need to be interpolated because the x index is right on a source pixel. This
prevents the out of bounds access for the second to last row in the last
column. Also the rightmost pixels where incorrectly interpolated with the
leftmost pixels of the next row. And it actually helps speed too of course.
* Added a compile time option to allocate the filter weighting and index
caches on the heap instead of the cache. I am not sure if it is a problem
though, I recall Haiku threads have quite a lot of stack space. The needed
memory depends on the target size. For a screen with 1920x1200, the caches
would need 12.5 KB. Allocating them on the stack saves about 0.2 ms on my
test system.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26656 a95241bf-73f2-0310-859d-f6bbb57e9c96
to provide asynchrounous (or only synchronous, if asynchronous is not
supported) I/O request support. It will eventually replace
{read,write}_pages(). None of the FS implementations implement them
yet.
* Implemented some support functions for request-based I/O. File system
implementations can use do_fd_io() which passes an I/O request to the
layer responsible for a given FD, and do_iterative_fd_io(), which
translates a request for a file to subrequests for the underlying
device and passes them on. Both fall back to synchrounous processing
when the io() hook is not supported.
Furthermore added vfs_synchronous_io() which should be handy for the
devfs to perform io_requests synchronously for devices that don't
support the io() hook.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26655 a95241bf-73f2-0310-859d-f6bbb57e9c96
added an IORequest::Init() version with that parameter. This makes
splitting an iovec array into IOBuffers/IORequests easier.
* Added IORequest::CreateSubRequest(). It creates and adds an IORequest
that covers a part of the range of the parent request, but may use
another file offset. This will be used e.g. in the way that the parent
request describes an I/O operation for a file while its subrequests
describe the same operation translated to the underlying device.
* Added IORequest::DeleteSubRequests(), which does the obvious. It's
also invoked in the destructor.
* Added method for iterating through subrequests.
* Made IORequestChunk::{Set,Reset}Status() protected. For both
subclasses some locking is needed (though different locking), so we
rather make this more explicit.
* Added IORequest::SetStatusAndNotify(), which is SetStatus() +
NotifyFinished() with proper locking.
* Changed the I/O request finished and iteration callback signatures.
The finished callback has got an additional "status" argument, since
the request itself may already be inaccessible at the time the
callback is executed.
* Changed IORequest::NotifyFinished(). The policy is now that if the
iteration callback fails, the method will do the finished
notifications. This simplifies things in the iteration callbacks.
* Fixed bug in IORequest::_CopyPhysical(): It didn't take into account
that the physical buffer could not be page aligned.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26654 a95241bf-73f2-0310-859d-f6bbb57e9c96
B_OP_COPY which is about 2.4 times faster than the AGG version (but of
course less generic). The speed up is even better for smaller and even
scales.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26652 a95241bf-73f2-0310-859d-f6bbb57e9c96
an invalid page state was supplied. The comment was misleading too,
as the pages weren't reserved upfront.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26650 a95241bf-73f2-0310-859d-f6bbb57e9c96
* When drawing BBitmaps with scaling in the app_server, use a bilinear
filter when a bitmap has this flag set. (Hope nobody objects, otherwise
I can revert or improve this. Performance can certainly be improved, since
the AGG implementation is too generic. But that goes for the nearest
neighbor implementation as well.)
* Flags are uint32, fix app_server side code to declare them correctly. Use
appropriate link methods in BBitmap and ServerApp.
* Enable the BeOS compatibility mode for B_RGB32 (works just like B_RGBA32
in B_OP_ALPHA mode).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26649 a95241bf-73f2-0310-859d-f6bbb57e9c96
* adjust all drivers to take that into account
* fix UpdateText() signature in JSDSlider to avoid warning
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26648 a95241bf-73f2-0310-859d-f6bbb57e9c96
children of the device and recreate them so that they are updated to any
changes. MainWindow::_Initialize() was using a stale BPartition pointer
after calling CommitModifications(). Now the pointer is recreated from the
id. Should fix#2548. Untested.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26643 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Renamed sModifiedTemporaryPages to sModifiedNoSwapPages to better
express what this variable is about.
