* Re: [PATCH net-next 4/4] net/sched: act_mirred: Implement ingress actions
From: David Miller @ 2016-09-27 13:44 UTC (permalink / raw)
To: daniel
Cc: shmulik.ladkani, jhs, xiyou.wangcong, edumazet, netdev,
shmulik.ladkani
In-Reply-To: <57EA4C66.8070907@iogearbox.net>
From: Daniel Borkmann <daniel@iogearbox.net>
Date: Tue, 27 Sep 2016 12:39:34 +0200
> Any reason why dev_forward_skb() is not preferred over direct
> netif_receive_skb() you're using? It would, for example, implicitly
> assure that pkt_type is always PACKET_HOST, etc.
dev_forward_skb() will pull the ethernet header.
And since a direct call to netif_receive_skb() will not, one of these
two choices won't work properly.
^ permalink raw reply
* [PATCH v7 8/8] thunderbolt: Adding maintainer entry
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
Add Amir Levy as maintainer for Thunderbolt(TM) ICM driver
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
---
MAINTAINERS | 8 +++++++-
1 file changed, 7 insertions(+), 1 deletion(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 01bff8e..a4a4614 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -10358,7 +10358,13 @@ F: include/uapi/linux/stm.h
THUNDERBOLT DRIVER
M: Andreas Noever <andreas.noever@gmail.com>
S: Maintained
-F: drivers/thunderbolt/
+F: drivers/thunderbolt/*
+
+THUNDERBOLT ICM DRIVER
+M: Amir Levy <amir.jer.levy@intel.com>
+S: Maintained
+F: drivers/thunderbolt/icm/
+F: Documentation/thunderbolt/networking.txt
TI BQ27XXX POWER SUPPLY DRIVER
R: Andrew F. Davis <afd@ti.com>
--
2.7.4
^ permalink raw reply related
* [PATCH v7 6/8] thunderbolt: Networking transmit and receive
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
This patch provides the handling interface for sending and receiving
network packets between the hosts over the full communication route
(using the communication path established in the previous patch).
The Thunderbolt Network driver interfaces the Linux network stack
and the hardware controller configuration to handle packet transmissions:
+----------------+ +----------------+
|Host 1 | |Host 2 |
| | | |
| +-------+ | | +-------+ |
| |Network| | | |Network| |
| |Stack | | | |Stack | |
| +-------+ | | +-------+ |
| ^ | | ^ |
| | | | | |
| v | | v |
| +-----------+ | | +-----------+ |
| |Thunderbolt| | | |Thunderbolt| |
| |Networking | | | |Networking | |
| |Driver | | | |Driver | |
| +-----------+ | | +-----------+ |
| ^ | | ^ |
| | | | | |
| v | | v |
| +-----------+ | | +-----------+ |
| |Thunderbolt| | | |Thunderbolt| |
| |Controller |<-+------------+->|Controller | |
| +-----------+ | | +-----------+ |
+----------------+ +----------------+
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
---
drivers/thunderbolt/icm/icm_nhi.c | 15 +
drivers/thunderbolt/icm/net.c | 1471 +++++++++++++++++++++++++++++++++++++
2 files changed, 1486 insertions(+)
diff --git a/drivers/thunderbolt/icm/icm_nhi.c b/drivers/thunderbolt/icm/icm_nhi.c
index 578eb14..c181abf 100644
--- a/drivers/thunderbolt/icm/icm_nhi.c
+++ b/drivers/thunderbolt/icm/icm_nhi.c
@@ -1036,6 +1036,7 @@ static irqreturn_t nhi_msi(int __always_unused irq, void *data)
{
struct tbt_nhi_ctxt *nhi_ctxt = data;
u32 isr0, isr1, imr0, imr1;
+ int i;
/* clear on read */
isr0 = ioread32(nhi_ctxt->iobase + REG_RING_NOTIFY_BASE);
@@ -1058,6 +1059,20 @@ static irqreturn_t nhi_msi(int __always_unused irq, void *data)
spin_unlock(&nhi_ctxt->lock);
+ for (i = 0; i < nhi_ctxt->num_ports; ++i) {
+ struct net_device *net_dev =
+ nhi_ctxt->net_devices[i].net_dev;
+ if (net_dev) {
+ u8 path = PATH_FROM_PORT(nhi_ctxt->num_paths, i);
+
+ if (isr0 & REG_RING_INT_RX_PROCESSED(
+ path, nhi_ctxt->num_paths))
+ tbt_net_rx_msi(net_dev);
+ if (isr0 & REG_RING_INT_TX_PROCESSED(path))
+ tbt_net_tx_msi(net_dev);
+ }
+ }
+
if (isr0 & REG_RING_INT_RX_PROCESSED(TBT_ICM_RING_NUM,
nhi_ctxt->num_paths))
schedule_work(&nhi_ctxt->icm_msgs_work);
diff --git a/drivers/thunderbolt/icm/net.c b/drivers/thunderbolt/icm/net.c
index acf30ad..a0f3c4a 100644
--- a/drivers/thunderbolt/icm/net.c
+++ b/drivers/thunderbolt/icm/net.c
@@ -134,6 +134,17 @@ struct approve_inter_domain_connection_cmd {
};
+struct tbt_frame_header {
+ /* size of the data with the frame */
+ __le32 frame_size;
+ /* running index on the frames */
+ __le16 frame_index;
+ /* ID of the frame to match frames to specific packet */
+ __le16 frame_id;
+ /* how many frames assembles a full packet */
+ __le32 frame_count;
+};
+
enum neg_event {
RECEIVE_LOGOUT = NUM_MEDIUM_STATUSES,
RECEIVE_LOGIN_RESPONSE,
@@ -141,15 +152,81 @@ enum neg_event {
NUM_NEG_EVENTS
};
+enum frame_status {
+ GOOD_FRAME,
+ GOOD_AS_FIRST_FRAME,
+ GOOD_AS_FIRST_MULTICAST_FRAME,
+ FRAME_NOT_READY,
+ FRAME_ERROR,
+};
+
+enum packet_filter {
+ /* all multicast MAC addresses */
+ PACKET_TYPE_ALL_MULTICAST,
+ /* all types of MAC addresses: multicast, unicast and broadcast */
+ PACKET_TYPE_PROMISCUOUS,
+ /* all unicast MAC addresses */
+ PACKET_TYPE_UNICAST_PROMISCUOUS,
+};
+
enum disconnect_path_stage {
STAGE_1 = BIT(0),
STAGE_2 = BIT(1)
};
+struct tbt_net_stats {
+ u64 tx_packets;
+ u64 tx_bytes;
+ u64 tx_errors;
+ u64 rx_packets;
+ u64 rx_bytes;
+ u64 rx_length_errors;
+ u64 rx_over_errors;
+ u64 rx_crc_errors;
+ u64 rx_missed_errors;
+ u64 multicast;
+};
+
+static const char tbt_net_gstrings_stats[][ETH_GSTRING_LEN] = {
+ "tx_packets",
+ "tx_bytes",
+ "tx_errors",
+ "rx_packets",
+ "rx_bytes",
+ "rx_length_errors",
+ "rx_over_errors",
+ "rx_crc_errors",
+ "rx_missed_errors",
+ "multicast",
+};
+
+struct tbt_buffer {
+ dma_addr_t dma;
+ union {
+ struct tbt_frame_header *hdr;
+ struct page *page;
+ };
+ u32 page_offset;
+};
+
+struct tbt_desc_ring {
+ /* pointer to the descriptor ring memory */
+ struct tbt_buf_desc *desc;
+ /* physical address of the descriptor ring */
+ dma_addr_t dma;
+ /* array of buffer structs */
+ struct tbt_buffer *buffers;
+ /* last descriptor that was associated with a buffer */
+ u16 last_allocated;
+ /* next descriptor to check for DD status bit */
+ u16 next_to_clean;
+};
+
/**
* struct tbt_port - the basic tbt_port structure
* @tbt_nhi_ctxt: context of the nhi controller.
* @net_dev: networking device object.
+ * @napi: network API
* @login_retry_work: work queue for sending login requests.
* @login_response_work: work queue for sending login responses.
* @work_struct logout_work: work queue for sending logout requests.
@@ -165,6 +242,11 @@ enum disconnect_path_stage {
* @login_retry_count: counts number of login retries sent.
* @local_depth: depth of the remote peer in the chain.
* @transmit_path: routing parameter for the icm.
+ * @tx_ring: transmit ring from where the packets are sent.
+ * @rx_ring: receive ring where the packets are received.
+ * @stats: network statistics of the rx/tx packets.
+ * @packet_filters: defines filters for the received packets.
+ * @multicast_hash_table: hash table of multicast addresses.
* @frame_id: counting ID of frames.
* @num: port number.
* @local_path: routing parameter for the icm.
@@ -174,6 +256,7 @@ enum disconnect_path_stage {
struct tbt_port {
struct tbt_nhi_ctxt *nhi_ctxt;
struct net_device *net_dev;
+ struct napi_struct napi;
struct delayed_work login_retry_work;
struct work_struct login_response_work;
struct work_struct logout_work;
@@ -189,6 +272,17 @@ struct tbt_port {
u8 login_retry_count;
u8 local_depth;
u8 transmit_path;
+ struct tbt_desc_ring tx_ring ____cacheline_aligned_in_smp;
+ struct tbt_desc_ring rx_ring;
+ struct tbt_net_stats stats;
+ u32 packet_filters;
+ /*
+ * hash table of 1024 boolean entries with hashing of
+ * the multicast address
+ */
+ u32 multicast_hash_table[DIV_ROUND_UP(
+ TBT_NET_MULTICAST_HASH_TABLE_SIZE,
+ BITS_PER_U32)];
u16 frame_id;
u8 num;
u8 local_path;
@@ -235,6 +329,8 @@ static void tbt_net_tear_down(struct net_device *net_dev, bool send_logout)
(port->local_path * REG_OPTS_STEP);
u32 rx_reg_val = ioread32(rx_reg) & ~REG_OPTS_E2E_EN;
+ napi_disable(&port->napi);
+
tx_reg = iobase + REG_TX_OPTIONS_BASE +
(port->local_path * REG_OPTS_STEP);
tx_reg_val = ioread32(tx_reg) & ~REG_OPTS_E2E_EN;
@@ -276,8 +372,1336 @@ static void tbt_net_tear_down(struct net_device *net_dev, bool send_logout)
port->nhi_ctxt->num_paths);
spin_unlock_irqrestore(&port->nhi_ctxt->lock, flags);
}
+
+ port->rx_ring.next_to_clean = 0;
+ port->rx_ring.last_allocated = TBT_NET_NUM_RX_BUFS - 1;
+
+}
+
+void tbt_net_tx_msi(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+ void __iomem *iobase = port->nhi_ctxt->iobase;
+ u32 prod_cons, prod, cons;
+
+ prod_cons = ioread32(TBT_RING_CONS_PROD_REG(iobase, REG_TX_RING_BASE,
+ port->local_path));
+ prod = TBT_REG_RING_PROD_EXTRACT(prod_cons);
+ cons = TBT_REG_RING_CONS_EXTRACT(prod_cons);
+ if (prod >= TBT_NET_NUM_TX_BUFS || cons >= TBT_NET_NUM_TX_BUFS)
+ return;
+
+ if (TBT_NUM_BUFS_BETWEEN(prod, cons, TBT_NET_NUM_TX_BUFS) >=
+ TX_WAKE_THRESHOLD) {
+ netif_wake_queue(port->net_dev);
+ } else {
+ spin_lock(&port->nhi_ctxt->lock);
+ /* enable TX interrupt */
+ RING_INT_ENABLE_TX(iobase, port->local_path);
+ spin_unlock(&port->nhi_ctxt->lock);
+ }
+}
+
+static irqreturn_t tbt_net_tx_msix(int __always_unused irq, void *data)
+{
+ struct tbt_port *port = data;
+ void __iomem *iobase = port->nhi_ctxt->iobase;
+ u32 prod_cons, prod, cons;
+
+ prod_cons = ioread32(TBT_RING_CONS_PROD_REG(iobase,
+ REG_TX_RING_BASE,
+ port->local_path));
+ prod = TBT_REG_RING_PROD_EXTRACT(prod_cons);
+ cons = TBT_REG_RING_CONS_EXTRACT(prod_cons);
+ if (prod < TBT_NET_NUM_TX_BUFS && cons < TBT_NET_NUM_TX_BUFS &&
+ TBT_NUM_BUFS_BETWEEN(prod, cons, TBT_NET_NUM_TX_BUFS) >=
+ TX_WAKE_THRESHOLD) {
+ spin_lock(&port->nhi_ctxt->lock);
+ /* disable TX interrupt */
+ RING_INT_DISABLE_TX(iobase, port->local_path);
+ spin_unlock(&port->nhi_ctxt->lock);
+
+ netif_wake_queue(port->net_dev);
+ }
+
+ return IRQ_HANDLED;
+}
+
+void tbt_net_rx_msi(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ napi_schedule_irqoff(&port->napi);
+}
+
+static irqreturn_t tbt_net_rx_msix(int __always_unused irq, void *data)
+{
+ struct tbt_port *port = data;
+
+ if (likely(napi_schedule_prep(&port->napi))) {
+ struct tbt_nhi_ctxt *nhi_ctx = port->nhi_ctxt;
+
+ spin_lock(&nhi_ctx->lock);
+ /* disable RX interrupt */
+ RING_INT_DISABLE_RX(nhi_ctx->iobase, port->local_path,
+ nhi_ctx->num_paths);
+ spin_unlock(&nhi_ctx->lock);
+
+ __napi_schedule_irqoff(&port->napi);
+ }
+
+ return IRQ_HANDLED;
+}
+
+static void tbt_net_pull_tail(struct sk_buff *skb)
+{
+ skb_frag_t *frag = &skb_shinfo(skb)->frags[0];
+ unsigned int pull_len;
+ unsigned char *va;
+
+ /*
+ * it is valid to use page_address instead of kmap since we are
+ * working with pages allocated out of the lomem pool
+ */
+ va = skb_frag_address(frag);
+
+ pull_len = eth_get_headlen(va, TBT_NET_RX_HDR_SIZE);
+
+ /* align pull length to size of long to optimize memcpy performance */
+ skb_copy_to_linear_data(skb, va, ALIGN(pull_len, sizeof(long)));
+
+ /* update all of the pointers */
+ skb_frag_size_sub(frag, pull_len);
+ frag->page_offset += pull_len;
+ skb->data_len -= pull_len;
+ skb->tail += pull_len;
+}
+
+static inline bool tbt_net_alloc_mapped_page(struct device *dev,
+ struct tbt_buffer *buf, gfp_t gfp)
+{
+ if (!buf->page) {
+ buf->page = alloc_page(gfp | __GFP_COLD);
+ if (unlikely(!buf->page))
+ return false;
+
+ buf->dma = dma_map_page(dev, buf->page, 0, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ if (dma_mapping_error(dev, buf->dma)) {
+ __free_page(buf->page);
+ buf->page = NULL;
+ return false;
+ }
+ buf->page_offset = 0;
+ }
+ return true;
+}
+
+static bool tbt_net_alloc_rx_buffers(struct device *dev,
+ struct tbt_desc_ring *rx_ring,
+ u16 cleaned_count, void __iomem *reg,
+ gfp_t gfp)
+{
+ u16 i = (rx_ring->last_allocated + 1) & (TBT_NET_NUM_RX_BUFS - 1);
+ bool res = false;
+
+ while (cleaned_count--) {
+ struct tbt_buf_desc *desc = &rx_ring->desc[i];
+ struct tbt_buffer *buf = &rx_ring->buffers[i];
+
+ /* making sure next_to_clean won't get old buffer */
+ desc->attributes = cpu_to_le32(DESC_ATTR_REQ_STS |
+ DESC_ATTR_INT_EN);
+ if (tbt_net_alloc_mapped_page(dev, buf, gfp)) {
+ res = true;
+ rx_ring->last_allocated = i;
+ i = (i + 1) & (TBT_NET_NUM_RX_BUFS - 1);
+ desc->phys = cpu_to_le64(buf->dma + buf->page_offset);
+ } else {
+ break;
+ }
+ }
+
+ if (res) {
+ iowrite32((rx_ring->last_allocated << REG_RING_CONS_SHIFT) &
+ REG_RING_CONS_MASK, reg);
+ }
+
+ return res;
+}
+
+static inline bool tbt_net_multicast_mac_set(const u32 *multicast_hash_table,
+ const u8 *ether_addr)
+{
+ u16 hash_val = TBT_NET_ETHER_ADDR_HASH(ether_addr);
+
+ return !!(multicast_hash_table[hash_val / BITS_PER_U32] &
+ BIT(hash_val % BITS_PER_U32));
+}
+
+static enum frame_status tbt_net_check_frame(struct tbt_port *port,
+ u16 frame_num, u32 *count,
+ u16 index, u16 *id, u32 *size)
+{
+ struct tbt_desc_ring *rx_ring = &port->rx_ring;
+ __le32 desc_attr = rx_ring->desc[frame_num].attributes;
+ enum frame_status res = GOOD_AS_FIRST_FRAME;
+ u32 len, frame_count, frame_size;
+ struct tbt_frame_header *hdr;
+
+ if (!(desc_attr & cpu_to_le32(DESC_ATTR_DESC_DONE)))
+ return FRAME_NOT_READY;
+
+ rmb(); /* read other fields from desc after checking DD */
+
+ if (unlikely(desc_attr & cpu_to_le32(DESC_ATTR_RX_CRC_ERR))) {
+ ++port->stats.rx_crc_errors;
+ goto err;
+ } else if (unlikely(desc_attr &
+ cpu_to_le32(DESC_ATTR_RX_BUF_OVRN_ERR))) {
+ ++port->stats.rx_over_errors;
+ goto err;
+ }
+
+ len = (le32_to_cpu(desc_attr) & DESC_ATTR_LEN_MASK)
+ >> DESC_ATTR_LEN_SHIFT;
+ if (len == 0)
+ len = TBT_RING_MAX_FRAME_SIZE;
+ /* should be greater than just header i.e. contains data */
+ if (unlikely(len <= sizeof(struct tbt_frame_header))) {
+ ++port->stats.rx_length_errors;
+ goto err;
+ }
+
+ prefetchw(rx_ring->buffers[frame_num].page);
+ hdr = page_address(rx_ring->buffers[frame_num].page) +
+ rx_ring->buffers[frame_num].page_offset;
+ /* prefetch first cache line of first page */
+ prefetch(hdr);
+
+ /* we are reusing so sync this buffer for CPU use */
+ dma_sync_single_range_for_cpu(&port->nhi_ctxt->pdev->dev,
+ rx_ring->buffers[frame_num].dma,
+ rx_ring->buffers[frame_num].page_offset,
+ TBT_RING_MAX_FRAME_SIZE,
+ DMA_FROM_DEVICE);
+
+ frame_count = le32_to_cpu(hdr->frame_count);
+ frame_size = le32_to_cpu(hdr->frame_size);
+
+ if (unlikely((frame_size > len - sizeof(struct tbt_frame_header)) ||
+ (frame_size == 0))) {
+ ++port->stats.rx_length_errors;
+ goto err;
+ }
+ /*
+ * In case we're in the middle of packet, validate the frame header
+ * based on first fragment of the packet
+ */
+ if (*count) {
+ /* check the frame count fits the count field */
+ if (frame_count != *count) {
+ ++port->stats.rx_length_errors;
+ goto check_as_first;
+ }
+
+ /*
+ * check the frame identifiers are incremented correctly,
+ * and id is matching
+ */
+ if ((le16_to_cpu(hdr->frame_index) != index) ||
+ (le16_to_cpu(hdr->frame_id) != *id)) {
+ ++port->stats.rx_missed_errors;
+ goto check_as_first;
+ }
+
+ *size += frame_size;
+ if (*size > TBT_NET_MTU) {
+ ++port->stats.rx_length_errors;
+ goto err;
+ }
+ res = GOOD_FRAME;
+ } else { /* start of packet, validate the frame header */
+ const u8 *addr;
+
+check_as_first:
+ rx_ring->next_to_clean = frame_num;
+
+ /* validate the first packet has a valid frame count */
+ if (unlikely(frame_count == 0 ||
+ frame_count > (TBT_NET_NUM_RX_BUFS / 4))) {
+ ++port->stats.rx_length_errors;
+ goto err;
+ }
+
+ /* validate the first packet has a valid frame index */
+ if (hdr->frame_index != 0) {
+ ++port->stats.rx_missed_errors;
+ goto err;
+ }
+
+ BUILD_BUG_ON(TBT_NET_RX_HDR_SIZE > TBT_RING_MAX_FRM_DATA_SZ);
+ if ((frame_count > 1) && (frame_size < TBT_NET_RX_HDR_SIZE)) {
+ ++port->stats.rx_length_errors;
+ goto err;
+ }
+
+ addr = (u8 *)(hdr + 1);
+
+ /* check the packet can go through the filter */
+ if (is_multicast_ether_addr(addr)) {
+ if (!is_broadcast_ether_addr(addr)) {
+ if ((port->packet_filters &
+ (BIT(PACKET_TYPE_PROMISCUOUS) |
+ BIT(PACKET_TYPE_ALL_MULTICAST))) ||
+ tbt_net_multicast_mac_set(
+ port->multicast_hash_table, addr))
+ res = GOOD_AS_FIRST_MULTICAST_FRAME;
+ else
+ goto err;
+ }
+ } else if (!(port->packet_filters &
+ (BIT(PACKET_TYPE_PROMISCUOUS) |
+ BIT(PACKET_TYPE_UNICAST_PROMISCUOUS))) &&
+ !ether_addr_equal(port->net_dev->dev_addr, addr)) {
+ goto err;
+ }
+
+ *size = frame_size;
+ *count = frame_count;
+ *id = le16_to_cpu(hdr->frame_id);
+ }
+
+#if (PREFETCH_STRIDE < 128)
+ prefetch((u8 *)hdr + PREFETCH_STRIDE);
+#endif
+
+ return res;
+
+err:
+ rx_ring->next_to_clean = (frame_num + 1) & (TBT_NET_NUM_RX_BUFS - 1);
+ return FRAME_ERROR;
+}
+
+static inline unsigned int tbt_net_max_frm_data_size(
+ __maybe_unused u32 frame_size)
+{
+#if (TBT_NUM_FRAMES_PER_PAGE > 1)
+ return ALIGN(frame_size + sizeof(struct tbt_frame_header),
+ L1_CACHE_BYTES) -
+ sizeof(struct tbt_frame_header);
+#else
+ return TBT_RING_MAX_FRM_DATA_SZ;
+#endif
+}
+
+static int tbt_net_poll(struct napi_struct *napi, int budget)
+{
+ struct tbt_port *port = container_of(napi, struct tbt_port, napi);
+ void __iomem *reg = TBT_RING_CONS_PROD_REG(port->nhi_ctxt->iobase,
+ REG_RX_RING_BASE,
+ port->local_path);
+ struct tbt_desc_ring *rx_ring = &port->rx_ring;
+ u16 cleaned_count = TBT_NUM_BUFS_BETWEEN(rx_ring->last_allocated,
+ rx_ring->next_to_clean,
+ TBT_NET_NUM_RX_BUFS);
+ unsigned long flags;
+ int rx_packets = 0;
+
+loop:
+ while (likely(rx_packets < budget)) {
+ struct sk_buff *skb;
+ enum frame_status status;
+ bool multicast = false;
+ u32 frame_count = 0, size;
+ u16 j, frame_id;
+ int i;
+
+ /*
+ * return some buffers to hardware, one at a time is too slow
+ * so allocate TBT_NET_RX_BUFFER_WRITE buffers at the same time
+ */
+ if (cleaned_count >= TBT_NET_RX_BUFFER_WRITE) {
+ tbt_net_alloc_rx_buffers(&port->nhi_ctxt->pdev->dev,
+ rx_ring, cleaned_count, reg,
+ GFP_ATOMIC);
+ cleaned_count = 0;
+ }
+
+ status = tbt_net_check_frame(port, rx_ring->next_to_clean,
+ &frame_count, 0, &frame_id,
+ &size);
+ if (status == FRAME_NOT_READY)
+ break;
+
+ if (status == FRAME_ERROR) {
+ ++cleaned_count;
+ continue;
+ }
+
+ multicast = (status == GOOD_AS_FIRST_MULTICAST_FRAME);
+
+ /*
+ * i is incremented up to the frame_count frames received,
+ * j cyclicly goes over the location from the next frame
+ * to clean in the ring
+ */
+ j = (rx_ring->next_to_clean + 1);
+ j &= (TBT_NET_NUM_RX_BUFS - 1);
+ for (i = 1; i < frame_count; ++i) {
+ status = tbt_net_check_frame(port, j, &frame_count, i,
+ &frame_id, &size);
+ if (status == FRAME_NOT_READY)
+ goto out;
+
+ j = (j + 1) & (TBT_NET_NUM_RX_BUFS - 1);
+
+ /* if a new frame is found, start over */
+ if (status == GOOD_AS_FIRST_FRAME ||
+ status == GOOD_AS_FIRST_MULTICAST_FRAME) {
+ multicast = (status ==
+ GOOD_AS_FIRST_MULTICAST_FRAME);
+ cleaned_count += i;
+ i = 0;
+ continue;
+ }
+
+ if (status == FRAME_ERROR) {
+ cleaned_count += (i + 1);
+ goto loop;
+ }
+ }
+
+ /* allocate a skb to store the frags */
+ skb = netdev_alloc_skb_ip_align(port->net_dev,
+ TBT_NET_RX_HDR_SIZE);
+ if (unlikely(!skb))
+ break;
+
+ /*
+ * we will be copying header into skb->data in
+ * tbt_net_pull_tail so it is in our interest to prefetch
+ * it now to avoid a possible cache miss
+ */
+ prefetchw(skb->data);
+
+ /*
+ * if overall size of packet smaller than TBT_NET_RX_HDR_SIZE
+ * which is a small buffer size we decided to allocate
+ * as the base to RX
+ */
+ if (size <= TBT_NET_RX_HDR_SIZE) {
+ struct tbt_buffer *buf =
+ &(rx_ring->buffers[rx_ring->next_to_clean]);
+ u8 *va = page_address(buf->page) + buf->page_offset +
+ sizeof(struct tbt_frame_header);
+
+ memcpy(__skb_put(skb, size), va,
+ ALIGN(size, sizeof(long)));
+
+ /*
+ * Reuse buffer as-is,
+ * just make sure it is local
+ * Access to local memory is faster than non-local
+ * memory so let's reuse.
