to the specs, the application may still use the FD without changing its state,
and only closedir() should finally close it.
* This fixes bug #5055.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34362 a95241bf-73f2-0310-859d-f6bbb57e9c96
as this fixes some more incorrect minimizations on the app_server side.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34361 a95241bf-73f2-0310-859d-f6bbb57e9c96
object is lost outside its scope.
Now vector data is copied instead.
Spotted by Pete Goodeve, thanks!
This should close#4562.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34358 a95241bf-73f2-0310-859d-f6bbb57e9c96
Display size is still kept to 32x32.
M patchbay/Jamfile
M patchbay/EndpointInfo.cpp
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34357 a95241bf-73f2-0310-859d-f6bbb57e9c96
* untested unsolicited response support
* added quirk support for vref and gpio
* vref are now enabled for all inputs, and gpio for some Apple Macs
* replaced dprintf with TRACE and ERROR macros
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34352 a95241bf-73f2-0310-859d-f6bbb57e9c96
the B_NOT_MINIMIZABLE flag. It should more properly fix#4337.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34351 a95241bf-73f2-0310-859d-f6bbb57e9c96
Bosii is a transcription of 802.11. It is a work-in-progress title to
point out parts of the wlan stack which are still in an experimental state.
For example the control codes within bosii_driver.h are moving targets.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34349 a95241bf-73f2-0310-859d-f6bbb57e9c96
The source is based on the FreeBSD RELEASE_8_0_0 code, found in Haiku's
freebsd vendor branch.
This driver enables the network card in my EeePC 1005HA-M, for example.
To compile it issue "jam atheros813x".
* Introducing the new build target.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34348 a95241bf-73f2-0310-859d-f6bbb57e9c96
The driver sources are based upon the FreeBSD RELEASE_8_0_0 source as found in
Haiku's freebsd vendor branch.
* Currently only the atheros driver is working and can be compiled with
jam atheros.
* Every driver contains a Jamfile already, so that the compilation process
can be started with jam <driver_name>. Also note, that linking of every
driver besides atheros and iprowifi2200 will fail at the moment.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34345 a95241bf-73f2-0310-859d-f6bbb57e9c96
This port is based on FreeBSD RELEASE_8_0_0 (r199625) found in Haiku's
freebsd vendor branch.
* Added freebsd_wlan directory to the Jamfile in HAIKU_TOP/src/libs/compat/
so that the stack can be built with jam libfreebsd_wlan.a.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34343 a95241bf-73f2-0310-859d-f6bbb57e9c96
recursive lock. I haven't investigated it closer, but the previous
implementation was even broken -- "strace /bin/true" showed two
release_sem() calls, but no acquire_sem().
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34342 a95241bf-73f2-0310-859d-f6bbb57e9c96
as a mutex, but allocates its semaphore lazily. This comes at the cost of an
additional atomic_add() when the semaphore has actually to be acquired, but
saves the semaphore creation completely in single-threaded programs and in
any program when there's no lock contention.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34339 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Fixed check in hoardSbrk(): resize_area() was invoked, even if the area was
already large enough.
* Increased the initial heap size to 64 pages. Apparently the hoard
implementation is rather generous and the first malloc() (caused by
__init_heap()) already required enlarging the area.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34336 a95241bf-73f2-0310-859d-f6bbb57e9c96
- Add more debug info for the KDL debug command
- Request channel pointer also for non l2cap signal frames
- Add locks for packet queues
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34333 a95241bf-73f2-0310-859d-f6bbb57e9c96
- Start to move some common code from the driver layer (not yet used)
- Style
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34332 a95241bf-73f2-0310-859d-f6bbb57e9c96
- Don't allow scrolling when a pad button is pressed because we are likely in a drag action.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34326 a95241bf-73f2-0310-859d-f6bbb57e9c96
preprocessed <syscalls.h> header in sources compiled for the build platform.
Instead we're generating macros for the return type and parameter type sizes
via the CreateAsmStructOffsetsHeader rule (no template magic anymore) and
thus get a clear separation of host and target code, resulting in better
portability (although surprisingly enough the build already worked on 64 bit
hosts).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34323 a95241bf-73f2-0310-859d-f6bbb57e9c96
architectures, there's now a arch/generic/generic_gensyscalls.h which is
included by the otherwise empty arch_gensyscalls.h files.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34320 a95241bf-73f2-0310-859d-f6bbb57e9c96
Moved the AboutRequested code to a static TermView method, called from
everywhere (now TermApp knows about TermView, but no one is perfect).
Reordered and removed includes.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34318 a95241bf-73f2-0310-859d-f6bbb57e9c96
notifications if an inode in a query result was moved/renamed - this time all
the information is correct, though.
* While I did not introduce B_ENTRY_MOVED for queries yet, this should make
adding it very simple (left as an excercise for the reader ;-))
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34317 a95241bf-73f2-0310-859d-f6bbb57e9c96
a file suitable for the CreateAsmStructOffsetsHeader rule. The file is then
used to generate the actual matypes.h header. This fixes issues when
cross-compiling from a 64 bit platform.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34312 a95241bf-73f2-0310-859d-f6bbb57e9c96
use for the asm_offsets.cpp file, so it can be reused elsewhere.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@34311 a95241bf-73f2-0310-859d-f6bbb57e9c96
Axel Dörfler