summaryrefslogtreecommitdiff
path: root/sys/arch/sgi
AgeCommit message (Expand)Author
2009-07-13Extend xbridge to support shared interrupt handlers, and perform PCI-PCIMiod Vallat
2009-07-11When rebooting, pass a request to the PROM for it to skip diagnostics andMiod Vallat
2009-07-06Almost rewrite xbridge PCI resource allocation:Miod Vallat
2009-07-03Try to be smarter when allocating pci resources, and use unused devio slotsMiod Vallat
2009-07-01The widget mapping code has been written back when I was only working onMiod Vallat
2009-06-28Attempt to map memory resources of unconfigured devices, if they do not haveMiod Vallat
2009-06-27Preliminary cleanup work in order to help the PCI resource management code.Miod Vallat
2009-06-27In the bus_space_*_raw_* functions, move the invariant computations out ofMiod Vallat
2009-06-27If a device which has not been initialized by ARCS has a non-zero BAR, ignoreMiod Vallat
2009-06-27Add an explicit widget locator to xbow children; allows a given widget toMiod Vallat
2009-06-24Even more pci ethernet devices, based on a diff from Brad.Miod Vallat
2009-06-24add some pci ethernet devices that some people, like miod, might be usingTheo de Raadt
2009-06-24for the isp(4) driver, invert the meaning of the options which pullTheo de Raadt
2009-06-21Disable memory above 2GB physical again, as we can end up with an unkillableMiod Vallat
2009-06-21Remove the ioc interrupt probe code, the heuristic is correct; origin 300Miod Vallat
2009-06-21Add a bunch of usb devices, per deraadt@'s pressure.Miod Vallat
2009-06-21- add and enable stge(4)Jasper Lievisse Adriaanse
2009-06-19Make the 64 bit bootblocks compile again after the <machine/mnode.h> changes.Miod Vallat
2009-06-17Make sure xbowmatch() only matches on xbow stanzas.Miod Vallat
2009-06-17Make sure we do not coalesce memory ranges from different nodes.Miod Vallat
2009-06-15Back out all the buffer cache changes I committed during c2k9. This reverts t...Bob Beck
2009-06-14Don't mess with the bounds in the lp if *partoffp is non-NULL, since thisTheo de Raadt
2009-06-13Enumerate all available nodes for hardware on IP27/IP35 systems; works to someMiod Vallat
2009-06-13Figure out which nodes are connected, and use all of their memory for now;Miod Vallat
2009-06-13Shuffle and rename HUB defines for consistency, no functional change yet.Miod Vallat
2009-06-12Make sure splsoftnet() masks softclock; fallout form switch to genericMiod Vallat
2009-06-11The MAC multicast filter is set in mec_iff() hence there is need to do thisJoel Sing
2009-06-10Switch sgi to per-process AST, and move ast() from interrupt.c to trap.cMiod Vallat
2009-06-10In the NMI handler, always print EPC even if we're going to drop into ddb.Miod Vallat
2009-06-10Add udl and url, verified to work on Fuel (url to the RAMDISK too)Miod Vallat
2009-06-05use the right accounting for bound edges (i think)Theo de Raadt
2009-06-04More fallback from struct disklabel changes, yawnMiod Vallat
2009-06-04Recycle four ancient fields in the disklabel structure, replacing them withTheo de Raadt
2009-06-04Don't bother setting RB_SERCONS in boothowto since this flag is not usedMiod Vallat
2009-06-03add kern.bufcachepercent sysctl to allow adjusting the buffer cacheBob Beck
2009-06-03- add hotplug deviceJasper Lievisse Adriaanse
2009-06-03Arla client rename from xfs to nnpfs for later upgrades. Tested on various ar...Janne Johansson
2009-06-01Enable zyd in more places now we know it is endian safe.Jonathan Gray
2009-05-31Remove BUS_DMAMEM_NOSYNC definition. Its name is not consistent withMiod Vallat
2009-05-30When booting in `install software' mode from the SGI boot menu, do not try toMiod Vallat
2009-05-28Be more paranoid in the IOC3 interrupt klugeMiod Vallat
2009-05-28Poor man's reset code for IP27/IP35 where returning through ARCBios doesn'tMiod Vallat
2009-05-28Handle NMI interrupts on IP27/IP35, gives us a change to play with ddb,Miod Vallat
2009-05-28If readsgilabel() is invoked in spoofing mode, only report the whole deviceMiod Vallat
2009-05-27Use the KL enumeration functions to try and figure out what kind of XBowMiod Vallat
2009-05-27Yet another attempt at a more reliable detection of the second interruptMiod Vallat
2009-05-27Rework KL board and component enumeration, to allow it to be used fromMiod Vallat
2009-05-27Make sure onboard devices on Octane get their DMA byteswap settings correct.Miod Vallat
2009-05-25Years ago, I fixed an R5000 O2 instability by implementing a workaround forMiod Vallat
2009-05-24Provide more xbridge-specific functions in the bus_dma_tag_t we use forMiod Vallat