info.
* Implemented arch_set_cpu_state() (Axel, please check, what registers
the debugger should actually be allowed to modify).
* Fixed arch_get_cpu_state(). We now explicitly copy the registers
manually instead of using memcpy(). Added support for extended
registers (FPU,...).
* Completed break- and watchpoint support.
* Implemented handling of debug and breakpoint exceptions.
The x86 specific part of the userland debugging support should be complete
now. That is in theory the entire debugging interface is available on x86
machines now. It's almost completely untested yet, though.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11526 a95241bf-73f2-0310-859d-f6bbb57e9c96
needed to get the topmost user iframe. Added i386_get_user_iframe() for
exactly that purpose.
* Added reinitialization of user debugging for the new thread when
switching the context.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11523 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Made the parameter for restoring the FPU state const.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11522 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added event callbacks for break- and watchpoints and single stepping.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11520 a95241bf-73f2-0310-859d-f6bbb57e9c96
functions for setting break-/watchpoints (implemented).
Break-/watchpoints don't work yet, though. Missing is still the setting
of the debug control register DR7 somewhere in the interrupt code and
the handling of debug exceptions.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11509 a95241bf-73f2-0310-859d-f6bbb57e9c96
* In team_delete_team() we no longer explicitly destroy the team debug
info. It is not necessary, since the terminating debug nub thread will
do that anyway. Moreover we access the debugger port in a safer manner
now.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11504 a95241bf-73f2-0310-859d-f6bbb57e9c96
by the team lock, but by a lock in the structure. This is necessary
since we can't acquire the team lock while thread switching, for then the
thread lock is already held.
* Added support for single stepping and break-/watchpoints.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11503 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Now a debugged thread and the debug nub thread are not longer blocking
on ports when they receive a SIGKILL[THR] (we couldn't kill a debugged
team before).
* Small fixes. More correct cleanup when the debug nub thread terminates.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11491 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added is_kill_signal_pending(). Can anyone tell me why the compiler
complains about a missing previous prototype for it? <ksignal.h> is
included, and when only preprocessing the file, the prototype is clearly
there. I'm clueless.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11488 a95241bf-73f2-0310-859d-f6bbb57e9c96
- If the thread's B_THREAD_DEBUG_STOP flag is set, it is now always
stopped before handling the signals. Unless a SIGKILL[THR] signal is
pending.
- Also notify the debugger, when a signal is ignored.
* The signal's sigaction structure is additionally passed to the debugger
now.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11478 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Now we deal correctly with thread debug flags (child threads inherit
the parent ones, save for syscall tracing in which case it needs to
be specified explicitly).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11477 a95241bf-73f2-0310-859d-f6bbb57e9c96
In the first case this is probably too early, since the image is not
really loaded at that time by the loader. We should probably introduce
a syscall the loader can invoke for that purpose.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11475 a95241bf-73f2-0310-859d-f6bbb57e9c96
convenience.
* Added/implemented callbacks for almost all missing debug events (team,
thread, image creation/deletion, exceptions/faults, signals).
* The debugger can now specify how to deal with the event that stopped an
event (ignore or handle signals, exceptions/faults).
* Implemented B_DEBUGGED_THREAD_GET_WHY_STOPPED debugger message.
* The cpu_state is now passed to the debugger with
B_DEBUGGER_MESSAGE_THREAD_STOPPED notifications.
* Completed _user_debugger() implementation.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11474 a95241bf-73f2-0310-859d-f6bbb57e9c96
keyboard controller method, and only if that fails we fall back to what
we have before.
At least my SiS laptop now properly reboots, and doesn't just halt
anymore - doesn't seem to have a negative impact on other machines, but
please report any regressions :-)
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11386 a95241bf-73f2-0310-859d-f6bbb57e9c96
console module which is used by the console driver, and will be used by
the on-screen KDL.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11350 a95241bf-73f2-0310-859d-f6bbb57e9c96
* The syscall arguments are copied into the respective messages to the
debugger now.
* Optimized the nub thread memory reading command. Now only as many bytes
are sent to the debugger as were actually read.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11343 a95241bf-73f2-0310-859d-f6bbb57e9c96
and thread_at_kernel_exit() are always called for userland threads at the appropriate
situation (note, I've renamed those from *_atkernel_*).
