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]) by smtp.lore.kernel.org (Postfix) with ESMTP id 41FD3E732D0 for ; Fri, 29 Sep 2023 09:05:07 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 735278D0072; Fri, 29 Sep 2023 05:05:06 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 6E5038D0023; Fri, 29 Sep 2023 05:05:06 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 5ACC18D0072; Fri, 29 Sep 2023 05:05:06 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0013.hostedemail.com [216.40.44.13]) by kanga.kvack.org (Postfix) with ESMTP id 47FFB8D0023 for ; Fri, 29 Sep 2023 05:05:06 -0400 (EDT) Received: from smtpin25.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay05.hostedemail.com (Postfix) with ESMTP id 12D3141277 for ; Fri, 29 Sep 2023 09:05:06 +0000 (UTC) X-FDA: 81289050612.25.0C2B9E5 Received: from dfw.source.kernel.org (dfw.source.kernel.org [139.178.84.217]) by imf17.hostedemail.com (Postfix) with ESMTP id 41A4140014 for ; Fri, 29 Sep 2023 09:05:04 +0000 (UTC) Authentication-Results: imf17.hostedemail.com; dkim=pass header.d=kernel.org header.s=k20201202 header.b=MBdWRV8B; spf=pass (imf17.hostedemail.com: domain of rppt@kernel.org designates 139.178.84.217 as permitted sender) smtp.mailfrom=rppt@kernel.org; dmarc=pass (policy=none) header.from=kernel.org ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1695978304; 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=Jd6t27ADvEzsLxdcnOgyEauHOhoGpmbq+RPKJ0HOIkg=; b=gdtKm8mC4mKgi6RNp75H0NpJIUdn+dn75DhEqitvJOm0hZCRblW4S/El2aAyc1stw6noEb ilt/twajm043BtHGVcWFl/cgpP89mFaBLSoeGqeU3Cu0+JCCZh5G8rc/z8ZUeUjScPJtXh Cz3vhRRMLrICFahXRp4Aw64kiz3cGFY= ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1695978304; a=rsa-sha256; cv=none; b=21sL9TzZaeHXj4tR9ls/6HWqoloiIP1PrHGTAT/JDTBFwcMAbL/B9fA1rVVVw20m6Bpbc0 3+u1X3OnEzBRBYwNGnOcx3NCi9pSJ+NmbWzvlGhBk4ZzCsECJWdZ26I6PfH0O2o2gBRLLi 1ky5+9c9u+IbyE+I5jcggVQOzugbOhY= ARC-Authentication-Results: i=1; imf17.hostedemail.com; dkim=pass header.d=kernel.org header.s=k20201202 header.b=MBdWRV8B; spf=pass (imf17.hostedemail.com: domain of rppt@kernel.org designates 139.178.84.217 as permitted sender) smtp.mailfrom=rppt@kernel.org; dmarc=pass (policy=none) header.from=kernel.org Received: from smtp.kernel.org (transwarp.subspace.kernel.org [100.75.92.58]) by dfw.source.kernel.org (Postfix) with ESMTP id 309CE61E43; Fri, 29 Sep 2023 09:05:03 +0000 (UTC) Received: by smtp.kernel.org (Postfix) with ESMTPSA id 17B88C433C7; Fri, 29 Sep 2023 09:05:00 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1695978302; bh=pr8p7EOkXa02e0yniOjEWTDShN7pEPVZkAF3BQVj6/w=; h=Date:From:To:Cc:Subject:References:In-Reply-To:From; b=MBdWRV8BDfLVCWP0zIXzuzJVo3HPrLR22jHQ/7BQuwQUJGs0RcKE/msuoXXS5Tlpn rP6YAAy1ziNWjiBj87TjF5QC06hS/aoyV0/npUU9KzHFbKSvkKGshqXQM768p/GOf2 DLV00jN2DjlLutMn6PWqRYCwF8yCLD9i60VEgvY5G44ilv/xlbn7b0W/BOSj7LHkhE KHdBxgBAJ3ea+tDQr52tvYUy0TmRMoykNxMJLRm/pEvvJeo7qIAzr+SElmJv/srHQv ymnCss/Z8EEW6YMPQiuYmLP1awSwuWBfItnrFR0nfSwxthzbU4ee0kyKWNBeQ2I/K2 0AAxsWb2b8TZA== Date: Fri, 29 Sep 2023 12:04:06 +0300 From: Mike Rapoport To: Yajun Deng Cc: akpm@linux-foundation.org, linux-mm@kvack.org, linux-kernel@vger.kernel.org Subject: Re: [PATCH] memblock: don't run loop in memblock_add_range() twice Message-ID: <20230929090406.GV3303@kernel.org> References: <20230927013752.2515238-1-yajun.deng@linux.dev> <20230928061619.GS3303@kernel.org> <3ee9c8e4-870c-4ab0-906a-7d214031d1a6@linux.dev> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <3ee9c8e4-870c-4ab0-906a-7d214031d1a6@linux.dev> X-Rspamd-Queue-Id: 41A4140014 X-Rspam-User: X-Stat-Signature: cz79g9idj5ibqk6wcp5f1tikihcx7os3 X-Rspamd-Server: rspam03 X-HE-Tag: 1695978304-682166 X-HE-Meta: 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 Ydwg83U5 F4v+6jKMS5zoYTsIwvHwt0frW3ZN+EJCb5JFO1H4OIB40dSsl5pV38ufvFtOXXdOIiXSwsyymrod7VUfUa5jb5+xNFSFBs9+YL9Y0XQSfqe1YYP6ZpI/GxHIIMIJHBKuWK5C47qVLGhE1X2VYTe/y6trsH0nnq5st6ngllbJg1P2xZECNT19aw5+N8glgvMOLwLcjEmwk07gBas9kCDZW86shPQWI4GHmPFcOHgJfbu3KMqvhV1wK3CYlYQ== X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On Thu, Sep 28, 2023 at 04:47:59PM +0800, Yajun Deng wrote: > > On 2023/9/28 14:16, Mike Rapoport wrote: > > On Wed, Sep 27, 2023 at 09:37:52AM +0800, Yajun Deng wrote: > > > There is round twice