summaryrefslogtreecommitdiff
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* This commit was manufactured by cvs2svn to create tagrelease/4.6.0cvs2svn2002-07-031-1/+1
| | | | | | | | 'RELENG_4_6_0_RELEASE'. Notes: svn path=/releng/4.6/; revision=99348 svn path=/release/4.6.0/; revision=99349; tag=release/4.6.0
* This commit was manufactured by cvs2svn to create branch 'RELENG_4_6'.cvs2svn2002-07-033566-0/+1665495
|\ | | | | | | Notes: svn path=/releng/4.6/; revision=99348
| * This commit was manufactured by cvs2svn to create branch 'RELENG_4'.cvs2svn2002-07-032-0/+681
| | | | | | | | Notes: svn path=/stable/4/; revision=99347
| * MFC rev 1.131 - add another chip rev.Mike Silbersack2002-07-031-0/+1
| | | | | | | | Notes: svn path=/stable/4/; revision=99328
| * MFC: Fix GCC warnings caused by initializing a zero length array.Archie Cobbs2002-07-0232-242/+155
| | | | | | | | | | | | | | | | | | In the process, simply things a bit by getting rid of 'struct ng_parse_struct_info' which was useless because it only contained one field. Notes: svn path=/stable/4/; revision=99312
| * Adjust the previous commit so that the value of SI_SUB_DRIVERS doesn't change,Archie Cobbs2002-07-021-2/+2
| | | | | | | | | | | | | | | | | | to avoid binary compatibility problems with 3rd party drivers. Requested by: julian Notes: svn path=/stable/4/; revision=99311
| * MFC: Const'ify variables to make it clear we're not writing to the mbuf data.Archie Cobbs2002-07-025-49/+53
| | | | | | | | Notes: svn path=/stable/4/; revision=99297
| * MFC: (rev. 1.70) Initialize mbuf subsystem before initializing drivers.Archie Cobbs2002-07-021-2/+2
| | | | | | | | | | | | | | PR: kern/38745 Notes: svn path=/stable/4/; revision=99281
| * MFC: (rev. 1.90) Fix M_TRAILINGSPACE() to return zero if !M_WRITABLE(m).Archie Cobbs2002-07-021-2/+9
| | | | | | | | Notes: svn path=/stable/4/; revision=99280
| * MFC: (rev. 1.92) Don't set "m->m_ext.ext_size" to zero in m_split().Archie Cobbs2002-07-021-1/+5
| | | | | | | | Notes: svn path=/stable/4/; revision=99279
| * MFC: (rev. 1.18) Fix bugs where mbuf was being accessed without m_pullup().Archie Cobbs2002-07-021-7/+41
| | | | | | | | Notes: svn path=/stable/4/; revision=99276
| * MFC: (rev. 1.24) Fix for checking M_WRITABLE(), eliminate some complexity.Archie Cobbs2002-07-021-57/+27
| | | | | | | | Notes: svn path=/stable/4/; revision=99275
| * MFC RLIMIT_VMEM resource Stage I - limit the total virtual memory use forMatthew Dillon2002-07-024-1/+26
| | | | | | | | | | | | | | a process (exec() to be handled in Stage II). Notes: svn path=/stable/4/; revision=99274
| * MFC: (rev. 1.23) Don't send packets out an interface unless IFF_UP|IFF_RUNNING.Archie Cobbs2002-07-021-2/+9
| | | | | | | | Notes: svn path=/stable/4/; revision=99272
| * MFC r1.22 - delaying ich calibration until system clock itself is calibrated.Orion Hodson2002-07-021-6/+22
| | | | | | | | Notes: svn path=/stable/4/; revision=99261
| * MFC: Fixes to asm statements in the Internet checksum code to makeJohn Polstra2002-07-022-20/+28
| | | | | | | | | | | | | | | | | | | | | | | | them safer at high optimization levels. Also, replace matching constraints with the "+" modifier where appropriate. This commit merges in the following revisions from the main branch: i386/in_cksum.c 1.23, 1.24 include/in_cksum.h 1.11, 1.13 Notes: svn path=/stable/4/; revision=99244
| * A bit of language touches to LINTTom Rhodes2002-07-021-2/+2
| | | | | | | | | | | | | | | | | | PR: 39823 (mostly) Submitted by: Chris Pepper <pepper@rockefeller.edu> Approved by: rwatson (mentor) Notes: svn path=/stable/4/; revision=99233
| * MFC: 1.72; avoid exposing struct if_clone and the sys/queue.h macrosIan Dowse2002-07-011-0/+4
| | | | | | | | | | | | | | to userland programs. Notes: svn path=/stable/4/; revision=99178
| * MFC: Fix two off-by-one errors when sanity-checking inode numbers.Ian Dowse2002-07-012-2/+2
| | | | | | | | Notes: svn path=/stable/4/; revision=99176
| * MFC: 1.137; fix a harmless typo in do_sigaction() where the wrongIan Dowse2002-07-011-3/+3
| | | | | | | | | | | | | | members of a union were used (but the members alias each other). Notes: svn path=/stable/4/; revision=99174
| * Now that the IF_HANDOFF macro has been MFC'd, don't define it here. AWarner Losh2002-06-301-0/+2
| | | | | | | | | | | | | | | | | | | | proper MFC of a newer wi driver will happen later, and this code was ifdef for 4.x. This removes some warnings in the building of the 4.x kernel. Notes: svn path=/stable/4/; revision=99167
| * MFC: Add missing splx call (if_bge.c revision 1.15).John Polstra2002-06-301-1/+3
