From patchwork Wed Mar 25 09:35:00 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Julian Wiedmann X-Patchwork-Id: 221893 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=-9.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham 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 5CFD3C43331 for ; Wed, 25 Mar 2020 09:35:31 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 2189520714 for ; Wed, 25 Mar 2020 09:35:31 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726262AbgCYJfa (ORCPT ); Wed, 25 Mar 2020 05:35:30 -0400 Received: from mx0a-001b2d01.pphosted.com ([148.163.156.1]:32816 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727492AbgCYJfR (ORCPT ); Wed, 25 Mar 2020 05:35:17 -0400 Received: from pps.filterd (m0098393.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 02P9Y5Sa087666 for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from e06smtp01.uk.ibm.com (e06smtp01.uk.ibm.com [195.75.94.97]) by mx0a-001b2d01.pphosted.com with ESMTP id 2ywbthr6k0-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from localhost by e06smtp01.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 25 Mar 2020 09:35:09 -0000 Received: from b06cxnps3075.portsmouth.uk.ibm.com (9.149.109.195) by e06smtp01.uk.ibm.com (192.168.101.131) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Wed, 25 Mar 2020 09:35:07 -0000 Received: from d06av23.portsmouth.uk.ibm.com (d06av23.portsmouth.uk.ibm.com [9.149.105.59]) by b06cxnps3075.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 02P9Z9Gj63176734 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Wed, 25 Mar 2020 09:35:09 GMT Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 95E99A4053; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 51BAEA4059; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) Received: from tuxmaker.boeblingen.de.ibm.com (unknown [9.152.85.9]) by d06av23.portsmouth.uk.ibm.com (Postfix) with ESMTP; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) From: Julian Wiedmann To: David Miller Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Julian Wiedmann Subject: [PATCH v2 net-next 04/11] s390/qdio: extend polling support to multiple queues Date: Wed, 25 Mar 2020 10:35:00 +0100 X-Mailer: git-send-email 2.17.1 In-Reply-To: <20200325093507.20831-1-jwi@linux.ibm.com> References: <20200325093507.20831-1-jwi@linux.ibm.com> X-TM-AS-GCONF: 00 x-cbid: 20032509-4275-0000-0000-000003B2A6F9 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 20032509-4276-0000-0000-000038C7E46F Message-Id: <20200325093507.20831-5-jwi@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.138, 18.0.645 definitions=2020-03-25_01:2020-03-23,2020-03-25 signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 mlxscore=0 malwarescore=0 phishscore=0 impostorscore=0 priorityscore=1501 spamscore=0 bulkscore=0 mlxlogscore=999 lowpriorityscore=0 clxscore=1015 adultscore=0 suspectscore=2 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2003020000 definitions=main-2003250077 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org When the support for polling drivers was initially added, it only considered Input Queue 0. But as QDIO interrupts are actually for the full device and not a single queue, this doesn't really fit for configurations where multiple Input Queues are used. Rework the qdio code so that interrupts for a polling driver are not split up into actions for each queue. Instead deliver the interrupt as a single event, and let the driver decide which queue needs what action. When re-enabling the QDIO interrupt via qdio_start_irq(), this means that the qdio code needs to (1) put _all_ eligible queues back into a state where they raise IRQs, (2) and afterwards check _all_ eligible queues for new work to bridge the race window. On the qeth side of things (as the only qdio polling driver), we can now add CQ polling support to the main NAPI poll routine. It doesn't consume NAPI budget, and to avoid hogging the CPU we yield control after completing one full queue worth of buffers. The subsequent qdio_start_irq() will check for any additional work, and have us re-schedule the NAPI instance accordingly. Signed-off-by: Julian Wiedmann Acked-by: Heiko Carstens --- arch/s390/include/asm/qdio.h | 9 ++-- drivers/s390/cio/qdio.h | 11 +++-- drivers/s390/cio/qdio_debug.c | 4 +- drivers/s390/cio/qdio_main.c | 50 +++++++++++------------ drivers/s390/cio/qdio_setup.c | 16 ++++---- drivers/s390/cio/qdio_thinint.c | 38 ++++++++--------- drivers/s390/net/qeth_core_main.c | 68 +++++++++++++------------------ 7 files changed, 87 insertions(+), 109 deletions(-) diff --git a/arch/s390/include/asm/qdio.h b/arch/s390/include/asm/qdio.h index 1e3517b0518b..e577f8533009 100644 --- a/arch/s390/include/asm/qdio.h +++ b/arch/s390/include/asm/qdio.h @@ -338,7 +338,7 @@ typedef void qdio_handler_t(struct ccw_device *, unsigned int, int, * @no_output_qs: number of output queues * @input_handler: handler to be called for input queues * @output_handler: handler to be called for output queues - * @queue_start_poll_array: polling handlers (one per input queue or NULL) + * @irq_poll: Data IRQ polling handler (NULL when not supported) * @scan_threshold: # of in-use buffers that triggers scan on output queue * @int_parm: interruption parameter * @input_sbal_addr_array: address of no_input_qs * 128 pointers @@ -359,8 +359,7 @@ struct qdio_initialize { unsigned int no_output_qs; qdio_handler_t *input_handler; qdio_handler_t *output_handler; - void (**queue_start_poll_array) (struct ccw_device *, int, - unsigned long); + void (*irq_poll)(struct ccw_device *cdev, unsigned long data); unsigned int scan_threshold; unsigned long int_parm; struct qdio_buffer **input_sbal_addr_array; @@ -415,8 +414,8 @@ extern int qdio_activate(struct ccw_device *); extern void qdio_release_aob(struct qaob *); extern int do_QDIO(struct ccw_device *, unsigned int, int, unsigned int, unsigned int); -extern int qdio_start_irq(struct ccw_device *, int); -extern int qdio_stop_irq(struct ccw_device *, int); +extern int qdio_start_irq(struct ccw_device *cdev); +extern int qdio_stop_irq(struct ccw_device *cdev); extern int qdio_get_next_buffers(struct ccw_device *, int, int *, int *); extern int qdio_inspect_queue(struct ccw_device *cdev, unsigned int nr, bool is_input, unsigned int *bufnr, diff --git a/drivers/s390/cio/qdio.h b/drivers/s390/cio/qdio.h index ff74eb5fce50..f72f961cc78f 100644 --- a/drivers/s390/cio/qdio.h +++ b/drivers/s390/cio/qdio.h @@ -177,8 +177,8 @@ struct qdio_queue_perf_stat { unsigned int nr_sbal_total; }; -enum qdio_queue_irq_states { - QDIO_QUEUE_IRQS_DISABLED, +enum qdio_irq_poll_states { + QDIO_IRQ_DISABLED, }; struct qdio_input_q { @@ -188,10 +188,6 @@ struct qdio_input_q { int ack_count; /* last time of noticing incoming data */ u64 timestamp; - /* upper-layer polling flag */ - unsigned long queue_irq_state; - /* callback to start upper-layer polling */ - void (*queue_start_poll) (struct ccw_device *, int, unsigned long); }; struct qdio_output_q { @@ -299,6 +295,9 @@ struct qdio_irq { struct qdio_q *input_qs[QDIO_MAX_QUEUES_PER_IRQ]; struct qdio_q *output_qs[QDIO_MAX_QUEUES_PER_IRQ]; + void (*irq_poll)(struct ccw_device *cdev, unsigned long data); + unsigned long poll_state; + debug_info_t *debug_area; struct mutex setup_mutex; struct qdio_dev_perf_stat perf_stat; diff --git a/drivers/s390/cio/qdio_debug.c b/drivers/s390/cio/qdio_debug.c index 9c0370b27426..00244607c8c0 100644 --- a/drivers/s390/cio/qdio_debug.c +++ b/drivers/s390/cio/qdio_debug.c @@ -128,8 +128,8 @@ static int qstat_show(struct seq_file *m, void *v) q->u.in.ack_start, q->u.in.ack_count); seq_printf(m, "DSCI: %x IRQs disabled: %u\n", *(u8 *)q->irq_ptr->dsci, - test_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state)); + test_bit(QDIO_IRQ_DISABLED, + &q->irq_ptr->poll_state)); } seq_printf(m, "SBAL states:\n"); seq_printf(m, "|0 |8 |16 |24 |32 |40 |48 |56 63|\n"); diff --git a/drivers/s390/cio/qdio_main.c b/drivers/s390/cio/qdio_main.c index 3475317c42e5..02ced5949287 100644 --- a/drivers/s390/cio/qdio_main.c +++ b/drivers/s390/cio/qdio_main.c @@ -950,19 +950,14 @@ static void qdio_int_handler_pci(struct qdio_irq *irq_ptr) if (unlikely(irq_ptr->state != QDIO_IRQ_STATE_ACTIVE)) return; - for_each_input_queue(irq_ptr, q, i) { - if (q->u.in.queue_start_poll) { - /* skip if polling is enabled or already in work */ - if (test_and_set_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state)) { - QDIO_PERF_STAT_INC(irq_ptr, int_discarded); - continue; - } - q->u.in.queue_start_poll(q->irq_ptr->cdev, q->nr, - q->irq_ptr->int_parm); - } else { + if (irq_ptr->irq_poll) { + if (!test_and_set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state)) + irq_ptr->irq_poll(irq_ptr->cdev, irq_ptr->int_parm); + else + QDIO_PERF_STAT_INC(irq_ptr, int_discarded); + } else { + for_each_input_queue(irq_ptr, q, i) tasklet_schedule(&q->tasklet); - } } if (!pci_out_supported(irq_ptr) || !irq_ptr->scan_threshold) @@ -1610,24 +1605,26 @@ EXPORT_SYMBOL_GPL(do_QDIO); /** * qdio_start_irq - process input buffers * @cdev: associated ccw_device for the qdio subchannel - * @nr: input queue number * * Return codes * 0 - success * 1 - irqs not started since new data is available */ -int qdio_start_irq(struct ccw_device *cdev, int nr) +int qdio_start_irq(struct ccw_device *cdev) { struct qdio_q *q; struct qdio_irq *irq_ptr = cdev->private->qdio_data; + unsigned int i; if (!irq_ptr) return -ENODEV; - q = irq_ptr->input_qs[nr]; clear_nonshared_ind(irq_ptr); - qdio_stop_polling(q); - clear_bit(QDIO_QUEUE_IRQS_DISABLED, &q->u.in.queue_irq_state); + + for_each_input_queue(irq_ptr, q, i) + qdio_stop_polling(q); + + clear_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state); /* * We need to check again to not lose initiative after @@ -1635,13 +1632,16 @@ int qdio_start_irq(struct ccw_device *cdev, int nr) */ if (test_nonshared_ind(irq_ptr)) goto rescan; - if (!qdio_inbound_q_done(q, q->first_to_check)) - goto rescan; + + for_each_input_queue(irq_ptr, q, i) { + if (!qdio_inbound_q_done(q, q->first_to_check)) + goto rescan; + } + return 0; rescan: - if (test_and_set_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state)) + if (test_and_set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state)) return 0; else return 1; @@ -1729,23 +1729,19 @@ EXPORT_SYMBOL(qdio_get_next_buffers); /** * qdio_stop_irq - disable interrupt processing for the device * @cdev: associated ccw_device for the qdio subchannel - * @nr: input queue number * * Return codes * 0 - interrupts were already disabled * 1 - interrupts successfully disabled */ -int qdio_stop_irq(struct ccw_device *cdev, int nr) +int qdio_stop_irq(struct ccw_device *cdev) { - struct qdio_q *q; struct qdio_irq *irq_ptr = cdev->private->qdio_data; if (!irq_ptr) return -ENODEV; - q = irq_ptr->input_qs[nr]; - if (test_and_set_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state)) + if (test_and_set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state)) return 0; else return 1; diff --git a/drivers/s390/cio/qdio_setup.c b/drivers/s390/cio/qdio_setup.c index 66e4bdca9d89..7b831bb4e229 100644 --- a/drivers/s390/cio/qdio_setup.c +++ b/drivers/s390/cio/qdio_setup.c @@ -224,15 +224,6 @@ static void setup_queues(struct qdio_irq *irq_ptr, setup_queues_misc(q, irq_ptr, qdio_init->input_handler, i); q->is_input_q = 1; - if (qdio_init->queue_start_poll_array && - qdio_init->queue_start_poll_array[i]) { - q->u.in.queue_start_poll = - qdio_init->queue_start_poll_array[i]; - set_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state); - } else { - q->u.in.queue_start_poll = NULL; - } setup_storage_lists(q, irq_ptr, input_sbal_array, i); input_sbal_array += QDIO_MAX_BUFFERS_PER_Q; @@ -483,6 +474,13 @@ int qdio_setup_irq(struct qdio_initialize *init_data) ccw_device_get_schid(irq_ptr->cdev, &irq_ptr->schid); setup_queues(irq_ptr, init_data); + if (init_data->irq_poll) { + irq_ptr->irq_poll = init_data->irq_poll; + set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state); + } else { + irq_ptr->irq_poll = NULL; + } + setup_qib(irq_ptr, init_data); qdio_setup_thinint(irq_ptr); set_impl_params(irq_ptr, init_data->qib_param_field_format, diff --git a/drivers/s390/cio/qdio_thinint.c b/drivers/s390/cio/qdio_thinint.c index 7c4e4ec08a12..8f315c53de23 100644 --- a/drivers/s390/cio/qdio_thinint.c +++ b/drivers/s390/cio/qdio_thinint.c @@ -135,28 +135,24 @@ static inline void tiqdio_call_inq_handlers(struct qdio_irq *irq) has_multiple_inq_on_dsci(irq)) xchg(irq->dsci, 