summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Fix pmap_page_set_memattr() behavior in the presence of fictitious pagesnwhitehorn2010-10-012-42/+20
| | | | | | by just caching the mode for later use by pmap_enter(), following amd64. While here, correct some mismerges from mmu_oea64 -> mmu_oea and clean up some dead code found while fixing the fictitious page behavior.
* Remove extra semicolon.yongari2010-10-011-2/+2
|
* Allow write DMA to request larger DMA burst size to get betteryongari2010-10-012-0/+5
| | | | | | performance on BCM5785. Obtained from: OpenBSD
* put back the assigment to sched_time. It was correct, andluigi2010-10-011-0/+1
| | | | | | it was necessary. Submitted by: Riccardo Panicucci
* Only hide the ifa and not the tp under #ifdef INET as the tp is neededbz2010-10-011-1/+1
| | | | | | for locking evenwhen there is no INET. MFC after: 3 days
* Change the output of mctest to give a summary of the results insteadgnn2010-10-012-13/+34
| | | | | | | | | | | | of printing a long list. Add a default base port, and default mulitcast address to the runner script. Add support for specifying a different local and remote interface in the runner script. MFC after: 1 week
* libc: Remove the i386 assembler version of strlen(3).jilles2010-10-012-55/+1
| | | | | | | | | | | | | On anything modern, the C version, which processes a word at a time, is much faster. The Intel optimization manual explicitly warns against using REP prefixes with SCAS or CMPS, which is exactly what the assembler version does. A simple test on a Phenom II showed the C version, compiled with -O2, to be about twice as fast determining the length of 100000 strings between 0 and 255 bytes long. MFC after: 2 weeks
* Proper bracketing.bz2010-10-011-2/+2
| | | | | | PR: kern/151100 Submitted by: SunMinghao (sunminghao hotmail.com) MFC after: 3 days
* - simplify by using one build scriptmr2010-10-018-678/+32
| | | | | | - use label for mounting - use GENERIC kernel - create empty Pkg directory
* i386 and amd64 mp_machdep: improve topology detection for Intel CPUsavg2010-10-012-172/+250
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This patch is significantly based on previous work by jkim. List of changes: - added comments that describe topology uniformity assumption - added reference to Intel Processor Topology Enumeration article - documented a few global variables that describe topology - retired weirdly set and used logical_cpus variable - changed fallback code for mp_ncpus > 0 case, so that CPUs are treated as being different packages rather than cores in a single package - moved AMD-specific code to topo_probe_amd [jkim] - in topo_probe_0x4() follow Intel-prescribed procedure of deriving SMT and core masks and match APIC IDs against those masks [started by jkim] - in topo_probe_0x4() drop code for double-checking topology parameters by looking at L1 cache properties [jkim] - in topo_probe_0xb() add fallback path to topo_probe_0x4() as prescribed by Intel [jkim] Still to do: - prepare for upcoming AMD CPUs by using new mechanism of uniform topology description [pointed by jkim] - probe cache topology in addition to CPU topology and probably use that for scheduler affinity topology; e.g. Core2 Duo and Athlon II X2 have the same CPU topology, but Athlon cores do not share L2 cache while Core2's do (no L3 cache in both cases) - think of supporting non-uniform topologies if they are ever implemented for platforms in question - think how to better described old HTT vs new HTT distinction, HTT vs SMT can be confusing as SMT is a generic term - more robust code for marking CPUs as "logical" and/or "hyperthreaded", use HTT mask instead of modulo operation - correct support for halting logical and/or hyperthreaded CPUs, let scheduler know that it shouldn't schedule any threads on those CPUs PR: kern/145385 (related) In collaboration with: jkim Tested by: Sergey Kandaurov <pluknet@gmail.com>, Jeremy Chadwick <freebsd@jdc.parodius.com>, Chip Camden <sterling@camdensoftware.com>, Steve Wills <steve@mouf.net>, Olivier Smedts <olivier@gid0.org>, Florian Smeets <flo@smeets.im> MFC after: 1 month
* sysctls in kern_shutdown: add twin tunablesavg2010-10-011-6/+9
| | | | | | | also make couple of sysctl-controlled variables static Reviewed by: rwatson MFC after: 1 week
* Fix background roaming to actually work in AUTO roaming mode.adrian2010-10-011-1/+1
| | | | | IEEE80211_F_BGSCAN is a vap flag, not a channel flag. So although bgscan occured, sta_roam_check() would never be called.
* - Check flag with the bitwise operator, not the logical operator.lulf2010-10-012-4/+4
| | | | | Submitted by: arundel MFC after: 1 week
* Implementaiton of man, manpath, whatis, and apropos written entirely in sh.gordon2010-10-016-0/+1510
| | | | | | | | | | | | | | | | | | Features of this new version in favor of the old one: BSD licensed -- old one is GPL. Imports configuration from /etc/man.conf and LOCALBASE/etc/man.d/*.conf allowing ports to extend the base functionality. The pluggable configuration can supplement the manual search path (retiring use.perl), add locales, and override language specific toolsets (attempt to merge the japanese/man port into the base system as much as possible). Much effort has been made to make this version mirror the functionality of the existing implementation. For 99% of users, it should be a drop in replacement. PR: gnu/143271, gnu/4419 Reviewed by: dougb (previous versions) Approved by: wes (mentor)
* Fix IFCAP_TXCSUM/IFCAP_RXCSUM handling. Previously bge(4) usedyongari2010-09-301-4/+8
| | | | | | IFCAP_HWCSUM to know which capability should be changed such that disabling RX checksun offloading resulted in disabling TX checksum offloading.
