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
2170 6395 d 9 h jermar /trunk/kernel/ Simplify synchronization in as_switch().
The function was oversynchronized, which
was causing deadlocks on the address
space mutex.

Now, address spaces can only be switched
when the asidlock is held. This also protects
stealing of ASIDs. No other synchronization
is necessary.
 
2161 6397 d 12 h jermar /trunk/kernel/arch/sparc64/src/mm/ Fix TSB bug during TSB refill.
When one wants to enable a TSB entry, he or she should set the
entry invalid bit to false, as opposed to setting it to true.
 
2144 6398 d 9 h jermar /trunk/kernel/arch/sparc64/src/mm/ Fix TSB size.  
2143 6398 d 16 h jermar /trunk/kernel/arch/sparc64/include/ Remove unneeded enum member.  
2141 6399 d 8 h jermar /trunk/ The Ultimate Solution To Illegal Virtual Aliases.
It is better to avoid them completely than to fight them.
Switch the sparc64 port to 16K pages. The TLBs and TSBs
continue to operate with 8K pages only. Page tables and
other generic parts operate with 16K pages.

Because the MMU doesn't support 16K directly, each 16K
page is emulated by a pair of 8K pages. With 16K pages,
illegal aliases cannot be created in 16K D-cache.
 
2134 6400 d 12 h jermar /trunk/kernel/ Reworked handling of illegal virtual aliases caused by frame reuse.
We moved the incomplete handling from backend's frame method to
backend's page_fault method. The page_fault method is the one that
can create an illegal alias if it writes the userspace frame using
kernel address with a different page color than the page to which is
this frame mapped in userspace. When we detect this, we do D-cache
shootdown on all processors (!!!).

If we add code that accesses userspace memory from kernel address
space, we will have to check for illegal virtual aliases at all such
places.

I tested this on a 4-way simulated E6500 and a real-world Ultra 5,
which has unfortunatelly only one processor.

This solves ticket #26.
 
2129 6435 d 11 h jermar /trunk/ Fix small problem in sparc64 port.

Bump version to 0.2.0.4 (Sunset).
 
2125 6443 d 11 h decky /trunk/ experimental support for Objective C
(disabled by default)
 
2107 6451 d 10 h jermar /trunk/ When clock() is called by an external interrupt dispatched by the IRQ dispatcher,
no spinlock can be held or the kernel will not be preemptive. This fixes Ticket #24.

Formating and indentation fixes.
 
2106 6451 d 18 h jermar /trunk/kernel/ Merge as_t structure into one and leave the differring parts in as_genarch_t.

Indentation and formatting changes in header files.
 
2105 6455 d 19 h decky /trunk/kernel/ move ipc structures to ipc.h  
2091 6457 d 14 h jermar /trunk/kernel/arch/sparc64/src/ Fix creating the kkbdpoll() kernel thread on sparc64 SMP machines.
This fixes Ticket #19.
 
2089 6457 d 18 h decky /trunk/ huge type system cleanup
remove cyclical type dependencies across multiple header files
many minor coding style fixes
 
2083 6463 d 18 h decky /trunk/kernel/ typedef elimination  
2082 6463 d 18 h decky /trunk/ __asm__ __volatile__ -> asm volatile  
2079 6464 d 13 h decky /trunk/kernel/ remove some typedefs in favor of anonymous structures  
2078 6464 d 15 h jermar /trunk/kernel/arch/sparc64/src/ When invalidating entire TLBs on sparc64, make sure to also invalidate any (locked) global entries.
This fixes Ticket #21.
Fix a comment in start.S stating that the kernel installs a global entry for itself. All entries installed by the kernel
are local to some memory context.
 
2076 6466 d 19 h jermar /trunk/kernel/ Beat the implicit illegal virtual alias created by reusing userspace frames.
In the anonymous and ELF backends, if the architecture has virtually indexed D-cache,
selectively flush cachelines belonging to the frame being freed.
This fixes Ticket #20.
 
2071 6468 d 11 h jermar /trunk/ (c) versus (C)  
2068 6475 d 18 h jermar /trunk/kernel/ Formatting and indentation fixes.  
2067 6478 d 17 h jermar /trunk/kernel/ More formatting changes.  
2065 6482 d 9 h jermar /trunk/kernel/ Coding style fixes and formatting improvements.  
2058 6494 d 20 h jermar /trunk/kernel/arch/sparc64/src/ Improve comments.  
2056 6495 d 23 h jermar /trunk/kernel/arch/sparc64/include/ Fix mismatched byteswapping macros on sparc64.  
2055 6495 d 23 h jermar /trunk/kernel/arch/sparc64/src/ sparc64 work:
- No need to lock kernel stack and userspace window buffer into DTLB.
 
2054 6496 d 9 h jermar /trunk/kernel/ Fix important comment in kernel/arch/sparc64/src/proc/scheduler.c.

Improve framebuffer code.

Formatting and indentation fixes.
 
2049 6499 d 19 h jermar /trunk/kernel/arch/sparc64/src/ More formatting and indentation changes.  
2048 6500 d 12 h jermar /trunk/ Formatting and indentation changes.  
2047 6502 d 10 h jermar /trunk/kernel/arch/sparc64/include/ Convert sparc64 critical section barriers to RMO memory model.
More portable, no confusion from documentation.
 
2046 6502 d 12 h jermar /trunk/kernel/arch/sparc64/include/ Another comment fix.  
2045 6502 d 12 h jermar /trunk/kernel/arch/sparc64/include/ Fix comment.  
2044 6502 d 13 h jermar /trunk/kernel/arch/sparc64/ On sparc64's tick interrupt, adjust the TICK_COMPARE register instead of the TICK
register. The TICK register now increments linearily and can be used by the get_cycle()
function.
 
2043 6502 d 13 h jermar /trunk/kernel/arch/sparc64/src/ Make the sparc64 kernel compile again.  
2038 6503 d 9 h jermar /trunk/kernel/arch/sparc64/ Remove old stuff.  
2037 6503 d 9 h jermar /trunk/kernel/arch/sparc64/include/ Backout change from revision 2034.
sparc64 is not yet ready for get_cycle().
 
2034 6503 d 17 h jermar /trunk/kernel/arch/sparc64/include/ get_cycle() for sparc64.  
2025 6505 d 8 h jermar /trunk/ Coding style fixes.  
2023 6505 d 12 h decky /trunk/kernel/arch/ add dummy cycle.h  
2016 6505 d 18 h jermar /trunk/kernel/arch/sparc64/src/mm/ The D-cache line size is actually 32 bytes on UltraSPARC II, IIi (and also IIe).  
2015 6507 d 11 h jermar /trunk/ Rework support for virtually indexed cache.
Instead of repeatedly flushing the data cache, which was a huge overkill, refuse to create an illegal address alias
in the kernel (again) and allocate appropriate page color in userspace instead. Extend the detection also to
SYS_PHYSMEM_MAP syscall.

Add support for tracking physical memory areas mappable by SYS_PHYSMEM_MAP.

Lots of coding style changes.