* Changed tracking of sModifiedNoSwapPages. It really counts
non-swappable pages only, now (if swap support is enabled).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26641 a95241bf-73f2-0310-859d-f6bbb57e9c96
- Enabling all net_buffers code & ACL
- Asume contiguusness for outgoing packets
- Fix incorrect cleaning of btDevices
- Posting ACL data to the net_device Kernel module.
Sumarizing this enables ACL tranmision mode in the driver which starts to be "real network data" of the bt protocols not configuration as we had before
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26638 a95241bf-73f2-0310-859d-f6bbb57e9c96
buffer lives in userland can easily be checked via IS_USER_ADDRESS.
* Added B_VIP_IO_REQUEST flag which will be used by the page writer and
should cause allocations to be made in a way that they cannot fail.
Not implemented yet, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26632 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Split IORequest::ChunkFinished() into OperationFinished() and
and SubrequestFinished(). Moved the notification part into a new
method NotifyFinished().
* Added new IOScheduler thread for notifying finished requests.
IOScheduler::_Finisher() hands over finished request to it, unless the
requests don't have callbacks. We need the separate thread, since the
callbacks can potentially reenter the scheduler and thus cause
deadlocks.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26631 a95241bf-73f2-0310-859d-f6bbb57e9c96
versions of the AddressSpaceReadLocker constructor and SetTo(). When
true it will get a new reference to the address space.
* Changed vm_soft_fault(): Now it gets the address space as a
parameter (for page faults getting it happens in vm_page_fault()).
This should fix lock_memory_etc() for other teams -- it used the
correct address space before, but the invoked vm_soft_fault() always
used the current team's address space.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26628 a95241bf-73f2-0310-859d-f6bbb57e9c96
- add 040 cpu and mmu stuff
- use leftover from the page root table to put interrupt vector table to set VBR to
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26627 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Init swap support in main().
* Added "bool swappable" parameter to
VMCacheFactory::CreateAnonymousCache(). A cache supporting swapping
is created when true. Adjusted invocations accordingly.
* The page writer does now write non-locked swappable pages (when
memory is low).
* Fixed header guard of VMAnonymousNoSwapCache.h.
* Swap support is compiled conditionally, controlled by the
ENABLE_SWAP_SUPPORT in src/system/kernel/vm/VMAnonymousCache.h. It is
disabled ATM. Since no swap files are added, it wouldn't have much
effect anyway.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26625 a95241bf-73f2-0310-859d-f6bbb57e9c96
swappable, since that happens before we can initialize swap support
(which requires the later initialized VFS).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26623 a95241bf-73f2-0310-859d-f6bbb57e9c96
the new team. Otherwise it would lose it's properties (full lock,
B_KERNEL_AREA).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26622 a95241bf-73f2-0310-859d-f6bbb57e9c96
- Fix PincodeWindow to send the pincode commands dissapear after clicking
- Improve the debug output of bluetooth_server
- Handle all needed events for the pairing
- Simple request could send and receive the event before adding the request to the events wanted list. Inverted the order of this sequence.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26614 a95241bf-73f2-0310-859d-f6bbb57e9c96
happens in the I/O scheduler thread, we need to use unlock_memory_etc().
Changed the IOBuffer::{Lock,Unlock}Memory() methods accordingly.
The test driver seems to be working stable, now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26609 a95241bf-73f2-0310-859d-f6bbb57e9c96
completed operations queue. That doesn't need to happen synchronously
either. So we have to restart the loop.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26607 a95241bf-73f2-0310-859d-f6bbb57e9c96
helps to avoid stupid mistakes. The driver works again, though things
are still not stable.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26605 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Factored an _UnmountAndEjectVolume() method that takes a partition and mount
path out of the method with the same name that gets a BMessage.
