Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752100AbbHRAMI (ORCPT ); Mon, 17 Aug 2015 20:12:08 -0400 Received: from mail-la0-f49.google.com ([209.85.215.49]:34552 "EHLO mail-la0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751181AbbHRAMF (ORCPT ); Mon, 17 Aug 2015 20:12:05 -0400 MIME-Version: 1.0 In-Reply-To: <1439548222-231611-3-git-send-email-liguozhu@hisilicon.com> References: <1439548222-231611-1-git-send-email-liguozhu@hisilicon.com> <1439548222-231611-3-git-send-email-liguozhu@hisilicon.com> Date: Tue, 18 Aug 2015 03:12:02 +0300 Message-ID: Subject: Re: [PATCH 2/5] net: add Hisilicon Network Subsystem hnae framework support From: Alexey Klimov To: Kenneth Lee Cc: robh+dt@kernel.org, pawel.moll@arm.com, Mark Rutland , ijc+devicetree@hellion.org.uk, Kumar Gala , Catalin Marinas , Will Deacon , Yisen.Zhuang@huawei.com, "David S. Miller" , paul.gortmaker@windriver.com, dingtianhong@huawei.com, zhangfei.gao@linaro.org, devicetree@vger.kernel.org, Linux Kernel Mailing List , linux-arm-kernel@lists.infradead.org, netdev@vger.kernel.org, linuxarm@huawei.com, salil.mehta@huawei.com, huangdaode@hisilicon.com, Kenneth Lee , Yury Norov Content-Type: text/plain; charset=UTF-8 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 41187 Lines: 1229 Hi Kenneth, just small minor question. On Fri, Aug 14, 2015 at 1:30 PM, Kenneth Lee wrote: > HNAE (Hisilicon Network Acceleration Engine) is a framework to provide a > unified ring buffer interface for Hisilicon Network Acceleration Engines. > > With the interface, upper layer can work as ethernet driver, ODP driver or > other service driver on purpose. > > Signed-off-by: Kenneth Lee > Signed-off-by: Yisen Zhuang > --- > drivers/net/ethernet/hisilicon/Kconfig | 33 +- > drivers/net/ethernet/hisilicon/Makefile | 1 + > drivers/net/ethernet/hisilicon/hns/Makefile | 15 + > drivers/net/ethernet/hisilicon/hns/hnae.c | 494 +++++++++++++++++++++++ > drivers/net/ethernet/hisilicon/hns/hnae.h | 582 ++++++++++++++++++++++++++++ > 5 files changed, 1124 insertions(+), 1 deletion(-) > create mode 100644 drivers/net/ethernet/hisilicon/hns/Makefile > create mode 100644 drivers/net/ethernet/hisilicon/hns/hnae.c > create mode 100644 drivers/net/ethernet/hisilicon/hns/hnae.h > > diff --git a/drivers/net/ethernet/hisilicon/Kconfig b/drivers/net/ethernet/hisilicon/Kconfig > index dead17b..1e4f5a7 100644 > --- a/drivers/net/ethernet/hisilicon/Kconfig > +++ b/drivers/net/ethernet/hisilicon/Kconfig > @@ -5,7 +5,7 @@ > config NET_VENDOR_HISILICON > bool "Hisilicon devices" > default y > - depends on ARM > + depends on ARM || ARM64 > ---help--- > If you have a network (Ethernet) card belonging to this class, say Y. > > @@ -31,4 +31,35 @@ config HIP04_ETH > If you wish to compile a kernel for a hardware with hisilicon p04 SoC and > want to use the internal ethernet then you should answer Y to this. > > +config HNS > + tristate "Hisilicon Network Subsystem Support (Framework)" > + ---help--- > + This selects the framework support for Hisilicon Network Subsystem. It > + is needed by any driver which provides HNS acceleration engine or make > + use of the engine > + > +config HNS_DSAF > + tristate "Hisilicon HNS DSAF device Support" > + select HNS > + select HNS_MDIO > + ---help--- > + This selects the DSAF (Distributed System Area Frabric) network > + acceleration engine support. The engine is used in Hisilicon P660, > + Hi1610 and further ICT SoC > + > +config HNS_MDIO > + tristate "Hisilicon HNS MDIO device Support" > + select MDIO > + ---help--- > + This selects the HNS MDIO support. It is needed by HNS_DSAF to access > + the PHY > + > +config HNS_ENET > + tristate "Hisilicon HNS Ethernet Device Support" > + select PHYLIB > + select HNS > + ---help--- > + This selects the general ethernet driver for HNS. This module make > + use of any HNS AE driver, such as HNS_DSAF > + > endif # NET_VENDOR_HISILICON > diff --git a/drivers/net/ethernet/hisilicon/Makefile b/drivers/net/ethernet/hisilicon/Makefile > index 6c14540..2503a9b 100644 > --- a/drivers/net/ethernet/hisilicon/Makefile > +++ b/drivers/net/ethernet/hisilicon/Makefile > @@ -4,3 +4,4 @@ > > obj-$(CONFIG_HIX5HD2_GMAC) += hix5hd2_gmac.o > obj-$(CONFIG_HIP04_ETH) += hip04_mdio.o hip04_eth.o > +obj-$(CONFIG_HNS) += hns/ > diff --git a/drivers/net/ethernet/hisilicon/hns/Makefile b/drivers/net/ethernet/hisilicon/hns/Makefile > new file mode 100644 > index 0000000..6680602 > --- /dev/null > +++ b/drivers/net/ethernet/hisilicon/hns/Makefile > @@ -0,0 +1,15 @@ > +# > +# Makefile for the HISILICON network device