The timing should be more accurate this way, and the thread::last_time_type field
is no longer needed: all interrupts are now added to the kernel time (where the
time is actually spent).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11331 a95241bf-73f2-0310-859d-f6bbb57e9c96
a respective message, when the debugged team has been deleted.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11316 a95241bf-73f2-0310-859d-f6bbb57e9c96
goes away; therefore we must not return B_BAD_PORT_ID in this case.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11305 a95241bf-73f2-0310-859d-f6bbb57e9c96
Fixed rename implementation - it should now actually work (and also notify
the move which didn't happen before).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11229 a95241bf-73f2-0310-859d-f6bbb57e9c96
notify_listener() and send_notification() can be dropped.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11219 a95241bf-73f2-0310-859d-f6bbb57e9c96
for the kernel - but since we still compile libroot.so with kernel build
rules, I place it there for now.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11210 a95241bf-73f2-0310-859d-f6bbb57e9c96
Removed *_create() and *_create_entry_ref() syscalls - they are now handled by
*_open() and *_open_entry_ref() calls which get another parameter for the permissions.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11205 a95241bf-73f2-0310-859d-f6bbb57e9c96
Actually, it would be quite possible for another thread to specifically
ask for space in that reserved range, so we're playing safe now.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11177 a95241bf-73f2-0310-859d-f6bbb57e9c96
it, so that it can guarantee readding it again, in case the transfer is failing.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11173 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Added resolve_mount_point_to_volume_root() version for external use
(by the node monitor).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11159 a95241bf-73f2-0310-859d-f6bbb57e9c96
space to be created - this replaces the internal ID they had before.
Now, team_id == aspace_id.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11085 a95241bf-73f2-0310-859d-f6bbb57e9c96
it readable/writable for all userland apps. This is not what we want here.
We want to make the area cloneable for the registrar and hence use the
new flag B_USER_CLONEABLE_AREA. But there seems to be a bug that causes
strange things to happen, when using that flag only, therefore we go with
all of them until it is fixed.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11046 a95241bf-73f2-0310-859d-f6bbb57e9c96
* The counter semaphore is now only released when writing a command into
an formerly empty area. The userland part is aware of that. :-)
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11026 a95241bf-73f2-0310-859d-f6bbb57e9c96
have subdirectories in order to find device drivers like /dev/audio/raw
without any drivers in /dev/audio.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@11017 a95241bf-73f2-0310-859d-f6bbb57e9c96
to userland applications. A userland server (probably the registrar) will
register as the server responsible for delivering the messages. The
messages are passed to it via shared memory.
The advantage over sending the messages directly will be, that they won't
need to be dropped, if the receiver port is temporarily full.
Currently only the kernel side is implemented. Completely untested yet.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10983 a95241bf-73f2-0310-859d-f6bbb57e9c96
append-only and hence has very compact storage and a comparatively small
footprint implementation. Can be used for kernel->userland messaging.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10958 a95241bf-73f2-0310-859d-f6bbb57e9c96
cache_io() should now finally work correctly in all cases.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10906 a95241bf-73f2-0310-859d-f6bbb57e9c96
this file still does the right thing as suggested by Michael Pfeiffer.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10882 a95241bf-73f2-0310-859d-f6bbb57e9c96
Added short description to find_reserved_area() as that name is not that good.
vm_reserve_address_range() now doesn't do anything if the requested size
of the reservation is zero.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10867 a95241bf-73f2-0310-859d-f6bbb57e9c96
and not on a page boundary had been executed with a I/O size too large,
and thus clobbering all sorts of memory (oh my god).
The problems with ResourceFile now seems to be gone.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10863 a95241bf-73f2-0310-859d-f6bbb57e9c96
up for the first time, it will try to load all drivers that may export
devices there. Those devices will currently stay loaded, though - this
will only be changed when load_kernel_add_on() is able to cache add-ons.
Renamed pnp_devfs_probe() to pnp_devfs_register_device().
Fixed debug output for devfs_create().
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10838 a95241bf-73f2-0310-859d-f6bbb57e9c96
con_init() now takes care of the initialization of its currently built-in
console drivers itself.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10836 a95241bf-73f2-0310-859d-f6bbb57e9c96
Also includes the loader of R5 styled drivers (taken from dev.c).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10833 a95241bf-73f2-0310-859d-f6bbb57e9c96
an error in case the vnode name could not be retrieved without it).
Comment cleanup for consistency.
Renamed remaining "omode" variables to openMode.
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10831 a95241bf-73f2-0310-859d-f6bbb57e9c96
Added kernel private calls load_module() and unload_module() that can be used to
load/unload modules by path.
load_module_image() now only mentions when it couldn't load a module when tracing
is enabled (because the device manager uses it to determine if an add-on is a
module or an old-style driver).
git-svn-id: file:///srv/svn/repos/haiku/trunk/current@10821 a95241bf-73f2-0310-859d-f6bbb57e9c96