+ * If not local, let's free it and reallocate later.
+ */
+ if (likely(page_to_nid(buf->page) == numa_node_id()))
+ /* sync the buffer for use by the device */
+ dma_sync_single_range_for_device(
+ &port->nhi_ctxt->pdev->dev,
+ buf->dma, buf->page_offset,
+ TBT_RING_MAX_FRAME_SIZE,
+ DMA_FROM_DEVICE);
+ else {
+ /* this page cannot be reused so discard it */
+ put_page(buf->page);
+ buf->page = NULL;
+ dma_unmap_page(&port->nhi_ctxt->pdev->dev,
+ buf->dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ }
+ rx_ring->next_to_clean = (rx_ring->next_to_clean + 1) &
+ (TBT_NET_NUM_RX_BUFS - 1);
+ } else {
+ for (i = 0; i < frame_count; ++i) {
+ struct tbt_buffer *buf = &(rx_ring->buffers[
+ rx_ring->next_to_clean]);
+ struct tbt_frame_header *hdr =
+ page_address(buf->page) +
+ buf->page_offset;
+ u32 frm_size = le32_to_cpu(hdr->frame_size);
+
+ unsigned int truesize =
+ tbt_net_max_frm_data_size(frm_size);
+
+ /* add frame to skb struct */
+ skb_add_rx_frag(skb, skb_shinfo(skb)->nr_frags,
+ buf->page,
+ sizeof(struct tbt_frame_header)
+ + buf->page_offset,
+ frm_size, truesize);
+
+#if (TBT_NUM_FRAMES_PER_PAGE > 1)
+ /* move offset up to the next cache line */
+ buf->page_offset += (truesize +
+ sizeof(struct tbt_frame_header));
+
+ /*
+ * we can reuse buffer if there is space
+ * available and it is local
+ */
+ if (page_to_nid(buf->page) == numa_node_id()
+ && buf->page_offset <=
+ PAGE_SIZE - TBT_RING_MAX_FRAME_SIZE) {
+ /*
+ * bump ref count on page before
+ * it is given to the stack
+ */
+ get_page(buf->page);
+ /*
+ * sync the buffer for use by the
+ * device
+ */
+ dma_sync_single_range_for_device(
+ &port->nhi_ctxt->pdev->dev,
+ buf->dma, buf->page_offset,
+ TBT_RING_MAX_FRAME_SIZE,
+ DMA_FROM_DEVICE);
+ } else
+#endif
+ {
+ buf->page = NULL;
+ dma_unmap_page(
+ &port->nhi_ctxt->pdev->dev,
+ buf->dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ }
+
+ rx_ring->next_to_clean =
+ (rx_ring->next_to_clean + 1) &
+ (TBT_NET_NUM_RX_BUFS - 1);
+ }
+ /*
+ * place header from the first
+ * fragment in linear portion of buffer
+ */
+ tbt_net_pull_tail(skb);
+ }
+
+ /*
+ * The Thunderbolt medium doesn't have any restriction on
+ * minimum frame size, thus doesn't need any padding in
+ * transmit.
+ * The network stack accepts Runt Ethernet frames,
+ * therefor there is neither padding in receive.
+ */
+
+ skb->protocol = eth_type_trans(skb, port->net_dev);
+ napi_gro_receive(&port->napi, skb);
+
+ ++rx_packets;
+ port->stats.rx_bytes += size;
+ if (multicast)
+ ++port->stats.multicast;
+ cleaned_count += frame_count;
+ }
+
+out:
+ port->stats.rx_packets += rx_packets;
+
+ if (cleaned_count)
+ tbt_net_alloc_rx_buffers(&port->nhi_ctxt->pdev->dev,
+ rx_ring, cleaned_count, reg,
+ GFP_ATOMIC);
+
+ /* If all work not completed, return budget and keep polling */
+ if (rx_packets >= budget)
+ return budget;
+
+ /* Work is done so exit the polling mode and re-enable the interrupt */
+ napi_complete(napi);
+
+ spin_lock_irqsave(&port->nhi_ctxt->lock, flags);
+ /* enable RX interrupt */
+ RING_INT_ENABLE_RX(port->nhi_ctxt->iobase, port->local_path,
+ port->nhi_ctxt->num_paths);
+
+ spin_unlock_irqrestore(&port->nhi_ctxt->lock, flags);
+
+ return 0;
+}
+
+static int tbt_net_open(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+ int res = 0;
+ int i, j;
+
+ /* change link state to off until path establishment finishes */
+ netif_carrier_off(net_dev);
+
+ /*
+ * if we previously succeeded to allocate msix entries,
+ * now request IRQ for them:
+ * 2=tx data port 0,
+ * 3=rx data port 0,
+ * 4=tx data port 1,
+ * 5=rx data port 1,
+ * ...
+ * if not, if msi is used, nhi_msi will handle icm & data paths
+ */
+ if (port->nhi_ctxt->msix_entries) {
+ char name[] = "tbt-net-xx-xx";
+
+ scnprintf(name, sizeof(name), "tbt-net-rx-%02u", port->num);
+ res = devm_request_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[3+(port->num*2)].vector,
+ tbt_net_rx_msix, 0, name, port);
+ if (res) {
+ netif_err(port, ifup, net_dev, "request_irq %s failed %d\n",
+ name, res);
+ goto out;
+ }
+ name[8] = 't';
+ res = devm_request_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[2+(port->num*2)].vector,
+ tbt_net_tx_msix, 0, name, port);
+ if (res) {
+ netif_err(port, ifup, net_dev, "request_irq %s failed %d\n",
+ name, res);
+ goto request_irq_failure;
+ }
+ }
+ /*
+ * Verifying that all buffer sizes are well defined.
+ * Starting with frame(s) will not tip over the
+ * page boundary
+ */
+ BUILD_BUG_ON(TBT_NUM_FRAMES_PER_PAGE < 1);
+ /*
+ * Just to make sure we have enough place for containing
+ * 3 max MTU packets for TX
+ */
+ BUILD_BUG_ON((TBT_NET_NUM_TX_BUFS * TBT_RING_MAX_FRAME_SIZE) <
+ (TBT_NET_MTU * 3));
+ /* make sure the number of TX Buffers is power of 2 */
+ BUILD_BUG_ON_NOT_POWER_OF_2(TBT_NET_NUM_TX_BUFS);
+ /*
+ * Just to make sure we have enough place for containing
+ * 3 max MTU packets for RX
+ */
+ BUILD_BUG_ON((TBT_NET_NUM_RX_BUFS * TBT_RING_MAX_FRAME_SIZE) <
+ (TBT_NET_MTU * 3));
+ /* make sure the number of RX Buffers is power of 2 */
+ BUILD_BUG_ON_NOT_POWER_OF_2(TBT_NET_NUM_RX_BUFS);
+
+ port->rx_ring.last_allocated = TBT_NET_NUM_RX_BUFS - 1;
+
+ port->tx_ring.buffers = vzalloc(TBT_NET_NUM_TX_BUFS *
+ sizeof(struct tbt_buffer));
+ if (!port->tx_ring.buffers)
+ goto ring_alloc_failure;
+ port->rx_ring.buffers = vzalloc(TBT_NET_NUM_RX_BUFS *
+ sizeof(struct tbt_buffer));
+ if (!port->rx_ring.buffers)
+ goto ring_alloc_failure;
+
+ /*
+ * Allocate TX and RX descriptors
+ * if the total size is less than a page, do a central allocation
+ * Otherwise, split TX and RX
+ */
+ if (TBT_NET_SIZE_TOTAL_DESCS <= PAGE_SIZE) {
+ port->tx_ring.desc = dmam_alloc_coherent(
+ &port->nhi_ctxt->pdev->dev,
+ TBT_NET_SIZE_TOTAL_DESCS,
+ &port->tx_ring.dma,
+ GFP_KERNEL | __GFP_ZERO);
+ if (!port->tx_ring.desc)
+ goto ring_alloc_failure;
+ /* RX starts where TX finishes */
+ port->rx_ring.desc = &port->tx_ring.desc[TBT_NET_NUM_TX_BUFS];
+ port->rx_ring.dma = port->tx_ring.dma +
+ (TBT_NET_NUM_TX_BUFS * sizeof(struct tbt_buf_desc));
+ } else {
+ port->tx_ring.desc = dmam_alloc_coherent(
+ &port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_TX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ &port->tx_ring.dma,
+ GFP_KERNEL | __GFP_ZERO);
+ if (!port->tx_ring.desc)
+ goto ring_alloc_failure;
+ port->rx_ring.desc = dmam_alloc_coherent(
+ &port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_RX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ &port->rx_ring.dma,
+ GFP_KERNEL | __GFP_ZERO);
+ if (!port->rx_ring.desc)
+ goto rx_desc_alloc_failure;
+ }
+
+ /* allocate TX buffers and configure the descriptors */
+ for (i = 0; i < TBT_NET_NUM_TX_BUFS; i++) {
+ port->tx_ring.buffers[i].hdr = dma_alloc_coherent(
+ &port->nhi_ctxt->pdev->dev,
+ TBT_NUM_FRAMES_PER_PAGE * TBT_RING_MAX_FRAME_SIZE,
+ &port->tx_ring.buffers[i].dma,
+ GFP_KERNEL);
+ if (!port->tx_ring.buffers[i].hdr)
+ goto buffers_alloc_failure;
+
+ port->tx_ring.desc[i].phys =
+ cpu_to_le64(port->tx_ring.buffers[i].dma);
+ port->tx_ring.desc[i].attributes =
+ cpu_to_le32(DESC_ATTR_REQ_STS |
+ TBT_NET_DESC_ATTR_SOF_EOF);
+
+ /*
+ * In case the page is bigger than the frame size,
+ * make the next buffer descriptor points
+ * on the next frame memory address within the page
+ */
+ for (i++, j = 1; (i < TBT_NET_NUM_TX_BUFS) &&
+ (j < TBT_NUM_FRAMES_PER_PAGE); i++, j++) {
+ port->tx_ring.buffers[i].dma =
+ port->tx_ring.buffers[i - 1].dma +
+ TBT_RING_MAX_FRAME_SIZE;
+ port->tx_ring.buffers[i].hdr =
+ (void *)(port->tx_ring.buffers[i - 1].hdr) +
+ TBT_RING_MAX_FRAME_SIZE;
+ /* move the next offset i.e. TBT_RING_MAX_FRAME_SIZE */
+ port->tx_ring.buffers[i].page_offset =
+ port->tx_ring.buffers[i - 1].page_offset +
+ TBT_RING_MAX_FRAME_SIZE;
+ port->tx_ring.desc[i].phys =
+ cpu_to_le64(port->tx_ring.buffers[i].dma);
+ port->tx_ring.desc[i].attributes =
+ cpu_to_le32(DESC_ATTR_REQ_STS |
+ TBT_NET_DESC_ATTR_SOF_EOF);
+ }
+ i--;
+ }
+
+ port->negotiation_status =
+ BIT(port->nhi_ctxt->net_devices[port->num].medium_sts);
+ if (port->negotiation_status == BIT(MEDIUM_READY_FOR_CONNECTION)) {
+ port->login_retry_count = 0;
+ queue_delayed_work(port->nhi_ctxt->net_workqueue,
+ &port->login_retry_work, 0);
+ }
+
+ netif_info(port, ifup, net_dev, "Thunderbolt(TM) Networking port %u - ready for ThunderboltIP negotiation\n",
+ port->num);
+ return 0;
+
+buffers_alloc_failure:
+ /*
+ * Rollback the Tx buffers that were already allocated
+ * until the failure
+ */
+ for (i--; i >= 0; i--) {
+ /* free only for first buffer allocation */
+ if (port->tx_ring.buffers[i].page_offset == 0)
+ dma_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NUM_FRAMES_PER_PAGE *
+ TBT_RING_MAX_FRAME_SIZE,
+ port->tx_ring.buffers[i].hdr,
+ port->tx_ring.buffers[i].dma);
+ port->tx_ring.buffers[i].hdr = NULL;
+ }
+ /*
+ * For central allocation, free all
+ * otherwise free RX and then TX separately
+ */
+ if (TBT_NET_SIZE_TOTAL_DESCS <= PAGE_SIZE) {
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_SIZE_TOTAL_DESCS,
+ port->tx_ring.desc,
+ port->tx_ring.dma);
+ port->rx_ring.desc = NULL;
+ } else {
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_RX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ port->rx_ring.desc,
+ port->rx_ring.dma);
+ port->rx_ring.desc = NULL;
+rx_desc_alloc_failure:
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_TX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ port->tx_ring.desc,
+ port->tx_ring.dma);
+ }
+ port->tx_ring.desc = NULL;
+ring_alloc_failure:
+ vfree(port->tx_ring.buffers);
+ port->tx_ring.buffers = NULL;
+ vfree(port->rx_ring.buffers);
+ port->rx_ring.buffers = NULL;
+ res = -ENOMEM;
+ netif_err(port, ifup, net_dev, "Thunderbolt(TM) Networking port %u - unable to allocate memory\n",
+ port->num);
+
+ if (!port->nhi_ctxt->msix_entries)
+ goto out;
+
+ devm_free_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[2 + (port->num * 2)].vector,
+ port);
+request_irq_failure:
+ devm_free_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[3 + (port->num * 2)].vector,
+ port);
+out:
+ return res;
+}
+
+static int tbt_net_close(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+ int i;
+
+ /*
+ * Close connection, disable rings, flow controls
+ * and interrupts
+ */
+ tbt_net_tear_down(net_dev, !(port->negotiation_status &
+ BIT(RECEIVE_LOGOUT)));
+
+ cancel_work_sync(&port->login_response_work);
+ cancel_work_sync(&port->logout_work);
+ cancel_work_sync(&port->status_reply_work);
+ cancel_work_sync(&port->approve_inter_domain_work);
+
+ /* Rollback the Tx buffers that were allocated */
+ for (i = 0; i < TBT_NET_NUM_TX_BUFS; i++) {
+ if (port->tx_ring.buffers[i].page_offset == 0)
+ dma_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NUM_FRAMES_PER_PAGE *
+ TBT_RING_MAX_FRAME_SIZE,
+ port->tx_ring.buffers[i].hdr,
+ port->tx_ring.buffers[i].dma);
+ port->tx_ring.buffers[i].hdr = NULL;
+ }
+ /* Unmap the Rx buffers that were allocated */
+ for (i = 0; i < TBT_NET_NUM_RX_BUFS; i++)
+ if (port->rx_ring.buffers[i].page) {
+ put_page(port->rx_ring.buffers[i].page);
+ port->rx_ring.buffers[i].page = NULL;
+ dma_unmap_page(&port->nhi_ctxt->pdev->dev,
+ port->rx_ring.buffers[i].dma, PAGE_SIZE,
+ DMA_FROM_DEVICE);
+ }
+
+ /*
+ * For central allocation, free all
+ * otherwise free RX and then TX separately
+ */
+ if (TBT_NET_SIZE_TOTAL_DESCS <= PAGE_SIZE) {
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_SIZE_TOTAL_DESCS,
+ port->tx_ring.desc,
+ port->tx_ring.dma);
+ port->rx_ring.desc = NULL;
+ } else {
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_RX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ port->rx_ring.desc,
+ port->rx_ring.dma);
+ port->rx_ring.desc = NULL;
+ dmam_free_coherent(&port->nhi_ctxt->pdev->dev,
+ TBT_NET_NUM_TX_BUFS *
+ sizeof(struct tbt_buf_desc),
+ port->tx_ring.desc,
+ port->tx_ring.dma);
+ }
+ port->tx_ring.desc = NULL;
+
+ vfree(port->tx_ring.buffers);
+ port->tx_ring.buffers = NULL;
+ vfree(port->rx_ring.buffers);
+ port->rx_ring.buffers = NULL;
+
+ devm_free_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[3 + (port->num * 2)].vector,
+ port);
+ devm_free_irq(&port->nhi_ctxt->pdev->dev,
+ port->nhi_ctxt->msix_entries[2 + (port->num * 2)].vector,
+ port);
+
+ netif_info(port, ifdown, net_dev, "Thunderbolt(TM) Networking port %u - is down\n",
+ port->num);
+
+ return 0;
+}
+
+static bool tbt_net_xmit_csum(struct sk_buff *skb,
+ struct tbt_desc_ring *tx_ring, u32 first,
+ u32 last, u32 frame_count)
+{
+
+ struct tbt_frame_header *hdr = tx_ring->buffers[first].hdr;
+ __wsum wsum = (__force __wsum)htonl(skb->len -
+ skb_transport_offset(skb));
+ int offset = skb_transport_offset(skb);
+ __sum16 *tucso; /* TCP UDP Checksum Segment Offset */
+ __be16 protocol = skb->protocol;
+ u8 *dest = (u8 *)(hdr + 1);
+ int len;
+
+ if (skb->ip_summed != CHECKSUM_PARTIAL) {
+ for (; first != last;
+ first = (first + 1) & (TBT_NET_NUM_TX_BUFS - 1)) {
+ hdr = tx_ring->buffers[first].hdr;
+ hdr->frame_count = cpu_to_le32(frame_count);
+ }
+ return true;
+ }
+
+ if (protocol == htons(ETH_P_8021Q)) {
+ struct vlan_hdr *vhdr, vh;
+
+ vhdr = skb_header_pointer(skb, ETH_HLEN, sizeof(vh), &vh);
+ if (!vhdr)
+ return false;
+
+ protocol = vhdr->h_vlan_encapsulated_proto;
+ }
+
+ /*
+ * Data points on the beginning of packet.
+ * Check is the checksum absolute place in the
+ * packet.
+ * ipcso will update IP checksum.
+ * tucso will update TCP/UPD checksum.