drivers. > +# > + > +obj-$(CONFIG_HNS) += hnae.o > + > +obj-$(CONFIG_HNS_DSAF) += hns_dsaf.o > +hns_dsaf-objs = hns_ae_adapt.o hns_dsaf_gmac.o hns_dsaf_mac.o hns_dsaf_misc.o \ > + hns_dsaf_main.o hns_dsaf_ppe.o hns_dsaf_rcb.o hns_dsaf_xgmac.o > + > +obj-$(CONFIG_HNS_MDIO) += hns_mdio.o > +hns_mdio-objs = hns_mdio_main.o > + > +obj-$(CONFIG_HNS_ENET) += hns_enet_drv.o > +hns_enet_drv-objs = hns_enet.o hns_ethtool.o > diff --git a/drivers/net/ethernet/hisilicon/hns/hnae.c b/drivers/net/ethernet/hisilicon/hns/hnae.c > new file mode 100644 > index 0000000..fd09768 > --- /dev/null > +++ b/drivers/net/ethernet/hisilicon/hns/hnae.c > @@ -0,0 +1,494 @@ > +/* > + * Copyright (c) 2014-2015 Hisilicon Limited. > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License as published by > + * the Free Software Foundation; either version 2 of the License, or > + * (at your option) any later version. > + */ > + > +#include > +#include > +#include > +#include > + > +#include "hnae.h" > + > +#define cls_to_ae_dev(dev) container_of(dev, struct hnae_ae_dev, cls_dev) > + > +static struct class *hnae_class; > + > +static inline void hnae_list_add(spinlock_t *lock, struct list_head *node, > + struct list_head *head) > +{ > + unsigned long flags; > + > + spin_lock_irqsave(lock, flags); > + list_add_tail_rcu(node, head); > + spin_unlock_irqrestore(lock, flags); > +} > + > +static inline void hnae_list_del(spinlock_t *lock, struct list_head *node) > +{ > + unsigned long flags; > + > + spin_lock_irqsave(lock, flags); > + list_del_rcu(node); > + spin_unlock_irqrestore(lock, flags); > +} > + > +static int hnae_alloc_buffer(struct hnae_ring *ring, struct hnae_desc_cb *cb) > +{ > + unsigned int order = hnae_page_order(ring); > + struct page *p = dev_alloc_pages(order); > + > + if (!p) > + return -ENOMEM; > + > + cb->priv = p; > + cb->page_offset = 0; > + cb->reuse_flag = 0; > + cb->buf = page_address(p); > + cb->length = hnae_page_size(ring); > + cb->type = DESC_TYPE_PAGE; > + > + return 0; > +} > + > +static void hnae_free_buffer(struct hnae_ring *ring, struct hnae_desc_cb *cb) > +{ > + if (cb->type == DESC_TYPE_SKB) > + dev_kfree_skb_any((struct sk_buff *)cb->priv); > + else if (unlikely(is_rx_ring(ring))) > + put_page((struct page *)cb->priv); > + memset(cb, 0, sizeof(*cb)); > +} > + > +static int hnae_map_buffer(struct hnae_ring *ring, struct hnae_desc_cb *cb) > +{ > + cb->dma = dma_map_page(ring_to_dev(ring), cb->priv, 0, > + cb->length, ring_to_dma_dir(ring)); > + > + if (dma_mapping_error(ring_to_dev(ring), cb->dma)) > + return -EIO; > + > + return 0; > +} > + > +static void hnae_unmap_buffer(struct hnae_ring *ring, struct hnae_desc_cb *cb) > +{ > + if (cb->type == DESC_TYPE_SKB) > + dma_unmap_single(ring_to_dev(ring), cb->dma, cb->length, > + ring_to_dma_dir(ring)); > + else > + dma_unmap_page(ring_to_dev(ring), cb->dma, cb->length, > + ring_to_dma_dir(ring)); > +} > + > +static struct hnae_buf_ops hnae_bops = { > + .alloc_buffer = hnae_alloc_buffer, > + .free_buffer = hnae_free_buffer, > + .map_buffer = hnae_map_buffer, > + .unmap_buffer = hnae_unmap_buffer, > +}; > + > +static int __ae_match(struct device *dev, const void *data) > +{ > + struct hnae_ae_dev *hdev = cls_to_ae_dev(dev); > + const char *ae_id = data; > + > + if (!strncmp(ae_id, hdev->name, AE_NAME_SIZE)) > + return 1; > + > + return 0; > +} > + > +static struct hnae_ae_dev *find_ae(const char *ae_id) > +{ > + struct device *dev; > + > + BUG_ON(!ae_id); > + > + dev = class_find_device(hnae_class, NULL, ae_id, __ae_match); > + > + return cls_to_ae_dev(dev); > +} Most entries of class_find_device() in kernel that i see perform if-check for NULL on return value. What do you think about such option for example: return dev ? cls_to_ae_dev(dev) : NULL; ? Is it possible that, someday, your dev will be NULL and creep in container_of() that doesn't protected from ptr = NULL and .... (let's go down a little in another function) > + > +static void hnae_free_buffers(struct hnae_ring *ring) > +{ > + int i; > + > + for (i = 0; i < ring->desc_num; i++) > + hnae_free_buffer_detach(ring, i); > +} > + > +/* Allocate memory for raw pkg, and map with dma */ > +static int hnae_alloc_buffers(struct hnae_ring *ring) > +{ > + int i, j, ret; > + > + for (i = 0; i < ring->desc_num; i++) { > + ret = hnae_alloc_buffer_attach(ring, i); > + if (ret) > + goto out_buffer_fail; > + } > + > + return 0; > + > +out_buffer_fail: > + for (j = i - 1; j >= 0; j--) > + hnae_free_buffer_detach(ring, j); > + return ret; > +} > + > +/* free desc along with its attached buffer */ > +static inline void