thread_block(), thread_unblock(),...) that allow a thread to wait for
something without needing a semaphore or condition variable. It can
simply block and another thread can unblock it. Supports timeouts and
interrupting. Both semaphores and condition variables use this
common mechanism, now.
* Semaphores:
- Some simplifications due to the thread blocking mechanism.
- Changed locking order to sem -> thread. It was the other way around
before and when introducing the wait_for_objects() support I had
also introduced a situation where the locking was reverse, which
could potentially cause a dead lock on SMP systems.
- Instead of queueing thread structures, a semaphore queues
queued_thread entries now, which are created on the stack. The
thread::sem structure could thus be removed.
- Added sem_entry::net_count, which is sem_entry::count plus the
acquisition count of all waiting threads. This number is needed in
remove_thread_from_sem() and instead of computing it there we
maintain it.
- Fixed remove_thread_from_sem(). It would not unblock threads, if
the sem count was <= 0.
- Made sem::last_acquirer unconditional. It is actually needed for
sem_info::latest_holder. Fixed fill_sem_info() accordingly.
- Added some optional tracing output, though only via ktrace_printf().
* Condition variables:
- Could be simplified significantly through the use of the thread
blocking mechanism. Removed a good deal of unnecessary code.
- Moved the ConditionVariableEntry "flags" parameter from Wait() to
Add(), and adjusted all places where condition variables are used
accordingly.
* snooze() uses thread_block_with_timeout() instead of a semaphore.
* Simplified thread interrupting in the signal and user debugger code.
Instead of separate functions for threads waiting on a semaphore or
condititon variable, we only have a single thread_interrupt(), now.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@25099 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Implemented temporary work-around for net_buffer append_cloned(),
which doesn't seem to work right in combination with remove_header().
Or maybe I'm just misunderstood how it is supposed to be used. Anyway,
this fixed invalid data in the stream when buffers were split by a
read.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@25070 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Also removed the header files that belong to those files.
* Only kept the userland_ipc.h header for now, to remember us about the
former userland server driver (that I also removed - it can always be
resurrected from SVN if needed).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@25045 a95241bf-73f2-0310-859d-f6bbb57e9c96
reference. The places where that is of relevance do that anyway, and
tty_ioctl(), where it isn't, failed before, although that was not
necessary. This prevented for instance ioctls() on the master tty
before any slave had been opened.
* If the tty has no process group set, don't check for background reads.
This was a problem with telnetd, respectively the executed login,
which couldn't access the tty, since telnetd makes sure neither itself
nor login has a controlling tty.
telnet still doesn't work, exactly because it has no controlling tty
and cannot set the tty process group. Not sure how that is supposed to
work. Furthermore the tty doesn't have the usual flags set, which is
apparrently the reason for the workaround (read_string()) in login.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@25038 a95241bf-73f2-0310-859d-f6bbb57e9c96
new filesystem interface.
* Renamed sISO* to gISO* though as they are in fact not static.
Briefly tested and works as expected. Thanks for your work Dustin!
Closes enhancement #2089.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24994 a95241bf-73f2-0310-859d-f6bbb57e9c96
was, thanks to the detailed instructions from Ingo, quite doable.
* Ported the (deprecated and now publically removed) notify_listener() calls
to the respective new notify_* calls.
* Fixed debug build.
* Added the FAT filesystem back to the image.
I only tested the filesystem briefly and everything seemed to work as expected
(mounted the FAT part of my bootable Haiku memory stick, opened a few files
and copied something to it). At least it should not be any more broken than
before the interface changes ;-).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24987 a95241bf-73f2-0310-859d-f6bbb57e9c96
1 second instead of half a second. Might help with slow BIOSes.
* Disable EHCI interrupts if the BIOS handover fails so that we do not flood
the system with interrupts if the BIOS later on decides to still comply with
our handover request.
* Add an unconditional success message to EHCI and UHCI to better see if
controllers are present/initialized correctly at all.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24980 a95241bf-73f2-0310-859d-f6bbb57e9c96
in. We do not enable this interrupt and if we still "handle" it in our
interrupt handler we might prevent the interrupt from reaching the real intended
handler in a shared interrupt setup. Not sure if this could have happend at
all, but this is more correct in any case.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24965 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Enabled testing the checksum of the EDID info.
* Added a version check of the EDID info.
* Added more debug output.
* In the ModeList class, changes were made to how the refresh rate is
computed and used. Previously, some of the basic VESA modes were not
added to the mode list because the computed and specified refresh rates
did not exactly match. Now if the computed refresh rate is within 1.2%
of the specified rate, the mode is added. With this change, all basic
VESA modes selected by the EDID info are now added to the mode list.
