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 X-Spam-Level: X-Spam-Status: No, score=-9.0 required=3.0 tests=DKIM_SIGNED,DKIM_VALID, DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_PASS,URIBL_BLOCKED,USER_AGENT_NEOMUTT autolearn=ham autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 99DA8C04EB8 for ; Thu, 6 Dec 2018 06:29:42 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 3F053214DA for ; Thu, 6 Dec 2018 06:29:42 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="FYTodboI" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 3F053214DA Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=none smtp.mailfrom=linux-kernel-owner@vger.kernel.org Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729031AbeLFG3l (ORCPT ); Thu, 6 Dec 2018 01:29:41 -0500 Received: from mail-ed1-f66.google.com ([209.85.208.66]:46517 "EHLO mail-ed1-f66.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728791AbeLFG3k (ORCPT ); Thu, 6 Dec 2018 01:29:40 -0500 Received: by mail-ed1-f66.google.com with SMTP id o10so19067936edt.13 for ; Wed, 05 Dec 2018 22:29:39 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=date:from:to:cc:subject:message-id:reply-to:references:mime-version :content-disposition:in-reply-to:user-agent; bh=lT2EWLB6aH536Usb+qgj+Ndb1wMs3lEr0sS4J7KTDKE=; b=FYTodboIRl0LvDcX2uWkRGLP2rVWuMjv0weYTADo7ZsF3gm/UTDtYDtAqScTtN0U7z QYj1J3EXsdwMo9xQmOYJioPAjPf1KuttDUQzb7twYofxTQrajbk+kbzjd3MKdy6PKj3V 8Cg54CGvKPBjBhd2fvZPzJABsRztxKffDlu3J2ntdN6NYbfRCHbrkBC2TjhkdWAAhKQj Bd/ZliVFMHeSQOD/PaCCGSxtM7vSmxpNs2Nx4L7sNRgyYjUaPqWLRmjguDhvHA5vm7ux 69SeTvnkN9SLZj5OkzAQhQXo82Z1L4HvUufihIC/bT7YEU0Smvp2nJdHPKvKf/yI/LRO mNOQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:date:from:to:cc:subject:message-id:reply-to :references:mime-version:content-disposition:in-reply-to:user-agent; bh=lT2EWLB6aH536Usb+qgj+Ndb1wMs3lEr0sS4J7KTDKE=; b=Zrc0okhbaX5vUyCshKlmLZM5UA3enl6FrlvDOZd7R++oeWV7QGH2Dq5Ro+a39USaby mCVQ7hz4w8/bhglvOReE7XHmhTfzwg93d3Jgj1wNw2ySeOmy9KEjdPAp4CAcvQrT+N+H eQ0pWsl9uXznZH/oUFTqf9UlMRP7yOlJBr/LesdyF5waBs9wg7UY86qRrZsNyNAHCWeZ VWzZ3WhQ8a0v7X2cml4V+XNHIQ+4DKNP5LxvKK+cKvtB3rvON/PruvHoz4JfQREoBj1w eODCwmNbk66iQ+1YCxfW7TA0oz10zZJ8kO3sxIGwtzVmBgK6wcpArRVlwnZZh6TVRZ7F Nzbw== X-Gm-Message-State: AA+aEWZXylMVfzwHD/PFx2XCmD5qTGDFWQFLsDdsIYzjTSGNhQGS7xg5 0dnf0ra4pLMWQaBDa54W+gs= X-Google-Smtp-Source: AFSGD/WIMQfbCnlglb91fsMRPrTLLBd2C7qfl8icZuniuFS0W9g0FdVPTS4Iep+bLeVKri2p46JOdg== X-Received: by 2002:a17:906:5394:: with SMTP id g20-v6mr20736357ejo.174.1544077778659; Wed, 05 Dec 2018 22:29:38 -0800 (PST) Received: from localhost ([185.92.221.13]) by smtp.gmail.com with ESMTPSA id c22-v6sm3441073ejo.32.2018.12.05.22.29.37 (version=TLS1_2 cipher=ECDHE-RSA-CHACHA20-POLY1305 bits=256/256); Wed, 05 Dec 2018 22:29:37 -0800 (PST) Date: Thu, 6 Dec 2018 06:29:37 +0000 From: Wei Yang To: Nicolas Boichat Cc: richard.weiyang@gmail.com, Will Deacon , Michal Hocko , Levin Alexander , linux-mm@kvack.org, Christoph Lameter , Huaisheng Ye , Matthew Wilcox , linux-arm Mailing List , David Rientjes , yingjoe.chen@mediatek.com, Vlastimil Babka , Tomasz Figa , Mike Rapoport , Matthias Brugger , Joonsoo Kim , Robin Murphy , lkml , Pekka Enberg , iommu@lists.linux-foundation.org, Andrew Morton , Mel Gorman Subject: Re: [PATCH v4 2/3] mm: Add support for kmem caches in DMA32 zone Message-ID: <20181206062937.bs4sjud6mm6gy47i@master> Reply-To: Wei Yang References: <20181205054828.183476-1-drinkcat@chromium.org> <20181205054828.183476-3-drinkcat@chromium.org> <20181205072528.l7blg6y24ggblh4m@master> <20181205121807.evmslrimsv4pdtza@master> <20181206033257.mmgh6efejee2i2ae@master> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: NeoMutt/20170113 (1.7.2) Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Thu, Dec 06, 2018 at 11:55:02AM +0800, Nicolas Boichat wrote: >On Thu, Dec 6, 2018 at 11:32 AM Wei Yang wrote: >> >> On Thu, Dec 06, 2018 at 08:41:36AM +0800, Nicolas Boichat wrote: >> >On Wed, Dec 5, 2018 at 8:18 PM Wei Yang wrote: >> >> >> >> On Wed, Dec 05, 2018 at 03:39:51PM +0800, Nicolas Boichat wrote: >> >> >On Wed, Dec 5, 2018 at 3:25 PM Wei Yang wrote: >> >> >> >> >> >> On Wed, Dec 05, 2018 at 01:48:27PM +0800, Nicolas Boichat wrote: >> >> >> >In some cases (e.g. IOMMU ARMv7s page allocator), we need to allocate >> >> >> >data structures smaller than a page with GFP_DMA32 flag. >> >> >> > >> >> >> >This change makes it possible to create a custom cache in DMA32 zone >> >> >> >using kmem_cache_create, then allocate memory using kmem_cache_alloc. >> >> >> > >> >> >> >We do not create a DMA32 kmalloc cache array, as there are currently >> >> >> >no users of kmalloc(..., GFP_DMA32). The new test in check_slab_flags >> >> >> >ensures that such calls still fail (as they do before this change). >> >> >> > >> >> >> >Fixes: ad67f5a6545f ("arm64: replace ZONE_DMA with ZONE_DMA32") >> >> >> >Signed-off-by: Nicolas Boichat >> >> >> >--- >> >> >> > >> >> >> >Changes since v2: >> >> >> > - Clarified commit message >> >> >> > - Add entry in sysfs-kernel-slab to document the new sysfs file >> >> >> > >> >> >> >(v3 used the page_frag approach) >> >> >> > >> >> >> >Documentation/ABI/testing/sysfs-kernel-slab | 9 +++++++++ >> >> >> > include/linux/slab.h | 2 ++ >> >> >> > mm/internal.h | 8 ++++++-- >> >> >> > mm/slab.c | 4 +++- >> >> >> > mm/slab.h | 3 ++- >> >> >> > mm/slab_common.c | 2 +- >> >> >> > mm/slub.c | 18 +++++++++++++++++- >> >> >> > 7 files changed, 40 insertions(+), 6 deletions(-) >> >> >> > >> >> >> >diff --git a/Documentation/ABI/testing/sysfs-kernel-slab b/Documentation/ABI/testing/sysfs-kernel-slab >> >> >> >index 29601d93a1c2ea..d742c6cfdffbe9 100644 >> >> >> >--- a/Documentation/ABI/testing/sysfs-kernel-slab >> >> >> >+++ b/Documentation/ABI/testing/sysfs-kernel-slab >> >> >> >@@ -106,6 +106,15 @@ Description: >> >> >> > are from ZONE_DMA. >> >> >> > Available when CONFIG_ZONE_DMA is enabled. >> >> >> > >> >> >> >+What: /sys/kernel/slab/cache/cache_dma32 >> >> >> >+Date: December 2018 >> >> >> >+KernelVersion: 4.21 >> >> >> >+Contact: Nicolas Boichat >> >> >> >+Description: >> >> >> >+ The cache_dma32 file is read-only and specifies whether objects >> >> >> >+ are from ZONE_DMA32. >> >> >> >+ Available when CONFIG_ZONE_DMA32 is enabled. >> >> >> >+ >> >> >> > What: /sys/kernel/slab/cache/cpu_slabs >> >> >> > Date: May 2007 >> >> >> > KernelVersion: 2.6.22 >> >> >> >diff --git a/include/linux/slab.h b/include/linux/slab.h >> >> >> >index 11b45f7ae4057c..9449b19c5f107a 100644 >> >> >> >--- a/include/linux/slab.h >> >> >> >+++ b/include/linux/slab.h >> >> >> >@@ -32,6 +32,8 @@ >> >> >> > #define SLAB_HWCACHE_ALIGN ((slab_flags_t __force)0x00002000U) >> >> >> > /* Use GFP_DMA memory */ >> >> >> > #define SLAB_CACHE_DMA ((slab_flags_t __force)0x00004000U) >> >> >> >+/* Use GFP_DMA32 memory */ >> >> >> >+#define SLAB_CACHE_DMA32 ((slab_flags_t __force)0x00008000U) >> >> >> > /* DEBUG: Store the last owner for bug