From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Original-To: caml-list@sympa.inria.fr Delivered-To: caml-list@sympa.inria.fr Received: from mail3-relais-sop.national.inria.fr (mail3-relais-sop.national.inria.fr [192.134.164.104]) by sympa.inria.fr (Postfix) with ESMTPS id 69B087FA5F for ; Thu, 19 Jan 2017 14:41:41 +0100 (CET) Authentication-Results: mail3-smtp-sop.national.inria.fr; spf=None smtp.pra=yminsky@janestreet.com; spf=Pass smtp.mailfrom=yminsky@janestreet.com; spf=None smtp.helo=postmaster@mxout3.mail.janestreet.com Received-SPF: None (mail3-smtp-sop.national.inria.fr: no sender authenticity information available from domain of yminsky@janestreet.com) identity=pra; client-ip=38.105.200.229; receiver=mail3-smtp-sop.national.inria.fr; envelope-from="yminsky@janestreet.com"; x-sender="yminsky@janestreet.com"; x-conformance=sidf_compatible Received-SPF: Pass (mail3-smtp-sop.national.inria.fr: domain of yminsky@janestreet.com designates 38.105.200.229 as permitted sender) identity=mailfrom; client-ip=38.105.200.229; receiver=mail3-smtp-sop.national.inria.fr; envelope-from="yminsky@janestreet.com"; x-sender="yminsky@janestreet.com"; x-conformance=sidf_compatible; x-record-type="v=spf1" Received-SPF: None (mail3-smtp-sop.national.inria.fr: no sender authenticity information available from domain of postmaster@mxout3.mail.janestreet.com) identity=helo; client-ip=38.105.200.229; receiver=mail3-smtp-sop.national.inria.fr; envelope-from="yminsky@janestreet.com"; x-sender="postmaster@mxout3.mail.janestreet.com"; x-conformance=sidf_compatible IronPort-PHdr: =?us-ascii?q?9a23=3A9wVYJBGxv6+ralh9bfMxvJ1GYnF86YWxBRYc798d?= =?us-ascii?q?s5kLTJ7yp8qwAkXT6L1XgUPTWs2DsrQf2raQ7/mrBT1IyK3CmUhKSIZLWR4BhJ?= =?us-ascii?q?detC0bK+nBN3fGKuX3ZTcxBsVIWQwt1Xi6NU9IBJS2PAWK8TW94jEIBxrwKxd+?= =?us-ascii?q?KPjrFY7OlcS30P2594HObwlSijewZbx/IA+5oAnMucUbjo9vIbstxxXUpXdFZ/?= =?us-ascii?q?5Yzn5yK1KJmBb86Maw/Jp9/ClVpvks6c1OX7jkcqohVbBXAygoPG4z5M3wqBnM?= =?us-ascii?q?VhCP6WcGUmUXiRVHHQ7I5wznU5jrsyv6su192DSGPcDzULs5Vyiu47ttRRT1ky?= =?us-ascii?q?oMKSI3/3/LhcxxlKJboQyupxpjw47PfYqZMONycr7Bcd8GQGZMWMFeWTFcAoOn?= =?us-ascii?q?d4sAEfYOPfpWoYn6olsBtxq+BQ+xD+/rxTJFgnr60Ksn2OojDA7GxhQtENAAsH?= =?us-ascii?q?rUotv7N7ocX/6pw6fH1jjDc+pW1C3h5ITUbhwso/eBVq9wf8rLzkkvEhvIgVeK?= =?us-ascii?q?poz/ODOV0PkGvW+a7+pmTuKviG4moBx2rzmvw8csi4/JhpkWylHE7ih5wpw6Jd?= =?us-ascii?q?umR05gfd6kCoVfuD+GN4dsXswiRGRotT88x7Ybt5C7ey0Kx44mxx7Zc/GHco6I?= =?us-ascii?q?4gjiVOmLOzt4imhldKq/hhmo8Uigzer8WtOo31ZNqypIlMTHuHMV1xHL5MWKSe?= =?us-ascii?q?Fx8lq91TuPzQzf9P1ILVwumabFNZIsxqY8moQPvUnHBCP7m0X7gLWIekk5+OWk?= =?us-ascii?q?8fnrb7Hlq5KaKoR6kBvxMr40lcy6Gek4MhYBX2yc+emk0b3i8kr5QKhWjvA4iK?= =?us-ascii?q?XZtYrVJcUfpqKjAQ9ayIAj5AikDze6ytQXg2cILE5fdB2biIjpPknCIPH+Dfih?= =?us-ascii?q?n1ShiDlmyvPcMrH8HpnALmLPnKn9cbph8UJQ1hQ/wNRH65JREL4BIfbzWkHrtN?= =?us-ascii?q?zfCx80Kwy1w+PmCNV7zYMeWmKOD7SbMKzIt1+H+PkvI/WWZIALpjn9N+Ip6OLp?= =?us-ascii?q?jX88gVMdZ7Wm3YMLaHCkGfRrO1mWbmDpgtcFCGsKugs+TPf2iFCZSj5SZ3OyX7?= =?us-ascii?q?om6T0hCYKmC53DRoG3j7Cb0ie7BM4eWmcTL1mWEj/IfpSIVvFEPCebOMxukjUV?= =?us-ascii?q?WbWnY54o1Rap8gT9zuw0APDT/3gpvIjk0pBQ5uvIlgB6oS1xDsKG0HClTWhyk3?= =?us-ascii?q?gUXTI7wOZ0pkkrmQTL6rRxn/ENTY8b3PhOSApvcMOFwg=3D=3D?