On Sat, Feb 08, 2025 at 09:37:34AM -0800, Jörn Engel wrote: > On Sat, Feb 08, 2025 at 04:02:47PM +0000, Lorenzo Stoakes wrote: > > > > To be clear, you won't get any kind of undefined behaviour (what that means > > wrt the kernel is not entirely clear - but if it were to mean as equivalent > > to the compiler sort 'anything might happen' - then no) or incomplete state. > > Going off on that tangent, I think compiler folks completely abuse > undefined behavior. Historically it simply meant that you might get two > or more different (and well-defined) results, depending on which > specific hardware you ran on. Somehow that was reinterpreted as "I have > license to do whatever I want". > > > > I guess you mean PROT_NONE? :) For the case in the thread you would have to > > have mapped a hugetlb area over the PROT_NONE one without MAP_NORESERVE and > > with insufficiently reserved hugetlb pages, a combination which should be > > expected to possibly fail. > > > > If you perform an mprotect() to R/W the range, you will end up with a 'one > > and done' operation. > > > > I'd also suggest that hugetlb doesn't seem to fit a malloc library like to > > me, as you rely on reserved pages, rather wouldn't it make more sense to > > try to allocate memory that gets THP pages? You could MADV_COLLAPSE to try > > to make sure... > > > > However, if aligned correctly, we should automagically give you those. > > We tried THP around 2012 and rejected it. The latency tail became a lot > longer and fatter. Various things have changed that might make THP less > bad today, but I am not aware of anyone reevaluating it. A _lot_ has changed. Try it again :) > > I think the problem with THP was the mmap_sem. Given a heavily threaded > process, the mmap_sem tends to be the one dominant lock in the kernel. A lot of work has been done on reducing mmap_sem contention. Again, worth another shot ;) > > A secondary problem might have been the background thread scanning for > pages that can be merged. Not sure about this part. We just disabled > THP and moved on to other problems. This sounds like really ancient problems that no longer exist (as well as a lot of FUD about THP at the time actually). We tend to proactively go THP if we can these days. > > > And yes, PROT_NONE/PROT_RW. Sorry! Haha no problem, just to clarify I understood you! > > Jörn > > -- > Every hour workers spend doing something else is an hour that they > aren't doing their jobs. > -- unknown >
