summaryrefslogtreecommitdiff
path: root/sys/arch/mvme88k
AgeCommit message (Collapse)Author
2004-07-19When we want to thaw all work queues we care about, do it correctly.Miod Vallat
2004-07-19Direct REQUEST_SENSE commands to the correct target, instead of alwaysMiod Vallat
target zero. This makes targets other than zero working now... Spotted by nick@
2004-07-18TyposMiod Vallat
2004-07-13errno changes, lib major version bumps, and general flag dayMarco S Hyman
To build you must: cd /usr/src && make obj && make includes cd lib/libc && make depend && make && NOMAN=1 sudo make install cd /usr/src && make build
2004-07-13Remove compiler options workarounds for savectx breakage.Miod Vallat
2004-07-02Make page-aligned variable contiguous. Saves almost one page of data.Miod Vallat
2004-07-02Switch mvme88k to evcount, but keep maintaining intrcnt[] too for now.Miod Vallat
2004-07-02Disable a -O2 specific optimization which causes kern_fork (and probably otherMiod Vallat
files) to be miscompiled, until the problem is fixed in the compiler.
2004-07-02Be sure to fill intrhand structures correctly.Miod Vallat
2004-07-02Remove unused extra intrhand.Miod Vallat
2004-06-26When altering mmu registers, disable interrupts rather than splhigh().Miod Vallat
2004-06-24Typos in DEBUG code.Miod Vallat
2004-06-23Assorted night cleanups:Miod Vallat
- only compile debug and ddb-specific code if option DEBUG or option DDB. - shorten code path in 88100 case. - do not enable interrupts before the C handler is called, for the 88110 NMI trap.
2004-06-23Remove useless debug code.Miod Vallat
2004-06-23When servicing 88110 exceptions, reset dsr and isr as fast as possible.Miod Vallat
While there, simplify badaddr() handling on 88110.
2004-06-23Since the NMI interrupt source on 88110 is, well, non maskable, it isMiod Vallat
possible to receive it during autoconf. In this case, do not enable _other_ interrupts while we are servicing it. Note that, on the MVME197, since we are using the BUG for console output during autoconf, it is still possible to abort the bootstrap and return to the BUG with the NMI/ABORT switch.
2004-06-22During exception processing, we are not supposed to fault immediatelyMiod Vallat
after re-enabling the FPU (unless exception handling code is suffering from bugs, of course). Should this happen, invoke the error fault handler, but at least clear the exception state first, so that it gets a chance to run.
2004-06-22Change various macro and symbol names in order to have recognizable *100*Miod Vallat
and *110* names for processor-dependent code. No functional change.
2004-06-22Turn m88k_protection() into a macro again, compensating for 88110Miod Vallat
quirks locally where necessary; pmap_enter() does most of the dirty work.
2004-06-22After 88110 exception processing, do not reset the in-frame [di][lp]arMiod Vallat
values, they won't be put back to registers anyways...
2004-06-22On 88110, simplify enip processing:Miod Vallat
- always fetch its value when building trapframe, it is faster than only doing it after checking if it will have a meaningful value. - don't bother setting it to a safe value if we change exip to a non-delayslot address.
2004-06-22Do not access [id][pl]ar registers from C code - we only care about theirMiod Vallat
values in trapframes.
2004-06-19Share common ddb files accross m88k platforms.Miod Vallat
2004-06-10Fix a logic error in sglist creation.Miod Vallat
2004-05-26Define correct EOF and EOL characters in the default firmware termiosMiod Vallat
structure.
2004-05-26Use physical addresses as seen in the A24 MVME332XT address space, thisMiod Vallat
allows the driver to really don't care about its VME address and simplifies some code.
2004-05-25Make bpp_send() return the packet error code when invoked with NOWAIT.Miod Vallat
2004-05-25Make this driver aware of non-1:1 mappings.Miod Vallat
This currently mimcs the existing code, which can be cleaned up further.
2004-05-23Put back the fixed NCPUS == 1 version of flush_atc_entry().Miod Vallat
2004-05-23Simplify iomap_mapin().Miod Vallat
2004-05-23Nuke kvtop(), preserving a private copy in mvme88k vs(4) for now.Miod Vallat
2004-05-22Reset the board during probe. As a result, we can now honor the SCSIMiod Vallat
controller ID jumpers, instead of reprogramming the board to 7 every time.
2004-05-22Safer work queue thaw sequence.Miod Vallat
2004-05-22Convert to bus_space; but it still uses kvtop()...Miod Vallat
2004-05-21More cleaning, better registers programming, better error messages.Miod Vallat
2004-05-20More "can't help but clean things while reading".Miod Vallat
2004-05-20General cleanup before attempting to move towards bus_space and bus_dma.Miod Vallat
Also fix an arithmetic bug which caused almost no harm by luck.
2004-05-20Properly flush instruction cache for ptrace(PT_WRTIE_{DI}, ...) on powerpcMark Kettenis
and m68k. ok drahn@, millert@
2004-05-19No need to force a kernel mapping for the sram, the sram driver will do itMiod Vallat
if necessary.
2004-05-17KNF and minor cleaning.Miod Vallat
2004-05-16Simplify bugtty_chkinput(), from mvmeppc.Miod Vallat
2004-05-09Eliminate verbosity parameter to scsi_print_sense and the associatedKenneth R Westerback
chunk of code. It was never executed unless you were debugging a mvmex8k 'vs' device. ok miod@ (mvmex8k bits) marco@ deraadt@
2004-05-08Do not duplicate fragments from <machine/conf.h>.Miod Vallat
2004-05-07Replace the old ssh(4) NCR53C710 scsi driver with a pcctwo attachment forMiod Vallat
the MI osiop(4) driver. This improves performance very slighly and should also improve reliability. Tested on 25MHz (50MHz osiop) and 33MHz (66MHz osiop) MVME187 boards.
2004-05-07bus_dma(9) implementation for mvme88k, mostly based upon powerpc.Miod Vallat
2004-05-07Introduce a new cmmu operation, cachectl_pa, similar to cachectl, but takingMiod Vallat
a pa instead of a pmap_kernel va. The cachectl operation is now deprecated and will disappear soon.
2004-05-07Do not rely upon a fictitious MAXPHYSMEM value, but rather the actualMiod Vallat
physical memory size, to decide the end of /dev/*mem, as well as default cacheability for mappings.
2004-05-07Compute the size of the kernel page table at runtime, depending upon theMiod Vallat
board we run on, and its default mappings from pmap_table, rather than trying to compute a "fits all" value at compile time.
2004-05-06Do not leak resources if bus_space_map() fails during attach, or we don'tMiod Vallat
recognize the board address.
2004-05-06When the NVRAM is toast and we have to guess the CPU speed, choose a saferMiod Vallat
value depending upon the board type. This really only affects on-board SCSI on 1x7 anyway.
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-12 00:25:33 +0000