summaryrefslogtreecommitdiff
path: root/sys/arch/i386
AgeCommit message (Collapse)Author
2005-11-23kill two old mach macros that are not used anywhere in the tree anymore (or ↵Michael Shalayeff
ever)
2005-11-23assume vtophys(vaddr_t) just what all the other archs expect; no functional ↵Michael Shalayeff
change
2005-11-23prioritise <16m memory for isa bus_dma'ble memorble memory allocationsMichael Shalayeff
2005-11-23resolve a couple of problems in mpbios-mapped interrupts:Michael Shalayeff
- synthesise isa mappings (as 1-1) should those be missing in mpbios; - for rcc osb* firce "special" ints into isa mappings always. niklas@ ok and testing by many since
2005-11-23merge probe and match since one is only ever called from the other. knf tooDavid Gwynne
yup deraadt@ ok marco@
2005-11-22Added probing for Dell ESM serverJordan Hargrave
ok marco@
2005-11-22support one of the powerunitsDavid Gwynne
2005-11-22add another backplaneDavid Gwynne
2005-11-22shrink a littleDavid Gwynne
2005-11-22do not manually inline kvtopte() macro; no func changeMichael Shalayeff
2005-11-22pm_pdirpa is apaddr_t and PTDpaddr (on the contrary) is u_int32_tMichael Shalayeff
2005-11-22bios gives 32bit units not paddr/psize; no func changeMichael Shalayeff
2005-11-22for every sensor esm looks after it has to talk to the hardware, whichDavid Gwynne
means it could busy wait while the hardware gets ready. all the sensors are updated out of a single timeout. doing this for the 30ish esm sensors on my machine is probably not a good thing... this changes the way updates are handled. instead of a single big update for all the sensors, we now update only one sensor in a timeout. after that sensor has been updated it sets up a very short timeout for the next sensor to run out of. we also use the short timeouts to retry the same sensor if the hardware isnt ready which means we avoid the (worst) busy wait.
2005-11-22repair verbose printfsMichael Shalayeff
2005-11-22get rid of bmc_wait and use a flag on bmc_ready to indicate if we wantDavid Gwynne
to wait for the port to settle. lengthen the amount of time we wait to what jordan had originally.
2005-11-21add support for the backplane found on my 2500David Gwynne
2005-11-21add a handful of Intel PCI interrupt routers to the table.Brad Smith
82440MX_ISA, 82801AB_LPC, 82801E_LPC, 82801FBM_LPC, 82801GBM_LPC and 82801GH_LPC ok mickey@ beck@
2005-11-21normalise the values read off the hw for the voltagesDavid Gwynne
2005-11-21spacingTheo de Raadt
2005-11-21Fix copyrights.Marco Peereboom
ok jordan@ dlg@
2005-11-21Move contents of sys/select.h to sys/selinfo.h in preparation for aTodd C. Miller
userland-visible sys/select.h. Consistent with what Net and Free do. OK deraadt@, tested with full ports build by naddy@.
2005-11-21Add ESM support glue but keep it disabled for now.Jordan Hargrave
ok dlg@ marco@
2005-11-21Add support for ESM (Embedded Server Management) for Dell PowerEdge servers.Jordan Hargrave
Lot's of work by dlg@ ok marco@
2005-11-19Remove unnecessary lockmgr() archaism that was costing too much in termsPedro Martelletto
of panics and bugfixes. Access curproc directly, do not expect a process pointer as an argument. Should fix many "process context required" bugs. Incentive and okay millert@, okay marc@. Various testing, thanks.
2005-11-19Remove superfluous statement-ending semi-colons. i.e., "return foo;;" shouldAaron Campbell
just be "return foo;". Cleaning out some M's in my tree.
2005-11-18add some comments explaining the general CPU reset code and the AMD GeodeBrad Smith
SC1100 specific reset code. From NetBSD ok mickey@
2005-11-18splimp -> splvm. no binary diff.Brad Smith
ok mickey@
2005-11-15add pcn on alpha and remove redundant pcn entry on i386 RAMDISK_CDBrad Smith
2005-11-15add ixgb(4) though commented out for now.Brad Smith
2005-11-14convert and remove the last traces of i386_round_page(),Martin Reindl
i386_trunc_page(), i386_btop() and i386_ptob()
2005-11-13move kgdb_connect and kgdb_panic into MI. looked by miod@ and kettenis@Federico G. Schwindt
2005-11-13need cast to vaddr_tMartin Reindl
found the hard way by deraadt@
2005-11-13convert more MD macros to their MI counterparts, this timeMartin Reindl
i386_round_page(), i386_trunc_page(), i386_btop() and i386_ptob()
2005-11-11disable ahc0 at isa until we find the bugs that cause this to blow upTheo de Raadt
on a few rare machines. of course, by the time we do, they might all have lots their smoke anyways. it is probably because of the vlbus/eisa addressing it tries to do. ok krw
2005-11-10only 4 bits of an irq number are valid to set (just in case)Michael Shalayeff
2005-11-10do not use the array index before a range check is doneMichael Shalayeff
2005-11-10make sure to ignore >4g and split on 4g memory segments from the bios map; ↵Michael Shalayeff
tom@ ok
2005-11-10sensors.h is not used here, so remove the includeDavid Gwynne
ok grange@
2005-11-04only build powernow pieces as appropriate for config'd cpu typesTed Unangst
2005-10-28lost a diff from gordon, only attach if we can do something usefulTed Unangst
2005-10-28malloc(M_WAITOK) is bad in init code, M_DEVBUF is nicer than M_TEMP,Ted Unangst
and if the cpu gets stuck, don't spin too long.
2005-10-28update support for powernow (cool and quiet) on k7, and add supportTed Unangst
for k8. preliminary and not well tested yet. from freebsd via gordon klok.
2005-10-27repair ops. looking at the doc i suppose had no doc and no good clue when ↵Michael Shalayeff
wrote it; tested on osb5
2005-10-27ironically everything past 8231 has 8 pci lines (we assume)Michael Shalayeff
2005-10-27apparently 8235 has 8 pci lines too (ommited from the docs); tested by jolan@Michael Shalayeff
2005-10-26add vt8233 and vt8235 (at least 8235 verified in docs)Michael Shalayeff
2005-10-26reverse 8237 check to only match 8237Michael Shalayeff
2005-10-26add via 8237 back, works nowHenning Brauer
2005-10-26in set_trigger only touch triggers not whole routing crapMichael Shalayeff
2005-10-26fix 8237 trigger maskMichael Shalayeff
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-27 02:15:14 +0000