From nobody Fri Oct 30 17:33:06 2026 X-Original-To: dev-commits-src-all@mlmmj.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mlmmj.nyi.freebsd.org (Postfix) with ESMTP id 4jGSqQ4zCdz6xFxN for ; Fri, 30 Oct 2026 17:33:06 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from mxrelay.nyi.freebsd.org (mxrelay.nyi.freebsd.org [IPv6:2610:1c1:1:606c::19:3]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange x25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (No client certificate requested) by mx1.freebsd.org (Postfix) with ESMTPS id 4jGSqQ1Fb7z4WdK for ; Fri, 30 Oct 2026 17:33:06 +0000 (UTC) (envelope-from git@FreeBSD.org) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=freebsd.org; s=dkim; t=1793381586; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=I/b4p7XNdx9UtTNldxBh87oa4pVIAYHePzfOpNamtzs=; b=hZC8Mid19S7KVqucwV71P88AsRTW4Wy0wPo8QadKiSmwYBrwS6HrYnvzn5Q3EFpLHd7IK1 GSYNi9QNVs6GlOn+N59BNBOZ1Zoqu/JiP3LEr/YIMnUXvEqpIlJ87nnHZ2GIXm9OtfCJs9 p8i+om6MRKdAuWnhUBdSu4Ov4h1jKyc+s6GYk1fD/FaB3dyQTHxzPRU/hj63CAz+yrWX+9 i0IbVkKGbyNNrXgXIAr/z0oOoxBF6LB+iV0ZN6fW3uq+ZNAeXdOM309jBwFDoh9M6Q5OQk PsqzTKFEttY2f3gIS6qlMy/KOU+f0FASGh4I5XXKY4w1bonW9T5lXUWVApLpDA== ARC-Seal: i=1; a=rsa-sha256; d=freebsd.org; s=dkim; cv=none; t=1793381586; b=CVigZOjsPTcRAkYE0Uegm/wwaXtHV8KbccTBA7lRg+mZXQH08ctmax53hFwBKbwoEjswEm EgB9dQcT2ZQ9DqMY7vG/xuQWIqIbHG1NCsem8fm8shxBZtugTSFFKgY/eA0fd+b7bVXPmC aH8537eZIgp/SqX0CKh6QRg4vmqM5LttTqBePo1oYHIqPY7uIWvhP87X/Dpd+QEJjafk4G NgsdsnPlPAt6xprfuohWZEEVKs0ouJFKU5T6Vb5/Nk0FWtQXKBz5lCdU3rSQPTYLd9xJu6 IThD7K8Bm5WpmSYuQJC9kvN9x1KMo4FPH86P1LNv8hlg7sA+WEIgi1PwdEFheA== ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=freebsd.org; s=dkim; t=1793381586; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=I/b4p7XNdx9UtTNldxBh87oa4pVIAYHePzfOpNamtzs=; b=to7Cpm3NiJE2BCDUhKyb6ZYVWN/ksqpZyT6IU6wVEAhzaBt1FdPEp18Nv+o4+j4AMCpXEL 4kK9i99FNlbNkV/C/X0v1szhiDw3L/TVz1lPC7PjDZm1Ix5gvYHDAZtET9DIeecv/NQYUF QdVWSP5dYhCRpLFYgFG5OCO1OZ7f2p7S8GxZbEyDWEp7sJ+xRE32dO+SZ1lGlqxxCLT2wR SKFDrOYNmF7O2alicWkE8vYsfMmCrNiAuYuxJHItcSG8uVCLs2/mhq+PFEM93MniLTV077 bWRCelTb5IrmOZwYY3MRt/uyjIFVhlyy2hhdsAw/aM8Q9cTcb7fnBOTpVQTHFw== ARC-Authentication-Results: i=1; mx1.freebsd.org; none Received: from gitrepo.freebsd.org (gitrepo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:5]) by mxrelay.nyi.freebsd.org (Postfix) with ESMTP id 4jGSqQ0R4szdkp for ; Fri, 30 Oct 2026 17:33:06 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from git (uid 1279) (envelope-from git@FreeBSD.org) id 4022c by gitrepo.freebsd.org (DragonFly Mail Agent v0.13+ on gitrepo.freebsd.org); Fri, 30 Oct 2026 17:33:06 +0000 To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-branches@FreeBSD.org Cc: =?utf-8?Q?J=C3=A9r=C3=A9mie?= Jourdin From: Kevin Bowling Subject: git: 81a4068cc15c - stable/15 - re(4): harden re_watchdog() recovery and log controller state List-Archive: https://lists.freebsd.org/archives/dev-commits-src-all X-BeenThere: dev-commits-src-all@freebsd.org Sender: owner-dev-commits-src-all@FreeBSD.org List-Id: Commit messages for all branches of the src repository List-Post: List-Help: List-Subscribe: List-Unsubscribe: List-Owner: Precedence: list MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit X-Git-Committer: kbowling X-Git-Repository: src X-Git-Refname: refs/heads/stable/15 X-Git-Reftype: branch X-Git-Commit: 81a4068cc15cea194d2a63e3a03321348b023bac Auto-Submitted: auto-generated Date: Fri, 30 Oct 2026 17:33:06 +0000 Message-Id: <6ae4d4d2.4022c.7b2dc62a@gitrepo.freebsd.org> The branch stable/15 has been updated by kbowling: URL: https://cgit.FreeBSD.org/src/commit/?id=81a4068cc15cea194d2a63e3a03321348b023bac commit 81a4068cc15cea194d2a63e3a03321348b023bac Author: Jérémie Jourdin AuthorDate: 2026-07-19 18:47:17 +0000 Commit: Kevin Bowling