summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
2007-04-02unmapiodev() takes the va as its first argument, not the pa.Mark Kettenis
ok gwk@
2007-04-02Remove bogus code. Turning on parity for pci busses may be a good idea, butMark Kettenis
this is not the way to do it. And given the way Sun host bridges crap out upon access to pci config space for nonexistent devices, this code is rightout harmful for machines where device 0 isn't present on some busses. Tested by many; ok deraadt@
2007-04-02Return EFBIG in case not enough segments are present to load a DMA requestClaudio Jeker
instead of ENOMEM and remove/shortcut the additional debug printfs. Some network drivers will try to load the mbuf chain and linearize the mbufs if EFBIG is returned. OK miod@ kettenis@
2007-04-01Add code to establish interrupts.Mark Kettenis
2007-04-01Shift pci tags for PCIe.Mark Kettenis
2007-04-01Provide (non functional) dma tag.Mark Kettenis
2007-04-01Shift tags when converting them into an address offset. Necessary to supportMark Kettenis
access to PCIe config space, which is larger than normal PCI config space.
2007-04-01Fix DEBUG built.Mark Kettenis
ok art@
2007-04-01Remove some leftovers from the userret cleanups.Artur Grabowski
No need to save sticks anymore. miod@ ok and test
2007-03-31do nbuf constraint based on KERNEND/KERNBASE to avoid integer overflow from ↵Theo de Raadt
new giant sparc64 kvm; ok kettenis, tested by pooleb
2007-03-31Map PCIe config space.Mark Kettenis
2007-03-31Fix DEBUG.Mark Kettenis
2007-03-31Make sure we not to match secondary cpu's on non-MP kernel. I accidentallyMark Kettenis
didn't commit this bit, so it is still ok drahn@, gwk@.
2007-03-31Detect secondary cpu's. Get rid of legacy ofroot code.Mark Kettenis
ok drahn@, gwk@
2007-03-30Add commented out entries for pyro(4).Mark Kettenis
2007-03-30First stab at a driver for the Host/PCIe bridge found on the Sun FireMark Kettenis
v215/v245/v445 systems.
2007-03-29Add power hook to scoop(4), replacing the Zaurus-specific scoop_* calls inUwe Stuehler
pxa2x0_apm.c. Set a flag to indicate that scoop is currently suspended. Test this flag in scoop_timeout() (and maybe later in other places) to avoid manipulating scoop registers while the driver is suspended. The scoop_checkdisk timeout is intentionally not stopped, as it should be safe to have timeouts running accross suspend/resume. Put splhigh() in scoop_gpio_pin_write() to make changing the pin level an atomic operation. Result of prodding and diffs from Robert and Theo.
2007-03-29ignore various i2c seeproms that sun puts on all their machine parts; ok ↵Theo de Raadt
miod kettenis
2007-03-29Enable acx(4) as it is finally working on bigendian boxes. OK deraadt@Claudio Jeker
2007-03-29Comment fixes from Daniel Dickman <didickman AT gmail DOT com>, relevant changesGordon Willem Klok
for amd64 added by myself and while here nuke some whitespace.
2007-03-29Equiv diff to amd64 as was done on i386, bump VM_PHYSSEG_MAX. Amd64 areTobias Weingartner
PCs too... Tested by dlg@ and brad.
2007-03-28disk addresses are in hex, not decimal. makes my LSILogic,sas@3/disk@13David Gwynne
correctly match target 19. ok deraadt@
2007-03-28Bump max number of physical segments. Machines in the wildTobias Weingartner
seem to have reached this limit. ok marco@, looks ok ckuethe@
2007-03-27only actually change the drive LED state if there was a real state change;Theo de Raadt
tested by robert and myself
2007-03-27When trap() causes a panic, have stacktrace() use ddb to find the symbolMiod Vallat
names if option DDB, instead of the fn_name() helper which only knows about trap() anyway. While there, do not attempt to print the function arguments after the first frame, since they are known to be clobbered.
2007-03-27Try and print symbol-relative addresses for branches, and displayMiod Vallat
delay slots instructions correctly.
