From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 76DFC531AF3; Wed, 23 Sep 2026 14:34:56 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1790174097; cv=none; b=c/hUDsVyK+GoCUuMS7tLfoLH2ajgrtEwNZz2kOWHuHwJhUV/U88GxPsachy7ljaB8DYibxKPK8bbVPBZVTbY+FZTEr9pGsjMg9HZnJcj9wbkVe96+Il2a228JkxqrBVGgI67842P+Vhq3SA6tKqw4c9u3EPu6T6I65baFXC0/hE= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1790174097; c=relaxed/simple; bh=+6DU3Rgpl0cAMCUQhXz7MMERTo4M8sp3v2N5DtU7KXg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Dn0Jf4OhHadKnLBCtDM7bqBwSb1RcNr25wbVdd0BcNsFAoL9SZTQE+IDJhMc1JoAPZs+qm5gmIfA0dJvHFYDZuDvS+Ey3+pxahL7F+fPMHrfu1PZMZE7Foj1O9ggERfiL/rk8cEfEoQoUVPt5JPIWDFgQGA+4UCnDXmFri7yFhU= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b=A1h4hUpp; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=linuxfoundation.org header.i=@linuxfoundation.org header.b="A1h4hUpp" Received: by smtp.kernel.org (Postfix) with ESMTPSA id 8143D1F000FF; Wed, 23 Sep 2026 14:34:55 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linuxfoundation.org; s=korg; t=1790174096; bh=UZumUiw3G2GEAz0zxG6mCapI/1eu3zGElEeIZDMwR5s=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=A1h4hUppg5VPJz8F2weGChxPGEOO+DqIBS/VHSXAXQajTg9HQysKZAZji5jOU1eNh /JKEABN2rIQpRZuJzrfVlZzYOcUW7jFOsg5HnUYV0ljC0fyfXDr7zshvzXO4aHAULq H9H1mdYYEkz7sXj6ER1Xg+11Q6szs5AIXjIZOmVo= From: Greg Kroah-Hartman To: stable@vger.kernel.org Cc: Greg Kroah-Hartman , patches@lists.linux.dev, Sesidhar Baddela , Arulprabhu Ponnusamy , Gian Carlo Boffa , Karan Tilak Kumar , Arun Easi , Laurence Oberman , "Martin K. Petersen (Oracle)" , Sasha Levin Subject: [PATCH 7.2 432/438] scsi: fnic: Fix missed link-up when critical IRQ targets offline CPU Date: Wed, 23 Sep 2026 16:07:33 +0200 Message-ID: <20260923140656.127060535@linuxfoundation.org> X-Mailer: git-send-email 2.55.0 In-Reply-To: <20260923140644.756254324@linuxfoundation.org> References: <20260923140644.756254324@linuxfoundation.org> User-Agent: quilt/0.69 X-stable: review X-Patchwork-Hint: ignore Precedence: bulk X-Mailing-List: patches@lists.linux.dev List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit 7.2-stable review patch. If anyone has any objections, please let me know. ------------------ From: Arun Easi [ Upstream commit 0cb1fd924126f1f581621a5e804df98a02be9dff ] When CPU Hyper Threading is disabled, sibling CPUs remain present but are reported offline. Managed MSI-X IRQs can still receive affinity masks that include those offline CPUs. If a driver-critical vector is managed, it can be parked on an offline CPU and the driver may miss critical events such as link-up. Keep driver-critical vectors unmanaged so they can be migrated by the IRQ core when their target CPU is offlined. Since HWQ-0 is unmanaged now, in some queue combinations there can be no mappings to it in mq_map. So without the blk-mq fix mentioned below, system may crash during cpu offline/online tests. Fixes: 8a8449ca5e33 ("scsi: fnic: Modify ISRs to support multiqueue (MQ)") Cc: stable@vger.kernel.org Depends-on: commit 10845a105bbc ("blk-mq: skip CPU offline notify on unmapped hctx") Reviewed-by: Sesidhar Baddela Reviewed-by: Arulprabhu Ponnusamy Reviewed-by: Gian Carlo Boffa Reviewed-by: Karan Tilak Kumar Signed-off-by: Arun Easi Reviewed-by: Laurence Oberman Link: https://patch.msgid.link/20260903175547.57971-1-aeasi@cisco.com Signed-off-by: Martin K. Petersen (Oracle) [ Inlined fnic_mq_init_queue_map() into fnic_mq_map_queues_cpus() using the existing cpu variable. ] Signed-off-by: Sasha Levin Signed-off-by: Greg Kroah-Hartman --- drivers/scsi/fnic/fnic.h | 2 +- drivers/scsi/fnic/fnic_isr.c | 13 ++++++++++--- drivers/scsi/fnic/fnic_main.c | 33 ++++++++++++++++++++++++++++++++- 3 files changed, 43 insertions(+), 5 deletions(-) --- a/drivers/scsi/fnic/fnic.h +++ b/drivers/scsi/fnic/fnic.h @@ -30,7 +30,7 @@ #define DRV_NAME "fnic" #define DRV_DESCRIPTION "Cisco FCoE HBA Driver" -#define DRV_VERSION "1.9.0.0" +#define DRV_VERSION "1.9.0.1" #define PFX DRV_NAME ": " #define DFX DRV_NAME "%d: " --- a/drivers/scsi/fnic/fnic_isr.c +++ b/drivers/scsi/fnic/fnic_isr.c @@ -245,7 +245,14 @@ int fnic_set_intr_mode_msix(struct fnic unsigned int m = ARRAY_SIZE(fnic->wq); unsigned int o = ARRAY_SIZE(fnic->hw_copy_wq); unsigned int min_irqs = n + m + 1 + 1; /*rq, raw wq, wq, err*/ - + /* + * Make driver critical vectors unmanaged, or else it can get tied + * to an offline CPU. This can happen when hyper-threading is off. + */ + struct irq_affinity affd = { + .pre_vectors = n + m + 1, /* rq, raw wq, 1 ioq */ + .post_vectors = 1, /* err */ + }; /* * We need n RQs, m WQs, o Copy WQs, n+m+o CQs, and n+m+o+1 INTRs * (last INTR is used for WQ/RQ errors and notification area) @@ -263,8 +270,8 @@ int fnic_set_intr_mode_msix(struct fnic int vec_count = 0; int vecs = fnic->rq_count + fnic->raw_wq_count + fnic->wq_copy_count + 1; - vec_count = pci_alloc_irq_vectors(fnic->pdev, min_irqs, vecs, - PCI_IRQ_MSIX | PCI_IRQ_AFFINITY); + vec_count = pci_alloc_irq_vectors_affinity(fnic->pdev, min_irqs, + vecs, PCI_IRQ_MSIX|PCI_IRQ_AFFINITY, &affd); FNIC_ISR_DBG(KERN_INFO, fnic, "allocated %d MSI-X vectors\n", vec_count); --- a/drivers/scsi/fnic/fnic_main.c +++ b/drivers/scsi/fnic/fnic_main.c @@ -680,6 +680,8 @@ static int fnic_scsi_drv_init(struct fni void fnic_mq_map_queues_cpus(struct Scsi_Host *host) { + const struct cpumask *mask; + unsigned int queue, cpu; struct fnic *fnic = *((struct fnic **) shost_priv(host)); struct pci_dev *l_pdev = fnic->pdev; int intr_mode = fnic->config.intr_mode; @@ -700,7 +702,36 @@ void fnic_mq_map_queues_cpus(struct Scsi return; } - blk_mq_map_hw_queues(qmap, &l_pdev->dev, FNIC_PCI_OFFSET); + for_each_possible_cpu(cpu) + qmap->mq_map[cpu] = 0; + + /* + * Setup CPU to Queue mapping for all managed MSI-X IRQs. + * Q0 is driver critical and non-managed, hence start from Q1. + */ + for (queue = 1; queue < qmap->nr_queues; queue++) { + int irq_num = pci_irq_vector(fnic->pdev, + queue + FNIC_PCI_OFFSET); + + if (irq_num < 0) + continue; + + mask = pci_irq_get_affinity(fnic->pdev, + queue + FNIC_PCI_OFFSET); + if (!mask) { + shost_printk(KERN_ERR, host, + "failed to get irq_affinity map for queue:%d\n", irq_num); + continue; + } + FNIC_MAIN_DBG(KERN_INFO, fnic, + "got irq_affinity map for %d:\n", irq_num); + for_each_cpu(cpu, mask) { + qmap->mq_map[cpu] = qmap->queue_offset + queue; + FNIC_MAIN_DBG(KERN_INFO, fnic, + "[Q%d] cpu:%d <=> irq:%d\n", + queue, cpu, irq_num); + } + } } static int fnic_probe(struct pci_dev *pdev, const struct pci_device_id *ent)