= X-IronPort-Anti-Spam-Filtered: true X-IronPort-Anti-Spam-Result: =?us-ascii?q?A0AEAwCNwYBYfeXIaSZEGhkBAQEBAQEBA?= =?us-ascii?q?QEBAQcBAQEBARQBAQEBAQEBAQEBAQcBAQEBAYMUAQEBAQF/gQkHg0qKepEQgja?= =?us-ascii?q?FTo80KoJCgzYCgXIHQhUBAQEBAQEBAQEBARIBAQkWCk2CMxuCHAEFIwQZAQEsB?= =?us-ascii?q?gUBDwsLBwYCAgkdAgIiEgEFAQoEDgYTEg2ISQMYAwstol8/ixpogWs6gwgBAQW?= =?us-ascii?q?EJQODBgEBAQEBAQEDAQEBAQEBAQEBFwgSeYVAg2WBCYJRO4EwDYJMDC6CXoctD?= =?us-ascii?q?IhzimU4hmGDGINtg3+Bd1KEPYloihuHDxQegRQ1gTkSCBUVOhQFhBMMAxyBflU?= =?us-ascii?q?BhyiCOwEBAQ?= X-IPAS-Result: =?us-ascii?q?A0AEAwCNwYBYfeXIaSZEGhkBAQEBAQEBAQEBAQcBAQEBARQ?= =?us-ascii?q?BAQEBAQEBAQEBAQcBAQEBAYMUAQEBAQF/gQkHg0qKepEQgjaFTo80KoJCgzYCg?= =?us-ascii?q?XIHQhUBAQEBAQEBAQEBARIBAQkWCk2CMxuCHAEFIwQZAQEsBgUBDwsLBwYCAgk?= =?us-ascii?q?dAgIiEgEFAQoEDgYTEg2ISQMYAwstol8/ixpogWs6gwgBAQWEJQODBgEBAQEBA?= =?us-ascii?q?QEDAQEBAQEBAQEBFwgSeYVAg2WBCYJRO4EwDYJMDC6CXoctDIhzimU4hmGDGIN?= =?us-ascii?q?tg3+Bd1KEPYloihuHDxQegRQ1gTkSCBUVOhQFhBMMAxyBflUBhyiCOwEBAQ?= X-IronPort-AV: E=Sophos;i="5.33,254,1477954800"; d="scan'208";a="209955475" Received: from mxout3.mail.janestreet.com ([38.105.200.229]) by mail3-smtp-sop.national.inria.fr with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 19 Jan 2017 14:41:39 +0100 Received: from tot-qpr-mailcore1.delacy.com ([172.27.56.68] helo=tot-qpr-mailcore1) by mxout3.mail.janestreet.com with esmtps (TLSv1.2:DHE-RSA-AES256-GCM-SHA384:256) (Exim 4.84_2) (envelope-from ) id 1cUCyA-0004uF-8s for caml-list@inria.fr; Thu, 19 Jan 2017 08:41:38 -0500 X-JS-Flow: external X-JS-Scanner-attachment: (ok) No attachments Received: by tot-qpr-mailcore1 with JS-mailcore (0.1) (envelope-from ) id BYgMIS-AAAAxU-HJ; 2017-01-19 08:41:38.230137-05:00 Received: from mail-ua0-f198.google.com ([209.85.217.198]) by mxgoog1.mail.janestreet.com with esmtps (TLSv1.2:AES128-GCM-SHA256:128) (Exim 4.84_2) (envelope-from ) id 1cUCyA-0006ZK-5D for caml-list@inria.fr; Thu, 19 Jan 2017 08:41:38 -0500 Received: by mail-ua0-f198.google.com with SMTP id d5so24746079uag.5 for ; Thu, 19 Jan 2017 05:41:38 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=janestreet.com; s=google; h=mime-version:in-reply-to:references:from:date:message-id:subject:to :cc:content-transfer-encoding; bh=vcfThnMk41Jtc6eydoIBv+5Bm0gE7n4uONUjlLQhXnQ=; b=tsxqf9/ejRTPufFIavLMju2A7woZdowuEmTttfkSYQuZGapGZAH/q7fCmluHjaj99A 2bIQ1iCtUR9ctq7pHkLzbg+aHRob/gTeoHSpivZbaIGvwH9jJ/mbz2kBgy3NI/QicPl1 m/rRzKOa2bm6E96o6aJAMZWO6g1XfMQ9ftR4M= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:in-reply-to:references:from:date :message-id:subject:to:cc:content-transfer-encoding; bh=vcfThnMk41Jtc6eydoIBv+5Bm0gE7n4uONUjlLQhXnQ=; b=uBM4jaapk1k3pcugJuv5XBLmqBiNj3uymVJSpJDXhtLVMpdqn8yZ7faGLiRX3VfJmy zjAlmF+vBunDhCP/0OdjsaLl8Pl0TcszFm6nR1Ji58zUz1m7pDrh1/+rJkwrDX2dkJII d+DB+yO5Ta4j3acXdYgnVrhYTg/nzl0YlsJNSNZBHVRzxYGbcxrS3So8JLx/V/0QV+Xx gX2tHedw6N+MkxEPqAZ0Dy4hSiGDSBibiUlrTmwPG1jYgVQeVboJ0ThSY6bQtczhsXvm bIRffL9HhoaMlZP3Ju2oiqcPx9FJ5s701Sb7vrv9U1I9hGzHnpPOw5tfDaRaH5ETpvqe 9jAQ== X-Gm-Message-State: AIkVDXL/gNDDmfFjtJyUNXckQlXB8XSUTFHWirTB5I9RjmVEo3deS27UGhIt966DM5FuuSgnID+se7U808EhTMUQM1Og2KQfJI3yJyDJjzCOksvwVOzxC880WsYs5aJuUhmPpBtuSG7FBk5zq28L X-Received: by 10.31.130.207 with SMTP id e198mr3534177vkd.149.1484833297599; Thu, 19 Jan 2017 05:41:37 -0800 (PST) X-Received: by 10.31.130.207 with SMTP id e198mr3534156vkd.149.1484833297162; Thu, 19 Jan 2017 05:41:37 -0800 (PST) MIME-Version: 1.0 Received: by 10.31.229.6 with HTTP; Thu, 19 Jan 2017 05:41:16 -0800 (PST) In-Reply-To: <2B595CCC-1121-4C8C-8F5F-A235D3AB19BB@yahoo.com> References: <2B595CCC-1121-4C8C-8F5F-A235D3AB19BB@yahoo.com> From:Yaron Minsky Date: Thu, 19 Jan 2017 08:41:16 -0500 Message-ID: To:Mark Hayden Cc:"caml-list@inria.fr" Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable X-JS-Processed-by: mailcore Subject: Re: [Caml-list] Ocaml optimizer pitfalls & work-arounds It seems like your primary issues are around lack of specialization around two features: - unboxing in float arrays - optimization of ad-hoc operations (e.g., polymorphic compare) My view on this is that it's best not to rely on float array specialization at all, and I think the best improvement we can make to OCaml is to remove the ad-hoc specialization of float arrays, and instead add a separate, specialized (and unboxed) type for arrays of floats, similar to the Bytes.t type which is effectively a specialized byte array. As you've observed, specialization is brittle, and it's best not to rely on it too much. Beyond that, the existence of float arrays complicate the runtime quite a bit, and make other bugs more likely. There isn't yet consensus that specialization of float array should be removed, but I'm still hopeful that we'll get there. It's probably Jane Street's highest priority ask for the compiler. We also avoid use of polymorphic compare and other ad-hoc operations, preferring to use type-specialized comparators everywhere. This is better for semantic as well as performance reasons, since we've seen a lot of subtle bugs from polymorphic compare doing the wrong thing on specific types. It's hard to deny that using type-specialized comparators is more verbose than polymorphic compare, but hopefully modular implicits will make this problem go away, and we can get the best of both worlds. And we hope that Flambda will be up to the job of inlining away the overhead of the more indirect calling conventions imposed by modular implicits. I think that with the above changes, we can probably get pretty far towards the goal of being able to write OCaml code that is both highly performant and pretty. Being able to delay specialization until later in the compilation pipeline would help more, but I believe we can do pretty well without it. y On Thu, Jan 19, 2017 at 1:51 AM, Mark Hayden wrote: > We recently upgraded our Ocaml toolchain from 4.02.3 to Ocaml 4.04.0. > We were looking forward to a performance boost from the optimization > improvements, especially from flambda. While we generally were able > to achieve significant performance improvement, we were somewhat > surprised by the effort required to avoid certain pitfalls in Ocaml. > > This note describes some issues we ran into. We filed several > reports on Ocaml Mantis regarding our findings. However it appears > the underlying issues we ran into are unlikely to change: > > Your three reports (0007440, 0007441, 0007442) are manifestations > of the same fact: the OCaml compiler performs type-based > optimizations first, then erases