* The "additional video modes" shown in the EDID dump are now added to
the mode list. Previously, this mode data was setup but not added to
the mode list. Code was also added here to set the vertical &
horizontal sync polarity according the EDID info. The sync polarities
are set according to a VESA document that I have.
* Fixed copy_str() warning, broken removal of trailing spaces, and null
termination.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24959 a95241bf-73f2-0310-859d-f6bbb57e9c96
constructor the object can be safely deleted (as documented for bug #1473)
* Some simplifications by caching the Pipe object for a transfer
* Minor cleanup
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24954 a95241bf-73f2-0310-859d-f6bbb57e9c96
handing them over to the USB stack. Otherwise the finisher (kernel) thread
would have no way to fill those packets when they came from a user space
address range. Completely untested though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24947 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Handles wrapping of the chunk buffer for the case that we need to read
additional chunk data, but still need the data from the last chunk
buffer as well.
* Don't skip the very first frame when calculating the timing.
* Includes some new code to handle seeking internally, but it is currently
disabled and incomplete.
* Prepared to recover from floating point exceptions generated in some rare
situations by xvid, but we are currently missing fenv.h in our public POSIX
headers, so it is disabled.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24945 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Fixed shutdown(). It was computing the wrong fifo flags and set the
wrong ones from the wrong variable on the peer fifo.
* Generally made the Unix sockets behave more like they should. E.g.
after closing one end, it must still be possible to read from the
other (as long as there are buffered data). Also fine-tuned when to
return what errors from recv()/send().
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24943 a95241bf-73f2-0310-859d-f6bbb57e9c96
have two more optional hooks for attaching supplied ancillary data to a
net_buffer and for processing received ancillary data. Not sure, if that
is flexible enough for all kinds of ancillary data, but it is for
SCM_RIGHTS and also should for SCM_CRED[ENTIAL]S (if we ever decide to
implement one of those) -- don't know any other types on other protocol
levels.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24941 a95241bf-73f2-0310-859d-f6bbb57e9c96
implementation ATM, since it malloc()s the required memory.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24940 a95241bf-73f2-0310-859d-f6bbb57e9c96
read_next_module_name(). This is ugly and less flexible but it allows us to
explicitly force an initialization order. As detailed by the added comment
there seem to be controllers/BIOSes that require the EHCI controller to be
initialized after the companion host controller or else they refuse to publish
any high speed devices. This was the case for bug #1236 and my laptop when
legacy emulation is enabled and any device was plugged in at boot. This change
fixed the problem for me and hopefully closes bug #1236 too.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24926 a95241bf-73f2-0310-859d-f6bbb57e9c96
initialization as suggested by the EHCI spec and do not touch it again
once the controller is running. Doing so would result in undefined behaviour
according to EHCI specs.
* Set the port routing policy after the controller has started and not before
as suggested by the EHCI specs.
* Explicitly set the segment register after resetting the host controller and
not before, as this likely has no real effect ;-)
Might help with some legacy support or UHCI/OHCI to EHCI handover problems.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24925 a95241bf-73f2-0310-859d-f6bbb57e9c96
B_{READ|WRITE}_AREA. Otherwise we'd allow any user process to shredder our
data as Axel put it so nicely ;-) Thanks for the clarification.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24921 a95241bf-73f2-0310-859d-f6bbb57e9c96
when they are not open or they are deleted on free when they are already
removed. This should fix the sudden crashes when you unplugged a device that
might not have been fully closed yet.
Also handle the case of removal correctly and don't use the usb_device anymore
after releasing it by returning from the device removed hook. Calls to the
device just return B_DEV_NOT_READY in that case.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24920 a95241bf-73f2-0310-859d-f6bbb57e9c96
syscall_restart_ioctl_is_restarted() to syscall_restart_is_restarted,
IoctlSyscallFlagUnsetter to SyscallFlagUnsetter, and
IoctlSyscallRestartWrapper to SyscallRestartWrapper, as they are no
longer only used for ioctl().
* Removed unused syscall_restart_ioctl_handle_post().
* Made SyscallRestartWrapper a lot fancier. Instead of storing a
reference to the result value, it stores the value itself, and it
features all the interesting operators that make it appear like that
value. This simplifies the use of the class quite a bit.
* THREAD_FLAGS_SYSCALL is now set for all socket function and the
read[v](), write[v]() syscalls.
* Added is_syscall() function and net_stack hook to the net stack.
