Subversion Repositories HelenOS

Rev

Go to most recent revision | Show changed files | Details | Compare with Previous | Blame | RSS feed

Filtering Options

Rev Age Author Path Log message Diff
780 6735 d 2 h palkovsky /kernel/trunk/ Small fix in slab allocator for large-sized slabs.  
773 6735 d 21 h palkovsky /kernel/trunk/ Added slab stress test.
Fixed race condition in slab allocator.
Moved initialization of slab to the point where we know correct config.cpu_count
Correctly passes tests on amd64, ia32 in both bochs & qemu.
 
772 6735 d 22 h palkovsky /kernel/trunk/ Fixed non-initialized cpu-cache.  
768 6736 d 23 h palkovsky /kernel/trunk/ Debugged slab allocator. It currently supports per-CPU cache on 1 cpu.  
767 6737 d 1 h palkovsky /kernel/trunk/ Partially working SLAB CPU cache.  
766 6737 d 2 h palkovsky /kernel/trunk/ Basic SLAB (without CPU-cache) passes test.  
764 6737 d 7 h palkovsky /kernel/trunk/ Tested basic non-cached slab allocation.  
762 6737 d 9 h palkovsky /kernel/trunk/ Currently not-working SLAB allocator.
- slightly changed interface to frame_alloc, allow zone preference
 
759 6737 d 21 h palkovsky /kernel/trunk/ Skeleton of SLAB allocator.  
756 6738 d 23 h jermar /kernel/trunk/ Memory management work.
Remove the last (i.e. 'root') argument from page_mapping_insert() and page_mapping_find().
Page table address is now extracted from the first (i.e. 'as') argument.
Add a lot of infrastructure to make the above possible.
sparc64 is now broken, most likely because of insufficient identity mapping of physical memory.
 
755 6739 d 23 h jermar /kernel/trunk/ Change page_mapping_find/insert interfaces to take as_t * as first argument
and not asid_t as second argument. This change was necessitated by the
removal of mapping array from as_area_t and the fact that an address
space doesn't have an ASID when it is created.
 
745 6745 d 2 h jermar /kernel/trunk/test/mm/ Fix and cleanup frame allocator tests.
falloc1:
- fix so that it compiles on ia64
- check that frames is not NULL
falloc2:
- threads cannot be passed argument in the way this test did
(otherwise some threads will use the same value for val)
- check that frames is not NULL
- free frames at the end of the test
 
735 6748 d 6 h bondari /kernel/trunk/ Fixes, comments, tests to frame allocator  
733 6748 d 8 h bondari /kernel/trunk/test/mm/falloc2/ frame allocator test #2  
724 6753 d 2 h palkovsky /kernel/trunk/ Commit of the falloc_bad branch to trunk (719:723).  
692 6762 d 6 h bondari /kernel/trunk/ Frame allocation test #2  
689 6762 d 7 h bondari /kernel/trunk/ Frame alloc test #1  
687 6762 d 7 h jermar /kernel/trunk/ Memory management work.
Pass address space identifier to functions in page_operations to support single global page hash table.
 
674 6766 d 12 h cejka /kernel/trunk/ Support for printing float numbers in kernel removed.  
627 6786 d 6 h jermar /kernel/trunk/ Fix ia64 and sparc64 to compile with new atomic_t.
Fix rwlock test #5 and semaphore test #1 to compile with new atomic_t.

sparc64 work.
TBA must be set before a function call when MMU is switched off.
 
625 6786 d 7 h palkovsky /kernel/trunk/ We are now almost -Wall clean.
- redefined atomic_t
- deleted many, many unused variables
- some minor code cleanups found using compiler warning.
 
623 6786 d 13 h jermar /kernel/trunk/ Add SPINLOCK_DECLARE and SPINLOCK_INITIALIZE macros.
SPINLOCK_DECLARE is to be used instead of direct spinlock_t declarations
in dynamically allocated structures on which spinlock_initialize() is called after
their creation.
SPINLOCK_INITIALIZE is to be used instead of direct spinlock_t declarations
of global spinlocks. It declares and initializes the spinlock.
Moreover, both macros are empty on UP so that -Wall warnings about unused structures
get supressed.
 
609 6789 d 6 h palkovsky /kernel/trunk/ added support for breakpoints to mips32  
548 6795 d 4 h jermar /kernel/trunk/ Cleanup and fixes.  
534 6797 d 4 h jermar / Rename HelenOS/SPARTAN to HelenOS/kernel again.  
532 6797 d 4 h jermar / Revert renaming of HelenOS/SPARTAN to HelenOS/kernel because of important pending patch.  
501 6809 d 8 h jermar / Rename HelenOS/SPARTAN to HelenOS/kernel.  
492 6810 d 1 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.
 
483 6817 d 4 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.
 
475 6821 d 9 h jermar /SPARTAN/trunk/ Define atomic_t type.  
413 6845 d 0 h jermar /SPARTAN/trunk/ Rename cpu_priority_{high|low|restore|read} functions to interrupts_{disable|enable|restore|read}.
Rename pri_t to ipl_t (Interrupt Priority Level).
Rename thread_t::pri to thread_t::priority.
 
401 6854 d 14 h jermar /SPARTAN/trunk/test/mm/mapping1/ Finalize mm mapping test #1 once again.  
400 6854 d 14 h jermar /SPARTAN/trunk/ Finalize mapping1 test.
Record omitted excMod bug in msim.
 
399 6855 d 0 h jermar /SPARTAN/trunk/ mips32 memory management work.
TLB exceptions seem to be working on at least simics.
msim handles TLB Refill and TLB Invalid exceptions well,
but seems to skip writes to the mapped addressies in mm
mapping test #1 (the data is not written and exception
is not triggered).
 
397 6855 d 13 h jermar /SPARTAN/trunk/ Small changes to mips32 mm.
Small changes in mm mapping test #1.
 
396 6856 d 11 h jermar /SPARTAN/trunk/ Add mm/mapping1 test.
(Will not make it past TLB Invalid exception on mips32.)
Fixes in asid.c.
Make TLB register types union with __u32 value.
Implement tlb_invalidate() for mips32.
(TLB invalidation and shootdown path will have to be revised.)
 
329 6882 d 6 h palkovsky /SPARTAN/trunk/ Basic changes to boot it on SGI Indy.  
326 6882 d 22 h palkovsky /SPARTAN/trunk/ Cleanup of makefiles to have common options in one main makefile.

Add simple build process for different simulators for MIPS.
Added FPU context & lazy FPU context switching to MIPS.
Cleanup of MIPS linker script.
Moved MIPS kernel above 1MB. Not tested on real machine yet, but it might help.

There is something broken with gcc inlined memcpy (either simulator or gcc), it is disabled on BigEndian mips now.
 
296 6887 d 12 h palkovsky /SPARTAN/trunk/ More logical fpu_context_engaged setting.
The sse1 test now correctly tests for thread migration settings.
 
295 6887 d 12 h palkovsky /SPARTAN/trunk/ Test that really tests SSE context switching.