summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2007-06-24rework sensor tasks to use the kernels generic workq rather than a specialDavid Gwynne
kernel thread of its own. the api has changed (which will be fixed in the manpage shortly) so all the users of sensor tasks that i can find have been fixed too. noone tested, so its going in to force people to run with it. "put it in" deraadt@
2007-06-23Enable workaround for decrementer exception problem.Dale Rahn
2007-06-23net80211 growth means vr(4) goes away for nowTheo de Raadt
2007-06-22PCI malo(4) worksMartin Reindl
from Janjaap van Velthooven
2007-06-21Brace got lost in rev 1.51.Miod Vallat
2007-06-21Enable memory above 256MB again now that bus_dma has been fixed.Miod Vallat
2007-06-21Extent sgi bus_dma to cope with different views of memory: non-contiguousMiod Vallat
for the cpu, contiguous from different bases for devices. This allows memory above 256MB to be used with bus_dma (and we had really been lucky with the first few large-memory builds). Information about memory accesses taken from Linux.
2007-06-21Fix microtime to not lose clocik ticks, gives us among other things, realMiod Vallat
ping times on slow links; from mickey
2007-06-21Simple optimizations:Miod Vallat
- in __pmap_pv_enter, only walk the pv list to search for writeable mappings if we are ading a readonly mapping. - im pmap_protect, do not invoke pmap_extract(), instead directly extract the paddr from the pte we have already computed a few lines above.
2007-06-21Attach interrupt counters to intc interrupt sources too, so that clockMiod Vallat
interrupts get counted; from mickey
2007-06-20Do not hardcode imask[] size when intializing it.Miod Vallat
2007-06-20These drivers were abusing b_cylinder to store device-specific informationMiod Vallat
in strategy(), and were reusing it in start(). I first considered introducing a b_rawblkno field in struct buf, as has been done in NetBSD, to stop this abuse. However, it does not cost more to simply move the device-specific ``was-b_cylinder'' computation to the start() routine. Plus we get type fixes (daddr64_t) for free, although this does not really matter for these pedro-sized devices. Thus we do not need to grow struct buf for these devices which really ought to live in your Attic (and my machineroom, of course). hp300 HP-IB tested, vax mfm not, but the logic is the same.
2007-06-20alpha kernel have wsmux but not the actual devices; PR #5469 from janjaapMiod Vallat
2007-06-20Add file such that a kernel with lpt at ssio actually compiles.Mark Kettenis
2007-06-20b_cylinder does not need to be set on the callpath down into drivers.Theo de Raadt
cpu_disklabel can go away, since nothing anymore needs to use it.. except these two drivers. knock out that code for now to let it compile while miod gets his fix ready for these drivers.
2007-06-20b_cylinder does not need to be set on the callpath down into drivers.Theo de Raadt
cpu_disklabel can go away, since nothing anymore needs to use it; ok miod
2007-06-20lpt at ssioMark Kettenis
2007-06-20In vunmapbuf(), explicitely remove mappings before invoking uvm_km_free().Miod Vallat
Even if the latter would end up removing the mappings by itself, it would do so using pmap_remove() because phys_map is not intrsafe; but some platforms use pmap_kenter_pa() in vmapbuf(). By removing the mappings ourselves, we can ensure the remove function used matches the enter function which has been used. Discussed and theoretical ok art@
2007-06-20Protect pool operations with splvm.Miod Vallat
2007-06-20Make sure IPL_CLOCK blocks device interrupts.Miod Vallat
2007-06-19com at ssioMark Kettenis
2007-06-19Add ssio(4), a driver for the castrated National Semiconductor PC87560Mark Kettenis
Sucky SuperIO chip. For now it only provides support for the i8259-compatible PICs on the chip, but that support is necessary to make USB interrupts work.
2007-06-19Empty another cpu_disklabel. In this case read in the bootKenneth R Westerback
block/native label as needed rather than keeping it in cpu_disklabel. Also a number of assorted minor tweaks to reduce the delta to sparc64 disksubr.c. 'get this in' deraadt@
2007-06-19Mark PDE entries with PG_U and PG_M. We never need that information from theArtur Grabowski
mmu, it slightly speeds up tlb misses and according to an errata from AMD it can actually work around a bug in the mmu. toby@ ok
2007-06-18Lookup interrupt routing info based on elroy(4) bus number.Mark Kettenis
Fix and simplify debug code that prints the interrupt routing info while I'm there.
2007-06-18Calculate the disklabel checksum *after* filling in all the fields inKenneth R Westerback
the disklabel. ok deraadt@
2007-06-18this is the sgi disklabel, mips64 is not the sameTheo de Raadt
2007-06-18this is the sgi disksubr, not some mythical "all mips64 architectures can ↵Theo de Raadt
use it" balony
2007-06-18Don't use cpu_disklabel as local storage. Use local variables and discardKenneth R Westerback
the unneeded info when done with it. Another empty struct cpu_disklabel. ok deraadt@
2007-06-18Don't use cpu_disklabel as local storage. Use local variables and discardKenneth R Westerback
the unneeded info when done with it. Another empty struct cpu_disklabel. ok deraadt@
2007-06-18Use a shorter form to load XKPHYS constants in .S code, shaves a few textMiod Vallat
bytes, no functional change.
2007-06-18Disable instruction reordering around cpl assignments.Miod Vallat
2007-06-18delete comment above cpu_disklabel that is not neededTheo de Raadt
2007-06-18move comment to right placeTheo de Raadt
2007-06-18split out dpme disklabel reading, simplifying things significantlyTheo de Raadt
and making it look a lot more like other architectures; tested by myself and drahn and gwk
2007-06-18simplify error return case from writedisklabel(), tooTheo de Raadt
2007-06-18simplify error return case from writedisklabel()Theo de Raadt
2007-06-18simplify error return case from writedisklabel()Theo de Raadt
2007-06-18more comments that are differentTheo de Raadt
2007-06-18using same words in similar code sequences is goodTheo de Raadt
2007-06-18this XXX has no place here anymore. extended partition label writingTheo de Raadt
might even work now (ie. if it works, it is because finding the label location is now MI code shared for the read & write paths)
2007-06-18include not neededTheo de Raadt
2007-06-18msg is already gauranteed to be NULLTheo de Raadt
2007-06-17make it compileMartin Reindl
2007-06-17Eliminate a few straggling 'labelsector' field declarations fromKenneth R Westerback
cpu_disklabel's. 'labelsector' is not used anywhere according to grep. 'go for it' deraadt@
2007-06-17Get interrupt information from PDC and use it to program the IO SAPIC.Mark Kettenis
Needed to make edge triggered interrupts work.
2007-06-17Disable memory above 256MB for now, it triggers bugs (probably related toMiod Vallat
cache operations).
2007-06-17Coerce this into compiling.Miod Vallat
2007-06-17Some alpha (or SRM) designs use level 3 for i/o interrupts, instead of theMiod Vallat
more commonly encountered level 4. Do not complain in splassert() in this case (this is similar to the older vsbus vax machines workaround).
2007-06-17(error in commit messages to other similar MI files; see position inTheo de Raadt
ChangeLog to see other files) avoid pulling in machine/disklabel.h when sys/disklabel.h is a better choice
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-06 01:01:25 +0000