+ */
+ if (protocol == htons(ETH_P_IP)) {
+ __sum16 *ipcso = (__sum16 *)(dest +
+ ((u8 *)&(ip_hdr(skb)->check) - skb->data));
+
+ *ipcso = 0;
+ *ipcso = ip_fast_csum(dest + skb_network_offset(skb),
+ ip_hdr(skb)->ihl);
+ if (ip_hdr(skb)->protocol == IPPROTO_TCP)
+ tucso = (__sum16 *)(dest +
+ ((u8 *)&(tcp_hdr(skb)->check) - skb->data));
+ else if (ip_hdr(skb)->protocol == IPPROTO_UDP)
+ tucso = (__sum16 *)(dest +
+ ((u8 *)&(udp_hdr(skb)->check) - skb->data));
+ else
+ return false;
+
+ *tucso = ~csum_tcpudp_magic(ip_hdr(skb)->saddr,
+ ip_hdr(skb)->daddr, 0,
+ ip_hdr(skb)->protocol, 0);
+ } else if (skb_is_gso(skb)) {
+ if (skb_is_gso_v6(skb)) {
+ tucso = (__sum16 *)(dest +
+ ((u8 *)&(tcp_hdr(skb)->check) - skb->data));
+ *tucso = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
+ &ipv6_hdr(skb)->daddr,
+ 0, IPPROTO_TCP, 0);
+ } else if ((protocol == htons(ETH_P_IPV6)) &&
+ (skb_shinfo(skb)->gso_type & SKB_GSO_UDP)) {
+ tucso = (__sum16 *)(dest +
+ ((u8 *)&(udp_hdr(skb)->check) - skb->data));
+ *tucso = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
+ &ipv6_hdr(skb)->daddr,
+ 0, IPPROTO_UDP, 0);
+ } else {
+ return false;
+ }
+ } else if (protocol == htons(ETH_P_IPV6)) {
+ tucso = (__sum16 *)(dest + skb_checksum_start_offset(skb) +
+ skb->csum_offset);
+ *tucso = ~csum_ipv6_magic(&ipv6_hdr(skb)->saddr,
+ &ipv6_hdr(skb)->daddr,
+ 0, ipv6_hdr(skb)->nexthdr, 0);
+ } else {
+ return false;
+ }
+
+ /* First frame was headers, rest of the frames is data */
+ for (; first != last; first = (first + 1) & (TBT_NET_NUM_TX_BUFS - 1),
+ offset = 0) {
+ hdr = tx_ring->buffers[first].hdr;
+ dest = (u8 *)(hdr + 1) + offset;
+ len = le32_to_cpu(hdr->frame_size) - offset;
+ wsum = csum_partial(dest, len, wsum);
+ hdr->frame_count = cpu_to_le32(frame_count);
+ }
+ *tucso = csum_fold(wsum);
+
+ return true;
+}
+
+static netdev_tx_t tbt_net_xmit_frame(struct sk_buff *skb,
+ struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+ void __iomem *iobase = port->nhi_ctxt->iobase;
+ void __iomem *reg = TBT_RING_CONS_PROD_REG(iobase,
+ REG_TX_RING_BASE,
+ port->local_path);
+ struct tbt_desc_ring *tx_ring = &port->tx_ring;
+ struct tbt_frame_header *hdr;
+ u32 prod_cons, prod, cons, first;
+ /* len equivalent to the fragment length */
+ unsigned int len = skb_headlen(skb);
+ /* data_len is overall packet length */
+ unsigned int data_len = skb->len;
+ u32 frm_idx, frag_num = 0;
+ const u8 *src = skb->data;
+ bool unmap = false;
+ __le32 *attr;
+ u8 *dest;
+
+ if (unlikely(data_len == 0 || data_len > TBT_NET_MTU))
+ goto invalid_packet;
+
+ prod_cons = ioread32(reg);
+ prod = TBT_REG_RING_PROD_EXTRACT(prod_cons);
+ cons = TBT_REG_RING_CONS_EXTRACT(prod_cons);
+ if (prod >= TBT_NET_NUM_TX_BUFS || cons >= TBT_NET_NUM_TX_BUFS)
+ goto tx_error;
+
+ if (data_len > (TBT_NUM_BUFS_BETWEEN(prod, cons, TBT_NET_NUM_TX_BUFS) *
+ TBT_RING_MAX_FRM_DATA_SZ)) {
+ unsigned long flags;
+
+ netif_stop_queue(net_dev);
+
+ spin_lock_irqsave(&port->nhi_ctxt->lock, flags);
+ /*
+ * Enable TX interrupt to be notified about available buffers
+ * and restart transmission upon this.
+ */
+ RING_INT_ENABLE_TX(iobase, port->local_path);
+ spin_unlock_irqrestore(&port->nhi_ctxt->lock, flags);
+
+ return NETDEV_TX_BUSY;
+ }
+
+ first = prod;
+ attr = &tx_ring->desc[prod].attributes;
+ hdr = tx_ring->buffers[prod].hdr;
+ dest = (u8 *)(hdr + 1);
+ /* if overall packet is bigger than the frame data size */
+ for (frm_idx = 0; data_len > TBT_RING_MAX_FRM_DATA_SZ; ++frm_idx) {
+ u32 size_left = TBT_RING_MAX_FRM_DATA_SZ;
+
+ *attr &= cpu_to_le32(~(DESC_ATTR_LEN_MASK |
+ DESC_ATTR_INT_EN |
+ DESC_ATTR_DESC_DONE));
+ hdr->frame_size = cpu_to_le32(TBT_RING_MAX_FRM_DATA_SZ);
+ hdr->frame_index = cpu_to_le16(frm_idx);
+ hdr->frame_id = cpu_to_le16(port->frame_id);
+
+ do {
+ if (len > size_left) {
+ /*
+ * Copy data onto tx buffer data with full
+ * frame size then break
+ * and go to next frame
+ */
+ memcpy(dest, src, size_left);
+ len -= size_left;
+ dest += size_left;
+ src += size_left;
+ break;
+ }
+
+ memcpy(dest, src, len);
+ size_left -= len;
+ dest += len;
+
+ if (unmap) {
+ kunmap_atomic((void *)src);
+ unmap = false;
+ }
+ /*
+ * Ensure all fragments have been processed
+ */
+ if (frag_num < skb_shinfo(skb)->nr_frags) {
+ const skb_frag_t *frag =
+ &(skb_shinfo(skb)->frags[frag_num]);
+ len = skb_frag_size(frag);
+ /* map and then unmap quickly */
+ src = kmap_atomic(skb_frag_page(frag)) +
+ frag->page_offset;
+ unmap = true;
+ ++frag_num;
+ } else if (unlikely(size_left > 0)) {
+ goto invalid_packet;
+ }
+ } while (size_left > 0);
+
+ data_len -= TBT_RING_MAX_FRM_DATA_SZ;
+ prod = (prod + 1) & (TBT_NET_NUM_TX_BUFS - 1);
+ attr = &tx_ring->desc[prod].attributes;
+ hdr = tx_ring->buffers[prod].hdr;
+ dest = (u8 *)(hdr + 1);
+ }
+
+ *attr &= cpu_to_le32(~(DESC_ATTR_LEN_MASK | DESC_ATTR_DESC_DONE));
+ /* Enable the interrupts, for resuming from stop queue later (if so) */
+ *attr |= cpu_to_le32(DESC_ATTR_INT_EN |
+ (((sizeof(struct tbt_frame_header) + data_len) <<
+ DESC_ATTR_LEN_SHIFT) & DESC_ATTR_LEN_MASK));
+ hdr->frame_size = cpu_to_le32(data_len);
+ hdr->frame_index = cpu_to_le16(frm_idx);
+ hdr->frame_id = cpu_to_le16(port->frame_id);
+
+ /* In case the remaining data_len is smaller than a frame */
+ while (len < data_len) {
+ memcpy(dest, src, len);
+ data_len -= len;
+ dest += len;
+
+ if (unmap) {
+ kunmap_atomic((void *)src);
+ unmap = false;
+ }
+
+ if (frag_num < skb_shinfo(skb)->nr_frags) {
+ const skb_frag_t *frag =
+ &(skb_shinfo(skb)->frags[frag_num]);
+ len = skb_frag_size(frag);
+ src = kmap_atomic(skb_frag_page(frag)) +
+ frag->page_offset;
+ unmap = true;
+ ++frag_num;
+ } else if (unlikely(data_len > 0)) {
+ goto invalid_packet;
+ }
+ }
+ memcpy(dest, src, data_len);
+ if (unmap) {
+ kunmap_atomic((void *)src);
+ unmap = false;
+ }
+
+ ++frm_idx;
+ prod = (prod + 1) & (TBT_NET_NUM_TX_BUFS - 1);
+
+ if (!tbt_net_xmit_csum(skb, tx_ring, first, prod, frm_idx))
+ goto invalid_packet;
+
+ if (port->match_frame_id)
+ ++port->frame_id;
+
+ prod_cons &= ~REG_RING_PROD_MASK;
+ prod_cons |= (prod << REG_RING_PROD_SHIFT) & REG_RING_PROD_MASK;
+ wmb(); /* make sure producer update is done after buffers are ready */
+ iowrite32(prod_cons, reg);
+
+ ++port->stats.tx_packets;
+ port->stats.tx_bytes += skb->len;
+
+ dev_consume_skb_any(skb);
+ return NETDEV_TX_OK;
+
+invalid_packet:
+ netif_err(port, tx_err, net_dev, "port %u invalid transmit packet\n",
+ port->num);
+tx_error:
+ ++port->stats.tx_errors;
+ dev_kfree_skb_any(skb);
+ return NETDEV_TX_OK;
}
+static void tbt_net_set_rx_mode(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+ struct netdev_hw_addr *ha;
+
+ if (net_dev->flags & IFF_PROMISC)
+ port->packet_filters |= BIT(PACKET_TYPE_PROMISCUOUS);
+ else
+ port->packet_filters &= ~BIT(PACKET_TYPE_PROMISCUOUS);
+ if (net_dev->flags & IFF_ALLMULTI)
+ port->packet_filters |= BIT(PACKET_TYPE_ALL_MULTICAST);
+ else
+ port->packet_filters &= ~BIT(PACKET_TYPE_ALL_MULTICAST);
+
+ /* if you have more than a single MAC address */
+ if (netdev_uc_count(net_dev) > 1)
+ port->packet_filters |= BIT(PACKET_TYPE_UNICAST_PROMISCUOUS);
+ /* if have a single MAC address */
+ else if (netdev_uc_count(net_dev) == 1) {
+ netdev_for_each_uc_addr(ha, net_dev)
+ /* checks whether the MAC is what we set */
+ if (ether_addr_equal(ha->addr, net_dev->dev_addr))
+ port->packet_filters &=
+ ~BIT(PACKET_TYPE_UNICAST_PROMISCUOUS);
+ else
+ port->packet_filters |=
+ BIT(PACKET_TYPE_UNICAST_PROMISCUOUS);
+ } else {
+ port->packet_filters &= ~BIT(PACKET_TYPE_UNICAST_PROMISCUOUS);
+ }
+
+ /* Populate the multicast hash table with received MAC addresses */
+ memset(port->multicast_hash_table, 0,
+ sizeof(port->multicast_hash_table));
+ netdev_for_each_mc_addr(ha, net_dev) {
+ u16 hash_val = TBT_NET_ETHER_ADDR_HASH(ha->addr);
+
+ port->multicast_hash_table[hash_val / BITS_PER_U32] |=
+ BIT(hash_val % BITS_PER_U32);
+ }
+
+}
+
+static struct rtnl_link_stats64 *tbt_net_get_stats64(
+ struct net_device *net_dev,
+ struct rtnl_link_stats64 *stats)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ memset(stats, 0, sizeof(*stats));
+ stats->tx_packets = port->stats.tx_packets;
+ stats->tx_bytes = port->stats.tx_bytes;
+ stats->tx_errors = port->stats.tx_errors;
+ stats->rx_packets = port->stats.rx_packets;
+ stats->rx_bytes = port->stats.rx_bytes;
+ stats->rx_length_errors = port->stats.rx_length_errors;
+ stats->rx_over_errors = port->stats.rx_over_errors;
+ stats->rx_crc_errors = port->stats.rx_crc_errors;
+ stats->rx_missed_errors = port->stats.rx_missed_errors;
+ stats->rx_errors = stats->rx_length_errors + stats->rx_over_errors +
+ stats->rx_crc_errors + stats->rx_missed_errors;
+ stats->multicast = port->stats.multicast;
+ return stats;
+}
+
+static int tbt_net_set_mac_address(struct net_device *net_dev, void *addr)
+{
+ struct sockaddr *saddr = addr;
+
+ if (!is_valid_ether_addr(saddr->sa_data))
+ return -EADDRNOTAVAIL;
+
+ memcpy(net_dev->dev_addr, saddr->sa_data, net_dev->addr_len);
+
+ return 0;
+}
+
+static int tbt_net_change_mtu(struct net_device *net_dev, int new_mtu)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ /* MTU < 68 is an error and causes problems on some kernels */
+ if (new_mtu < 68 || new_mtu > (TBT_NET_MTU - ETH_HLEN))
+ return -EINVAL;
+
+ netif_info(port, probe, net_dev, "Thunderbolt(TM) Networking port %u - changing MTU from %u to %d\n",
+ port->num, net_dev->mtu, new_mtu);
+
+ net_dev->mtu = new_mtu;
+
+ return 0;
+}
+
+static const struct net_device_ops tbt_netdev_ops = {
+ /* called when the network is up'ed */
+ .ndo_open = tbt_net_open,
+ /* called when the network is down'ed */
+ .ndo_stop = tbt_net_close,
+ .ndo_start_xmit = tbt_net_xmit_frame,
+ .ndo_set_rx_mode = tbt_net_set_rx_mode,
+ .ndo_get_stats64 = tbt_net_get_stats64,
+ .ndo_set_mac_address = tbt_net_set_mac_address,
+ .ndo_change_mtu = tbt_net_change_mtu,
+ .ndo_validate_addr = eth_validate_addr,
+};
+
+static int tbt_net_get_settings(__maybe_unused struct net_device *net_dev,
+ struct ethtool_cmd *ecmd)
+{
+ ecmd->supported |= SUPPORTED_20000baseKR2_Full;
+ ecmd->advertising |= ADVERTISED_20000baseKR2_Full;
+ ecmd->autoneg = AUTONEG_DISABLE;
+ ecmd->transceiver = XCVR_INTERNAL;
+ ecmd->supported |= SUPPORTED_FIBRE;
+ ecmd->advertising |= ADVERTISED_FIBRE;
+ ecmd->port = PORT_FIBRE;
+ ethtool_cmd_speed_set(ecmd, SPEED_20000);
+ ecmd->duplex = DUPLEX_FULL;
+
+ return 0;
+}
+
+
+static u32 tbt_net_get_msglevel(struct net_device *net_dev)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ return port->msg_enable;
+}
+
+static void tbt_net_set_msglevel(struct net_device *net_dev, u32 data)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ port->msg_enable = data;
+}
+
+static void tbt_net_get_strings(__maybe_unused struct net_device *net_dev,
+ u32 stringset, u8 *data)
+{
+ if (stringset == ETH_SS_STATS)
+ memcpy(data, tbt_net_gstrings_stats,
+ sizeof(tbt_net_gstrings_stats));
+}
+
+static void tbt_net_get_ethtool_stats(struct net_device *net_dev,
+ __maybe_unused struct ethtool_stats *sts,
+ u64 *data)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ memcpy(data, &port->stats, sizeof(port->stats));
+}
+
+static int tbt_net_get_sset_count(__maybe_unused struct net_device *net_dev,
+ int sset)
+{
+ if (sset == ETH_SS_STATS)
+ return sizeof(tbt_net_gstrings_stats) / ETH_GSTRING_LEN;
+ return -EOPNOTSUPP;
+}
+
+static void tbt_net_get_drvinfo(struct net_device *net_dev,
+ struct ethtool_drvinfo *drvinfo)
+{
+ struct tbt_port *port = netdev_priv(net_dev);
+
+ strlcpy(drvinfo->driver, "Thunderbolt(TM) Networking",
+ sizeof(drvinfo->driver));
+ strlcpy(drvinfo->version, DRV_VERSION, sizeof(drvinfo->version));
+
+ strlcpy(drvinfo->bus_info, pci_name(port->nhi_ctxt->pdev),
+ sizeof(drvinfo->bus_info));
+ drvinfo->n_stats = tbt_net_get_sset_count(net_dev, ETH_SS_STATS);
+}
+
+static const struct ethtool_ops tbt_net_ethtool_ops = {
+ .get_settings = tbt_net_get_settings,
+ .get_drvinfo = tbt_net_get_drvinfo,
+ .get_link = ethtool_op_get_link,
+ .get_msglevel = tbt_net_get_msglevel,
+ .set_msglevel = tbt_net_set_msglevel,
+ .get_strings = tbt_net_get_strings,
+ .get_ethtool_stats = tbt_net_get_ethtool_stats,
+ .get_sset_count = tbt_net_get_sset_count,
+};
+
static inline int send_message(struct tbt_port *port, const char *func,
enum pdf_value pdf, u32 msg_len,
const void *msg)
@@ -506,6 +1930,10 @@ void negotiation_events(struct net_device *net_dev,
/* configure TX ring */
reg = iobase + REG_TX_RING_BASE +
(port->local_path * REG_RING_STEP);
+ iowrite32(lower_32_bits(port->tx_ring.dma),
+ reg + REG_RING_PHYS_LO_OFFSET);
+ iowrite32(upper_32_bits(port->tx_ring.dma),
+ reg + REG_RING_PHYS_HI_OFFSET);
tx_ring_conf = (TBT_NET_NUM_TX_BUFS << REG_RING_SIZE_SHIFT) &
REG_RING_SIZE_MASK;
@@ -548,6 +1976,10 @@ void negotiation_events(struct net_device *net_dev,
*/
reg = iobase + REG_RX_RING_BASE +
(port->local_path * REG_RING_STEP);
+ iowrite32(lower_32_bits(port->rx_ring.dma),
+ reg + REG_RING_PHYS_LO_OFFSET);
+ iowrite32(upper_32_bits(port->rx_ring.dma),
+ reg + REG_RING_PHYS_HI_OFFSET);
rx_ring_conf = (TBT_NET_NUM_RX_BUFS << REG_RING_SIZE_SHIFT) &
REG_RING_SIZE_MASK;
@@ -557,6 +1989,17 @@ void negotiation_events(struct net_device *net_dev,
REG_RING_BUF_SIZE_MASK;
iowrite32(rx_ring_conf, reg + REG_RING_SIZE_OFFSET);
+ /* allocate RX buffers and configure the descriptors */
+ if (!tbt_net_alloc_rx_buffers(&port->nhi_ctxt->pdev->dev,
+ &port->rx_ring,
+ TBT_NET_NUM_RX_BUFS,
+ reg + REG_RING_CONS_PROD_OFFSET,
+ GFP_KERNEL)) {
+ netif_err(port, link, net_dev, "Thunderbolt(TM) Networking port %u - no memory for receive buffers\n",
+ port->num);
+ tbt_net_tear_down(net_dev, true);
+ break;
+ }
spin_lock_irqsave(&port->nhi_ctxt->lock, flags);
/* enable RX interrupt */
@@ -569,6 +2012,7 @@ void negotiation_events(struct net_device *net_dev,
netif_info(port, link, net_dev, "Thunderbolt(TM) Networking port %u - ready\n",
port->num);
+ napi_enable(&port->napi);
netif_carrier_on(net_dev);
netif_start_queue(net_dev);
break;
@@ -779,15 +2223,42 @@ struct net_device *nhi_alloc_etherdev(struct tbt_nhi_ctxt *nhi_ctxt,
scnprintf(net_dev->name, sizeof(net_dev->name), "tbtnet%%dp%hhu",
port_num);
+ net_dev->netdev_ops = &tbt_netdev_ops;
+
+ netif_napi_add(net_dev, &port->napi, tbt_net_poll, NAPI_POLL_WEIGHT);
+
+ net_dev->hw_features = NETIF_F_SG |
+ NETIF_F_ALL_TSO |
+ NETIF_F_UFO |
+ NETIF_F_GRO |
+ NETIF_F_IP_CSUM |
+ NETIF_F_IPV6_CSUM;
+ net_dev->features = net_dev->hw_features;
+ if (nhi_ctxt->pci_using_dac)
+ net_dev->features |= NETIF_F_HIGHDMA;
+
INIT_DELAYED_WORK(&port->login_retry_work, login_retry);
INIT_WORK(&port->login_response_work, login_response);
INIT_WORK(&port->logout_work, logout);
INIT_WORK(&port->status_reply_work, status_reply);
INIT_WORK(&port->approve_inter_domain_work, approve_inter_domain);
+ net_dev->ethtool_ops = &tbt_net_ethtool_ops;
+
+ tbt_net_change_mtu(net_dev, TBT_NET_MTU - ETH_HLEN);
+
+ if (register_netdev(net_dev))
+ goto err_register;
+
+ netif_carrier_off(net_dev);
+
netif_info(port, probe, net_dev,
"Thunderbolt(TM) Networking port %u - MAC Address: %pM\n",
port_num, net_dev->dev_addr);
return net_dev;
+
+err_register:
+ free_netdev(net_dev);
+ return NULL;
}
--
2.7.4
^ permalink raw reply related
* [PATCH v7 4/8] thunderbolt: Communication with the ICM (firmware)
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
This patch provides the communication protocol between the
Intel Connection Manager(ICM) firmware that is operational in the
Thunderbolt controller in non-Apple hardware.
The ICM firmware-based controller is used for establishing and maintaining
the Thunderbolt Networking connection - we need to be able to communicate
with it.
