summaryrefslogtreecommitdiff
path: root/sys/arch
AgeCommit message (Collapse)Author
1997-07-07only depend SFILES if there are some (to avoid a cc warning: no input files)Thomas Graichen
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...
1997-07-07DOH! I got this right in the mem case but not in the io one.Niklas Hallqvist
1997-07-07Too many things include <machine/ansi.h> w/o <sys/types.h> soTodd C. Miller
make time_t int not int32_t.
1997-07-06DDB (crippled), BOOT_CONFIG (not yet working), DISKLABEL_*, fxp, wd & acdNiklas Hallqvist
1997-07-06DDB files + some major number specifications that were missingNiklas Hallqvist
1997-07-06Safer implementations of *_raw_multi_* funcsNiklas Hallqvist
1997-07-06wd and acd supportNiklas Hallqvist
1997-07-06Fix the register saving when entering the debugger. Present the trap inNiklas Hallqvist
clear text too.
1997-07-06Use a string array to get the name of a trapNiklas Hallqvist
1997-07-06Enter DDB via CTL+ALT+ESC + style nitsNiklas Hallqvist
1997-07-06DDB support (much not yet implemented).Niklas Hallqvist
1997-07-06Add kernel versions of setjmp/longjmp, needed for DDB. Add an esym pointerNiklas Hallqvist
too that should be initialized by the bootloader with the end of the symtab.
1997-07-06If -b specified as bootflag (-d was taken) enter DDB. If using a PC keyboardNiklas Hallqvist
this does not yet work as the keyboard is not enough setup to work even in polled mode.
1997-07-06Make traps go into DDB if configuredNiklas Hallqvist
1997-07-06Make DDB support more complete (still faking a lot)Niklas Hallqvist
1997-07-06Erase the correct amount of columns. Remove static specifiers.Niklas Hallqvist
1997-07-06Match changes to other configs, missed in last commit.Jason Downs
1997-07-06Sync with changes from NetBSD, up to 970705.Jason Downs
This includes all of the new frodo drivers, from Michael Smith, hacked up by Jason R. Thorpe; other changes, mostly configuration and/or m68k abstraction related, from Thorpe. A few others of my own. This compiles and runs, have not yet tested Domain kbd or apci ports.
1997-07-06Sync with NetBSD changes, 970415 - 970705.Jason Downs
This includes a new asm.h, as well as even more code abstracted from hp300. These changes are likely to break ports that don't know about them; hp300 runs at the moment.
1997-07-06maxpartitions 16Todd C. Miller
1997-07-06Remove -Wcast-qual unless someone has a way to make it not careTodd C. Miller
about discarding "volatile". Implicit casts via prototypes cause major headaches in conjunction with volatile variables.
1997-07-05disable fxp for nowTheo de Raadt
1997-07-05Make _BSD_TIME_T_ int32_t on all platforms. You can't writeTodd C. Miller
lint-free code when these differ in name accross platforms. Fundamnetal type has not changed (was int on alpha and long on others).
1997-07-05implement unimplemented.....Michael Shalayeff
implement unimplemented..... implement unimplemented..... in cc -S it looks like it should.... in cc -S it looks like it should.... in cc -S it looks like it should....
1997-07-05Remove obsolete filesNiklas Hallqvist
1997-07-05The vga console bug is now fixed, argument ordering problem.Niklas Hallqvist
1997-07-04NWD for wd devices, include wd.hJason Downs
1997-07-03fxp support, turned off in alpha until testedTheo de Raadt
1997-07-02Adjust selection time out (all of my devices still work, does everyone elses?)Jason Downs
1997-07-01gestures in the direction of TurboSpPARC compatibilitygrr
1997-07-01make powerdown conditionalgrr
1997-07-01smooth char waiting loopMichael Shalayeff
1997-06-30remove obj@ dependency againkstailey
1997-06-30Add lib* so obj links get made. Duh, Todd.Todd C. Miller
1997-06-30install libs mode 444 like other libskstailey
1997-06-30Add MBR & i386/disklabel probing in a generic fashion usable for moreNiklas Hallqvist
disklabel types, and for other architectures if they chose. Both read/write supported for both alpha and i386/arc disklabels. ISO9660 spoofing kept and tested. Add "option DISKLABEL_ALL" to your config to use.
1997-06-27Add a missing space in usage message.Todd C. Miller
1997-06-26Hard-code definition of HZ if undefined (not spec'ed in an 'option'gene
statement). This is fallout from the recent config changes; pointed out by Steven Campbel.
1997-06-26fixup crazy indentation in Bread()kstailey
1997-06-26Compile sun4m-only kernels with -mv8.Jason Downs
1997-06-26Obviously, those assignments weren't wanted at all...Jason Downs
1997-06-25Switch to genassym.cf and sync Makefile with other ports.Jason Downs
1997-06-25Add proto.Jason Downs
1997-06-25== is not an assignment.Jason Downs
1997-06-25Add power.h for powerdown() proto.Jason Downs
1997-06-25Add netinet/in.h for in_cksum() proto.Jason Downs
1997-06-25Kill unused variables.Jason Downs
1997-06-25remove unused variableJason Downs
1997-06-25Include usage notes in the rest of the endian.h files for consistancy.grr