0); + if (irq->irq_poll) { + if (!test_and_set_bit(QDIO_IRQ_DISABLED, &irq->poll_state)) + irq->irq_poll(irq->cdev, irq->int_parm); + else + QDIO_PERF_STAT_INC(irq, int_discarded); + + return; + } + for_each_input_queue(irq, q, i) { - if (q->u.in.queue_start_poll) { - /* skip if polling is enabled or already in work */ - if (test_and_set_bit(QDIO_QUEUE_IRQS_DISABLED, - &q->u.in.queue_irq_state)) { - QDIO_PERF_STAT_INC(irq, int_discarded); - continue; - } - - /* avoid dsci clear here, done after processing */ - q->u.in.queue_start_poll(irq->cdev, q->nr, - irq->int_parm); - } else { - if (!shared_ind(irq)) - xchg(irq->dsci, 0); - - /* - * Call inbound processing but not directly - * since that could starve other thinint queues. - */ - tasklet_schedule(&q->tasklet); - } + if (!shared_ind(irq)) + xchg(irq->dsci, 0); + + /* + * Call inbound processing but not directly + * since that could starve other thinint queues. + */ + tasklet_schedule(&q->tasklet); } } diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c index 13facf2d602b..68bba8d057d1 100644 --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -548,14 +548,6 @@ static void qeth_qdio_handle_aob(struct qeth_card *card, qdio_release_aob(aob); } -static inline int qeth_is_cq(struct qeth_card *card, unsigned int queue) -{ - return card->options.cq == QETH_CQ_ENABLED && - card->qdio.c_q != NULL && - queue != 0 && - queue == card->qdio.no_in_queues - 1; -} - static void qeth_setup_ccw(struct ccw1 *ccw, u8 cmd_code, u8 flags, u32 len, void *data) { @@ -3469,8 +3461,7 @@ static void qeth_check_outbound_queue(struct qeth_qdio_out_q *queue) } } -static void qeth_qdio_start_poll(struct ccw_device *ccwdev, int queue, - unsigned long card_ptr) +static void qeth_qdio_poll(struct ccw_device *cdev, unsigned long card_ptr) { struct qeth_card *card = (struct qeth_card *)card_ptr; @@ -3508,9 +3499,6 @@ static void qeth_qdio_cq_handler(struct qeth_card *card, unsigned int qdio_err, int i; int rc; - if (!qeth_is_cq(card, queue)) - return; - QETH_CARD_TEXT_(card, 5, "qcqhe%d", first_element); QETH_CARD_TEXT_(card, 5, "qcqhc%d", count); QETH_CARD_TEXT_(card, 5, "qcqherr%d", qdio_err); @@ -3556,9 +3544,7 @@ static void qeth_qdio_input_handler(struct ccw_device *ccwdev, QETH_CARD_TEXT_(card, 2, "qihq%d", queue); QETH_CARD_TEXT_(card, 2, "qiec%d", qdio_err); - if (qeth_is_cq(card, queue)) - qeth_qdio_cq_handler(card, qdio_err, queue, first_elem, count); - else if (qdio_err) + if (qdio_err) qeth_schedule_recovery(card); } @@ -4805,10 +4791,7 @@ static void qeth_determine_capabilities(struct qeth_card *card) } static void qeth_qdio_establish_cq(struct qeth_card *card, - struct qdio_buffer **in_sbal_ptrs, - void (**queue_start_poll) - (struct ccw_device *, int, - unsigned long)) + struct qdio_buffer **in_sbal_ptrs) { int i; @@ -4819,8 +4802,6 @@ static void qeth_qdio_establish_cq(struct qeth_card *card, for (i = 0; i < QDIO_MAX_BUFFERS_PER_Q; i++) in_sbal_ptrs[offset + i] = card->qdio.c_q->bufs[i].buffer; - - queue_start_poll[card->qdio.no_in_queues - 1] = NULL; } } @@ -4829,7 +4810,6 @@ static int qeth_qdio_establish(struct qeth_card *card) struct qdio_initialize init_data; char *qib_param_field; struct qdio_buffer **in_sbal_ptrs; - void (**queue_start_poll) (struct ccw_device *, int, unsigned long); struct qdio_buffer **out_sbal_ptrs; int i, j, k; int rc = 0; @@ -4856,16 +4836,7 @@ static int qeth_qdio_establish(struct qeth_card *card) for (i = 0; i < QDIO_MAX_BUFFERS_PER_Q; i++) in_sbal_ptrs[i] = card->qdio.in_q->bufs[i].buffer; - queue_start_poll = kcalloc(card->qdio.no_in_queues, sizeof(void *), - GFP_KERNEL); - if (!queue_start_poll) { - rc = -ENOMEM; - goto out_free_in_sbals; - } - for (i = 0; i < card->qdio.no_in_queues; ++i) - queue_start_poll[i] = qeth_qdio_start_poll; - - qeth_qdio_establish_cq(card, in_sbal_ptrs, queue_start_poll); + qeth_qdio_establish_cq(card, in_sbal_ptrs); out_sbal_ptrs = kcalloc(card->qdio.no_out_queues * QDIO_MAX_BUFFERS_PER_Q, @@ -4873,7 +4844,7 @@ static int qeth_qdio_establish(struct qeth_card *card) GFP_KERNEL); if (!out_sbal_ptrs) { rc = -ENOMEM; - goto out_free_queue_start_poll; + goto out_free_in_sbals; } for (i = 0, k = 0; i < card->qdio.no_out_queues; ++i) @@ -4891,7 +4862,7 @@ static int qeth_qdio_establish(struct qeth_card *card) init_data.no_output_qs = card->qdio.no_out_queues; init_data.input_handler = qeth_qdio_input_handler; init_data.output_handler = qeth_qdio_output_handler; - init_data.queue_start_poll_array = queue_start_poll; + init_data.irq_poll = qeth_qdio_poll; init_data.int_parm = (unsigned long) card; init_data.input_sbal_addr_array = in_sbal_ptrs; init_data.output_sbal_addr_array = out_sbal_ptrs; @@ -4924,8 +4895,6 @@ static int qeth_qdio_establish(struct qeth_card *card) } out: kfree(out_sbal_ptrs); -out_free_queue_start_poll: - kfree(queue_start_poll); out_free_in_sbals: kfree(in_sbal_ptrs); out_free_qib_param: @@ -5581,6 +5550,24 @@ static unsigned int qeth_rx_poll(struct qeth_card *card, int budget) return work_done; } +static void qeth_cq_poll(struct qeth_card *card) +{ + unsigned int work_done = 0; + + while (work_done < QDIO_MAX_BUFFERS_PER_Q) { + unsigned int start, error; + int completed; + + completed = qdio_inspect_queue(CARD_DDEV(card), 1, true, &start, + &error); + if (completed <= 0) + return; + + qeth_qdio_cq_handler(card, error, 1, start, completed); + work_done += completed; + } +} + int qeth_poll(struct napi_struct *napi, int budget) { struct qeth_card *card = container_of(napi, struct qeth_card, napi); @@ -5588,12 +5575,15 @@ int qeth_poll(struct napi_struct *napi, int budget) work_done = qeth_rx_poll(card, budget); + if (card->options.cq == QETH_CQ_ENABLED) + qeth_cq_poll(card); + /* Exhausted the RX budget. Keep IRQ disabled, we get called again. */ if (budget && work_done >= budget) return work_done; if (napi_complete_done(napi, work_done) && - qdio_start_irq(CARD_DDEV(card), 0)) + qdio_start_irq(CARD_DDEV(card))) napi_schedule(napi); return work_done; @@ -6756,7 +6746,7 @@ int qeth_stop(struct net_device *dev) } napi_disable(&card->napi); - qdio_stop_irq(CARD_DDEV(card), 0); + qdio_stop_irq(CARD_DDEV(card)); return 0; } From patchwork Wed Mar 25 09:35:01 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Julian Wiedmann X-Patchwork-Id: 221894 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=-9.