From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-lf1-f48.google.com (mail-lf1-f48.google.com [209.85.167.48]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1DA0D130ADA for ; Wed, 4 Sep 2024 07:22:26 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.167.48 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1725434549; cv=none; b=ggzWc+RP/rd660nN5Rd3nm0cm4m65Q+xmrxj/PSADxCuc6rGb10Tst5IrnC3csaX2PycLBxLcypps9OLyrQlJVEmD0uJlzOcypc6trbAftWjhj0SB6mNTwPbH5/Y89QsyVKfyuU89ZOBuLlpOOLfIQrkBmCiLIbgxmGXmdT0lHQ= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1725434549; c=relaxed/simple; bh=bbOsculdpTItYrfgREp6F0Q6OrQeZ+N8WXdyQ9WaxZY=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=lZqhDwV/BWrY6P94n+x6W3tGId1pPqEy/yAEujk6K8nLX6TKqaUfVisZ4KHi7PMT2Ju+bHbDn48qu6y4v8tPy8rtGThh+IAf9pbTmZujGuU3O8LtMi+JN1PX9bjo9+UbiUWl7F2wNEtg/ojA7x70ztWeBHPi97GzVgplsE4PLSo= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=suse.com; spf=pass smtp.mailfrom=suse.com; dkim=pass (2048-bit key) header.d=suse.com header.i=@suse.com header.b=YZ2v9DyY; arc=none smtp.client-ip=209.85.167.48 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=quarantine dis=none) header.from=suse.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=suse.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=suse.com header.i=@suse.com header.b="YZ2v9DyY" Received: by mail-lf1-f48.google.com with SMTP id 2adb3069b0e04-5334c4d6829so7833712e87.2 for ; Wed, 04 Sep 2024 00:22:26 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.com; s=google; t=1725434545; x=1726039345; darn=lists.linux.dev; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:from:date:from:to :cc:subject:date:message-id:reply-to; bh=Wh+WjO/eVoggYon1VI8AmiDiYT97jA+3Yp6qqo8Bz0o=; b=YZ2v9DyY5lkYM63My1ThLj2dO1zzd2+Kn7zTNt7kxn1JcPvcC9LfWHxlnVQ9papOXh vjPSLE/pz992IDwRy1sBVF90/KqfxolyT+tE8mmbjLWWy6ujTJP5gRQHfv1eQXSEDGaa xDDkXz3BtIBpPIh6SoCWeGNknGjMKOr+8ARyyDdYP3kNVuEi91D16xU2kPHUwoInSJBn cSb2ems37/zC4E4sZlQbVdGJANRa/6tI4ilqAfO9mIWZjxQNLLnG4WINxTzBxZe3wDV7 lo6sqt1hvogEYGnsZBVabxFqQMh8IsSVEbKZBqx95S5TShM0GwSxfv8WrLfePt5QOyJs JyjA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1725434545; x=1726039345; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:from:date :x-gm-message-state:from:to:cc:subject:date:message-id:reply-to; bh=Wh+WjO/eVoggYon1VI8AmiDiYT97jA+3Yp6qqo8Bz0o=; b=vltnm26xbNhmrJlN/r6QAuxR7gYZRJle8eP9cb4qOiHe8eFXgjU91olEiLJlIw8SY/ kln0HddDpHj7DMasG3d4+9fi3536ij49TJ1tkXpNIaAdl0qjqFCPlhaOYsLEOkFBT2m8 ikAxZmkiD3S9RoYejLl2yJ1q9xxeve7jAjUYFz1STNr0smIs28D5QDfCLa832MR4tUHM 2T87lUjpBDih33YHwSO+0jBur/Ero0BnDgJBI8xIrM4ZQ2yqbjG/DQZUTHjnO7OUPx/d TYma4cKgSa386p7r9YvPTRez29JC0FyqF36hIQoQK7SgQ4Nz/gKi2Av9lAYX2XsIX90D Bx0A== X-Forwarded-Encrypted: i=1; AJvYcCXsHz70F72CkTse5xmyOmbS5Q7cZe4vP3m5wa0XbiM8YWoBgyh5tnrcJY3pe0IdDSklj08doYjIOE6dKLIbnA==@lists.linux.dev X-Gm-Message-State: AOJu0Yyw6MnDkVxsNpcgrMTQvx3zM1NcZSr8O6d8LiVSxaUxtGBklojw bLB6WHpsNfZcDUBL5kNa4O+avLGhsA07JV7bsp9W5XhRWe52g2dk/vrInovvQcA= X-Google-Smtp-Source: AGHT+IG8z0kaSoq4W+20UCPAYQ/4EbWJkmc6rjz/n9sR4m5vvJuQveaRcmhYFgcpH0WFXJDIfkD49A== X-Received: by 2002:a05:6512:3350:b0:534:543e:1895 with SMTP id 2adb3069b0e04-53546bab3damr7960271e87.39.1725434544882; Wed, 04 Sep 2024 00:22:24 -0700 (PDT) Received: from localhost (109-81-82-19.rct.o2.cz. [109.81.82.19]) by smtp.gmail.com with ESMTPSA id a640c23a62f3a-a898919686bsm764838666b.134.2024.09.04.00.22.23 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 04 Sep 2024 00:22:24 -0700 (PDT) Date: Wed, 4 Sep 2024 09:22:23 +0200 From: Michal Hocko To: Barry Song <21cnbao@gmail.com> Cc: 42.hyeyoo@gmail.com, akpm@linux-foundation.org, cl@linux.com, david@redhat.com, hailong.liu@oppo.com, hch@infradead.org, iamjoonsoo.kim@lge.com, laoar.shao@gmail.com, linux-mm@kvack.org, penberg@kernel.org, rientjes@google.com, roman.gushchin@linux.dev, torvalds@linux-foundation.org, urezki@gmail.com, v-songbaohua@oppo.com, vbabka@suse.cz, virtualization@lists.linux.dev