types, then performs all the other > optimizations. This is very unlikely to change in the near future, > as it would require a total rewrite of the compiler. > > [X Leroy, https://caml.inria.fr/mantis/view.php?id=3D7440] > > I encourage readers to review the problem reports we submitted, which > include more concrete examples. I'm posting this note in case there > are others running into similar performance issues with their Ocaml > software and who might find it helpful in working around those > issues. I'm not aware of them being documented elsewhere and there > appears to be little prospect of the issues being addressed in the > compiler in the forseeable future. Please chime in if any of this is > inaccurate or there is something I missed. > > As an initial example, consider the following Ocaml code to find the > maximum floating point value in an array (that is at least 0.0): > > [Array.fold_left max 0.0 arr] > > Now compile this with the latest compiler and maximum optimization. > Because of how the Ocaml optimization works, this will run about > 10-15x slower (and allocate 2-3 words per array element) than a more > carefully written version that uses specialized operations and avoids > allocation. See below for one way to achieve this (while still using > a functional-programming style). > > (* Same as Array.fold_left, but with type casting. > *) > let [@inline] array_fold_leftf f (x:float) (a:float array) =3D > let r =3D ref x in > for i =3D 0 to Array.length a - 1 do > r :=3D f !r (Array.unsafe_get a i) > done; > !r > ;; > > let [@inline] float_max (v0:float) v1 =3D > if v0 > v1 then v0 else v1 > ;; > > let array_float_max a =3D > array_fold_leftf float_max 0.0 a > ;; > > The assembly for the "inner loop" for the two examples are below. > They were compiled with Ocaml 4.05.dev+flambda, "-O3 > -unbox-closures", MacOS 12.2, AMD64. > > Unoptimized example. Note test/branch for array tag. Allocation for > boxing (we did not include the calls to trigger a minor gc). There > is a call to Ocaml runtime for polymorphic greater-equal. This is > probably not what one would expect from an optimizing/inline compiler > for a simple case such as this. Note that to create this we used our > own definition of Array.fold_left which had an "[@inline]" > annotation. > > L215: > movq (%rsp), %rbx > .loc 1 38 14 > movzbq -8(%rbx), %rax > cmpq $254, %rax > je L219 > .loc 1 38 14 > movq -4(%rbx,%rdx,4), %rsi > movq %rsi, 24(%rsp) > jmp L218 > .align 2 > L219: > .loc 1 38 14 > L221: > subq $16, %r15 > movq _caml_young_limit@GOTPCREL(%rip), %rax > cmpq (%rax), %r15 > jb L222 > leaq 8(%r15), %rsi > movq $1277, -8(%rsi) > .loc 1 38 14 > movsd -4(%rbx,%rdx,4), %xmm0 > movsd %xmm0, (%rsi) > movq %rsi, 24(%rsp) > L218: > movq %rdi, 32(%rsp) > .file 5 "pervasives.ml" > .loc 5 65 17 > movq _caml_greaterequal@GOTPCREL(%rip), %rax > call _caml_c_call > L213: > movq _caml_young_ptr@GOTPCREL(%rip), %r11 > movq (%r11), %r15 > cmpq $1, %rax > je L217 > movq 32(%rsp), %rdi > jmp L216 > .align 2 > L217: > movq 24(%rsp), %rdi > L216: > movq 8(%rsp), %rdx > movq %rdx, %rax > addq $2, %rdx > movq %rdx, 8(%rsp) > movq 16(%rsp), %rbx > cmpq %rbx, %rax > jne L215 > > > The assembly for the more carefully writting case