summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2002-05-11Avoid multiple definitions of macros in float.h and limits.h. ThisTodd C. Miller
is a long-standing problem that recently showed up on vax because due to some indirection cpp no longer believes the definitions are equivalent (though they really are). From espie@
2002-05-11Enable dz at uba, since it seems to move characters back and forthHugh Graham
in the SIMH/VAX simulator OK. If anyone has an actual card to test, a report would be appreciated.
2002-05-11Add missing libraries to bsd.prog.mk (mostly kerberosV)Marc Espie
Use them in DPADD throughout the tree. Fix a few mispells (LIBMATH -> LIBM...) Wipe obsolete lib (LIBRESOLV) Sort added missing libraries, move obsolete stuff apart. Synch documentation in bsd.README ok deraadt@
2002-05-10on ServerWorks CNB20HE Host:Michael Shalayeff
do not synthesyse secondary pci bus on the pchb0 rev 0x23 OR pchb1 w/ any revision. prompted by frisco@blackant.net in pr#2642
2002-05-10Some protection from userland and some __asm paranoia.Artur Grabowski
From NetBSD.
2002-05-10Mark all those __asm directives as clobbering memory, becauseArtur Grabowski
that's what they are supposed to do.
2002-05-10uhid changesTheo de Raadt
2002-05-09ukbd (and others) at uhidev, not uhubMarco S Hyman
2002-05-09On the hid -> hiddev changes, this file was missed.Dale Rahn
2002-05-09ukbd now at uhidev, not uhub. Ditto for ums (commented out)Marco S Hyman
2002-05-09sid max should be such that after conversion into prot-id it will not overflowMichael Shalayeff
2002-05-09add stats gathering for tlb handlers, count the calls forMichael Shalayeff
itlb, dtlb and tlb-dirty traps and cumulative time spent in each of those, being average 37 for the dtlb handler of 24 insns. move unrelated to traps code out of the traps way since profiled tlb handlers grew up and shifted the trap-all too far. insert required nops/sync and implied regs here and there. encode diag-reg insns in .word to see better what is being generated for the code.
2002-05-09Sync hid stuff including ukbd, ums, and uhid with NetBSDNathan Binkert
This adds a uhidev device which can be thought of as something like a uhid bus. It allows more than one ukbd, ums, or uhid to attach to the same device instance. This functionality is found on many of the newer keyboards that have extra buttons. (The extra buttons show up as uhid device(s)). Tested by me on i386, dale on macppc, and jason on sparc64
2002-05-09Make the paragraph degree key on a swedish iBook keyboard work in X.Mats O Jansson
The only problem is that its switched with the less greater key. Add an option FIX_SV_X_KBDBUG to fix that. -moj
2002-05-08Overhaul of the french layout for adb keyboards, fixing errors, adding a lotMiod Vallat
more useful symbols (such as ~ | \ { } etc) that were not available before. Requested by some French users, checked against MacOS's keyboard applet.
2002-05-08print parisc level w/ the the parisc version, not for each cpuMichael Shalayeff
2002-05-08kill debugging printfskiMichael Shalayeff
2002-05-08regenMichael Shalayeff
2002-05-08a few more boards and corrections, from pg-hwdbMichael Shalayeff
2002-05-07bring over our mods; the cnv*c have a problem now, described in the osf ↵Michael Shalayeff
version of the cnv_float.h, at the end
2002-05-07Add url(4) and urlphy(4)Nathan Binkert
ok deraadt
2002-05-07move ether_crc32_le to if_ethersubr.c. Add ether_crc32_beNathan Binkert
2002-05-07Spring pmap cleaning:Miod Vallat
- update and fix comments - clean the batc code (still not used by default) - there was some KNF left to do - fix incorrect locking behaviour in pmap_remove_all() - fix incorrect pmap_remove_range() call which would cause the kernel to enter an infinite loop sometimes at shutdown - fix an off-by-one page loop in pmap_protect() This should bring the extra bits of stability I need to resume working on the compiler...