Subject: Re: [PATCH v4 3/3] mm: warn about illegal __GFP_NOFAIL usage in a more appropriate location and manner Message-ID: References: <20240903223935.1697-1-21cnbao@gmail.com> Precedence: bulk X-Mailing-List: virtualization@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <20240903223935.1697-1-21cnbao@gmail.com> On Wed 04-09-24 10:39:35, Barry Song wrote: > On Mon, Sep 2, 2024 at 7:58 PM Michal Hocko wrote: > > > > On Sat 31-08-24 08:28:23, Barry Song wrote: > > > From: Barry Song > > > > > > Three points for this change: > > > > > > 1. We should consolidate all warnings in one place. Currently, the > > >    order > 1 warning is in the hotpath, while others are in less > > >    likely scenarios. Moving all warnings to the slowpath will reduce > > >    the overhead for order > 1 and increase the visibility of other > > >    warnings. > > > > > > 2. We currently have two warnings for order: one for order > 1 in > > >    the hotpath and another for order > costly_order in the laziest > > >    path. I suggest standardizing on order > 1 since it’s been in > > >    use for a long time. > > > > > > 3. We don't need to check for __GFP_NOWARN in this case. __GFP_NOWARN > > >    is meant to suppress allocation failure reports, but here we're > > >    dealing with bug detection, not allocation failures. So replace > > >    WARN_ON_ONCE_GFP by WARN_ON_ONCE. > > > > > > Suggested-by: Vlastimil Babka > > > Signed-off-by: Barry Song > > > > Acked-by: Michal Hocko > > > > Updating the doc about order > 1 sounds like it would still fall into > > the scope of this patch. I don not think we absolutely have to document > > each unsupported gfp flags combination for GFP_NOFAIL but the order is a > > good addition with a note that kvmalloc should be used instead in such a > > case. > > Hi Andrew, > If there are no objections from Michal and David, could you please > squash the following: > > >From fc7a2a49e8d0811d706d13d2080393274f316806 Mon Sep 17 00:00:00 2001 > From: Barry Song > Date: Wed, 4 Sep 2024 10:26:19 +1200 > Subject: [PATCH] mm: also update the doc for __GFP_NOFAIL with order > 1 > > Obviously we only support order <= 1 __GFP_NOFAIL allocation and if > someone wants larger memory, they should consider using kvmalloc() > instead. > > Suggested-by: David Hildenbrand > Suggested-by: Michal Hocko > Signed-off-by: Barry Song LGTM. Thanks! > --- > include/linux/gfp_types.h | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/include/linux/gfp_types.h b/include/linux/gfp_types.h > index 4a1fa7706b0c..65db9349f905 100644 > --- a/include/linux/gfp_types.h > +++ b/include/linux/gfp_types.h > @@ -253,7 +253,8 @@ enum { > * used only when there is no reasonable failure policy) but it is > * definitely preferable to use the flag rather than opencode endless > * loop around allocator. > - * Using this flag for costly allocations is _highly_ discouraged. > + * Allocating pages from the buddy with __GFP_NOFAIL and order > 1 is > + * not supported. Please consider using kvmalloc() instead. > */ > #define __GFP_IO ((__force gfp_t)___GFP_IO) > #define __GFP_FS ((__force gfp_t)___GFP_FS) > -- > 2.34.1 > > > > > > > --- > > >  mm/page_alloc.c | 50 ++++++++++++++++++++++++------------------------- > > >  1 file changed, 25 insertions(+), 25 deletions(-) > > > > > > diff --git a/mm/page_alloc.c b/mm/page_alloc.c > > > index c81ee5662cc7..e790b4227322 100644 > > > --- a/mm/page_alloc.c > > > +++ b/mm/page_alloc.c > > > @@ -3033,12 +3033,6 @@ struct page *rmqueue(struct zone *preferred_zone, > > >  { > > >       struct page *page; > > > > > > -     /* > > > -      * We most definitely don't want callers attempting to > > > -      * allocate greater than order-1 page units with __GFP_NOFAIL. > > > -      */ > > > -     WARN_ON_ONCE((gfp_flags & __GFP_NOFAIL) && (order > 1)); > > > - > > >       if (likely(pcp_allowed_order(order))) { > > >               page = rmqueue_pcplist(preferred_zone, zone, order, > > >                                      migratetype, alloc_flags); > > > @@ -4175,6 +4169,7 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, > > >  { > > >       bool can_direct_reclaim = gfp_mask & __GFP_DIRECT_RECLAIM; > > >       bool can_compact = gfp_compaction_allowed(gfp_mask); > > > +     bool nofail = gfp_mask & __GFP_NOFAIL; > > >       const bool costly_order = order > PAGE_ALLOC_COSTLY_ORDER; > > >       struct page *page = NULL; > > >       unsigned int alloc_flags; > > > @@ -4187,6 +4182,25 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, > > >       unsigned int zonelist_iter_cookie; > > >       int reserve_flags; > > > > > > +     if (unlikely(nofail)) { > > > +             /* > > > +              * We most definitely don't want callers attempting to > > > +              * allocate greater than order-1 page units with __GFP_NOFAIL. > > > +              */ > > > +             WARN_ON_ONCE(order > 1); > > > +             /* > > > +              * Also we don't support __GFP_NOFAIL without __GFP_DIRECT_RECLAIM, > > > +              * otherwise, we may result in lockup. > > > +              */ > > > +             WARN_ON_ONCE(!can_direct_reclaim); > > > +             /* > > > +              * PF_MEMALLOC request from this context is rather bizarre > > > +              * because we cannot reclaim anything and only can loop waiting > > > +              * for somebody to do a work for us. > > > +              */ > > > +             WARN_ON_ONCE(current->flags & PF_MEMALLOC); > > > +     } > > > + > > >  restart: > > >       compaction_retries = 0; > > >       no_progress_loops = 0; > > > @@ -4404,29 +4418,15 @@ __alloc_pages_slowpath(gfp_t gfp_mask, unsigned int order, > > >        * Make sure that __GFP_NOFAIL request doesn't leak out and make sure > > >        * we always retry > > >        */ > > > -     if (gfp_mask & __GFP_NOFAIL) { > > > +     if (unlikely(nofail)) { > > >               /* > > > -              * All existing users of the __GFP_NOFAIL are blockable, so warn > > > -              * of any new users that actually require GFP_NOWAIT > > > +              * Lacking direct_reclaim we can't do anything to reclaim memory, > > > +              * we disregard these unreasonable nofail requests and still > > > +              * return NULL > > >                */ > > > -             if (WARN_ON_ONCE_GFP(!can_direct_reclaim, gfp_mask)) > > > +             if (!can_direct_reclaim) > > >                       goto fail; > > > > > > -             /* > > > -              * PF_MEMALLOC request from this context is rather bizarre > > > -              * because we cannot reclaim anything and only can loop waiting > > > -              * for somebody to do a work for us > > > -              */ > > > -             WARN_ON_ONCE_GFP(current->flags & PF_MEMALLOC, gfp_mask); > > > - > > > -             /* > > > -              * non failing costly orders are a hard requirement which we > > > -              * are not prepared for much so let's warn about these users > > > -              * so that we can identify them and convert them to something > > > -              * else. > > > -              */ > > > -             WARN_ON_ONCE_GFP(costly_order, gfp_mask); > > > - > > >               /* > > >                * Help non-failing allocations by giving some access to memory > > >                * reserves normally used for high priority non-blocking > > > -- > > > 2.34.1 > > > > -- > > Michal Hocko > > SUSE Labs > > Thanks > Barry -- Michal Hocko SUSE Labs