* Remove the mount point only if it's in rootfs.
* Minor cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26604 a95241bf-73f2-0310-859d-f6bbb57e9c96
Something is still fishy though: Opening the dma_resources_test device
in DiskProbe, the wrong data is shown. do_io() seems to be invoked with
the correct physical address, vm_get_physical_page() succeeds, and after
memcpy() the correct data are in the virtual address it returned, but
db in the kernel debugger shows that the data in the physical page have
not been changed. When quitting DiskProbe the "page still has mappings"
assert is triggered for the page.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26602 a95241bf-73f2-0310-859d-f6bbb57e9c96
the scheduler thread doing that has no direct access to the
destination team's address space.
* Improved some debug output.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26601 a95241bf-73f2-0310-859d-f6bbb57e9c96
get_memory_map(), but has a saner semantics and allows specifying a
team.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26600 a95241bf-73f2-0310-859d-f6bbb57e9c96
- rewrote early_query() to use the TT0 mapping to remove the page_hole stuff.
- fixed natfeat, using a page set up from the bootloader for now as it wants physical address. At least it's enough to see from the debugger:
load kernel...
kernel entry at 8003711a
Welcome to kernel debugger output!
Haiku revision: 26582
PANIC: unknown cpu_type 68040
Welcome to Kernel Debugging Land...
Running on CPU 0
kdebug>
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26599 a95241bf-73f2-0310-859d-f6bbb57e9c96
driver name didn't work here.
* Enabled using the I/O scheduler in the read() and write() hooks.
Something's still broken, though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26598 a95241bf-73f2-0310-859d-f6bbb57e9c96
using a lock and I'm not very much in favor of a global one.
* Added "finished" callback to IORequest.
* IOOperation::Finish() no longer invokes its parent request's
ChunkFinished(). The finisher does this instead. ChunkFinished()
can optionally remove the chunk from the parent.
* Added IORequest::Wait() which waits for the completion of the request.
* Introduced IORequestChunk::ResetStatus() to make setting the status to
"pending" somewhat more explicit.
* Implemented the missing IOScheduler::SetCallback() methods.
* The NotifyAll() calls on the IOScheduler's condition variables were
missing, so it just waited forever.
* Added some more debug output.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26597 a95241bf-73f2-0310-859d-f6bbb57e9c96
thread. Interrupting a thread only works when it is already waiting.
We do now use a flag to indicate whether the scheduler thread is
waiting (avoids thread_interrupt() calls when the thread is in driver
code). Furthermore before starting to wait, we check whether any
finisher work has to be done -- we do that (and the addition of the
entry to the condition variable) with the finisher lock being held to
avoid the race condition.
* Moved waiting for and getting the next unscheduled request into new
method _GetNextUnscheduledRequest().
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26594 a95241bf-73f2-0310-859d-f6bbb57e9c96
* as on darwin platform we cannot use resources for the boot sector data. Instead use attributes to store the data.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26589 a95241bf-73f2-0310-859d-f6bbb57e9c96
* additional includes for cygwin
* some automatic whitespace cleanup
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26588 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Cygwin does not have regex support, so we use the one we already have in our glibc.
* Keymap.h has a automatic forward declaration of re_registers, which causes a compile break on Cygwin, but not on other platforms. So add a include for regex.h in the header and remove it in the source file.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26587 a95241bf-73f2-0310-859d-f6bbb57e9c96
* rc uses a global instance of BResources. On destruction it calls Unset() of the BFile member fFile. BNode::Unset then calls close_fd(), which parses the static global DescriptorMap to actually close the descriptor. However on Cygwin the DescriptorMap has been destroyed already for some reason, thus accessing it is invalid. The "fix" is to put the map on the heap and delete it as soon as it gets empty. This way the global BResources instance can be freed without a crash.
* If someone likes to review, feel free to, it didn't cause any issues here
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26585 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Cygwin needs some additional defines compared to other platforms
* Additionally stpcpy and strcasestr are unknown on Cygwin. Thus we need to use the one from our posix library.
