"Windows 95 was 30 MB" is such an ignorant, obnoxious, trite take. a triple buffered framebuffer (which you want for smooth scrolling) for my 4K display is 70 MB in *pixels alone*. Obviously a complete webpage with precomposed textures would take more.https://twitter.com/julienPauli/status/1042113172143067138 …
-
-
yeah, but then fork stops to work, and there’re lots of things depending of fork. Actually understanding “there’s an option but nobody uses it” doesn’t make for a good argument though
-
I use it and it works, tell me where I'm wrong
-
So you are simply assuming *he* (implying *everyone* using linux kernel) can turn it off because it works for *you*. Are you suggesting this mechanism is actually not needed in the kernel as everyone should have it turned off?
-
nope. he claimed that setting vm.overcommit_memory=2 inherently breaks fork, which is obviously false, as anyone can verify by, for example, doing that on one of their machines and observing the results
-
so there’s no problems with fork and vm.overcommit_memory=2? Why isn’t it the default or even recommended setting then?
-
there are obviously no problems with fork and without overcommit because Linux is the only Unix-like that has overcommit enabled by default in the first place, yes.
-
i guess it's too much sarcasm to me. I have no idea what are you trying to say anymore
-
no sarcasm. I'm saying that you can just disable overcommit and have Linux work like every other OS with fork()...
- 1 more reply
New conversation -
-
-
I'm sure he understands that. There are better solutions than the oom killer, even for overcommit. And overcommit is so common because of a Unix design error (fork)
-
40-year-old Unix also had fork though
-
It didn't have the oom killer, so it had to have sensible page eviction/swapping solutions or just swap to death on overcommit
-
What are the better than OOM solution when low in memory? You could argue that OOM killer is worse than no solution, but AFAICT neither Windows nor Solaris has any solution. They just let you die while paging out, you can do that with Linux too.
-
one could argue that a better solution is to return an error from malloc/mmap/fork when a page needs to be reserved but doesn't have anything to back it.
-
I see, but I would argue that returning an error from malloc is not too different than killing the process, so I'm not sure if the extra effort really worth it for desktop applications. (should mmap do that with signal? Not entirely sure how would that work).
-
mmap can return an error too. the benefit of returning an error is that you can actually handle it, and e.g. free some caches and try again.
-
IIRC there are solutions for these usecases too: https://lwn.net/Articles/590960/ … I thought you had in mind signaling the app that accessed anonymous page w/o backing physical page. If you're unwilling to virtually allocate without physical pages, isn't that possible with Linux?
- 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.