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=-18.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=unavailable 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 F3D6BC43603 for ; Tue, 26 Jan 2021 20:46:27 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id CF9D9206D4 for ; Tue, 26 Jan 2021 20:46:27 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1731108AbhAZFFE (ORCPT ); Tue, 26 Jan 2021 00:05:04 -0500 Received: from mail.kernel.org ([198.145.29.99]:40802 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1731338AbhAYSyZ (ORCPT ); Mon, 25 Jan 2021 13:54:25 -0500 Received: by mail.kernel.org (Postfix) with ESMTPSA id 6AF8920719; Mon, 25 Jan 2021 18:53:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=linuxfoundation.org; s=korg; t=1611600825; bh=YaoFlUXHhcaBDbyC+y3UWkB3R7NV44U7PV0JzBxp0U0=; h=From:To:Cc:Subject:Date:In-Reply-To:References:From; b=W/rgiZU3KtMJsaadcu3blCmvN1iDjeN2fDBtWUU8XOUGb/8yJhq7mSoTXyhIzd8g1 11yds1MDw61XXp/oFRLy8bKneLoYW8Tbqqr3D49ASBhuEu99UJX4DBQm2p10ATQTq0 OiRQIJSp87X3k801Q4gZoBiH5it2Vj+GZ6ObpjnQ= From: Greg Kroah-Hartman To: linux-kernel@vger.kernel.org Cc: Greg Kroah-Hartman , stable@vger.kernel.org, Alexander Lobakin , Jakub Kicinski Subject: [PATCH 5.10 166/199] skbuff: back tiny skbs with kmalloc() in __netdev_alloc_skb() too Date: Mon, 25 Jan 2021 19:39:48 +0100 Message-Id: <20210125183223.200711279@linuxfoundation.org> X-Mailer: git-send-email 2.30.0 In-Reply-To: <20210125183216.245315437@linuxfoundation.org> References: <20210125183216.245315437@linuxfoundation.org> User-Agent: quilt/0.66 MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: stable@vger.kernel.org From: Alexander Lobakin commit 66c556025d687dbdd0f748c5e1df89c977b6c02a upstream. Commit 3226b158e67c ("net: avoid 32 x truesize under-estimation for tiny skbs") ensured that skbs with data size lower than 1025 bytes will be kmalloc'ed to avoid excessive page cache fragmentation and memory consumption. However, the fix adressed only __napi_alloc_skb() (primarily for virtio_net and napi_get_frags()), but the issue can still be achieved through __netdev_alloc_skb(), which is still used by several drivers. Drivers often allocate a tiny skb for headers and place the rest of the frame to frags (so-called copybreak). Mirror the condition to __netdev_alloc_skb() to handle this case too. Since v1 [0]: - fix "Fixes:" tag; - refine commit message (mention copybreak usecase). [0] https://lore.kernel.org/netdev/20210114235423.232737-1-alobakin@pm.me Fixes: a1c7fff7e18f ("net: netdev_alloc_skb() use build_skb()") Signed-off-by: Alexander Lobakin Link: https://lore.kernel.org/r/20210115150354.85967-1-alobakin@pm.me Signed-off-by: Jakub Kicinski Signed-off-by: Greg Kroah-Hartman --- net/core/skbuff.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-) --- a/net/core/skbuff.c +++ b/net/core/skbuff.c @@ -432,7 +432,11 @@ struct sk_buff *__netdev_alloc_skb(struc len += NET_SKB_PAD; - if ((len > SKB_WITH_OVERHEAD(PAGE_SIZE)) || + /* If requested length is either too small or too big, + * we use kmalloc() for skb->head allocation. + */ + if (len <= SKB_WITH_OVERHEAD(1024) || + len > SKB_WITH_OVERHEAD(PAGE_SIZE) || (gfp_mask & (__GFP_DIRECT_RECLAIM | GFP_DMA))) { skb = __alloc_skb(len, gfp_mask, SKB_ALLOC_RX, NUMA_NO_NODE); if (!skb)