summaryrefslogtreecommitdiff
path: root/sys/arch/sh
AgeCommit message (Collapse)Author
2006-11-20serial drivers should use spltty, not splserial.Dale Rahn
2006-11-19Use symbol information, when available, to display branch destinations.Miod Vallat
2006-11-17sorry miod, but this version locks the machine up rather fastTheo de Raadt
2006-11-16pagemove() this time without an accidental line removalMiod Vallat
2006-11-16backout pagemove() diff; does not workTheo de Raadt
2006-11-14A less awkward pagemove() implementation.Miod Vallat
2006-11-14match other architectures: _spinlock_lock_t is not volatile itself, butTheo de Raadt
code that uses it declares a volatile form
2006-11-10relocations belong in reloc.hDale Rahn
2006-11-10Correct order of the bitfields of float and double. Fixes ldexp() (andOtto Moerbeek
some more), which fixeds drand48(), which fixes perl build. ok miod@ drahn@ deraadt@
2006-11-10Remove debug code.Miod Vallat
2006-11-10Enforce correct userspace bounds in copy{in,out}str().Miod Vallat
2006-11-10Will be needed at some point.Dale Rahn
2006-11-09landisk has no DCD line, so we must force TS_CARR_ON. This is a gruesomeTheo de Raadt
hack, otherwise the console tty is not quite right. Solve better someday.
2006-11-09no newline at end of cpu_modelTheo de Raadt
2006-11-08fix sigmask handling bug that caused SIGCHLD to be almost alwaysOtto Moerbeek
masked. DOH! miod@
2006-11-05Use TF_SPC instead of hardcoding its value.Miod Vallat
2006-11-05Correctly release resources in intc_free_ih().Miod Vallat
2006-11-05Fix a couple of comments.Miod Vallat
2006-11-05Some basic fpu trap handling.Miod Vallat
2006-11-05define the bit that sh4 uses for FPU disable.Dale Rahn
2006-11-04make compile w/ debugMichael Shalayeff
2006-11-04make stack trace look as other flavoured stack traces look likeMichael Shalayeff
2006-11-03cleanup some devicesMichael Shalayeff
2006-11-03Return address is way too useful to not print here.Dale Rahn
2006-11-03For now use 3 pages like netbsd, allows the kernel to get _MUCH_ farther.Dale Rahn
I suspect this is really a bug in the P1_STACK code path.
2006-11-03must steal pages from the end of the physsegMichael Shalayeff
2006-11-03mdproc has to be initialisedMichael Shalayeff
2006-11-03step and stuff works in ddb now tooMichael Shalayeff
2006-11-02Remove unused code and move the knowledge of struct switchframe to a singleMiod Vallat
place.
2006-11-02proper brkpt detection so continue does restart properMichael Shalayeff
2006-10-23Do not match on non-shpcic config stanza.Miod Vallat
2006-10-23Do not panic on a ddb invocation with curproc == NULL, and don't try toMiod Vallat
userret() either on return.
2006-10-23Add interrupt name, however need to determine if the passed parameter is safeDale Rahn
to store or might be on the stack.
2006-10-20Try again with local variable to avoid alignment issues.Kenneth R Westerback
"The dosparts member of cpu_disklabel was not used for any persistant data. Eliminate it, and just use the data being read in while processing MBR and EBR records. Should be no functional change. ok weingart@ deraadt@"
2006-10-19Hack around a problem found on px-eh systems where the SH IPL+g initializesDale Rahn
part of the PCI addresses, but not to values reasonable for us.
2006-10-19pci_address_fixup code to do bus address allocation, 'firmware' appearentlyDale Rahn
doesn't touch pci. ohci version and re mac address probe correctly.
2006-10-18back out disksubr changes ... since they violate strict alignment onTheo de Raadt
some architectures (kaboom); that will be fixed in the next iteration hopefully!
2006-10-17The dosparts member of cpu_disklabel was not used for any persistantKenneth R Westerback
data. Eliminate it, and just use the data being read in while processing MBR and EBR records. Should be no functional change. ok weingart@ deraadt@
2006-10-17for now, we want traces.Dale Rahn
2006-10-17Dont clobber tf->tf_spc, we might want to print it in do_panic.Dale Rahn
2006-10-16pciide compat bits, stolen from armish.Dale Rahn
2006-10-16busname should be set.Dale Rahn
2006-10-14ps uses KERNBASE so define it to the correct value (pointed out by miod)Dale Rahn
2006-10-11tsk, tskDale Rahn
2006-10-10Match the SH4 FPSCR bits.Miod Vallat
2006-10-07More bits in progress, adapted from NetBSD.Miod Vallat
2006-10-06debugging delay offMichael Shalayeff
2006-10-06few more filesMichael Shalayeff
2006-10-06Sync with recent disklabel.h changes. Zap dkbad.Kenneth R Westerback
"Free commits" miod@
2006-10-06a few fixes to get thru the autoconfMichael Shalayeff