summaryrefslogtreecommitdiffstats
path: root/sys
Commit message (Collapse)AuthorAgeFilesLines
* Make intr_nesting_level per-process, rather than per-cpu. Setupjake2001-01-2154-96/+124
| | | | | | | | interrupt threads to run with it always >= 1, so that malloc can detect M_WAITOK from "interrupt" context. This is also necessary in order to context switch from sched_ithd() directly. Reviewed By: peter
* wi has been converted to NEWCARD, so included it when pccard isimp2001-01-212-2/+3
| | | | | | | included. Also, I forgot to update this to the new cis[] structure last night, mainly due to the above omission.
* Correct botchup in last commit that made older VIA chips fail...sos2001-01-211-2/+2
|
* permit icmp6 type <= 256 (was 32).ume2001-01-211-1/+1
| | | | Obtained from: KAME
* Make the order of the static initializer for all_mtx match the order ofjasone2001-01-213-6/+6
| | | | | | fields in struct mtx. Found by: jake
* Update the iwic driver: fix error handling for rx errors on the D-channelhm2001-01-218-111/+103
| | | | | which prevents erroneous packets from being put onto the protocol stack; enhance error detection for B-channel HDLC errors; remove old cvs id´s.
* Add a bunch of entries from NetBSD's if_ne_pcmcia.c v1.67. We stillimp2001-01-211-5/+371
| | | | | need to verify that all the entries in pccard.conf are present in this table.
* Use PCCARD_CIS_xxx #defines for the table of oem ids. These usuallyimp2001-01-213-16/+22
| | | | translate to all NULLs (as for all the ones in this commit).
* Convert the vendor/product cis strings to an array of CIS entries. Thisimp2001-01-212-6/+6
| | | | | allows us to use the PCCARD_CIS #defines generated to distinguish cards that don't have a Manufacturer and OEM.
* Remove APIC_INTR_DIAGNOSTIC - this has been disabled for some time now.peter2001-01-2111-471/+0
| | | | Remove some leftovers of removed SMP options.
* Remove MUTEX_DECLARE() and MTX_COLD. Instead, postpone full mutexjasone2001-01-2122-150/+300
| | | | | | | | initialization until after malloc() is safe to call, then iterate through all mutexes and complete their initialization. This change is necessary in order to avoid some circular bootstrapping dependencies.
* Remove the MAYBE_PANIC and GUARD_CPU macros - unused.peter2001-01-211-34/+0
|
* Remove the per-cpu pages used for copy and zero-ing pages of memoryjake2001-01-2115-262/+11
| | | | | | for SMP; just use the same ones as UP. These weren't used without holding Giant anyway, and the routines that use them would have to be protected from pre-emption to avoid migrating cpus.
* Break the isa attachment of the Crystal Semiconductor 89x0 into twoimp2001-01-215-144/+347
| | | | | | | | | parts: isa and pccard. The isa one is known to work with an IBM EtherJet ISA card. The pccard one isn't known to work because the EtherJet pccard I purchased recently arrived DOA :-(. I'll commit the pccard.conf entry when the replacement card arrives. I plan on MFC this in a week or two.
* remove duplicate entry that snuck in during the merge.imp2001-01-212-8/+0
|
* Implement optimized PCPU accessors. These are necessary to support ajake2001-01-211-15/+90
| | | | | | pre-emptable kernel. For variables of size 4 bytes or less they compile to a single instruction, which does not allow a process to migrate cpus in the middle, and get the value for the "wrong" cpu.
* Make residuals signed.mjacob2001-01-211-2/+8
| | | | PR: 24221
* - All of proc_compare needs sched_lock, so hold it for the for loop thatjhb2001-01-201-40/+20
| | | | | | | | | | calls it rather than obtaining and releasing it a lot in proc_compare. - Collect all of the data gathering and stick it just after the proc_compare loop. This way, we only have to grab sched_lock once now when handling SIGINFO. All the printf's are done after the values are calculated. Submitted mostly by: bde
* When ip6_fw_ctl() or soopt_mcopyout() return without success,ume2001-01-201-1/+1
| | | | | | don't free mbuf. It is already freed by these routins. PR: kern/24248
* When short of mbufs or mbuf clusters, we sleep on appropriate "counters."bmilekic2001-01-202-9/+6
| | | | | | | | | The counters are incremented when a thread goes to sleep and decremented either when a thread is woken up by another thread or when the sleep times out. There existed a race where the sleep count could be decremented twice resulting in an eventual underflow. Move the decrementing of the "counters" to the thread initiating the sleep and thus remedy the problem.
* Document some more options.asmodai2001-01-202-22/+52
| | | | | Apologies to Bruce for not yet cleaning it up in sections. Coming soon.
* Rename the ASSYM MTX_RECURSE to MTX_RECURSECNT in order to not conflictjake2001-01-205-8/+8
| | | | with the flag of the same name.
* Simplify the i386 asm MTX_{ENTER,EXIT} macros to just call thejake2001-01-208-225/+40
| | | | | | appropriate function, rather than doing a horse-and-buggy acquire. They now take the mutex type as an arg and can be used with sleep as well as spin mutexes.
* Correct check for partition c. Previously the check was for drive 2,grog2001-01-201-1/+1
| | | | | | which did not exactly have the desired result. Submitted by: Akira Watanabe <akira@myaw.ei.meisei-u.ac.jp>
* Do not recursively include <sys/selinfo.h> from <sys/select.h>; all in-treewollman2001-01-201-6/+0
| | | | | | clients have been fixed. (So far as I can tell, all of the user-land clients of <sys/select.h> do so vacuously, in the expectation that select() might be declared there.)
