From mboxrd@z Thu Jan 1 00:00:00 1970 X-Spam-Checker-Version: SpamAssassin 3.4.5-pre1 (2020-06-20) on ip-172-31-74-118.ec2.internal X-Spam-Level: X-Spam-Status: No, score=-1.9 required=3.0 tests=BAYES_00 autolearn=ham autolearn_force=no version=3.4.5-pre1 Path: eternal-september.org!reader02.eternal-september.org!feeder.eternal-september.org!news.gegeweb.eu!gegeweb.org!usenet-fr.net!proxad.net!feeder1-2.proxad.net!cleanfeed2-b.proxad.net!nnrp1-1.free.fr!not-for-mail Newsgroups: comp.lang.ada References: <00cd3aaa-d518-43a2-b321-58d6fae70aebo@googlegroups.com> <57eb7a65-51ea-4624-b9dc-9c4dda0fee59n@googlegroups.com> <5f70fd3b$0$13541$426a74cc@news.free.fr> <87wo0d3iac.fsf@nightsong.com> <87sgb02l7b.fsf@nightsong.com> <875z7vyy1u.fsf@nightsong.com> <87wo0bkns3.fsf@nightsong.com> <87sgayl6o0.fsf@nightsong.com> <87o8lml1md.fsf@nightsong.com> From: DrPi <314@drpi.fr> Subject: Re: is there a version of unix written in Ada Date: Thu, 1 Oct 2020 21:02:17 +0200 User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:78.0) Gecko/20100101 Thunderbird/78.3.1 MIME-Version: 1.0 In-Reply-To: <87o8lml1md.fsf@nightsong.com> Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: en-US Content-Transfer-Encoding: 8bit Message-ID: <5f7627bd$0$5863$426a74cc@news.free.fr> Organization: Guest of ProXad - France NNTP-Posting-Date: 01 Oct 2020 21:02:21 CEST NNTP-Posting-Host: 82.65.30.55 X-Trace: 1601578941 news-2.free.fr 5863 82.65.30.55:64972 X-Complaints-To: abuse@proxad.net Xref: reader02.eternal-september.org comp.lang.ada:60370 List-Id: Le 01/10/2020 à 11:46, Paul Rubin a écrit : > "Dmitry A. Kazakov" writes: >>> You have that either way. >> That is the point. > > With a conventional OS you have those synchronization delays AND > you have address translation delays. With Singularity you still > have the first, but you get rid of the second. Address translation is also a security feature. One process can't access data of another process. What about this with Singularity ? > >> So, where is performance gain? You still need storing/restoring >> registers and other context's data upon preemting. > > You run millions of instructions between preemptions, but you take the > address translation delay on EVERY memory access. TLB tables are reloaded when switching from one process to another. If the switched in process is swapped on disk, then you get a big performance hit. I don't get the point on translation delay on every memory access since the address translation is done by efficient specialized hardware. > >> Are you going to recompile and re-link everything in >> absolute addresses every time anything changes and then reboot? > > Position independent code is a thing. > >> Neither can do anything about it. You either have abstraction, like >> flat contiguous address space, however implemented, or you do not. The >> penalty is always there. You can have some relief from the hardware or >> none. > > Shrug, maybe there is some kind of block allocator like in the old days. > The original purpose of virtual memory was to allow simulating big ram > by paging to disk. Nobody cares about that any more. Really ? Have you had a look at Ressource manager, memory tab in Windows 10 ? You might be surprised. Even when everything fits in physical RAM, there are parts of (not used) allocated memory stored on disk. This way launching a new process is faster. At work, I use programs that require big amounts of RAM. Without disk swapping, I could not run them (16GB of physical RAM on the machine). We still do care virtual memory. > > That said, I don't know much about Singularity (never used it, haven't > read the papers) so maybe I'm missing something important. But the guys > who told me about it were knowledgeable and they were impressed by it. > > Anyway, this tangent started from the claim that nothing different had > been done in OS's in a while. I don't claim Singularity is great, only > that it's different. >