| | | | | | | | | | | | | | Submitted by: Mark Peek <mp@FreeBSD.org> Notes: svn path=/stable/4/; revision=99145
| * MFC: Work around an apparent chip bug involving the BCM5701 on aJohn Polstra2002-06-302-2/+37
| | | | | | | | | | | | | | PCI-X bus and unaligned receive buffers. Notes: svn path=/stable/4/; revision=99144
| * Add one function and one macro to timestamp events in the kernel usingLuigi Rizzo2002-06-303-0/+56
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | the TSC register and which are just too conveninent not to have them around. The macro (defined in sys/i386/include/param.h) is called TSTMP(class:4, unit:4, event:8, par:16) which in turn calls _TSTMP(arg) (function defined in sys/i386/isa/clock.c) which stores a pair of u_int32_t values <timestamp, arg> in a circular buffer exported to userland as the sysctl variable debug.timestamp The content of the buffer can be printed with something like sysctl -b debug.timestamp | hexdump -e '"%15u %15u\n"' This functionality must be enabled with "options KERN_TIMESTAMP", otherwise both the macro and the function are redefined as empty blocks. This code is i586+ specific, but it does not make any attempt to grab locks, identify which CPU it is running on, etc. etc. -- it is a debugging tool, and a very very simple and low level one, so you are supposed to know how to use it if you do not want to get random numbers out of it. I am committing this to RELENG_4 only because on -current parts of the kernel might run concurrently, so the use of a single unprotected static variable as the index of the next entry in the buffer might lead to inconsistent results. Plus, there might be similar stuff in -current already. If someone feels like porting this code to -current, it might be a good idea to keep separate per-cpu buffers to avoid the cost of protecting against concurrent accesses. Notes: svn path=/stable/4/; revision=99125
| * MFC (all: 1.11 (minor style fixes) and 1.12 (return the old mask and notBruce Evans2002-06-301-9/+18
| | | | | | | | | | | | | | | | | | its complement in fpsetmask()). PR: 38170 Notes: svn path=/stable/4/; revision=99096
| * Re-enable send_sessionid().Brian Somers2002-06-291-2/+2
| | | | | | | | | | | | | | | | | | The problem was that ppp(8) was only reading at most one netgraph message per second. The extra pre-connect message was producing too much of a delay and breaking the initial negotiation. Notes: svn path=/stable/4/; revision=99089
| * MFC 1.12: In m_aux_delete, no need to chase beyond victim.Hajimu UMEMOTO2002-06-291-0/+1
| | | | | | | | Notes: svn path=/stable/4/; revision=99084
| * MFC: Revised MLD-related definitionsHajimu UMEMOTO2002-06-293-69/+94
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | - Used mld_xxx and MLD_xxx instead of mld6_xxx and MLD6_xxx according to the official defintions in rfc2292bis (macro definitions for backward compatibility were provided) - Changed the first member of mld_hdr{} from mld_hdr to mld_icmp6_hdr to avoid name space conflict in C++ sys/netinet/icmp6.h: 1.8 sys/netinet6/icmp6.c: 1.18 sys/netinet6/mld6.c: 1.8 Notes: svn path=/stable/4/; revision=99083
| * MFC v1.21: Add additional cpuid feature flags and put into a canonical format.Mark Peek2002-06-291-18/+33
| | | | | | | | Notes: svn path=/stable/4/; revision=99071
| * MFC v1.114: Clock frequencies reported by sysctl should be unsigned values.Mark Peek2002-06-291-2/+2
| | | | | | | | Notes: svn path=/stable/4/; revision=99070
| * MFC v1.187: Clock frequencies reported by sysctl should be unsigned values.Mark Peek2002-06-291-2/+2
| | | | | | | | Notes: svn path=/stable/4/; revision=99069
| * MFC 1.313 - fix a bug in an optimization made by vfs_bio_clrbuf().Matthew Dillon2002-06-291-6/+11
| | | | | | | | Notes: svn path=/stable/4/; revision=99067
| * MFC rev 1.70:Mike Silbersack2002-06-291-0/+1
| | | | | | | | | | | | | | | | | | Ensure that packet counts are always reset to 0 when a route is cloned. Previously, they took on the count of their parent route (which was sometimes nonzero.) Notes: svn path=/stable/4/; revision=99011
| * MFC: r1.23; delay referencing next entry in timer chain.Jonathan Lemon2002-06-281-1/+7
| | | | | | | | Notes: svn path=/stable/4/; revision=98983
| * Add IF_HANDOFF() and _IF_QFULL() macros to ease merging code from -CURRENTLuigi Rizzo2002-06-281-90/+92
| | | | | | | | | | | | | | | | | | (and also make the code a lot more readable). While we are on this file, remove some unused interfaces and __P() Notes: svn path=/stable/4/; revision=98974
| * MFC 1.176 - fix a defect in wakeup() and a more serious defect inMatthew Dillon2002-06-281-2/+8