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
---
drivers/thunderbolt/Makefile | 1 +
drivers/thunderbolt/icm/Makefile | 2 +
drivers/thunderbolt/icm/icm_nhi.c | 1324 +++++++++++++++++++++++++++++++++++++
drivers/thunderbolt/icm/icm_nhi.h | 92 +++
drivers/thunderbolt/icm/net.h | 227 +++++++
5 files changed, 1646 insertions(+)
create mode 100644 drivers/thunderbolt/icm/Makefile
create mode 100644 drivers/thunderbolt/icm/icm_nhi.c
create mode 100644 drivers/thunderbolt/icm/icm_nhi.h
create mode 100644 drivers/thunderbolt/icm/net.h
diff --git a/drivers/thunderbolt/Makefile b/drivers/thunderbolt/Makefile
index 7a85bd1..b6aa6a3 100644
--- a/drivers/thunderbolt/Makefile
+++ b/drivers/thunderbolt/Makefile
@@ -1,3 +1,4 @@
obj-${CONFIG_THUNDERBOLT_APPLE} := thunderbolt.o
thunderbolt-objs := nhi.o ctl.o tb.o switch.o cap.o path.o tunnel_pci.o eeprom.o
+obj-${CONFIG_THUNDERBOLT_ICM} += icm/
diff --git a/drivers/thunderbolt/icm/Makefile b/drivers/thunderbolt/icm/Makefile
new file mode 100644
index 0000000..f0d0fbb
--- /dev/null
+++ b/drivers/thunderbolt/icm/Makefile
@@ -0,0 +1,2 @@
+obj-${CONFIG_THUNDERBOLT_ICM} += thunderbolt-icm.o
+thunderbolt-icm-objs := icm_nhi.o
diff --git a/drivers/thunderbolt/icm/icm_nhi.c b/drivers/thunderbolt/icm/icm_nhi.c
new file mode 100644
index 0000000..984aa7c
--- /dev/null
+++ b/drivers/thunderbolt/icm/icm_nhi.c
@@ -0,0 +1,1324 @@
+/*******************************************************************************
+ *
+ * Intel Thunderbolt(TM) driver
+ * Copyright(c) 2014 - 2016 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program. If not, see <http://www.gnu.org/licenses/>.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
+ *
+ * Contact Information:
+ * Intel Thunderbolt Mailing List <thunderbolt-software@lists.01.org>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ ******************************************************************************/
+
+#include <linux/printk.h>
+#include <linux/crc32.h>
+#include <linux/delay.h>
+#include <linux/dmi.h>
+#include "icm_nhi.h"
+#include "net.h"
+
+#define NHI_GENL_VERSION 1
+#define NHI_GENL_NAME "thunderbolt"
+
+#define DEVICE_DATA(num_ports, dma_port, nvm_ver_offset, nvm_auth_on_boot,\
+ support_full_e2e) \
+ ((num_ports) | ((dma_port) << 4) | ((nvm_ver_offset) << 10) | \
+ ((nvm_auth_on_boot) << 22) | ((support_full_e2e) << 23))
+#define DEVICE_DATA_NUM_PORTS(device_data) ((device_data) & 0xf)
+#define DEVICE_DATA_DMA_PORT(device_data) (((device_data) >> 4) & 0x3f)
+#define DEVICE_DATA_NVM_VER_OFFSET(device_data) (((device_data) >> 10) & 0xfff)
+#define DEVICE_DATA_NVM_AUTH_ON_BOOT(device_data) (((device_data) >> 22) & 0x1)
+#define DEVICE_DATA_SUPPORT_FULL_E2E(device_data) (((device_data) >> 23) & 0x1)
+
+#define USEC_TO_256_NSECS(usec) DIV_ROUND_UP((usec) * NSEC_PER_USEC, 256)
+
+/* NHI genetlink commands */
+enum {
+ NHI_CMD_UNSPEC,
+ NHI_CMD_SUBSCRIBE,
+ NHI_CMD_UNSUBSCRIBE,
+ NHI_CMD_QUERY_INFORMATION,
+ NHI_CMD_MSG_TO_ICM,
+ NHI_CMD_MSG_FROM_ICM,
+ NHI_CMD_MAILBOX,
+ NHI_CMD_APPROVE_TBT_NETWORKING,
+ NHI_CMD_ICM_IN_SAFE_MODE,
+ __NHI_CMD_MAX,
+};
+#define NHI_CMD_MAX (__NHI_CMD_MAX - 1)
+
+/* NHI genetlink policy */
+static const struct nla_policy nhi_genl_policy[NHI_ATTR_MAX + 1] = {
+ [NHI_ATTR_DRV_VERSION] = { .type = NLA_NUL_STRING, },
+ [NHI_ATTR_NVM_VER_OFFSET] = { .type = NLA_U16, },
+ [NHI_ATTR_NUM_PORTS] = { .type = NLA_U8, },
+ [NHI_ATTR_DMA_PORT] = { .type = NLA_U8, },
+ [NHI_ATTR_SUPPORT_FULL_E2E] = { .type = NLA_FLAG, },
+ [NHI_ATTR_MAILBOX_CMD] = { .type = NLA_U32, },
+ [NHI_ATTR_PDF] = { .type = NLA_U32, },
+ [NHI_ATTR_MSG_TO_ICM] = { .type = NLA_BINARY,
+ .len = TBT_ICM_RING_MAX_FRAME_SIZE },
+ [NHI_ATTR_MSG_FROM_ICM] = { .type = NLA_BINARY,
+ .len = TBT_ICM_RING_MAX_FRAME_SIZE },
+};
+
+/* NHI genetlink family */
+static struct genl_family nhi_genl_family = {
+ .id = GENL_ID_GENERATE,
+ .hdrsize = FIELD_SIZEOF(struct tbt_nhi_ctxt, id),
+ .name = NHI_GENL_NAME,
+ .version = NHI_GENL_VERSION,
+ .maxattr = NHI_ATTR_MAX,
+};
+
+static LIST_HEAD(controllers_list);
+static DEFINE_MUTEX(controllers_list_mutex);
+static atomic_t subscribers = ATOMIC_INIT(0);
+static u32 portid;
+
+static bool nhi_nvm_authenticated(struct tbt_nhi_ctxt *nhi_ctxt)
+{
+ enum icm_operation_mode op_mode;
+ u32 *msg_head, port_id, reg;
+ struct sk_buff *skb;
+ int i;
+
+ if (!nhi_ctxt->nvm_auth_on_boot)
+ return true;
+
+ /*
+ * The check for NVM authentication can take time for iCM,
+ * especially in low power configuration.
+ */
+ for (i = 0; i < 5; i++) {
+ u32 status = ioread32(nhi_ctxt->iobase + REG_FW_STS);
+
+ if (status & REG_FW_STS_NVM_AUTH_DONE)
+ break;
+
+ msleep(30);
+ }
+ /*
+ * The check for authentication is done after checking if iCM
+ * is present so it shouldn't reach the max tries (=5).
+ * Anyway, the check for full functionality below covers the error case.
+ */
+ reg = ioread32(nhi_ctxt->iobase + REG_OUTMAIL_CMD);
+ op_mode = (reg & REG_OUTMAIL_CMD_OP_MODE_MASK) >>
+ REG_OUTMAIL_CMD_OP_MODE_SHIFT;
+ if (op_mode == FULL_FUNCTIONALITY)
+ return true;
+
+ dev_warn(&nhi_ctxt->pdev->dev, "controller id %#x is in operation mode %#x status %#lx\n",
+ nhi_ctxt->id, op_mode,
+ (reg & REG_OUTMAIL_CMD_STS_MASK)>>REG_OUTMAIL_CMD_STS_SHIFT);
+
+ skb = genlmsg_new(NLMSG_ALIGN(nhi_genl_family.hdrsize), GFP_KERNEL);
+ if (!skb) {
+ dev_err(&nhi_ctxt->pdev->dev, "genlmsg_new failed: not enough memory to send controller operational mode\n");
+ return false;
+ }
+
+ /* keeping port_id into a local variable for next use */
+ port_id = portid;
+ msg_head = genlmsg_put(skb, port_id, 0, &nhi_genl_family, 0,
+ NHI_CMD_ICM_IN_SAFE_MODE);
+ if (!msg_head) {
+ nlmsg_free(skb);
+ dev_err(&nhi_ctxt->pdev->dev, "genlmsg_put failed: not enough memory to send controller operational mode\n");
+ return false;
+ }
+
+ *msg_head = nhi_ctxt->id;
+
+ genlmsg_end(skb, msg_head);
+
+ genlmsg_unicast(&init_net, skb, port_id);
+
+ return false;
+}
+
+int nhi_send_message(struct tbt_nhi_ctxt *nhi_ctxt, enum pdf_value pdf,
+ u32 msg_len, const void *msg, bool ignore_icm_resp)
+{
+ u32 prod_cons, prod, cons, attr;
+ struct tbt_icm_ring_shared_memory *shared_mem;
+ void __iomem *reg = TBT_RING_CONS_PROD_REG(nhi_ctxt->iobase,
+ REG_TX_RING_BASE,
+ TBT_ICM_RING_NUM);
+
+ dev_dbg(&nhi_ctxt->pdev->dev,
+ "send msg: controller id %#x pdf %u cmd %hhu msg len %u\n",
+ nhi_ctxt->id, pdf, ((u8 *)msg)[3], msg_len);
+
+ if (nhi_ctxt->d0_exit) {
+ dev_notice(&nhi_ctxt->pdev->dev,
+ "controller id %#x is exiting D0\n",
+ nhi_ctxt->id);
+ return -ENODEV;
+ }
+
+ prod_cons = ioread32(reg);
+ prod = TBT_REG_RING_PROD_EXTRACT(prod_cons);
+ cons = TBT_REG_RING_CONS_EXTRACT(prod_cons);
+ if (prod >= TBT_ICM_RING_NUM_TX_BUFS) {
+ dev_warn(&nhi_ctxt->pdev->dev,
+ "controller id %#x producer %u out of range\n",
+ nhi_ctxt->id, prod);
+ return -ENODEV;
+ }
+ if (TBT_TX_RING_FULL(prod, cons, TBT_ICM_RING_NUM_TX_BUFS)) {
+ dev_err(&nhi_ctxt->pdev->dev,
+ "controller id %#x TX ring full\n",
+ nhi_ctxt->id);
+ return -ENOSPC;
+ }
+
+ attr = (msg_len << DESC_ATTR_LEN_SHIFT) & DESC_ATTR_LEN_MASK;
+ attr |= (pdf << DESC_ATTR_EOF_SHIFT) & DESC_ATTR_EOF_MASK;
+
+ shared_mem = nhi_ctxt->icm_ring_shared_mem;
+ shared_mem->tx_buf_desc[prod].attributes = cpu_to_le32(attr);
+
+ memcpy(shared_mem->tx_buf[prod], msg, msg_len);
+
+ prod_cons &= ~REG_RING_PROD_MASK;
+ prod_cons |= (((prod + 1) % TBT_ICM_RING_NUM_TX_BUFS) <<
+ REG_RING_PROD_SHIFT) & REG_RING_PROD_MASK;
+
+ if (likely(!nhi_ctxt->wait_for_icm_resp))
+ nhi_ctxt->wait_for_icm_resp = true;
+ else
+ dev_dbg(&nhi_ctxt->pdev->dev,
+ "controller id %#x wait_for_icm_resp should have been cleared\n",
+ nhi_ctxt->id);
+
+ nhi_ctxt->ignore_icm_resp = ignore_icm_resp;
+
+ iowrite32(prod_cons, reg);
+
+ return 0;
+}
+
+static int nhi_send_driver_ready_command(struct tbt_nhi_ctxt *nhi_ctxt)
+{
+ struct driver_ready_command {
+ __be32 req_code;
+ __be32 crc;
+ } drv_rdy_cmd = {
+ .req_code = cpu_to_be32(CC_DRV_READY),
+ };
+ u32 crc32;
+
+ crc32 = __crc32c_le(~0, (unsigned char const *)&drv_rdy_cmd,
+ offsetof(struct driver_ready_command, crc));
+
+ drv_rdy_cmd.crc = cpu_to_be32(~crc32);
+
+ return nhi_send_message(nhi_ctxt, PDF_SW_TO_FW_COMMAND,
+ sizeof(drv_rdy_cmd), &drv_rdy_cmd, false);
+}
+
+/**
+ * nhi_search_ctxt - search by id the controllers_list.
+ * Should be called under controllers_list_mutex.
+ *
+ * @id: id of the controller
+ *
+ * Return: driver context if found, NULL otherwise.
+ */
+static struct tbt_nhi_ctxt *nhi_search_ctxt(u32 id)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+
+ list_for_each_entry(nhi_ctxt, &controllers_list, node)
+ if (nhi_ctxt->id == id)
+ return nhi_ctxt;
+
+ return NULL;
+}
+
+static int nhi_genl_subscribe(__always_unused struct sk_buff *u_skb,
+ struct genl_info *info)
+ __acquires(&nhi_ctxt->send_sem)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+
+ /*
+ * To send driver ready command to iCM, need at least one subscriber
+ * that will handle the response.
+ * Currently the assumption is one user mode daemon as subscriber
+ * so one portid global variable (without locking).
+ */
+ if (atomic_inc_return(&subscribers) >= 1) {
+ portid = info->snd_portid;
+ if (mutex_lock_interruptible(&controllers_list_mutex)) {
+ atomic_dec_if_positive(&subscribers);
+ return -ERESTART;
+ }
+ list_for_each_entry(nhi_ctxt, &controllers_list, node) {
+ int res;
+
+ if (nhi_ctxt->d0_exit ||
+ !nhi_nvm_authenticated(nhi_ctxt))
+ continue;
+
+ res = down_timeout(&nhi_ctxt->send_sem,
+ msecs_to_jiffies(10*MSEC_PER_SEC));
+ if (res) {
+ dev_err(&nhi_ctxt->pdev->dev,
+ "%s: controller id %#x timeout on send semaphore\n",
+ __func__, nhi_ctxt->id);
+ continue;
+ }
+
+ if (!mutex_trylock(&nhi_ctxt->d0_exit_send_mutex)) {
+ up(&nhi_ctxt->send_sem);
+ continue;
+ }
+
+ res = nhi_send_driver_ready_command(nhi_ctxt);
+
+ mutex_unlock(&nhi_ctxt->d0_exit_send_mutex);
+ if (res)
+ up(&nhi_ctxt->send_sem);
+ }
+ mutex_unlock(&controllers_list_mutex);
+ }
+
+ return 0;
+}
+
+static int nhi_genl_unsubscribe(__always_unused struct sk_buff *u_skb,
+ __always_unused struct genl_info *info)
+{
+ atomic_dec_if_positive(&subscribers);
+
+ return 0;
+}
+
+static int nhi_genl_query_information(__always_unused struct sk_buff *u_skb,
+ struct genl_info *info)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+ struct sk_buff *skb;
+ bool msg_too_long;
+ int res = -ENODEV;
+ u32 *msg_head;
+
+ if (!info || !info->userhdr)
+ return -EINVAL;
+
+ skb = genlmsg_new(NLMSG_ALIGN(nhi_genl_family.hdrsize) +
+ nla_total_size(sizeof(DRV_VERSION)) +
+ nla_total_size(sizeof(nhi_ctxt->nvm_ver_offset)) +
+ nla_total_size(sizeof(nhi_ctxt->num_ports)) +
+ nla_total_size(sizeof(nhi_ctxt->dma_port)) +
+ nla_total_size(0), /* nhi_ctxt->support_full_e2e */
+ GFP_KERNEL);
+ if (!skb)
+ return -ENOMEM;
+
+ msg_head = genlmsg_put_reply(skb, info, &nhi_genl_family, 0,
+ NHI_CMD_QUERY_INFORMATION);
+ if (!msg_head) {
+ res = -ENOMEM;
+ goto genl_put_reply_failure;
+ }
+
+ if (mutex_lock_interruptible(&controllers_list_mutex)) {
+ res = -ERESTART;
+ goto genl_put_reply_failure;
+ }
+
+ nhi_ctxt = nhi_search_ctxt(*(u32 *)info->userhdr);
+ if (nhi_ctxt && !nhi_ctxt->d0_exit) {
+ *msg_head = nhi_ctxt->id;
+
+ msg_too_long = !!nla_put_string(skb, NHI_ATTR_DRV_VERSION,
+ DRV_VERSION);
+
+ msg_too_long = msg_too_long ||
+ nla_put_u16(skb, NHI_ATTR_NVM_VER_OFFSET,
+ nhi_ctxt->nvm_ver_offset);
+
+ msg_too_long = msg_too_long ||
+ nla_put_u8(skb, NHI_ATTR_NUM_PORTS,
+ nhi_ctxt->num_ports);
+
+ msg_too_long = msg_too_long ||
+ nla_put_u8(skb, NHI_ATTR_DMA_PORT,
+ nhi_ctxt->dma_port);
+
+ if (msg_too_long) {
+ res = -EMSGSIZE;
+ goto release_ctl_list_lock;
+ }
+
+ if (nhi_ctxt->support_full_e2e &&
+ nla_put_flag(skb, NHI_ATTR_SUPPORT_FULL_E2E)) {
+ res = -EMSGSIZE;
+ goto release_ctl_list_lock;
+ }
+ mutex_unlock(&controllers_list_mutex);
+
+ genlmsg_end(skb, msg_head);
+
+ return genlmsg_reply(skb, info);
+ }
+
+release_ctl_list_lock:
+ mutex_unlock(&controllers_list_mutex);
+ genlmsg_cancel(skb, msg_head);
+
+genl_put_reply_failure:
+ nlmsg_free(skb);
+
+ return res;
+}
+
+static int nhi_genl_msg_to_icm(__always_unused struct sk_buff *u_skb,
+ struct genl_info *info)
+ __acquires(&nhi_ctxt->send_sem)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+ int res = -ENODEV;
+ int msg_len;
+ void *msg;
+
+ if (!info || !info->userhdr || !info->attrs ||
+ !info->attrs[NHI_ATTR_PDF] || !info->attrs[NHI_ATTR_MSG_TO_ICM])
+ return -EINVAL;
+
+ msg_len = nla_len(info->attrs[NHI_ATTR_MSG_TO_ICM]);
+ if (msg_len > TBT_ICM_RING_MAX_FRAME_SIZE)
+ return -ENOBUFS;
+
+ msg = nla_data(info->attrs[NHI_ATTR_MSG_TO_ICM]);
+
+ if (mutex_lock_interruptible(&controllers_list_mutex))
+ return -ERESTART;
+
+ nhi_ctxt = nhi_search_ctxt(*(u32 *)info->userhdr);
+ if (nhi_ctxt && !nhi_ctxt->d0_exit) {
+ /*
+ * waiting 10 seconds to receive a FW response
+ * if not, just give up and pop up an error
+ */
+ res = down_timeout(&nhi_ctxt->send_sem,
+ msecs_to_jiffies(10 * MSEC_PER_SEC));
+ if (res) {
+ void __iomem *rx_prod_cons = TBT_RING_CONS_PROD_REG(
+ nhi_ctxt->iobase,
+ REG_RX_RING_BASE,
+ TBT_ICM_RING_NUM);
+ void __iomem *tx_prod_cons = TBT_RING_CONS_PROD_REG(
+ nhi_ctxt->iobase,
+ REG_TX_RING_BASE,
+ TBT_ICM_RING_NUM);
+ dev_err(&nhi_ctxt->pdev->dev,
+ "controller id %#x timeout on send semaphore\n",
+ nhi_ctxt->id);
+ dev_dbg(&nhi_ctxt->pdev->dev,
+ "controller id %#x, tx prod&cons=%#x, rx prod&cons=%#x\n",
+ nhi_ctxt->id,
+ ioread32(tx_prod_cons),
+ ioread32(rx_prod_cons));
+ goto release_ctl_list_lock;
+ }
+
+ if (!mutex_trylock(&nhi_ctxt->d0_exit_send_mutex)) {
+ up(&nhi_ctxt->send_sem);
+ dev_notice(&nhi_ctxt->pdev->dev,
+ "controller id %#x is exiting D0\n",
+ nhi_ctxt->id);
+ goto release_ctl_list_lock;
+ }
+
+ mutex_unlock(&controllers_list_mutex);
+
+ res = nhi_send_message(nhi_ctxt,
+ nla_get_u32(info->attrs[NHI_ATTR_PDF]),
+ msg_len, msg, false);
+
+ mutex_unlock(&nhi_ctxt->d0_exit_send_mutex);
+ if (res)
+ up(&nhi_ctxt->send_sem);
+
+ return res;
+ }
+
+release_ctl_list_lock:
+ mutex_unlock(&controllers_list_mutex);
+ return res;
+}
+
+int nhi_mailbox(struct tbt_nhi_ctxt *nhi_ctxt, u32 cmd, u32 data, bool deinit)
+{
+ u32 delay = deinit ? U32_C(20) : U32_C(100);
+ int i;
+
+ dev_dbg(&nhi_ctxt->pdev->dev, "controller id %#x mbox command %#x\n",
+ nhi_ctxt->id, cmd);
+ iowrite32(data, nhi_ctxt->iobase + REG_INMAIL_DATA);
+ iowrite32(cmd, nhi_ctxt->iobase + REG_INMAIL_CMD);
+
+#define NHI_INMAIL_CMD_RETRIES 50
+ /*
+ * READ_ONCE fetches the value of nhi_ctxt->d0_exit every time
+ * and avoid optimization.
+ * deinit = true to continue the loop even if D3 process has been
+ * carried out.
+ */
+ for (i = 0; (i < NHI_INMAIL_CMD_RETRIES) &&
+ (deinit || !READ_ONCE(nhi_ctxt->d0_exit)); i++) {
+ cmd = ioread32(nhi_ctxt->iobase + REG_INMAIL_CMD);
+
+ if (cmd & REG_INMAIL_CMD_ERROR) {
+ /*
+ * when deinit this just informative as it may
+ * due to unplug of the cable
+ */
+ if (!deinit)
+ dev_err(&nhi_ctxt->pdev->dev,
+ "inmail error after %d msecs\n",
+ i * delay);
+ else
+ dev_info(&nhi_ctxt->pdev->dev,
+ "inmail error after %d msecs\n",
+ i * delay);
+
+ return -EIO;
+ }
+
+ if (!(cmd & REG_INMAIL_CMD_REQUEST))
+ break;
+
+ msleep(delay);
+ }
+
+ if (i == NHI_INMAIL_CMD_RETRIES) {
+ if (!deinit)
+ dev_err(&nhi_ctxt->pdev->dev, "inmail timeout\n");
+ else
+ dev_info(&nhi_ctxt->pdev->dev, "inmail timeout\n");
+ return -ETIMEDOUT;
+ }
+
+ return 0;
+}
+
+static int nhi_mailbox_generic(struct tbt_nhi_ctxt *nhi_ctxt, u32 mb_cmd)
+ __releases(&controllers_list_mutex)
+{
+ int res = -ENODEV;
+
+ if (mutex_lock_interruptible(&nhi_ctxt->mailbox_mutex)) {
+ res = -ERESTART;
+ goto release_ctl_list_lock;
+ }
+
+ if (!mutex_trylock(&nhi_ctxt->d0_exit_mailbox_mutex)) {
+ mutex_unlock(&nhi_ctxt->mailbox_mutex);
+ dev_notice(&nhi_ctxt->pdev->dev,
+ "controller id %#x is exiting D0\n",
+ nhi_ctxt->id);
+ goto release_ctl_list_lock;
+ }
+
+ mutex_unlock(&controllers_list_mutex);
+
+ res = nhi_mailbox(nhi_ctxt, mb_cmd, 0, false);
+ mutex_unlock(&nhi_ctxt->d0_exit_mailbox_mutex);
+ mutex_unlock(&nhi_ctxt->mailbox_mutex);
+
+ return res;
+
+release_ctl_list_lock:
+ mutex_unlock(&controllers_list_mutex);
+ return res;
+}
+
+static int nhi_genl_mailbox(__always_unused struct sk_buff *u_skb,
+ struct genl_info *info)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+ u32 cmd, mb_cmd;
+
+ if (!info || !info->userhdr || !info->attrs ||
+ !info->attrs[NHI_ATTR_MAILBOX_CMD])
+ return -EINVAL;
+
+ cmd = nla_get_u32(info->attrs[NHI_ATTR_MAILBOX_CMD]);
+ mb_cmd = ((cmd << REG_INMAIL_CMD_CMD_SHIFT) &
+ REG_INMAIL_CMD_CMD_MASK) | REG_INMAIL_CMD_REQUEST;
+
+ if (mutex_lock_interruptible(&controllers_list_mutex))
+ return -ERESTART;
+
+ nhi_ctxt = nhi_search_ctxt(*(u32 *)info->userhdr);
+ if (nhi_ctxt && !nhi_ctxt->d0_exit)
+ return nhi_mailbox_generic(nhi_ctxt, mb_cmd);
+
+ mutex_unlock(&controllers_list_mutex);
+ return -ENODEV;
+}
+
+
+static int nhi_genl_send_msg(struct tbt_nhi_ctxt *nhi_ctxt, enum pdf_value pdf,
+ const u8 *msg, u32 msg_len)
+{
+ u32 *msg_head, port_id;
+ struct sk_buff *skb;
+ int res;
+
+ if (atomic_read(&subscribers) < 1) {
+ dev_notice(&nhi_ctxt->pdev->dev, "no subscribers for controller id %#x, dropping message - pdf %u cmd %hhu msg len %u\n",
+ nhi_ctxt->id, pdf, msg[3], msg_len);
+ return -ENOTCONN;
+ }
+
+ skb = genlmsg_new(NLMSG_ALIGN(nhi_genl_family.hdrsize) +
+ nla_total_size(msg_len) +
+ nla_total_size(sizeof(pdf)),
+ GFP_KERNEL);
+ if (!skb)
+ return -ENOMEM;
+
+ port_id = portid;
+ msg_head = genlmsg_put(skb, port_id, 0, &nhi_genl_family, 0,
+ NHI_CMD_MSG_FROM_ICM);
+ if (!msg_head) {
+ res = -ENOMEM;
+ goto genl_put_reply_failure;
+ }
+
+ *msg_head = nhi_ctxt->id;
+
+ if (nla_put_u32(skb, NHI_ATTR_PDF, pdf) ||
+ nla_put(skb, NHI_ATTR_MSG_FROM_ICM, msg_len, msg)) {
+ res = -EMSGSIZE;
+ goto nla_put_failure;
+ }
+
+ genlmsg_end(skb, msg_head);
+
+ return genlmsg_unicast(&init_net, skb, port_id);
+
+nla_put_failure:
+ genlmsg_cancel(skb, msg_head);
+genl_put_reply_failure:
+ nlmsg_free(skb);
+
+ return res;
+}
+
+static bool nhi_msg_from_icm_analysis(struct tbt_nhi_ctxt *nhi_ctxt,
+ enum pdf_value pdf,
+ const u8 *msg, u32 msg_len)
+{
+ /*
+ * preparation for messages that won't be sent,
+ * currently unused in this patch.