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham 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 327D8C1975A for ; Wed, 25 Mar 2020 09:35:26 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 106FE20772 for ; Wed, 25 Mar 2020 09:35:26 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727634AbgCYJfZ (ORCPT ); Wed, 25 Mar 2020 05:35:25 -0400 Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]:63428 "EHLO mx0b-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727612AbgCYJfX (ORCPT ); Wed, 25 Mar 2020 05:35:23 -0400 Received: from pps.filterd (m0127361.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 02P9XaPu168479 for ; Wed, 25 Mar 2020 05:35:22 -0400 Received: from e06smtp01.uk.ibm.com (e06smtp01.uk.ibm.com [195.75.94.97]) by mx0a-001b2d01.pphosted.com with ESMTP id 2ywdr74g03-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Wed, 25 Mar 2020 05:35:20 -0400 Received: from localhost by e06smtp01.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 25 Mar 2020 09:35:10 -0000 Received: from b06cxnps3075.portsmouth.uk.ibm.com (9.149.109.195) by e06smtp01.uk.ibm.com (192.168.101.131) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Wed, 25 Mar 2020 09:35:08 -0000 Received: from d06av23.portsmouth.uk.ibm.com (d06av23.portsmouth.uk.ibm.com [9.149.105.59]) by b06cxnps3075.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 02P9ZABu57999384 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Wed, 25 Mar 2020 09:35:10 GMT Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id D682FA404D; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id A1109A4055; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) Received: from tuxmaker.boeblingen.de.ibm.com (unknown [9.152.85.9]) by d06av23.portsmouth.uk.ibm.com (Postfix) with ESMTP; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) From: Julian Wiedmann To: David Miller Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Julian Wiedmann Subject: [PATCH v2 net-next 05/11] s390/qeth: simplify L3 dev_id logic Date: Wed, 25 Mar 2020 10:35:01 +0100 X-Mailer: git-send-email 2.17.1 In-Reply-To: <20200325093507.20831-1-jwi@linux.ibm.com> References: <20200325093507.20831-1-jwi@linux.ibm.com> X-TM-AS-GCONF: 00 x-cbid: 20032509-4275-0000-0000-000003B2A6FB X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 20032509-4276-0000-0000-000038C7E471 Message-Id: <20200325093507.20831-6-jwi@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.138, 18.0.645 definitions=2020-03-25_01:2020-03-23,2020-03-25 signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 malwarescore=0 bulkscore=0 adultscore=0 impostorscore=0 lowpriorityscore=0 phishscore=0 priorityscore=1501 clxscore=1015 suspectscore=0 mlxlogscore=857 mlxscore=0 spamscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2003020000 definitions=main-2003250077 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org The logic that deals with errors from qeth_l3_get_unique_id() is quite complex: it sets card->unique_id to 0xfffe, additionally flags it as UNIQUE_ID_NOT_BY_CARD and later takes this flag as cue to not propagate card->unique_id to dev->dev_id. With dev->dev_id thus holding 0, addrconf_ifid_eui48() applies its default behaviour. Get rid of all the special bit masks, and just return the old uid in case of an error. For the vast majority of cases this will be 0 (and so we still get the desired default behaviour) - with the rare exception where qeth_l3_get_unique_id() might have been called earlier but the initialization then failed at a later point. Signed-off-by: Julian Wiedmann --- drivers/s390/net/qeth_core.h | 5 ----- drivers/s390/net/qeth_l3_main.c | 30 +++++++++++++----------------- 2 files changed, 13 insertions(+), 22 deletions(-) diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h index 9840d4fab010..257b7f3c5558 100644 --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -178,10 +178,6 @@ struct qeth_vnicc_info { #define QETH_RECLAIM_WORK_TIME HZ #define QETH_MAX_PORTNO 15 -/*IPv6 address autoconfiguration stuff*/ -#define UNIQUE_ID_IF_CREATE_ADDR_FAILED 0xfffe -#define UNIQUE_ID_NOT_BY_CARD 0x10000 - /*****************************************************************************/ /* QDIO queue and buffer handling */ /*****************************************************************************/ @@ -687,7 +683,6 @@ struct qeth_card_info { enum qeth_card_types type; enum qeth_link_types link_type; int broadcast_capable; - int unique_id; bool layer_enforced; struct qeth_card_blkt blkt; __u32 diagass_support; diff --git a/drivers/s390/net/qeth_l3_main.c b/drivers/s390/net/qeth_l3_main.c index 83ae75cf1389..b48cd0df3e31 100644 --- a/drivers/s390/net/qeth_l3_main.c +++ b/drivers/s390/net/qeth_l3_main.c @@ -949,39 +949,36 @@ static int qeth_l3_get_unique_id_cb(struct qeth_card *card, struct qeth_reply *reply, unsigned long data) { struct qeth_ipa_cmd *cmd = (struct qeth_ipa_cmd *) data; + u16 *uid = reply->param; if (cmd->hdr.return_code == 0) { - card->info.unique_id = cmd->data.create_destroy_addr.uid; + *uid = cmd->data.create_destroy_addr.uid; return 0; } - card->info.unique_id = UNIQUE_ID_IF_CREATE_ADDR_FAILED | - UNIQUE_ID_NOT_BY_CARD; dev_warn(&card->gdev->dev, "The network adapter failed to generate a unique ID\n"); return -EIO; } -static int qeth_l3_get_unique_id(struct qeth_card *card) +static u16 qeth_l3_get_unique_id(struct qeth_card *card, u16 uid) { - int rc = 0; struct qeth_cmd_buffer *iob; QETH_CARD_TEXT(card, 2, "guniqeid"); - if (!qeth_is_supported(card, IPA_IPV6)) { - card->info.unique_id = UNIQUE_ID_IF_CREATE_ADDR_FAILED | - UNIQUE_ID_NOT_BY_CARD; - return 0; - } + if (!qeth_is_supported(card, IPA_IPV6)) + goto out; iob = qeth_ipa_alloc_cmd(card, IPA_CMD_CREATE_ADDR, QETH_PROT_IPV6, IPA_DATA_SIZEOF(create_destroy_addr)); if (!iob) - return -ENOMEM; + goto out; - __ipa_cmd(iob)->data.create_destroy_addr.uid = card->info.unique_id; - rc = qeth_send_ipa_cmd(card, iob, qeth_l3_get_unique_id_cb, NULL); - return rc; + __ipa_cmd(iob)->data.create_destroy_addr.uid = uid; + qeth_send_ipa_cmd(card, iob, qeth_l3_get_unique_id_cb, &uid); + +out: + return uid; } static int @@ -1920,6 +1917,7 @@ static const struct net_device_ops qeth_l3_osa_netdev_ops = { static int qeth_l3_setup_netdev(struct qeth_card *card) { + struct net_device *dev = card->dev; unsigned int headroom; int rc; @@ -1937,9 +1935,7 @@ static int qeth_l3_setup_netdev(struct qeth_card *card) card->dev->netdev_ops = &qeth_l3_osa_netdev_ops; /*IPv6 address autoconfiguration stuff*/ - qeth_l3_get_unique_id(card); - if (!