Start the screensaver in the window thread instead of the runner
thread so that there is no lock contention for the window lock in
the runner thread when the saver starts.
The view that gets drawn into is assumed to have been prepared before
being passed to the runner thread, and this assumption has been made
true for the screensaver preview and screen_blanker apps.
Eliminate fHasStarted and the corresponding HasStarted() method in
ScreenSaverRunner as they are no longer needed.
Drawing still happens in the runner thread, and still needs to lock
the window thread potentially causing contention, yet, there
is a timeout here so the contention won't freeze the screensaver window,
only delay drawing the screensaver.
Drawing could be moved to the window thread via message passing to avoid
lock contention with the window but this would defeat a big part of the
purpose of having a separate rendering thread.
This fixes#10125 and #4260
* pin idle threads to their specific CPUs
* allow scheduler to implement SMP_MSG_RESCHEDULE handler
* scheduler_set_thread_priority() reworked
* at reschedule: enqueue old thread after dequeueing the new one
* Thread::scheduler_lock protects thread state, priority, etc.
* sThreadCreationLock protects thread creation and removal and list of
threads in team.
* Team::signal_lock and Team::time_lock protect list of threads in team
as well.
* Scheduler uses its own internal locking.
* The UNMAP command is theoretically much faster, as it can get many block
ranges instead of just a single range.
* Furthermore, the ATA TRIM command resembles it much better.
* Therefore, fs_trim_data now gets an array of ranges, and we use SCSI UNMAP
to trim.
* Updated BFS code to collect array ranges to fully support the new
fs_trim_data possibilities.
* No need for the atomically changed variables to be declared as
volatile.
* Drop support for atomically getting and setting unaligned data.
* Introduce atomic_get_and_set[64]() which works the same as
atomic_set[64]() used to. atomic_set[64]() does not return the
previous value anymore.
The new functions are meant to replace many uses of find_directory():
* find_paths() is supposed to be used when the directories of a certain
kind in all installation directories are needed (e.g. font
directories, add-on directory, etc.). Using this API makes code
robust wrt addition or removal of installation locations.
* find_path() is supposed to be used when files/directories associated
with a loaded program, library, or add-on need to be found (e.g. data
files or global settings).
* find_path_for_path() is similar to find_path(), but it starts from a
given path instead of an image.
The flag main purpose is to avoid race conditions between event handler
and cancel_timer(). However, cancel_timer() is safe even without
using gSchedulerLock.
If the event is scheduled to happen on other CPU than the CPU that
invokes cancel_timer() then cancel_timer() either disables the event
before its handler starts executing or waits until the event handler
is done.
If the event is scheduled on the same CPU that calls cancel_timer()
then, since cancel_timer() disables interrupts, the event is either
executed before cancel_timer() or when the timer interrupt handler
starts running the event is already disabled.
* Replace ports list mutex with R/W-lock.
* Move team port list protection to separate array of mutexes.
Relieve contention on sPortsLock by removing Team::port_list from its
protected items. With this, set_port_owner() only needs to acquire the
sPortsLock for reading.
* Add another hash table holding the ports by name. Used by find_port()
so it doesn't have to iterate over the list anymore.
* Use slab-based memory allocator for port messages. sPortQuotaLock was
acquired on every message send or receive and was thus another point
of contention. The lock is not necessary anymore.
* Lock for port hashes and Port::lock are no longer locked in a nested
fashion to reduce chances of blocking other threads.
* Make operations concurrency-safe by adding an atomically accessed
Port::state which provides linearization points to port creation and
deletion. Both operations are now divided into logical and physical
parts, the logical part just updating the state and the physical part
adding/remove it to/from the port hash and team port list.
* set_port_owner() is the only remaining function which still locks
Port::lock and one or two of sTeamListLock[] in a nested fashion.
Since it needs to move the port from one team list to another and
change Port::owner, there's no way around.
* Ports are now reference counted to make accesses to already-deleted
ports safe.
* Should fix#8007.
When removing a string attribute, decrement the referenced string's
usage count in the string cache. This fixes the potentially incorrect
usage counts in update mode. Not a serious problem, but it could lead
to only singly (or no longer) used strings to be written to the string
subsection instead of encoding them inline and thus to slightly greater
file sizes.