* Temporarily disable the printf() for micruptime() going backwards, thejhb2001-01-201-0/+5
| | | | | | | SIGXCPU signal, and killing of processes that exceed their allowed run time until they can play nice with sched_lock. Right now they are just potentital panics waiting to happen. The printf() has bitten several people.
* - Make npx_intr INTR_MPSAFE and move acquiring Giant into thejake2001-01-2010-71/+15
| | | | | | function itself. - Remove a hack to allow acquiring Giant from the npx asm trap vector.
* Finish deprecating <sys/select.h> in favor of <sys/selinfo.h> in kernel code.wollman2001-01-208-6/+14
|
* Be more careful with sched_lock in the SIGINFO handler. Specifically, dojhb2001-01-201-10/+21
| | | | | | | | not hold sched_lock while calling ttyprintf(). If we are on a serial console, then ttyprintf() will end up getting the sio lock, resulting in a lock order violation. Noticed by: des
* Sync to pccarddevs 1.9imp2001-01-201-2/+17
|
* Sync with NetBSD (again, after I got the most up to date file):imp2001-01-201-1/+10
| | | | | | | o 1.111 imp Sync with FreeBSD. o 1.110 nonaka Add Roland SCP-55 o 1.109 ichiro add NANOSPEED ROOT-RZ2000 WLAN Card o 1.108 ichiro add EMTAC A2424i 11Mbps WLAN Card
* sync with pccarddevs 1.8.imp2001-01-201-2/+2
|
* Correct entry for RATOC REX_R280 to comply with style used in the restimp2001-01-201-1/+1
| | | | of the file.
* Sync with pccarddevs.imp2001-01-201-9/+45
|
* Sync with NetBSD. v1.98->1.107:imp2001-01-201-4/+22
| | | | | | | | | | | | | | o 1.107 haya Add IO Data CBIDE2 ata interface card. o 1.106 jhawk s/NULL}/NULL }/ o 1.105 thorpej " " -> "&sp" in two entries, per Rafal Boni. o 1.104 thorpej Add SMC 2632W. From Rafal Boni, kern/11775. o 1.103 drochner add IBM microdrive o 1.102 soren Typo. o 1.101 hubertf Add ELSA WaveLAN card & a noname clone(?) o 1.100 toddpw Socket Comm. PC Card Ethernet, and tidy up naming. o 1.99 msaitoh MELCO LPC2-TX, Telecom Device TCD-HPC100, MACNICA ME1-JEIDA o 1.98 imp Sort BAY into numerical order. (already in FreeBSD)
* Check the return value of sf_encap() and handle errors accordingly.wpaul2001-01-202-6/+22
|
* Set the OACTIVE flag if vr_encap() fails.wpaul2001-01-202-0/+2
|
* Silence compiler warnings.wpaul2001-01-202-4/+4
|
* Bug fixes that I've put together while working on a project in the office:wpaul2001-01-194-18/+38
| | | | | | | | | | | | | | if_vr: handle the case where vr_encap() returns failure: bust out of the packet sending loop instead of panicking. Also add some missing newlines to some printf()s. if_dc: The miibus_read and miibus_write methods keep swapping in and out of MII mode by fiddling with CSR6 for cards with MII PHYs. This is a hack to support the original Macronix 98713 card which has built-in NWAY that uses an MII-like management interface even though it uses serial transceivers. Conditionalize this so that we only do this on 98713 chips, since it does bad things to genuine tulip chips (and maybe other clones).
* Make some attempt to accomodate kern/24221- propate residual to mt_residmjacob2001-01-191-0/+41
| | | | | | | even if it might overflow. If we do a read or set of hardware or logical block position, we also clear Tape Frozen status.
* Change the reinstall.debug target to depend onjhb2001-01-195-20/+40
| | | | | | {kernel,modules}-reinstall.debug rather than {kernel,modules}-reinstall. Otherwise, the '.debug' portion of the target is lost, and you end up reinstalling the non debug version instead of the debug version.
* Sync with GENERIC - drop I386_CPUpeter2001-01-191-1/+0
|
* Convert apm from a bogus 'count' into a plain option. Clean out somepeter2001-01-1916-40/+34
| | | | other cruft from the files.alpha and files.ia64 that were related to this.
* Zap unused #include "apm.h"peter2001-01-194-4/+2
|
* Zap some cut-paste code that isn't used. It #included "apm.h" andpeter2001-01-193-16/+0
| | | | an apm include file but had no other apm references.
* Zap #include "apm.h" - it was not used.peter2001-01-191-1/+0
|
* Use #ifdef DEV_NPX from opt_npx.h instead of #if NNPX > 0 from npx.hpeter2001-01-1921-106/+112
|
* EEK! vm86bios.s has got #if NNPX > 0 code without a correspondingpeter2001-01-191-0/+2
| | | | | #include "npx.h" - the code has been dead for a while and vm86 calls have not been saving FPU context it seems.
* At great personal risk to my sanity, turn off COMPAT_OLDISA and thepeter2001-01-192-6/+4
| | | | | two drivers that depend on it - ie and le. The compat code has not been disabled.
* Add missing twe (3ware) and ahb (adaptec 174x) devices (!)peter2001-01-192-0/+14
|
OpenPOWER on IntegriCloud