+ */
+ bool send_event = true;
+
+ switch (pdf) {
+ case PDF_ERROR_NOTIFICATION:
+ dev_err(&nhi_ctxt->pdev->dev,
+ "controller id %#x PDF_ERROR_NOTIFICATION %hhu msg len %u\n",
+ nhi_ctxt->id, msg[11], msg_len);
+ /* fallthrough */
+ case PDF_WRITE_CONFIGURATION_REGISTERS:
+ /* fallthrough */
+ case PDF_READ_CONFIGURATION_REGISTERS:
+ if (nhi_ctxt->wait_for_icm_resp) {
+ nhi_ctxt->wait_for_icm_resp = false;
+ up(&nhi_ctxt->send_sem);
+ }
+ break;
+
+ case PDF_FW_TO_SW_RESPONSE:
+ if (nhi_ctxt->wait_for_icm_resp) {
+ nhi_ctxt->wait_for_icm_resp = false;
+ up(&nhi_ctxt->send_sem);
+ }
+ break;
+
+ default:
+ dev_warn(&nhi_ctxt->pdev->dev,
+ "controller id %#x pdf %u isn't handled/expected\n",
+ nhi_ctxt->id, pdf);
+ break;
+ }
+
+ return send_event;
+}
+
+static void nhi_msgs_from_icm(struct work_struct *work)
+ __releases(&nhi_ctxt->send_sem)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt = container_of(work, typeof(*nhi_ctxt),
+ icm_msgs_work);
+ void __iomem *reg = TBT_RING_CONS_PROD_REG(nhi_ctxt->iobase,
+ REG_RX_RING_BASE,
+ TBT_ICM_RING_NUM);
+ u32 prod_cons, prod, cons;
+
+ prod_cons = ioread32(reg);
+ prod = TBT_REG_RING_PROD_EXTRACT(prod_cons);
+ cons = TBT_REG_RING_CONS_EXTRACT(prod_cons);
+ if (prod >= TBT_ICM_RING_NUM_RX_BUFS) {
+ dev_warn(&nhi_ctxt->pdev->dev,
+ "controller id %#x producer %u out of range\n",
+ nhi_ctxt->id, prod);
+ return;
+ }
+ if (cons >= TBT_ICM_RING_NUM_RX_BUFS) {
+ dev_warn(&nhi_ctxt->pdev->dev,
+ "controller id %#x consumer %u out of range\n",
+ nhi_ctxt->id, cons);
+ return;
+ }
+
+ while (!TBT_RX_RING_EMPTY(prod, cons, TBT_ICM_RING_NUM_RX_BUFS) &&
+ !nhi_ctxt->d0_exit) {
+ struct tbt_buf_desc *rx_desc;
+ u8 *msg;
+ u32 msg_len;
+ enum pdf_value pdf;
+ bool send_event;
+
+ cons = (cons + 1) % TBT_ICM_RING_NUM_RX_BUFS;
+ rx_desc = &(nhi_ctxt->icm_ring_shared_mem->rx_buf_desc[cons]);
+ if (!(le32_to_cpu(rx_desc->attributes) &
+ DESC_ATTR_DESC_DONE)) {
+ usleep_range(10, 20);
+ if (unlikely(!(le32_to_cpu(rx_desc->attributes) &
+ DESC_ATTR_DESC_DONE)))
+ dev_err(&nhi_ctxt->pdev->dev,
+ "controller id %#x buffer %u might not completely processed\n",
+ nhi_ctxt->id, cons);
+ }
+
+ rmb(); /* read the descriptor and the buffer after DD check */
+ pdf = (le32_to_cpu(rx_desc->attributes) & DESC_ATTR_EOF_MASK)
+ >> DESC_ATTR_EOF_SHIFT;
+ msg = nhi_ctxt->icm_ring_shared_mem->rx_buf[cons];
+ msg_len = (le32_to_cpu(rx_desc->attributes)&DESC_ATTR_LEN_MASK)
+ >> DESC_ATTR_LEN_SHIFT;
+
+ dev_dbg(&nhi_ctxt->pdev->dev,
+ "%s: controller id %#x pdf %u cmd %hhu msg len %u\n",
+ __func__, nhi_ctxt->id, pdf, msg[3], msg_len);
+
+ send_event = nhi_msg_from_icm_analysis(nhi_ctxt, pdf, msg,
+ msg_len);
+
+ if (send_event)
+ nhi_genl_send_msg(nhi_ctxt, pdf, msg, msg_len);
+
+ /* set the descriptor for another receive */
+ rx_desc->attributes = cpu_to_le32(DESC_ATTR_REQ_STS |
+ DESC_ATTR_INT_EN);
+ rx_desc->time = 0;
+ }
+
+ /* free the descriptors for more receive */
+ prod_cons &= ~REG_RING_CONS_MASK;
+ prod_cons |= (cons << REG_RING_CONS_SHIFT) & REG_RING_CONS_MASK;
+ iowrite32(prod_cons, reg);
+
+ if (!nhi_ctxt->d0_exit) {
+ unsigned long flags;
+
+ spin_lock_irqsave(&nhi_ctxt->lock, flags);
+ /* enable RX interrupt */
+ RING_INT_ENABLE_RX(nhi_ctxt->iobase, TBT_ICM_RING_NUM,
+ nhi_ctxt->num_paths);
+
+ spin_unlock_irqrestore(&nhi_ctxt->lock, flags);
+ }
+}
+
+static irqreturn_t nhi_icm_ring_rx_msix(int __always_unused irq, void *data)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt = data;
+
+ spin_lock(&nhi_ctxt->lock);
+ /*
+ * disable RX interrupt
+ * We like to allow interrupt mitigation until the work item
+ * will be completed.
+ */
+ RING_INT_DISABLE_RX(nhi_ctxt->iobase, TBT_ICM_RING_NUM,
+ nhi_ctxt->num_paths);
+
+ spin_unlock(&nhi_ctxt->lock);
+
+ schedule_work(&nhi_ctxt->icm_msgs_work);
+
+ return IRQ_HANDLED;
+}
+
+static irqreturn_t nhi_msi(int __always_unused irq, void *data)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt = data;
+ u32 isr0, isr1, imr0, imr1;
+
+ /* clear on read */
+ isr0 = ioread32(nhi_ctxt->iobase + REG_RING_NOTIFY_BASE);
+ isr1 = ioread32(nhi_ctxt->iobase + REG_RING_NOTIFY_BASE +
+ REG_RING_NOTIFY_STEP);
+ if (unlikely(!isr0 && !isr1))
+ return IRQ_NONE;
+
+ spin_lock(&nhi_ctxt->lock);
+
+ imr0 = ioread32(nhi_ctxt->iobase + REG_RING_INTERRUPT_BASE);
+ imr1 = ioread32(nhi_ctxt->iobase + REG_RING_INTERRUPT_BASE +
+ REG_RING_INTERRUPT_STEP);
+ /* disable the arrived interrupts */
+ iowrite32(imr0 & ~isr0,
+ nhi_ctxt->iobase + REG_RING_INTERRUPT_BASE);
+ iowrite32(imr1 & ~isr1,
+ nhi_ctxt->iobase + REG_RING_INTERRUPT_BASE +
+ REG_RING_INTERRUPT_STEP);
+
+ spin_unlock(&nhi_ctxt->lock);
+
+ if (isr0 & REG_RING_INT_RX_PROCESSED(TBT_ICM_RING_NUM,
+ nhi_ctxt->num_paths))
+ schedule_work(&nhi_ctxt->icm_msgs_work);
+
+ return IRQ_HANDLED;
+}
+
+/**
+ * nhi_set_int_vec - Mapping of the MSIX vector entry to the ring
+ * @nhi_ctxt: contains data on NHI controller
+ * @path: ring to be mapped
+ * @msix_msg_id: msix entry to be mapped
+ */
+static inline void nhi_set_int_vec(struct tbt_nhi_ctxt *nhi_ctxt, u32 path,
+ u8 msix_msg_id)
+{
+ void __iomem *reg;
+ u32 step, shift, ivr;
+
+ if (msix_msg_id % 2)
+ path += nhi_ctxt->num_paths;
+
+ step = path / REG_INT_VEC_ALLOC_PER_REG;
+ shift = (path % REG_INT_VEC_ALLOC_PER_REG) *
+ REG_INT_VEC_ALLOC_FIELD_BITS;
+ reg = nhi_ctxt->iobase + REG_INT_VEC_ALLOC_BASE +
+ (step * REG_INT_VEC_ALLOC_STEP);
+ ivr = ioread32(reg) & ~(REG_INT_VEC_ALLOC_FIELD_MASK << shift);
+ iowrite32(ivr | (msix_msg_id << shift), reg);
+}
+
+/* NHI genetlink operations array */
+static const struct genl_ops nhi_ops[] = {
+ {
+ .cmd = NHI_CMD_SUBSCRIBE,
+ .policy = nhi_genl_policy,
+ .doit = nhi_genl_subscribe,
+ },
+ {
+ .cmd = NHI_CMD_UNSUBSCRIBE,
+ .policy = nhi_genl_policy,
+ .doit = nhi_genl_unsubscribe,
+ },
+ {
+ .cmd = NHI_CMD_QUERY_INFORMATION,
+ .policy = nhi_genl_policy,
+ .doit = nhi_genl_query_information,
+ },
+ {
+ .cmd = NHI_CMD_MSG_TO_ICM,
+ .policy = nhi_genl_policy,
+ .doit = nhi_genl_msg_to_icm,
+ .flags = GENL_ADMIN_PERM,
+ },
+ {
+ .cmd = NHI_CMD_MAILBOX,
+ .policy = nhi_genl_policy,
+ .doit = nhi_genl_mailbox,
+ .flags = GENL_ADMIN_PERM,
+ },
+};
+
+static int nhi_suspend(struct device *dev) __releases(&nhi_ctxt->send_sem)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt = pci_get_drvdata(to_pci_dev(dev));
+ void __iomem *rx_reg, *tx_reg;
+ u32 rx_reg_val, tx_reg_val;
+
+ /* must be after negotiation_events, since messages might be sent */
+ nhi_ctxt->d0_exit = true;
+
+ rx_reg = nhi_ctxt->iobase + REG_RX_OPTIONS_BASE +
+ (TBT_ICM_RING_NUM * REG_OPTS_STEP);
+ rx_reg_val = ioread32(rx_reg) & ~REG_OPTS_E2E_EN;
+ tx_reg = nhi_ctxt->iobase + REG_TX_OPTIONS_BASE +
+ (TBT_ICM_RING_NUM * REG_OPTS_STEP);
+ tx_reg_val = ioread32(tx_reg) & ~REG_OPTS_E2E_EN;
+ /* disable RX flow control */
+ iowrite32(rx_reg_val, rx_reg);
+ /* disable TX flow control */
+ iowrite32(tx_reg_val, tx_reg);
+ /* disable RX ring */
+ iowrite32(rx_reg_val & ~REG_OPTS_VALID, rx_reg);
+
+ mutex_lock(&nhi_ctxt->d0_exit_mailbox_mutex);
+ mutex_lock(&nhi_ctxt->d0_exit_send_mutex);
+
+ cancel_work_sync(&nhi_ctxt->icm_msgs_work);
+
+ if (nhi_ctxt->wait_for_icm_resp) {
+ nhi_ctxt->wait_for_icm_resp = false;
+ nhi_ctxt->ignore_icm_resp = false;
+ /*
+ * if there is response, it is lost, so unlock the send
+ * for the next resume.
+ */
+ up(&nhi_ctxt->send_sem);
+ }
+
+ mutex_unlock(&nhi_ctxt->d0_exit_send_mutex);
+ mutex_unlock(&nhi_ctxt->d0_exit_mailbox_mutex);
+
+ /* wait for all TX to finish */
+ usleep_range(5 * USEC_PER_MSEC, 7 * USEC_PER_MSEC);
+
+ /* disable all interrupts */
+ iowrite32(0, nhi_ctxt->iobase + REG_RING_INTERRUPT_BASE);
+ /* disable TX ring */
+ iowrite32(tx_reg_val & ~REG_OPTS_VALID, tx_reg);
+
+ return 0;
+}
+
+static int nhi_resume(struct device *dev) __acquires(&nhi_ctxt->send_sem)
+{
+ dma_addr_t phys;
+ struct tbt_nhi_ctxt *nhi_ctxt = pci_get_drvdata(to_pci_dev(dev));
+ struct tbt_buf_desc *desc;
+ void __iomem *iobase = nhi_ctxt->iobase;
+ void __iomem *reg;
+ int i;
+
+ if (nhi_ctxt->msix_entries) {
+ iowrite32(ioread32(iobase + REG_DMA_MISC) |
+ REG_DMA_MISC_INT_AUTO_CLEAR,
+ iobase + REG_DMA_MISC);
+ /*
+ * Vector #0, which is TX complete to ICM,
+ * isn't been used currently.
+ */
+ nhi_set_int_vec(nhi_ctxt, 0, 1);
+
+ for (i = 2; i < nhi_ctxt->num_vectors; i++)
+ nhi_set_int_vec(nhi_ctxt, nhi_ctxt->num_paths - (i/2),
+ i);
+ }
+
+ /* configure TX descriptors */
+ for (i = 0, phys = nhi_ctxt->icm_ring_shared_mem_dma_addr;
+ i < TBT_ICM_RING_NUM_TX_BUFS;
+ i++, phys += TBT_ICM_RING_MAX_FRAME_SIZE) {
+ desc = &nhi_ctxt->icm_ring_shared_mem->tx_buf_desc[i];
+ desc->phys = cpu_to_le64(phys);
+ desc->attributes = cpu_to_le32(DESC_ATTR_REQ_STS);
+ }
+ /* configure RX descriptors */
+ for (i = 0;
+ i < TBT_ICM_RING_NUM_RX_BUFS;
+ i++, phys += TBT_ICM_RING_MAX_FRAME_SIZE) {
+ desc = &nhi_ctxt->icm_ring_shared_mem->rx_buf_desc[i];
+ desc->phys = cpu_to_le64(phys);
+ desc->attributes = cpu_to_le32(DESC_ATTR_REQ_STS |
+ DESC_ATTR_INT_EN);
+ }
+
+ /* configure throttling rate for interrupts */
+ for (i = 0, reg = iobase + REG_INT_THROTTLING_RATE;
+ i < NUM_INT_VECTORS;
+ i++, reg += REG_INT_THROTTLING_RATE_STEP) {
+ iowrite32(USEC_TO_256_NSECS(128), reg);
+ }
+
+ /* configure TX for ICM ring */
+ reg = iobase + REG_TX_RING_BASE + (TBT_ICM_RING_NUM * REG_RING_STEP);
+ phys = nhi_ctxt->icm_ring_shared_mem_dma_addr +
+ offsetof(struct tbt_icm_ring_shared_memory, tx_buf_desc);
+ iowrite32(lower_32_bits(phys), reg + REG_RING_PHYS_LO_OFFSET);
+ iowrite32(upper_32_bits(phys), reg + REG_RING_PHYS_HI_OFFSET);
+ iowrite32((TBT_ICM_RING_NUM_TX_BUFS << REG_RING_SIZE_SHIFT) &
+ REG_RING_SIZE_MASK,
+ reg + REG_RING_SIZE_OFFSET);
+
+ reg = iobase + REG_TX_OPTIONS_BASE + (TBT_ICM_RING_NUM*REG_OPTS_STEP);
+ iowrite32(REG_OPTS_RAW | REG_OPTS_VALID, reg);
+
+ /* configure RX for ICM ring */
+ reg = iobase + REG_RX_RING_BASE + (TBT_ICM_RING_NUM * REG_RING_STEP);
+ phys = nhi_ctxt->icm_ring_shared_mem_dma_addr +
+ offsetof(struct tbt_icm_ring_shared_memory, rx_buf_desc);
+ iowrite32(lower_32_bits(phys), reg + REG_RING_PHYS_LO_OFFSET);
+ iowrite32(upper_32_bits(phys), reg + REG_RING_PHYS_HI_OFFSET);
+ iowrite32(((TBT_ICM_RING_NUM_RX_BUFS << REG_RING_SIZE_SHIFT) &
+ REG_RING_SIZE_MASK) |
+ ((TBT_ICM_RING_MAX_FRAME_SIZE << REG_RING_BUF_SIZE_SHIFT) &
+ REG_RING_BUF_SIZE_MASK),
+ reg + REG_RING_SIZE_OFFSET);
+ iowrite32(((TBT_ICM_RING_NUM_RX_BUFS - 1) << REG_RING_CONS_SHIFT) &
+ REG_RING_CONS_MASK,
+ reg + REG_RING_CONS_PROD_OFFSET);
+
+ reg = iobase + REG_RX_OPTIONS_BASE + (TBT_ICM_RING_NUM*REG_OPTS_STEP);
+ iowrite32(REG_OPTS_RAW | REG_OPTS_VALID, reg);
+
+ /* enable RX interrupt */
+ RING_INT_ENABLE_RX(iobase, TBT_ICM_RING_NUM, nhi_ctxt->num_paths);
+
+ if (likely((atomic_read(&subscribers) > 0) &&
+ nhi_nvm_authenticated(nhi_ctxt))) {
+ down(&nhi_ctxt->send_sem);
+ nhi_ctxt->d0_exit = false;
+ mutex_lock(&nhi_ctxt->d0_exit_send_mutex);
+ /*
+ * interrupts are enabled here before send due to
+ * implicit barrier in mutex
+ */
+ nhi_send_driver_ready_command(nhi_ctxt);
+ mutex_unlock(&nhi_ctxt->d0_exit_send_mutex);
+ } else {
+ nhi_ctxt->d0_exit = false;
+ }
+
+ return 0;
+}
+
+static void icm_nhi_shutdown(struct pci_dev *pdev)
+{
+ nhi_suspend(&pdev->dev);
+}
+
+static void icm_nhi_remove(struct pci_dev *pdev)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt = pci_get_drvdata(pdev);
+ int i;
+
+ nhi_suspend(&pdev->dev);
+
+ if (nhi_ctxt->net_workqueue)
+ destroy_workqueue(nhi_ctxt->net_workqueue);
+
+ /*
+ * disable irq for msix or msi
+ */
+ if (likely(nhi_ctxt->msix_entries)) {
+ /* Vector #0 isn't been used currently */
+ devm_free_irq(&pdev->dev, nhi_ctxt->msix_entries[1].vector,
+ nhi_ctxt);
+ pci_disable_msix(pdev);
+ } else {
+ devm_free_irq(&pdev->dev, pdev->irq, nhi_ctxt);
+ pci_disable_msi(pdev);
+ }
+
+ /*
+ * remove controller from the controllers list
+ */
+ mutex_lock(&controllers_list_mutex);
+ list_del(&nhi_ctxt->node);
+ mutex_unlock(&controllers_list_mutex);
+
+ nhi_mailbox(
+ nhi_ctxt,
+ ((CC_DRV_UNLOADS_AND_DISCONNECT_INTER_DOMAIN_PATHS
+ << REG_INMAIL_CMD_CMD_SHIFT) &
+ REG_INMAIL_CMD_CMD_MASK) |
+ REG_INMAIL_CMD_REQUEST,
+ 0, true);
+
+ usleep_range(1 * USEC_PER_MSEC, 5 * USEC_PER_MSEC);
+ iowrite32(1, nhi_ctxt->iobase + REG_HOST_INTERFACE_RST);
+
+ mutex_destroy(&nhi_ctxt->d0_exit_send_mutex);
+ mutex_destroy(&nhi_ctxt->d0_exit_mailbox_mutex);
+ mutex_destroy(&nhi_ctxt->mailbox_mutex);
+ for (i = 0; i < nhi_ctxt->num_ports; i++)
+ mutex_destroy(&(nhi_ctxt->net_devices[i].state_mutex));
+}
+
+static int icm_nhi_probe(struct pci_dev *pdev, const struct pci_device_id *id)
+{
+ struct tbt_nhi_ctxt *nhi_ctxt;
+ void __iomem *iobase;
+ int i, res;
+ bool enable_msi = false;
+
+ res = pcim_enable_device(pdev);
+ if (res) {
+ dev_err(&pdev->dev, "cannot enable PCI device, aborting\n");
+ return res;
+ }
+
+ res = pcim_iomap_regions(pdev, 1 << NHI_MMIO_BAR, pci_name(pdev));
+ if (res) {
+ dev_err(&pdev->dev, "cannot obtain PCI resources, aborting\n");
+ return res;
+ }
+
+ /* cannot fail - table is allocated in pcim_iomap_regions */
+ iobase = pcim_iomap_table(pdev)[NHI_MMIO_BAR];
+
+ /* check if ICM is running */
+ if (!(ioread32(iobase + REG_FW_STS) & REG_FW_STS_ICM_EN)) {
+ dev_err(&pdev->dev, "ICM isn't present, aborting\n");
+ return -ENODEV;
+ }
+
+ nhi_ctxt = devm_kzalloc(&pdev->dev, sizeof(*nhi_ctxt), GFP_KERNEL);
+ if (!nhi_ctxt)
+ return -ENOMEM;
+
+ nhi_ctxt->pdev = pdev;
+ nhi_ctxt->iobase = iobase;
+ nhi_ctxt->id = (PCI_DEVID(pdev->bus->number, pdev->devfn) << 16) |
+ id->device;
+ /*
+ * Number of paths represents the number of rings available for
+ * the controller.
