for struct ip. To be able to use some general checksum functions like in_addword() in a non-IPv4 context, limit the (also exported to user space) IPv4 specific functions to the times, when the ip.h header is present and IPVERSION is defined (to 4). Approved by: re | 08 July 2012, 10:46:55 UTC |
7c731bd | Bjoern A. Zeeb | 08 July 2012, 10:29:01 UTC | MFC r235924: Introduce a (for now copied stripped down) in6_cksum_pseudo() function. We should be able to use this from in6_cksum() but we should also ponder possible MD specific improvements. It takes an extra csum argument to allow for easy checks as will be done by the upper layer protocol input paths. Approved by: re | 08 July 2012, 10:29:01 UTC |
22fd1d1 | Bjoern A. Zeeb | 08 July 2012, 10:04:30 UTC | MFC r235921: Optimize in6_cksum(), re-ordering work and limiting variable initialization, removing a bzero() for mostly re-initialized struct values, making use of the newly introduced in6_getscope(), as well as converting an if/panic to a KASSERT(). Approved by: re | 08 July 2012, 10:04:30 UTC |
9321617 | Bjoern A. Zeeb | 08 July 2012, 09:49:01 UTC | MFC r235916: Introduce in6_getscope() to allow more effective checksum computations without the need to copy the address to clear the scope. Approved by: re | 08 July 2012, 09:49:01 UTC |
4d2d2c1 | Bjoern A. Zeeb | 08 July 2012, 09:06:25 UTC | MFC r235671: Allow tso4 and tso6 be set individually given we have the bits. This will help with drivers not working as expected during the transition time and later. Approved by: re | 08 July 2012, 09:06:25 UTC |
665dade | Dimitry Andric | 06 July 2012, 17:42:34 UTC | MFC r238071: Fix clang warning, introduced in the recent dtrace import. Approved by: re (kib) | 06 July 2012, 17:42:34 UTC |
60398de | Benjamin Kaduk | 06 July 2012, 03:56:45 UTC | MFC r237581: Explicitly mention that setting the change and expiry times to zero is equivalent to leaving the time unset. [1] Wordsmith in the compat support section. Use a full path to nologin(8) in the context of setting it as a user's shell, keeping a separate cross-reference. PR: docs/169354 [1] Approved by: re (hrs), hrs (mentor) | 06 July 2012, 03:56:45 UTC |
ede3d3a | Andrey Chernov | 06 July 2012, 01:32:53 UTC | MFC r237777 Call set_terminal_mode() after video mode change, not before, because video mode change always reset terminal mode to xterm. It allows things like 'vidcontrol -T cons25 80x30' works as supposed, and allows 'allscreens_flags="-T cons25 80x30"' in /etc/rc.conf too. Approved by: re (hrs) | 06 July 2012, 01:32:53 UTC |
6c1387b | Edward Tomasz Napierala | 05 July 2012, 16:02:14 UTC | MFC r237636: Add manual page for bxe(4). Approved by: re (kib) | 05 July 2012, 16:02:14 UTC |