When joining with a single range, firstRange would be the same as
RangeAt(endIndex - 1) and we would overwrite its offset field before
getting its end offset, thus possibly resulting in a wrong joined range
size.
Simple scheduler behaves exactly the same as affine scheduler with a
single core. Obviously, affine scheduler is more complicated thus
introduces greater overhead but quite a lot of multicore logic has been
disabled on single core systems in the previous commit.
The method is supposed to return B_OK as long as the _result object has
been initialized, even if committing the transaction failed. Fixes the
unhelpful error messages of pkgman when committing the transaction
failed for some reason.
Update BKeymap::GetModifiedCharacters() to translate a given character
and set of modifiers filling out a list of all characters that match for another
set of modifiers.
This allows us to, for example, get all characters in the normal map that
have the '+' character in the corresponding shift map.
It is fully generic allowing one to get a list of characters in any map given
a character and modifiers of another map.
Also I converted from using a BList to using a BObjectList.
With this, along with BWindow::HasShortcut(), the semantic shortcuts now
work not only with Command+'=', but any key in the normal map that has
'+' in it's shift map as long as it isn't already taken by another shortcut.
This method fills out the passed-in BList of modified utf-8 characters for
a given utf-8 character and set of modifiers.
For example if you pass in "=" and B_SHIFT_KEY the list will get filled
out with each character in the shift map that has "=" in the normal map.
Each supported keymap modifier combination is available.
The reason this is useful will soon become apparent.
A BList is used because the character might be mapped multiple times,
for example if you have a Mac keyboard you've got two "=" keys, one in
0x1d and one in 0x6a.
The caller is responsible for creating the BList and destroying it as well as
freeing the resulting character strings.
This builds off of hrev46243 adding add-on directories all in one place
in AddOnMonitorHandler instead of repeating the code 3 times in
IndexServer, AddOnManager, and MediaAddOnServer.
The safe mode checking in InputServer is now redundant since it all
gets funneled into AddOnMonitorHandler::AddAddOnDirectories()
and the safe mode flags are checked there.
We should probably remove the InputServer::SafeMode() method, but,
I didn't want to break anything that depended on it so I left it.
There is a global heap of cores, where the key is the highest priority
of threads running on that core. Moreover, for each core there is
a heap of logical processors on this core where the key is the priority
of currently running thread.
The per-core heap is used for load balancing among logical processors
on that core. The global heap is used in initial decision where to put
the thread (note that the algorithm that makes this decision is not
complete yet).
Simple scheduler is used when we do not have to worry about cache affinity
(i.e. single core with or without SMT, multicore with all cache levels
shared).
When we replace gSchedulerLock with more fine grained locking affine
scheduler should also be chosen when logical CPU count is high (regardless
of cache).
In SMP systems simple scheduler will be used only when all logical
processors share all levels of cache and the number of CPUs is low.
In such systems we do not have to care about cache affinity and
the contention on the lock protecting shared run queue is low. Single
run queue makes load balancing very simple.
Kernel support for yielding to all (including lower priority) threads
has been removed. POSIX sched_yield() remains unchanged.
If a thread really needs to yield to everyone it can reduce its priority
to the lowest possible and then yield (it will then need to manually
return to its prvious priority upon continuing).
Each thread has its minimal priority that depends on the static priority.
However, it is still able to starve threads with even lower priority
(e.g. CPU bound threads with lower static priority). To prevent this
another penalty is introduced. When the minimal priority is reached
penalty (count mod minimal_priority) is added, where count is the number
of time slices since the thread reached its minimal priority. This prevents
starvation of lower priorirt threads (since all CPU bound threads may have
their priority temporaily reduced to 1) but preserves relation between
static priorities - when there are two CPU bound threads the one with
higher static priority would get more CPU time.
Should already have been done back when the semantics for the
B_COMMON_*DIRECTORY constants was changed.
Currently old and new version behave the same. So this is just a
contingency measure ATM.
* This does intentionally break source compatibility, so that a review
of concerned code is forced.
