of the image to be skipped. Magnify therefore showed a white frame as long as
the area around the cursor would not change. Fixes bug #598 (again). I wonder
why this didn't show up earlier though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24156 a95241bf-73f2-0310-859d-f6bbb57e9c96
vector are unhandled, it is disabled if it is not a shared interrupt vector
(as before). Otherwise a message is printed to indicate the condition. I
reduced the amount of interrupts to 10000, as with 100000 it would take over
half an hour on my machine to trigger. Feel free to adjust.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24146 a95241bf-73f2-0310-859d-f6bbb57e9c96
unhandled. Otherwise it brings more trouble than it solves as it also disables
the interrupt of other devices (most often USB or network).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24120 a95241bf-73f2-0310-859d-f6bbb57e9c96
entire screen when exiting the kernel debugger. It sets up a thread that sends
a message to the (currently hardcoded) desktop message looper. The desktop then
does mark the whole screen dirty which causes a full redraw.
Since interrupts need to be enabled I went with an asynchronous thread and
releasing a request sem in the add-ons' exit hook.
Added the add-on to the image as it shouldn't hurt to have it for now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24025 a95241bf-73f2-0310-859d-f6bbb57e9c96
size of a reallocated block. If you had kernel heap leak checking on, this
could have caused the first four bytes of the next block to be overwritten
with the size of the reallocation of the previous block.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24011 a95241bf-73f2-0310-859d-f6bbb57e9c96
This eliminates the edge case where the grow thread would not be able to create
a new area because no memory could be allocated for the allocation of the area.
As this case cannot happen anymore, it is also not possible to deadlock in
memalign. Therefore the timeout (which would only have prevented the deadlock
but wouldn't have solved the edge case anyway) has been removed too.
Add options to dump the dedicated grow heap and to only print the current heap
count to the "heap" debugger command.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23994 a95241bf-73f2-0310-859d-f6bbb57e9c96
total count of allocations and bytes.
* Also add a few more bin sizes (for 8, 24 and 48 bytes) turns out especially
allocations of 20-24 bytes are pretty common. And as it only wastes a few
bytes per page this doesn't hurt at all.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23961 a95241bf-73f2-0310-859d-f6bbb57e9c96
heap leak check info would otherwise be overwritten for allocations that still
fit the 16 byte bin (i.e. allocations of 0-4 bytes).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23956 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Tracing of allocations, reallocations and frees
* Leak checking infrastructure to dump allocations
The leak checking code records the team and thread id when an allocation is
made as well as stores the originally requested size. It also adds the
"allocations" debugger command that can dump all current allocations (usually
a huge list) or filter by either a team or thread id. This way it's easily
possible to find leftover allocations of no more active teams/threads.
Combined with the tracing support one might be able to track down the time and
reason of an allocation and possibly find the corresponding leak if it is one.
Note that kernel heap leak checking has to be enabled manually by setting the
KERNEL_HEAP_LEAK_CHECK define to 1.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23953 a95241bf-73f2-0310-859d-f6bbb57e9c96
and pages are now kept in lists as well. This allows to return free pages once
a bin does not need them anymore. Partially filled pages are kept in a sorted
linked list so that allocation will always happen on the fullest page - this
favours having full pages and makes it more likely lightly used pages will get
completely empty so they can be returned. Generally this now goes more in the
direction of a slab allocator.
The allocation logic has been extracted, so a heap is now simply attachable to
a region of memory. This allows for multiple heaps and for dynamic growing. In
case the allocator runs out of free pages, an asynchronous growing thread is
notified to create a new area and attach a new heap to it.
By default the kernel heap is now set to 16MB and grows by 8MB each time all
heaps run full.
This should solve quite a few issues, like certain bins just claiming all pages
so that even if there is free space nothing can be allocated. Also it obviously
does aways with filling the heap page by page until it overgrows.