hnae_free_desc(struct hnae_ring *ring) > +{ > + hnae_free_buffers(ring); > + dma_unmap_single(ring_to_dev(ring), ring->desc_dma_addr, > + ring->desc_num * sizeof(ring->desc[0]), > + ring_to_dma_dir(ring)); > + ring->desc_dma_addr = 0; > + kfree(ring->desc); > + ring->desc = NULL; > +} > + > +/* alloc desc, without buffer attached */ > +static inline int hnae_alloc_desc(struct hnae_ring *ring) > +{ > + int size = ring->desc_num * sizeof(ring->desc[0]); > + > + ring->desc = kzalloc(size, GFP_KERNEL); > + if (!ring->desc) > + return -ENOMEM; > + > + ring->desc_dma_addr = dma_map_single(ring_to_dev(ring), > + ring->desc, size, ring_to_dma_dir(ring)); > + if (dma_mapping_error(ring_to_dev(ring), ring->desc_dma_addr)) { > + ring->desc_dma_addr = 0; > + kfree(ring->desc); > + ring->desc = NULL; > + return -ENOMEM; > + } > + > + return 0; > +} > + > +/* fini ring, also free the buffer for the ring */ > +static inline void hnae_fini_ring(struct hnae_ring *ring) > +{ > + hnae_free_desc(ring); > + kfree(ring->desc_cb); > + ring->desc_cb = NULL; > + ring->next_to_clean = 0; > + ring->next_to_use = 0; > +} > + > +/* init ring, and with buffer for rx ring */ > +static inline int hnae_init_ring(struct hnae_queue *q, struct hnae_ring *ring, > + int flags) > +{ > + int ret; > + > + if (ring->desc_num <= 0 || ring->buf_size <= 0) > + return -EINVAL; > + > + ring->q = q; > + ring->flags = flags; > + assert(!ring->desc && !ring->desc_cb && !ring->desc_dma_addr); > + > + /* not matter for tx or rx ring, the ntc and ntc start from 0 */ > + assert(ring->next_to_use == 0); > + assert(ring->next_to_clean == 0); > + > + ring->desc_cb = kcalloc(ring->desc_num, sizeof(ring->desc_cb[0]), > + GFP_KERNEL); > + if (!ring->desc_cb) { > + ret = -ENOMEM; > + goto out; > + } > + > + ret = hnae_alloc_desc(ring); > + if (ret) > + goto out_with_desc_cb; > + > + if (is_rx_ring(ring)) { > + ret = hnae_alloc_buffers(ring); > + if (ret) > + goto out_with_desc; > + } > + > + return 0; > + > +out_with_desc: > + hnae_free_desc(ring); > +out_with_desc_cb: > + kfree(ring->desc_cb); > + ring->desc_cb = NULL; > +out: > + return ret; > +} > + > +static inline int hnae_init_queue(struct hnae_handle *h, struct hnae_queue *q, > + struct hnae_ae_dev *dev) > +{ > + int ret; > + > + q->dev = dev; > + q->handle = h; > + > + ret = hnae_init_ring(q, &q->tx_ring, q->tx_ring.flags | RINGF_DIR); > + if (ret) > + goto out; > + > + ret = hnae_init_ring(q, &q->rx_ring, q->rx_ring.flags & ~RINGF_DIR); > + if (ret) > + goto out_with_tx_ring; > + > + if (dev->ops->init_queue) > + dev->ops->init_queue(q); > + > + return 0; > + > +out_with_tx_ring: > + hnae_fini_ring(&q->tx_ring); > +out: > + return ret; > +} > + > +static inline void hnae_fini_queue(struct hnae_queue *q) > +{ > + if (q->dev->ops->fini_queue) > + q->dev->ops->fini_queue(q); > + > + hnae_fini_ring(&q->tx_ring); > + hnae_fini_ring(&q->rx_ring); > +} > + > +/** > + * ae_chain - define ae chain head > + */ > +static RAW_NOTIFIER_HEAD(ae_chain); > + > +int hnae_register_notifier(struct notifier_block *nb) > +{ > + return raw_notifier_chain_register(&ae_chain, nb); > +} > +EXPORT_SYMBOL(hnae_register_notifier); > + > +void hnae_unregister_notifier(struct notifier_block *nb) > +{ > + if (raw_notifier_chain_unregister(&ae_chain, nb)) > + dev_err(NULL, "notifier chain unregister fail\n"); > +} > +EXPORT_SYMBOL(hnae_unregister_notifier); > + > +int hnae_reinit_handle(struct hnae_handle *handle) > +{ > + int i, j; > + int ret; > + > + for (i = 0; i < handle->q_num; i++) /* free ring*/ > + hnae_fini_queue(handle->qs[i]); > + > + for (i = 0; i < handle->q_num; i++) {/* reinit ring*/ > + ret = hnae_init_queue(handle, handle->qs[i], handle->dev); > + if (ret) > + goto out_when_init_queue; > + } > + return 0; > +out_when_init_queue: > + for (j = i - 1; j >= 0; j--) > + hnae_fini_queue(handle->qs[j]); > + return ret; > +} > +EXPORT_SYMBOL(hnae_reinit_handle); > + > +/* hnae_get_handle - get a handle from the AE > + * @owner_dev: the dev use this handle > + * @ae_id: the id of the ae to be used > + * @ae_opts: the options set for the handle > + * @bops: the callbacks for buffer management > + * > + * return handle ptr or ERR_PTR > + */ > +struct hnae_handle *hnae_get_handle(struct device *owner_dev, > + const char *ae_id, const char *ae_opts, > + struct hnae_buf_ops *bops) > +{ > + struct hnae_ae_dev *dev; > + struct hnae_handle *handle; > + int i, j; > + int ret; > + > + dev = find_ae(ae_id); > + if (!dev) > + return ERR_PTR(-ENODEV); .. and here you might not receive NULL after container_of() works on that