2007-03-27grok LSILogic,sas controllers in bootpath; as well, become aware that (atTheo de Raadt
least) "scsi" controllers give a target/port number, not a kernel sd#. So, translate. We might later want to do the same translation for IDE as well, but perhaps not until after people test a bit more; ok dlg
2007-03-26software-driven drive light code, using a timeout. uwe wrote some codeTheo de Raadt
initially that used a disk_busy() hook, but this uses a lot less cpu and is MD; ok uwe
2007-03-26Rip out the KERN_ error codes.Artur Grabowski
ok otto@
2007-03-26i shouldn't be shy about commiting ahci to this.David Gwynne
lots of poking by jsg@
2007-03-26ichss_cookie is a bad name, and should not be a void pointer.Gordon Willem Klok
Spotted by Stefan Sperling <stsp AT tsp.in-berlin.de> when reviewing his piix speedstep diff. ok tedu@
2007-03-24Call the new ohci_checkrev function on xscale, otherwise the USB versionChristopher Pascoe
is not set correctly and the USB bus does not attach. Tested ckuethe@
2007-03-23Do not enter ddb from the interrupt handler unless option DEBUG.Miod Vallat
2007-03-23Real atomic_{set,clear}bits_int implementation, and replace similarMiod Vallat
{set,clr}_ipending with the above routines. ok kettenis@
2007-03-23Define PROC_PC and simplify userret(); ok kettenis@ (and I think drahn@ too)Miod Vallat
2007-03-23Clean up the failure path of the attach routine, check the return value ofGordon Willem Klok
dbdma_alloc, and mapiodev, and some whitespace/long line cleanup. Tested on 9500MP and 9600MP. ok martin@
2007-03-23Fix compiliation of MULTIPROCESSOR without DDB, being defined.Gordon Willem Klok
From: Mike Belopuhov <mkb AT crypt.org DOT ru>. ok kettenis@
2007-03-23Remove duplicated defines from intr.h that are already defined inArtur Grabowski
intrdefs.h that is inlucded first thing in intr.h. eyeballed for sanity by pedro@
2007-03-22Do not define ISA_DMA_STATS by default; ok deraadtMiod Vallat
2007-03-22Move powerpc to __HAVE_MUTEX. With help from drahn@. Tested by nick@, xsa@,Mark Kettenis
deraadt@. "reads right" deraadt@
2007-03-22Change yet another kernel message starting with `<'.Miod Vallat
2007-03-22When registering VME interrupts, maintain a direct ipl->vector table, as longMiod Vallat
as no more than one interrupt is registered for a given level. Then, if the VME interrupt vector reading cycle fails on the 188 interrupt arbiter, we can use this table as a hint if it has a valid entry, since we know on which ipl line the interrupt occured. This basically silences the m188_ext_int: timeout getting VME interrupt vector, level 3, mask 0x400<IRQ3> occasional messages appearing when the MVME376 is overloaded.
2007-03-22In cmmu routines, replace splhigh() with disable_interrupts(), saves a functionMiod Vallat
pointer indirection for a similar result; also move the interrupt disabling code to the public routines, so that we do not end altering the psr more than necessary.
2007-03-22split userland & kernel struct sensor/sensordev so that the additionTheo de Raadt
of new fields in the future is less disruptive. This is done similar to how struct proc is handled for ps(1). ok jmc (man page changes) tested fkr simon, and more suggestions from millert
2007-03-21no need to include uvm_extern.h hereMartin Reindl
2007-03-21Map the kernel text read only. Because we can.Miod Vallat
2007-03-21Uninitialized variable in db_write_bytes().Miod Vallat
2007-03-21ahci seems to work fine on i386. thanks to marc winiger for figuring thatDavid Gwynne
out for me.
2007-03-21Real atomic_{set,clear}bits_int implementation, ok deraadtMiod Vallat
2007-03-21Real atomic_{set,clear}bits_int implementation, no performance hit on 060Miod Vallat
unless used on non-aligned data; ok deraadt@
generated by cgit v1.2.3 (git 2.25.1) at 2024-12-29 19:20:05 +0000