[ruby-core:80514] [Ruby trunk Feature#12589] VM performance improvement proposal

Issue #12589 has been updated by vmakarov (Vladimir Makarov).


magaudet (Matthew Gaudet) wrote:

> >  You can not implement JIT without consuming additional memory.  May be 
> >  for some MRI environments like Heroku the additional memory consumption 
> >  can be critical. And for such environment it might be better not to use 
> >  JIT at all.  Still there are other Ruby environments where people can 
> >  spare memory consumption for faster code.
> 
> Indeed. I spoke at Ruby Kaigi 2016 [1] trying very hard to encouraging thinking 
> about exactly what it is that 3x3 should accomplish, and how to measure. As 
> I am sure you are aware, the selection of benchmark and benchmarking methodology
> is key to making sure you actually achieve your aims. 
> 
> [1]: http://rubykaigi.org/2016/presentations/MattStudies.html 

By the way, I did some memory consumption measurements using size of
max (peak) resident area for a small Ruby program (about 15 lines) and
its sub-processes on free x86-64 machine with 32GB memory using
`(j)ruby --disable-gems`. Here are the numbers:

```
Ruby trunk:                                 6.4MB
RTL:                                        6.5MB
RTL+GCC JIT:                               26.9MB
RTL+LLVM JIT:                              52.1MB
OMR:                                        6.5MB
OMR+JIT:                                   18.0MB
jruby:                                    244.5MB
Graal:                                    771.0MB
```

It can give a rough idea what JIT memory consumption costs are.

The numbers should be taken with a grain of salt.  It includes
all code size too.  As I wrote multiple running program copies
share the code.  And in case of GCC (cc1) it is about 20MB (so
in a case of 20 running GCC on a server, the average size of max
resident area could be about 7.9MB).

I have no idea what is the code size of jruby and Graal as they
use sub-processes and I know nothing about them.



----------------------------------------
Feature #12589: VM performance improvement proposal
https://bugs.ruby-lang.org/issues/12589#change-64017

* Author: vmakarov (Vladimir Makarov)
* Status: Open
* Priority: Normal
* Assignee: 
* Target version: 
----------------------------------------
  Hello.  I'd like to start a big MRI project but I don't want to
disrupt somebody else plans.  Therefore I'd like to have MRI
developer's opinion on the proposed project or information if somebody
is already working on an analogous project.

  Basically I want to improve overall MRI VM performance:

  * First of all, I'd like to change VM insns and move from
    **stack-based** insns to **register transfer** ones.  The idea behind
    it is to decrease VM dispatch overhead as approximately 2 times
    less RTL insns are necessary than stack based insns for the same
    program (for Ruby it is probably even less as a typical Ruby program
    contains a lot of method calls and the arguments are passed through
    the stack).

    But *decreasing memory traffic* is even more important advantage
    of RTL insns as an RTL insn can address temporaries (stack) and
    local variables in any combination.  So there is no necessity to
    put an insn result on the stack and then move it to a local
    variable or put variable value on the stack and then use it as an
    insn operand.  Insns doing more also provide a bigger scope for C
    compiler optimizations.

    The biggest changes will be in files compile.c and insns.def (they
    will be basically rewritten).  **So the project is not a new VM
    machine.  MRI VM is much more than these 2 files.**

    The disadvantage of RTL insns is a bigger insn memory footprint
    (which can be upto 30% more) although as I wrote there are fewer
    number of RTL insns.

    Another disadvantage of RTL insns *specifically* for Ruby is that
    insns for call sequences will be basically the same stack based
    ones but only bigger as they address the stack explicitly.

  * Secondly, I'd like to **combine some frequent insn sequences** into
    bigger insns.  Again it decreases insn dispatch overhead and
    memory traffic even more.  Also it permits to remove some type
    checking.

    The first thing on my mind is a sequence of a compare insn and a
    branch and using immediate operands besides temporary (stack) and
    local variables.  Also it is not a trivial task for Ruby as the
    compare can be implemented as a method.

