From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id A8F75F3380C for ; Tue, 17 Mar 2026 08:33:38 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id EC0616B0089; Tue, 17 Mar 2026 04:33:37 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id E6FC26B008A; Tue, 17 Mar 2026 04:33:37 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id D851C6B008C; Tue, 17 Mar 2026 04:33:37 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0016.hostedemail.com [216.40.44.16]) by kanga.kvack.org (Postfix) with ESMTP id C29C06B0089 for ; Tue, 17 Mar 2026 04:33:37 -0400 (EDT) Received: from smtpin15.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay04.hostedemail.com (Postfix) with ESMTP id 806B31A05A2 for ; Tue, 17 Mar 2026 08:33:37 +0000 (UTC) X-FDA: 84554891274.15.385E5A0 Received: from va-1-113.ptr.blmpb.com (va-1-113.ptr.blmpb.com [209.127.230.113]) by imf28.hostedemail.com (Postfix) with ESMTP id 88B50C0005 for ; Tue, 17 Mar 2026 08:33:34 +0000 (UTC) Authentication-Results: imf28.hostedemail.com; dkim=pass header.d=bytedance.com header.s=2212171451 header.b=UNdSsuft; dmarc=pass (policy=quarantine) header.from=bytedance.com; spf=pass (imf28.hostedemail.com: domain of changfengnan@bytedance.com designates 209.127.230.113 as permitted sender) smtp.mailfrom=changfengnan@bytedance.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1773736416; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:dkim-signature; bh=ouY2DlmxHP8dSdFOfc8RggJly7xAXy6YWI/q1/g6vaI=; b=h8rkSk8nvoeiCecH6lxmLUPbg3gi96MXFT3k1QCnASKip9KyBVQ83I7T87+o483aRHilv9 S2o2x5lL9cqwa/Hqr+KOpyC51OwJMx1GJSGItPXnmwnzKLf/puJEC8fJ9b6fPIqm64BSZn +rvvOZ+NznJ49Whg3UxkmIpAQADujsE= ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1773736416; a=rsa-sha256; cv=none; b=KM6pfb+7kEujcqOMakSV/KCkctRbEslXk5fpq16Ys60K1iXD8uAzS3/5Iqp8BOeTP01T8Q lX33AXyyroXw4QLV4GX8qDWllsiY28kkTJXkIO2Vih2jjtmR3Oxim669YKCT1osdXockU2 sTZfgpvUicXQCGOZG50xpKk3x6l7Fmk= ARC-Authentication-Results: i=1; imf28.hostedemail.com; dkim=pass header.d=bytedance.com header.s=2212171451 header.b=UNdSsuft; dmarc=pass (policy=quarantine) header.from=bytedance.com; spf=pass (imf28.hostedemail.com: domain of changfengnan@bytedance.com designates 209.127.230.113 as permitted sender) smtp.mailfrom=changfengnan@bytedance.com DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; s=2212171451; d=bytedance.com; t=1773736409; h=from:subject: mime-version:from:date:message-id:subject:to:cc:reply-to:content-type: mime-version:in-reply-to:message-id; bh=ouY2DlmxHP8dSdFOfc8RggJly7xAXy6YWI/q1/g6vaI=; b=UNdSsuftD9G9ZDe1RiM5S1tQwhHyI8KCu3XHAXVRMcK1kKxpaCne5ibOQUoSEReqUIhzQ5 Lp1OWbn/Z+Z8nzj6xNDZiqM67IwG4KZCT4DiV9lrD05kkjdhhtmilgxCb1Tq5AeVfGm+H0 Gaw7fp5Hl+JaqFEQi/EGON2btc2ReFB2mint/9+DLNyo+1XOPxkpG1Zx0Hjm+nrtRN3AqU efgVhGd2l+kTqdscuzRWRKO+kACt1NPdWT9/lzeOZMd8khZgZLgxaLHTo+rs/gB1IGUpD3 K20DbMYJPb2Coe8+FYtS/UZncgvud4L9r9tvqygFx9QuUlh+jRjXC5kras6JUw== Date: Tue, 17 Mar 2026 16:33:24 +0800 Message-Id: To: "Vlastimil Babka (SUSE)" Subject: Re: [PATCH v3] iomap: add allocation cache for iomap_dio References: <20260115021108.1913695-1-guzebing1612@gmail.com> In-Reply-To: X-Lms-Return-Path: Cc: "Dave Chinner" , "Harry Yoo" , "Hao Li" , "guzebing" , , , , , , , , , From: "changfengnan" Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable X-Rspamd-Server: rspam02 X-Rspamd-Queue-Id: 88B50C0005 X-Stat-Signature: dg5qer6poizukr7r7u6p5mju7nybszj8 X-Rspam-User: X-HE-Tag: 1773736414-161254 X-HE-Meta: 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 8OEJgRE2 H+zpe Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: List-Subscribe: List-Unsubscribe: > From: "Vlastimil Babka (SUSE)" > Date:=C2=A0 Tue, Mar 17, 2026, 16:28 > Subject:=C2=A0 Re: [PATCH v3] iomap: add allocation cache for iomap_dio > To: "changfengnan" > Cc: "Dave Chinner", "Harry Yoo", "Hao Li", "guzebing", , , , , , , = , , > On 3/17/26 08:28, changfengnan wrote: > >=C2=A0 > >> That suggests in that test you used larger capacity than the automatic= ally > >> calculated. > > The 10% improvement is due to the every cache has sheaves. > > When I tested 256-byte objects, default sheaf_capacity is 26, allocatin= g and > > freeing 32 objects did not show a noticeable difference, but allocating= and > > freeing 128 objects resulted in a significant improvement, about 3-4x i= n a=C2=A0 > > multithreaded environment.=C2=A0 about 12% improvement in single thread= . >=C2=A0 > Great! >=C2=A0 > >> =C2=A0 > >> > I'm thinking that maybe these improvements may not be significant en= ough to > >> > see the effect in the io flow. > >> > Using a simple list seems to be the most efficient approach. > >>=C2=A0 > >> I think the question is, what improvement do you now see with your add= ed > >> pcpu cache vs kmalloc() when 7.0-rc4 is used as the baseline? > >=C2=A0 > > On 7.0-rc4, pcpu get 1.20M IOPS , kmalloc get 1.19M IOPS, new cache wit= h set sheaf_capacity 256, 1.19M IOPS > > On 6.19, pcpu get 1.20M IOPS,=C2=A0 kmalloc get 1.17M IOPS, new cache w= ith set sheaf_capacity 256, 1.19M IOPS. >=C2=A0 > Thanks a lot for that data. My conclusion is that kmalloc before sheaves = did > indeed worse and custom pcpu cache improved it relatively more. Kmalloc w= ith > sheaves does better, and the improvement of custom pcpu cache is smaller. > Also the default sheaf capacity seems to be enough for this workload. Agree. >=C2=A0 > IO is not my area but getting from 1.19M to 1.20M doesn't look like it's > worth the custom code? (possibly from 1.17M to 1.20M it also wasn't). Yes, at least for now, there=E2=80=99s no need for a per-CPU. It might be better to replace kmalloc with a new cache, but my tests so far haven=E2=80=99t shown any performance improvements.=C2=A0 I=E2=80=99ll look= into it further. >=C2=A0