summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Add __elfN(dump_thread). This function is called from __elfN(coredump)marcel2004-08-1112-79/+122
| | | | | | | | | to allow dumping per-thread machine specific notes. On ia64 we use this function to flush the dirty registers onto the backingstore before we write out the PRSTATUS notes. Tested on: alpha, amd64, i386, ia64 & sparc64 Not tested on: arm, powerpc
* Add ADAPTIVE_GIANT to GENERIC on i386, with the intent of making itrwatson2004-08-111-0/+1
| | | | | | | | | | a standard configuration similar to [NO_]ADAPTIVE_MUTEXES. This feature causes Giant to be included in the set of mutexes adaptively spun on. It appears to have a positive effect on performance on SMP across several workloads, including measurements of a 16% improvement on buildworld, and 30%+ improvement for MySQL using the supersmack benchmark with Giant over the network stack; a 6% improvement without Giant on the network stack (as a result of less giant contention).
* In v_addpollinfo(), we allocate storage to back vp->v_pollinfo. However,rwatson2004-08-111-1/+7
| | | | | | | | | we may sleep when doing so; check that we didn't race with another thread allocating storage for the vnode after allocation is made to a local pointer, and only update the vnode pointer if it's still NULL. Otherwise, accept that another thread got there first, and release the local storage. Discussed with: jmg
* Update links to the bluez-firmware packageemax2004-08-111-4/+4
| | | | Update md5 hash
* Set IFF_RUNNING flag on the interface as soon as the control device is opened.emax2004-08-111-3/+9
|
* Modify vnode locking key: the v_pollinfo pointer itself is protectedrwatson2004-08-101-1/+2
| | | | | by Giant; the contents are protected by the pollinfo mutex. We rely on Giant to prevent races in assigning the value of v_pollinfo.
* Add id tags to sections which does not already have one to make itsimon2004-08-102-22/+22
| | | | simpler to link directly into the Hardware Notes.
* Auto generate device listings the following drivers: mlx, mly, and ncv.simon2004-08-103-18/+11
|
* Add a HARDWARE section which lists supported devices.simon2004-08-101-3/+5
|
* Don't call DEVICE_RESUME a second time if DEVICE_SUSPEND fails. Thenjl2004-08-101-11/+10
| | | | | bus_generic_suspend method does this for us. Disable interrupts before entering S1. This may help some systems suspend to S1 successfully.
* Eliminate the acquisition and release of Giant within physio(). Removealc2004-08-101-6/+0
| | | | | | | the spl calls. Reviewed by: phk@ Discussed with: scottl@
* Add a comment describing pmap_extract_and_hold() noting that the protectionalc2004-08-101-0/+7
| | | | check still needs implementation on arm.
* Add a HARDWARE section which lists supported devices. Clean up devicesimon2004-08-102-29/+38
| | | | lists a bit while I'm here anyway.
* Add pmap locking to many of the functions.alc2004-08-104-20/+72
| | | | | | | | | | | | | Implement the protection check required by the pmap_extract_and_hold() specification. Remove the acquisition and release of Giant from pmap_extract_and_hold() and pmap_protect(). Many thanks to Ken Smith for resolving a sparc64-specific initialization problem in my original patch. Tested by: kensmith@
* If we kill the worklist thread of a RAID5 plex we can destroyle2004-08-102-1/+1
| | | | | the worklist mutex at the same time, so move the mtx_destroy() call to gv_kill_thread().
* Lock the topology before calling gv_parse_config, not afterwards.le2004-08-101-1/+1
|
* - Recognize HARDCODED flag when dumping consumer configuration.pjd2004-08-101-39/+27
| | | | - Improve code readabilty a bit.
* Forgot to commit those: introduce hardcoded provider functionality,pjd2004-08-102-17/+64
| | | | | which allow to store provider's name in the metadata and avoid problems when few providers share the same last sector.
* Comment-out the debugging printf I left in in case there were somemarcel2004-08-101-1/+1
| | | | packet related problems. No problems have been reported.
* Fix one of the lastest commit. This bio_caller1 should also be changed topjd2004-08-101-1/+1
| | | | | | | bio_driver1 (as all the rest). This introduced a small memory leak, but it wasn't really critical, because maximum memory for g_stripe_zone is always set, so after few requests gstripe was working in "economic" mode.
* typowilko2004-08-101-1/+0
|
* There is nothing but real hardware to verify things: 164sx can reallywilko2004-08-101-4/+7
| | | | | | run on non-ECC, plain vanilla PC133 PC SDRAM DIMMs. Thanks to: obrien@ for the donation of a 164sx mlb.
* Teach sysinstall about the "srescue" distribution, which contains thecperciva2004-08-106-1/+11
| | | | | | | | contents of /usr/src/rescue. Until now, the files were shipped with releases but sysinstall would ignore them (resulting in a non-buildable source tree). Sanity checked by: jhb
* Synchronize the extra SA threading checks and return value handling ofjhb2004-08-101-24/+50
| | | | | | condition variables with that of msleep(). Reviewed by: davidxu
* Update "documentation date" fields.green2004-08-103-3/+3
|
* Document exactly how vslock(9) is broken.green2004-08-101-0/+6
|
* Update mlock(2) manpage to cross-reference m{,un}lockall(2), removegreen2004-08-101-1/+5
| | | | | a case where ENOMEM could be returned by munlock(2), and add possible system deadlock to the BUGS section.
* Update contigmalloc(9)'s manpage to reflect usage of the malloc_typegreen2004-08-101-9/+2
| | | | | argument and that allocation occurs from the end of the address space backward (still first-fit).