I think this is now a well performing and scalable allocator we can live with
for quite some time. It is well tested under emulation and real hardware and
performs as expected. If problems come up there is an extensive sanity checker
that can be enabled by PARANOID_VALIDATION that covers most aspects of the
allocator. For normal operation this is not necessary though and is therefore
disabled by default.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23939 a95241bf-73f2-0310-859d-f6bbb57e9c96
not use a single static variable to synchronize CPUs at two points. In an
environment where CPUs do not really run concurently (in emulation or with
logical processors) it would be possible for CPUs to get trapped in the first
synchronization while another CPU might just do its thing and change the
sync variable again. These CPUs would then never leave the first loop as the
exit condition has already passed again. The key is to use two different sync
variables like it is done in early kernel initialization. As I didn't manage
to trigger this code though I am not sure if this is gonna work.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23926 a95241bf-73f2-0310-859d-f6bbb57e9c96
This is not safe when already freed memory is overwritten. But since we also
store the next pointer of the freelist in there, overwriting would break the
freelist and cause a crash in that case. This gives a drastic performance
boost when freelists grow during use and especially when opening and closing
a lot of programs.
* Optimize filling the freed element with 0xdeadbeef by writing 4 bytes at a
time instead of using single byte writes. Works as all our bins have an
element size that is a multiple of four. Put a panic in there just in case
this assumption isn't met for some reason.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23879 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Stumbled upon a possible bug while trying to understand the reuse of large
allocations. The "first" variable was always set to the current index at the
end of the loop, even if it was already set. This should have caused that
the success condition to never be reached.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23866 a95241bf-73f2-0310-859d-f6bbb57e9c96
When specified it desigantes that the interrupt handler should not lock the
vector with a spinlock when executing the installed interrupt handlers. This
is necessary to allow the same interrupt vector to be handled in parallel on
different CPUs. And it is required for the CPU halt to work synchronously when
there is more than one AP CPU. Though the acquire_spinlock() should cause IPIs
to be processed, only this fixed the SMP_MSG_FLAG_SYNC problem for me.
Not locking is safe as long as it is guaranteed that no interrupt handler is
registered or removed while the interrupt handler is running. We can guarantee
this for the SMP interrupt handlers we install in arch_smp_init() as they are
never uninstalled. Probably this flag should be made private though.
Restored the SMP_MSG_FLAG_SYNC when entering the kernel debugger.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23838 a95241bf-73f2-0310-859d-f6bbb57e9c96
confirmation after sending off the startup IPIs. We simply moved on and setup
the temporary stack address for the next CPU to be started. With this it was
possible that the trampoline code did not manage to load the address before
we overwrote it. So for configurations with more than two CPUs it was possible
that two CPUs were setup to the same kernel stack which could have caused all
sorts of things - most likely a tripple fault and a reboot. On real hardware
this seems very unlikely but it was easily reproducible with QEMU and -smp >2.
We now use the shared trampoline stack to implement a notification mechanism.
The trampoline code will clear the stack location variable once it has loaded
everything it needs from the trampoline stack. On the other side
smp_boot_other_cpus() will wait for this variable to be cleared after it sent
the startup IPIs so that it knows when it can safely move on and overwrite the
area to boot the next CPU.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23800 a95241bf-73f2-0310-859d-f6bbb57e9c96
was a race condition when more than one CPU would enter the debugger at the
same time (or rather before one CPU could stop all the others). We now use the
inDebugger variable to tell if someone is already in the debugger and then
only process inter CPU messages and retry entering the debugger.
Since sending the synchronous broadcast most of the time hung over here with
SMP enabled I removed the synchronous flag and added a simple spin to give the
other CPUs a chance to process the halt request. Added comments that explain
the reasons and a ToDo to revert to synchronous delivery once we fixed the
problem. The kernel debugger is now usable on my quad core.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23751 a95241bf-73f2-0310-859d-f6bbb57e9c96
systems it can easily happen that the thread gets removed from the queue (when
it times out for example) during the time we don't hold the sem lock.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23749 a95241bf-73f2-0310-859d-f6bbb57e9c96
at the end of that loop we guaranteed a crash when this special handling was
triggered.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23748 a95241bf-73f2-0310-859d-f6bbb57e9c96
OS-owned bit. Otherwise we overwrite exactly the bit we later have to wait for.
