[ruby-core:116449] [Ruby master Feature#20205] Enable `frozen_string_literal` by default

["byroot (Jean Boussier) via ruby-core" <ruby-core@ml.ruby-lang.org>]

Issue #20205 has been updated by byroot (Jean Boussier).


Alright, I finally got hold of the benchmarking server. For the record it's an AWS c5n.metal, with various tuning like disabling frequency scaling etc to hopefully get more stable results.

- `yjit-bench` revision: https://github.com/Shopify/yjit-bench/commit/95e1a3caddc7281fbaf5bd0f20b197561453993f
- `ruby` revision: https://github.com/Shopify/ruby/commit/bb0cee8daba4b70cedb40b36af05d796956475d6

Both rubyes are ran with YJIT enabled. `mutable` has `MUTABLE_STRINGS=1` which makes the `frozen_string_literal: true` comment become essentially a noop.

I ran headline benchmark twice for good measures, most of them seem to be consistent with 1% between the two runs, except for `ruby-lsp` which has widely inconsistent results (`+/-12%`??), and `hexapdf` too to some extent (`+/-4%`).

```
mutable: ruby 3.4.0dev (2024-01-24T08:24:16Z mutable-strings bb0cee8dab) +YJIT dev [x86_64-linux]
frozen: ruby 3.4.0dev (2024-01-24T08:24:16Z mutable-strings bb0cee8dab) +YJIT dev [x86_64-linux]

--------------  ------------  ----------  -----------  ----------  --------------  --------------
bench           mutable (ms)  stddev (%)  frozen (ms)  stddev (%)  frozen 1st itr  mutable/frozen
activerecord    72.8          1.2         70.8         1.2         1.03            1.03          
chunky-png      1666.3        0.1         1672.7       0.2         1.00            1.00          
erubi-rails     2330.1        0.3         2312.2       0.2         1.01            1.01          
hexapdf         3960.1        8.9         3614.7       7.3         1.04            1.10          
liquid-c        107.1         1.3         101.3        1.4         0.95            1.06          
liquid-compile  100.0         2.8         98.5         3.1         1.00            1.02          
liquid-render   153.0         1.1         137.8        1.1         0.97            1.11          
lobsters        1346.6        8.5         1239.2       7.6         1.00            1.09          
mail            202.8         0.8         198.9        1.0         1.02            1.02          
psych-load      3723.2        0.1         3661.3       0.1         1.01            1.02          
railsbench      2669.8        0.1         2519.6       0.2         1.02            1.06          
rubocop         278.6         5.9         272.5        6.1         1.00            1.02          
ruby-lsp        217.5         8.7         250.7        10.4        0.98            0.87          
sequel          111.8         0.7         111.8        0.8         1.01            1.00          
--------------  ------------  ----------  -----------  ----------  --------------  --------------
Legend:
- frozen 1st itr: ratio of mutable/frozen time for the first benchmarking iteration.
- mutable/frozen: ratio of mutable/frozen time. Higher is better for frozen. Above 1 represents a speedup.



mutable: ruby 3.4.0dev (2024-01-24T08:24:16Z mutable-strings bb0cee8dab) +YJIT dev [x86_64-linux]
frozen: ruby 3.4.0dev (2024-01-24T08:24:16Z mutable-strings bb0cee8dab) +YJIT dev [x86_64-linux]

--------------  ------------  ----------  -----------  ----------  --------------  --------------
bench           mutable (ms)  stddev (%)  frozen (ms)  stddev (%)  frozen 1st itr  mutable/frozen
activerecord    72.8          1.2         70.8         1.3         1.00            1.03          
chunky-png      1664.1        0.2         1672.5       0.2         1.01            0.99          
erubi-rails     2351.9        0.2         2311.4       0.2         1.01            1.02          
hexapdf         3759.9        7.0         3696.5       5.8         1.04            1.02          
liquid-c        108.0         1.1         101.3        1.4         0.97            1.07          
liquid-compile  100.5         3.1         98.5         3.1         0.99            1.02          
liquid-render   152.8         1.0         138.0        1.1         0.97            1.11          
lobsters        1339.0        9.8         1244.2       7.9         1.00            1.08          
mail            202.6         0.7         198.9        1.0         1.00            1.02          
psych-load      3627.0        0.1         3659.6       0.0         0.99            0.99          
railsbench      2670.7        0.1         2560.3       1.7         1.02            1.04          
rubocop         279.5         6.3         271.4        6.1         1.00            1.03          
ruby-lsp        274.2         9.5         245.0        9.8         0.98            1.12          
sequel          111.9         0.7         112.0        1.3         1.01            1.00          
--------------  ------------  ----------  -----------  ----------  --------------  --------------
Legend:
- frozen 1st itr: ratio of mutable/frozen time for the first benchmarking iteration.
- mutable/frozen: ratio of mutable/frozen time. Higher is better for frozen. Above 1 represents a speedup.
```

Overall the performance benefit seem much more important than on my initial benchmark, likely in part because YJIT is enabled and also likely in part because of the different architecture (`X86_64` vs `ARM64`).