* ECANCELED is not defined on Cygwin, so only add error in case it is.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26584 a95241bf-73f2-0310-859d-f6bbb57e9c96
* As this was the last user of the READ_COUNT/WRITE_COUNT definitions in
vm_priv.h, I removed those as well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26582 a95241bf-73f2-0310-859d-f6bbb57e9c96
It does its job by using part of libsupc++, with fake malloc and friends to make sure it won't double fault.
Works here, but cp-demangle must be extracted by hand from the lib, can't get the rule to work as I want, Ingo ?
Maybe using it directly without malloc hack would be ok with 16KB buffer, but I'm not sure of that.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26581 a95241bf-73f2-0310-859d-f6bbb57e9c96
fs_vnode_ops::write_pages() to be called with fsReenter = true. Since
this is no longer the case, the argument has become superfluous. For
read_pages() it always was. Removed the argument from the functions
and all functions that propagated it.
* Some whitespace at the end of lines was removed.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26579 a95241bf-73f2-0310-859d-f6bbb57e9c96
will take care of the locking.
* Use new(std::nothrow) over new, and malloc (only in kernel_interface.cpp);
the kernel_cpp.h header isn't really necessary anymore, so there is no reason
BFS should continue to use it.
* Removed superfluous NULL checks.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26574 a95241bf-73f2-0310-859d-f6bbb57e9c96
introduces the following relevant changes:
* VMCache:
- Renamed vm_cache to VMCache, merged it with vm_store and made it a
C++ class with virtual methods (replacing the store operations).
Turned the different store implementations into subclasses.
- Introduced MergeStore() callback, changed semantics of Commit().
- Changed locking and referencing semantics. A reference can only be
acquired/released with the cache locked. An unreferenced cache is
deleted and a mergeable cache merged when it is unlocked. This
removes the "busy" state of a cache and simplifies the page fault
code.
* Added VMAnonymousCache, which will implement swap support (work by
Zhao Shuai). It is not integrated and used yet, though.
* Enabled the mutex/recursive lock holder asserts.
* Fixed DoublyLinkedList::Swap().
* Generalized the low memory handler to a low resource handler. And made
semaphores and reserved memory handled resources. Made
vm_try_resource_memory() optionally wait (with timeout), and used that
feature to reserve memory for areas.
...
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26572 a95241bf-73f2-0310-859d-f6bbb57e9c96
* move libprint headers into libs headers folder accordingly
* merge all shared folders sources into kits print, we might build later on a
real print kit, propably also to access cups from an nicely API, atm static
* move all shared headers into private print, also pr_server.h from interface
* adjust build to work with the changed folder layout
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26570 a95241bf-73f2-0310-859d-f6bbb57e9c96
really work yet (there are some bits missing in the IORequest and IOScheduler
to able to use them this way).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26565 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Fixed crash when it couldn't find the device.
* Added a "-l|--list-types" option that list all file systems that support
initializing (currently only BFS).
* Added a "-q|--dont-ask" option that prevents mkfs from questioning your
command :-)
* Fixed return value - would return failure to the shell when it actually
succeeded and vice versa.
* Improved/clarified some output and fixed spelling errors.
* Errors now have the command name as prefix, as that's how it is usually
done (and very useful for batch processing).
* Cleanup, though I kept the FsCreator class which is kinda superfluous.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26564 a95241bf-73f2-0310-859d-f6bbb57e9c96
did not set the "devicesOnly" flag to false when calling RegisterDevice().
* ddm_userland_interface.cpp incorrectly wrote to userland memory when it
assigned "neededSize" in several places.
* Replaced on-stack path with the UserStringParameter class where appropriate.
* Made the UserStringParameter class castable to char*.