(card->info.unique_id & UNIQUE_ID_NOT_BY_CARD)) - card->dev->dev_id = card->info.unique_id & 0xffff; + dev->dev_id = qeth_l3_get_unique_id(card, dev->dev_id); if (!IS_VM_NIC(card)) { card->dev->features |= NETIF_F_SG; From patchwork Wed Mar 25 09:35:02 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Julian Wiedmann X-Patchwork-Id: 221892 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=-9.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham 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 CEA02C54FD0 for ; Wed, 25 Mar 2020 09:35:34 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id AF94420714 for ; Wed, 25 Mar 2020 09:35:34 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727658AbgCYJfd (ORCPT ); Wed, 25 Mar 2020 05:35:33 -0400 Received: from mx0a-001b2d01.pphosted.com ([148.163.156.1]:39314 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727494AbgCYJfQ (ORCPT ); Wed, 25 Mar 2020 05:35:16 -0400 Received: from pps.filterd (m0098393.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 02P9Y5Li087676 for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from e06smtp01.uk.ibm.com (e06smtp01.uk.ibm.com [195.75.94.97]) by mx0a-001b2d01.pphosted.com with ESMTP id 2ywbthr6ks-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from localhost by e06smtp01.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 25 Mar 2020 09:35:10 -0000 Received: from b06cxnps4074.portsmouth.uk.ibm.com (9.149.109.196) by e06smtp01.uk.ibm.com (192.168.101.131) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Wed, 25 Mar 2020 09:35:08 -0000 Received: from d06av23.portsmouth.uk.ibm.com (d06av23.portsmouth.uk.ibm.com [9.149.105.59]) by b06cxnps4074.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 02P9ZAo219398856 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Wed, 25 Mar 2020 09:35:10 GMT Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 22A43A4053; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id E18A2A4051; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) Received: from tuxmaker.boeblingen.de.ibm.com (unknown [9.152.85.9]) by d06av23.portsmouth.uk.ibm.com (Postfix) with ESMTP; Wed, 25 Mar 2020 09:35:09 +0000 (GMT) From: Julian Wiedmann To: David Miller Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Julian Wiedmann Subject: [PATCH v2 net-next 06/11] s390/qeth: clean up the mac_bits Date: Wed, 25 Mar 2020 10:35:02 +0100 X-Mailer: git-send-email 2.17.1 In-Reply-To: <20200325093507.20831-1-jwi@linux.ibm.com> References: <20200325093507.20831-1-jwi@linux.ibm.com> X-TM-AS-GCONF: 00 x-cbid: 20032509-4275-0000-0000-000003B2A6FA X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 20032509-4276-0000-0000-000038C7E470 Message-Id: <20200325093507.20831-7-jwi@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.138, 18.0.645 definitions=2020-03-25_01:2020-03-23,2020-03-25 signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 mlxscore=0 malwarescore=0 phishscore=0 impostorscore=0 priorityscore=1501 spamscore=0 bulkscore=0 mlxlogscore=999 lowpriorityscore=0 clxscore=1015 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2003020000 definitions=main-2003250077 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org We're down to a single bit flag for MAC-address related status, reflect that in the info struct. Also set up the flag during initialization instead of clearing it during shutdown - one more little step towards unifying the shutdown code. Signed-off-by: Julian Wiedmann --- drivers/s390/net/qeth_core.h | 3 +-- drivers/s390/net/qeth_l2_main.c | 13 +++++++------ 2 files changed, 8 insertions(+), 8 deletions(-) diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h index 257b7f3c5558..911dcef6adc6 100644 --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -668,18 +668,17 @@ struct qeth_card_blkt { #define QETH_BROADCAST_WITH_ECHO 0x01 #define QETH_BROADCAST_WITHOUT_ECHO 0x02 -#define QETH_LAYER2_MAC_REGISTERED 0x02 struct qeth_card_info { unsigned short unit_addr2; unsigned short cula; u8 chpid; __u16 func_level; char mcl_level[QETH_MCL_LENGTH + 1]; + u8 dev_addr_is_registered:1; u8 open_when_online:1; u8 promisc_mode:1; u8 use_v1_blkt:1; u8 is_vm_nic:1; - int mac_bits; enum qeth_card_types type; enum qeth_link_types link_type; int broadcast_capable; diff --git a/drivers/s390/net/qeth_l2_main.c b/drivers/s390/net/qeth_l2_main.c index 73cb363b1fab..249b00d91d46 100644 --- a/drivers/s390/net/qeth_l2_main.c +++ b/drivers/s390/net/qeth_l2_main.c @@ -291,7 +291,6 @@ static void qeth_l2_stop_card(struct qeth_card *card) qeth_qdio_clear_card(card, 0); qeth_clear_working_pool_list(card); flush_workqueue(card->event_wq); - card->info.mac_bits &= ~QETH_LAYER2_MAC_REGISTERED; card->info.promisc_mode = 0; } @@ -337,14 +336,16 @@ static void qeth_l2_register_dev_addr(struct qeth_card *card) qeth_l2_request_initial_mac(card); if (!IS_OSN(card) && !qeth_l2_send_setmac(card, card->dev->dev_addr)) - card->info.mac_bits |= QETH_LAYER2_MAC_REGISTERED; + card->info.dev_addr_is_registered = 1; + else + card->info.dev_addr_is_registered = 0; } static int qeth_l2_validate_addr(struct net_device *dev) { struct qeth_card *card = dev->ml_priv; - if (card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED) + if (card->info.dev_addr_is_registered) return eth_validate_addr(dev); QETH_CARD_TEXT(card, 4, "nomacadr"); @@ -370,7 +371,7 @@ static int qeth_l2_set_mac_address(struct net_device *dev, void *p) /* don't register the same address twice */ if (ether_addr_equal_64bits(dev->dev_addr, addr->sa_data) && - (card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED)) + card->info.dev_addr_is_registered) return 0; /* add the new address, switch over, drop the old */ @@ -380,9 +381,9 @@ static int qeth_l2_set_mac_address(struct net_device *dev, void *p) ether_addr_copy(old_addr, dev->dev_addr); ether_addr_copy(dev->dev_addr, addr->sa_data); - if (card->info.mac_bits & QETH_LAYER2_MAC_REGISTERED) + if (card->info.dev_addr_is_registered) qeth_l2_remove_mac(card, old_addr); - card->info.mac_bits |= QETH_LAYER2_MAC_REGISTERED; + card->info.dev_addr_is_registered = 1; return 0; } From patchwork Wed Mar 25 09:35:03 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Julian Wiedmann X-Patchwork-Id: 221896 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=-9.