returned NULL after class_find_device(). I guess that might happen after someone change order of fields in struct hnae_ae_dev. I understand that it's unlikely to happen but could you please check or comment? > + > + handle = dev->ops->get_handle(dev, ae_opts); > + if (IS_ERR(handle)) > + return handle; > + > + handle->dev = dev; > + handle->owner_dev = owner_dev; > + handle->bops = bops ? bops : &hnae_bops; > + handle->ae_opts = ae_opts; > + > + for (i = 0; i < handle->q_num; i++) { > + ret = hnae_init_queue(handle, handle->qs[i], dev); > + if (ret) > + goto out_when_init_queue; > + } > + > + __module_get(dev->owner); > + > + hnae_list_add(&dev->lock, &handle->node, &dev->handle_list); > + > + return handle; > + > +out_when_init_queue: > + for (j = i - 1; j >= 0; j--) > + hnae_fini_queue(handle->qs[j]); > + > + return ERR_PTR(-ENOMEM); > +} > +EXPORT_SYMBOL(hnae_get_handle); > + > +void hnae_put_handle(struct hnae_handle *h) > +{ > + struct hnae_ae_dev *dev = h->dev; > + int i; > + > + for (i = 0; i < h->q_num; i++) > + hnae_fini_queue(h->qs[i]); > + > + hnae_list_del(&dev->lock, &h->node); > + > + if (dev->ops->put_handle) > + dev->ops->put_handle(h); > + > + module_put(dev->owner); > +} > +EXPORT_SYMBOL(hnae_put_handle); > + > +static void __hnae_release(struct device *dev) > +{ > +} > + > +/** > + * hnae_ae_register - register a AE engine to hnae framework > + * @hdev: the hnae ae engine device > + * @owner: the module who provides this dev > + * NOTE: the duplicated name will not be checked > + */ > +int hnae_ae_register(struct hnae_ae_dev *hdev, struct module *owner) > +{ > + static atomic_t id = ATOMIC_INIT(-1); > + int ret; > + > + if (!hdev->dev) > + return -ENODEV; > + > + if (!hdev->ops || !hdev->ops->get_handle || > + !hdev->ops->toggle_ring_irq || > + !hdev->ops->toggle_queue_status || > + !hdev->ops->get_status || !hdev->ops->adjust_link) > + return -EINVAL; > + > + hdev->owner = owner; > + hdev->id = (int)atomic_inc_return(&id); > + hdev->cls_dev.parent = hdev->dev; > + hdev->cls_dev.class = hnae_class; > + hdev->cls_dev.release = __hnae_release; > + dev_set_name(&hdev->cls_dev, "hnae%d", hdev->id); > + ret = device_register(&hdev->cls_dev); > + if (ret) > + return ret; > + > + __module_get(THIS_MODULE); > + > + INIT_LIST_HEAD(&hdev->handle_list); > + spin_lock_init(&hdev->lock); > + > + ret = raw_notifier_call_chain(&ae_chain, HNAE_AE_REGISTER, NULL); > + if (ret) > + dev_dbg(hdev->dev, > + "has not notifier for AE: %s\n", hdev->name); > + > + return 0; > +} > +EXPORT_SYMBOL(hnae_ae_register); > + > +/** > + * hnae_ae_unregister - unregisters a HNAE AE engine > + * @cdev: the device to unregister > + */ > +void hnae_ae_unregister(struct hnae_ae_dev *hdev) > +{ > + device_unregister(&hdev->cls_dev); > + module_put(THIS_MODULE); > +} > +EXPORT_SYMBOL(hnae_ae_unregister); > + > +static ssize_t handles_show(struct device *dev, > + struct device_attribute *attr, char *buf) > +{ > + ssize_t s = 0; > + struct hnae_ae_dev *hdev = cls_to_ae_dev(dev); > + struct hnae_handle *h; > + int i = 0, j; > + > + list_for_each_entry_rcu(h, &hdev->handle_list, node) { > + s += sprintf(buf + s, "%s(%s): ", dev_name(h->owner_dev), > + h->ae_opts); > + for (j = 0; j < h->q_num; j++) { > + if (h->qs[i]) { > + s += sprintf(buf + s, > + "(%llu,%llu; ", > + h->qs[i]->tx_ring.stats.sw_err_cnt, > + h->qs[i]->tx_ring.stats.io_err_cnt); > + s += sprintf(buf + s, > + "%llu,%llu,%llu), ", > + h->qs[i]->rx_ring.stats.sw_err_cnt, > + h->qs[i]->rx_ring.stats.io_err_cnt, > + h->qs[i]->rx_ring.stats.seg_pkt_cnt); > + } else { > + s += sprintf(buf + s, "(null)\n"); > + } > + } > + s += sprintf(buf + s, "\n"); > + } > + > + return s; > +} > + > +static DEVICE_ATTR_RO(handles); > +static struct attribute *hnae_class_attrs[] = { > + &dev_attr_handles.attr, > + NULL, > +}; > +ATTRIBUTE_GROUPS(hnae_class); > + > +static int __init hnae_init(void) > +{ > + hnae_class = class_create(THIS_MODULE, "hnae"); > + if (IS_ERR(hnae_class)) > + return PTR_ERR(hnae_class); > + > + hnae_class->dev_groups = hnae_class_groups; > + return 0; > +} > + > +static void __exit hnae_exit(void) > +{ > + class_destroy(hnae_class); > +} > + > +subsys_initcall(hnae_init); > +module_exit(hnae_exit); > + > +MODULE_AUTHOR("Hisilicon, Inc."); > +MODULE_LICENSE("GPL"); > +MODULE_DESCRIPTION("Hisilicon Network Acceleration Engine Framework"); > + > +/* vi: set tw=78 noet: */ > diff --git a/drivers/net/ethernet/hisilicon/hns/hnae.h b/drivers/net/ethernet/hisilicon/hns/hnae.h > new file mode 100644 > index 0000000..e7856cf > --- /dev/null > +++ b/drivers/net/ethernet/hisilicon/hns/hnae.h > @@ -0,0 +1,582 @@ > +/* > + * Copyright (c) 2014-2015 Hisilicon Limited. > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License as published by > + * the Free Software Foundation; either version 2 of the License, or > + * (at your option) any later version. > + */ > + > +#ifndef __HNAE_H > +#define __HNAE_H > + > +/* Names used in this framework: > + * ae handle (handle): > + * a set of queues provided by AE > + * ring buffer queue (rbq): > + * the channel between upper layer and the AE, can do tx and rx > + * ring: > + * a tx or rx channel within a rbq > + * ring description (desc): > + * an element in the ring with packet information > + * buffer: > + * a memory region referred by desc with the full packet payload > + * > + * "num" means a static number set as a parameter, "count" mean a dynamic > + * number set while running > + * "cb" means control block > + */ > + > +#include > +#include > +#include > +#include > +#include > +#include > + > +#define HNAE_DRIVER_VERSION "1.1.0" > +#define HNAE_DRIVER_NAME "hns" > +#define HNAE_COPYRIGHT "Copyright(c) 2015 Huawei Corporation." > +#define HNAE_DRIVER_STRING "Hisilicon Network Subsystem Driver" > +#define HNAE_DEFAULT_DEVICE_DESCR "Hisilicon Network Subsystem" > + > +#ifdef DEBUG > + > +#ifndef assert > +#define assert(expr) \ > +do { \ > + if (!(expr)) { \ > + pr_err("Assertion failed! %s, %s, %s, line %d\n", \ > + #expr, __FILE__, __func__, __LINE__); \ > + } \ > +} while (0) > +#endif > + > +#else > + > +#ifndef assert > +#define assert(expr) > +#endif > + > +#endif > + > +#define AE_NAME_SIZE 16 > +#define AE_OPTS_SIZE 64 > + > +/* some said the RX and TX RCB format should not be the same in the future. But > + * it is the same now... > + */ > +#define RCB_REG_BASEADDR_L 0x00 /* P660 support only 32bit accessing */ > +#define RCB_REG_BASEADDR_H 0x04 > +#define RCB_REG_BD_NUM 0x08 > +#define RCB_REG_BD_LEN 0x0C > +#define RCB_REG_PKTLINE 0x10 > +#define RCB_REG_TAIL 0x18 > +#define RCB_REG_HEAD 0x1C > +#define RCB_REG_FBDNUM 0x20 > +#define RCB_REG_OFFSET 0x24 /* pkt num to be handled */ > +#define RCB_REG_PKTNUM_RECORD 0x2C /* total pkt received */ > + > +#define HNS_RX_HEAD_SIZE 256 > + > +#define HNAE_AE_REGISTER 0x1 > + > +#define RCB_RING_NAME_LEN 16 > + > +enum hnae_led_state { > + HNAE_LED_INACTIVE, > + HNAE_LED_ACTIVE, > + HNAE_LED_ON, > + HNAE_LED_OFF > +}; > + > +#define HNS_RX_FLAG_VLAN_PRESENT 0x1 > +#define HNS_RX_FLAG_L3ID_IPV4 0x0 > +#define HNS_RX_FLAG_L3ID_IPV6 0x1 > +#define HNS_RX_FLAG_L4ID_UDP 0x0 > +#define HNS_RX_FLAG_L4ID_TCP 0x1 > + > +#define HNS_TXD_ASID_S 0 > +#define HNS_TXD_ASID_M (0xff << HNS_TXD_ASID_S) > +#define HNS_TXD_BUFNUM_S 8 > +#define HNS_TXD_BUFNUM_M (0x3 << HNS_TXD_BUFNUM_S) > +#define HNS_TXD_PORTID_S 10 > +#define HNS_TXD_PORTID_M (0x7 << HNS_TXD_PORTID_S) > + > +#define HNS_TXD_RA_B 8 > +#define HNS_TXD_RI_B 9 > +#define HNS_TXD_L4CS_B 10 > +#define HNS_TXD_L3CS_B 11 > +#define HNS_TXD_FE_B 12 > +#define HNS_TXD_VLD_B 13 > +#define HNS_TXD_IPOFFSET_S 14 > +#define HNS_TXD_IPOFFSET_M (0xff << HNS_TXD_IPOFFSET_S) > + > +#define HNS_RXD_IPOFFSET_S 0 > +#define HNS_RXD_IPOFFSET_M (0xff << HNS_TXD_IPOFFSET_S) > +#define HNS_RXD_BUFNUM_S 8 > +#define HNS_RXD_BUFNUM_M (0x3 << HNS_RXD_BUFNUM_S) > +#define HNS_RXD_PORTID_S 10 > +#define HNS_RXD_PORTID_M (0x7 << HNS_RXD_PORTID_S) > +#define HNS_RXD_DMAC_S 13 > +#define HNS_RXD_DMAC_M (0x3 << HNS_RXD_DMAC_S) > +#define HNS_RXD_VLAN_S 15 > +#define HNS_RXD_VLAN_M (0x3 << HNS_RXD_VLAN_S) > +#define HNS_RXD_L3ID_S 17 > +#define HNS_RXD_L3ID_M (0xf << HNS_RXD_L3ID_S) > +#define HNS_RXD_L4ID_S 21 > +#define HNS_RXD_L4ID_M (0xf << HNS_RXD_L4ID_S) > +#define HNS_RXD_FE_B 25 > +#define HNS_RXD_FRAG_B 26 > +#define HNS_RXD_VLD_B 27 > +#define HNS_RXD_L2E_B 28 > +#define HNS_RXD_L3E_B 29 > +#define HNS_RXD_L4E_B 30 > +#define HNS_RXD_DROP_B 31 > + > +#define HNS_RXD_VLANID_S 8 > +#define HNS_RXD_VLANID_M (0xfff << HNS_RXD_VLANID_S) > +#define HNS_RXD_CFI_B 20 > +#define HNS_RXD_PRI_S 21 > +#define HNS_RXD_PRI_M (0x7 << HNS_RXD_PRI_S) > +#define HNS_RXD_ASID_S 24 > +#define HNS_RXD_ASID_M (0xff << HNS_RXD_ASID_S) > + > +/* hardware spec ring buffer format */ > +struct __packed hnae_desc { > + __le64 addr; > + union { > + struct { > + __le16 asid_bufnum_pid; > + __le16 send_size; > + __le32 flag_ipoffset; > + __le32 reserved_3[4]; > + } tx; > + > + struct { > + __le32 ipoff_bnum_pid_flag; > + __le16 pkt_len; > + __le16 size; > + __le32 vlan_pri_asid; > + __le32 reserved_2[3]; > + } rx; > + }; > +}; > + > +struct hnae_desc_cb { > + dma_addr_t dma; /* dma address of this desc */ > + void *buf; /* cpu addr for a desc */ > + > + /* priv data for the desc, e.g. skb when use with ip stack*/ > + void *priv; > + u16 page_offset; > + u16 reuse_flag; > + > + u16 length; /* length of the buffer */ > + > + /* desc type, used by the ring user to mark the type of the priv data */ > + u16 type; > +}; > + > +#define setflags(flags, bits) ((flags) |= (bits)) > +#define unsetflags(flags, bits) ((flags) &= ~(bits)) > + > +/* hnae_ring->flags fields */ > +#define RINGF_DIR 0x1 /* TX or RX ring, set if TX */ > +#define is_tx_ring(ring) ((ring)->flags & RINGF_DIR) > +#define is_rx_ring(ring) (!is_tx_ring(ring)) > +#define ring_to_dma_dir(ring) (is_tx_ring(ring) ? \ > + DMA_TO_DEVICE : DMA_FROM_DEVICE) > + > +struct ring_stats { > + u64 io_err_cnt; > + u64 sw_err_cnt; > + u64 seg_pkt_cnt; > + u64 reuse_pg_cnt; > + u64 err_pkt_len; > + u64 non_vld_descs; > + u64 err_bd_num; > + u64 csum_err; > + union { > + u64 tx_pkts; > + u64 rx_pkts; > + } pkts; > + union { > + u64 tx_bytes; > + u64 rx_bytes; > + } bytes; > + union { > + u64 tx_err_cnt; > + u64 rx_err_cnt; > + } err_cnt; > +}; > + > +struct hnae_queue; > + > +struct hnae_ring { > + u8 __iomem *io_base; /* base io address for the ring */ > + struct hnae_desc *desc; /* dma map address space */ > + struct hnae_desc_cb *desc_cb; > + struct hnae_queue *q; > + int irq; > + char ring_name[RCB_RING_NAME_LEN]; > + > + /* statistic */ > + struct ring_stats stats; > + > + dma_addr_t desc_dma_addr; > + u32 buf_size; /* size for hnae_desc->addr, preset by AE */ > + u16 desc_num; /* total number of desc */ > + u16 max_desc_num_per_pkt; > + u16 max_raw_data_sz_per_desc; > + u16 max_pkt_size; > + int next_to_use; /* idx of next spare desc */ > + > + /* idx of lastest sent desc, the ring is empty when equal to > + * next_to_use > + */ > + int next_to_clean; > + > + int flags; /* ring attribute */ > +}; > + > +#define ring_ptr_move_fw(ring, p) \ > + ((ring)->p = ((ring)->p + 1) % (ring)->desc_num) > +#define ring_ptr_move_bw(ring, p) \ > + ((ring)->p = ((ring)->p - 1 + (ring)->desc_num) % (ring)->desc_num) > + > +enum hns_desc_type { > + DESC_TYPE_SKB, > + DESC_TYPE_PAGE, > +}; > + > +#define assert_is_ring_idx(ring, idx) \ > + assert((idx) >= 0 && (idx) < (ring)->desc_num) > + > +/* the distance between [begin, end) in a ring buffer > + * note: there is a unuse slot between the begin and the end > + */ > +static inline int ring_dist(struct hnae_ring *ring, int begin, int end) > +{ > + assert_is_ring_idx(ring, begin); > + assert_is_ring_idx(ring, end); > + > + return (end - begin + ring->desc_num) % ring->desc_num; > +} > + > +static inline int ring_space(struct hnae_ring *ring) > +{ > + return ring->desc_num - > + ring_dist(ring, ring->next_to_clean, ring->next_to_use) - 1; > +} > + > +static inline int is_ring_empty(struct hnae_ring *ring) > +{ > + assert_is_ring_idx(ring, ring->next_to_use); > + assert_is_ring_idx(ring, ring->next_to_clean); > + > + return ring->next_to_use == ring->next_to_clean; > +} > + > +#define hnae_buf_size(_ring) ((_ring)->buf_size) > +#define hnae_page_order(_ring) (get_order(hnae_buf_size(_ring))) > +#define hnae_page_size(_ring) (PAGE_SIZE << hnae_page_order(_ring)) > + > +struct hnae_handle; > + > +/* allocate and dma map space for hnae desc */ > +struct hnae_buf_ops { > + int (*alloc_buffer)(struct hnae_ring *ring, struct hnae_desc_cb *cb); > + void (*free_buffer)(struct hnae_ring *ring, struct hnae_desc_cb *cb); > + int (*map_buffer)(struct hnae_ring *ring, struct hnae_desc_cb *cb); > + void (*unmap_buffer)(struct hnae_ring *ring, struct hnae_desc_cb *cb); > +}; > + > +struct hnae_queue { > + void __iomem *io_base; > + struct hnae_ae_dev *dev; /* the device who use this queue */ > + struct hnae_ring rx_ring, tx_ring; > + struct hnae_handle *handle; > +}; > + > +/*hnae loop mode*/ > +enum hnae_loop { > + MAC_LOOP_NONE = 0, > + MAC_INTERNALLOOP_MAC, > + MAC_INTERNALLOOP_SERDES, > + MAC_INTERNALLOOP_PHY, > +}; > + > +/*hnae port type*/ > +enum hnae_port_type { > + HNAE_PORT_SERVICE = 0, > + HNAE_PORT_DEBUG > +}; > + > +/* This struct defines the operation on the handle. > + * > + * get_handle(): (mandatory) > + * Get a handle from AE according to its name and options. > + * the AE driver should manage the space used by handle and its queues while > + * the HNAE framework will allocate desc and desc_cb for all rings in the > + * queues. > + * put_handle(): > + * Release the handle. > + * start(): > + * Enable the hardware, include all queues > + * stop(): > + * Disable