+ */
+ nhi_ctxt->num_paths = ioread32(iobase + REG_HOP_COUNT) &
+ REG_HOP_COUNT_TOTAL_PATHS_MASK;
+
+ nhi_ctxt->nvm_auth_on_boot = DEVICE_DATA_NVM_AUTH_ON_BOOT(
+ id->driver_data);
+ nhi_ctxt->support_full_e2e = DEVICE_DATA_SUPPORT_FULL_E2E(
+ id->driver_data);
+
+ nhi_ctxt->dma_port = DEVICE_DATA_DMA_PORT(id->driver_data);
+ /*
+ * Number of ports in the controller
+ */
+ nhi_ctxt->num_ports = DEVICE_DATA_NUM_PORTS(id->driver_data);
+ nhi_ctxt->nvm_ver_offset = DEVICE_DATA_NVM_VER_OFFSET(id->driver_data);
+
+ mutex_init(&nhi_ctxt->d0_exit_send_mutex);
+ mutex_init(&nhi_ctxt->d0_exit_mailbox_mutex);
+ mutex_init(&nhi_ctxt->mailbox_mutex);
+
+ sema_init(&nhi_ctxt->send_sem, 1);
+
+ INIT_WORK(&nhi_ctxt->icm_msgs_work, nhi_msgs_from_icm);
+
+ spin_lock_init(&nhi_ctxt->lock);
+
+ nhi_ctxt->net_devices = devm_kcalloc(&pdev->dev,
+ nhi_ctxt->num_ports,
+ sizeof(struct port_net_dev),
+ GFP_KERNEL);
+ if (!nhi_ctxt->net_devices)
+ return -ENOMEM;
+
+ for (i = 0; i < nhi_ctxt->num_ports; i++)
+ mutex_init(&(nhi_ctxt->net_devices[i].state_mutex));
+
+ /*
+ * allocating RX and TX vectors for ICM and per port
+ * for thunderbolt networking.
+ * The mapping of the vector is carried out by
+ * nhi_set_int_vec and looks like:
+ * 0=tx icm, 1=rx icm, 2=tx data port 0,
+ * 3=rx data port 0...
+ */
+ nhi_ctxt->num_vectors = (1 + nhi_ctxt->num_ports) * 2;
+ nhi_ctxt->msix_entries = devm_kcalloc(&pdev->dev,
+ nhi_ctxt->num_vectors,
+ sizeof(struct msix_entry),
+ GFP_KERNEL);
+ if (likely(nhi_ctxt->msix_entries)) {
+ for (i = 0; i < nhi_ctxt->num_vectors; i++)
+ nhi_ctxt->msix_entries[i].entry = i;
+ res = pci_enable_msix_exact(pdev,
+ nhi_ctxt->msix_entries,
+ nhi_ctxt->num_vectors);
+
+ if (res ||
+ /*
+ * Allocating ICM RX only.
+ * vector #0, which is TX complete to ICM,
+ * isn't been used currently
+ */
+ devm_request_irq(&pdev->dev,
+ nhi_ctxt->msix_entries[1].vector,
+ nhi_icm_ring_rx_msix, 0, pci_name(pdev),
+ nhi_ctxt)) {
+ devm_kfree(&pdev->dev, nhi_ctxt->msix_entries);
+ nhi_ctxt->msix_entries = NULL;
+ enable_msi = true;
+ }
+ } else {
+ enable_msi = true;
+ }
+ /*
+ * In case allocation didn't succeed, use msi instead of msix
+ */
+ if (enable_msi) {
+ res = pci_enable_msi(pdev);
+ if (res) {
+ dev_err(&pdev->dev, "cannot enable MSI, aborting\n");
+ return res;
+ }
+ res = devm_request_irq(&pdev->dev, pdev->irq, nhi_msi, 0,
+ pci_name(pdev), nhi_ctxt);
+ if (res) {
+ dev_err(&pdev->dev,
+ "request_irq failed %d, aborting\n", res);
+ return res;
+ }
+ }
+ /*
+ * try to work with address space of 64 bits.
+ * In case this doesn't work, work with 32 bits.
+ */
+ if (!dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64))) {
+ nhi_ctxt->pci_using_dac = true;
+ } else {
+ res = dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(32));
+ if (res) {
+ dev_err(&pdev->dev,
+ "No suitable DMA available, aborting\n");
+ return res;
+ }
+ }
+
+ BUILD_BUG_ON(sizeof(struct tbt_buf_desc) != 16);
+ BUILD_BUG_ON(sizeof(struct tbt_icm_ring_shared_memory) > PAGE_SIZE);
+ nhi_ctxt->icm_ring_shared_mem = dmam_alloc_coherent(
+ &pdev->dev, sizeof(*nhi_ctxt->icm_ring_shared_mem),
+ &nhi_ctxt->icm_ring_shared_mem_dma_addr,
+ GFP_KERNEL | __GFP_ZERO);
+ if (nhi_ctxt->icm_ring_shared_mem == NULL) {
+ dev_err(&pdev->dev, "dmam_alloc_coherent failed, aborting\n");
+ return -ENOMEM;
+ }
+
+ nhi_ctxt->net_workqueue = create_singlethread_workqueue("thunderbolt");
+ if (!nhi_ctxt->net_workqueue) {
+ dev_err(&pdev->dev, "create_singlethread_workqueue failed, aborting\n");
+ return -ENOMEM;
+ }
+
+ pci_set_master(pdev);
+ pci_set_drvdata(pdev, nhi_ctxt);
+
+ nhi_resume(&pdev->dev);
+ /*
+ * Add the new controller at the end of the list
+ */
+ mutex_lock(&controllers_list_mutex);
+ list_add_tail(&nhi_ctxt->node, &controllers_list);
+ mutex_unlock(&controllers_list_mutex);
+
+ return res;
+}
+
+/*
+ * The tunneled pci bridges are siblings of us. Use resume_noirq to reenable
+ * the tunnels asap. A corresponding pci quirk blocks the downstream bridges
+ * resume_noirq until we are done.
+ */
+static const struct dev_pm_ops icm_nhi_pm_ops = {
+ SET_SYSTEM_SLEEP_PM_OPS(nhi_suspend, nhi_resume)
+};
+
+static const struct pci_device_id nhi_pci_device_ids[] = {
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_REDWOOD_RIDGE_2C_NHI),
+ DEVICE_DATA(1, 5, 0xa, false, false) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_REDWOOD_RIDGE_4C_NHI),
+ DEVICE_DATA(2, 5, 0xa, false, false) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_FALCON_RIDGE_2C_NHI),
+ DEVICE_DATA(1, 5, 0xa, false, false) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_FALCON_RIDGE_4C_NHI),
+ DEVICE_DATA(2, 5, 0xa, false, false) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_WIN_RIDGE_2C_NHI),
+ DEVICE_DATA(1, 3, 0xa, false, false) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_2C_NHI),
+ DEVICE_DATA(1, 5, 0xa, true, true) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_4C_NHI),
+ DEVICE_DATA(2, 5, 0xa, true, true) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_LP_NHI),
+ DEVICE_DATA(1, 3, 0xa, true, true) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_2C_NHI),
+ DEVICE_DATA(1, 5, 0xa, true, true) },
+ { PCI_VDEVICE(INTEL, PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_4C_NHI),
+ DEVICE_DATA(2, 5, 0xa, true, true) },
+ { 0, }
+};
+
+MODULE_DEVICE_TABLE(pci, nhi_pci_device_ids);
+MODULE_LICENSE("GPL");
+MODULE_VERSION(DRV_VERSION);
+
+static struct pci_driver icm_nhi_driver = {
+ .name = "thunderbolt",
+ .id_table = nhi_pci_device_ids,
+ .probe = icm_nhi_probe,
+ .remove = icm_nhi_remove,
+ .shutdown = icm_nhi_shutdown,
+ .driver.pm = &icm_nhi_pm_ops,
+};
+
+static int __init icm_nhi_init(void)
+{
+ int rc;
+
+ if (dmi_match(DMI_BOARD_VENDOR, "Apple Inc."))
+ return -ENODEV;
+
+ rc = genl_register_family_with_ops(&nhi_genl_family, nhi_ops);
+ if (rc)
+ goto failure;
+
+ rc = pci_register_driver(&icm_nhi_driver);
+ if (rc)
+ goto failure_genl;
+
+ return 0;
+
+failure_genl:
+ genl_unregister_family(&nhi_genl_family);
+
+failure:
+ pr_debug("nhi: error %d occurred in %s\n", rc, __func__);
+ return rc;
+}
+
+static void __exit icm_nhi_unload(void)
+{
+ genl_unregister_family(&nhi_genl_family);
+ pci_unregister_driver(&icm_nhi_driver);
+}
+
+module_init(icm_nhi_init);
+module_exit(icm_nhi_unload);
diff --git a/drivers/thunderbolt/icm/icm_nhi.h b/drivers/thunderbolt/icm/icm_nhi.h
new file mode 100644
index 0000000..30a5657
--- /dev/null
+++ b/drivers/thunderbolt/icm/icm_nhi.h
@@ -0,0 +1,92 @@
+/*******************************************************************************
+ *
+ * Intel Thunderbolt(TM) driver
+ * Copyright(c) 2014 - 2016 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program. If not, see <http://www.gnu.org/licenses/>.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
+ *
+ * Contact Information:
+ * Intel Thunderbolt Mailing List <thunderbolt-software@lists.01.org>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ ******************************************************************************/
+
+#ifndef ICM_NHI_H_
+#define ICM_NHI_H_
+
+#include <linux/pci.h>
+#include "../nhi_regs.h"
+
+#define DRV_VERSION "16.1.54.1"
+
+#define PCI_DEVICE_ID_INTEL_WIN_RIDGE_2C_NHI 0x157d /* Tbt 2 Low Pwr */
+#define PCI_DEVICE_ID_INTEL_WIN_RIDGE_2C_BRIDGE 0x157e
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_LP_NHI 0x15bf /* Tbt 3 Low Pwr */
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_LP_BRIDGE 0x15c0
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_4C_NHI 0x15d2 /* Thunderbolt 3 */
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_4C_BRIDGE 0x15d3
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_2C_NHI 0x15d9
+#define PCI_DEVICE_ID_INTEL_ALPINE_RIDGE_C_2C_BRIDGE 0x15da
+
+#define TBT_ICM_RING_MAX_FRAME_SIZE 256
+#define TBT_ICM_RING_NUM 0
+#define TBT_RING_MAX_FRM_DATA_SZ (TBT_RING_MAX_FRAME_SIZE - \
+ sizeof(struct tbt_frame_header))
+
+enum icm_operation_mode {
+ SAFE_MODE,
+ AUTHENTICATION_MODE_FUNCTIONALITY,
+ ENDPOINT_OPERATION_MODE,
+ FULL_FUNCTIONALITY,
+};
+
+#define TBT_ICM_RING_NUM_TX_BUFS TBT_RING_MIN_NUM_BUFFERS
+#define TBT_ICM_RING_NUM_RX_BUFS ((PAGE_SIZE - (TBT_ICM_RING_NUM_TX_BUFS * \
+ (sizeof(struct tbt_buf_desc) + TBT_ICM_RING_MAX_FRAME_SIZE))) / \
+ (sizeof(struct tbt_buf_desc) + TBT_ICM_RING_MAX_FRAME_SIZE))
+
+/* struct tbt_icm_ring_shared_memory - memory area for DMA */
+struct tbt_icm_ring_shared_memory {
+ u8 tx_buf[TBT_ICM_RING_NUM_TX_BUFS][TBT_ICM_RING_MAX_FRAME_SIZE];
+ u8 rx_buf[TBT_ICM_RING_NUM_RX_BUFS][TBT_ICM_RING_MAX_FRAME_SIZE];
+ struct tbt_buf_desc tx_buf_desc[TBT_ICM_RING_NUM_TX_BUFS];
+ struct tbt_buf_desc rx_buf_desc[TBT_ICM_RING_NUM_RX_BUFS];
+} __aligned(TBT_ICM_RING_MAX_FRAME_SIZE);
+
+/* mailbox data from SW */
+#define REG_INMAIL_DATA 0x39900
+
+/* mailbox command from SW */
+#define REG_INMAIL_CMD 0x39904
+#define REG_INMAIL_CMD_CMD_SHIFT 0
+#define REG_INMAIL_CMD_CMD_MASK GENMASK(7, REG_INMAIL_CMD_CMD_SHIFT)
+#define REG_INMAIL_CMD_ERROR BIT(30)
+#define REG_INMAIL_CMD_REQUEST BIT(31)
+
+/* mailbox command from FW */
+#define REG_OUTMAIL_CMD 0x3990C
+#define REG_OUTMAIL_CMD_STS_SHIFT 0
+#define REG_OUTMAIL_CMD_STS_MASK GENMASK(7, REG_OUTMAIL_CMD_STS_SHIFT)
+#define REG_OUTMAIL_CMD_OP_MODE_SHIFT 8
+#define REG_OUTMAIL_CMD_OP_MODE_MASK \
+ GENMASK(11, REG_OUTMAIL_CMD_OP_MODE_SHIFT)
+#define REG_OUTMAIL_CMD_REQUEST BIT(31)
+
+#define REG_FW_STS 0x39944
+#define REG_FW_STS_ICM_EN GENMASK(1, 0)
+#define REG_FW_STS_NVM_AUTH_DONE BIT(31)
+
+#endif
diff --git a/drivers/thunderbolt/icm/net.h b/drivers/thunderbolt/icm/net.h
new file mode 100644
index 0000000..a1e43f6
--- /dev/null
+++ b/drivers/thunderbolt/icm/net.h
@@ -0,0 +1,227 @@
+/*******************************************************************************
+ *
+ * Intel Thunderbolt(TM) driver
+ * Copyright(c) 2014 - 2016 Intel Corporation.
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along
+ * with this program. If not, see <http://www.gnu.org/licenses/>.
+ *
+ * The full GNU General Public License is included in this distribution in
+ * the file called "COPYING".
+ *
+ * Contact Information:
+ * Intel Thunderbolt Mailing List <thunderbolt-software@lists.01.org>
+ * Intel Corporation, 5200 N.E. Elam Young Parkway, Hillsboro, OR 97124-6497
+ *
+ ******************************************************************************/
+
+#ifndef NET_H_
+#define NET_H_
+
+#include <linux/pci.h>
+#include <linux/netdevice.h>
+#include <linux/mutex.h>
+#include <linux/semaphore.h>
+#include <net/genetlink.h>
+
+/*
+ * Each physical port contains 2 channels.
+ * Devices are exposed to user based on physical ports.
+ */
+#define CHANNELS_PER_PORT_NUM 2
+/*
+ * Calculate host physical port number (Zero-based numbering) from
+ * host channel/link which starts from 1.
+ */
+#define PORT_NUM_FROM_LINK(link) (((link) - 1) / CHANNELS_PER_PORT_NUM)
+
+#define TBT_TX_RING_FULL(prod, cons, size) ((((prod) + 1) % (size)) == (cons))
+#define TBT_TX_RING_EMPTY(prod, cons) ((prod) == (cons))
+#define TBT_RX_RING_FULL(prod, cons) ((prod) == (cons))
+#define TBT_RX_RING_EMPTY(prod, cons, size) ((((cons) + 1) % (size)) == (prod))
+
+#define PATH_FROM_PORT(num_paths, port_num) (((num_paths) - 1) - (port_num))
+
+/* Protocol Defined Field values for SW<->FW communication in raw mode */
+enum pdf_value {
+ PDF_READ_CONFIGURATION_REGISTERS = 1,
+ PDF_WRITE_CONFIGURATION_REGISTERS,
+ PDF_ERROR_NOTIFICATION,
+ PDF_ERROR_ACKNOWLEDGMENT,
+ PDF_PLUG_EVENT_NOTIFICATION,
+ PDF_INTER_DOMAIN_REQUEST,
+ PDF_INTER_DOMAIN_RESPONSE,
+ PDF_CM_OVERRIDE,
+ PDF_RESET_CIO_SWITCH,
+ PDF_FW_TO_SW_NOTIFICATION,
+ PDF_SW_TO_FW_COMMAND,
+ PDF_FW_TO_SW_RESPONSE
+};
+
+/*
+ * SW->FW commands
+ * CC = Command Code
+ */
+enum {
+ CC_GET_THUNDERBOLT_TOPOLOGY = 1,
+ CC_GET_VIDEO_RESOURCES_DATA,
+ CC_DRV_READY,
+ CC_APPROVE_PCI_CONNECTION,
+ CC_CHALLENGE_PCI_CONNECTION,
+ CC_ADD_DEVICE_AND_KEY,
+ CC_APPROVE_INTER_DOMAIN_CONNECTION = 0x10
+};
+
+/*
+ * FW->SW responses
+ * RC = response code
+ */
+enum {
+ RC_GET_TBT_TOPOLOGY = 1,
+ RC_GET_VIDEO_RESOURCES_DATA,
+ RC_DRV_READY,
+ RC_APPROVE_PCI_CONNECTION,
+ RC_CHALLENGE_PCI_CONNECTION,
+ RC_ADD_DEVICE_AND_KEY,
+ RC_INTER_DOMAIN_PKT_SENT = 8,
+ RC_APPROVE_INTER_DOMAIN_CONNECTION = 0x10
+};
+
+/*
+ * FW->SW notifications
+ * NC = notification code
+ */
+enum {
+ NC_DEVICE_CONNECTED = 3,
+ NC_DEVICE_DISCONNECTED,
+ NC_DP_DEVICE_CONNECTED_NOT_TUNNELED,
+ NC_INTER_DOMAIN_CONNECTED,
+ NC_INTER_DOMAIN_DISCONNECTED
+};
+
+/*
+ * SW -> FW mailbox commands
+ * CC = Command Code
+ */
+enum {
+ CC_STOP_CM_ACTIVITY,
+ CC_ENTER_PASS_THROUGH_MODE,
+ CC_ENTER_CM_OWNERSHIP_MODE,
+ CC_DRV_LOADED,
+ CC_DRV_UNLOADED,
+ CC_SAVE_CURRENT_CONNECTED_DEVICES,
+ CC_DISCONNECT_PCIE_PATHS,
+ CC_DRV_UNLOADS_AND_DISCONNECT_INTER_DOMAIN_PATHS,
+ DISCONNECT_PORT_A_INTER_DOMAIN_PATH = 0x10,
+ DISCONNECT_PORT_B_INTER_DOMAIN_PATH,
+ DP_TUNNEL_MODE_IN_ORDER_PER_CAPABILITIES = 0x1E,
+ DP_TUNNEL_MODE_MAXIMIZE_SNK_SRC_TUNNELS,
+ CC_SET_FW_MODE_FD1_D1_CERT = 0x20,
+ CC_SET_FW_MODE_FD1_D1_ALL,
+ CC_SET_FW_MODE_FD1_DA_CERT,
+ CC_SET_FW_MODE_FD1_DA_ALL,
+ CC_SET_FW_MODE_FDA_D1_CERT,
+ CC_SET_FW_MODE_FDA_D1_ALL,
+ CC_SET_FW_MODE_FDA_DA_CERT,
+ CC_SET_FW_MODE_FDA_DA_ALL
+};
+
+
+/* NHI genetlink attributes */
+enum {
+ NHI_ATTR_UNSPEC,
+ NHI_ATTR_DRV_VERSION,
+ NHI_ATTR_NVM_VER_OFFSET,
+ NHI_ATTR_NUM_PORTS,
+ NHI_ATTR_DMA_PORT,
+ NHI_ATTR_SUPPORT_FULL_E2E,
+ NHI_ATTR_MAILBOX_CMD,
+ NHI_ATTR_PDF,
+ NHI_ATTR_MSG_TO_ICM,
+ NHI_ATTR_MSG_FROM_ICM,
+ __NHI_ATTR_MAX,
+};
+#define NHI_ATTR_MAX (__NHI_ATTR_MAX - 1)
+
+struct port_net_dev {
+ struct net_device *net_dev;
+ struct mutex state_mutex;
+};
+
+/**
+ * struct tbt_nhi_ctxt - thunderbolt native host interface context
+ * @node: node in the controllers list.
+ * @pdev: pci device information.
+ * @iobase: address of I/O.
+ * @msix_entries: MSI-X vectors.
+ * @icm_ring_shared_mem: virtual address of iCM ring.
+ * @icm_ring_shared_mem_dma_addr: DMA addr of iCM ring.
+ * @send_sem: semaphore for sending messages to iCM
+ * one at a time.
+ * @mailbox_mutex: mutex for sending mailbox commands to
+ * iCM one at a time.
+ * @d0_exit_send_mutex: synchronizing the d0 exit with messages.
+ * @d0_exit_mailbox_mutex: synchronizing the d0 exit with mailbox.
+ * @lock: synchronizing the interrupt registers
+ * access.
+ * @icm_msgs_work: work queue for handling messages
+ * from iCM.
+ * @net_devices: net devices per port.
+ * @net_workqueue: work queue to send net messages.
+ * @id: id of the controller.
+ * @num_paths: number of paths supported by controller.
+ * @nvm_ver_offset: offset of NVM version in NVM.
+ * @num_vectors: number of MSI-X vectors.
+ * @num_ports: number of ports in the controller.
+ * @dma_port: DMA port.
+ * @d0_exit: whether controller exit D0 state.
+ * @nvm_auth_on_boot: whether iCM authenticates the NVM
+ * during boot.
+ * @wait_for_icm_resp: whether to wait for iCM response.
+ * @ignore_icm_resp: whether to ignore iCM response.
+ * @pci_using_dac: whether using DAC.
+ * @support_full_e2e: whether controller support full E2E.
+ */
+struct tbt_nhi_ctxt {
+ struct list_head node;
+ struct pci_dev *pdev;
+ void __iomem *iobase;
+ struct msix_entry *msix_entries;
+ struct tbt_icm_ring_shared_memory *icm_ring_shared_mem;
+ dma_addr_t icm_ring_shared_mem_dma_addr;
+ struct semaphore send_sem;
+ struct mutex mailbox_mutex;
+ struct mutex d0_exit_send_mutex;
+ struct mutex d0_exit_mailbox_mutex;
+ spinlock_t lock;
+ struct work_struct icm_msgs_work;
+ struct port_net_dev *net_devices;
+ struct workqueue_struct *net_workqueue;
+ u32 id;
+ u32 num_paths;
+ u16 nvm_ver_offset;
+ u8 num_vectors;
+ u8 num_ports;
+ u8 dma_port;
+ bool d0_exit;
+ bool nvm_auth_on_boot : 1;
+ bool wait_for_icm_resp : 1;
+ bool ignore_icm_resp : 1;
+ bool pci_using_dac : 1;
+ bool support_full_e2e : 1;
+};
+
+int nhi_send_message(struct tbt_nhi_ctxt *nhi_ctxt, enum pdf_value pdf,
+ u32 msg_len, const void *msg, bool ignore_icm_resp);
+int nhi_mailbox(struct tbt_nhi_ctxt *nhi_ctxt, u32 cmd, u32 data, bool deinit);
+
+#endif
--
2.7.4
^ permalink raw reply related
* [PATCH v7 3/8] thunderbolt: Kconfig for Thunderbolt Networking
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
Update to the Kconfig Thunderbolt description to add
Thunderbolt networking as an option.