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham 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 A2BDAC1975A for ; Wed, 25 Mar 2020 09:35:20 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 7B7BE20714 for ; Wed, 25 Mar 2020 09:35:20 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727590AbgCYJfT (ORCPT ); Wed, 25 Mar 2020 05:35:19 -0400 Received: from mx0b-001b2d01.pphosted.com ([148.163.158.5]:2906 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1727399AbgCYJfS (ORCPT ); Wed, 25 Mar 2020 05:35:18 -0400 Received: from pps.filterd (m0098421.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 02P9YeLm074466 for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from e06smtp03.uk.ibm.com (e06smtp03.uk.ibm.com [195.75.94.99]) by mx0a-001b2d01.pphosted.com with ESMTP id 2ywewv52j4-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Wed, 25 Mar 2020 05:35:16 -0400 Received: from localhost by e06smtp03.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 25 Mar 2020 09:35:13 -0000 Received: from b06cxnps4074.portsmouth.uk.ibm.com (9.149.109.196) by e06smtp03.uk.ibm.com (192.168.101.133) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Wed, 25 Mar 2020 09:35:10 -0000 Received: from d06av23.portsmouth.uk.ibm.com (d06av23.portsmouth.uk.ibm.com [9.149.105.59]) by b06cxnps4074.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 02P9ZAdb39059692 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Wed, 25 Mar 2020 09:35:10 GMT Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 6493CA4053; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 2DDE1A405E; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) Received: from tuxmaker.boeblingen.de.ibm.com (unknown [9.152.85.9]) by d06av23.portsmouth.uk.ibm.com (Postfix) with ESMTP; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) From: Julian Wiedmann To: David Miller Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Julian Wiedmann Subject: [PATCH v2 net-next 07/11] s390/qeth: collect more TX statistics Date: Wed, 25 Mar 2020 10:35:03 +0100 X-Mailer: git-send-email 2.17.1 In-Reply-To: <20200325093507.20831-1-jwi@linux.ibm.com> References: <20200325093507.20831-1-jwi@linux.ibm.com> X-TM-AS-GCONF: 00 x-cbid: 20032509-0012-0000-0000-0000039760F4 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 20032509-0013-0000-0000-000021D457F3 Message-Id: <20200325093507.20831-8-jwi@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.138, 18.0.645 definitions=2020-03-25_01:2020-03-23,2020-03-25 signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 lowpriorityscore=0 suspectscore=0 spamscore=0 clxscore=1015 malwarescore=0 impostorscore=0 bulkscore=0 phishscore=0 mlxscore=0 adultscore=0 mlxlogscore=999 priorityscore=1501 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2003020000 definitions=main-2003250077 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Count the number of TX doorbells we issue to the qdio layer. Also count the number of actual frames in a TX buffer, and then use this data along with the byte count during TX completion. We'll make additional use of the frame count in a subsequent patch. Signed-off-by: Julian Wiedmann --- drivers/s390/net/qeth_core.h | 2 ++ drivers/s390/net/qeth_core_main.c | 20 ++++++++++++++------ drivers/s390/net/qeth_ethtool.c | 1 + 3 files changed, 17 insertions(+), 6 deletions(-) diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h index 911dcef6adc6..f56670bfcd0a 100644 --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -402,6 +402,7 @@ struct qeth_qdio_out_buffer { struct qdio_buffer *buffer; atomic_t state; int next_element_to_fill; + unsigned int frames; unsigned int bytes; struct sk_buff_head skb_list; int is_header[QDIO_MAX_ELEMENTS_PER_BUFFER]; @@ -457,6 +458,7 @@ struct qeth_out_q_stats { u64 tso_bytes; u64 packing_mode_switch; u64 stopped; + u64 doorbell; u64 completion_yield; u64 completion_timer; diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c index 68bba8d057d1..108dd9a34f30 100644 --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -1155,17 +1155,20 @@ static void qeth_tx_complete_buf(struct qeth_qdio_out_buffer *buf, bool error, QETH_TXQ_STAT_INC(queue, bufs); QETH_TXQ_STAT_ADD(queue, buf_elements, buf->next_element_to_fill); + if (error) { + QETH_TXQ_STAT_ADD(queue, tx_errors, buf->frames); + } else { + QETH_TXQ_STAT_ADD(queue, tx_packets, buf->frames); + QETH_TXQ_STAT_ADD(queue, tx_bytes, buf->bytes); + } + while ((skb = __skb_dequeue(&buf->skb_list)) != NULL) { unsigned int bytes = qdisc_pkt_len(skb); bool is_tso = skb_is_gso(skb); unsigned int packets; packets = is_tso ? skb_shinfo(skb)->gso_segs : 1; - if (error) { - QETH_TXQ_STAT_ADD(queue, tx_errors, packets); - } else { - QETH_TXQ_STAT_ADD(queue, tx_packets, packets); - QETH_TXQ_STAT_ADD(queue, tx_bytes, bytes); + if (!error) { if (skb->ip_summed == CHECKSUM_PARTIAL) QETH_TXQ_STAT_ADD(queue, skbs_csum, packets); if (skb_is_nonlinear(skb)) @@ -1202,6 +1205,7 @@ static void qeth_clear_output_buffer(struct qeth_qdio_out_q *queue, qeth_scrub_qdio_buffer(buf->buffer, queue->max_elements); buf->next_element_to_fill = 0; + buf->frames = 0; buf->bytes = 0; atomic_set(&buf->state, QETH_QDIO_BUF_EMPTY); } @@ -3389,6 +3393,7 @@ static void qeth_flush_buffers(struct qeth_qdio_out_q *queue, int index, } } + QETH_TXQ_STAT_INC(queue, doorbell); qdio_flags = QDIO_FLAG_SYNC_OUTPUT; if (atomic_read(&queue->set_pci_flags_count)) qdio_flags |= QDIO_FLAG_PCI_OUT; @@ -3942,6 +3947,7 @@ static int __qeth_xmit(struct qeth_card *card, struct qeth_qdio_out_q *queue, next_element = qeth_fill_buffer(buffer, skb, hdr, offset, hd_len); buffer->bytes += bytes; + buffer->frames += skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1; queue->prev_hdr = hdr; flush = __netdev_tx_sent_queue(txq, bytes, @@ -4032,6 +4038,8 @@ int qeth_do_send_packet(struct