----------------------------------------
Feature #20205: Enable `frozen_string_literal` by default
https://bugs.ruby-lang.org/issues/20205#change-106473

* Author: byroot (Jean Boussier)
* Status: Open
* Priority: Normal
----------------------------------------
### Context

The `frozen_string_literal: true` pragma was introduced in Ruby 2.3, and as far as I'm aware the plan was initially to make it the default for Ruby 3.0, but this plan was abandoned because it would be too much of a breaking change without any real further notice.

According to Matz, he still wishes to enable `frozen_string_literal` by default in the future, but a reasonable migration plan is required. 

The main issue is backward compatibility, flipping the switch immediately would break a lot of code, so there must be some deprecation period.

The usual the path forward for this kind of change is to emit deprecation warnings one of multiple versions in advance.

One example of that was the Ruby 2.7 keyword argument deprecation. It was quite verbose, and some users were initially annoyed, but I think the community pulled through it and I don't seem to hear much about it anymore.

So for frozen string literals, the first step would be to start warning when a string that would be frozen in the future is mutated.

### Deprecation Warning Implementation

I implemented a quick proof of concept with @etienne in https://github.com/Shopify/ruby/pull/549

In short:

- Files with `# frozen_string_literal: true` or `# frozen_string_literal: false` don't change in behavior at all.
- Files with no `# frozen_string_literal` comment are compiled to use `putchilledstring` opcode instead of regular `putstring`.
- This opcode mark the string with a user flag, when these strings are mutated, a warning is issued.

Currently the proof of concept issue the warning at the mutation location, which in some case can make locating where the string was allocated a bit hard.

But it is possible to improve it so the message also include the location at which the literal string was allocated, and learning from the keyword argument warning experience,
we can record which warnings were already issued to avoid spamming users with duplicated warnings.

As currently implemented, there is almost no overhead. If we modify the implementation to record the literal location,
we'd incur a small memory overhead for each literal string in a file without an explicit `frozen_string_literal` pragma.

But I believe we could do it in a way that has no overhead if `Warning[:deprecated] = false`.

### Timeline

The migration would happen in 3 steps, each step can potentially last multiple releases. e.g. `R0` could be `3.4`, `R1` be `3.7` and `R2` be `4.0`.
I don't have a strong opinion on the pace.

- Release `R0`: introduce the deprecation warning (only if deprecation warnings enabled).
- Release `R1`: make the deprecation warning show up regardless of verbosity level.
- Release `R2`: make string literals frozen by default.

### Impact

Given that `rubocop` is quite popular in the community and it has enforced the usage of `# frozen_string_literal: true` for years now,
I suspect a large part of the actively maintained codebases in the wild wouldn't see any warnings.

And with recent versions of `minitest` enabling deprecation warnings by default (and [potentially RSpec too](https://github.com/rspec/rspec-core/issues/2867)),
the few that didn't migrate will likely be made compatible quickly.

The real problem of course are the less actively developed libraries and applications. For such cases, any codebase can remain compatible by setting `RUBYOPT="--disable=frozen_string_literal"`,
and so even after `R2` release. The flag would never be removed any legacy codebase can continue upgrading Ruby without changing a single line of cod by just flipping this flag.

### Workflow for library maintainers

As a library maintainer, fixing the deprecation warnings can be as simple as prepending `# frozen_string_literal: false` at the top of all their source files, and this will keep working forever.

Alternatively they can of course make their code compatible with frozen string literals.

Code that is frozen string literal compatible doesn't need to explicitly declare it. Only code that need it turned of need to do so.

### Workflow for application owners

For application owners, the workflow is the same than for libraries.

However if they depend on a gem that hasn't updated, or that they can't upgrade it, they can run their application with `RUBYOPT="--disable=frozen_string_literal"` and it will keep working forever.

Any user running into an incompatibility issue can set `RUBYOPT="--disable=frozen_string_literal"` forever, even in `4.x`, the only thing changing is the default value.

And any application for which all dependencies have been made fully frozen string literal compatible can set `RUBYOPT="--enable=frozen_string_literal"` and start immediately removing magic comment from their codebase.




-- 
https://bugs.ruby-lang.org/
 ______________________________________________
 ruby-core mailing list -- ruby-core@ml.ruby-lang.org
 To unsubscribe send an email to ruby-core-leave@ml.ruby-lang.org
 ruby-core info -- https://ml.ruby-lang.org/mailman3/postorius/lists/ruby-core.ml.ruby-lang.org/