is below. No > allocation. No call to external C code. No test/branch for array > tag. This matches what I think most people would like to see. It is > compact enough that (maybe) it would benefit from unrolling: > > l225: > .loc 1 46 14 > movsd -4(%rax,%rdi,4), %xmm1 > comisd %xmm1, %xmm0 > jbe l227 > jmp l226 > .align 2 > l227: > movapd %xmm1, %xmm0 > l226: > movq %rdi, %rsi > addq $2, %rdi > cmpq %rbx, %rsi > jne l225 > > > The two main learnings we found were: > > * Polymorphic primitives ([Array.get], [compare], [>=3D], [min]) are > only specialized if they appear in a context where the types can be > determined at their exact call site, otherwise a polymorphic > version is used. If the use of the primitive is later inlined in a > context where the type is no longer polymorphic, the function will > _not_ be specialized by the compiler. > > * Use of abstract data types prevents specialization. In particular, > defining an abstract data type in a module ("type t ;;") will > prevent specialization (even after inlining) for any polymorphic > primitives (eg, "caml_equal") used with that type. For instance, > if the underlying type for [t] is actually [int], other modules > will still use polymorphic equality instead of a single machine > instruction. You can prevent this behavior with the "private" > keyword in order to export the type information, "type t =3D private > int". Alternatively, the module can include its own specialized > operations and other modules can be careful to use them. > > It bears emphasizing that the issues described in this note apply > even when all of the code is "fully inlined" and uses highest level > of optimization. Specialization in the Ocaml compiler occurs in a > stage prior to inlining. If it hasn=E2=80=99t happened before inlining, = it > won=E2=80=99t happen afterwards. > > What kind of effect does lack of specialization have on performance? > Calling the "caml_compare" Ocaml C runtime function to compare > integers can be 10-20x times slower than using a single integer > comparison machine instruction. Ditto for floating point values. > The unspecialized [Array.get], [Array.set], and (on 32-bit) > [Array.length] have to check the tag on the array to determine if the > array uses the unboxed floating-point represntation (I wish Ocaml > didn't use this!). For instance, the polymorphic [Array.get] checks > the tag on the array and (for floating point arrays) reads the value > and boxes the floating point value (ie, allocate 2-3 words on the > heap). Note that when iterating over an array, the check on the tag > will be included in _each_ loop iteration. > > Other impacts of using non-specialized functions: > > * Use of polymorphic primitives means floating point values have to > be boxed, requiring heap allocation. Through use of specialized > specialized primitives, in many cases floats can remain unboxed. > > * All the extra native code from using polymorphic primitives > (checking array tags, conditionally boxing floats, calling out to > Ocaml C runtime) can have follow-on effects for further inlining. > In other words, when native code can be kept compact, then more > code can be inlined and/or loops unrolled and this can in turn > allow further