only some sub requests worked. Previously we would have simply canceled all
of the subrequests in the loop because we cancel after the first error.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32914 a95241bf-73f2-0310-859d-f6bbb57e9c96
* An off by one error prevented the very last block of a device to be accessed
through IO.
* In case of error the request wasn't notified causing anyone (the page writer
for example) to wait forever for the request to complete.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32913 a95241bf-73f2-0310-859d-f6bbb57e9c96
capabilities to aid in debugging memory corruption issues.
It does:
* Initialize memory to 0xcc to help turn up use of uninitialized memory
* Set freed memory to 0xdeadbeef to help find accesses of freed memory
* Use the paranoid heap validation to turn up many cases of memory corruption
* Use a simplistic wall check to turn up memory overwrites past allocations
* Take extra steps to validate freed addresses to turn up misaligned frees
It has an interface to en-/disable paranoid validation and to start/stop regular
wall checking. Both are currently just enabled. At a later stage a debug version
of libroot could be used by an application and the checks enabled at will. Note
that due to the paranoid validation and the suboptimal locking this allocator
will perform horribly. Still to find memory corruption issues in the system or
also in your applications it can be helpful to build your installation with it
turned on. To enable it you currently need to edit the Jamfile to sub-include
the malloc_debug instead of the malloc directory.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32894 a95241bf-73f2-0310-859d-f6bbb57e9c96
for the page list but not the area one. Since multiple pages can be allocated
at once, even an area that is not at the top of the list can become empty. In
such a case the area list would previously have lost entries. Also because
we can remove more than one page from any area, not just the top one, we may
need to move forward in the list so that it stays ordered by free pages.
+alphabranch
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32880 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Got rid of <dirent_private.h> -- the __DIR structure is private to dirent.c,
now. The attribute directory, index directory, and query functions use the
the public POSIX API, so does the kernel module code. Those components were
not initializing the structure correctly anymore since the introduction of
telldir()/seekdir().
+alphabranch
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32819 a95241bf-73f2-0310-859d-f6bbb57e9c96
to " &" as really the whole rest of the file uses that style.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32813 a95241bf-73f2-0310-859d-f6bbb57e9c96
switch /boot to a different volume in two operations, unless you have first
linked /system/lib into /bin. This patch assumes that / will always have the
ID 1. Don't know if that is proper. Note that I also thought about solving
this in the VFS, since perhaps it isn't the job of root-fs to know about
/boot, but that would of course introduce another check for every rename
operation, which I decided against.
+alphabranch
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32806 a95241bf-73f2-0310-859d-f6bbb57e9c96
Seems like we didn't remember to change it back, even
with the #warning. :-)
I tested ftp and it works fine. This also closes ticket #4176.
+alphabranch
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32766 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Moved device_manager/settings.cpp to debug/safemode_settings.cpp.
* Removed the somewhat hacky concatenation of kernel settings and safemode
settings in the boot loader. Instead, get_safemode_option() will now fall
back to the kernel settings, if it couldn't spot a setting in the safemode
settings. This allows for more control, and also makes enabling serial
debug output actually work (ie. overriding the kernel settings via safemode
options).
* Adjusted debug_init_post_vm(), and smp_init_other_cpus() to use
get_safemode_boolean().
* Therefore, I added safemode_settings.cpp to the boot loader as well.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32685 a95241bf-73f2-0310-859d-f6bbb57e9c96
So either +alphabranch or remove the declarations from <dirent.h>.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32679 a95241bf-73f2-0310-859d-f6bbb57e9c96
into the scope of the lock. Saves an InterruptsSpinLocker.
* Use an InterruptsSpinLocker() as in other places.
* When creating/forking a team fails because the kernel thread cannot be spawned
balance the already sent TEAM_ADDED notification by sending a TEAM_REMOVED one.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32595 a95241bf-73f2-0310-859d-f6bbb57e9c96
been removed. That includes CPU disabling and thread pinning, as that becomes
pointless with only one CPU.
* Return a proper reschedule hint on enqueing a thread, based on the priority
of the current thread vs. the enqueued one.
* Enable dynamic scheduler selection. With one CPU the simple scheduler will
be used, otherwise affine is selected.
* Removed the scheduler type define as we now always auto-select it.
* Some cleanup.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32573 a95241bf-73f2-0310-859d-f6bbb57e9c96
* scheduler_enqueue_in_runqueue() now allows the scheduler to return a hint as to whether a reschedule is desirable or not. This is used in a few other places in order to relegate scheduling decisions entirely to the scheduler rather than the priority hacks previously used. There are probably other places in the kernel that could now make use of that information to more intelligently call reschedule() though.
* Switch over the default scheduler to scheduler_affine().
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32554 a95241bf-73f2-0310-859d-f6bbb57e9c96
you can easily produce a dead lock. This should fix the system hang
experienced as part of bug #4223.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32526 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Keep track of the currently running threads.
* Make use of that info to decide if a thread that becomes ready should preempt
the running thread.
* If we should preempt we send the target CPU a reschedule message.
* This preemption strategy makes keeping track of idle CPUs by means of a bitmap
superflous and it is therefore removed.
* Right now only other CPUs are preempted though, not the current one.
* Add missing initialization of the quantum tracking code.
* Do not extend the quantum of the idle thread based quantum tracking as we want
it to not run longer than necessary. Once the preemption works completely
adding a quantum timer for the idle thread will become unnecessary though.
* Fix thread stealing code, it did missed the last thread in the run queue.
* When stealing, try to steal the highest priority thread that is currently
waiting by taking priorities into account when finding the target run queue.
* Simplify stealing code a bit as well.
* Minor cleanups.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32503 a95241bf-73f2-0310-859d-f6bbb57e9c96
re-insert it at a new place, but by only setting the priority and not the
next_priority field, the thread would actually be enqueued at the same priority
level as before. Didn't cause any real damage, guess it was just an oversight.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32494 a95241bf-73f2-0310-859d-f6bbb57e9c96
not hold the bin lock while setting and only protect the actual freeing. May
reduce bin lock contention a little bit.
* Tiny optimization for force-clearing 0xdeadbeef. Do it after setting to 0xcc,
so that it's less likely we have to do it.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32473 a95241bf-73f2-0310-859d-f6bbb57e9c96
* Always select POLLNVAL. We wouldn't get notified, if the FD was closed while
waiting.
* Skip negative FDs (as per POSIX standard).
* When encountering an invalid FD, we must not wait. Previously we did that as
long as any FD was valid.
* The return count must consider all FDs with non-null revents, including
invalid ones (revents is POLLNVAL then).
common_poll()/common_wait_for_objects():
* Don't bail out early, when no FDs/objects have been selected. We always want
to wait.
_user_poll()/_user_wait_for_objects():
* Copy the array back to userland even if common_*() returned an error. The
functions do now always fill in the resulting events.
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32418 a95241bf-73f2-0310-859d-f6bbb57e9c96
* B_EVENT_INVALID is no longer passed to the FD's select()/deselect() hooks.
* Now we always attach the select info to the I/O context, even if no event has
been selected. The reasoning is that B_EVENT_INVALID is always automatically
selected and handled by the VFS, so we need the handle to notify on close().
git-svn-id: file:///srv/svn/repos/haiku/haiku/trunk@32417 a95241bf-73f2-0310-859d-f6bbb57e9c96
