summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* Get the build bits right for the new Architecture Independant null- andmarkm2000-06-2510-1/+23
| | | | | entropy drivers. Reviewed by: dfr(mostly)
* Strip out the machine-independant parts of the memory device.markm2000-06-254-514/+100
| | | | | | /dev/(u)random, /dev/null, /dev/zero are all moving to machine-independant drivers. Reviewed by: dfr
* Strip out the machine-independant parts of the memory device.markm2000-06-252-243/+50
| | | | | /dev/(u)random, /dev/null, /dev/zero are all moving to machine-independant drivers.
* New machine-independant /dev/random driver.markm2000-06-256-0/+800
| | | | | | | | | | | | | | | | This is work-in-progress, and the entropy-gathering routines are not yet present. As such, this should be viewed as a pretty reasonable PRNG with _ABSOLUTELY_NO_ security!! Entropy gathering will be the subject of ongoing work. This is written as a module, and as such is unloadable, but there is no refcounting done. I would like to use something like device_busy(9) to achieve this (eventually). Lots of useful ideas from: bde, phk, Jeroen van Gelderen Reviewed by: dfr
* New machine independant /dev/null and /dev/zero driver. This device ismarkm2000-06-252-0/+274
| | | | | | | | | | | | | | | severely stripped down compared with its predecessor, and is measurably a _lot_ faster. Many thanks to Jeroen van Gelderen for lots of good ideas. There is still a problem with this; it is written as a mudule, and as such is theoretically unloadable. However, there is no refcounting done as I would prefer to do that a'la device_busy(9), rather than some "home-rolled" scheme. The point is pretty moot, as /dev/null is effectively compulsory. Reviewed by: dfr
* Typo fix: tunnable.mph2000-06-251-3/+2
| | | | | Backslash-parsing is not the role of printf(3), but is provided by the C language.
* Remove obsolete comment.bp2000-06-252-10/+0
| | | | Submitted by: Marius Bendiksen <mbendiks@eunet.no>
* If swdpio1 doesn't clear, we have a reversed (or disconnected) cable. Changemjacob2000-06-251-3/+3
| | | | | | the message to indicate that it could also be a disconnected cable, and return okay from wx_hw_intialize *anyway*. This allows us to contineu to set the station address and when we do get link up, we're ready to roll.
* Fixed atpic_attach() for the SMP (specifically APIC_IO) case.fsmp2000-06-244-4/+8
| | | | Approved by: msmith@freebsd.org
* Remove obsoleted info about linking from contribache2000-06-241-16/+2
|
* Stupid me, I put the opt_tdfx.h underneath a test for TDFX_LINUX, whichcokane2000-06-242-4/+9
| | | | | resides in opt_tdfx. I also cleand up that large define. Compile, tell me if it does, and I'll re-enable the tdfx entry in the makefile.
* Back out previous commit until the build of the 'tdfx' module is unbroken.archie2000-06-231-1/+1
|
* There is no atdevbase on the alpha. Hide this case statement on alpha untilmjacob2000-06-231-0/+3
| | | | the design/implementation is complete.
* Fix the breakage that occurred when somebody didn't do a completemjacob2000-06-231-2/+2
| | | | checkin of the changes to kern_prot.c.
* Add tdfx to modules Makefile, so it gets made.cokane2000-06-231-1/+1
|
* Oops! Disabled the ed driver becasue it cannot be compiled.kato2000-06-231-1/+1
| | | | Pointed out by: nyan
* Include pc98/pc98/pc98.h in which M_EPSON_PC98 is defined when thekato2000-06-231-1/+3
| | | | EPSON_MEMWIN option is specified.
* Make the PnP 'slopsucker' quiet in the !bootverbose case - the real NPXmsmith2000-06-233-3/+15
| | | | | probe happens much earlier, and may come to very different conclusions about the system's NPX setup.
* Add a stub driver to consume the PnP "system resource" items, and hidemsmith2000-06-234-0/+196
| | | | them in the !bootverbose case.
* Add PnP probe methods to some common AT hardware drivers. In each case,msmith2000-06-2319-2/+1014
| | | | | | | | | the PnP probe is merely a stub as we make assumptions about some of this hardware before we have probed it. Since these devices (with the exception of the speaker) are 'standard', suppress output in the !bootverbose case to clean up the probe messages somewhat.
* Collect the ISA DMA defines from the MI ISA code, not a private copy.msmith2000-06-231-3/+2
|
* Stop trying to do anything funny with the interrupt resource range. Themsmith2000-06-234-52/+16
| | | | AT PIC will consume IRQ 2 correctly in the !APIC_IO case.
* Fix typo (inT -> int)dima2000-06-231-1/+1
|
* Finish up the fixing of the linux ioctl stuff. Add line to remove the devcokane2000-06-232-2/+4
| | | | | | entry upon unload. Fix the module makefile.
