- When guessing the initial thread state for an unschedule event also check
the previous event, so we can decide whether the thread is still ready.
Previously the time to the first schedule event could be accounted
incorrectly.
- Made the main loop a bit more robust with respect to unexpected thread
states.
* The check boxes for latency and preemption time were labeled the wrong way
around.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30680 a95241bf-73f2-0310-859d-f6bbb57e9c96
For some reason, it was resizing the window to the size of the dragger frame :-) (the frame variable was reused to a different purpose)
Fixes ticket #3896
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30679 a95241bf-73f2-0310-859d-f6bbb57e9c96
-> Significant speed up, since there is much less app_server communication.
-> Much improved looks, since the line is now drawn as one connected line.
The StrokeLine(BPoint to) version cannot do this.
Hope I am not interfering, Ingo!
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30676 a95241bf-73f2-0310-859d-f6bbb57e9c96
publish themselves in busses/ide, so you cannot just move them to busses/ata
without changing that in the source as well. The object files are still built
in the separate busses/ata under generated.
* src/system/kernel/device_manager/device_manager.cpp also hardcodes
busses/ide in two places.
I tried changing all of this to add busses/ata, but my system remains
unbootable with the new ATA stack. I do have another system, and because
of the previous mixup in HaikuImage, there it installed the drivers in
busses/ide when building for the new ATA stack (by mistake), but those
drivers actually publish themselves in busses/ide, so this system actually
boots with the new ATA stack, because of the mixup. Therefore My change here
to install into busses/ide for either stack should be correct for now and should
actually fix building the *old* stack.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30671 a95241bf-73f2-0310-859d-f6bbb57e9c96
value -- make sure to ignore feedback when setting the range (happens when the
old value doesn't fit the range anymore). Fixes unexpected jumps when zooming
out.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30666 a95241bf-73f2-0310-859d-f6bbb57e9c96
so should be reusable for bluethooth HID as well (which is the same). The only
missing part so far is the logical collections that would allow nicer
enumeration of the report structure but is otherwise not useful. It should
support all of the HID specs except for usage aliases (even long items that
aren't actually defined should just work if they ever are). Not integrated into
the USB specific device framework and there are no actual drivers making use
of provided functionallity. The parsing was tested and works for all of the 3
devices I had available, but actual interpretation of data is not tested as the
driver side is missing. Will close that gap as a next step and then port the
mouse and keyboard drivers to that framework. Eventually a generic driver that
makes unknown fields available to userland apps in some way should be fairly
easy to implement with that.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30664 a95241bf-73f2-0310-859d-f6bbb57e9c96
from the same "associated file" problem as the latter. This now finally fixes
bug #3861. This badly needs some cleanup.
* Fixed a possible problem I introduced in ISOReadDir() (did not read the next
block even if it should have).
* Fixed warnings with debug output turned on.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30663 a95241bf-73f2-0310-859d-f6bbb57e9c96
* The scrolling related computations used a slightly off chart width/height.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30658 a95241bf-73f2-0310-859d-f6bbb57e9c96
disappear.
* When parsing rock ridge attributes, we no longer stop when we encounter an
unknown one. Instead, we just parse through until the end. The ISO image as
part of #3861 also made this visible.
* Minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30655 a95241bf-73f2-0310-859d-f6bbb57e9c96
* fixed a couple of memory leaks in the BeOS-specific (attribute related)
code of zip
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30653 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Set the displayed domain to 0.5 s for debugging purposes.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30649 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added support for scrolling.
* Automatic white space cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30648 a95241bf-73f2-0310-859d-f6bbb57e9c96
explaining the reasons behind it. Oh well, at least the commit count raises...
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30644 a95241bf-73f2-0310-859d-f6bbb57e9c96
* A CHS read error would previously be hidden by a successful disk system reset.
Not that it matters much because it'd have resulted in a non-booting system
anyway.
* Add some more debug output, minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30643 a95241bf-73f2-0310-859d-f6bbb57e9c96
children. Otherwise a driver that builds up a device hierarchy could run into
trouble when the parent hub is removed before its child devices. Not that there
were any drivers that do so, but it just seems more correct.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30641 a95241bf-73f2-0310-859d-f6bbb57e9c96
* ISOReadDirEnt() did happily read after the 2048 byte block. Now we check after
having processed an entry if the position is on the 2048 block boundary, and
skip to the next block directly.
* fs_walk() assumed a 2 block set (and accessed memory therein), but only has
access to a single block. I haven't looked at the specs, so I'm not really
sure what the old code tried to achieve. In any case, it doesn't crash
anymore.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30640 a95241bf-73f2-0310-859d-f6bbb57e9c96
(missing spaces after commas) that were pointed out by Stephan
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30636 a95241bf-73f2-0310-859d-f6bbb57e9c96
rootfs_find_in_dir() was wrong, leading to never be able to find the fromName
in the directory. Furthermore, the parent of the root directory is itself, but
the check to see whether or not the target is valid did not take this into
account, and therefore ran into an endless loop. This fixes bug #3864.
* Rearranged rootfs_rename() to be clearer.
* Style cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30633 a95241bf-73f2-0310-859d-f6bbb57e9c96
it anymore after having called vfs_asynchronous_read_pages().
* Now, Prepare() does all the preparation work, and ReadAsync() does the actual
work - this must be called without having the cache locked. This also fixes
another bug where the callback would be deleted twice in case the I/O request
failed.
* This fixes bug #3847.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@30632 a95241bf-73f2-0310-859d-f6bbb57e9c96