hunting */ >> >> >> > #define SLAB_STORE_USER ((slab_flags_t __force)0x00010000U) >> >> >> > /* Panic if kmem_cache_create() fails */ >> >> >> >diff --git a/mm/internal.h b/mm/internal.h >> >> >> >index a2ee82a0cd44ae..fd244ad716eaf8 100644 >> >> >> >--- a/mm/internal.h >> >> >> >+++ b/mm/internal.h >> >> >> >@@ -14,6 +14,7 @@ >> >> >> > #include >> >> >> > #include >> >> >> > #include >> >> >> >+#include >> >> >> > #include >> >> >> > >> >> >> > /* >> >> >> >@@ -34,9 +35,12 @@ >> >> >> > #define GFP_CONSTRAINT_MASK (__GFP_HARDWALL|__GFP_THISNODE) >> >> >> > >> >> >> > /* Check for flags that must not be used with a slab allocator */ >> >> >> >-static inline gfp_t check_slab_flags(gfp_t flags) >> >> >> >+static inline gfp_t check_slab_flags(gfp_t flags, slab_flags_t slab_flags) >> >> >> > { >> >> >> >- gfp_t bug_mask = __GFP_DMA32 | __GFP_HIGHMEM | ~__GFP_BITS_MASK; >> >> >> >+ gfp_t bug_mask = __GFP_HIGHMEM | ~__GFP_BITS_MASK; >> >> >> >+ >> >> >> >+ if (!IS_ENABLED(CONFIG_ZONE_DMA32) || !(slab_flags & SLAB_CACHE_DMA32)) >> >> >> >+ bug_mask |= __GFP_DMA32; >> >> >> >> >> >> The original version doesn't check CONFIG_ZONE_DMA32. >> >> >> >> >> >> Do we need to add this condition here? >> >> >> Could we just decide the bug_mask based on slab_flags? >> >> > >> >> >We can. The reason I did it this way is that when we don't have >> >> >CONFIG_ZONE_DMA32, the compiler should be able to simplify to: >> >> > >> >> >bug_mask = __GFP_HIGHMEM | ~__GFP_BITS_MASK; >> >> >if (true || ..) => if (true) >> >> > bug_mask |= __GFP_DMA32; >> >> > >> >> >Then just >> >> >bug_mask = __GFP_HIGHMEM | ~__GFP_BITS_MASK | __GFP_DMA32; >> >> > >> >> >And since the function is inline, slab_flags would not even need to be >> >> >accessed at all. >> >> > >> >> >> >> Hmm, I get one confusion. >> >> >> >> This means if CONFIG_ZONE_DMA32 is not enabled, bug_mask will always >> >> contains __GFP_DMA32. This will check with cachep->flags. >> >> >> >> If cachep->flags has GFP_DMA32, this always fail? >> >> >> >> Is this possible? >> > >> >Not fully sure to understand the question, but the code is: >> >if (!IS_ENABLED(CONFIG_ZONE_DMA32) || !(slab_flags & SLAB_CACHE_DMA32)) >> > bug_mask |= __GFP_DMA32; >> > >> >IS_ENABLED(CONFIG_ZONE_DMA32) == true: >> > - (slab_flags & SLAB_CACHE_DMA32) => bug_mask untouched, __GFP_DMA32 >> >is allowed. >> > - !(slab_flags & SLAB_CACHE_DMA32) => bug_mask |= __GFP_DMA32;, >> >__GFP_DMA32 triggers warning >> >IS_ENABLED(CONFIG_ZONE_DMA32) == false: >> > => bug_mask |= __GFP_DMA32;, __GFP_DMA32 triggers warning (as >> >expected, GFP_DMA32 does not make sense if there is no DMA32 zone). >> >> This is the case I am thinking. >> >> The warning is reasonable since there is no DMA32. While the >> kmem_cache_create() user is not easy to change their code. >> >> For example, one writes code and wants to have a kmem_cache with DMA32 >> capability, so he writes kmem_cache_create(__GFP_DMA32). The code is >> there and not easy to change. But one distro builder decides to disable >> DMA32. This will leads to all the kmem_cache_create() through warning? > >I don't think CONFIG_ZONE_DMA32 can be enabled/disabled by >distro/user? IIUC this is a property of the architecture, some have it >enabled, some don't. Ok, thanks. > >> This behavior is what we expect? >> >> > >> >Does that clarify? >> > >> >> >> >> -- >> >> Wei Yang >> >> Help you, Help me >> >> -- >> Wei Yang >> Help you, Help me -- Wei Yang Help you, Help me