ruby-core@ruby-lang.org archive (unofficial mirror)
 help / color / mirror / Atom feed
From: "ioquatix (Samuel Williams)" <noreply@ruby-lang.org>
To: ruby-core@ruby-lang.org
Subject: [ruby-core:109088] [Ruby master Bug#18886] Struct aref and aset don't trigger any tracepoints.
Date: Wed, 29 Jun 2022 06:05:25 +0000 (UTC)	[thread overview]
Message-ID: <redmine.journal-98234.20220629060525.3344@ruby-lang.org> (raw)
In-Reply-To: redmine.issue-18886.20220629021928.3344@ruby-lang.org

Issue #18886 has been updated by ioquatix (Samuel Williams).


With some minor changes the best performance I could get is around 20% cost:

```
|                       |compare-ruby|built-ruby|
|:----------------------|-----------:|---------:|
|vm_struct_big_aref_hi  |    149.051M|  121.849M|
|                       |       1.22x|         -|
|vm_struct_big_aref_lo  |    147.353M|  121.167M|
|                       |       1.22x|         -|
|vm_struct_big_aset     |       7.107|     6.469|
|                       |       1.10x|         -|
|vm_struct_big_href_hi  |     27.081M|   28.269M|
|                       |           -|     1.04x|
|vm_struct_big_href_lo  |     27.642M|   27.630M|
|                       |       1.00x|         -|
|vm_struct_big_hset     |       3.106|     3.150|
|                       |           -|     1.01x|
|vm_struct_small_aref   |    139.761M|  113.250M|
|                       |       1.23x|         -|
|vm_struct_small_aset   |       7.086|     6.775|
|                       |       1.05x|         -|
|vm_struct_small_href   |     30.977M|   31.479M|
|                       |           -|     1.02x|
|vm_struct_small_hset   |     28.565M|   28.554M|
|                       |       1.00x|         -|
```

----------------------------------------
Bug #18886: Struct aref and aset don't trigger any tracepoints.
https://bugs.ruby-lang.org/issues/18886#change-98234

* Author: ioquatix (Samuel Williams)
* Status: Open
* Priority: Normal
* Assignee: ko1 (Koichi Sasada)
* Backport: 2.7: UNKNOWN, 3.0: UNKNOWN, 3.1: UNKNOWN
----------------------------------------
Given the following program, `thing.name` and `thing.shape` don't trigger `c_call` trace points (or any trace points actually).

```ruby
pp RUBY_VERSION

trace_point = TracePoint.new(:line, :call, :c_call, :a_call) do |trace|
  puts trace.event

  if trace.event == :call
    # Ruby doesn't always mark call-sites in sub-expressions, so we use this approach to compute a call site and mark it:
    if location = caller_locations(2, 1).first and path = location.path
      puts "> #{path}:#{location.lineno}:#{trace.event}"
    end
  end
  
  if path = trace.path
    puts "= #{path}:#{trace.lineno}:#{trace.event}"
  end
end

trace_point.enable

# This will trigger call trace points
class Thing
  def name
    :cat
  end
  
  def shape
    :square
  end
end
thing = Thing.new

# Thing = Struct.new(:name, :shape)
# thing = Thing.new(:cat, :rectangle)

[
  name: thing.name,
  shape: thing.shape,
]
```

## Current HEAD

```
= ../test.rb:30:line:
= ../test.rb:30:c_call:new
= ../test.rb:30:c_call:inherited
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:const_added
= ../test.rb:31:line:
= ../test.rb:31:c_call:new
= ../test.rb:31:c_call:initialize
= ../test.rb:34:line:
```

## Proposed PR

```
= ../test.rb:30:line:
= ../test.rb:30:c_call:new
= ../test.rb:30:c_call:inherited
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:singleton_method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:method_added
= ../test.rb:30:c_call:const_added
= ../test.rb:31:line:
= ../test.rb:31:c_call:new
= ../test.rb:31:c_call:initialize
= ../test.rb:34:line:
= ../test.rb:34:c_call:name
= ../test.rb:35:c_call:shape
```

The reason is the internal implementation of struct doesn't have trace point instrumentation in `vm_call_opt_struct_aset` or `vm_call_opt_struct_aref`.

Proposed fix: https://github.com/ruby/ruby/pull/6071 but this would need a review, maybe @jeremyevans0 and @ko1 can help.



-- 
https://bugs.ruby-lang.org/

  parent reply	other threads:[~2022-06-29  6:05 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-06-29  2:19 [ruby-core:109083] [Ruby master Bug#18886] Struct aref and aset don't trigger any tracepoints ioquatix (Samuel Williams)
2022-06-29  3:23 ` [ruby-core:109084] " ioquatix (Samuel Williams)
2022-06-29  6:05 ` ioquatix (Samuel Williams) [this message]
2022-09-21 20:43 ` [ruby-core:109983] " ioquatix (Samuel Williams)
2023-10-27  1:40 ` [ruby-core:115184] " jeremyevans0 (Jeremy Evans) via ruby-core
2023-11-28  4:32 ` [ruby-core:115494] " mame (Yusuke Endoh) via ruby-core
2023-11-28 18:46 ` [ruby-core:115519] " jeremyevans0 (Jeremy Evans) via ruby-core
2023-12-07  5:18 ` [ruby-core:115625] " ko1 (Koichi Sasada) via ruby-core

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-list from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

  List information: https://www.ruby-lang.org/en/community/mailing-lists/

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=redmine.journal-98234.20220629060525.3344@ruby-lang.org \
    --to=ruby-core@ruby-lang.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).