* Minor cleanup in KDiskDeviceManager.cpp.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26563 a95241bf-73f2-0310-859d-f6bbb57e9c96
sub menu and quit menu tracking. This closes#1826. I tested a bit with
various different menu situations and there seem to be no negative side
effects.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26561 a95241bf-73f2-0310-859d-f6bbb57e9c96
This makes the kernel loadable by the bootloader, which does load it, and all the modules \o/
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26558 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Moved data buffer creation to a TestSuiteContext class.
* Added checks if the I/O operation does the correct thing, ie. reads/writes
the data to the right offset.
* Rearranged DMA translation: we now handle the partial write case correctly
(bounce buffer must always span over the whole block), and are able to join
adjacent bounce buffers together.
* The new _AddBounceBuffer() method also respects boundary and segment size
restrictions for bounce buffers.
* IOOperation now prepares the outgoing vecs/offset/length to contain the
right data for the current phase (partial read begin/end/do-all); it will
also make sure that the lengths of the vecs are of the same size than the
whole request.
* All tests are now passed, the I/O request implementation seems to be ready
for integration now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26556 a95241bf-73f2-0310-859d-f6bbb57e9c96
Hide USB errors better from the upper layers: Only try the "clear feature"
command if the error is B_DEV_STALLED. On other errors, the clear feature
was likely to fail, as it should only be used on stalled devices. Errors
intended to be ignored were thus not ignored because of the "clear feature"
failure.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26555 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Don't spam the syslog when the device was not a Wacom.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26553 a95241bf-73f2-0310-859d-f6bbb57e9c96
behind highspeed hubs with multiple transaction translators not to work.
* The parent hub isn't necessarily the one providing the transaction translator
for a full- or lowspeed device. Therefore we need to inherit the hub/port
addresses to lower devices unless we are a highspeed hub that might be the
actual provider of the transaction translator. This logic error caused full-
and lowspeed devices not to work if they were behind a fullspeed hub that
itself was behind a highspeed hub.
This fixes bugs #1889 and #2501.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26552 a95241bf-73f2-0310-859d-f6bbb57e9c96
automounter already mounted these partitions. Since this happens asynchronously,
it sometimes worked and sometimes not. The very simply and non-hacky fix for this
problem is to send a message from the automounter to the application looper to
have it open the previous windows after the initial mount scan is done.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26549 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Fixed mounting previously mounted partitions. fSettings was never initialized when restoring
the settings. The code I removed earlier didn't do that either.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26548 a95241bf-73f2-0310-859d-f6bbb57e9c96
from the settings message just after having restored it. This should fix
restoring the previously mounted partitions, but I have not tested it yet.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26547 a95241bf-73f2-0310-859d-f6bbb57e9c96
It works by adjusting the display_mode to have twice the virtual height to
cause the graphics driver to allocate enough frame buffer. Since hardware
acceleration calls are based on geometry locations instead of memory locations,
accelerated calls can work in the offscreen buffer with this method.
Similarily, the new CopyBackToFront() method copies the back buffer into the
front buffer with hardware acceleration. The code is currently disabled, since
not all graphics drivers handle the virtual versus display size correctly,
for example the intel_extreme driver. It works for example with the radeon
driver, but another problem prevents me to judge the benefit of this method.
Most types of screen redraws are flicker free, though, which is the whole point
of the excercise. :-)
Another problem with the current code is the initial mode switch. It does
not try again to double the frame buffer in the fall back code paths. It could
also check the frame buffer memory at all, before even trying to save some
cycles.
To see it in action, uncomment the code at line 508.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26545 a95241bf-73f2-0310-859d-f6bbb57e9c96
hardware accelerated functions could be double buffered.
* Align the rectangle used for arc drawing like those for ellipsis.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26544 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added a flag "fIsOffscreenBuffer", which is used to shift the frame buffer
pointer to the position after the visible frame buffer.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26543 a95241bf-73f2-0310-859d-f6bbb57e9c96
the front buffer so that derived classes could override it.
* Minor coding style changes.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@26542 a95241bf-73f2-0310-859d-f6bbb57e9c96