in memblock_add_range(). The first counts the number > > > of regions needed to accommodate the new area. The second actually inserts > > > them. But the first round isn't really needed, we just need to check the > > > counts before inserting them. > > > > > > Check the count before calling memblock_insert_region(). If the count is > > > equal to the maximum value, it needs to resize the array. Otherwise, > > > insert it directly. > > > > > > To avoid nested calls to memblock_add_range(), we need to call > > > memblock_reserve() out of memblock_double_array(). > > memblock_add_range() does an extra loop once in a while, but I don't think > > removing it will have any actual effect on the boot time. > > > Yes, it has no obvious actual effect on the boot time,  but it does reduce > the number of unnecessary loop. > > The actual effect on the boot time should not be the only criterion for > whether a patch is accepted or not. > > Since the comment in the previous code, it tells the user that it would be > executed twice, this can be misleading to users. > > So the new code will be simpler and clearer. It not just change the code, > but also remove the comment Adding return-by-pointer parameters to memblock_double_array() and pulling memblock_reserve() out of this function is in no way simpler and clearer that having an extra loop. If the comment is wrong, just fix the comment. > about "executed twice",  it obviously tells the user only resize the array > if it is equal to the maximum value > > and doesn't need to be executed twice. > > > > Signed-off-by: Yajun Deng > > > --- > > > mm/memblock.c | 117 ++++++++++++++++++++++++-------------------------- > > > 1 file changed, 57 insertions(+), 60 deletions(-) > > > > > > diff --git a/mm/memblock.c b/mm/memblock.c > > > index 5a88d6d24d79..3f44c84f5d0b 100644 > > > --- a/mm/memblock.c > > > +++ b/mm/memblock.c > > > @@ -400,6 +400,8 @@ void __init memblock_discard(void) > > > * @type: memblock type of the regions array being doubled > > > * @new_area_start: starting address of memory range to avoid overlap with > > > * @new_area_size: size of memory range to avoid overlap with > > > + * @new_reserve_base: starting address of new array > > > + * @new_reserve_size: size of new array > > > * > > > * Double the size of the @type regions array. If memblock is being used to > > > * allocate memory for a new reserved regions array and there is a previously > > > @@ -412,7 +414,9 @@ void __init memblock_discard(void) > > > */ > > > static int __init_memblock memblock_double_array(struct memblock_type *type, > > > phys_addr_t new_area_start, > > > - phys_addr_t new_area_size) > > > + phys_addr_t new_area_size, > > > + phys_addr_t *new_reserve_base, > > > + phys_addr_t *new_reserve_size) > > > { > > > struct memblock_region *new_array, *old_array; > > > phys_addr_t old_alloc_size, new_alloc_size; > > > @@ -490,11 +494,13 @@ static int __init_memblock memblock_double_array(struct memblock_type *type, > > > memblock_free(old_array, old_alloc_size); > > > /* > > > - * Reserve the new array if that comes from the memblock. Otherwise, we > > > - * needn't do it > > > + * Keep the address and size if that comes from the memblock. Otherwise, > > > + * we needn't do it. > > > */ > > > - if (!use_slab) > > > - BUG_ON(memblock_reserve(addr, new_alloc_size)); > > > + if (!use_slab) { > > > + *new_reserve_base = addr; > > > + *new_reserve_size = new_alloc_size; > > > + } > > > /* Update slab flag */ > > > *in_slab = use_slab; > > > @@ -588,11 +594,12 @@ static int __init_memblock memblock_add_range(struct memblock_type *type, > > > phys_addr_t base, phys_addr_t size, > > > int nid, enum memblock_flags flags) > > > { > > > - bool insert = false; > > > phys_addr_t obase = base; > > > phys_addr_t end = base + memblock_cap_size(base, &size); > > > - int idx, nr_new, start_rgn = -1, end_rgn; > > > + phys_addr_t new_base = 0, new_size; > > > + int idx, start_rgn = -1, end_rgn; > > > struct memblock_region *rgn; > > > + unsigned long ocnt = type->cnt; > > > if (!size) > > > return 0; > > > @@ -608,25 +615,6 @@ static int __init_memblock memblock_add_range(struct memblock_type *type, > > > return 0; > > > } > > > - /* > > > - * The worst case is when new range overlaps