tag:blogger.com,1999:blog-3971202189709462152.post7777576128992250197..comments2024-03-11T12:50:02.036+01:00Comments on PyPy Status Blog: Software Transactional Memory lisp experimentsCarl Friedrich Bolz-Tereickhttp://www.blogger.com/profile/00518922641059511014noreply@blogger.comBlogger7125tag:blogger.com,1999:blog-3971202189709462152.post-445343620257710652013-07-17T18:20:28.280+02:002013-07-17T18:20:28.280+02:00This is a really great news.
Wish you all the bes...This is a really great news.<br /><br />Wish you all the best with further work!Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-64557313736606039802013-07-14T07:31:37.927+02:002013-07-14T07:31:37.927+02:00@glen: we're focusing right now on the machine...@glen: we're focusing right now on the machines we have, which are standard Intels with 4, 8, or at most 12 cores. I believe it is interesting too, and it's what people have right now in their own desktop or laptop computers. Obviously the scalability to larger numbers of cores is important as well, but we can't simply disregard any result involving less than 64 cores.Armin Rigohttps://www.blogger.com/profile/06300515270104686574noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-27409709072692019922013-07-13T19:19:02.010+02:002013-07-13T19:19:02.010+02:00You really need to go above 4 threads: 8,16,32, an...You really need to go above 4 threads: 8,16,32, and 64 at least. Then plot out the overhead of the STM related to this level of threading. If your benchmark is too small, alter it so that it makes sense to try and solve it with 64 threads.Glen Newtonhttps://www.blogger.com/profile/03639309432955855745noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-6192523689336878582013-07-12T14:39:09.491+02:002013-07-12T14:39:09.491+02:00All your arguments are valid, and I didn't rea...All your arguments are valid, and I didn't really expect much from hinting, just decided to try. It would seem that Haswell is still inching towards higher multicore scalability - probably thanks to improved atomic and fence ops in general. It's a benefit for those workloads that should conceptually scale well...Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-90242921498120541832013-07-12T14:15:48.011+02:002013-07-12T14:15:48.011+02:00Cool that you hacked on it! Note however that spi...Cool that you hacked on it! Note however that spinlock acquisition is not a blocker in these examples --- we implement STM mostly without locks, and locks are acquired rarely. Running independent code without getting STM conflicts means that each thread will in practice only acquire its own lock. And a single global lock is used for major GC --- but there, the large amount of work done means that using the Haswell xacquire/xrelease hints is just counterproductive.<br /><br />"Resulting scaling from 1 to 4 threads" doesn't mean anything, as in some examples it scales perfectly, and in other examples it doesn't scale at all (as expected).Armin Rigohttps://www.blogger.com/profile/06300515270104686574noreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-44441897768293509292013-07-12T14:12:23.990+02:002013-07-12T14:12:23.990+02:00Just to clarify my above comment: those were avera...Just to clarify my above comment: those were average factors of scaling per doubling of threads. So, 4-thread version ran actually 3.67, 3.50 and 3.54 times faster than single-threaded version.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-3971202189709462152.post-73034049586178144812013-07-12T14:06:34.422+02:002013-07-12T14:06:34.422+02:00I hacked a bit; inserted likely hint on early exit...I hacked a bit; inserted likely hint on early exit on spinlock acquisition, Haswell xacquire/xrelease hints on spinlock acquisition and release, and compiled with Haswell optimized flags.<br /><br />Resulting scaling from 1 to 4 threads for tests were 1.92, 1.87 and 1.88. I think that's already quite close to 2.<br /><br />I think this is OK, but not extraordinary.Anonymousnoreply@blogger.com