* Finish up the fixing of the linux ioctl stuff. Add line to remove the devcokane2000-06-232-1/+6
| | | | entry upon unload.
* Minor redundant #include fix.cokane2000-06-231-1/+0
|
* Fixed trouble with linux emu, not it should work in the module.cokane2000-06-233-10/+6
| | | | Removed unnecessary warning message too.
* Fixed problem with linux ioctl code, module loading should work now.cokane2000-06-232-0/+59
|
* Remove these here, these were repocopied to src/sys/ufs/ffs.peter2000-06-233-5532/+0
|
* fix races in the uidinfo subsystem, several problems existed:alfred2000-06-2212-82/+120
| | | | | | | | | | | | | | | | | | | | | | | | 1) while allocating a uidinfo struct malloc is called with M_WAITOK, it's possible that while asleep another process by the same user could have woken up earlier and inserted an entry into the uid hash table. Having redundant entries causes inconsistancies that we can't handle. fix: do a non-waiting malloc, and if that fails then do a blocking malloc, after waking up check that no one else has inserted an entry for us already. 2) Because many checks for sbsize were done as "test then set" in a non atomic manner it was possible to exceed the limits put up via races. fix: instead of querying the count then setting, we just attempt to set the count and leave it up to the function to return success or failure. 3) The uidinfo code was inlining and repeating, lookups and insertions and deletions needed to be in their own functions for clarity. Reviewed by: green
* fix warning, declare function static.alfred2000-06-221-1/+1
| | | | Reviewed by: dfr
* Rename macros to all-uppercase. Get rid of a comment that was ironicgreen2000-06-221-10/+4
| | | | | (I goofed on the bitshifts myself long ago ;) and a bit redundant: code should be clear enough that it seldom needs comments at all.
* Make the generated set headers depend on Makefile as well (that is wherepeter2000-06-224-8/+8
| | | | | | | the list of .o files comes from - if we change the list of files that are built, we need to rescan the files) Obtained from: bde (indirectly)
* Put RF_SHAREABLE into the bus_alloc_resource call.cokane2000-06-221-1/+1
|
* Switch LINUX_TDFX to TDFX_LINUX, that's what i get for typing thiscokane2000-06-221-1/+1
| | | | at 2 in the morning.
* Inhibit successful DAD messages and "no default interface" messages.ume2000-06-221-2/+4
| | | | | | | It seems that people find them too noisy. (ND6_DEBUG will enable them) Obtained from: KAME Project
* Add a hack to fail registration of kq events on a non-ufs filesystem, asjlemon2000-06-221-0/+8
| | | | support for those is non-existent at the moment.
* Add code so that the udata field is preserved across a TRACK event.jlemon2000-06-221-9/+18
| | | | | | | | | When re-adding an event, do not reset the event state. If the event was pending, it will remain pending. This allows the user to change the udata field after the event was registered, while not losing any events which have already occurred. Reported by: jmg
* correct bad TTL with packets generated by v4 mapped udp. from kameitojun2000-06-221-0/+7
|
* Add 'kern.disks', a sysctl which returns the list of disks fromnbm2000-06-221-0/+29
| | | | | | | disk_enumerate(), space delimited. This allows non-root users to get a list of disks and will simplify libdisk's Disk_Names(). Reviewed by: phk
* Sync with sys/dev/syscons/scterm-sc.c revisions 1.6 and 1.7.nyan2000-06-222-14/+34
|
* Whitespace-only changes: apply accepted line breaking style as asheldonh2000-06-221-91/+138
| | | | | | pre-cursor to other markup changes. Accepted in principle by the author.
* Functions may be static and __inline, but not extern and __inline. Thisgreen2000-06-221-1/+1
| | | | should fix broken no-"-O" kernel builds.
* Add my name next to the majro for 3dfx (107), to reflect new info contact,cokane2000-06-221-1/+1
| | | | due to recent commit.
* Add SOFTUPDATES to GENERIC (BOOTMFS has this filtered out)peter2000-06-222-0/+2
|
* Took TDFX_VERBOSE out of Makefile, added $FreeBSD$ and option for DEBUG.cokane2000-06-223-8/+18
| | | | | Added a notice for linux users that the kld isn't working for them yet, so no linux emu from the kld. It compiles statically alright though.
* Converted all TDFX_VERBOSE ifdefs to DEBUG, now it gets debugged with thecokane2000-06-221-1/+1
| | | | | | rest of the kernel. Added new option TDFX_LINUX for optional LINUX_EMULATION, so that perhaps some people don't have to use linux emu to run the driver if they don't want to.
* Stick in header, $FreeBSD$.cokane2000-06-224-28/+188
|
* Stick the tdfx entry in here, after some cleaning.cokane2000-06-222-0/+2
|
* Fix the 'file not found' in the load target, someonecokane2000-06-221-1/+1
| | | | forgot to add the '.ko' to the end of the module filename.
OpenPOWER on IntegriCloud