all existing regions, > > > - * then we'll need type->cnt + 1 empty regions in @type. So if > > > - * type->cnt * 2 + 1 is less than or equal to type->max, we know > > > - * that there is enough empty regions in @type, and we can insert > > > - * regions directly. > > > - */ > > > - if (type->cnt * 2 + 1 <= type->max) > > > - insert = true; > > > - > > > -repeat: > > > - /* > > > - * The following is executed twice. Once with %false @insert and > > > - * then with %true. The first counts the number of regions needed > > > - * to accommodate the new area. The second actually inserts them. > > > - */ > > > - base = obase; > > > - nr_new = 0; > > > - > > > for_each_memblock_type(idx, type, rgn) { > > > phys_addr_t rbase = rgn->base; > > > phys_addr_t rend = rbase + rgn->size; > > > @@ -644,15 +632,23 @@ static int __init_memblock memblock_add_range(struct memblock_type *type, > > > WARN_ON(nid != memblock_get_region_node(rgn)); > > > #endif > > > WARN_ON(flags != rgn->flags); > > > - nr_new++; > > > - if (insert) { > > > - if (start_rgn == -1) > > > - start_rgn = idx; > > > - end_rgn = idx + 1; > > > - memblock_insert_region(type, idx++, base, > > > - rbase - base, nid, > > > - flags); > > > - } > > > + > > > + /* > > > + * If type->cnt is equal to type->max, it means there's > > > + * not enough empty region and the array needs to be > > > + * resized. Otherwise, insert it directly. > > > + */ > > > + if ((type->cnt == type->max) && > > > + memblock_double_array(type, obase, size, > > > + &new_base, &new_size)) > > > + return -ENOMEM; > > > + > > > + if (start_rgn == -1) > > > + start_rgn = idx; > > > + end_rgn = idx + 1; > > > + memblock_insert_region(type, idx++, base, > > > + rbase - base, nid, > > > + flags); > > > } > > > /* area below @rend is dealt with, forget about it */ > > > base = min(rend, end); > > > @@ -660,33 +656,28 @@ static int __init_memblock memblock_add_range(struct memblock_type *type, > > > /* insert the remaining portion */ > > > if (base < end) { > > > - nr_new++; > > > - if (insert) { > > > - if (start_rgn == -1) > > > - start_rgn = idx; > > > - end_rgn = idx + 1; > > > - memblock_insert_region(type, idx, base, end - base, > > > - nid, flags); > > > - } > > > + if ((type->cnt == type->max) && > > > + memblock_double_array(type, obase, size, > > > + &new_base, &new_size)) > > > + return -ENOMEM; > > > + > > > + if (start_rgn == -1) > > > + start_rgn = idx; > > > + end_rgn = idx + 1; > > > + memblock_insert_region(type, idx, base, end - base, > > > + nid, flags); > > > } > > > - if (!nr_new) > > > + if (ocnt == type->cnt) > > > return 0; > > > - /* > > > - * If this was the first round, resize array and repeat for actual > > > - * insertions; otherwise, merge and return. > > > - */ > > > - if (!insert) { > > > - while (type->cnt + nr_new > type->max) > > > - if (memblock_double_array(type, obase, size) < 0) > > > - return -ENOMEM; > > > - insert = true; > > > - goto repeat; > > > - } else { > > > - memblock_merge_regions(type, start_rgn, end_rgn); > > > - return 0; > > > - } > > > + memblock_merge_regions(type, start_rgn, end_rgn); > > > + > > > + /* Reserve the new array */ > > > + if (new_base) > > > + memblock_reserve(new_base, new_size); > > > + > > > + return 0; > > > } > > > /** > > > @@ -755,6 +746,7 @@ static int __init_memblock memblock_isolate_range(struct memblock_type *type, > > > int *start_rgn, int *end_rgn) > > > { > > > phys_addr_t end = base + memblock_cap_size(base, &size); > > > + phys_addr_t new_base = 0, new_size; > > > int idx; > > > struct memblock_region *rgn; > > > @@ -764,10 +756,15 @@ static int __init_memblock memblock_isolate_range(struct memblock_type *type, > > > return 0; > > > /* we'll create at most two more regions */ > > > - while (type->cnt + 2 > type->max) > > > - if (memblock_double_array(type, base, size) < 0) > > > + if (type->cnt + 2 > type->max) { > > > + if (memblock_double_array(type, base, size, > > > + &new_base, &new_size)) > > > return -ENOMEM; > > > + if (new_base) > > > + memblock_reserve(new_base, new_size); > > > + } > > > + > > > for_each_memblock_type(idx, type, rgn) { > > > phys_addr_t rbase = rgn->base; > > > phys_addr_t rend = rbase + rgn->size; > > > -- > > > 2.25.1 > > > -- Sincerely yours, Mike.