<html><head><meta http-equiv="content-type" content="text/html; charset=utf-8"></head><body dir="auto"><div dir="ltr"></div><div dir="ltr">Hi all,</div><div dir="ltr"><br></div><div dir="ltr">To get full process malloc stats I would use to <a href="https://linux.die.net/man/3/mallinfo">https://linux.die.net/man/3/mallinfo</a></div><div dir="ltr"><br></div><div dir="ltr">This will not include ofc anon pages allocated by mmap, but should cover stats.</div><div dir="ltr"><br></div><div dir="ltr">Kind regards,</div><div dir="ltr">Rado</div><div dir="ltr"><br></div><div dir="ltr"><blockquote type="cite">On 1 Aug 2023, at 19:10, Brian S O'Neill <bronee@gmail.com> wrote:<br><br></blockquote></div><blockquote type="cite"><div dir="ltr"><span>On 2023-08-01 09:52 AM, Maurizio Cimadamore wrote:</span><br><blockquote type="cite"><span>On 01/08/2023 17:19, mail@smogura.eu wrote:</span><br></blockquote><blockquote type="cite"><blockquote type="cite"><span>* Should all memory allocations even with custom allocators be tracked - this can be a little bit error prone.</span><br></blockquote></blockquote><blockquote type="cite"><span>Custom allocators/arenas came up twice.</span><br></blockquote><blockquote type="cite"><span>I don't think custom arenas would defeat the counting - I'd expect most custom arenas (like the PoolArena used in the benchmark) to just delegate to a "true" confined arena under the hood - so, some "tracked allocation would still occur.</span><br></blockquote><span></span><br><span>I was thinking it might be useful to have a region-based arena not backed by an ordinary MemorySegment. Reporting "true" memory usage might need to be handled differently.</span><br></div></blockquote></body></html>