qeth_card *card, struct qeth_qdio_out_q *queue, } next_element = qeth_fill_buffer(buffer, skb, hdr, offset, hd_len); + buffer->bytes += qdisc_pkt_len(skb); + buffer->frames += skb_is_gso(skb) ? skb_shinfo(skb)->gso_segs : 1; if (queue->do_pack) QETH_TXQ_STAT_INC(queue, skbs_pack); @@ -5668,7 +5676,7 @@ static int qeth_tx_poll(struct napi_struct *napi, int budget) unsigned int bidx = QDIO_BUFNR(i); buffer = queue->bufs[bidx]; - packets += skb_queue_len(&buffer->skb_list); + packets += buffer->frames; bytes += buffer->bytes; qeth_handle_send_error(card, buffer, error); diff --git a/drivers/s390/net/qeth_ethtool.c b/drivers/s390/net/qeth_ethtool.c index 31e019085fc3..6f0cc6fcc759 100644 --- a/drivers/s390/net/qeth_ethtool.c +++ b/drivers/s390/net/qeth_ethtool.c @@ -39,6 +39,7 @@ static const struct qeth_stats txq_stats[] = { QETH_TXQ_STAT("TSO bytes", tso_bytes), QETH_TXQ_STAT("Packing mode switches", packing_mode_switch), QETH_TXQ_STAT("Queue stopped", stopped), + QETH_TXQ_STAT("Doorbell", doorbell), QETH_TXQ_STAT("Completion yield", completion_yield), QETH_TXQ_STAT("Completion timer", completion_timer), }; From patchwork Wed Mar 25 09:35:04 2020 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Julian Wiedmann X-Patchwork-Id: 221895 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=-9.8 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH, MAILING_LIST_MULTI, SIGNED_OFF_BY, SPF_HELO_NONE, SPF_PASS, USER_AGENT_GIT autolearn=ham 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 993F6C43331 for ; Wed, 25 Mar 2020 09:35:22 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 6533520714 for ; Wed, 25 Mar 2020 09:35:22 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727566AbgCYJfW (ORCPT ); Wed, 25 Mar 2020 05:35:22 -0400 Received: from mx0a-001b2d01.pphosted.com ([148.163.156.1]:49730 "EHLO mx0a-001b2d01.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1727285AbgCYJfT (ORCPT ); Wed, 25 Mar 2020 05:35:19 -0400 Received: from pps.filterd (m0098393.ppops.net [127.0.0.1]) by mx0a-001b2d01.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 02P9Y4TR087551 for ; Wed, 25 Mar 2020 05:35:18 -0400 Received: from e06smtp02.uk.ibm.com (e06smtp02.uk.ibm.com [195.75.94.98]) by mx0a-001b2d01.pphosted.com with ESMTP id 2ywbthr6ms-1 (version=TLSv1.2 cipher=AES256-GCM-SHA384 bits=256 verify=NOT) for ; Wed, 25 Mar 2020 05:35:18 -0400 Received: from localhost by e06smtp02.uk.ibm.com with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted for from ; Wed, 25 Mar 2020 09:35:13 -0000 Received: from b06cxnps4074.portsmouth.uk.ibm.com (9.149.109.196) by e06smtp02.uk.ibm.com (192.168.101.132) with IBM ESMTP SMTP Gateway: Authorized Use Only! Violators will be prosecuted; (version=TLSv1/SSLv3 cipher=AES256-GCM-SHA384 bits=256/256) Wed, 25 Mar 2020 09:35:09 -0000 Received: from d06av23.portsmouth.uk.ibm.com (d06av23.portsmouth.uk.ibm.com [9.149.105.59]) by b06cxnps4074.portsmouth.uk.ibm.com (8.14.9/8.14.9/NCO v10.0) with ESMTP id 02P9ZAtt54919404 (version=TLSv1/SSLv3 cipher=DHE-RSA-AES256-GCM-SHA384 bits=256 verify=OK); Wed, 25 Mar 2020 09:35:10 GMT Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id AACE4A405B; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) Received: from d06av23.portsmouth.uk.ibm.com (unknown [127.0.0.1]) by IMSVA (Postfix) with ESMTP id 6FE9FA4055; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) Received: from tuxmaker.boeblingen.de.ibm.com (unknown [9.152.85.9]) by d06av23.portsmouth.uk.ibm.com (Postfix) with ESMTP; Wed, 25 Mar 2020 09:35:10 +0000 (GMT) From: Julian Wiedmann To: David Miller Cc: netdev , linux-s390 , Heiko Carstens , Ursula Braun , Julian Wiedmann , Jakub Kicinski Subject: [PATCH v2 net-next 08/11] s390/qeth: add TX IRQ coalescing support for IQD devices Date: Wed, 25 Mar 2020 10:35:04 +0100 X-Mailer: git-send-email 2.17.1 In-Reply-To: <20200325093507.20831-1-jwi@linux.ibm.com> References: <20200325093507.20831-1-jwi@linux.ibm.com> X-TM-AS-GCONF: 00 x-cbid: 20032509-0008-0000-0000-000003638B59 X-IBM-AV-DETECTION: SAVI=unused REMOTE=unused XFE=unused x-cbparentid: 20032509-0009-0000-0000-00004A84FA88 Message-Id: <20200325093507.20831-9-jwi@linux.ibm.com> X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.138, 18.0.645 definitions=2020-03-25_01:2020-03-23,2020-03-25 signatures=0 X-Proofpoint-Spam-Details: rule=outbound_notspam policy=outbound score=0 mlxscore=0 malwarescore=0 phishscore=0 impostorscore=0 priorityscore=1501 spamscore=0 bulkscore=0 mlxlogscore=999 lowpriorityscore=0 clxscore=1015 adultscore=0 suspectscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2003020000 definitions=main-2003250077 Sender: netdev-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org Since IQD devices complete (most of) their transmissions synchronously, they don't offer TX completion IRQs and have no HW coalescing controls. But we can fake the easy parts in SW, and give the user some control wrt to how often the TX NAPI code should be triggered to process the TX completions. Having per-queue controls can in particular help the dedicated mcast queue, as it likely benefits from different fine-tuning than what the ucast queues need. CC: Jakub Kicinski Signed-off-by: Julian Wiedmann --- drivers/s390/net/qeth_core.h | 15 ++++--- drivers/s390/net/qeth_core_main.c | 20 +++++++-- drivers/s390/net/qeth_ethtool.c | 75 +++++++++++++++++++++++++++++++ 3 files changed, 102 insertions(+), 8 deletions(-) diff --git a/drivers/s390/net/qeth_core.h b/drivers/s390/net/qeth_core.h index f56670bfcd0a..a6553e78af08 100644 --- a/drivers/s390/net/qeth_core.h +++ b/drivers/s390/net/qeth_core.h @@ -459,6 +459,7 @@ struct qeth_out_q_stats { u64 packing_mode_switch; u64 stopped; u64 doorbell; + u64 coal_frames; u64 completion_yield; u64 completion_timer; @@ -469,6 +470,8 @@ struct qeth_out_q_stats { u64 tx_dropped; }; +#define QETH_TX_MAX_COALESCED_FRAMES 1 +#define QETH_TX_COALESCE_USECS 25 #define QETH_TX_TIMER_USECS 500 struct qeth_qdio_out_q { @@ -492,9 +495,13 @@ struct qeth_qdio_out_q { struct napi_struct napi; struct timer_list timer; struct qeth_hdr *prev_hdr; + unsigned int coalesced_frames; u8 bulk_start; u8 bulk_count; u8 bulk_max; + + unsigned int coalesce_usecs; + unsigned int max_coalesced_frames; }; #define