Could theoretically help on legacy support issues, but I doubt it since most
hardware probably does not even use EHCI legacy support as it's normally not
needed to support full/low-speed input devices.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23747 a95241bf-73f2-0310-859d-f6bbb57e9c96
useful soon so that you can actually do something after entering KDL.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23742 a95241bf-73f2-0310-859d-f6bbb57e9c96
This should be in line with all uses of scheduler_enqueue_in_run_queue() and
simplifies a few places where it is used.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23738 a95241bf-73f2-0310-859d-f6bbb57e9c96
controller ownership in EHCI to aid debugging USB legacy support issues.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23735 a95241bf-73f2-0310-859d-f6bbb57e9c96
enqueued into the run_queue again. Modified the workaround to a panic in the
scheduler so we notice when something else does the same.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23731 a95241bf-73f2-0310-859d-f6bbb57e9c96
A thread that just missed a semaphore and put itself into the sems notify
queue could be enqueued into the run queue by release_sem_etc() of another
CPU before the CPU running the thread had a chance to reschedule it. Therefore
there is a timeframe where a thread can be running on one CPU and already be
in the run queue again. In this case no other CPU may schedule this thread
because then it would overwrite the threads' CPU pointer which kills the rest
of the scheduler logic, smp_get_current_cpu() and everything that depends on
that (like the kernel debugger). The more CPUs you have the easier this could
happen, up to the point where it was always triggered during boot on my quad
core system. The system would freeze and you could not enter the kernel
debugger, because two CPUs thought they were the same and disabled each other
through SMP communication. This makes booting my system stable and might fix
the occasional hang on boot for other SMP systems with only 2 CPUs/cores.
I've put a ToDo comment that details this above the workaround. Maybe we
should fix this in another way. Reviews, comments and suggestions welcome ;-)
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23721 a95241bf-73f2-0310-859d-f6bbb57e9c96
Notify the finisher thread after canceling transfers. It was possible that freeing the allocated
transfer resources would be delayed unnecessarily until some other transfer completed and woke up
the finisher thread. Found by Salvatore Benedetto, thanks!
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23714 a95241bf-73f2-0310-859d-f6bbb57e9c96
transfer resources would be delayed unnecessarily until some other transfer completed and woke up
the finisher thread. Found by Salvatore Benedetto, thanks!
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23713 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Refactored everything to C++ with the different devices as subclasses
* Added proper ACM detection with parsing of the ACM descriptors
* Added device transfer error handling and fixed some concurency issues
* Big cleanup to conform to our style guide
This should make at least ACM stable to use. Commiting this over my K850i with
ACM compliant USB modem and UMTS data connection. Note that support for all
other device classes (Prolific, FTDI and KLSI) is untested but should work the
same as before. Note also that since we currently lack a TTY module this will
only build/work for R5 or Dano with the proper TTY headers.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23634 a95241bf-73f2-0310-859d-f6bbb57e9c96
any devices anymore it gets unloaded. Also made sure the function pointers
are properly initialized and disabled debug output.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23494 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Fixed wrong start and size used in the runtime_loader
* Fixed off by one error in the matching loop of the cleanup hook
* Make sure we successfully acquire the locking sem of the exit stack
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23493 a95241bf-73f2-0310-859d-f6bbb57e9c96
As (our) gcc unfortunately uses atexit() to clean up lazily initialized static
variables inside functions we have to ensure that we do the right thing with
unloadable shared objects. In case a shared object was unloaded that installed
an atexit() hook the application would crash on exit. We now implement a
callback into libroot that is used to call all the atexit() hooks of a
component that is to be unloaded. Most prominently this fixes the media_server
crash at shutdown.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23486 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Removed now unnecessary devfs calls again
* Add an explanation in device_removed()
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23442 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Cache the looked up image symbols as well as the API version in the driver entry
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23441 a95241bf-73f2-0310-859d-f6bbb57e9c96
mix up the resizingMode and flags field in the BView constructor. This does
not cause problems under BeOS but was causing wrong follow modes when those
applications were run under Haiku. In case such an application is detetected
a corresponding warning is printed out. Also added a comment explaining the
rational behind the change and a ToDo for the later cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23424 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Move most of MIME database support out of libbe and into registrar
* Use the (async) MessageDeliverer instead of a synchronous SendMessage in _SendMonitorUpdate
This fixes a deadlock when the message port of a MIME database watching
application gets full as documented in bug #1311.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23423 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Beginnings of devfs interaction to properly publish/unpublish under Haiku
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23412 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Some restructuring and cleanup
This keeps the input_server from polling a removed device continuously.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23388 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Actually enable the periodic schedule so that any interrupts work over EHCI
For those who wait for OHCI: You can now attach your USB 1.1 mouse or keyboard