  I already did some experiments.  RTL insns & combining insns permits
to speed the following micro-benchmark in more 2 times:

```
i = 0
while i<30_000_000 # benchmark loop 1
  i += 1
end
```

The generated RTL insns for the benchmark are

```
== disasm: #<ISeq:<main>@while.rb>======================================
== catch table
| catch type: break  st: 0007 ed: 0020 sp: 0000 cont: 0020
| catch type: next   st: 0007 ed: 0020 sp: 0000 cont: 0005
| catch type: redo   st: 0007 ed: 0020 sp: 0000 cont: 0007
|------------------------------------------------------------------------
local table (size: 2, temp: 1, argc: 0 [opts: 0, rest: -1, post: 0, block: -1, kw: -1@-1, kwrest: -1])
[ 2] i
0000 set_local_val    2, 0                                            (   1)
0003 jump             13                                              (   2)
0005 jump             13
0007 plusi            <callcache>, 2, 2, 1, -1                        (   3)
0013 btlti            7, <callcache>, -1, 2, 30000000, -1             (   2)
0020 local_ret        2, 0                                            (   3)
```

In this experiment I ignored trace insns (that is another story) and a
complication that a integer compare insn can be re-implemented as a
Ruby method.  Insn bflti is combination of LT immediate compare and
branch true.

A modification of fib benchmark is sped up in 1.35 times:

```
def fib_m n
  if n < 1
    1
  else
    fib_m(n-1) * fib_m(n-2)
  end
end

fib_m(40)
```

The RTL code of fib_m looks like

```
== disasm: #<ISeq:[e-mail]>==========================================
local table (size: 2, temp: 3, argc: 1 [opts: 0, rest: -1, post: 0, block: -1, kw: -1@-1, kwrest: -1])
[ 2] n<Arg>
0000 bflti            10, <callcache>, -1, 2, 1, -1                   (   2)
0007 val_ret          1, 16
0010 minusi           <callcache>, -2, 2, 1, -2                       (   5)
0016 simple_call_self <callinfo!mid:fib_m, argc:1, FCALL|ARGS_SIMPLE>, <callcache>, -1
0020 minusi           <callcache>, -3, 2, 2, -3
0026 simple_call_self <callinfo!mid:fib_m, argc:1, FCALL|ARGS_SIMPLE>, <callcache>, -2
0030 mult             <callcache>, -1, -1, -2, -1
0036 temp_ret         -1, 16
```

In reality, the improvement of most programs probably will be about
10%.  That is because of very dynamic nature of Ruby (a lot of calls,
checks for redefinition of basic type operations, checking overflows
to switch to GMP numbers).  For example, integer addition can not be
less than about x86-64 17 insns out of the current 50 insns on the
fast path.  So even if you make the rest (33) insns 2 times faster,
the improvement will be only 30%.

A very important part of MRI performance improvement is to make calls
fast because there are a lot of them in Ruby but as I read in some
Koichi Sasada's presentations he pays a lot of attention to it.  So I
don't want to touch it.

  * Thirdly.  I want to implement the insns as small inline functions
    for future AOT compiler, of course, if the projects described
    above are successful.  It will permit easy AOT generation of C code
    which will be basically calls of the functions.

    I'd like to implement AOT compiler which will generate a Ruby
    method code, call a C compiler to generate a binary shared code
    and load it into MRI for subsequent calls.  The key is to minimize
    the compilation time.  There are many approaches to do it but I
    don't want to discuss it right now.

    C generation is easy and most portable implementation of AOT but
    in future it is possible to use GCC JIT plugin or LLVM IR to
    decrease overhead of C scanner/parser.

    C compiler will see a bigger scope (all method insns) to do
    optimizations.  I think using AOT can give another 10%
    improvement.  It is not that big again because of dynamic nature
    of Ruby and any C compiler is not smart enough to figure out
    aliasing for typical generated C program.

    The life with the performance point of view would be easy if Ruby
    did not permit to redefine basic operations for basic types,
    e.g. plus for integer.  In this case we could evaluate types of
    operands and results using some data flow analysis and generate
    faster specialized insns.  Still a gradual typing if it is
    introduced in future versions of Ruby would help to generate such
    faster insns.

  Again I wrote this proposal for discussion as I don't want to be in
a position to compete with somebody else ongoing big project.  It
might be counterproductive for MRI development.  Especially I don't
want it because the project is big and long and probably will have a
lot of tehcnical obstacles and have a possibilty to be a failure.




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

Unsubscribe: <mailto:[e-mail]?subject=unsubscribe>
<http://lists.ruby-lang.org/cgi-bin/mailman/options/ruby-core>