qeth_for_each_output_queue(card, q, i) \ @@ -503,12 +510,10 @@ struct qeth_qdio_out_q { #define qeth_napi_to_out_queue(n) container_of(n, struct qeth_qdio_out_q, napi) -static inline void qeth_tx_arm_timer(struct qeth_qdio_out_q *queue) +static inline void qeth_tx_arm_timer(struct qeth_qdio_out_q *queue, + unsigned long usecs) { - if (timer_pending(&queue->timer)) - return; - mod_timer(&queue->timer, usecs_to_jiffies(QETH_TX_TIMER_USECS) + - jiffies); + timer_reduce(&queue->timer, usecs_to_jiffies(usecs) + jiffies); } static inline bool qeth_out_queue_is_full(struct qeth_qdio_out_q *queue) diff --git a/drivers/s390/net/qeth_core_main.c b/drivers/s390/net/qeth_core_main.c index 108dd9a34f30..0c9f1464a778 100644 --- a/drivers/s390/net/qeth_core_main.c +++ b/drivers/s390/net/qeth_core_main.c @@ -2404,6 +2404,8 @@ static int qeth_alloc_qdio_queues(struct qeth_card *card) queue->card = card; queue->queue_no = i; timer_setup(&queue->timer, qeth_tx_completion_timer, 0); + queue->coalesce_usecs = QETH_TX_COALESCE_USECS; + queue->max_coalesced_frames = QETH_TX_MAX_COALESCED_FRAMES; /* give outbound qeth_qdio_buffers their qdio_buffers */ for (j = 0; j < QDIO_MAX_BUFFERS_PER_Q; ++j) { @@ -2762,6 +2764,7 @@ static int qeth_init_qdio_queues(struct qeth_card *card) queue->next_buf_to_fill = 0; queue->do_pack = 0; queue->prev_hdr = NULL; + queue->coalesced_frames = 0; queue->bulk_start = 0; queue->bulk_count = 0; queue->bulk_max = qeth_tx_select_bulk_max(card, queue); @@ -3357,6 +3360,7 @@ static void qeth_flush_buffers(struct qeth_qdio_out_q *queue, int index, buf = queue->bufs[bidx]; buf->buffer->element[buf->next_element_to_fill - 1].eflags |= SBAL_EFLAGS_LAST_ENTRY; + queue->coalesced_frames += buf->frames; if (queue->bufstates) queue->bufstates[bidx].user = buf; @@ -3401,8 +3405,18 @@ static void qeth_flush_buffers(struct qeth_qdio_out_q *queue, int index, queue->queue_no, index, count); /* Fake the TX completion interrupt: */ - if (IS_IQD(card)) - napi_schedule(&queue->napi); + if (IS_IQD(card)) { + unsigned int frames = READ_ONCE(queue->max_coalesced_frames); + unsigned int usecs = READ_ONCE(queue->coalesce_usecs); + + if (frames && queue->coalesced_frames >= frames) { + napi_schedule(&queue->napi); + queue->coalesced_frames = 0; + QETH_TXQ_STAT_INC(queue, coal_frames); + } else if (usecs) { + qeth_tx_arm_timer(queue, usecs); + } + } if (rc) { /* ignore temporary SIGA errors without busy condition */ @@ -5667,7 +5681,7 @@ static int qeth_tx_poll(struct napi_struct *napi, int budget) if (completed <= 0) { /* Ensure we see TX completion for pending work: */ if (napi_complete_done(napi, 0)) - qeth_tx_arm_timer(queue); + qeth_tx_arm_timer(queue, QETH_TX_TIMER_USECS); return 0; } diff --git a/drivers/s390/net/qeth_ethtool.c b/drivers/s390/net/qeth_ethtool.c index 6f0cc6fcc759..ebdc03210608 100644 --- a/drivers/s390/net/qeth_ethtool.c +++ b/drivers/s390/net/qeth_ethtool.c @@ -40,6 +40,7 @@ static const struct qeth_stats txq_stats[] = { QETH_TXQ_STAT("Packing mode switches", packing_mode_switch), QETH_TXQ_STAT("Queue stopped", stopped), QETH_TXQ_STAT("Doorbell", doorbell), + QETH_TXQ_STAT("IRQ for frames", coal_frames), QETH_TXQ_STAT("Completion yield", completion_yield), QETH_TXQ_STAT("Completion timer", completion_timer), }; @@ -109,6 +110,38 @@ static void qeth_get_ethtool_stats(struct net_device *dev, txq_stats, TXQ_STATS_LEN); } +static void __qeth_set_coalesce(struct net_device *dev, + struct qeth_qdio_out_q *queue, + struct ethtool_coalesce *coal) +{ + WRITE_ONCE(queue->coalesce_usecs, coal->tx_coalesce_usecs); + WRITE_ONCE(queue->max_coalesced_frames, coal->tx_max_coalesced_frames); + + if (coal->tx_coalesce_usecs && + netif_running(dev) && + !qeth_out_queue_is_empty(queue)) + qeth_tx_arm_timer(queue, coal->tx_coalesce_usecs); +} + +static int qeth_set_coalesce(struct net_device *dev, + struct ethtool_coalesce *coal) +{ + struct qeth_card *card = dev->ml_priv; + struct qeth_qdio_out_q *queue; + unsigned int i; + + if (!IS_IQD(card)) + return -EOPNOTSUPP; + + if (!coal->tx_coalesce_usecs && !coal->tx_max_coalesced_frames) + return -EINVAL; + + qeth_for_each_output_queue(card, queue, i) + __qeth_set_coalesce(dev, queue, coal); + + return 0; +} + static void qeth_get_ringparam(struct net_device *dev, struct ethtool_ringparam *param) { @@ -244,6 +277,43 @@ static int qeth_set_tunable(struct net_device *dev, } } +static int qeth_get_per_queue_coalesce(struct net_device *dev, u32 __queue, + struct ethtool_coalesce *coal) +{ + struct qeth_card *card = dev->ml_priv; + struct qeth_qdio_out_q *queue; + + if (!IS_IQD(card)) + return -EOPNOTSUPP; + + if (__queue >= card->qdio.no_out_queues) + return -EINVAL; + + queue = card->qdio.out_qs[__queue]; + + coal->tx_coalesce_usecs = queue->coalesce_usecs; + coal->tx_max_coalesced_frames = queue->max_coalesced_frames; + return 0; +} + +static int qeth_set_per_queue_coalesce(struct net_device *dev, u32 queue, + struct ethtool_coalesce *coal) +{ + struct qeth_card *card = dev->ml_priv; + + if (!IS_IQD(card)) + return -EOPNOTSUPP; + + if (queue >= card->qdio.no_out_queues) + return -EINVAL; + + if (!coal->tx_coalesce_usecs && !coal->tx_max_coalesced_frames) + return -EINVAL; + + __qeth_set_coalesce(dev, card->qdio.out_qs[queue], coal); + return 0; +} + /* Helper function to fill 'advertising' and 'supported' which are the same. */ /* Autoneg and full-duplex are supported and advertised unconditionally. */ /* Always advertise and support all speeds up to specified, and only one */ @@ -443,7 +513,10 @@ static int qeth_get_link_ksettings(struct net_device *netdev, } const struct ethtool_ops qeth_ethtool_ops = { + .supported_coalesce_params = ETHTOOL_COALESCE_TX_USECS | + ETHTOOL_COALESCE_TX_MAX_FRAMES, .get_link = ethtool_op_get_link, + .set_coalesce = qeth_set_coalesce, .get_ringparam = qeth_get_ringparam, .get_strings = qeth_get_strings, .get_ethtool_stats = qeth_get_ethtool_stats, @@ -454,6 +527,8 @@ const struct ethtool_ops qeth_ethtool_ops = { .get_ts_info = qeth_get_ts_info, .get_tunable = qeth_get_tunable, .set_tunable = qeth_set_tunable, + .get_per_queue_coalesce = qeth_get_per_queue_coalesce, + .set_per_queue_coalesce = qeth_set_per_queue_coalesce, .get_link_ksettings = qeth_get_link_ksettings, };