RFR: 8272807: Permit use of memory concurrent with pretouch
Aleksey Shipilev
shade at openjdk.java.net
Fri Feb 4 11:11:05 UTC 2022
On Fri, 4 Feb 2022 10:47:18 GMT, Kim Barrett <kbarrett at openjdk.org> wrote:
> [Trying again, after a couple previous proposals.]
>
> Please review this change to os::pretouch_memory to permit use of the memory
> being pretouched concurrently with the pretouch operation. This will allow
> future changes that can make use of such overlap.
>
> Testing:
> mach5 tier1-3 with -XX:+AlwaysPreTouch.
>
> Verified no performance impact by comparing startup times with Epsilon GC and
> -XX:+AlwaysPretouch on a large enough heap that the pretouch takes a
> significant amount of time. Tested on both linux-x64 and linux-aarch64.
>
> None of these test concurrent use; that will wait for future changes that use
> the new feature.
Looks okay to me, with minor nits.
src/hotspot/share/runtime/os.cpp line 1758:
> 1756: void os::pretouch_memory(void* start, void* end, size_t page_size) {
> 1757: assert(is_power_of_2(page_size), "page size misaligned: %zu", page_size);
> 1758: assert(page_size >= sizeof(int), "page size too small: %zu", page_size);
`SIZE_FORMAT` instead of `%zu`.
src/hotspot/share/runtime/os.cpp line 1771:
> 1769: Atomic::add(reinterpret_cast<int*>(cur), 0, memory_order_relaxed);
> 1770: if (cur == last) break;
> 1771: }
Took me some time to check the loop. Suggestion:
do {
Atomic::add(reinterpret_cast<int*>(cur), 0, memory_order_relaxed);
cur += page_size;
} while (cur < last);
-------------
PR: https://git.openjdk.java.net/jdk/pull/7343
More information about the hotspot-runtime-dev
mailing list