* Back out all behavioral chnages.green2004-08-102-36/+17
|
* Describe better what the initial /usr/obj is expected to be populated with.ru2004-08-101-2/+4
| | | | Prodded by: grehan
* Fix recent breakage in rescue. We need to build a new crunchgenharti2004-08-101-1/+1
| | | | | | that will not emit the bad MAKE=make line that caused the breakage. Submitted by: ru
* Initialize thread_db module.davidxu2004-08-101-0/+2
|
* As AMD64 architecture volume 1 chapter 3.1.2 says, high 32 bits of %rflagsdavidxu2004-08-101-3/+4
| | | | | | are resevered, they can be written with anything, but they always read as zero, we should simulate it in set_regs() as we are reading/writting real hardware %rflags register.
* This commit was generated by cvs2svn to compensate for changes in r133429,harti2004-08-101-1/+1
|\ | | | | | | which included commits to RCS files with non-trunk default branches.
| * Fix the 4.X -> 5.X upgrade path by include <inttypes.h> instead ofharti2004-08-101-1/+1
| | | | | | | | stdint.h.
* | Add documentation for log(9).joerg2004-08-102-5/+46
| | | | | | | | MFC after: 1 week
* | - Use a new flag, KEF_XFERABLE, to record with certainty that this kse hadjeff2004-08-101-34/+76
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | contributed to the transferable load count. This prevents any potential problems with sched_pin() being used around calls to setrunqueue(). - Change the sched_add() load balancing algorithm to try to migrate on wakeup. This attempts to place threads that communicate with each other on the same CPU. - Don't clear the idle counts in kseq_transfer(), let the cpus do that when they call sched_add() from kseq_assign(). - Correct a few out of date comments. - Make sure the ke_cpu field is correct when we preempt. - Call kseq_assign() from sched_clock() to catch any assignments that were done without IPI. Presently all assignments are done with an IPI, but I'm trying a patch that limits that. - Don't migrate a thread if it is still runnable in sched_add(). Previously, this could only happen for KSE threads, but due to changes to sched_switch() all threads went through this path. - Remove some code that was added with preemption but is not necessary.
* | Favor building the cdrtools port over fetching the package. Theremarcel2004-08-101-24/+22
| | | | | | | | | | | | | | hasn't been any package build for ia64 for a long time and in this case it means that the cdrtools package on the FTP sites is one that doesn't have mkisofs. This breaks the release build. By trying to build the port first, we protect ourselves a bit more against this.
* | Update wording regarding geography to match the man page and web site.murray2004-08-101-1/+1
| | | | | | | | | | | | | | Specifically, change the second level menu title from 'Country' to 'Country or Region', since e.g. Hong Kong is not a country. Submitted by: Xin LI <delphij@frontfree.net>
* | Disable the snmp_atm module until the source gets fixed. This will unbreakscottl2004-08-102-2/+2
| | | | | | | | this portion of the build.
* | Don't acquire Giant in cryptof_close(), as the code is intended to berwatson2004-08-101-7/+0
| | | | | | | | | | | | able to run MPsafe (and appears to be MPsafe). Discussed with (some time ago): sam
* | Inline umich license from COPYRIGHT to make it clear what license therwatson2004-08-102-7/+92
| | | | | | | | | | | | | | | | | | umich copyright is asserting. Clarify that the copyright I'm asserting is the standard Berkeley license. Remove Giant assertions from AARP and DDP input routines.
* | Initial pass at shifting sysinstall(8) to install X.org instead of XFree86.kensmith2004-08-1010-463/+312
| | | | | | | | | | | | | | There are still a few nits to work out (graphical config non-functional). Patches by: anholt@, adapted slightly Reviewed by: 'make release'
* | Catch up with change to <machine/pte.h>.marcel2004-08-102-18/+16
| |
* | By default, the watch utility will attempt to open /dev/snp0, ifcsjp2004-08-101-3/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | another process already has /dev/snp0 open, the snp(4) will return EBUSY, in which case watch will try to open /dev/snp1..9. Currently watch does not check errno to see if the failure was a result of EBUSY. This results in watch making futile attempts to open snp0..snp9 even though devices may not exist or the caller does not have permissions to access the device. In addition to this, it attempts to setup the screen for snooping even though it may not ever get an snp device. So this patch does two things 1) Checks errno for EBUSY, if open(2) fails for another reason print that reason and exit. 2) setup the terminal for snooping after the snp descriptor has been obtained. Approved by: bmilekic (mentor)
* | Skip the syncing disks loop if there are no dirty buffers. Remove anjl2004-08-102-6/+14
| | | | | | | | | | | | variable used to flag the initial printf. Submitted by: truckman (earlier version)
* | Introduce ng_hci_inquiry_response structure and use it in the hccontrol(8)emax2004-08-102-15/+11
| |
* | Add a temporary debugging hack to detect a deadlock in setrunqueue(). Thisscottl2004-08-101-0/+7
| | | | | | | | | | | | is here so that we can gather stats on the nature of the recent rash of hard lockups, and in this particular case panic the machine instead of letting it deadlock forever.
* | In syscall, always make a copy of parameters from trapframe, thisdavidxu2004-08-091-12/+9
| | | | | | | | | | | | | | | | becauses some syscalls using set_mcontext can sneakily change parameters and later when those syscalls references parameters, they will wrongly use register values in mcontext_t. Approved by: peter
* | Clean up and add coments.julian2004-08-091-35/+66
| | | | | | | | | | reserve some bits in flag words for upcoming work. Note the unused bits so they can be found easier.
OpenPOWER on IntegriCloud