Stream Fusion, to Completeness

Presentation of Stream Fusion, to Completeness at POPL 2017 in Paris.

for more information about the project follow: http://strymonas.github.io

Transcript

1. Stream Fusion, to Completeness Oleg Kiselyov Aggelos Biboudis Nick Palladinos Yannis Smaragdakis 18/1/2017 POPL 2017 Paris University of Athens Nessos IT Tohoku University University of Athens
2. Stream Fusion, to Completeness Design a library for fast streams … • stream of elements, functionally • no-storage, lazy, finite/infinite, one-shot => bulk … that supports a wide range and complex combinations of operators … … and generates loop-based, fused code with zero allocations. 2 `ZHCUVGT
3. Staging Stream Fusion 3 UVCIKPI also our main example for this talk
4. Staging Stream Fusion 4 CPFOWEJOQTGEQORNGZ
5. Guaranteed Performance ✓ no intermediate results, no buffers ✓ no closure creation ✓ function calls should get inlined ✓ no deconstructions and constructions of tuples at 
 run-time 5
6. Benchmarks 6 OCaml/BER MetaOCaml
7. Benchmarks 7 Scala/LMS
8. Multi-Stage Programming • think of code templates • brackets to create well-{formed, scoped, typed} templates 
 let c = .< 1 + 2 >. • create holes 
 let cf x = .< .~x + .~x >. • synthesize code 
 cf c ~> .< (1 + 2) + (1 + 2) >. 8
9. Step 0: Naive Staging • start from an F-co-algebras signature (an Unfold) • sprinkle the code with staging annotations 9 type α stream = ∃σ. σ code * (σ code ! (α,σ) stream_shape code) type ('a,'z) stream_shape = | Nil | Cons of 'a * 'z binding time analysis
10. let map : ('a code -> 'b code) -> 'a stream -> 'b stream = fun f (s,step) -> let new_step = fun s -> .< match .~(step s) with | Nil -> Nil | Cons (a,t) -> Cons (.~(f .<a>.), t)>. in (s,new_step);; 10 Step 0: Naive Staging
11. Result (step 0) let rec loop_1 z_2 s_3 = match match match s_3 with | (i_4, arr_5) -> if i_4 < (Array.length arr_5) then Cons ((arr_5.(i_4)),((i_4 + 1), arr_5)) else Nil with | Nil -> Nil | Cons (a_6,t_7) -> Cons ((a_6 * a_6), t_7) with | Nil -> z_2 | Cons (a_8,t_9) -> loop_1 (z_2 + a_8) t_9 of_arr map sum 11 PQKPVGTOGFKCVG✓ HWPEVKQPKPNKPKPI✓ XCTKQWUQXGTJGCFU✗ ✗ ✗
12. Step 1: fusing the stepper let map : ('a code -> 'b code) -> 'a st_stream -> 'b st_stream = fun f (s, step) -> let new_step s k = step s @@ function | Nil -> k Nil | Cons (a,t) -> .<let a' = .~(f a) in .~(k @@ Cons (.<a'>., t))>. in (s, new_step) ;; 12 stream_shape is static and factored out of the dynamic code * Anders Bondorf. 1992. Improving binding times without explicit CPS-conversion. In LFP ’92 * Oleg Kiselyov, Why a program in CPS specializes better, http://okmij.org/ftp/meta-programming/#bti • stepper has known structure though!
13. Result let rec loop_1 z_2 s_3 = match s_3 with | (i_4, arr_5) -> if i_4 < (Array.length arr_5) then let el_6 = arr_5.(i_4) in let a'_7 = el_6 * el_6 in loop_1 (z_2 + a'_7) ((i_4 + 1), arr_5) else z_2 13 UVGRRGTKPNKPGF✓ RCVVGTPOCVEJKPI✗ ✗
14. Step 2: fusing the state let of_arr : 'a array code -> 'a st_stream = let init arr k = .< let i = ref 0 and arr = .~arr in .~(k (.<i>.,.<arr>.))>. and step (i,arr) k = .< if !(.~i) < Array.length .~arr then let el = (.~arr).(!(.~i)) in incr .~i; .~(k @@ Cons (.<el>., ())) else .~(k Nil)>. in fun arr -> (init arr,step) (int * α array) code ~> int ref code * α array code 14 • no pair-allocation in loop: state passed in and mutated
15. Result let i_8 = ref 0 and arr_9 = [|0;1;2;3;4|] in let rec loop_10 z_11 = if ! i_8 < Array.length arr_9 then let el_12 = arr_9.(! i_8) in incr i_8; let a'_13 = el_12 * el_12 in loop_10 (z_11+a'_13) else z_11 15 PQRCVVGTPOCVEJKPI✓ TGEWTUKQP✗ ✗ ✗
16. Step 3: generating imperative loops let of_arr : 'a array code -> 'a stream = fun arr -> let init k = .<let arr = .~arr in .~(k .<arr>.)>. and upper_bound arr = .<Array.length .~arr - 1>. and index arr i k = .<let el = (.~arr).(.~i) in .~(k .<el>.)>. in (init, For {upb;index}) 16 start with For-form and if needed transform to Unfold
17. Result let s_1 = ref 0 in let arr_2 = [|0;1;2;3;4|] in for i_3 = 0 to (Array.length arr_2) - 1 do let el_4 = arr_2.(i_3) in let t_5 = el_4 * el_4 in s_1 := !s_1 + t_5 done; !s_1 17 NQQRDCUGFHWUGF✓
18. type card_t = AtMost1 | Many type (α,σ) producer_t = | For of {upb: σ ! int code; index: σ ! int code ! (α ! unit code) ! unit code} | Unfold of {term: σ ! bool code; card: card_t; step: σ ! (α ! unit code) ! unit code} and α producer = ∃σ. (∀ω. (σ ! ω code) ! ω code) * (α,σ) producer_t and α st_stream = | Linear of α producer | Nested of ∃β. β producer * (β ! α st_stream) and α stream = α code st_stream Final Datatype 18 • Linearity (filter and flat_map) • Sub-ranging and infinite streams 
 (take and unfold) • Fusing parallel streams (zip)
19. The takeaways • stream-fusion is a domain-specific optimization • domain-specific optimizations are better tackled outside the general purpose compiler • multi-stage programming is not a trivial sprinkling of staging annotations 19
20. Thanks! 20 strymonas.github.io