the hardware > + * set_opts(): (mandatory) > + * Set options to the AE > + * get_opts(): (mandatory) > + * Get options from the AE > + * get_status(): > + * Get the carrier state of the back channel of the handle, 1 for ok, 0 for > + * non-ok > + * toggle_ring_irq(): (mandatory) > + * Set the ring irq to be enabled(0) or disable(1) > + * toggle_queue_status(): (mandatory) > + * Set the queue to be enabled(1) or disable(0), this will not change the > + * ring irq state > + * adjust_link() > + * adjust link status > + * set_loopback() > + * set loopback > + * get_ring_bdnum_limit() > + * get ring bd number limit > + * get_pauseparam() > + * get tx and rx of pause frame use > + * set_autoneg() > + * set auto autonegotiation of pause frame use > + * get_autoneg() > + * get auto autonegotiation of pause frame use > + * set_pauseparam() > + * set tx and rx of pause frame use > + * get_coalesce_usecs() > + * get usecs to delay a TX interrupt after a packet is sent > + * get_rx_max_coalesced_frames() > + * get Maximum number of packets to be sent before a TX interrupt. > + * set_coalesce_usecs() > + * set usecs to delay a TX interrupt after a packet is sent > + * set_coalesce_frames() > + * set Maximum number of packets to be sent before a TX interrupt. > + * get_ringnum() > + * get RX/TX ring number > + * get_max_ringnum() > + * get RX/TX ring maximum number > + * get_mac_addr() > + * get mac address > + * set_mac_addr() > + * set mac address > + * set_mc_addr() > + * set multicast mode > + * set_mtu() > + * set mtu > + * update_stats() > + * update Old network device statistics > + * get_ethtool_stats() > + * get ethtool network device statistics > + * get_strings() > + * get a set of strings that describe the requested objects > + * get_sset_count() > + * get number of strings that @get_strings will write > + * update_led_status() > + * update the led status > + * set_led_id() > + * set led id > + * get_regs() > + * get regs dump > + * get_regs_len() > + * get the len of the regs dump > + */ > +struct hnae_ae_ops { > + struct hnae_handle *(*get_handle)(struct hnae_ae_dev *dev, > + const char *opts); > + void (*put_handle)(struct hnae_handle *handle); > + void (*init_queue)(struct hnae_queue *q); > + void (*fini_queue)(struct hnae_queue *q); > + int (*start)(struct hnae_handle *handle); > + void (*stop)(struct hnae_handle *handle); > + void (*reset)(struct hnae_handle *handle); > + int (*set_opts)(struct hnae_handle *handle, int type, void *opts); > + int (*get_opts)(struct hnae_handle *handle, int type, void **opts); > + int (*get_status)(struct hnae_handle *handle); > + int (*get_info)(struct hnae_handle *handle, > + u8 *auto_neg, u16 *speed, u8 *duplex); > + int (*set_info)(struct hnae_handle *handle, > + u8 auto_neg, u16 speed, u8 duplex); > + void (*toggle_ring_irq)(struct hnae_ring *ring, u32 val); > + void (*toggle_queue_status)(struct hnae_queue *queue, u32 val); > + void (*adjust_link)(struct hnae_handle *handle, int speed, int duplex); > + int (*set_loopback)(struct hnae_handle *handle, > + enum hnae_loop loop_mode, int en); > + void (*get_ring_bdnum_limit)(struct hnae_queue *queue, > + u32 *uplimit, u32 *lowlimit); > + void (*get_pauseparam)(struct hnae_handle *handle, > + u32 *auto_neg, u32 *rx_en, u32 *tx_en); > + int (*set_autoneg)(struct hnae_handle *handle, u8 enable); > + int (*get_autoneg)(struct hnae_handle *handle); > + void (*set_pauseparam)(struct hnae_handle *handle, > + u32 rx_en, u32 tx_en); > + void (*get_coalesce_usecs)(struct hnae_handle *handle, > + u32 *tx_usecs, u32 *rx_usecs); > + void (*get_rx_max_coalesced_frames)(struct hnae_handle *handle, > + u32 *tx_frames, u32 *rx_frames); > + void (*set_coalesce_usecs)(struct hnae_handle *handle, u32 timeout); > + int (*set_coalesce_frames)(struct hnae_handle *handle, > + u32 coalesce_frames); > + void (*get_ringnum)(struct hnae_handle *handle, u32 *ringnum); > + void (*get_max_ringnum)(struct hnae_handle *handle, u32 *max_ringnum); > + int (*get_mac_addr)(struct hnae_handle *handle, void **p); > + int (*set_mac_addr)(struct hnae_handle *handle, void *p); > + int (*set_mc_addr)(struct hnae_handle *handle, void *addr); > + int (*set_mtu)(struct hnae_handle *handle, int new_mtu); > + void (*update_stats)(struct hnae_handle *handle, > + struct net_device_stats *net_stats); > + void (*get_stats)(struct hnae_handle *handle, u64 *data); > + void (*get_strings)(struct hnae_handle *handle, > + u32 stringset, u8 *data); > + int (*get_sset_count)(struct hnae_handle *handle, int stringset); > + void (*update_led_status)(struct hnae_handle *handle); > + int (*set_led_id)(struct hnae_handle *handle, > + enum hnae_led_state status); > + void (*get_regs)(struct hnae_handle *handle, void *data); > + int (*get_regs_len)(struct hnae_handle *handle); > +}; > + > +struct hnae_ae_dev { > + struct device cls_dev; /* the class dev */ > + struct device *dev; /* the presented dev */ > + struct hnae_ae_ops *ops; > + struct list_head node; > + struct module *owner; /* the module who provides this dev */ > + int id; > + char name[AE_NAME_SIZE]; > + struct list_head handle_list; > + spinlock_t lock; /* lock to protect the handle_list */ > +}; > + > +struct hnae_handle { > + struct device *owner_dev; /* the device which make use of this handle */ > + struct hnae_ae_dev *dev; /* the device who provides this handle */ > + struct device_node *phy_node; > + phy_interface_t phy_if; > + u32 if_support; > + int q_num; > + const char *ae_opts; > + enum hnae_port_type port_type; > + struct list_head node; /* list to hnae_ae_dev->handle_list */ > + struct hnae_buf_ops *bops; /* operation for the buffer */ > + struct net_device_stats net_stats; > + struct hnae_queue **qs; /* array base of all queues */ > +}; > + > +#define ring_to_dev(ring) ((ring)->q->dev->dev) > + > +struct hnae_handle *hnae_get_handle(struct device *owner_dev, const char *ae_id, > + const char *ae_opts, > + struct hnae_buf_ops *bops); > +void hnae_put_handle(struct hnae_handle *handle); > +int hnae_ae_register(struct hnae_ae_dev *dev, struct module *owner); > +void hnae_ae_unregister(struct hnae_ae_dev *dev); > + > +int hnae_register_notifier(struct notifier_block *nb); > +void hnae_unregister_notifier(struct notifier_block *nb); > +int hnae_reinit_handle(struct hnae_handle *handle); > + > +#define hnae_queue_xmit(q, buf_num) writel_relaxed(buf_num, \ > + (q)->tx_ring.io_base + RCB_REG_TAIL) > + > +#ifndef assert > +#define assert(cond) > +#endif > + > +static inline int hnae_reserve_buffer_map(struct hnae_ring *ring, > + struct hnae_desc_cb *cb) > +{ > + struct hnae_buf_ops *bops = ring->q->handle->bops; > + int ret; > + > + ret = bops->alloc_buffer(ring, cb); > + if (ret) > + goto out; > + > + ret = bops->map_buffer(ring, cb); > + if (ret) > + goto out_with_buf; > + > + return 0; > + > +out_with_buf: > + bops->free_buffer(ring, cb); > +out: > + return ret; > +} > + > +static inline int hnae_alloc_buffer_attach(struct hnae_ring *ring, int i) > +{ > + int ret = hnae_reserve_buffer_map(ring, &ring->desc_cb[i]); > + > + if (ret) > + return ret; > + > + ring->desc[i].addr = (__le64)ring->desc_cb[i].dma; > + > + return 0; > +} > + > +static inline void hnae_buffer_detach(struct hnae_ring *ring, int i) > +{ > + ring->q->handle->bops->unmap_buffer(ring, &ring->desc_cb[i]); > + ring->desc[i].addr = 0; > +} > + > +static inline void hnae_free_buffer_detach(struct hnae_ring *ring, int i) > +{ > + struct hnae_buf_ops *bops = ring->q->handle->bops; > + struct hnae_desc_cb *cb = &ring->desc_cb[i]; > + > + if (!ring->desc_cb[i].dma) > + return; > + > + hnae_buffer_detach(ring, i); > + bops->free_buffer(ring, cb); > +} > + > +/* detach a in-used buffer and replace with a reserved one */ > +static inline void hnae_replace_buffer(struct hnae_ring *ring, int i, > + struct hnae_desc_cb *res_cb) > +{ > + struct hnae_buf_ops *bops = ring->q->handle->bops; > + struct hnae_desc_cb tmp_cb = ring->desc_cb[i]; > + > + bops->unmap_buffer(ring, &ring->desc_cb[i]); > + ring->desc_cb[i] = *res_cb; > + *res_cb = tmp_cb; > + ring->desc[i].addr = (__le64)ring->desc_cb[i].dma; > + ring->desc[i].rx.ipoff_bnum_pid_flag = 0; > +} > + > +static inline void hnae_reuse_buffer(struct hnae_ring *ring, int i) > +{ > + ring->desc_cb[i].reuse_flag = 0; > + ring->desc[i].addr = (__le64)(ring->desc_cb[i].dma > + + ring->desc_cb[i].page_offset); > + ring->desc[i].rx.ipoff_bnum_pid_flag = 0; > +} > + > +#define hnae_set_field(origin, mask, shift, val) \ > + do { \ > + (origin) &= (~(mask)); \ > + (origin) |= ((val) << (shift)) & (mask); \ > + } while (0) > + > +#define hnae_set_bit(origin, shift, val) \ > + hnae_set_field((origin), (0x1 << (shift)), (shift), (val)) > + > +#define hnae_get_field(origin, mask, shift) (((origin) & (mask)) >> (shift)) > + > +#define hnae_get_bit(origin, shift) \ > + hnae_get_field((origin), (0x1 << (shift)), (shift)) > + > +#endif > -- > 1.9.1 > > > _______________________________________________ > linux-arm-kernel mailing list > linux-arm-kernel@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-arm-kernel -- Best regards, Klimov Alexey -- 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/