2002-05-06Increase MAXPHYS from 63K to 64K which brings it in line with all ourTodd C. Miller
other architectures and matches NetBSD-current. hugh@ OK
2002-05-06Try to deal with the pci shadow bus issue by disabling a specific chipNathan Binkert
revision of the CNB20HE that always seems to be involved in the shadow bus. This is definately not the right way to do things. The correct way is to try to figure out that the bus is being detected for the second time. ok deraadt
2002-05-06drop out of any intr fixup activities if was disabled in the config flagsMichael Shalayeff
2002-05-06tuner driver for the fms cards.Michael Shalayeff
attach radio at the fms(4), instead of a separate device, which would never work out fine. from Vladimir Popov <jumbo@narod.ru> rework the radio config stuff (radio at radiobus, instead of radio at radio), allowing to tag source files in the files.* files w/ radio attribute. from mickey@
2002-05-03disable ahc again for now, since it does not fit; ok miodPeter Valchev
2002-05-03ptrdiff_t and size_t fixes.Marc Espie
okay mickey@
2002-05-03no guarantee that ptrdiff_t is int and not long, so cast.Marc Espie
ok mickey@
2002-05-03le_ioasic.h is historyTodd C. Miller
2002-05-03Need a dummy tc_fb_cnattach() until we have real fb support onTodd C. Miller
TURBOchannel machines.
2002-05-02Oops, it's too early for this chunk of code... spotted by millert@.Miod Vallat
2002-05-02Add a type specifier for intrnames, eintrnames, intrcnt, and eintrcnt.Todd C. Miller
Without this, n_type in struct nlist ends up as N_UNDF for those symbols which makes vmstat -i unhappy. mido@ OK
2002-05-02Big TURBOchannel support catchup from NetBSD, part 1.Miod Vallat
A few local changes and tweaks remain. This bring DEC 3000 machines back in the game, but framebuffers are still not supported at the moment. Thanks to ericj@ and nate@ for supplying me a DEC 3000 for testing.
2002-05-02update the nbsd rcsid to unconfuse some lamersMichael Shalayeff
2002-04-30unused stuffingMichael Shalayeff
2002-04-30uninitialized variable usage was hereMichael Shalayeff
2002-04-30yup, ahc fits easily again; requested by bradTheo de Raadt
2002-04-30Fix an ancient problem in how sparc interrupts are handled.Artur Grabowski
There are many interrupt handlers that assume that they don't need to do any spl protection in their code because the interrupt of some level can't be interrupted by an interrupt of the same level. The problem is that some interrupt handlers have hardware levels that are lower then their "software" levels. Fix this by adding an additional field to struct intrhand that specifies which "software" level an interrupt handler has and blocks that level while handling the interrupt. This new field is initialized in intr_establish which gets an additional argument (which can be -1 meaning that the interrupt handler doesn't need to block any additional level). ok deraadt@
2002-04-30dmesg logs show that many NICs run at prio 7. Bump IPL_NET to 7.Artur Grabowski
deraadt@ ok.
2002-04-29Enforce the bio <= net <= tty <= imp <= clock spl level ordering by fixingMiod Vallat
the tty spl, too low.
2002-04-29I want option DEBUG to rock my world.Miod Vallat
(and that means it should compile, for starters)
2002-04-29Better colors from rasops cmap, and enable all rasops features.Dale Rahn
ok miod, mattheiu
2002-04-29Use netisr_dispatch.Miod Vallat
2002-04-29Sync /dev/mem prototypes with the real interface here too.Miod Vallat
2002-04-29Do not forget to invoke iso_disklabelspoof() to deal correctly with cd-rom.Miod Vallat
2002-04-29Define placeholders for art's splassert() debugging stuff on all arches.Miod Vallat
Currently as no-ops everywhere. ok art@, deraadt@
2002-04-29wierd -> weirdPeter Valchev
2002-04-29Change macppc vgafb from rcons to rasops, a faster color supporting virtualDale Rahn
terminal emulator for gfx console. tested by mattheiu, miod.
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-04 20:08:44 +0000