diff --git a/drivers/s390/net/Kconfig b/drivers/s390/net/Kconfig index bf236d474538722ad19f7335e70d9eba7dde0d5e..4508c49610e245ebdfb18b4657434edcd2df9096 100644 --- a/drivers/s390/net/Kconfig +++ b/drivers/s390/net/Kconfig @@ -111,11 +111,12 @@ config CCWGROUP config ISM tristate "Support for ISM vPCI Adapter" - depends on PCI && SMC + depends on PCI + imply SMC default n help Select this option if you want to use the Internal Shared Memory - vPCI Adapter. + vPCI Adapter. The adapter can be used with the SMC network protocol. To compile as a module choose M. The module name is ism. If unsure, choose N. diff --git a/drivers/s390/net/ism.h b/drivers/s390/net/ism.h index 38fe90c2597d1862565711c08f844bbffe874b24..047fa6101555d0721cf87c2ab0cc0c0299e862eb 100644 --- a/drivers/s390/net/ism.h +++ b/drivers/s390/net/ism.h @@ -5,6 +5,7 @@ #include #include #include +#include #include #include @@ -15,8 +16,6 @@ */ #define ISM_DMB_WORD_OFFSET 1 #define ISM_DMB_BIT_OFFSET (ISM_DMB_WORD_OFFSET * 32) -#define ISM_NR_DMBS 1920 -#define ISM_IDENT_MASK 0x00FFFF #define ISM_REG_SBA 0x1 #define ISM_REG_IEQ 0x2 @@ -177,7 +176,7 @@ struct ism_eq_header { struct ism_eq { struct ism_eq_header header; - struct smcd_event entry[15]; + struct ism_event entry[15]; }; struct ism_sba { @@ -189,30 +188,9 @@ struct ism_sba { u16 dmbe_mask[ISM_NR_DMBS]; }; -struct ism_dev { - spinlock_t lock; - struct pci_dev *pdev; - struct smcd_dev *smcd; - - struct ism_sba *sba; - dma_addr_t sba_dma_addr; - DECLARE_BITMAP(sba_bitmap, ISM_NR_DMBS); - - struct ism_eq *ieq; - dma_addr_t ieq_dma_addr; - - int ieq_idx; -}; - #define ISM_CREATE_REQ(dmb, idx, sf, offset) \ ((dmb) | (idx) << 24 | (sf) << 23 | (offset)) -struct ism_systemeid { - u8 seid_string[24]; - u8 serial_number[4]; - u8 type[4]; -}; - static inline void __ism_read_cmd(struct ism_dev *ism, void *data, unsigned long offset, unsigned long len) { diff --git a/drivers/s390/net/ism_drv.c b/drivers/s390/net/ism_drv.c index 26cc943d2034046e2ef8e956a53dbe46c0228b8d..2b48af04084ef621f65d93d8489cece38139447b 100644 --- a/drivers/s390/net/ism_drv.c +++ b/drivers/s390/net/ism_drv.c @@ -15,9 +15,6 @@ #include #include #include -#include - -#include #include "ism.h" @@ -35,6 +32,101 @@ MODULE_DEVICE_TABLE(pci, ism_device_table); static debug_info_t *ism_debug_info; +#define NO_CLIENT 0xff /* must be >= MAX_CLIENTS */ +static struct ism_client *clients[MAX_CLIENTS]; /* use an array rather than */ + /* a list for fast mapping */ +static u8 max_client; +static DEFINE_MUTEX(clients_lock); +static bool ism_v2_capable; +struct ism_dev_list { + struct list_head list; + struct mutex mutex; /* protects ism device list */ +}; + +static struct ism_dev_list ism_dev_list = { + .list = LIST_HEAD_INIT(ism_dev_list.list), + .mutex = __MUTEX_INITIALIZER(ism_dev_list.mutex), +}; + +static void ism_setup_forwarding(struct ism_client *client, struct ism_dev *ism) +{ + unsigned long flags; + + spin_lock_irqsave(&ism->lock, flags); + ism->subs[client->id] = client; + spin_unlock_irqrestore(&ism->lock, flags); +} + +int ism_register_client(struct ism_client *client) +{ + struct ism_dev *ism; + int i, rc = -ENOSPC; + + mutex_lock(&ism_dev_list.mutex); + mutex_lock(&clients_lock); + for (i = 0; i < MAX_CLIENTS; ++i) { + if (!clients[i]) { + clients[i] = client; + client->id = i; + if (i == max_client) + max_client++; + rc = 0; + break; + } + } + mutex_unlock(&clients_lock); + + if (i < MAX_CLIENTS) { + /* initialize with all devices that we got so far */ + list_for_each_entry(ism, &ism_dev_list.list, list) { + ism->priv[i] = NULL; + client->add(ism); + ism_setup_forwarding(client, ism); + } + } + mutex_unlock(&ism_dev_list.mutex); + + return rc; +} +EXPORT_SYMBOL_GPL(ism_register_client); + +int ism_unregister_client(struct ism_client *client) +{ + struct ism_dev *ism; + unsigned long flags; + int rc = 0; + + mutex_lock(&ism_dev_list.mutex); + list_for_each_entry(ism, &ism_dev_list.list, list) { + spin_lock_irqsave(&ism->lock, flags); + /* Stop forwarding IRQs and events */ + ism->subs[client->id] = NULL; + for (int i = 0; i < ISM_NR_DMBS; ++i) { + if (ism->sba_client_arr[i] == client->id) { + WARN(1, "%s: attempt to unregister '%s' with registered dmb(s)\n", + __func__, client->name); + rc = -EBUSY; + goto err_reg_dmb; + } + } + spin_unlock_irqrestore(&ism->lock, flags); + } + mutex_unlock(&ism_dev_list.mutex); + + mutex_lock(&clients_lock); + clients[client->id] = NULL; + if (client->id + 1 == max_client) + max_client--; + mutex_unlock(&clients_lock); + return rc; + +err_reg_dmb: + spin_unlock_irqrestore(&ism->lock, flags); + mutex_unlock(&ism_dev_list.mutex); + return rc; +} +EXPORT_SYMBOL_GPL(ism_unregister_client); + static int ism_cmd(struct ism_dev *ism, void *cmd) { struct ism_req_hdr *req = cmd; @@ -193,36 +285,19 @@ static int ism_read_local_gid(struct ism_dev *ism) if (ret) goto out; - ism->smcd->local_gid = cmd.response.gid; + ism->local_gid = cmd.response.gid; out: return ret; } -static int ism_query_rgid(struct smcd_dev *smcd, u64 rgid, u32 vid_valid, - u32 vid) -{ - struct ism_dev *ism = smcd->priv; - union ism_query_rgid cmd; - - memset(&cmd, 0, sizeof(cmd)); - cmd.request.hdr.cmd = ISM_QUERY_RGID; - cmd.request.hdr.len = sizeof(cmd.request); - - cmd.request.rgid = rgid; - cmd.request.vlan_valid = vid_valid; - cmd.request.vlan_id = vid; - - return ism_cmd(ism, &cmd); -} - -static void ism_free_dmb(struct ism_dev *ism, struct smcd_dmb *dmb) +static void ism_free_dmb(struct ism_dev *ism, struct ism_dmb *dmb) { clear_bit(dmb->sba_idx, ism->sba_bitmap); dma_free_coherent(&ism->pdev->dev, dmb->dmb_len, dmb->cpu_addr, dmb->dma_addr); } -static int ism_alloc_dmb(struct ism_dev *ism, struct smcd_dmb *dmb) +static int ism_alloc_dmb(struct ism_dev *ism, struct ism_dmb *dmb) { unsigned long bit; @@ -250,10 +325,11 @@ static int ism_alloc_dmb(struct ism_dev *ism, struct smcd_dmb *dmb) return dmb->cpu_addr ? 0 : -ENOMEM; } -static int ism_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) +int ism_register_dmb(struct ism_dev *ism, struct ism_dmb *dmb, + struct ism_client *client) { - struct ism_dev *ism = smcd->priv; union ism_reg_dmb cmd; + unsigned long flags; int ret; ret = ism_alloc_dmb(ism, dmb); @@ -277,14 +353,18 @@ static int ism_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) goto out; } dmb->dmb_tok = cmd.response.dmb_tok; + spin_lock_irqsave(&ism->lock, flags); + ism->sba_client_arr[dmb->sba_idx - ISM_DMB_BIT_OFFSET] = client->id; + spin_unlock_irqrestore(&ism->lock, flags); out: return ret; } +EXPORT_SYMBOL_GPL(ism_register_dmb); -static