* Binary compatibility should be maintained in most cases. The values
of the constants for the writable directories are now used for the
writable system directories. The values for the non-writable
directories are mapped to "/boot/system/data/empty/...", an empty or
non-existent directory, so that they will simply be skipped in search
paths. Only code that explicitly expects to find something in a
B_COMMON_* directory, will fail.
* Remove support for the "common" installation location from packagefs,
package kit, package daemon, package managers.
* Rename the B_COMMON_*_DIRECTORY constants referring to writable
directories to B_SYSTEM_*_DIRECTORY.
* Remove/adjust the use of various B_COMMON_*_DIRECTORY constants.
I'm sure some occurrence still remain. They can be adjusted when the
remaining B_COMMON_*_DIRECTORY constants are removed.
* find_directory() and hard-coded paths use /boot/system instead of
/boot/common.
* The build system creates the writable directories in /boot/system
instead of /boot/common.
* The build system no longer installs any packages in /boot/common.
* The new class is called DriverSettingsMessageAdapter which can translate
between a driver_settings file, and a BMessage.
* The net_server Settings class is now just using this class.
- BJobStateListener: Add progress state and corresponding hook.
- FetchFileJob: Notify job progress hook on libcurl notifications.
- UserInteractionHandler: Add hooks for download progress and checksum
validation progress.
- PackageManager: inherit from JobStateListener and watch for job
notifications for internally generated jobs. Forward to corresponding
UserInteractionHandler hooks as needed.
- Adapt pkgman, HaikuDepot and package_daemon to above changes.
Neither HaikuDepot nor package_daemon's progress hooks are wired up to
do anything yet though.
- Pull functionality back into package manager itself since the extra
indirection doesn't really buy us anything in this case, as neither
request that it handles requires a decision provider.
- Adjust pkgman and HaikuDepot accordingly.
* imported asc-num.txt as a reference, was used to generate the asc sense table.
* use the sense asc and key tables to know which action and status codes are
to be applied.
* tested with an hard disk and a dvd reader.
* these tables could be reused by the scsi_periph module.
* Add NotifyDone() to all repository-attribute handlers and invoke that
to notify any listeners.
* Unify deletion to a single implementation of Delete() in the base
class. Before, the root handler for a repository didn't do that, but
just triggered the notification.
Support for 64-bit atomic operations for ARMv7+ is currently stubbed
out in libroot, but our current targets do not use it anyway.
We now select atomics-as-syscalls automatically based on the ARM
architecture we're building for. The intent is to do away with
most of the board specifics (at the very least on the kernel side)
and just specify the lowest ARMvX version you want to build for.
This will give flexibility in being able to distribute a single
image for a wide range of devices, and building a tuned system
for one specific core type.
This adds the -mapcs-frame compiler flag for ARM to have "stable"
stack frames, adds support to the kernel for dumping stack crawls,
and initial support for iframes. There' much more functionality
to unlock in KDL, but this makes debugging already a lot more
comfortable.....
Since both platforms can boot the same kernel we must accept either
arg, so we make sure they are identical for now.
TODO: use a union or KMessage maybe?
As korli suggested use B_PAGE_SIZE for defining stack size related
definitions what seems to be more natural for them and also may
help if we ever support an architecture with page size different than
4kB.
* Seems like there was no easy way to simply invalidate
a given BRow. Introduced BColumnListView::InvalidateRow().
* BRow::SetField() tried to invalidate the row, but invalidated
the listview instead of the BOutlineView responsible for
drawing the list contents. Use the new InvaalidateRow().
As korli suggested use B_PAGE_SIZE for defining stack size related
definitions what seems to be more natural for them and also may
help if we ever support an architecture with page size different than
4kB.
* Seems like there was no easy way to simply invalidate
a given BRow. Introduced BColumnListView::InvalidateRow().
* BRow::SetField() tried to invalidate the row, but invalidated
the listview instead of the BOutlineView responsible for
drawing the list contents. Use the new InvaalidateRow().
* It simplifies putting regular layout-aware views or layout items into
a BScrollView.
* Not quite complete yet: Height-for-width support is missing, but that
also requires fixing BScrollView in this respect. Scroll bar auto-hide
support would be nice as well.
