summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2002-03-14Remove a bunch of #if 0 codeJason Wright
2002-03-14Make the sbus ipl handling more like psycho. Some devices (Aurora 210sj, eg)Jason Wright
use Sbus level 5 for interrupts which means spltty() is not blocked correctly because the sbus interrupt handling uses the implied spl of the interrupt instead of the level passed to it.
2002-03-14From NetBSD: driver for lpt at ebusJason Wright
2002-03-14add apio* at sbus? and lpt* at apio?Jason Wright
2002-03-14Do not include headers twice. Pointed out by Dries Schellekens.Dale Rahn
2002-03-14protect from overflows and null derefs in cpu probing; from ↵Michael Shalayeff
wd@ics.nara-wu.ac.jp via pr#2457
2002-03-14this mod has nothing to do w/ ptrace option addition, did not even compileMichael Shalayeff
2002-03-14Properly conditionalize uvm history dumpJason Wright
2002-03-14use getprop, not getpropstring to fetch 'magma_prom' property...Jason Wright
ra->ra_name is needed and is the result of getpropstring
2002-03-14add asio* at sbus? and com* at asio?Jason Wright
2002-03-14Final __P removal plus some cosmetic fixupsTodd C. Miller
2002-03-14First round of __P removal in sysTodd C. Miller
2002-03-14Remove PTRACE support out of the ramdisks/miniroots.Miod Vallat
2002-03-14Turn the ptrace(2) syscall into a kernel compile option, option PTRACE inMiod Vallat
your kernel configuration file. By default, GENERIC will enable this. When PTRACE is not enabled, several ptrace-like features of the procfs filesystem will be disabled as well (namely, the ability to read and write any process' registers, as well as attching, single stepping and detaching to/from processes). This should help paranoid people build better sandboxens, and us to build smaller ramdisks.
2002-03-13Complete rewrite of the powerpc pmap handling, Instead of keepingDale Rahn
the spill list for each PTEG, the V->P translations are stored in trees for each pmap. All valid kernel mappings are preallocated in 1-1 memory so that tlb spill/loads for kernel accesses can be looked up while physical, user mappings are not guaranteed to be 1-1 mapped, thus the kernel must go virtual to look up user mappings. While this is more expensive, the tree search is much lower cost than the long linked list search. Also on each pmap_remove() it was necessary to search the linked lists for each possible mapping, now it just looks up the entry in the tree. This change gives a 25-36% speedup in 'make build' time. What was around 2:50 is now around 1:55 on a 733MHz G4. This change causes a likely existing bug to appear quite often, it deals with the segment register invalidation in kernel mode. Because of that problem, currently this change limits the physical memory used to 256MB. This limitation will be fixed soon, it is not an error in the pmap code. * Effort sponsored in part by the Defense Advanced Research Projects * Agency (DARPA) and Air Force Research Laboratory, Air Force * Materiel Command, USAF, under agreement number F30602-01-2-0537.
2002-03-13On sparc, PAGE_SIZE and friends might not be a compile-time constant.Miod Vallat
Instead of using a homegrown set of variables in this case, rely on uvmexp fields once uvm has been initialized. This requires a few #include <uvm/uvm_extern.h> here and there in the kernel as well. Idea from art@, changes by me.
2002-03-12rest of copy_on_fault proper handlingsMichael Shalayeff
2002-03-12sparc64's aren't likely to be CPU_ISSUN4CJason Wright
2002-03-12Look up the TTE and only do the streaming cache dance if the mapping is ↵Jason Wright
marked as streaming.
2002-03-12extraneous 'or'; from Sven Kirmess <sven.kirmess@gmx.net>Michael Shalayeff
2002-03-12Emulate ptrace(PT_STEP in software.Artur Grabowski
More or less directly from FreeBSD.
2002-03-12sync with KAMEKenjiro Cho
ALTQify more drivers. ok millert@
2002-03-12oopsMichael Shalayeff
2002-03-11If the clock has lost over 1000 days, use the filesystem time insteadDale Rahn
of the hardware clock. Prevent clock screwage due to lost battery on laptops where the hardware clock goes back to 0. From NetBSD.
2002-03-11Add <machine/ieeefp.h>, will be needed by libc shortly.Miod Vallat
2002-03-11Add a <machine/varargs.h> skeleton for userland.Miod Vallat
2002-03-10rdboot -> hdbootHugh Graham
2002-03-09Completely rewritten keyboard detection mechanism. Instead of using theDale Rahn
openfirmware commands (which apple seems to have removed), walk the device tree attempting to detect what type of keyboard(s) are present. Give preference to USB if USB and ADB both found. If none found hope for USB. This does have a problem on RevC iMacs booting without USB keyboard, it seems apple left the adb device tree information which indicates an ADB kebyoard is present. This fixes newer ibook and PBG4 keyboard detection.
2002-03-09Add case for BTYP_1305.Hugh Graham
2002-03-08Remove the code which reenables interrupts in these interrupt controllers,Dale Rahn
they do not appear to be capable of handling all possible nested interrupts.
2002-03-08have to set p_md as well, used to shoot clouds beforeMichael Shalayeff
2002-03-08last arg to the copystr could be 0, also correct onfault handlingMichael Shalayeff
2002-03-08Move the message buffer out of low memory, Openfirmware clears the areaDale Rahn
on reboot. perhaps OF uses it at other times? Since OF always use the same memory addresses, this should always allocate the same ram to the msgbuf, and allow it to be preserved across reboot.
2002-03-07Clean up BUS_DMA_* flags, and make sure all arch's define:Jason Wright
BUS_DMA_READ, BUS_DMA_WRITE, and BUS_DMA_STREAMING
2002-03-07Remove some debugging code accidentally committed in previousJason Wright
2002-03-07From NetBSD:Jason Wright
Fix calculation of dma segment length when the DVMA range is crossing the boundary.
2002-03-06- recognize serial speeds over 9600 bauds, and allow users to set such aMiod Vallat
speed (beware, old PDC will not support this!) - increase the arbitrary device limit for the lists; some machines (743 and 748 come to mind) can have lots of heads and serial ports... - be more clever when sorting the devices in our lists, so as to get the correct order on machines where the built-in serial ports are GIO but on different busses.
2002-03-06Support serial speed over 9600 bauds.Miod Vallat
2002-03-06no need to maskMichael Shalayeff
2002-03-06expand fpu name, makes line overflow a bit, maybe deal w/ it laterMichael Shalayeff
2002-03-06regenMichael Shalayeff
2002-03-06cleanup the fpu names a bitMichael Shalayeff
2002-03-05rearrange stuff, chat about cacheMichael Shalayeff
2002-03-05Kill prom stuff duplicates from bugio.[ch].Miod Vallat
2002-03-05Insert missing newline for autoconf prettiness.Jason Wright
2002-03-05Provide a few more constants to get on par with other arches and getMiod Vallat
raidframe tools to compile; ok mickey@
2002-03-04ok, so apparently wx doesn't work so wellNathan Binkert
2002-03-04dc is not listed twice. Put de back, and move dc out of the untested section.Nathan Binkert
While we're at it, move wx out of the untested section too.
2002-03-04since we can't fit both dc and de, we've got to have de so we can have supportNathan Binkert
for 2104x chips which dc does not support at all.
2002-03-04Move dc from the untested group since it is tested, though a commit willNathan Binkert
soon make de win over dc on alpha in all cases, so de would have to be disabled to use dc. (For 21140 and 21142/3 chips. Clones will use dc bydefault.)
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-04 21:09:58 +0000