Subversion Repositories HelenOS

Rev

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

Filtering Options

Rev Age Author Path Log message Diff
3596 5653 d 17 h svoboda /branches/tracing/kernel/generic/src/mm/ Update obsolete comment.  
3431 5717 d 16 h svoboda /branches/tracing/kernel/ Add configuration option CONFIG_UDEBUG to compile udebug in the kernel or not.  
3425 5718 d 3 h svoboda /branches/tracing/ Merge latest trunk changes into tracing (broken)  
3424 5718 d 15 h svoboda /branches/tracing/ Merge trunk changes up to r3180 into tracing  
3127 5815 d 15 h svoboda /branches/tracing/kernel/generic/src/ Fix rounding error in as_debug_write() - prevent panic when writing on page boundary  
3036 5824 d 20 h svoboda /branches/tracing/kernel/generic/ Make sure area written by debugger is private and anonymous by copying it, if necessary. (Not atomic atm)  
3013 5829 d 4 h svoboda /branches/tracing/kernel/generic/ Bypass area write-protection for udebug memory writes.  
3011 5830 d 15 h svoboda /branches/tracing/ Merged latest trunk changes to tracing  
2787 5890 d 21 h decky /branches/tracing/ add dynamic linking, debugging and tracing branch  
2745 5920 d 1 h decky /trunk/ code cleanup (mostly signed/unsigned)
allow extra compiler warnings
 
2725 5940 d 22 h decky /trunk/kernel/ remove config.memory_size, get_memory_size() and memory_init.{c|d}
the amount of available memory can be calculated from the sizes of the zones
add FRAMES2SIZE, SIZE2KB and SIZE2MB functions/macros (code readability)
 
2712 5945 d 23 h decky /trunk/kernel/ prettyprint output  
2647 5992 d 18 h jermar /trunk/kernel/generic/src/mm/ Fix a bug in anonymous address space area sharing. Call the backend's share
function only if the source area is not already shared. Otherwise increment the
sh_info refcount. Without this fix, the kernel will try to recreate the pagemap
if three and more address space areas attempt to share the same data.
 
2556 6071 d 2 h jermar /trunk/kernel/generic/ Rename as_get_size() to as_area_get_size() and add a doxygen comment.  
2467 6178 d 19 h jermar /trunk/ Improve comments for arch-specific implementations of hierarchical
4-level page tables. Improve formatting.
 
2222 6244 d 16 h decky /trunk/kernel/ map kernel pages explicitly as writable (this solves compatibility issues with Intel Core 2)
make VESA framebuffer initialization more robust
 
2183 6247 d 21 h jermar /trunk/kernel/generic/ Continue to de-oversynchronize the kernel.
- replace as->refcount with an atomic counter; accesses to this
reference counter are not to be done when the as->lock mutex is held;
this gets us rid of mutex_lock_active();

Remove the possibility of a deadlock between TLB shootdown and asidlock.
- get rid of mutex_lock_active() on as->lock
- when locking the asidlock spinlock, always do it conditionally and with
preemption disabled; in the unsuccessful case, enable interrupts and try again
- there should be no deadlock between TLB shootdown and the as->lock mutexes
- PLEASE REVIEW !!!

Add DEADLOCK_PROBE's to places where we have spinlock_trylock() loops.
 
2170 6252 d 15 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.
 
2141 6256 d 13 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 6257 d 17 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.
 

Show All