From mboxrd@z Thu Jan 1 00:00:00 1970 From: Eric Dumazet Subject: Re: [PATCH v3 net-next 08/14] mlx4: use order-0 pages for RX Date: Thu, 16 Feb 2017 09:11:01 -0800 Message-ID: References: <20170213195858.5215-1-edumazet@google.com> <20170213195858.5215-9-edumazet@google.com> <20170214131206.44b644f6@redhat.com> <1487087488.8227.53.camel@edumazet-glaptop3.roam.corp.google.com> <37bc04eb-71c9-0433-304d-87fcf8b06be3@mellanox.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Cc: Tariq Toukan , Jesper Dangaard Brouer , Eric Dumazet , Alexander Duyck , "David S . Miller" , netdev , Martin KaFai Lau , Saeed Mahameed , Willem de Bruijn , Brenden Blanco , Alexei Starovoitov , linux-mm To: Tom Herbert Return-path: Received: from mail-it0-f42.google.com ([209.85.214.42]:37163 "EHLO mail-it0-f42.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932193AbdBPRLD (ORCPT ); Thu, 16 Feb 2017 12:11:03 -0500 Received: by mail-it0-f42.google.com with SMTP id x75so32177129itb.0 for ; Thu, 16 Feb 2017 09:11:02 -0800 (PST) In-Reply-To: Sender: netdev-owner@vger.kernel.org List-ID: > You're admitting that Eric's patches improve driver quality, > stability, and performance but you're not allowing this in the kernel > because "we know what benchmarks our customers are going to run". Note that I do not particularly care if these patches go in 4.11 or 4.12 really. I already backported them into our 4.3 based kernel. I guess that we could at least propose the trivial patch for stable releases, since PowerPC arches really need it. diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h index cec59bc264c9ac197048fd7c98bcd5cf25de0efd..0f6d2f3b7d54f51de359d4ccde21f4585e6b7852 100644 --- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h +++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h @@ -102,7 +102,8 @@ /* Use the maximum between 16384 and a single page */ #define MLX4_EN_ALLOC_SIZE PAGE_ALIGN(16384) -#define MLX4_EN_ALLOC_PREFER_ORDER PAGE_ALLOC_COSTLY_ORDER +#define MLX4_EN_ALLOC_PREFER_ORDER min_t(int, get_order(32768), \ + PAGE_ALLOC_COSTLY_ORDER) /* Receive fragment sizes; we use at most 3 fragments (for 9600 byte MTU * and 4K allocations) */