optimization. > > Some suggestions others may find helpful: > > * Consider using the "private" keyword for any abstract types in your > modules. We added over 50 of these to our code base. It is an > ugly but effective work-around. > > * The min/max functions in standard library Pervasives are > particularly problematic. They are polymorphic so their comparison > will never be specialized. It can be helpful to define specialized > functions such as: > > let [@inline] float_max (v0:float) (v1:float) =3D > if v0 > v1 then v0 else v1 > ;; > > let [@inline] int_max (v0:int) (v1:int) =3D > if v0 > v1 then v0 else v1 > ;; > > These will be compiled to use native machine code and unboxed > values. > > * Any use of polymorphism can negatively affect performance. Be > careful about inadvertently introducing polymorphism into your > program, such as this helper function: > > let [@inline] getter v ofs =3D Array.get v ofs ;; > > This will result in unspecialized version of [Array.get] being > inlined at all call-sites. Note that if your .mli file defines the > function as non-polymorphic that will still _not_ affect how > [getter] is compiled.: > > type getter : t -> int -> int ;; (* does not affect optimization *) > > You must cast the type in the implementation in order for [Array.get] > to be specialized: > > let [@inline] getter (v:int array) ofs =3D Array.get v ofs ;; > > * All the iterators in the Array module (eg, [Array.iter]) in the > standard library are polymorphic, so will use unspecialized > accessors and be affected by the issues described here. Using the > following to sum and array of floats may seem elegant: > > Array.fold_left (+) 0.0 arr > > However, the resulting code is much slower (and allocates 2 floats > per array entry, ie 4-6 words) than a "specialized" version. Note > that even if [Array.fold_left] and surrounding code were "fully > inlined," it is still a polymorphic function so the above > performance penalty for checking the array tag and boxing the float > is present. See also the earlier example. > > * It can be helpful to review the compiled assembly code (using "-S" > option for ocamlopt) and look for tell-tale signs of lack of > specialization, such as calls to [_caml_greaterequal] or allocation > [caml_call_gc] in cases where they are not expected. You can refer > to the assembly code for the original implementation, know that > that code will not be specialized when inlined. > > As I said, by examining our hot spots and following the suggestions > above, we found the resulting native code could be comparable to what > we would expect from C. It is unfortunate these issues were > (apparently) designed into the Ocaml compiler architecture, because > otherwise it would have seemed this would be a natural area of > improvement for the compiler. I would have thought a staticly typed > language such as Ocaml would (through its type checker) be > well-suited for the simple types of function specialization described > in this note. > > > -- > Caml-list mailing list. Subscription management and archives: > https://sympa.inria.fr/sympa/arc/caml-list > Beginner's list: http://groups.yahoo.com/group/ocaml_beginners > Bug reports: http://caml.inria.fr/bin/caml-bugs