* Removed "kernel" parameter from all net_stack_interface and net_socket
module hooks. They aren't need any longer, since is_syscall() can be
used instead.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24914 a95241bf-73f2-0310-859d-f6bbb57e9c96
- The net_stack driver is no longer used. Instead we have a kernel
module which is directly used by syscall implementations in the
kernel. I.e. we no longer tunnel those functions through ioctls, but
have normal syscalls.
- Removed the superfluous net starter module.
- Implemented the FDTYPE_SOCKET type file_descriptors, that is sockets
are no longer vnode based.
- Adjusted libnetwork (the socket function implementations)
accordingly.
- Adjusted netstat accordingly.
* Socket module:
- Implemented socketpair().
- Added "kernel" parameter to the control hook. Quite a few more hooks
would actually need the parameter, but I didn't change those yet,
since that would probably also require changes to the protocol
module interface.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24908 a95241bf-73f2-0310-859d-f6bbb57e9c96
KDEBUG on by default which causes benaphores to always use their semaphore.
Therefore when initing the benaphore with the inline compatibility function
and then using it using the lock.h provided inline would always cause a
deadlock and a hanging system under BeOS. Also we now (re-)define the
B_KERNEL_{READ|WRITE}_AREA defines for non Haiku targets to 0 apparently which
would probably also cause the stack to misbehave. Therefore they are now
just redefined (again) to plain B_{READ|WRITE}_AREA in BeOSCompatibility.h.
Change the inclusion order in some places so things work as expected.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24894 a95241bf-73f2-0310-859d-f6bbb57e9c96
to ignore a trailing null byte, which the code now didn't do anymore.
* This caused bug #2054.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24888 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Flags and timeout arguments to acquire_sem_etc() were swapped.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24886 a95241bf-73f2-0310-859d-f6bbb57e9c96
After successfully copying the data from the receive queue into a new
buffer, we should actually return that buffer and update the queue size.
recv() doesn't KDL anymore when reading less data than queued.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24885 a95241bf-73f2-0310-859d-f6bbb57e9c96
is almost complete, but still quite buggy (receiving data has a good
chance to drop into KDL).
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24884 a95241bf-73f2-0310-859d-f6bbb57e9c96
it's a no-op, and I removed the prototype from KernelExport.h.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24875 a95241bf-73f2-0310-859d-f6bbb57e9c96
with B_USB_* and the command union is now usb_raw_command.
* Changed usb_raw and the BUSB* classes accordingly.
* Moved the raw_device struct into usb_raw.cpp as suggested by Francois as it
was the only thing usb_raw_private.h defined.
* Removed usb_raw_private.h again because of the above.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24861 a95241bf-73f2-0310-859d-f6bbb57e9c96
name, as it in fact is like a kit in the kit. It combines the different BUSB*
classes as compared to SerialPort.h where there really is a 1:1 relationship
between the header and the (one) class. If someone has a better fitting name
please shout.
Separated the usb_raw.h into usb_raw.h which defines the protocol and
usb_raw_private.h which holds the internal device structure for usb_raw. This
reduces the header dependencies.
Hopefully cought all occurencies of USBKit.a usage (in usb_dev_info and
the usb_webcam media add-on) but might have missed something that is not in
the image.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24858 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Made all debugging functions endian-safe.
* Added tracing for log entries.
* Added new KDL command "bfs_journal" to dump all pending log entries.
* When BFS_DEBUGGER_COMMANDS is defined, the LogEntry class will also track
the transaction ID it belonged to.
* The "bfs" KDL command now sets some useful debugger variables.
* The "bfs_allocator" KDL command now accepts the group index as 3rd argument.
* Renamed Journal::_TransactionListener() to _TransactionIdle(), as that's
all it is for.
* Removed TODO comment in Volume::WriteSuperBlock(), as it's actually not true.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24852 a95241bf-73f2-0310-859d-f6bbb57e9c96
usb_interface_descriptor of an alternate interface without having to switch
to it.
* Add some reserve bytes to all classes and add some reserved virtual slots
where the objects are publically constructable.
* Remove the source compatibilty defines that were briding the old USB* with
the new BUSB* class names.
* Implement the usb_raw side of getting an alternate interface descriptor.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24842 a95241bf-73f2-0310-859d-f6bbb57e9c96
clearity as suggested by Philippe Houdoin once and brought to my attention by
Slavatore. As this definition is completely private to the USB stack it would
not have caused any conflicts though.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@24828 a95241bf-73f2-0310-859d-f6bbb57e9c96