* MFV of 213310, tzdata2010medwin2010-09-302-5/+5
| | | | | - Fix historical data for Hongkong - Fix location description for Antarctica/Vostok
* Retire the amd64 and i386 specific inline assembly versions of ldexp.c,dim2010-09-302-136/+0
| | | | | | | | as they are slower than the generic version in C, at least on modern hardware. This leaves us with just five implementations. Suggested by: bde Approved by: rpaulo (mentor)
* Add support for memory attributes (pmap_mapdev_attr() and friends) onnwhitehorn2010-09-307-51/+290
| | | | | PowerPC/AIM. This is currently stubbed out on Book-E, since I have no idea how to implement it there.
* Rename rl_setmulti() to rl_rxfilter() as rl_rxfilter() will handleyongari2010-09-301-66/+42
| | | | | | | | | | | | | | | | | | | | IFF_ALLMULTI/IFF_PROMISC as well as multicast filter configuration. Rewrite RX filter logic to reduce number of register accesses and make it handle promiscuous/allmulti toggling without controller reinitialization. Previously rl(4) counted on controller reinitialization to reprogram promiscuous configuration but r211767 resulted in avoiding controller reinitialization whenever promiscuous mode is toggled. To address this, keep track of driver's view of interface state and handle IFF_ALLMULTI/IFF_PROMISC changes without reinitializing controller. This should fix a regression introduced in r211267. While I'm here remove unnecessary variable reassignment in ioctl handler. PR: kern/151079 MFC after: 1 week
* there must be only one SYSINIT with SI_SUB_RUN_SCHEDULER+SI_ORDER_ANY orderavg2010-09-303-3/+3
| | | | | | | | | | | | | | | SI_SUB_RUN_SCHEDULER+SI_ORDER_ANY should only be used to call scheduler() function which turns the initial thread into swapper proper and thus there is no further SYSINIT processing. Other SYSINITs with SI_SUB_RUN_SCHEDULER+SI_ORDER_ANY may get ordered after scheduler() and thus never executed. That particular relative order is semi-arbitrary. Thus, change such places to use SI_ORDER_MIDDLE. Also, use SI_ORDER_MIDDLE instead of correct, but less appealing, SI_ORDER_ANY - 1. MFC after: 1 week
* debug.kdb.stop_cpus sysctl: hint that this is also a tunableavg2010-09-301-1/+1
| | | | MFC after: 1 week
* kmem_size* sysctls: hint that these are also tunablesavg2010-09-301-4/+4
| | | | MFC after: 1 week
* Do not use regular interrupts on NVidia HPETs. NVidia MCP5x chipsets havemav2010-09-301-0/+7
| | | | | | number of unexplained interrupt problems. For some reason, using HPET interrupts there breaks HDA sound. Legacy route mode interrupts reported to work fine there.
* Revert r132291.mav2010-09-301-2/+1
| | | | | Restore setting PIO/WDMA timings for VIA UDMA133 controllers. Linux disables only AST register writing there, but no all timings.
* If an RCS file is truncated, rcsfile_getdelta() will return NULL. Insteadjhb2010-09-301-0/+8
| | | | | | | | | | of faulting, check for NULL. However, returning an error would cause csup to just abort the entire update. Instead, break out of the loop and return ok. The attempts to update the file will trigger a MD5 failure which will cause csup to download the entire file as a fixup. Reviewed by: lulf MFC after: 1 week
* Fix printing of the "rw" mount option in fstab(5) format (-p option).jh2010-09-301-12/+12
| | | | | | | | | fstab(5) format requires that one of "rw", "rq" or "ro" is always specified. PR: bin/123021 Reviewed by: keramida, rodrigc MFC after: 2 weeks
* change code to use unwind.h.davidxu2010-09-302-5/+6
|
* Remove target peigen.c so that make doesn't warn us about it.rpaulo2010-09-301-2/+0
|
* Also the stop function should accept 64-bit exception class.davidxu2010-09-301-1/+1
|
* - Sizeof exception class type should always be 64-bit.davidxu2010-09-301-3/+5
| | | | - Make some functions be visible to BSD source code.
* Import unwind.h from libunwind-0.99.davidxu2010-09-301-0/+154
| | | | Approved by: core (kib, rwatson, imp, brooks)
* Add advice to new fathers from Thomas Ptacek.cperciva2010-09-291-0/+6
| | | | | | Suggested by: wilko Approved by: Thomas Ptacek Security: New parents often feel very insecure.