| | | | | | | | | | | | | | wakeup_one(). Notes: svn path=/stable/4/; revision=98949
| * MFC: remove a printf and add a comment about an assumption which mightLuigi Rizzo2002-06-271-3/+10
| | | | | | | | | | | | | | be occasionally violated by some device drivers. Notes: svn path=/stable/4/; revision=98947
| * Remove calls to send_sessionid().Brian Somers2002-06-271-2/+2
| | | | | | | | | | | | | | | | | | | | | | I have no idea what the problem is (yet). Andre Albsmeier has reported that removing these fixes his PPPoE connection again - although the calls don't seem to affect PPPoE here. This is a temporary fix 'till I figure out what's broken. Notes: svn path=/stable/4/; revision=98936
| * MFC: `sed -i'.David E. O'Brien2002-06-261-1/+3
| | | | | | | | | | | | | | | | | | This also gets deregisterization, de_P, and all others changes between 5-CURRENT and RELENG_4. Bump __FreeBSD_version for bsd.port.mk's use. Notes: svn path=/stable/4/; revision=98880
| * Bump revsion to 460100 after consultation with re@ and so@ to reflectWarner Losh2002-06-261-1/+1
| | | | | | | | | | | | | | dns fix. Notes: svn path=/stable/4/; revision=98857
| * MFC:Prafulla Deuskar2002-06-254-5322/+0
| | | | | | | | | | | | | | | | | | Removed unneeded files. if_em_fxhw.[c,h] and if_em_phy.[c,h] have been merged into one [c,h] file. Notes: svn path=/stable/4/; revision=98797
| * Unbreak LINT by finishing the move of sr(4) to sys/dev.Tony Finch2002-06-243-3/+4
| | | | | | | | | | | | | | Reviewed by: jhay Notes: svn path=/stable/4/; revision=98761
| * MFC (1.225: if trap() is called when ddb is active, then go directly toBruce Evans2002-06-241-0/+8
| | | | | | | | | | | | | | | | | | trap_fatal(); do not blunder around enabling interrupts and running trap handlers). This is more of a rewrite than a merge because trap() has has fewer but more harmful goto's in RELENG_4. Notes: svn path=/stable/4/; revision=98741
| * Bump FreeBSD_version for sendfile fix.Alfred Perlstein2002-06-241-1/+1
| | | | | | | | Notes: svn path=/stable/4/; revision=98730
| * Fix sendfile bug where headers counted against amount of file to be sent.Alfred Perlstein2002-06-241-3/+5
| | | | | | | | | | | | | | MFC: 1.103 Notes: svn path=/stable/4/; revision=98729
| * Enforce a 2^31 * PAGE_SIZE file size limit to ensure that largeIan Dowse2002-06-231-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | files cannot overflow the vm_pindex_t type. This is not an MFC, since increasing the size of vm_pindex_t as was done in -CURRENT is not useful or appropriate for RELENG_4. Note that on the alpha platform this change is not necessary because vm_pindex_t is 64-bit, but limiting files to 16TB there is unlikely to cause any problems. Submitted by: tegge Notes: svn path=/stable/4/; revision=98711
| * MFC: add PACKET_TAG_* #defines to implement packet annotations.Luigi Rizzo2002-06-231-63/+132
| | | | | | | | | | | | | | | | | | | | | | | | While at it, remove __P() macros and align some comments and indentation to the form used in -current to minimize differences (which are still large though). Apart from the PACKET_TAG_* #defines, for all practical purposes this commit is a whitespace change. Notes: svn path=/stable/4/; revision=98672
| * MFC: NUL terminate the ACNAME passed to userland.Brian Somers2002-06-221-3/+4
| | | | | | | | Notes: svn path=/stable/4/; revision=98637
| * MFC 1.90. Fix a major performance issue that was introduced in lateMatthew Dillon2002-06-221-7/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | 2000. Setting B_CLRBUF for the BALLOC call unexpectedly caused BALLOC to issue a read-before-write for each fs block, resulting in a huge (4x) performance penalty due to extra I/Os and lack of clustering. This workaround avoids the unnecessary read resulting in full write clustering performance to media. It is expected to have major visible effects only on large news systems (e.g. INN cyclic buffers), and databases, especially on hardware RAID systems (10x), and has already been shown to have a huge positive effect on INN. Most other programs use write-append which is already optimal. This patch does not fix the double-memory-copy to the filesystem buffer. That is a more difficult problem and is being worked on as a separate issue. Reviewed by: mckusick, tegge Notes: svn path=/stable/4/; revision=98626
| * Somewhat belatedly bump __FreeBSD_version for 4.6-STABLE.Bruce A. Mah2002-06-211-1/+1
| | | | | | | | | | | | | | Approved by: imp Notes: svn path=/stable/4/; revision=98571
generated by cgit v1.3 (git 2.53.0) at 2026-06-02 12:41:17 +0000