The menu item "Thunderbolt support" now offers:
"Apple Hardware Support" (existing)
and/or
"Thunderbolt Networking" (new)
You can choose the driver for your platform or build both drivers -
each driver will detect if it can run on the specific platform.
If the Thunderbolt Networking option is chosen, Thunderbolt Networking
will be enabled between Linux non-Apple systems, macOS and
Windows based systems.
Thunderbolt Networking will not affect any other Thunderbolt feature that
was previous available to Linux users on either Apple or
non-Apple platforms.
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
---
drivers/thunderbolt/Kconfig | 27 +++++++++++++++++++++++----
drivers/thunderbolt/Makefile | 2 +-
2 files changed, 24 insertions(+), 5 deletions(-)
diff --git a/drivers/thunderbolt/Kconfig b/drivers/thunderbolt/Kconfig
index c121acc..376e5bb 100644
--- a/drivers/thunderbolt/Kconfig
+++ b/drivers/thunderbolt/Kconfig
@@ -1,13 +1,32 @@
-menuconfig THUNDERBOLT
- tristate "Thunderbolt support for Apple devices"
+config THUNDERBOLT
+ tristate "Thunderbolt support"
depends on PCI
select CRC32
help
- Cactus Ridge Thunderbolt Controller driver
+ Thunderbolt Controller driver
+
+if THUNDERBOLT
+
+config THUNDERBOLT_APPLE
+ tristate "Apple hardware support"
+ help
This driver is required if you want to hotplug Thunderbolt devices on
Apple hardware.
Device chaining is currently not supported.
- To compile this driver a module, choose M here. The module will be
+ To compile this driver as a module, choose M here. The module will be
called thunderbolt.
+
+config THUNDERBOLT_ICM
+ tristate "Thunderbolt Networking"
+ help
+ This driver is required if you want Thunderbolt Networking on
+ non-Apple hardware.
+ It creates a virtual Ethernet device that enables computer to
+ computer communication over a Thunderbolt cable.
+
+ To compile this driver as a module, choose M here. The module will be
+ called thunderbolt_icm.
+
+endif
diff --git a/drivers/thunderbolt/Makefile b/drivers/thunderbolt/Makefile
index 5d1053c..7a85bd1 100644
--- a/drivers/thunderbolt/Makefile
+++ b/drivers/thunderbolt/Makefile
@@ -1,3 +1,3 @@
-obj-${CONFIG_THUNDERBOLT} := thunderbolt.o
+obj-${CONFIG_THUNDERBOLT_APPLE} := thunderbolt.o
thunderbolt-objs := nhi.o ctl.o tb.o switch.o cap.o path.o tunnel_pci.o eeprom.o
--
2.7.4
^ permalink raw reply related
* [PATCH v7 2/8] thunderbolt: Updating the register definitions
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
Adding more Thunderbolt(TM) register definitions
and some helper macros.
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
---
drivers/thunderbolt/nhi_regs.h | 109 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 109 insertions(+)
diff --git a/drivers/thunderbolt/nhi_regs.h b/drivers/thunderbolt/nhi_regs.h
index 75cf069..b8e961f 100644
--- a/drivers/thunderbolt/nhi_regs.h
+++ b/drivers/thunderbolt/nhi_regs.h
@@ -9,6 +9,11 @@
#include <linux/types.h>
+#define NHI_MMIO_BAR 0
+
+#define TBT_RING_MIN_NUM_BUFFERS 2
+#define TBT_RING_MAX_FRAME_SIZE (4 * 1024)
+
enum ring_flags {
RING_FLAG_ISOCH_ENABLE = 1 << 27, /* TX only? */
RING_FLAG_E2E_FLOW_CONTROL = 1 << 28,
@@ -39,6 +44,33 @@ struct ring_desc {
u32 time; /* write zero */
} __packed;
+/**
+ * struct tbt_buf_desc - TX/RX ring buffer descriptor.
+ * This is same as struct ring_desc, but without the use of bitfields and
+ * with explicit endianity.
+ */
+struct tbt_buf_desc {
+ __le64 phys;
+ __le32 attributes;
+ __le32 time;
+};
+
+#define DESC_ATTR_LEN_SHIFT 0
+#define DESC_ATTR_LEN_MASK GENMASK(11, DESC_ATTR_LEN_SHIFT)
+#define DESC_ATTR_EOF_SHIFT 12
+#define DESC_ATTR_EOF_MASK GENMASK(15, DESC_ATTR_EOF_SHIFT)
+#define DESC_ATTR_SOF_SHIFT 16
+#define DESC_ATTR_SOF_MASK GENMASK(19, DESC_ATTR_SOF_SHIFT)
+#define DESC_ATTR_TX_ISOCH_DMA_EN BIT(20) /* TX */
+#define DESC_ATTR_RX_CRC_ERR BIT(20) /* RX after use */
+#define DESC_ATTR_DESC_DONE BIT(21)
+#define DESC_ATTR_REQ_STS BIT(22) /* TX and RX before use */
+#define DESC_ATTR_RX_BUF_OVRN_ERR BIT(22) /* RX after use */
+#define DESC_ATTR_INT_EN BIT(23)
+#define DESC_ATTR_OFFSET_SHIFT 24
+#define DESC_ATTR_OFFSET_MASK GENMASK(31, DESC_ATTR_OFFSET_SHIFT)
+
+
/* NHI registers in bar 0 */
/*
@@ -60,6 +92,30 @@ struct ring_desc {
*/
#define REG_RX_RING_BASE 0x08000
+#define REG_RING_STEP 16
+#define REG_RING_PHYS_LO_OFFSET 0
+#define REG_RING_PHYS_HI_OFFSET 4
+#define REG_RING_CONS_PROD_OFFSET 8 /* cons - RO, prod - RW */
+#define REG_RING_CONS_SHIFT 0
+#define REG_RING_CONS_MASK GENMASK(15, REG_RING_CONS_SHIFT)
+#define REG_RING_PROD_SHIFT 16
+#define REG_RING_PROD_MASK GENMASK(31, REG_RING_PROD_SHIFT)
+#define REG_RING_SIZE_OFFSET 12
+#define REG_RING_SIZE_SHIFT 0
+#define REG_RING_SIZE_MASK GENMASK(15, REG_RING_SIZE_SHIFT)
+#define REG_RING_BUF_SIZE_SHIFT 16
+#define REG_RING_BUF_SIZE_MASK GENMASK(27, REG_RING_BUF_SIZE_SHIFT)
+
+#define TBT_RING_CONS_PROD_REG(iobase, ringbase, ringnumber) \
+ ((iobase) + (ringbase) + \
+ ((ringnumber) * REG_RING_STEP) + \
+ REG_RING_CONS_PROD_OFFSET)
+
+#define TBT_REG_RING_PROD_EXTRACT(val) (((val) & REG_RING_PROD_MASK) >> \
+ REG_RING_PROD_SHIFT)
+
+#define TBT_REG_RING_CONS_EXTRACT(val) (((val) & REG_RING_CONS_MASK) >> \
+ REG_RING_CONS_SHIFT)
/*
* 32 bytes per entry, one entry for every hop (REG_HOP_COUNT)
* 00: enum_ring_flags
@@ -77,6 +133,19 @@ struct ring_desc {
* ..: unknown
*/
#define REG_RX_OPTIONS_BASE 0x29800
+#define REG_RX_OPTS_TX_E2E_HOP_ID_SHIFT 12
+#define REG_RX_OPTS_TX_E2E_HOP_ID_MASK \
+ GENMASK(22, REG_RX_OPTS_TX_E2E_HOP_ID_SHIFT)
+#define REG_RX_OPTS_MASK_OFFSET 4
+#define REG_RX_OPTS_MASK_EOF_SHIFT 0
+#define REG_RX_OPTS_MASK_EOF_MASK GENMASK(15, REG_RX_OPTS_MASK_EOF_SHIFT)
+#define REG_RX_OPTS_MASK_SOF_SHIFT 16
+#define REG_RX_OPTS_MASK_SOF_MASK GENMASK(31, REG_RX_OPTS_MASK_SOF_SHIFT)
+
+#define REG_OPTS_STEP 32
+#define REG_OPTS_E2E_EN BIT(28)
+#define REG_OPTS_RAW BIT(30)
+#define REG_OPTS_VALID BIT(31)
/*
* three bitfields: tx, rx, rx overflow
@@ -86,6 +155,7 @@ struct ring_desc {
*/
#define REG_RING_NOTIFY_BASE 0x37800
#define RING_NOTIFY_REG_COUNT(nhi) ((31 + 3 * nhi->hop_count) / 32)
+#define REG_RING_NOTIFY_STEP 4
/*
* two bitfields: rx, tx
@@ -94,8 +164,47 @@ struct ring_desc {
*/
#define REG_RING_INTERRUPT_BASE 0x38200
#define RING_INTERRUPT_REG_COUNT(nhi) ((31 + 2 * nhi->hop_count) / 32)
+#define REG_RING_INT_TX_PROCESSED(ring_num) BIT(ring_num)
+#define REG_RING_INT_RX_PROCESSED(ring_num, num_paths) BIT((ring_num) + \
+ (num_paths))
+#define RING_INT_DISABLE(base, val) iowrite32( \
+ ioread32((base) + REG_RING_INTERRUPT_BASE) & ~(val), \
+ (base) + REG_RING_INTERRUPT_BASE)
+#define RING_INT_ENABLE(base, val) iowrite32( \
+ ioread32((base) + REG_RING_INTERRUPT_BASE) | (val), \
+ (base) + REG_RING_INTERRUPT_BASE)
+#define RING_INT_DISABLE_TX(base, ring_num) \
+ RING_INT_DISABLE(base, REG_RING_INT_TX_PROCESSED(ring_num))
+#define RING_INT_DISABLE_RX(base, ring_num, num_paths) \
+ RING_INT_DISABLE(base, REG_RING_INT_RX_PROCESSED(ring_num, num_paths))
+#define RING_INT_ENABLE_TX(base, ring_num) \
+ RING_INT_ENABLE(base, REG_RING_INT_TX_PROCESSED(ring_num))
+#define RING_INT_ENABLE_RX(base, ring_num, num_paths) \
+ RING_INT_ENABLE(base, REG_RING_INT_RX_PROCESSED(ring_num, num_paths))
+#define RING_INT_DISABLE_TX_RX(base, ring_num, num_paths) \
+ RING_INT_DISABLE(base, REG_RING_INT_TX_PROCESSED(ring_num) | \
+ REG_RING_INT_RX_PROCESSED(ring_num, num_paths))
+
+#define REG_RING_INTERRUPT_STEP 4
+
+#define REG_INT_THROTTLING_RATE 0x38c00
+#define REG_INT_THROTTLING_RATE_STEP 4
+#define NUM_INT_VECTORS 16
+
+#define REG_INT_VEC_ALLOC_BASE 0x38c40
+#define REG_INT_VEC_ALLOC_STEP 4
+#define REG_INT_VEC_ALLOC_FIELD_BITS 4
+#define REG_INT_VEC_ALLOC_FIELD_MASK (BIT(REG_INT_VEC_ALLOC_FIELD_BITS) - 1)
+#define REG_INT_VEC_ALLOC_PER_REG ((BITS_PER_BYTE * sizeof(u32)) / \
+ REG_INT_VEC_ALLOC_FIELD_BITS)
/* The last 11 bits contain the number of hops supported by the NHI port. */
#define REG_HOP_COUNT 0x39640
+#define REG_HOP_COUNT_TOTAL_PATHS_MASK GENMASK(10, 0)
+
+#define REG_HOST_INTERFACE_RST 0x39858
+
+#define REG_DMA_MISC 0x39864
+#define REG_DMA_MISC_INT_AUTO_CLEAR BIT(2)
#endif
--
2.7.4
^ permalink raw reply related
* [PATCH v7 1/8] thunderbolt: Macro rename
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
In-Reply-To: <1474983821-2313-1-git-send-email-amir.jer.levy@intel.com>
This first patch updates the NHI Thunderbolt controller registers file to
reflect that it is not only for Cactus Ridge.
No functional change intended.
Signed-off-by: Amir Levy <amir.jer.levy@intel.com>
Signed-off-by: Andreas Noever <andreas.noever@gmail.com>
---
drivers/thunderbolt/nhi_regs.h | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/thunderbolt/nhi_regs.h b/drivers/thunderbolt/nhi_regs.h
index 86b996c..75cf069 100644
--- a/drivers/thunderbolt/nhi_regs.h
+++ b/drivers/thunderbolt/nhi_regs.h
@@ -1,11 +1,11 @@
/*
- * Thunderbolt Cactus Ridge driver - NHI registers
+ * Thunderbolt driver - NHI registers
*
* Copyright (c) 2014 Andreas Noever <andreas.noever@gmail.com>
*/
-#ifndef DSL3510_REGS_H_
-#define DSL3510_REGS_H_
+#ifndef NHI_REGS_H_
+#define NHI_REGS_H_
#include <linux/types.h>
--
2.7.4
^ permalink raw reply related
* [PATCH v7 0/8] thunderbolt: Introducing Thunderbolt(TM) Networking
From: Amir Levy @ 2016-09-27 13:43 UTC (permalink / raw)
To: gregkh
Cc: andreas.noever, bhelgaas, corbet, linux-kernel, linux-pci, netdev,
linux-doc, mario_limonciello, thunderbolt-linux, mika.westerberg,
tomas.winkler, xiong.y.zhang, Amir Levy
This driver enables Thunderbolt Networking on non-Apple platforms
running Linux.
Thunderbolt Networking provides peer-to-peer connections to transfer
files between computers, perform PC migrations, and/or set up small
workgroups with shared storage.
This is a virtual connection that emulates an Ethernet adapter that
enables Ethernet networking with the benefit of Thunderbolt superfast
medium capability.
Thunderbolt Networking enables two hosts and several devices that
have a Thunderbolt controller to be connected together in a linear
(Daisy chain) series from a single port.
Thunderbolt Networking for Linux is compatible with Thunderbolt
Networking on systems running macOS or Windows and also supports
Thunderbolt generation 2 and 3 controllers.
Note that all pre-existing Thunderbolt generation 3 features, such as
USB, Display and other Thunderbolt device connectivity will continue
to function exactly as they did prior to enabling Thunderbolt Networking.
Code and Software Specifications:
This kernel code creates a virtual ethernet device for computer to
computer communication over a Thunderbolt cable.
The new driver is a separate driver to the existing Thunderbolt driver.
It is designed to work on systems running Linux that
interface with Intel Connection Manager (ICM) firmware based
Thunderbolt controllers that support Thunderbolt Networking.
The kernel code operates in coordination with the Thunderbolt user-
space daemon to implement full Thunderbolt networking functionality.
Hardware Specifications:
Thunderbolt Hardware specs have not yet been published but are used
where necessary for register definitions.
Changes since v6:
- Removed (TM) from Kconfig
- Removed redundant DRV_NAME macro
- Removed copyright notice from Makefile
- Moved networking.txt file to Documentation/Thunderbolt/
- Edited patch descriptions
- Edited documentation in networking.txt
These patches were pushed to GitHub where they can be reviewed more
comfortably with green/red highlighting:
https://github.com/01org/thunderbolt-software-kernel-tree
Daemon code:
https://github.com/01org/thunderbolt-software-daemon
For reference, here's a link to version 6:
[v6]: https://lkml.org/lkml/2016/8/1/233
Amir Levy (8):
thunderbolt: Macro rename
thunderbolt: Updating the register definitions
thunderbolt: Kconfig for Thunderbolt Networking
thunderbolt: Communication with the ICM (firmware)
thunderbolt: Networking state machine
thunderbolt: Networking transmit and receive
thunderbolt: Networking doc
thunderbolt: Adding maintainer entry
Documentation/00-INDEX | 2 +
Documentation/thunderbolt/networking.txt | 132 ++
MAINTAINERS | 8 +-
drivers/thunderbolt/Kconfig | 27 +-
drivers/thunderbolt/Makefile | 3 +-
drivers/thunderbolt/icm/Makefile | 2 +
drivers/thunderbolt/icm/icm_nhi.c | 1622 +++++++++++++++++++++
drivers/thunderbolt/icm/icm_nhi.h | 92 ++
drivers/thunderbolt/icm/net.c | 2264 ++++++++++++++++++++++++++++++
drivers/thunderbolt/icm/net.h | 297 ++++
drivers/thunderbolt/nhi_regs.h | 115 +-
11 files changed, 4555 insertions(+), 9 deletions(-)
create mode 100644 Documentation/thunderbolt/networking.txt
create mode 100644 drivers/thunderbolt/icm/Makefile
create mode 100644 drivers/thunderbolt/icm/icm_nhi.c
create mode 100644 drivers/thunderbolt/icm/icm_nhi.h
create mode 100644 drivers/thunderbolt/icm/net.c
create mode 100644 drivers/thunderbolt/icm/net.h
--
2.7.4
^ permalink raw reply
* Re: [PATCH net-next 0/2] net: ethernet: mediatek: some bug fixes for PDAM and HW LRO
From: David Miller @ 2016-09-27 13:42 UTC (permalink / raw)
To: nelson.chang; +Cc: john, nbd, netdev, linux-mediatek, nelsonch.tw
In-Reply-To: <1474871630-4134-1-git-send-email-nelson.chang@mediatek.com>
From: Nelson Chang <nelson.chang@mediatek.com>
Date: Mon, 26 Sep 2016 14:33:48 +0800
> 1) Add to stop PDMA while stopping the frame engine
> 2) Modify the register settings for LRO relinquishments
> 3) Jump out from the waiting loop while LRO relinquishments are done
Series applied, but like Sergei I think you should have split patch
#2 into two separate patches.
You even list the changes individually here in your header
posting.
^ permalink raw reply
* [PATCH nf-next v2 2/2] nf_set_hooks_head: acommodate different kconfig
From: Aaron Conole @ 2016-09-27 13:38 UTC (permalink / raw)
To: netfilter-devel, netdev; +Cc: Florian Westphal, Pablo Neira Ayuso
In-Reply-To: <1474983515-4115-1-git-send-email-aconole@bytheb.org>
When CONFIG_NETFILTER_INGRESS is unset (or no), we need to handle
the request for registration properly by dropping the hook. This
releases the entry during the set.
Signed-off-by: Aaron Conole <aconole@bytheb.org>
---
net/netfilter/core.c | 15 +++++++++++----
1 file changed, 11 insertions(+), 4 deletions(-)
diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index e58e420..61e8a9d 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -90,10 +90,12 @@ static void nf_set_hooks_head(struct net *net, const struct nf_hook_ops *reg,
{
switch (reg->pf) {
case NFPROTO_NETDEV:
+#ifdef CONFIG_NETFILTER_INGRESS
/* We already checked in nf_register_net_hook() that this is
* used from ingress.
*/
rcu_assign_pointer(reg->dev->nf_hooks_ingress, entry);
+#endif
break;
default:
rcu_assign_pointer(net->nf.hooks[reg->pf][reg->hooknum],
@@ -107,10 +109,15 @@ int nf_register_net_hook(struct net *net, const struct nf_hook_ops *reg)
struct nf_hook_entry *hooks_entry;
struct nf_hook_entry *entry;
- if (reg->pf == NFPROTO_NETDEV &&
- (reg->hooknum != NF_NETDEV_INGRESS ||
- !reg->dev || dev_net(reg->dev) != net))
- return -EINVAL;
+ if (reg->pf == NFPROTO_NETDEV) {
+#ifndef CONFIG_NETFILTER_INGRESS
+ if (reg->hooknum == NF_NETDEV_INGRESS)
+ return -EOPNOTSUPP;
+#endif
+ if (reg->hooknum != NF_NETDEV_INGRESS ||
+ !reg->dev || dev_net(reg->dev) != net)
+ return -EINVAL;
+ }
entry = kmalloc(sizeof(*entry), GFP_KERNEL);
if (!entry)
--
2.7.4
^ permalink raw reply related
* [PATCH nf-next v2 1/2] netfilter: Fix potential null pointer dereference
From: Aaron Conole @ 2016-09-27 13:38 UTC (permalink / raw)
To: netfilter-devel, netdev; +Cc: Florian Westphal, Pablo Neira Ayuso
In-Reply-To: <1474983515-4115-1-git-send-email-aconole@bytheb.org>
It's possible for nf_hook_entry_head to return NULL if two
nf_unregister_net_hook calls happen simultaneously with a single hook
entry in the list. This fix ensures that no null pointer dereference
could occur when such a race happens.
Signed-off-by: Aaron Conole <aconole@bytheb.org>
---
net/netfilter/core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/netfilter/core.c b/net/netfilter/core.c
index 360c63d..e58e420 100644
--- a/net/netfilter/core.c
+++ b/net/netfilter/core.c
@@ -160,7 +160,7 @@ void nf_unregister_net_hook(struct net *net, const struct nf_hook_ops *reg)
mutex_lock(&nf_hook_mutex);
hooks_entry = nf_hook_entry_head(net, reg);
- if (hooks_entry->orig_ops == reg) {
+ if (hooks_entry && hooks_entry->orig_ops == reg) {
nf_set_hooks_head(net, reg,
nf_entry_dereference(hooks_entry->next));
goto unlock;
--
2.7.4
^ permalink raw reply related
* [PATCH nf-next v2 0/2] fixes for recent nf_compact hooks
From: Aaron Conole @ 2016-09-27 13:38 UTC (permalink / raw)
To: netfilter-devel, netdev; +Cc: Florian Westphal, Pablo Neira Ayuso
Two possible error conditions were caught during an extended testing
session, and by a build robot. These patches fix the two issues (a
missing handler when config is changed, and a potential NULL
dereference).
Aaron Conole (2):
netfilter: Fix potential null pointer dereference
nf_set_hooks_head: acommodate different kconfig
net/netfilter/core.c | 6 +++++-
1 file changed, 5 insertions(+), 1 deletion(-)
--
2.5.5
^ permalink raw reply
* Re: [PATCH v2] fs/select: add vmalloc fallback for select(2)
From: Eric Dumazet @ 2016-09-27 13:34 UTC (permalink / raw)
To: Vlastimil Babka
Cc: Andrew Morton, Alexander Viro, linux-fsdevel, linux-kernel,
linux-mm, Michal Hocko, netdev
In-Reply-To: <6e62a278-4ac3-a866-51c6-e32511406aba@suse.cz>
On Tue, 2016-09-27 at 10:13 +0200, Vlastimil Babka wrote:
> I doubt anyone runs that in production, especially if performance is of concern.
>
I doubt anyone serious runs select() on a large fd set in production.
