[swift-corelibs-dev] Issue understanding libdispatch runloop poking
phabouzit at apple.com
Mon Oct 16 13:39:38 CDT 2017
> On Oct 16, 2017, at 8:03 AM, Rogiel Sulzbach via swift-corelibs-dev <swift-corelibs-dev at swift.org> wrote:
> Thank you for all the help. I will try to take a look at this during the week.
> I am not even running it in swift yet, I just got it to compile (without actually working) so that I could get Foundation to build without having to disable the entire URL system, but I would like to use dispatch.
> The reason I am going after runloops is because it was the first test to fail and many tests seem to depend on it. It seemed like a logical place to start porting.
It really isn'tm it's a high level construct that needs a TON of dispatch working for it to work, I'd start with simpler tests first. and sample code you write here and there, that only use pure dispatch, no runloops.
Focus on sources and the thread pool, this is where your issues will arise first. Once that is stable, you can move to runloops as it is built atop of the rest.
> I think i get it now. Ill have to manually pass the fd to the kevent implementation. Makes sense. As soon as I get things starting to work i'll fork and keep a separate branch with these changes to ease review.
It's also for your sake, if during review was ask you to do things differently or find some issue, then it's very likely that there will be ripple effect in the rest of your porting.
I encourage you to do focused changes (not small in terms of number of files touched, but small in terms of scope). It'll also allow us to check that it doesn't regress other ports which is always a concern ;)
> On Oct 12, 2017, 9:08 PM -0300, Pierre Habouzit via swift-corelibs-dev <swift-corelibs-dev at swift.org>, wrote:
>> oh and last, while working on FreeBSD, we are killing support for pthread_workqueue_additem_np, pthread_workqueue_setdispatch_np and similar functions that are ancient and do not support all the required prioirties. in your port, please do not try to repair this support if it is broken, because we're killing it anyway, that would be wasted time.
>> The minimal workqueue implementation that we'll support requires implementation of:
>> - _pthread_workqueue_init
>> - _pthread_workqueue_addthreads
>> Until FreeBSD supports this, the same internal workqueue implementation Linux has can be used. These interfaces above are not that different in terms of kernel requirements from the old interfaces and shouldn't be too difficult to support upstream long term from the original pthread_workqueue_additem_np era APIs FreeBSD had support for originally.
>> Anything older than that is about to be removed.
>>> On Oct 12, 2017, at 4:56 PM, Pierre Habouzit via swift-corelibs-dev <swift-corelibs-dev at swift.org <mailto:swift-corelibs-dev at swift.org>> wrote:
>>>> On Oct 12, 2017, at 4:52 PM, Pierre Habouzit via swift-corelibs-dev <swift-corelibs-dev at swift.org <mailto:swift-corelibs-dev at swift.org>> wrote:
>>>>> On Oct 12, 2017, at 7:48 AM, Rogiel Sulzbach <rogiel at rogiel.com <mailto:rogiel at rogiel.com>> wrote:
>>>>> I am trying to get libdispatch working on FreeBSD but I am hitting a wall: I cannot get my head around how the runloop implementation dispatches a event to the kqueue once a new job is submitted. So far I was able to get the library to fully compile but most tests are timing out due to a issue getting the kqueue event handler to awake when the poke method is called.
>>>>> Looking at the Linux implementation (https://github.com/apple/swift-corelibs-libdispatch/blob/master/src/queue.c#L6119) <https://github.com/apple/swift-corelibs-libdispatch/blob/master/src/queue.c#L6119)> it appears to be creating a new eventfd descriptor and calls eventfd_write on the poke method. A eventfd_read call is only made in the event_epoll.c implementation file BUT I am unable to find where the specific fd (created in queue.c) is forwarded to the epool handler. In other words, how is the event queue implementation informed of that file descriptor?
>>>>> On the Darwin side of things, it appears to use a mix of kqueue (on the event handling side) and mach ports (on the signaling side), but i still cannot understand how the event_kevent.c implementation becomes aware of that port allocated on queue.c.
>>>>> In order do get dispatch working on FreeBSD I need to be able to understand the flow that happens between _dispatch_runloop_queue_class_poke and a event getting fired in the kqueue implementation (which I believe should be happening in _dispatch_kq_poll).
>>>>> Pierre Habouzit replied to me on the bug tracker:
>>>>> The equivalent of the eventfd on kevent systems is to use an EVFILT_USER. It is very likely that the current implementation using kevent is badly broken on FreeBSD it has never run for a very long time. I have no access to FreeBSD at this time to help though but I can help with questions on the list.
>>>>> Indeed, the entire library needed changes in 39 files to get compiling, let alone working! But I am getting there!
>>>> 39 seems a bit excessive so I'd encourage you when you submit this awesome porting effort to try to check in individual "porting fixes" in separate branches.
>>>> You should also try to get it working at the C level before you try to integrate with Swift, runloop queues are not required to get the basic library working
>>>>> If i am understanding it correctly, I must send a EVFILT_USER and upon receiving that event I get the corresponding unote and call dux_merge_evt. Right?
>>>> Hmm actually when answering on Jira I had missed that you were talking about runllops. However, the equivalent of eventfd on kevent will be EVFILT_USER.
>>>> You will need to first work on the implementation of CFRunloop itself to understand how it "blocks". It will likely park in kevent() with a kqueue fd.
>>>> The CFRunloop in question will prepare an EVFILT USER that represents "availability" of dispatch items to drain in the runloop queue. the difficulty is that unlike epoll of portsets on the Mac, the runloop queue has to know your kqueue fd. So you need to devise a way to pass it down.
>>>> Once you have done that _dispatch_runloop_queue_class_poke() on FreeBSD will issue a NOTE_TRIGGER of the EVFILT_USER which will break the CFRunloo parked in kqueue() out, and this will recognize the EVFILT_USER as being the hint that you need to drain it.
>>>> I would suggest that on FreeBSD you add a _dispatch_runloop_queue_set_kqueue() which you will store where the fd/mach port is stored on other platforms.
>>> it also means that you likely need to export some function shared between Swift & Dispatch to "recognize" the knote in question, you'll need to pick an ident and a udata field. You probably want to have defines in <dispatch/private.h> when on FreeBSD for this purpose, since both dispatch & CF/Swift will need to know how to form the knote in question.
>>> swift-corelibs-dev mailing list
>>> swift-corelibs-dev at swift.org <mailto:swift-corelibs-dev at swift.org>
>>> https://lists.swift.org/mailman/listinfo/swift-corelibs-dev <https://lists.swift.org/mailman/listinfo/swift-corelibs-dev>
> swift-corelibs-dev mailing list
> swift-corelibs-dev at swift.org
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the swift-corelibs-dev