Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752229AbbBVU55 (ORCPT ); Sun, 22 Feb 2015 15:57:57 -0500 Received: from shards.monkeyblade.net ([149.20.54.216]:49215 "EHLO shards.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751601AbbBVU5z (ORCPT ); Sun, 22 Feb 2015 15:57:55 -0500 Date: Sun, 22 Feb 2015 15:57:50 -0500 (EST) Message-Id: <20150222.155750.495884608409425947.davem@davemloft.net> To: szepe@pinerecords.com Cc: eric.dumazet@gmail.com, fw@strlen.de, romieu@fr.zoreil.com, hayeswang@realtek.com, edumazet@google.com, therbert@google.com, pupilla@hotmail.com, linux-kernel@vger.kernel.org, netdev@vger.kernel.org Subject: Re: 1e918876 breaks r8169 (linux-3.18+) From: David Miller In-Reply-To: <20150222004151.GA4947@louise.pinerecords.com> References: <20150221190507.GA25687@louise.pinerecords.com> <1424548447.5565.48.camel@edumazet-glaptop2.roam.corp.google.com> <20150222004151.GA4947@louise.pinerecords.com> X-Mailer: Mew version 6.6 on Emacs 24.4 / Mule 6.0 (HANACHIRUSATO) Mime-Version: 1.0 Content-Type: Text/Plain; charset=us-ascii Content-Transfer-Encoding: 7bit X-Greylist: Sender succeeded SMTP AUTH, not delayed by milter-greylist-4.5.7 (shards.monkeyblade.net [149.20.54.216]); Sun, 22 Feb 2015 12:57:54 -0800 (PST) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 4492 Lines: 139 From: Tomas Szepe Date: Sun, 22 Feb 2015 01:41:51 +0100 >> > Sure, just did. Unfortunately, 3.19.0 + 0bec3b70 + this patch results >> > in a driver that retains the problem. >> >> OK, could you test following patch instead ? > > Yup, but tough luck: 3.19.0 + 0bec3b70 + this patch -> problem present. I'm reverting the two commits for now, as below. We can put them back in if we can resolve the problems. ==================== [PATCH] r8169: Revert BQL and xmit_more support. There are certain regressions which are pointing to these two commits which we are having a hard time resolving. So revert them for now. Specifically this reverts: commit 0bec3b700d106a8b0a34227b2976d1a582f1aab7 Author: Florian Westphal Date: Wed Jan 7 10:49:49 2015 +0100 r8169: add support for xmit_more and commit 1e918876853aa85435e0f17fd8b4a92dcfff53d6 Author: Florian Westphal Date: Wed Oct 1 13:38:03 2014 +0200 r8169: add support for Byte Queue Limits There were some attempts by Eric Dumazet to address some obvious problems in the TX flow, to see if they would fix the problems, but none of them seem to help for the regression reporters. Signed-off-by: David S. Miller --- drivers/net/ethernet/realtek/r8169.c | 30 +++++++----------------------- 1 file changed, 7 insertions(+), 23 deletions(-) diff --git a/drivers/net/ethernet/realtek/r8169.c b/drivers/net/ethernet/realtek/r8169.c index ad0020a..b156092 100644 --- a/drivers/net/ethernet/realtek/r8169.c +++ b/drivers/net/ethernet/realtek/r8169.c @@ -5067,8 +5067,6 @@ static void rtl_hw_reset(struct rtl8169_private *tp) RTL_W8(ChipCmd, CmdReset); rtl_udelay_loop_wait_low(tp, &rtl_chipcmd_cond, 100, 100); - - netdev_reset_queue(tp->dev); } static void rtl_request_uncached_firmware(struct rtl8169_private *tp) @@ -7049,7 +7047,6 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb, u32 status, len; u32 opts[2]; int frags; - bool stop_queue; if (unlikely(!TX_FRAGS_READY_FOR(tp, skb_shinfo(skb)->nr_frags))) { netif_err(tp, drv, dev, "BUG! Tx Ring full when queue awake!\n"); @@ -7090,8 +7087,6 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb, txd->opts2 = cpu_to_le32(opts[1]); - netdev_sent_queue(dev, skb->len); - skb_tx_timestamp(skb); /* Force memory writes to complete before releasing descriptor */ @@ -7106,16 +7101,11 @@ static netdev_tx_t rtl8169_start_xmit(struct sk_buff *skb, tp->cur_tx += frags + 1; - stop_queue = !TX_FRAGS_READY_FOR(tp, MAX_SKB_FRAGS); + RTL_W8(TxPoll, NPQ); - if (!skb->xmit_more || stop_queue || - netif_xmit_stopped(netdev_get_tx_queue(dev, 0))) { - RTL_W8(TxPoll, NPQ); - - mmiowb(); - } + mmiowb(); - if (stop_queue) { + if (!TX_FRAGS_READY_FOR(tp, MAX_SKB_FRAGS)) { /* Avoid wrongly optimistic queue wake-up: rtl_tx thread must * not miss a ring update when it notices a stopped queue. */ @@ -7198,7 +7188,6 @@ static void rtl8169_pcierr_interrupt(struct net_device *dev) static void rtl_tx(struct net_device *dev, struct rtl8169_private *tp) { unsigned int dirty_tx, tx_left; - unsigned int bytes_compl = 0, pkts_compl = 0; dirty_tx = tp->dirty_tx; smp_rmb(); @@ -7222,8 +7211,10 @@ static void rtl_tx(struct net_device *dev, struct rtl8169_private *tp) rtl8169_unmap_tx_skb(&tp->pci_dev->dev, tx_skb, tp->TxDescArray + entry); if (status & LastFrag) { - pkts_compl++; - bytes_compl += tx_skb->skb->len; + u64_stats_update_begin(&tp->tx_stats.syncp); + tp->tx_stats.packets++; + tp->tx_stats.bytes += tx_skb->skb->len; + u64_stats_update_end(&tp->tx_stats.syncp); dev_kfree_skb_any(tx_skb->skb); tx_skb->skb = NULL; } @@ -7232,13 +7223,6 @@ static void rtl_tx(struct net_device *dev, struct rtl8169_private *tp) } if (tp->dirty_tx != dirty_tx) { - netdev_completed_queue(tp->dev, pkts_compl, bytes_compl); - - u64_stats_update_begin(&tp->tx_stats.syncp); - tp->tx_stats.packets += pkts_compl; - tp->tx_stats.bytes += bytes_compl; - u64_stats_update_end(&tp->tx_stats.syncp); - tp->dirty_tx = dirty_tx; /* Sync with rtl8169_start_xmit: * - publish dirty_tx ring index (write barrier) -- 2.1.0 -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/