Last time I used it was in last century.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
^ permalink raw reply
* Re: [PATCH] Fix link error in 32bit arch because of 64bit division
From: Eric Dumazet @ 2016-09-27 13:32 UTC (permalink / raw)
To: Vishwanath Pai
Cc: pablo, kaber, kadlec, johunt, netfilter-devel, coreteam, netdev,
pai.vishwain, zlpnobody
In-Reply-To: <20160927074203.GA3968@akamai.com>
On Tue, 2016-09-27 at 03:42 -0400, Vishwanath Pai wrote:
> Fix link error in 32bit arch because of 64bit division
>
> Division of 64bit integers will cause linker error undefined reference
> to `__udivdi3'. Fix this by replacing divisions with div64_64
>
> Signed-off-by: Vishwanath Pai <vpai@akamai.com>
>
> ---
> net/netfilter/xt_hashlimit.c | 15 ++++++++-------
> 1 file changed, 8 insertions(+), 7 deletions(-)
>
> diff --git a/net/netfilter/xt_hashlimit.c b/net/netfilter/xt_hashlimit.c
> index 44a095e..7fc694e 100644
> --- a/net/netfilter/xt_hashlimit.c
> +++ b/net/netfilter/xt_hashlimit.c
> @@ -465,19 +465,20 @@ static u64 user2credits(u64 user, int revision)
> {
> if (revision == 1) {
> /* If multiplying would overflow... */
> - if (user > 0xFFFFFFFF / (HZ*CREDITS_PER_JIFFY_v1))
> + if (user > div64_u64(0xFFFFFFFF, (HZ*CREDITS_PER_JIFFY_v1)))
How can this be needed ? 0xFFFFFFFF is 32bits, compiler knows how to
compute 0xFFFFFFFF / (HZ*CREDITS_PER_JIFFY_v1) itself, without using a
64 bit divide !
Please be selective.
> /* Divide first. */
> - return (user / XT_HASHLIMIT_SCALE) *\
> + return div64_u64(user, XT_HASHLIMIT_SCALE) *\
> HZ * CREDITS_PER_JIFFY_v1;
>
> - return (user * HZ * CREDITS_PER_JIFFY_v1) \
> - / XT_HASHLIMIT_SCALE;
> + return div64_u64((user * HZ * CREDITS_PER_JIFFY_v1),
> + XT_HASHLIMIT_SCALE);
> } else {
> - if (user > 0xFFFFFFFFFFFFFFFF / (HZ*CREDITS_PER_JIFFY))
> - return (user / XT_HASHLIMIT_SCALE_v2) *\
Probably same remark here.
> + if (user > div64_u64(0xFFFFFFFFFFFFFFFF, (HZ*CREDITS_PER_JIFFY)))
> + return div64_u64(user, XT_HASHLIMIT_SCALE_v2) *\
> HZ * CREDITS_PER_JIFFY;
>
> - return (user * HZ * CREDITS_PER_JIFFY) / XT_HASHLIMIT_SCALE_v2;
> + return div64_u64((user * HZ * CREDITS_PER_JIFFY),
> + XT_HASHLIMIT_SCALE_v2);
> }
> }
>
^ permalink raw reply
* Re: [Gigaset307x-common] ISDN-Gigaset: Release memory in gigaset_initcs() after an allocation failure
From: SF Markus Elfring @ 2016-09-27 12:52 UTC (permalink / raw)
To: Karsten Keil
Cc: Tilman Schmidt, Paul Bolle, gigaset307x-common, netdev,
kernel-janitors, LKML, Julia Lawall
In-Reply-To: <baa46a3d-1f73-2f7e-6402-98763b159739@linux-pingi.de>
>> * Is it still correct nowadays that the function "gigaset_initcs" did not
>> call the function "kfree" after a later function call failed?
>
> Yes, if it is handled in another place, Paul already did show you the place.
To which source code place do you refer here?
>> * Do you expect that allocated memory will be automatically reclaimed
>> after it would return a null pointer?
>>
> Of course not
Thanks for this acknowledgement.
Regards,
Markus
^ permalink raw reply
* Re: [Gigaset307x-common] ISDN-Gigaset: Release memory in gigaset_initcs() after an allocation failure
From: isdn @ 2016-09-27 12:16 UTC (permalink / raw)
To: SF Markus Elfring, Tilman Schmidt
Cc: Paul Bolle, gigaset307x-common, netdev, kernel-janitors, LKML,
Julia Lawall
In-Reply-To: <5974a2c5-2739-cd69-6e37-04a645bc08d7@users.sourceforge.net>
Am 27.09.2016 um 13:32 schrieb SF Markus Elfring:
>>> I got the impression that the exception handling was incomplete in the
>>> implementation of the function "gigaset_initcs".
>>
>> That impression is wrong. Careful reading of the code will confirm that.
>
> * Is it still correct nowadays that the function "gigaset_initcs" did not
> call the function "kfree" after a later function call failed?
>
Yes, if it is handled in another place, Paul already did show you the place.
> * Do you expect that allocated memory will be automatically reclaimed
> after it would return a null pointer?
>
Of course not
Best regards
Karsten
^ permalink raw reply
* [PATCH] ipv6 addrconf: enable use of proc_dointvec_minmax in addrconf_sysctl
From: Maciej Żenczykowski @ 2016-09-27 12:16 UTC (permalink / raw)
To: Maciej Żenczykowski, David S . Miller
Cc: netdev, Erik Kline, Lorenzo Colitti, Hannes Frederic Sowa
From: Maciej Żenczykowski <maze@google.com>
Signed-off-by: Maciej Żenczykowski <maze@google.com>
---
net/ipv6/addrconf.c | 10 ++++++++--
1 file changed, 8 insertions(+), 2 deletions(-)
diff --git a/net/ipv6/addrconf.c b/net/ipv6/addrconf.c
index 2f1f5d439788..11fa1a5564d4 100644
--- a/net/ipv6/addrconf.c
+++ b/net/ipv6/addrconf.c
@@ -6044,8 +6044,14 @@ static int __addrconf_sysctl_register(struct net *net, char *dev_name,
for (i = 0; table[i].data; i++) {
table[i].data += (char *)p - (char *)&ipv6_devconf;
- table[i].extra1 = idev; /* embedded; no ref */
- table[i].extra2 = net;
+ /* If one of these is already set, then it is not safe to
+ * overwrite either of them: this makes proc_dointvec_minmax
+ * usable.
+ */
+ if (!table[i].extra1 && !table[i].extra2) {
+ table[i].extra1 = idev; /* embedded; no ref */
+ table[i].extra2 = net;
+ }
}
snprintf(path, sizeof(path), "net/ipv6/conf/%s", dev_name);
--
2.8.0.rc3.226.g39d4020
^ permalink raw reply related
* Re: [1/3] ath10k: use devm_clk_get() instead of clk_get()
From: Kalle Valo @ 2016-09-27 12:14 UTC (permalink / raw)
To: Masahiro Yamada
Cc: linux-wireless-u79uwXL29TY76Z2rM5mHXA, Luis R. Rodriguez,
netdev-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
ath10k-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, Masahiro Yamada,
Johannes Berg
In-Reply-To: <1473165598-28683-2-git-send-email-yamada.masahiro-uWyLwvC0a2jby3iVrkZq2A@public.gmane.org>
Masahiro Yamada <yamada.masahiro-uWyLwvC0a2jby3iVrkZq2A@public.gmane.org> wrote:
> Use the managed variant of clk_get() to simplify the failure path
> and the .remove callback.
>
> Signed-off-by: Masahiro Yamada <yamada.masahiro-uWyLwvC0a2jby3iVrkZq2A@public.gmane.org>
3 patches applied to ath-next branch of ath.git, thanks.
828662753d60 ath10k: use devm_clk_get() instead of clk_get()
c5d8a34675d9 ath10k: use devm_reset_control_get() instead of reset_control_get()
65901a9e7058 ath10k: do not check if reset is NULL
--
https://patchwork.kernel.org/patch/9316579/
Documentation about submitting wireless patches and checking status
from patchwork:
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply
* [PATCH V2 4.9] brcmfmac: use correct skb freeing helper when deleting flowring
From: Rafał Miłecki @ 2016-09-27 12:11 UTC (permalink / raw)
To: Kalle Valo
Cc: Arend van Spriel, Franky Lin, Hante Meuleman,
Pieter-Paul Giesberts, Franky Lin,
linux-wireless-u79uwXL29TY76Z2rM5mHXA,
brcm80211-dev-list.pdl-dY08KVG/lbpWk0Htik3J/w,
netdev-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA, Rafał Miłecki
In-Reply-To: <20160927091428.19223-1-zajec5-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
From: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
Flowrings contain skbs waiting for transmission that were passed to us
by netif. It means we checked every one of them looking for 802.1x
Ethernet type. When deleting flowring we have to use freeing function
that will check for 802.1x type as well.
Freeing skbs without a proper check was leading to counter not being
properly decreased. This was triggering a WARNING every time
brcmf_netdev_wait_pend8021x was called.
Signed-off-by: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
Acked-by: Arend van Spriel <arend-dY08KVG/lbpWk0Htik3J/w@public.gmane.org>
Cc: stable-u79uwXL29TY76Z2rM5mHXA@public.gmane.org # 4.5+
---
V2: Add Cc for stable 4.5+. It doesn't apply cleanly to 4.4 and is not
possible for 4.3- due to missing brcmf_get_ifp.
---
drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c
index b16b367..d0b738d 100644
--- a/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c
+++ b/drivers/net/wireless/broadcom/brcm80211/brcmfmac/flowring.c
@@ -234,13 +234,20 @@ static void brcmf_flowring_block(struct brcmf_flowring *flow, u16 flowid,
void brcmf_flowring_delete(struct brcmf_flowring *flow, u16 flowid)
{
+ struct brcmf_bus *bus_if = dev_get_drvdata(flow->dev);
struct brcmf_flowring_ring *ring;
+ struct brcmf_if *ifp;
u16 hash_idx;
+ u8 ifidx;
struct sk_buff *skb;
ring = flow->rings[flowid];
if (!ring)
return;
+
+ ifidx = brcmf_flowring_ifidx_get(flow, flowid);
+ ifp = brcmf_get_ifp(bus_if->drvr, ifidx);
+
brcmf_flowring_block(flow, flowid, false);
hash_idx = ring->hash_id;
flow->hash[hash_idx].ifidx = BRCMF_FLOWRING_INVALID_IFIDX;
@@ -249,7 +256,7 @@ void brcmf_flowring_delete(struct brcmf_flowring *flow, u16 flowid)
skb = skb_dequeue(&ring->skblist);
while (skb) {
- brcmu_pkt_buf_free_skb(skb);
+ brcmf_txfinalize(ifp, skb, false);
skb = skb_dequeue(&ring->skblist);
}
--
2.9.3
^ permalink raw reply related
* Re: [PATCH 4.9] brcmfmac: use correct skb freeing helper when deleting flowring
From: Kalle Valo @ 2016-09-27 12:08 UTC (permalink / raw)
To: Rafał Miłecki
Cc: Arend Van Spriel, Franky Lin, Hante Meuleman,
Pieter-Paul Giesberts, Franky Lin, linux-wireless@vger.kernel.org,
open list:BROADCOM BRCM80211 IEEE802.11n WIRELESS DRIVER,
Network Development, Linux Kernel Mailing List,
Rafał Miłecki
In-Reply-To: <CACna6rxDb_KMwGTFKu1Mf9azmk_PLO0aLR3foRtkqyywa1DNVQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
Rafał Miłecki <zajec5-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> writes:
>>>>> Kalle: this isn't important enough for 4.8 as it's too late for that.
>>>>>
>>>>> I'd like to get it for 4.9 however, as this fixes bug that could lead
>>>>> to WARNING on every add_key/del_key call. We was struggling with these
>>>>> WARNINGs for some time and this fixes one of two problems causing them.
>>>
>>> Ok, I'll queue this for 4.9.
>>>
>>>> Please mark it for stable as well.
>>>
>>> I can add that. Any ideas how old releases stable releases should this
>>> go to?
>>
>> I was analyzing this.
>> 1) This patch uses brcmf_get_ifp which is available in 4.4+ only.
>> 2) It applies cleanly to 4.5+ only due to 32f90caa7debd ("brcmfmac:
>> Increase nr of supported flowrings.")
>> 3) 4.4 would also require applying to the patch without broadcom/ subdir
>>
>> That said I suggest 4.5+. Any objections?
>
> Let me see if patchwork with pick Cc tag as it does for others.
>
> Cc: stable-u79uwXL29TY76Z2rM5mHXA@public.gmane.org # 4.5+
An excellent idea but no luck:
Signed-off-by: Rafa? Mi?ecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
Acked-by: Arend van Spriel <arend-dY08KVG/lbpWk0Htik3J/w@public.gmane.org>
I'll add this to my patchwork wishlist though, I think it would be a
really useful feature to have.
(The question marks are because of my buggy copy paste, ignore those)
--
Kalle Valo
^ permalink raw reply
* Re: [Gigaset307x-common] ISDN-Gigaset: Release memory in gigaset_initcs() after an allocation failure
From: Tilman Schmidt @ 2016-09-27 12:08 UTC (permalink / raw)
To: SF Markus Elfring
Cc: Paul Bolle, Karsten Keil, gigaset307x-common, netdev,
kernel-janitors, LKML, Julia Lawall
In-Reply-To: <5974a2c5-2739-cd69-6e37-04a645bc08d7@users.sourceforge.net>
On Tue, Sep 27, 2016, at 13:32, SF Markus Elfring wrote:
> >> I got the impression that the exception handling was incomplete in the
> >> implementation of the function "gigaset_initcs".
> >
> > That impression is wrong. Careful reading of the code will confirm that.
>
> * Is it still correct nowadays that the function "gigaset_initcs" did not
> call the function "kfree" after a later function call failed?
Wrong premise. That statement was never correct.
> * Do you expect that allocated memory will be automatically reclaimed
> after it would return a null pointer?
No. Should I? Do you?
Regards,
Tilman
--
Tilman Schmidt
tilman@imap.cc
^ permalink raw reply
* Re: [PATCH 4.9] brcmfmac: use correct skb freeing helper when deleting flowring
From: Rafał Miłecki @ 2016-09-27 12:07 UTC (permalink / raw)
To: Arend Van Spriel
Cc: Kalle Valo, Franky Lin, Hante Meuleman, Pieter-Paul Giesberts,
Franky Lin,
linux-wireless-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
open list:BROADCOM BRCM80211 IEEE802.11n WIRELESS DRIVER,
Network Development, Linux Kernel Mailing List,
Rafał Miłecki
In-Reply-To: <41e5f45f-1956-8c9d-49d4-cd73e2c5c7c0-dY08KVG/lbpWk0Htik3J/w@public.gmane.org>
On 27 September 2016 at 14:04, Arend Van Spriel
<arend.vanspriel-dY08KVG/lbpWk0Htik3J/w@public.gmane.org> wrote:
> On 27-9-2016 13:58, Rafał Miłecki wrote:
>> On 27 September 2016 at 13:44, Rafał Miłecki <zajec5-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>>> On 27 September 2016 at 13:27, Kalle Valo <kvalo-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org> wrote:
>>>> Arend Van Spriel <arend.vanspriel-dY08KVG/lbpWk0Htik3J/w@public.gmane.org> writes:
>>>>
>>>>> On 27-9-2016 11:14, Rafał Miłecki wrote:
>>>>>> From: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
>>>>>>
>>>>>> Flowrings contain skbs waiting for transmission that were passed to us
>>>>>> by netif. It means we checked every one of them looking for 802.1x
>>>>>> Ethernet type. When deleting flowring we have to use freeing function
>>>>>> that will check for 802.1x type as well.
>>>>>>
>>>>>> Freeing skbs without a proper check was leading to counter not being
>>>>>> properly decreased. This was triggering a WARNING every time
>>>>>> brcmf_netdev_wait_pend8021x was called.
>>>>>
>>>>> Acked-by: Arend van Spriel <arend-dY08KVG/lbpWk0Htik3J/w@public.gmane.org>
>>>>>> Signed-off-by: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
>>>>>> ---
>>>>>> Kalle: this isn't important enough for 4.8 as it's too late for that.
>>>>>>
>>>>>> I'd like to get it for 4.9 however, as this fixes bug that could lead
>>>>>> to WARNING on every add_key/del_key call. We was struggling with these
>>>>>> WARNINGs for some time and this fixes one of two problems causing them.
>>>>
>>>> Ok, I'll queue this for 4.9.
>>>>
>>>>> Please mark it for stable as well.
>>>>
>>>> I can add that. Any ideas how old releases stable releases should this
>>>> go to?
>>>
>>> I was analyzing this.
>>> 1) This patch uses brcmf_get_ifp which is available in 4.4+ only.
>>> 2) It applies cleanly to 4.5+ only due to 32f90caa7debd ("brcmfmac:
>>> Increase nr of supported flowrings.")
>>> 3) 4.4 would also require applying to the patch without broadcom/ subdir
>>>
>>> That said I suggest 4.5+. Any objections?
>
> No objections. Just a tip. I tend to look at kernel.org main page to see
> the stable and long-term kernel listed. So 4.7+ and 4.5+ have same
> meaning as 4.5 and 4.6 are not stable/long-term kernels.
Some projects may work on their own stable kernels, e.g. Ubuntu, see:
https://wiki.ubuntu.com/Kernel/Dev/ExtendedStable
That's why I don't always look strictly at upstream stable releases only.
--
Rafał
^ permalink raw reply
* Re: [PATCH 4.9] brcmfmac: use correct skb freeing helper when deleting flowring
From: Arend Van Spriel @ 2016-09-27 12:04 UTC (permalink / raw)
To: Rafał Miłecki, Kalle Valo
Cc: Franky Lin, Hante Meuleman, Pieter-Paul Giesberts, Franky Lin,
linux-wireless-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
open list:BROADCOM BRCM80211 IEEE802.11n WIRELESS DRIVER,
Network Development, Linux Kernel Mailing List,
Rafał Miłecki
In-Reply-To: <CACna6rxDb_KMwGTFKu1Mf9azmk_PLO0aLR3foRtkqyywa1DNVQ-JsoAwUIsXosN+BqQ9rBEUg@public.gmane.org>
On 27-9-2016 13:58, Rafał Miłecki wrote:
> On 27 September 2016 at 13:44, Rafał Miłecki <zajec5-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org> wrote:
>> On 27 September 2016 at 13:27, Kalle Valo <kvalo-sgV2jX0FEOL9JmXXK+q4OQ@public.gmane.org> wrote:
>>> Arend Van Spriel <arend.vanspriel-dY08KVG/lbpWk0Htik3J/w@public.gmane.org> writes:
>>>
>>>> On 27-9-2016 11:14, Rafał Miłecki wrote:
>>>>> From: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
>>>>>
>>>>> Flowrings contain skbs waiting for transmission that were passed to us
>>>>> by netif. It means we checked every one of them looking for 802.1x
>>>>> Ethernet type. When deleting flowring we have to use freeing function
>>>>> that will check for 802.1x type as well.
>>>>>
>>>>> Freeing skbs without a proper check was leading to counter not being
>>>>> properly decreased. This was triggering a WARNING every time
>>>>> brcmf_netdev_wait_pend8021x was called.
>>>>
>>>> Acked-by: Arend van Spriel <arend-dY08KVG/lbpWk0Htik3J/w@public.gmane.org>
>>>>> Signed-off-by: Rafał Miłecki <rafal-g1n6cQUeyibVItvQsEIGlw@public.gmane.org>
>>>>> ---
>>>>> Kalle: this isn't important enough for 4.8 as it's too late for that.
>>>>>
>>>>> I'd like to get it for 4.9 however, as this fixes bug that could lead
>>>>> to WARNING on every add_key/del_key call. We was struggling with these
>>>>> WARNINGs for some time and this fixes one of two problems causing them.
>>>
>>> Ok, I'll queue this for 4.9.
>>>
>>>> Please mark it for stable as well.
>>>
>>> I can add that. Any ideas how old releases stable releases should this
>>> go to?
>>
>> I was analyzing this.
>> 1) This patch uses brcmf_get_ifp which is available in 4.4+ only.
>> 2) It applies cleanly to 4.5+ only due to 32f90caa7debd ("brcmfmac:
>> Increase nr of supported flowrings.")
>> 3) 4.4 would also require applying to the patch without broadcom/ subdir
>>
>> That said I suggest 4.5+. Any objections?
No objections. Just a tip. I tend to look at kernel.org main page to see
the stable and long-term kernel listed. So 4.7+ and 4.5+ have same
meaning as 4.5 and 4.6 are not stable/long-term kernels.
Regards,
Arend
> Let me see if patchwork with pick Cc tag as it does for others.
>
> Cc: stable-u79uwXL29TY76Z2rM5mHXA@public.gmane.org # 4.5+
>
> This may be worth backporting to 4.4 as well (as it's longterm), but
> I'll do it separately due to patch not applying cleanly.
^ permalink raw reply
* Re: ath10k: Spelling and miscellaneous neatening
From: Kalle Valo @ 2016-09-27 12:02 UTC (permalink / raw)
To: Joe Perches; +Cc: netdev, linux-wireless, linux-kernel, ath10k
In-Reply-To: <95a6b65277914d88178ed2b15e182455067cca60.1472490319.git.joe@perches.com>
Joe Perches <joe@perches.com> wrote:
> Correct some trivial comment typos.
> Remove unnecessary parentheses in a long line.
> Convert a return; before the end of a void function definition to just ;
>
> Signed-off-by: Joe Perches <joe@perches.com>
> Reviewed-by: Julian Calaby <julian.calaby@gmail.com>
Patch applied to ath-next branch of ath.git, thanks.
e13dbead976d ath10k: spelling and miscellaneous neatening
--
https://patchwork.kernel.org/patch/9304171/
Documentation about submitting wireless patches and checking status
from patchwork:
https://wireless.wiki.kernel.org/en/developers/documentation/submittingpatches
^ permalink raw reply
* Re: [PATCH] VSOCK: Don't dec ack backlog twice for rejected connections
From: David Miller @ 2016-09-27 11:59 UTC (permalink / raw)
To: jhansen; +Cc: pv-drivers, netdev, linux-kernel, virtualization, stefanha,
gregkh
In-Reply-To: <1474959593-15311-1-git-send-email-jhansen@vmware.com>
From: Jorgen Hansen <jhansen@vmware.com>
Date: Mon, 26 Sep 2016 23:59:53 -0700
> If a pending socket is marked as rejected, we will decrease the
> sk_ack_backlog twice. So don't decrement it for rejected sockets
> in vsock_pending_work().
>
> Testing of the rejected socket path was done through code
> modifications.
>
> Reported-by: Stefan Hajnoczi <stefanha@redhat.com>
> Signed-off-by: Jorgen Hansen <jhansen@vmware.com>
> Reviewed-by: Adit Ranadive <aditr@vmware.com>
> Reviewed-by: Aditya Sarwade <asarwade@vmware.com>
Applied, thanks.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox