[libdispatch-dev] libdispatch port to FreeBSD (was: [libdispatch-changes] [27] trunk/src)

Bobby Powers bobbypowers at gmail.com
Mon Sep 14 13:51:15 PDT 2009


Hi Robert,

On Mon, Sep 14, 2009 at 4:05 PM, Robert Watson <robert at fledge.watson.org> wrote:
>
> Dear all--
>
> I have committed an initial autoconf/automake/libtool build framework to
> libdispatch svn, and a first cut at conditionally compiling Apple-specific
> pieces of libdispatch.  This allows libdispatch to build on both FreeBSD and
> Mac OS X Snow Leopard.

Thats great, from the looks of it it must have taken a lot of time to
split it up so much!

> Some notes for FreeBSD hackers:
>
> - Several kqueue extensions (EVFILT_USER, NOTE_TRIGGER, EV_RECEIPT, and
>  EV_DISPATCH) are required in the FreeBSD kernel, and hopefully will go into
>  the FreeBSD tree in the next couple of days.
>
> - Blocks are definitely not yet supported -- there seems to be a lot of
>  interest, so I'm optimistic.  As compilers aren't my forte, any assistance
>  in this department would be most appreciated. :-)
>
> - On FreeBSD, there is no support for Apple's pthread work queue primitive,
> so
>  I'm using the fallback worker code in libdispatch.  The FreeBSD kernel
>  community needs to have a conversation about the right way to address this
>  requirement, but in the mean time the worker model should be fine.
>
> - There are also some kqueue events present in FreeBSD but not in Mac OS X;
>  these may require tweaks to libdispatch to fully support by GCD-aware
>  applications using these FreeBSD events.
>
> - I have developed and tested on a 32-bit Intel system; the comments below
> on
>  -march on Mac OS X likely apply for FreeBSD/amd64 users as well.
>
> Some notes for Mac OS X hackers:
>
> - For now, you will need to hand-trim the -march line from src/Makefile.
>  This should be fixed soon by some further autoconf work.
>
> - When I use the externally-built dylib on Snow Leopard, test programs spin.
>  This may be due to interference between the libSystem dispatch code and the
>  dylib -- possibly forcing the use of pthread_create_key rather than
> reserved
>  keys would help.
>
> - I have not attempted to configure/build for Leopard as yet, but this
> change
>  should make that a lot easier.  EVFILT_USER is likely the biggest obstacle.
>
> Some notes for Linux hackers:
>
> - The lack of kqueue support is the critical obstacle in a port to Linux.
>  It
>  may be that a kqueue emulation library based on epoll (or even libevent)
>  could fill this gap.  If someone wants to look at this, pay very careful
>  attention to kqueue semantics, as libdispatch relies on them heavily.

I am quite interested in this, but don't have much experience yet with
event interfaces.  I'll try to prepare some thoughts by the end of the
week, but libevent seems very interesting.

Thanks for all the work so far.

yours,
Bobby

> Please consider this a work-in-progress -- don't be surprised when it
> doesn't build, let alone work.  :-)
>
> Finally: this work would not have been possible without the support of
> Jordan Hubbard, Kevin Van Vechten, Dave Zarzycki, Bill Siegrist, and Apple.
>  I am surprised by how easily the porting has gone so far -- it's a credit
> to their careful planning and quality of implementation that it has gone so
> smoothly.
>
> Thanks also to Peter O'Gorman and Colin Percival for their assistance with
> auto*, my nemesis.
>
> With any luck, there will be more good porting news to report in the near
> future!
>
> Thanks,
>
> Robert N M Watson
> Computer Laboratory
> University of Cambridge
>
> ---------- Forwarded message ----------
> Date: Mon, 14 Sep 2009 11:46:19 -0700 (PDT)
> From: source_changes at macosforge.org
> Reply-To: libdispatch-dev at lists.macosforge.org
> To: libdispatch-changes at lists.macosforge.org
> Subject: [libdispatch-changes] [27] trunk/src
>
>
>  Revision
>      27
>
>  Author
>      robert at fledge.watson.org
>
>  Date
>      2009-09-14 11:46:18 -0700 (Mon, 14 Sep 2009)
>
>  Log Message
>
> Port of libdispatch to the FreeBSD operating system, which consisted of
> disabling various Mac OS X-specific portions of the library (such as Mach),
> providing portable versions of certain services internally (POSIX semaphores
> instead of Mach semaphores, clock_gettime() instead of mac_absolute_time()),
> and providing an autoconf/automake/libtool build infrastructure.  This is
> just a first cut, refinement to follow.
>
> Conditionalize portions of the libdispatch build as follows:
>
> HAVE_MACH
>        If Mach is present, support Mach ports throughout libdispatch.
>
> USE_LIBDISPATCH_INIT_CONSTRUCTOR
>        If defined, add the 'constructor' attribute to libdispatch_init so
>        that it runs as part of library initialization.  This is not needed
>        in Mac OS X libSystem due to explicit invocation of libpthread_init
>        by libsyscall, but is for a stand-alone library.
>
> HAVE_DECL_EVFILT_SESSION, HAVE_DECL_NOTE_SIGNAL, HAVE_DECL_NOTE_REAP,
> HAVE_DECL_NOTE_NONE, HAVE_DECL_VQ_UPDATE, HAVE_DECL_VQ_VERYLOWDISK
>        Conditionally support kqueue events that are not present on FreeBSD.
>
> DISPATCH_NO_LEGACY
>        Expand the scope of this existing macro to allow libdispatch to
>        build without legacy.c.
>
> HAVE_MACH_ABSOLUTE_TIME
>        Wrap mach_absolute_time() in _dispatch_absolute_time(), and
>        provide a portable clock_gettime() implementation.  Some further
>        refinment will be required here.
>
> USE_APPLE_SEMAPHORE_OPTIMIZATIONS
>        Only use non-portable assembly optimizations of dispatch semaphores
>        if supported.
>
> HAVE_SEM_INIT
>        Use POSIX semaphores if Mach semaphores aren't present; further
>        ifdef'ing possibly required here.  If neither HAVE_MACH nor
>        HAVE_SEM_INIT are present, the compile will fail.
>
> USE_APPLE_CRASHREPORTER_INFO
>        Only define Mac OS X crashreporter state if requested.
>
> HAVE_PTHREAD_WORKQUEUES
>        Only compile in support for pthread work queues if present;
>        otherwise force use of normal worker threads.
>
> HAVE_PTHREAD_KEY_INIT_NP
>        If defined, use non-portable allocation of pthread keys--otherwise,
>        use pthread_create_key().
>
> USE_APPLE_TSD_OPTIMIZATIONS
>        Only use non-portable assembly optimizations if dispatch thread-
>        specific data if this is defined.
>
> HAVE_MALLOC_CREATE_ZONE
>        If not present, use compat/malloc_zone.h which implements the malloc
>        zone APIs using normal malloc(3).
>
> Also fix some portability nits for building, such as casting before using
> long as a format string for several types.  Note that this port will not
> build or run on a stock FreeBSD release or svn distribution yet, as it
> relies on kqueue extensions not committed back to FreeBSD yet.
>
> There are several known issues, such as how best to map CLOCK_UPTIME or
> CLOCK_MONOTONIC struct timespec timestamps into a Mach absolute time
> int64_t number space.  Also, the POSIX semaphore ifdefing is adequate but
> not complete.
>
>  Modified Paths
>
>  *  trunk/src/benchmark.c
>  *  trunk/src/dispatch.h
>  *  trunk/src/internal.h
>  *  trunk/src/legacy.h
>  *  trunk/src/os_shims.h
>  *  trunk/src/private.h
>  *  trunk/src/queue.c
>  *  trunk/src/queue_internal.h
>  *  trunk/src/semaphore.c
>  *  trunk/src/semaphore_internal.h
>  *  trunk/src/shims.c
>  *  trunk/src/source.c
>  *  trunk/src/source.h
>  *  trunk/src/source_private.h
>  *  trunk/src/time.c
>
>  Diff
>
>    Modified: trunk/src/benchmark.c (26 => 27)
> _______________________________________________
> libdispatch-dev mailing list
> libdispatch-dev at lists.macosforge.org
> http://lists.macosforge.org/mailman/listinfo.cgi/libdispatch-dev
>



More information about the libdispatch-dev mailing list