CommitDate: 2026-10-30 17:32:43 +0000 re(4): harden re_watchdog() recovery and log controller state Distinguish the two failure classes from the PR in a single log line (ring indices, ISR/IMR, TXCFG, interrupt mode): lost interrupt vs genuine DMA stall. Bail out instead of re-initializing when the controller reads back all-ones (fallen off the bus; reinit cannot help). Re-assert the driver's existing ASPM-disabled policy before reinit, since firmware/power transitions re-arming L0s/L1 is a documented stall trigger. Diagnostics-only on the recovered path; no fast-path change. * Field diagnostics running on an RTL8168H production fleet; the log format distinguishes lost-doorbell / DMA-stall / dead-controller without a debug build. Reviewed by: adrian Differential Revision: https://reviews.freebsd.org/D58279 PR: kern/166724 (cherry picked from commit 42f3088bd980a469dbbc3e6b568edb7ce0ab2bc3) --- sys/dev/re/if_re.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++--- 1 file changed, 58 insertions(+), 3 deletions(-) diff --git a/sys/dev/re/if_re.c b/sys/dev/re/if_re.c index e8e344c233d5..faf7922a02df 100644 --- a/sys/dev/re/if_re.c +++ b/sys/dev/re/if_re.c @@ -3600,6 +3600,8 @@ re_watchdog(struct rl_softc *sc) { struct epoch_tracker et; if_t ifp; + const char *imode; + uint32_t isr, imr, txcfg; RL_LOCK_ASSERT(sc); @@ -3607,18 +3609,71 @@ re_watchdog(struct rl_softc *sc) return; ifp = sc->rl_ifp; + /* + * Snapshot the controller state before reclaim for the diagnostics + * below. Reading RL_ISR is side-effect free: the interrupt status + * is cleared by writing ones, not by reading. + */ + imode = (sc->rl_flags & RL_FLAG_MSIX) ? "MSI-X" : + (sc->rl_flags & RL_FLAG_MSI) ? "MSI" : "INTx"; + isr = CSR_READ_2(sc, RL_ISR); + imr = CSR_READ_2(sc, RL_IMR); re_txeof(sc); if (sc->rl_ldata.rl_tx_free == sc->rl_ldata.rl_tx_desc_cnt) { - if_printf(ifp, "watchdog timeout (missed Tx interrupts) " - "-- recovering\n"); + /* + * The retry reclaim drained the ring: a Tx completion interrupt + * was lost. Log the interrupt state so the lost-interrupt path + * can be diagnosed without a debug build. + */ + if_printf(ifp, "watchdog timeout (missed Tx interrupts) -- " + "recovering (ISR 0x%04x IMR 0x%04x %s)\n", + isr, imr, imode); if (!if_sendq_empty(ifp)) re_start_locked(ifp); return; } - if_printf(ifp, "watchdog timeout\n"); + txcfg = CSR_READ_4(sc, RL_TXCFG); + /* + * A genuine Tx stall. Log the Tx ring position and controller + * state in a single line -- enough to distinguish a lost doorbell, + * a DMA stall, and a controller that has fallen off the bus. + */ + if_printf(ifp, "watchdog timeout -- resetting (tx %d/%d cons %d " + "prod %d ISR 0x%04x IMR 0x%04x TXCFG 0x%08x %s)\n", + sc->rl_ldata.rl_tx_free, sc->rl_ldata.rl_tx_desc_cnt, + sc->rl_ldata.rl_tx_considx, sc->rl_ldata.rl_tx_prodidx, + isr, imr, txcfg, imode); if_inc_counter(ifp, IFCOUNTER_OERRORS, 1); + /* + * If the controller reads back all-ones it has fallen off the bus; + * re_init_locked() cannot recover it, so flag it and bail rather + * than spinning through reset after reset. + */ + if (txcfg == 0xFFFFFFFF) { + if_printf(ifp, + "controller not responding; not reinitializing\n"); + return; + } + + /* + * ASPM is a common cause of these Tx timeouts. Make sure L0s/L1 + * and CLKREQ are still disabled on the link before reinitializing; + * firmware or a power transition may have re-armed them. + */ + if (sc->rl_expcap != 0) { + uint16_t ctl; + + ctl = pci_read_config(sc->rl_dev, + sc->rl_expcap + PCIER_LINK_CTL, 2); + if ((ctl & (PCIEM_LINK_CTL_ECPM | PCIEM_LINK_CTL_ASPMC)) != 0) { + ctl &= ~(PCIEM_LINK_CTL_ECPM | PCIEM_LINK_CTL_ASPMC); + pci_write_config(sc->rl_dev, + sc->rl_expcap + PCIER_LINK_CTL, ctl, 2); + } + } + NET_EPOCH_ENTER(et); re_rxeof(sc, NULL); NET_EPOCH_EXIT(et);