I read the blog posts about pointer compression and the advantages, it seems like you get some performance benefit too?
The end goal is to set a maximum heap limit for a script execution (often short-lived), and when breached, raise a nice "out of memory" error that an ordinary user could understand. If the end users chooses, they can reexecute with a higher ceiling. Sometimes a lot of RAM would be nice to use, but you really want to set some kind of expected limit so all the memory isn't consumed by a runaway calculation. With full pointers, that would involve turning off pointer compression, but that would also disable cppgc_young_generation and a lot of the other cppgc features, right? I think I saw the young-generation requires a cage which requires pointer compression? On Wed, 22 Mar 2023 at 22:53, dinf...@chromium.org <dinfu...@chromium.org> wrote: > > > On Wednesday, March 22, 2023 at 2:47:15 PM UTC+1 harr...@gmail.com wrote: > > Ok thanks Dominik and Camillo. > I think RequestTermination can fail to actually interrupt execution if > there is no additional callouts into the engine where it can throw > something. But better than nothing! > > > I believe in such cases we could introduce additional checks to fix those > issues. That would likely help other embedders as well (e.g. DevTools uses > the same callback for a similar purpose). > > > > How big can the heap be these days? > I've seen mention of 2GB, 4GB, 8GB ... more? > > > With ptr compression we are limited to 4GB but with full pointers the > limit could be manually set to something that's larger than that. > > > > On Tuesday, March 21, 2023 at 4:23:21 PM UTC+8 dinf...@chromium.org wrote: > > Hi, > > I don't think it's possible to do that from the near heap limit callback > as our C++ code doesn't doesn't handle allocation failures gracefully. I > believe on OOM you have to do something like: 1) extend the limit using > that callback first and then 2) use Isolate::RequestTermination to cancel > execution. > > Hope that helps! > > Cheers, > Dominik > > On Tuesday, March 21, 2023 at 8:37:46 AM UTC+1 Camillo Bruni wrote: > > Hi Paul, > > You could try using v8::Isolate::TerminateExecution(). It's a best-effort > approach to limit further JS execution and allocations. > > cheers, > Camillo > > On Tue, 21 Mar 2023 at 08:00, Paul Harris <harr...@gmail.com> wrote: > > Hi again, > > Is there a nice way I could cancel a JS execution (eg raise an exception) > from within eg the NearHeapCallback to avoid OOMs? > > I'd rather if the script is consuming more RAM than expected, that it > raises exceptions and unwinds rather than bombing out. > > Thanks, > Paul > > -- > -- > v8-dev mailing list > v8-...@googlegroups.com > http://groups.google.com/group/v8-dev > --- > You received this message because you are subscribed to the Google Groups > "v8-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to v8-dev+un...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/v8-dev/9b1749b5-ccda-4a74-b846-d0347121c7fen%40googlegroups.com > <https://groups.google.com/d/msgid/v8-dev/9b1749b5-ccda-4a74-b846-d0347121c7fen%40googlegroups.com?utm_medium=email&utm_source=footer> > . > > Camillo Bruni | Software Engineer, V8 | Google Germany GmbH | Erika-Mann > Str. 33, 80636 München > > Registergericht und -nummer: Hamburg, HRB 86891 | Sitz der Gesellschaft: > Hamburg | Geschäftsführer: Paul Manicle, Halimah DeLaine Prado > > Diese E-Mail ist vertraulich. Falls Ssie diese fälschlicherweise erhalten > haben sollten, leiten Sie diese bitte nicht an jemand anderes weiter, > löschen Sie alle Kopien und Anhänge davon und lassen Sie mich bitte wissen, > dass die E-Mail an die falsche Person gesendet wurde. This e-mail is > confidential. If you received this communication by mistake, please don't > forward it to anyone else, please erase all copies and attachments, and > please let me know that it has gone to the wrong person. > > -- > -- > v8-dev mailing list > v8-dev@googlegroups.com > http://groups.google.com/group/v8-dev > --- > You received this message because you are subscribed to a topic in the > Google Groups "v8-dev" group. > To unsubscribe from this topic, visit > https://groups.google.com/d/topic/v8-dev/OeJ_DlMj3v8/unsubscribe. > To unsubscribe from this group and all its topics, send an email to > v8-dev+unsubscr...@googlegroups.com. > To view this discussion on the web visit > https://groups.google.com/d/msgid/v8-dev/62d4fd80-7280-4f14-ba4a-a3eafdeb52f8n%40googlegroups.com > <https://groups.google.com/d/msgid/v8-dev/62d4fd80-7280-4f14-ba4a-a3eafdeb52f8n%40googlegroups.com?utm_medium=email&utm_source=footer> > . > -- -- v8-dev mailing list v8-dev@googlegroups.com http://groups.google.com/group/v8-dev --- You received this message because you are subscribed to the Google Groups "v8-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to v8-dev+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/v8-dev/CAMSUDdaDNFSW%3DU9Pz%2BbKr8K3-OqqgiPxCedaje_4QFU_OKrk9Q%40mail.gmail.com.