* Rename TransactionHandler -> InstallationInterface and
DaemonClientTransactionHandler -> ClientInstallationInterface.
* Add InstallationInterface::InitInstalledRepository(). Use in
_AddInstalledRepository() to get the repository packages instead of
using the package roster. The ClientInstallationInterface
implementation does it that way.
If we fail to lock the window in the kInitialTickRate time, quit the thread.
We were deadlocking causing #4260 because you could open several
threads by moving through the screen saver list quickly all trying to lock
the same window at the same time, classic deadlock.
* Move RepositoryBuilder class to libpackage and add B* prefix to name.
* Pull BPackageManager class out of PackageManager and move to
libpackage. The base class is customizable via three handler objects
responsible for transaction handling, request execution, respectively
user interaction.
* Reorganize _ApplyPackageChanges(): Now we first prepare the
transactions for all affected installation locations (downloading
files etc.) and then commit them.
* The default module is replaced by the Virtio RNG module when found.
* This can have the undesired effect of rendering /dev/urandom slow.
* Tested with the following QEmu command line option:
-device virtio-rng-pci,rng=rng0 -object rng-random,filename=/dev/random,id=rng0
* moved random.h to private/drivers headers.
It provides the functionality to copy file system entries (also
recursively). The code originates from the copyattr sources. Some
copyattr specific functionality has been removed and the code has been
adjusted for library use (i.e. no exit()s or fprintf()s). An optional
controller object can be set to customize the behavior.
* Mostly useful for virtualization at the moment. Works in QEmu.
* Can be enabled by safemode settings/menu.
* Please note that x2APIC normally requires use of VT-d interrupt remapping feature
on real hardware, which we don't support yet.
Get rid of unused fRunner variable.
It is very basic now, it just works.
Also, renamed msg to message in MessageReceived() and
declared MakeFocus() above it (alphabetically).
the child menu bar or the child menu bar's menu is enabled/disabled.
This means that there is just one status we have to check, the menu
fields, and the child menus agree. This change takes practical form
in the Backgrounds preflet which disables the placement menu when
the image is set to "None", but, only the menu got disabled and not the
parent menu field so the label was erroneously still drawn as enabled.
* the processing of requests in drivers is eased a bit with this change, but
this could be improved for instance by enabling a driver to dequeue items
in a service thread instead of the interrupt handler.
* made a few methods const.
* make use of MSI/MSI-X PCI x86 API
* MSI support untested because QEmu only offers MSI-X
* changed a bit the Virtio bus API by adding a queue count parameter
for the setup_interrupt() hook.
* They are all over the place.. I give up
* Going off of engineering names and DCE is more accurate
* A lot of this info came from the x.org wiki
* I'd like to transition some of the engineering
name checks to use DCE versions.. they tend to be more
accurate and exact. (in some cases we can't, but most of
the time we can)
* Introduce new package attribute B_HPKG_ATTRIBUTE_ID_PACKAGE (valid
only in a repository file) to group the attributes belonging to a
package.
* BRepositoryContentHandler:
- No longer derive from BPackageContentHandler.
- Add hooks HandlePackage() and HandlePackageDone() that bracket the
attributes for a package. This is more explicit and robust than
handlers having to guess when one package ended and the next began.
* BRepositoryCache: Make use of BPackageInfoContentHandler. No need to
duplicate the code for reading a package info from package info
attributes.
This resolves all issues the test suite uncovered. It should also deal
with hard links correctly, though that hasn't been tested. Still
unsupported are:
* changes due to mounting/unmounting a volume,
* tracking of symlinks in the path components.
Add inner class BWatchingInterface and method SetWatchingInterface().
This abstracts the calls to watch_node() and stop_watching(), thus
making it possible to use the path monitor in Tracker.
Fixes#9816
It is no longer necessary, or even desirable for us to set the max
content width of the menu bar of a BMenuField now that BMenuItem
truncation and BMenuField sizing are working.
The user may, however, wish to set the max content width of the menu
bar of a BMenuField themselves like so:
menuField->MenuBar()->SetMaxContentWidth(width);
and the Interface Kit will automatically deduct the left and right
margins from the width including the space used by the drop down arrow.
John Scipione