* - Fix values of CS1_EN and CS2_EN flagsgonzo2010-09-292-7/+6
| | | | | | | - Unbreak kernel build by fixing naming convention of GPIO_FUNC flags Spotted by: Luiz Otavio O Souza, Andrew Thompson
* Commit the rest of r213270.emaste2010-09-291-0/+1
| | | | Thanks to Anonymous <swell dot k at gmail.com> for spotting this.
* tr: Fix '[=]=]' equivalence class.jilles2010-09-294-2/+6
| | | | | | | | | | | A closing bracket immediately after '[=' should not be treated as special. Different from the submitted patch, a string ending with '[=' does not cause access beyond the terminating '\0'. PR: bin/150384 Submitted by: Richard Lowe MFC after: 2 weeks
* Implement hardware MAC statistics for BCM5705 or newer Broadcomyongari2010-09-292-49/+339
| | | | | | | controllers. bge(4) exported MAC statistics on controllers that maintain the statistics in the NIC's internal memory. Newer controllers require register access to fetch these values. These counters provide useful information to diagnose driver issues.
* Fix bogus error message from bus_dmamem_alloc() about incorrect alignment.neel2010-09-295-5/+5
| | | | | | | | | The check for alignment should be made against the physical address and not the virtual address that maps it. Sponsored by: NetApp Submitted by: Will McGovern (will at netapp dot com) Reviewed by: mjacob, jhb
* Apply the same workaround for clang to amd64's version of ldexp.c (as indim2010-09-291-3/+9
| | | | | | | | | | | | | r212976): order the incoming arguments to fscale as st(0), st(1), and mark temp2 volatile (only in case of compilation with clang) to force clang to pop it correctly. No binary change when compiled with gcc. This fixes ldexp() when compiled with clang on amd64, which makes drand48() and friends work correctly again, and this in turn fixes perl's tempfile(). Reported by: Renato Botelho, Derek Tattersall Approved by: rpaulo (mentor)
* After r207391, brgphy(4) passes resolved flow-control settings toyongari2010-09-291-2/+13
| | | | | | | | parent driver. Use that information to configure flow-control. One drawback is there is no way to disable flow-control as we still don't have proper way to not advertise RX/TX pause capability to link partner. But I don't think it would cause severe problems and users can selectively disable flow-control in switch port.
* remove an unnecessary (and wrong) assignment.luigi2010-09-291-1/+0
| | | | | | | | It was meant to reset idle_time (and it was not needed), but i even used the wrong field. Obtained from: Oleg MFC after: 3 days
* AR71XX_GPIO_* defines were introduced by adrian@ a while ago,gonzo2010-09-291-15/+0
| | | | remove duplicated.
* Fix legal staff in GPIO sources:gonzo2010-09-295-40/+70
| | | | | | | | - license clause now contains "AUTHOR AND CONTRIBUTORS" instead of just "AUTHOR" - Add license/copyright to gpioc.c Spotted by: Edward Tomasz Napierala, Andrew Turner
* Since local variable 'i' is used only in a KASSERT, declare andmckusick2010-09-291-2/+3
| | | | | | | initialize it only if INVARIANTS is defined to avoid a declared but unused warning. Suggested by: Brian Somers <brian@FreeBSD.org>
* Add support for the BFD target efi-app-x86_64.rpaulo2010-09-296-15/+981
| | | | | This uses only GPL2 source code and is a requirement for a 64 bit EFI boot loader.
* Previously, the aac driver did not handle enclosure management AIFs,emaste2010-09-294-0/+80
| | | | | | | which were raised during hot-swap events. Now such events trigger cam rescans, as is done in the mps driver. Submitted by: Mark Johnston <mjohnston at sandvine dot com>
* Account for unlocking a spin mutex in the lock profiling code in the !SMPjhb2010-09-291-1/+4
| | | | | | | case. Submitted by: truckman MFC after: 3 days
* /etc/rc.d/defaultroute currently bails immediately if all interfacesemaste2010-09-291-17/+26
| | | | | | | | | | | | | | | | set to use DHCP have no carrier. This can cause grief as it may take some time for link to be established, and defaultroute may terminate before this happens. Introduce a defaultroute_carrier_delay variable and then wait that long in defaultroute before bailing if no interfaces have carrier. With the default settings defaultroute will wait for five seconds for this, and the original 30 second wait for a default route to appear is unchanged. Note that there is in discussion an alternative approach to the broader problem of waiting for DHCP-configured routes. However, this change addresses a real problem in the current defaultroute script. Discussed on: freebsd-rc@
* If rt2560_bbp_init() fails, don't drop the lock as the callers ofjhb2010-09-291-2/+1
| | | | | | | rt2560_init_locked() expect the lock to be held on return. Reported by: Anton Shterenlikht mexas of bristol ac uk MFC after: 1 week
* whitespace changes in preparation for future commitsluigi2010-09-295-8/+15
|
* fix handling of initial credit for an idle pipe.luigi2010-09-291-1/+4
| | | | | | | | | This fixes the bug where setting bw > 1 MTU/tick resulted in infinite bandwidth if io_fast=1 PR: 147245 148429 Obtained from: Riccardo Panicucci MFC after: 3 days
OpenPOWER on IntegriCloud