summaryrefslogtreecommitdiff
path: root/sys/arch/sparc
AgeCommit message (Collapse)Author
1998-02-26Change order of cpu initialization to force the correct dmesg output.Jason Wright
1998-02-24compensate for new gasTheo de Raadt
1998-02-18installboot onto partition cTheo de Raadt
1998-02-12Look for installboot in /usr/mdec unless user specified otherwise.Magnus Holmberg
1998-02-11From NetBSD: sun4m anti crash code in IOMMU_FLUSH{PAGE,ALL}Marco S Hyman
1998-02-05sync with netbsd: 4/300 dma fixJason Wright
1998-02-05oops, need include fileTheo de Raadt
1998-02-05rename ddb.panic_ddb, and add ddb.console. Now you can stop console ddb entryTheo de Raadt
with a sysctl. There will be architectures and drivers that lack function, and I trust the maintainers of those will forget to add the code..
1998-01-28Undo the 4/300 dmaselect hack and fix it correctlyJason Wright
1998-01-14For some reason the 4/300 doesn't like ncr53c9x_dmaselect.Jason Wright
Added code to check to see if parent device == obio, if so, disable dmaselect More research to be done here for a real fix (if possible)
1998-01-13Workaround for fpu_cleanup panic. We just ignore it. I'm going toJason Wright
look further into really fixing this problem.
1997-12-25do not break the bootblocksTheo de Raadt
1997-12-23use raster console by default since the sun prom is buggy.chuck
1997-12-21ifdef some definitions of NBPG that use kernel variables, when theAngelos D. Keromytis
file is included from userland programs (such as lsof)
1997-11-11blush!Niklas Hallqvist
1997-11-11make tty col/row size be unsignedNiklas Hallqvist
1997-11-11whoops typo in last commit + accept 0 in col/row values on sun4 mean full szNiklas Hallqvist
1997-11-11Limit the console size by the frambuffer size. My 4/100 seemsNiklas Hallqvist
to lie about the console size
1997-11-11A vital variable was not initialized! Thanks DDB, and you...Niklas Hallqvist
remember to kill those static specifiers.
1997-11-11I might as well say it here: please do not make functions in theNiklas Hallqvist
kernel static, unless they are also inline. Debugging in DDB becomes a nightmare when the symbols are not there. Instead use good non-clashing names, maybe prefixed by the devicename, subsystemname or similar.
1997-11-11Remove thos evil statics, some KNFNiklas Hallqvist
1997-11-11unused variable under certain conditionsNiklas Hallqvist
1997-11-11Remove those static specifiersNiklas Hallqvist
1997-11-09New endian.h design. All architectures now use a common file for the bulkNiklas Hallqvist
part. Some ports can optimize the swap operations. This also means the general API is extended with swap16, swap32, htobe16, htobe32, betoh16, betoh32, htole16, htole32, letoh16 and letoh32.
1997-11-07simple_lock api changed slightlyTheo de Raadt
1997-10-23correct obj buildingTheo de Raadt
1997-10-01make rawpart FS_UNUSEDTheo de Raadt
1997-09-20only eject root floppiesTheo de Raadt
1997-09-17Double default NMBCLUSTERS settings.Jason Downs
1997-09-17Reorganizaed bootblocks. Builds like hp300 now.Jason Downs
1997-09-17NETBSD_CURRENT_970916. Lot's just ID changes, since changes don't apply toJason Downs
us. Includes some pmap changes, for which I don't have the original commit message(s) handy.
1997-09-15kill Locore.cTheo de Raadt
1997-09-11Turn off "branch folding" in microsparc II processors. This seems toTheo de Raadt
avoid strange panics happening. Patch suggested by Chris Torek.
1997-09-11let's at least list all scsi devices, ok?Todd T. Fries
1997-09-01Thanks go to Jason Downs <downsj@openbsd.org> who virtually handed me thisTodd T. Fries
patch.
1997-08-25Implement LED blinking as a run-time configurable option via the sysctlJason Downs
machdep.led_blink. Rename auxreg.h because we need config to generate one now.
1997-08-24-msupersparc is bad news on the microSparc.Jason Downs
1997-08-08Change the bounds_check_with_label API to also take a cpu_disklabelNiklas Hallqvist
reference for making transferral of meta-information possible from readdisklabel to bounds_check_with_label. The first (and maybe only) thing that will use this is the multi-disklabel-format code on the alpha where the labelsector is passed via cpu_disklabel so the label write-protection can work correctly no matter what label was found. Also use a new macro DKBAD to get at the dkbad field of the cpu_disklabel implementations that contain it. This too is for multi-disklabel architectures where the "bad" field can be inside a union. Use this macro as a means for a driver to check if an architecture supports dkbad constructs. Remove proto of bounds_check_with_label from all MD disklabel.h as it is in sys/disklabel.h. I have not been able to test the changes everywhere, if I break anything I apologize, and promise to fix it as soon as I become aware of it.
1997-08-08Updated from GENERIC.Jason Downs
1997-08-08Update GENERIC_SCSI3 from GENERIC and add tags.Jason Downs
1997-08-08Missed these...Jason Downs
1997-08-08Mostly sync to NetBSD-current 970804.Jason Downs
GENERIC currently compiles and runs; some devices (isp) are not complete and not yet enabled.
1997-08-01proto __flt_rounds correctlyTheo de Raadt
1997-07-26Must include netatalk/at_extern to get atintr() prototype.denny
1997-07-24multiple inclusion protection; brenner@umr.eduTheo de Raadt
1997-07-23Add AppleTalk netisr. Only tested (indeed, only compiled) on mac68k port.denny
Add options NETATALK to your config to try it out.
1997-07-13Sync with the new audio changes.Angelos D. Keromytis
1997-07-08Added -msupersparc for SUN4M-only configs, (right after -mv8) -msupersparcjohns
supposedly adds additional instruction scheduling for better performance, where -mv8 seems to only "allow" the multiply and div instructions. This is what I got from the gcc docs anyway.. Reading the RTL for sparc seems to agree with this hypothesis, and the man pages..
1997-07-07Fix a major sun4m stability problem showed up as random segv's and othergrr
failures in innocent programs. Typically something like a make build of libc or something executing a large number of shell commands would fail, dumping core in sh or random faults in as or cc*. According to Chris Torek, the problem is that he make the destination pages for zero and copy be non-cached to prevent wiping otherwise relevant stuff from the cache. On systems with a L2 cache like the SS10 modules with cache, setting the pages as non-cacheable inhibits updating the chache, and also disable PA chache snooping, doesn't do anything about possible dirty lines in the L2 cache, which may eventually get flushed and corrupt the nice new page. It's not clear that he's got the failure mode 100% correct, but making the pages cacheable does seem to avoid the symptoms and testing the e-cache flush hypothesis is difficult. Aaron Brown proposed a fix that was conditional on mmutype, and netbsd-current has it conditional on a cpu.quirk, at the moment this is unconditional, the performance hit is arguable...
1997-07-07Too many things include <machine/ansi.h> w/o <sys/types.h> soTodd C. Miller
make time_t int not int32_t.
generated by cgit v1.2.3 (git 2.25.1) at 2025-01-09 10:22:23 +0000