| Age | Commit message (Collapse) | Author |
|---|
|
We should probably split tip and cu into separate man pages with
the common bits included.
|
|
dealing with output source selection (doesn't work right yet, so its disabled).
|
|
this was done in -two- places is beyone me); spotted and tested by Nick.
|
|
|
|
ok hin@
|
|
|
|
is written. I'm not sure how to -correctly- handle this for the fullduplex
case, but this seems to make playback sound MUCH better...
|
|
|
|
of the thread regressions tests pass on i386, sparc, sparc64
(save the siginfo test on sparc64 due to a kernel issue) and alpha.
The tests should also pass on ppc.
In addition, it fixes the problems with the "mysql torture test"
provided by one of our users. The python port also appears to
work correctly with these changes.
Summary of changes:
* check_pending removed from thread structure, no longer used.
* unused elements of sigstatus structure removed. The sigstatus
structure is now used to keep track of siginfo data instead of
defining a local array in uthread_sig.c.
* _thread_kern_sched_sig removed
* _thread_sig_process goes away -- can't have a lock active when
signal handlers are called. Functions now call _thread_sig_handle
directly.
* _thread_clear_pending now used lib wide to clear pending flags.
It was named _clean_pending_flag and only used in uthread_sig.c.
The function clears both per thread signals, and per process signals.
* _thread_sig_handle now returns a value.
* unused debugging macros removed from the thread kernel
* dispatch pending signals after switching to a thread providing
that signal handling hasn't been deferred by the thread.
* process thread switchhooks before dispatching pending sigs
* check for thread cancellation before dispatching pending sigs
* re-wrote pthread-kill to do the correct thing. It now does
minimal thread-kill-specific processing and then calls the
existing code in uthread_sig to process the generated signal.
* shut the compiler up when compiling uthread_mutex.c
* no more "signal_lock". It does more harm than good.
* keep track of "per-process" signals.
* don't bother saving siginfo_t data for the scheduling signal.
* per posix: SIGSTOP cleared when SIGCONT received and vice versa.
* add _dispatch_signal to properly dispatch a signal to a thread.
It sets the appropriate signal mask, something that was missing
in the previous implementation. This fixes several bugs.
The previous method held a lock. If the signal handler longjmp-ed
the lock was never cleared and no more signals were processed.
|
|
half-sized areas. fixes preemption_float on powerpc.
|
|
based on netbsd.
|
|
|
|
|
|
|
|
|
|
This allows us to remove some ambiguities on how some functions are called,
remove some diagnostic checks for conditions that can never happen and
remove the ugly hack with "pmap_initialized".
Then, rework internal function interfaces and some logic so as to stop
fetching vm_page from a pa and the reverse every now and then - this makes
some pmap operations run much faster.
per art@'s idea.
|
|
is still a bit spotty, but at least interrupts are not enabled constantly
any more.
Also discovered a feature: messing with the LINE/MIC bit in GAINCTL has
the side effect of wedging the chip under some circumstances, but I think
this can be worked around by resetting the codec... later.
|
|
ok millert@
|
|
|
|
|
|
bug #46, map@appgate.com; ok henning@, provos@, stevesk@
|
|
to replace queues (or options) which are global and can't be loaded into
an anchor. Found by j@pureftpd.org.
|
|
|
|
hbhopt_print() and dstopt_print() can return 0 if
the option is located just one byte short of snapend
this would cause an infinite loop in ip6_print().
|
|
|
|
handlers will be called with the current signal masked unless the
handler was installed with the SA_NODEFER flag. The test did
not check this (and the pthread code was incorrect by not setting
the mask). This fixes the test. The pthread lib part of the fix
will be made in a short while. Untill then the test will fail.
|
|
|
|
chunk of this driver will be rewritten once I've gotten some zzz's.
|
|
|
|
|
|
This is annoying because the chip will keep dma'n, but everytime I disable
interrupts in halt_*, they never come back. Grr.
|
|
|
|
|
|
|
|
|
|
|
|
from NetBSD kern/19614
|
|
|
|
|
|
|
|
pintr/parg)
|
|
but the timing appears to get off some how and buffers just stay on the
chip. Oh well, at least its keeping me from playing in the traffic.
|
|
|
|
|
|
|
|
millert.
|
|
(longnames). This ensures the install sets are listed in lowercase and
are thus visible to the install/upgrade scripts.
Installing/upgrading from sets on msdos partitions should now work
(again?).
Problem noticed by Michael Lucas.
deraadt@ ok
|
|
Simplify and clarify logic around selecting/installing sets.
Rename do_selection_list() to select_sets(), and install_get_files()
to install_files().
Eliminate glob_selection() and get_get_files() and the global
GET_FILES, folding logic into select_sets() and install_files().
Show the user entire 'url' when no sets are found, rather than trying
to hide mount points, etc.
Eliminate VERSION_MAJOR and VERSION_MINOR since they were always used
together. Instead, simply construct a few variables (SETDIR, FTPDIR,
OBSD) holding the strings these were used to construct.
Various 'standard' cleanups of quotes, braces, etc. in modified areas.
Also shrinks install.sub by 681 bytes.
|
|
|
|
yet, but the knobs and interrupt routine seem to work. More to come.
|
