Subversion Repositories HelenOS-historic

Rev

Go to most recent revision | Show changed files | Directory listing | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
506 6749 d 16 h decky /kernel/trunk/ preparation for init task loading  
503 6749 d 18 h jermar /kernel/trunk/ Rename arch_late_init() to arch_pre_smp_init().
Add arch_post_smp_init().
Enable IRQ_KBD on ia-32 (doesn't work on Bochs SMP, works on Bochs UP).		  
501 6749 d 20 h jermar / Rename HelenOS/SPARTAN to HelenOS/kernel.  
493 6750 d 12 h jermar /SPARTAN/trunk/ Move genarch/{src|include}/firmware/ofw to genarch/{src|include}/ofw.  
492 6750 d 13 h jermar /SPARTAN/trunk/ Rename map_page_to_frame() to page_mapping_insert() and find_mapping() to page_mapping_find().
New names should make it more apparent that these two belong together.		  
486 6757 d 12 h jermar /SPARTAN/trunk/ Revert to -O3 for all architectures as the mips32 bug got fixed.  
484 6757 d 13 h palkovsky /SPARTAN/trunk/arch/mips32/src/ Fixed typo in context save.  
483 6757 d 15 h jermar /SPARTAN/trunk/ For now, each architecture must use its own -O switch (-O2 doesn't work for ia64, -O3 doesn't work for mips32).

New mips32 atomic_add() function.

Cleanup.		  
481 6759 d 12 h vana /SPARTAN/trunk/arch/ia32/include/ Revert of =m in ia32 atomic operations. In according to gcc docs +m should be right but as Jakub said it is not working
with amd64 but =m is... and I'm not able to build tools for that platform to test it now :-( . (Maybe bug in gcc ???)		  
478 6760 d 18 h jermar /SPARTAN/trunk/ Re-aply mistakenly reverted changes.  
477 6760 d 21 h vana /SPARTAN/trunk/ Atomic inc & dec functions synchronized on all ia32,ia64 and mips platforms. Now there are 3 versions which returns no value, new value and old value och changed variable.  
476 6760 d 22 h jermar /SPARTAN/trunk/arch/ia64/ ia64 CPU identification.  
475 6761 d 21 h jermar /SPARTAN/trunk/ Define atomic_t type.  
473 6762 d 12 h jermar /SPARTAN/trunk/arch/ia64/src/ ia64 work.
Comment some offending steps of heavyweight interrupt handler regarding RSE switching.
With this change, ia64 will do well on tests rwlock #4 and thread #1.
Cleanup.		  
472 6762 d 20 h jermar /SPARTAN/trunk/ Changes in build system.
For .S targets, always do -D__ASM__.
Remove unnecessary #define __ASM__ from various *.S files.
At the end of build, generate disassembler dump for kernel.raw.

ia64 work.
Better General Exception handler.		  
470 6763 d 9 h jermar /SPARTAN/trunk/ ia64 work.
Big cleanup of IA-64 interrupt processing.
Merge of interrupt.c and interrupt_handler.c.
Rewrite of ivt.S and interrupt.c.
Higher level interrupt handlers are now passed a vector number and a pointer to stack structure.

ia32 work.
ia32 has ordered writes. Until it deploys weaker memory ordering model, write_barrier() can be empty statement.		  
469 6763 d 17 h jermar /SPARTAN/trunk/arch/ia32/include/ Get the memory barriers on ia32 right.  
468 6763 d 17 h decky /SPARTAN/trunk/ fix IA-32 fences (1/2)  
467 6763 d 18 h jermar /SPARTAN/trunk/arch/ Add sparc64 context_offset.h.
Fix symbolic links in amd64 to be relative.		  
465 6763 d 20 h jermar /SPARTAN/trunk/arch/sparc64/ Add missing arch sources for sparc64.  

←Prev12Next→

Show All