[#86787] [Ruby trunk Feature#14723] [WIP] sleepy GC — ko1@...

Issue #14723 has been updated by ko1 (Koichi Sasada).

13 messages 2018/05/01
[#86790] Re: [Ruby trunk Feature#14723] [WIP] sleepy GC — Eric Wong <normalperson@...> 2018/05/01

ko1@atdot.net wrote:

[#87095] [Ruby trunk Feature#14767] [PATCH] gc.c: use monotonic counters for objspace_malloc_increase — ko1@...

Issue #14767 has been updated by ko1 (Koichi Sasada).

9 messages 2018/05/17

[ruby-core:86862] Re: [Ruby trunk Feature#14723] [WIP] sleepy GC

From: Eric Wong <normalperson@...>
Date: 2018-05-03 07:07:25 UTC
List: ruby-core #86862
sam.saffron@gmail.com wrote:
> perf change seems a tiny bit more noticable. 

Thanks for benchmarking!  Disappointing results, though.

Is this is with my latest updates up thread with do_select
and gc_*_continue functions?

Can you try #define RUBY_GC_SLEEPY_MARK 0 in gc.h to disable
incremental marking on sleep?

I wonder if incremental marking is causing too many objects
to be marked when it is triggered deep in the stack.

Marking is best done when the stack is shallow (where unicorn
calls IO.select), but could be harmful when the stack is deep
(where Pg calls rb_thread_fd_select).

Unsubscribe: <mailto:ruby-core-request@ruby-lang.org?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>

In This Thread