written the page, we now do it before, so that it cannot lose any changed data
anymore; it doesn't matter if the page is written to while writing it back, the
worst thing that can happen is that we write the same page twice. Also, we don't
rely on the PAGE_MODIFIED bit anymore, we now check all mappings of that page
to find all modified pages, no matter how far the (currently disabled) page
daemon had come.
Also, destroying an area will now result in writing back changed pages - this
is only really important for memory mapped files, though, and should probably
be avoided for other vm_store types.
Minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15597 a95241bf-73f2-0310-859d-f6bbb57e9c96
Moved some headers around and cleaned them up.
Fixed threading code in libbind.so. Don't know why I hadn't noticed this before. Now it is reliable, but uses an even uglier hack.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15585 a95241bf-73f2-0310-859d-f6bbb57e9c96
mappings yet, that's a good way to test this functionality.
* Turns out the directory of the mappings must be known already - they
should be added automatically, but I've added them manually for now
(which is okay for the default system directory).
* Having more than one style with the same family in the mappings didn't
work as planned - it now does.
* On my current system, time spend in the font manager constructor went
down from 1.5 secs to around 12000 usecs (and I have only a moderate
number of fonts installed, or so I thought - looks like the mappings
were a good idea :-)) - and that directly cuts down the boot time.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15580 a95241bf-73f2-0310-859d-f6bbb57e9c96
service depends on it, it doesn't make any sense to call it that early in the
game.
* The VFS now has a low memory handler for vnodes as well. If there is enough
memory left, it won't free any vnodes anymore.
* Potential crashing bug fix: some functions did not check if the FD passed
in belonged to the right type; they just assumed it had a valid vnode, but
it could have had a mount structure associated as well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15566 a95241bf-73f2-0310-859d-f6bbb57e9c96
memory status.
Added new B_NO_LOW_MEMORY constant for the usual case.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15551 a95241bf-73f2-0310-859d-f6bbb57e9c96
kernel TLBs from being flushed on context switch.
* new arch_cpu_user_TLB_invalidate() that now does what arch_cpu_global_TLB_invalidate()
did before.
* arch_cpu_global_TLB_invalidate() will now flush all TLBs, even those from the
kernel.
* some cleanups.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15535 a95241bf-73f2-0310-859d-f6bbb57e9c96
* The boot loader now checks the CPU for the cpuid and rdtsc features, which we
currently both rely on.
* Removed old and no longer used stage2_priv.h header
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15534 a95241bf-73f2-0310-859d-f6bbb57e9c96
has not yet been tested, though - I'll do this after this commit):
* Removed the arch_memory_type stuff from vm_area; since there are only 8 memory
ranges on x86, it's simply overkill. The MTRR code now remembers the area ID
and finds the MTRR that way (it could also iterate over the existing MTRRs).
* Introduced some post_modules() init functions.
* If the other x86 CPUs out there don't differ a lot, MTRR functionality might
be put back into the kernel.
* x86_write_msr() was broken, it wrote the 64 bit number with the 32 bit words
switched - it took me some time (and lots of #GPs) to figure that one out.
* Removed the macro read_ebp() and introduced a function x86_read_ebp()
(it's not really a time critical call).
* Followed the Intel docs on how to change MTRRs (symmetrically on all CPUs
with caches turned off).
* Asking for memory types will automatically change the requested length to
a power of two - note that BeOS seems to behave in the same, although that's
not really very clean.
* fixed MTRRs are ignored for now - we should make sure at least, though,
that they are identical on all CPUs (or turn them off, even though I'd
prefer the BIOS stuff to be uncacheable, which we don't enforce yet, though).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15528 a95241bf-73f2-0310-859d-f6bbb57e9c96
* finally nailed that much hated bug where the
wrong background underneath the cursor was restored.
I always looked for this in the wrong place -
of course, you cannot blit regions on top
of the cursor and expect the backup of the
background to be magically valid still... ;-)
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15461 a95241bf-73f2-0310-859d-f6bbb57e9c96
is known by the server, anyway.
* B_MODAL_SUBSET_WINDOW_FEEL now also works as expected.
* Renamed AS_REM_FROM_SUBSET to AS_REMOVE_FROM_SUBSET.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15285 a95241bf-73f2-0310-859d-f6bbb57e9c96
* renamed some Decorator functions like GetLook() to Look().
* renamed _look to fLook and _flags to fFlags.
* removed the feel from decorators - they don't need to know
or care about the feel.
* the DefaultDecorator didn't allow resizing of modal windows.
* DefaultDecorator::SetTitle() no longer clears the passed
in updateRegion, it just fills it (so that you can call
more than one of these methods after the other, and still
get a correct update region).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15266 a95241bf-73f2-0310-859d-f6bbb57e9c96
know what's going on, but this appears to fix it. And since we'll get a new glyph
layout backend sooner or later, I don't want to mess with that stuff again.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15176 a95241bf-73f2-0310-859d-f6bbb57e9c96
an app_server code. It's always up-to-date as it just reads the header
directly.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15148 a95241bf-73f2-0310-859d-f6bbb57e9c96
* RootLayer's mouse event processing is now at its minimum - the
EventDispatcher handles them now. As a result, a window will now
get only one message per event.
* RootLayer adds "_view_token" to mouse moved messages that specify
the view currently under the cursor.
* There is now a mouse event layer in RootLayer that gets preferred
when it's set - this is now used for the window moving instead of
the previous mechanism.
* changed the previous DistributeMessage() to an UnpackMessage()
method following Adi's suggestion.
* caveat: some things might be functionally broken in RootLayer now
because of removing the mouse notification stuff.
* "be:transit" handling is now done completely client side by
BWindow::_SanitizeMessage(() (similar to what the input_server does).
This should also make the mechanism pretty robust, since every
B_MOUSE_MOVED message can now trigger the view transit (in case a
message is lost). B_WINDOW_ACTIVATED messages should be generated
client side as well.
* renamed AS_LAYER_GET_MOUSE_COORDS to AS_GET_MOUSE as it's not a
layer specific command, and also gets the mouse buttons.
* B_MOUSE_* messages from the up server now contain only a "screen_where"
field; "where" (in window's coordinates) and "be:view_where" are
added in BMessage::_SanitizeMessage().
* messages that don't have a valid target in the looper are now
dropped instead of being sent to the looper - this should be done
in BLooper as well, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15087 a95241bf-73f2-0310-859d-f6bbb57e9c96
task_looper() again.
* removed BMessenger::fPreferred - whenever you had to specify "usePreferred" separately,
you don't have to do that anymore - use B_PREFERRED_TOKEN instead.
* fixed BTokenSpace::GetToken() semantics: it will no longer touch the "object" argument
in case of failure.
* Introduced a BWindow::_DistributeMessage() that will be part of the event dispatcher
counterpart to the app_server (the other will be _DetermineTarget()).
* Made it easier to use Michael's Message4 implementation: just add the following line
to your UserBuildConfig:
AppendToConfigVar DEFINES : HAIKU_TOP src : USING_MESSAGE4 : global ;
* Introduced ServerWindow::HandlerMessenger() and FocusMessenger() - the first will
target the client handler, while the other will target the preferred handler of the
client looper (usually the view having focus).
* Fixed dano message unflattening in the Message4 code.
* Changed BMessage::PrintToStream() to no longer use macros in the Message4 implementation.
* I hope that's all - it's a huge change, but it's all connected.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@15046 a95241bf-73f2-0310-859d-f6bbb57e9c96
