On Sat, Jan 18, 2020 at 12:26:53AM +0300, Kirill A. Shutemov wrote: > On Fri, Jan 17, 2020 at 09:32:39AM -0800, Minchan Kim wrote: > > On Fri, Jan 17, 2020 at 06:58:37PM +0300, Kirill A. Shutemov wrote: > > > On Fri, Jan 17, 2020 at 12:52:25PM +0100, Michal Hocko wrote: > > > > On Thu 16-01-20 15:59:50, Minchan Kim wrote: > > > > > There is usecase that System Management Software(SMS) want to give > > > > > a memory hint like MADV_[COLD|PAGEEOUT] to other processes and > > > > > in the case of Android, it is the ActivityManagerService. > > > > > > > > > > It's similar in spirit to madvise(MADV_WONTNEED), but the information > > > > > required to make the reclaim decision is not known to the app. Instead, > > > > > it is known to the centralized userspace daemon(ActivityManagerService), > > > > > and that daemon must be able to initiate reclaim on its own without > > > > > any app involvement. > > > > > > > > > > To solve the issue, this patch introduces new syscall process_madvise(2). > > > > > It uses pidfd of an external processs to give the hint. > > > > > > > > > > int process_madvise(int pidfd, void *addr, size_t length, int advise, > > > > > unsigned long flag); > > > > > > > > > > Since it could affect other process's address range, only privileged > > > > > process(CAP_SYS_PTRACE) or something else(e.g., being the same UID) > > > > > gives it the right to ptrace the process could use it successfully. > > > > > The flag argument is reserved for future use if we need to extend the > > > > > API. > > > > > > > > > > I think supporting all hints madvise has/will supported/support to > > > > > process_madvise is rather risky. Because we are not sure all hints make > > > > > sense from external process and implementation for the hint may rely on > > > > > the caller being in the current context so it could be error-prone. > > > > > Thus, I just limited hints as MADV_[COLD|PAGEOUT] in this patch. > > > > > > > > > > If someone want to add other hints, we could hear hear the usecase and > > > > > review it for each hint. It's more safe for maintainace rather than > > > > > introducing a buggy syscall but hard to fix it later. > > > > > > > > I have brought this up when we discussed this in the past but there is > > > > no reflection on that here so let me bring that up again. > > > > > > > > I believe that the interface has an inherent problem that it is racy. > > > > The external entity needs to know the address space layout of the target > > > > process to do anyhing useful on it. The address space is however under > > > > the full control of the target process though and the external entity > > > > has no means to find out that the layout has changed. So > > > > time-to-check-time-to-act is an inherent problem. > > > > > > > > This is a serious design flaw and it should be explained why it doesn't > > > > matter or how to use the interface properly to prevent that problem. > > > > > > I agree, it looks flawed. > > > > > > Also I don't see what System Management Software can generically do on > > > sub-process level. I mean how can it decide which part of address space is > > > less important than other. > > > > > > I see how a manager can indicate that this process (or a group of > > > processes) is less important than other, but on per-addres-range basis? > > > > For example, memory ranges shared by several processes or critical for the > > latency, we could avoid those ranges to be cold/pageout to prevent > > unncecessary CPU burning/paging. > > Hmm.. I still don't see why any external entity has a better (or any) > knowledge about the matter. The process has to do this, no? I think Sandeep already gave enough information in other thread. > > > I also think people don't want to give an KSM hint to non-mergeable area. > > And how the manager knows which data is mergable? Oleksandr, could you say your thought why you need address range based API? > > If you are intimate enough with the process' internal state feel free to > inject syscall into the process with ptrace. Why bother with half-measures? Concern is we want to act the hint in caller's context, not calle because calle is usually very limited in cpuset/cgroups or even freezed state so they couldn't act by themselves quick enough, which makes many problems. One of efforts to solve the issue was "Expedited memory reclaim" https://lwn.net/Articles/785709/ That could be also a good candidate for process_madvise API.