to a USB 2.0 hub and it should work.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23382 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Minor style cleanup in the conversion factor calculation
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23217 a95241bf-73f2-0310-859d-f6bbb57e9c96
* When disabling SMP initialization of those parts is avoided so it works more cleanly
* Moved smp_find_mp_config() into smp_init() and moved MPS specific parts into smp_do_mp_config()
* Removed the global floating struct, if found it is now passed into smp_do_mp_config() like it is done for ACPI
* Made the file a bit more C++, applied coding style to some variable names and unified/improved some of the debug output
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23216 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Insert the pause op in all spin wait loops (as macro for platform independent sources or as inline assembly on x86 only files)
* Fix some warnings with tracing on and extended some output
* Minor cleanups here and there
The pause instruction is implemented since P4 systems but is fully backwards compatible (it's a no-op prior to P4). According to Intel specs it reduces performance penalties as memory order violations can be avoided. Also power consumption is reduced. Most of all this will be beneficial to hyper-threading systems as it frees resources to the other logical processor when one logical processor executes a fast spinning loop.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23141 a95241bf-73f2-0310-859d-f6bbb57e9c96
* ACPI is evaluated first as it also handles things like multi core or hyper threading setups
* Removed other (disabled) hyper threading code per the notes in the corresponding ToDo
* Limit the detected CPU count to 2 for now as I wasn't able to get it working in either emulation nor real hardware with more than 2 CPUs
* Added a reserved byte to the mp_config_table struct, it worked only by luck as the compiler did padding there to get to the same size
I can now boot my Core 2 Quad with two out of four processors active :-)
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23100 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Move waiting for the first explore to complete
* Small cleanup
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@23099 a95241bf-73f2-0310-859d-f6bbb57e9c96
With this it is possible to use low/fullspeed (USB 1.1) devices below USB 2.0 Hubs.
Note though that you cannot yet use mice or keyboards attached to USB 2.0 Hubs as split transactions for interrupt pipes are still missing.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22973 a95241bf-73f2-0310-859d-f6bbb57e9c96
* All other init is done in InitCommon() which only needs to be present in the base class
* Adding the hub port a device is attached to to the Device class
* Add hub port and hub address to the Pipe class (will be used for split transactions in EHCI)
* Update the root hubs to reflect the changes to hub creation
Sadly we need pipes that do not belong to devices (default pipe for addressing and to get initial device descriptor). Therefore we cannot resolve hub address and port from the parent device.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22972 a95241bf-73f2-0310-859d-f6bbb57e9c96
* The sparation based on async and periodic schedule made no sense as it is a question of queued/non-queued
* Therefore removed the near complete code duplication for adding interrupt transfers (that are also queued)
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22943 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Uses a "collapsed binary tree" (for lack of a better name) to support the different intervals
* Remove a leftover variable declaration that was hiding error conditions away...
* Some cleanup
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22942 a95241bf-73f2-0310-859d-f6bbb57e9c96
Changes are now collected during explore and notifications as well as rescans are done at once. Through this a driver is also not rescanned multiple times anymore.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22929 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Always check the return of the resize function
* Handle resize errors gracefully and ensure that the message stays intact
* Don't crash when printing a message that contains a field with no items (unlikely but possible in low memory situations)
* Fixed renaming fields - was completely broken and would have missed up field names and contents
* Some cleanup
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@22047 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Check for failed allocations and set errno correspondingly in malloc(), calloc(), memalign() and realloc()
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21911 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Remove the leftover user_area - this is handeled inside the Transfer class by now
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21906 a95241bf-73f2-0310-859d-f6bbb57e9c96
This also ensures that the rescans do not happen while the devices are not yet added or already removed.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21905 a95241bf-73f2-0310-859d-f6bbb57e9c96
Someone please review. Should errno be set to ENOMEM here?
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21903 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added a todo about unsafe access of the buffer
* Removed some types from is_type_swapped() to exactly mirror R5
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21902 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Also swap the flattened size when creating the read buffer
* Define specialized byte_swap()s for unsigned types too so that type_code and the like get swapped correctly
This should fix bug #1371.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21901 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Transfer removal is now done only from the finisher
* Processing is only done after the transfer has been removed from the list
* Canceling transfers only sets a flag so the finisher can remove the transfer
* The callback at cancel is done synchronously though
This fixes quite a few race conditions where transfers were processed while they where already removed. It could happen before that callbacks would have been called twice or freed descriptors/callback buffers were accessed.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@21882 a95241bf-73f2-0310-859d-f6bbb57e9c96
Michael Lotz