int ism_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) +int ism_unregister_dmb(struct ism_dev *ism, struct ism_dmb *dmb) { - struct ism_dev *ism = smcd->priv; union ism_unreg_dmb cmd; + unsigned long flags; int ret; memset(&cmd, 0, sizeof(cmd)); @@ -293,6 +373,10 @@ static int ism_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) cmd.request.dmb_tok = dmb->dmb_tok; + spin_lock_irqsave(&ism->lock, flags); + ism->sba_client_arr[dmb->sba_idx - ISM_DMB_BIT_OFFSET] = NO_CLIENT; + spin_unlock_irqrestore(&ism->lock, flags); + ret = ism_cmd(ism, &cmd); if (ret && ret != ISM_ERROR) goto out; @@ -301,10 +385,10 @@ static int ism_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) out: return ret; } +EXPORT_SYMBOL_GPL(ism_unregister_dmb); -static int ism_add_vlan_id(struct smcd_dev *smcd, u64 vlan_id) +static int ism_add_vlan_id(struct ism_dev *ism, u64 vlan_id) { - struct ism_dev *ism = smcd->priv; union ism_set_vlan_id cmd; memset(&cmd, 0, sizeof(cmd)); @@ -316,9 +400,8 @@ static int ism_add_vlan_id(struct smcd_dev *smcd, u64 vlan_id) return ism_cmd(ism, &cmd); } -static int ism_del_vlan_id(struct smcd_dev *smcd, u64 vlan_id) +static int ism_del_vlan_id(struct ism_dev *ism, u64 vlan_id) { - struct ism_dev *ism = smcd->priv; union ism_set_vlan_id cmd; memset(&cmd, 0, sizeof(cmd)); @@ -330,44 +413,15 @@ static int ism_del_vlan_id(struct smcd_dev *smcd, u64 vlan_id) return ism_cmd(ism, &cmd); } -static int ism_set_vlan_required(struct smcd_dev *smcd) -{ - return ism_cmd_simple(smcd->priv, ISM_SET_VLAN); -} - -static int ism_reset_vlan_required(struct smcd_dev *smcd) -{ - return ism_cmd_simple(smcd->priv, ISM_RESET_VLAN); -} - -static int ism_signal_ieq(struct smcd_dev *smcd, u64 rgid, u32 trigger_irq, - u32 event_code, u64 info) -{ - struct ism_dev *ism = smcd->priv; - union ism_sig_ieq cmd; - - memset(&cmd, 0, sizeof(cmd)); - cmd.request.hdr.cmd = ISM_SIGNAL_IEQ; - cmd.request.hdr.len = sizeof(cmd.request); - - cmd.request.rgid = rgid; - cmd.request.trigger_irq = trigger_irq; - cmd.request.event_code = event_code; - cmd.request.info = info; - - return ism_cmd(ism, &cmd); -} - static unsigned int max_bytes(unsigned int start, unsigned int len, unsigned int boundary) { return min(boundary - (start & (boundary - 1)), len); } -static int ism_move(struct smcd_dev *smcd, u64 dmb_tok, unsigned int idx, - bool sf, unsigned int offset, void *data, unsigned int size) +int ism_move(struct ism_dev *ism, u64 dmb_tok, unsigned int idx, bool sf, + unsigned int offset, void *data, unsigned int size) { - struct ism_dev *ism = smcd->priv; unsigned int bytes; u64 dmb_req; int ret; @@ -388,46 +442,13 @@ static int ism_move(struct smcd_dev *smcd, u64 dmb_tok, unsigned int idx, return 0; } - -static struct ism_systemeid SYSTEM_EID = { - .seid_string = "IBM-SYSZ-ISMSEID00000000", - .serial_number = "0000", - .type = "0000", -}; - -static void ism_create_system_eid(void) -{ - struct cpuid id; - u16 ident_tail; - char tmp[5]; - - get_cpu_id(&id); - ident_tail = (u16)(id.ident & ISM_IDENT_MASK); - snprintf(tmp, 5, "%04X", ident_tail); - memcpy(&SYSTEM_EID.serial_number, tmp, 4); - snprintf(tmp, 5, "%04X", id.machine); - memcpy(&SYSTEM_EID.type, tmp, 4); -} - -static void ism_get_system_eid(struct smcd_dev *smcd, u8 **eid) -{ - *eid = &SYSTEM_EID.seid_string[0]; -} - -static u16 ism_get_chid(struct smcd_dev *smcd) -{ - struct ism_dev *ismdev; - - ismdev = (struct ism_dev *)smcd->priv; - if (!ismdev || !ismdev->pdev) - return 0; - - return to_zpci(ismdev->pdev)->pchid; -} +EXPORT_SYMBOL_GPL(ism_move); static void ism_handle_event(struct ism_dev *ism) { - struct smcd_event *entry; + struct ism_event *entry; + struct ism_client *clt; + int i; while ((ism->ieq_idx + 1) != READ_ONCE(ism->ieq->header.idx)) { if (++(ism->ieq_idx) == ARRAY_SIZE(ism->ieq->entry)) @@ -435,7 +456,11 @@ static void ism_handle_event(struct ism_dev *ism) entry = &ism->ieq->entry[ism->ieq_idx]; debug_event(ism_debug_info, 2, entry, sizeof(*entry)); - smcd_handle_event(ism->smcd, entry); + for (i = 0; i < max_client; ++i) { + clt = ism->subs[i]; + if (clt) + clt->handle_event(ism, entry); + } } } @@ -444,6 +469,8 @@ static irqreturn_t ism_handle_irq(int irq, void *data) struct ism_dev *ism = data; unsigned long bit, end; unsigned long *bv; + u16 dmbemask; + u8 client_id; bv = (void *) &ism->sba->dmb_bits[ISM_DMB_WORD_OFFSET]; end = sizeof(ism->sba->dmb_bits) * BITS_PER_BYTE - ISM_DMB_BIT_OFFSET; @@ -457,9 +484,13 @@ static irqreturn_t ism_handle_irq(int irq, void *data) break; clear_bit_inv(bit, bv); + dmbemask = ism->sba->dmbe_mask[bit + ISM_DMB_BIT_OFFSET]; ism->sba->dmbe_mask[bit + ISM_DMB_BIT_OFFSET] = 0; barrier(); - smcd_handle_irq(ism->smcd, bit + ISM_DMB_BIT_OFFSET); + client_id = ism->sba_client_arr[bit]; + if (unlikely(client_id == NO_CLIENT || !ism->subs[client_id])) + continue; + ism->subs[client_id]->handle_irq(ism, bit + ISM_DMB_BIT_OFFSET, dmbemask); } if (ism->sba->e) { @@ -471,33 +502,24 @@ static irqreturn_t ism_handle_irq(int irq, void *data) return IRQ_HANDLED; } -static const struct smcd_ops ism_ops = { - .query_remote_gid = ism_query_rgid, - .register_dmb = ism_register_dmb, - .unregister_dmb = ism_unregister_dmb, - .add_vlan_id = ism_add_vlan_id, - .del_vlan_id = ism_del_vlan_id, - .set_vlan_required = ism_set_vlan_required, - .reset_vlan_required = ism_reset_vlan_required, - .signal_event = ism_signal_ieq, - .move_data = ism_move, - .get_system_eid = ism_get_system_eid, - .get_chid = ism_get_chid, -}; - static int ism_dev_init(struct ism_dev *ism) { struct pci_dev *pdev = ism->pdev; - int ret; + int i, ret; ret = pci_alloc_irq_vectors(pdev, 1, 1, PCI_IRQ_MSI); if (ret <= 0) goto out; + ism->sba_client_arr = kzalloc(ISM_NR_DMBS, GFP_KERNEL); + if (!ism->sba_client_arr) + goto free_vectors; + memset(ism->sba_client_arr, NO_CLIENT, ISM_NR_DMBS); + ret = request_irq(pci_irq_vector(pdev, 0), ism_handle_irq, 0, pci_name(pdev), ism); if (ret) - goto free_vectors; + goto free_client_arr; ret = register_sba(ism); if (ret) @@ -511,13 +533,24 @@ static int ism_dev_init(struct ism_dev *ism) if (ret) goto unreg_ieq; - if (!ism_add_vlan_id(ism->smcd, ISM_RESERVED_VLANID)) + if (!ism_add_vlan_id(ism, ISM_RESERVED_VLANID)) /* hardware is V2 capable */ - ism_create_system_eid(); + ism_v2_capable = true; + else + ism_v2_capable = false; + + mutex_lock(&ism_dev_list.mutex); + mutex_lock(&clients_lock); + for (i = 0; i < max_client; ++i) { + if (clients[i]) { + clients[i]->add(ism); + ism_setup_forwarding(clients[i], ism); + } + } + mutex_unlock(&clients_lock); - ret = smcd_register_dev(ism->smcd); - if (ret) - goto unreg_ieq; + list_add(&ism->list, &ism_dev_list.list); + mutex_unlock(&ism_dev_list.mutex); query_info(ism); return 0; @@ -528,12 +561,23 @@ static int ism_dev_init(struct ism_dev *ism) unregister_sba(ism); free_irq: free_irq(pci_irq_vector(pdev, 0), ism); +free_client_arr: + kfree(ism->sba_client_arr); free_vectors: pci_free_irq_vectors(pdev); out: return ret; } +static void ism_dev_release(struct device *dev) +{ + struct ism_dev *ism; + + ism = container_of(dev, struct ism_dev, dev); + + kfree(ism); +} + static int ism_probe(struct pci_dev *pdev, const struct pci_device_id *id) { struct ism_dev *ism; @@ -546,6 +590,13 @@ static int ism_probe(struct pci_dev *pdev, const struct pci_device_id *id) spin_lock_init(&ism->lock); dev_set_drvdata(&pdev->dev, ism); ism->pdev = pdev; + ism->dev.parent = &pdev->dev; + ism->dev.release = ism_dev_release; + device_initialize(&ism->dev); + dev_set_name(&ism->dev, dev_name(&pdev->dev)); + ret = device_add(&ism->dev); + if (ret) + goto err_dev; ret = pci_enable_device_mem(pdev); if (ret) @@ -563,44 +614,53 @@ static int ism_probe(struct pci_dev *pdev, const struct pci_device_id *id) dma_set_max_seg_size(&pdev->dev, SZ_1M); pci_set_master(pdev); - ism->smcd = smcd_alloc_dev(&pdev->dev, dev_name(&pdev->dev), &ism_ops, - ISM_NR_DMBS); - if (!ism->smcd) { - ret = -ENOMEM; - goto err_resource; - } - - ism->smcd->priv = ism; ret = ism_dev_init(ism); if (ret) - goto err_free; + goto err_resource; return 0; -err_free: - smcd_free_dev(ism->smcd); err_resource: pci_release_mem_regions(pdev); err_disable: pci_disable_device(pdev); err: - kfree(ism); + device_del(&ism->dev); +err_dev: dev_set_drvdata(&pdev->dev, NULL); + put_device(&ism->dev); + return ret; } static void ism_dev_exit(struct ism_dev *ism) { struct pci_dev *pdev = ism->pdev; + unsigned long flags; + int i; + + spin_lock_irqsave(&ism->lock, flags); + for (i = 0; i < max_client; ++i) + ism->subs[i] = NULL; + spin_unlock_irqrestore(&ism->lock, flags); + + mutex_lock(&ism_dev_list.mutex); + mutex_lock(&clients_lock); + for (i = 0; i < max_client; ++i) { + if (clients[i]) + clients[i]->remove(ism); + } + mutex_unlock(&clients_lock); - smcd_unregister_dev(ism->smcd); - if (SYSTEM_EID.serial_number[0] != '0' || - SYSTEM_EID.type[0] != '0') - ism_del_vlan_id(ism->smcd, ISM_RESERVED_VLANID); + if (ism_v2_capable) + ism_del_vlan_id(ism, ISM_RESERVED_VLANID); unregister_ieq(ism); unregister_sba(ism); free_irq(pci_irq_vector(pdev, 0), ism); + kfree(ism->sba_client_arr); pci_free_irq_vectors(pdev); + list_del_init(&ism->list); + mutex_unlock(&ism_dev_list.mutex); } static void ism_remove(struct pci_dev *pdev) @@ -609,11 +669,11 @@ static void ism_remove(struct pci_dev *pdev) ism_dev_exit(ism); - smcd_free_dev(ism->smcd); pci_release_mem_regions(pdev); pci_disable_device(pdev); + device_del(&ism->dev); dev_set_drvdata(&pdev->dev, NULL); - kfree(ism); + put_device(&ism->dev); } static struct pci_driver ism_driver = { @@ -631,6 +691,8 @@ static int __init ism_init(void) if (!ism_debug_info) return -ENODEV; + memset(clients, 0, sizeof(clients)); + max_client = 0; debug_register_view(ism_debug_info, &debug_hex_ascii_view); ret = pci_register_driver(&ism_driver); if (ret) @@ -647,3 +709,150 @@ static void __exit ism_exit(void) module_init(ism_init); module_exit(ism_exit); + +/*************************** SMC-D Implementation *****************************/ + +#if IS_ENABLED(CONFIG_SMC) +static int ism_query_rgid(struct ism_dev *ism, u64 rgid, u32 vid_valid, + u32 vid) +{ + union ism_query_rgid cmd; + + memset(&cmd, 0, sizeof(cmd)); + cmd.request.hdr.cmd = ISM_QUERY_RGID; + cmd.request.hdr.len = sizeof(cmd.request); + + cmd.request.rgid = rgid; + cmd.request.vlan_valid = vid_valid; + cmd.request.vlan_id = vid; + + return ism_cmd(ism, &cmd); +} + +static int smcd_query_rgid(struct smcd_dev *smcd, struct smcd_gid *rgid, + u32 vid_valid, u32 vid) +{ + return ism_query_rgid(smcd->priv, rgid->gid, vid_valid, vid); +} + +static int smcd_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb, + void *client) +{ + return ism_register_dmb(smcd->priv, (struct ism_dmb *)dmb, client); +} + +static int smcd_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) +{ + return ism_unregister_dmb(smcd->priv, (struct ism_dmb *)dmb); +} + +static int smcd_add_vlan_id(struct smcd_dev *smcd, u64 vlan_id) +{ + return ism_add_vlan_id(smcd->priv, vlan_id); +} + +static int smcd_del_vlan_id(struct smcd_dev *smcd, u64 vlan_id) +{ + return ism_del_vlan_id(smcd->priv, vlan_id); +} + +static int smcd_set_vlan_required(struct smcd_dev *smcd) +{ + return ism_cmd_simple(smcd->priv, ISM_SET_VLAN); +} + +static int smcd_reset_vlan_required(struct smcd_dev *smcd) +{ + return ism_cmd_simple(smcd->priv, ISM_RESET_VLAN); +} + +static int ism_signal_ieq(struct ism_dev *ism, u64 rgid, u32 trigger_irq, + u32 event_code, u64 info) +{ + union ism_sig_ieq cmd; + + memset(&cmd, 0, sizeof(cmd)); + cmd.request.hdr.cmd = ISM_SIGNAL_IEQ; + cmd.request.hdr.len = sizeof(cmd.request); + + cmd.request.rgid = rgid; + cmd.request.trigger_irq = trigger_irq; + cmd.request.event_code = event_code; + cmd.request.info = info; + + return ism_cmd(ism, &cmd); +} + +static int smcd_signal_ieq(struct smcd_dev *smcd, struct smcd_gid *rgid, + u32 trigger_irq, u32 event_code, u64 info) +{ + return ism_signal_ieq(smcd->priv, rgid->gid, + trigger_irq, event_code, info); +} + +static int smcd_move(struct smcd_dev *smcd, u64 dmb_tok, unsigned int idx, + bool sf, unsigned int offset, void *data, + unsigned int size) +{ + return ism_move(smcd->priv, dmb_tok, idx, sf, offset, data, size); +} + +static int smcd_supports_v2(void) +{ + return ism_v2_capable; +} + +static u64 ism_get_local_gid(struct ism_dev *ism) +{ + return ism->local_gid; +} + +static void smcd_get_local_gid(struct smcd_dev *smcd, + struct smcd_gid *smcd_gid) +{ + smcd_gid->gid = ism_get_local_gid(smcd->priv); + smcd_gid->gid_ext = 0; +} + +static u16 ism_get_chid(struct ism_dev *ism) +{ + if (!ism || !ism->pdev) + return 0; + + return to_zpci(ism->pdev)->pchid; +} + +static u16 smcd_get_chid(struct smcd_dev *smcd) +{ + return ism_get_chid(smcd->priv); +} + +static inline struct device *smcd_get_dev(struct smcd_dev *dev) +{ + struct ism_dev *ism = dev->priv; + + return &ism->dev; +} + +static const struct smcd_ops ism_ops = { + .query_remote_gid = smcd_query_rgid, + .register_dmb = smcd_register_dmb, + .unregister_dmb = smcd_unregister_dmb, + .add_vlan_id = smcd_add_vlan_id, + .del_vlan_id = smcd_del_vlan_id, + .set_vlan_required = smcd_set_vlan_required, + .reset_vlan_required = smcd_reset_vlan_required, + .signal_event = smcd_signal_ieq, + .move_data = smcd_move, + .supports_v2 = smcd_supports_v2, + .get_local_gid = smcd_get_local_gid, + .get_chid = smcd_get_chid, + .get_dev = smcd_get_dev, +}; + +const struct smcd_ops *ism_get_smcd_ops(void) +{ + return &ism_ops; +} +EXPORT_SYMBOL_GPL(ism_get_smcd_ops); +#endif diff --git a/include/linux/ism.h b/include/linux/ism.h new file mode 100644 index 0000000000000000000000000000000000000000..5428edd9098231f80cbcb297846f19ec584dcc7f --- /dev/null +++ b/include/linux/ism.h @@ -0,0 +1,92 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Internal Shared Memory + * + * Definitions for the ISM module + * + * Copyright IBM Corp. 2022 + */ +#ifndef _ISM_H +#define _ISM_H + +#include + +struct ism_dmb { + u64 dmb_tok; + u64 rgid; + u32 dmb_len; + u32 sba_idx; + u32 vlan_valid; + u32 vlan_id; + void *cpu_addr; + dma_addr_t dma_addr; +}; + +/* Unless we gain unexpected popularity, this limit should hold for a while */ +#define MAX_CLIENTS 8 +#define ISM_NR_DMBS 1920 + +struct ism_dev { + spinlock_t lock; /* protects the ism device */ + struct list_head list; + struct pci_dev *pdev; + + struct ism_sba *sba; + dma_addr_t sba_dma_addr; + DECLARE_BITMAP(sba_bitmap, ISM_NR_DMBS); + u8 *sba_client_arr; /* entries are indices into 'clients' array */ + void *priv[MAX_CLIENTS]; + + struct ism_eq *ieq; + dma_addr_t ieq_dma_addr; + + struct device dev; + u64 local_gid; + int ieq_idx; + + struct ism_client *subs[MAX_CLIENTS]; +}; + +struct ism_event { + u32 type; + u32 code; + u64 tok; + u64 time; + u64 info; +}; + +struct ism_client { + const char *name; + void (*add)(struct ism_dev *dev); + void (*remove)(struct ism_dev *dev); + void (*handle_event)(struct ism_dev *dev, struct ism_event *event); + /* Parameter dmbemask contains a bit vector with updated DMBEs, if sent + * via ism_move_data(). Callback function must handle all active bits + * indicated by dmbemask. + */ + void (*handle_irq)(struct ism_dev *dev, unsigned int bit, u16 dmbemask); + /* Private area - don't touch! */ + u8 id; +}; + +int ism_register_client(struct ism_client *client); +int ism_unregister_client(struct ism_client *client); +static inline void *ism_get_priv(struct ism_dev *dev, + struct ism_client *client) { + return dev->priv[client->id]; +} + +static inline void ism_set_priv(struct ism_dev *dev, struct ism_client *client, + void *priv) { + dev->priv[client->id] = priv; +} + +int ism_register_dmb(struct ism_dev *dev, struct ism_dmb *dmb, + struct ism_client *client); +int ism_unregister_dmb(struct ism_dev *dev, struct ism_dmb *dmb); +int ism_move(struct ism_dev *dev, u64 dmb_tok, unsigned int idx, bool sf, + unsigned int offset, void *data, unsigned int size); + +const struct smcd_ops *ism_get_smcd_ops(void); + +#endif /* _ISM_H */ diff --git a/include/net/smc.h b/include/net/smc.h index e441aa97ad61b1834f4983d10e15a34a7936f916..db84e4e35080f4792409993d0dfa62de8621e8a3 100644 --- a/include/net/smc.h +++ b/include/net/smc.h @@ -11,6 +11,14 @@ #ifndef _SMC_H #define _SMC_H +#include +#include +#include +#include +#include "linux/ism.h" + +struct sock; + #define SMC_MAX_PNETID_LEN 16 /* Max. length of PNET id */ struct smc_hashinfo { @@ -18,9 +26,6 @@ struct smc_hashinfo { struct hlist_head ht; }; -int smc_hash_sk(struct sock *sk); -void smc_unhash_sk(struct sock *sk); - /* SMCD/ISM device driver interface */ struct smcd_dmb { u64 dmb_tok; @@ -41,39 +46,43 @@ struct smcd_dmb { #define ISM_ERROR 0xFFFF -struct smcd_event { - u32 type; - u32 code; - u64 tok; - u64 time; - u64 info; -}; - struct smcd_dev; +struct smcd_gid { + u64 gid; + u64 gid_ext; +}; + struct smcd_ops { - int (*query_remote_gid)(struct smcd_dev *dev, u64 rgid, u32 vid_valid, - u32 vid); - int (*register_dmb)(struct smcd_dev *dev, struct smcd_dmb *dmb); + int (*query_remote_gid)(struct smcd_dev *dev, struct smcd_gid *rgid, + u32 vid_valid, u32 vid); + int (*register_dmb)(struct smcd_dev *dev, struct smcd_dmb *dmb, + void *client); int (*unregister_dmb)(struct smcd_dev *dev, struct smcd_dmb *dmb); - int (*add_vlan_id)(struct smcd_dev *dev, u64 vlan_id); - int (*del_vlan_id)(struct smcd_dev *dev, u64 vlan_id); - int (*set_vlan_required)(struct smcd_dev *dev); - int (*reset_vlan_required)(struct smcd_dev *dev); - int (*signal_event)(struct smcd_dev *dev, u64 rgid, u32 trigger_irq, - u32 event_code, u64 info); int (*move_data)(struct smcd_dev *dev, u64 dmb_tok, unsigned int idx, bool sf, unsigned int offset, void *data, unsigned int size); - void (*get_system_eid)(struct smcd_dev *dev, u8 **eid); + int (*supports_v2)(void); + void (*get_local_gid)(struct smcd_dev *dev, struct smcd_gid *gid); u16 (*get_chid)(struct smcd_dev *dev); + struct device* (*get_dev)(struct smcd_dev *dev); + + /* optional operations */ + int (*add_vlan_id)(struct smcd_dev *dev, u64 vlan_id); + int (*del_vlan_id)(struct smcd_dev *dev, u64 vlan_id); + int (*set_vlan_required)(struct smcd_dev *dev); + int (*reset_vlan_required)(struct smcd_dev *dev); + int (*signal_event)(struct smcd_dev *dev, struct smcd_gid *rgid, + u32 trigger_irq, u32 event_code, u64 info); + int (*support_dmb_nocopy)(struct smcd_dev *dev); + int (*attach_dmb)(struct smcd_dev *dev, struct smcd_dmb *dmb); + int (*detach_dmb)(struct smcd_dev *dev, u64 token); }; struct smcd_dev { const struct smcd_ops *ops; - struct device dev; void *priv; - u64 local_gid; + void *client; struct list_head list; spinlock_t lock; struct smc_connection **conn; @@ -88,11 +97,4 @@ struct smcd_dev { u8 going_away : 1; }; -struct smcd_dev *smcd_alloc_dev(struct device *parent, const char *name, - const struct smcd_ops *ops, int max_dmbs); -int smcd_register_dev(struct smcd_dev *smcd); -void smcd_unregister_dev(struct smcd_dev *smcd); -void smcd_free_dev(struct smcd_dev *smcd); -void smcd_handle_event(struct smcd_dev *dev, struct smcd_event *event); -void smcd_handle_irq(struct smcd_dev *dev, unsigned int bit); #endif /* _SMC_H */ diff --git a/include/net/sock.h b/include/net/sock.h index 30e128286a863792b6030531dc68f6208ce49d00..9795d680a280465307e45f9aaee499dbd9895334 100644 --- a/include/net/sock.h +++ b/include/net/sock.h @@ -1460,13 +1460,31 @@ proto_memory_pressure(struct proto *prot) #ifdef CONFIG_PROC_FS +#define PROTO_INUSE_NR 64 /* should be enough for the first time */ +struct prot_inuse { + int val[PROTO_INUSE_NR]; +}; /* Called with local bh disabled */ -void sock_prot_inuse_add(struct net *net, struct proto *prot, int inc); +static inline void sock_prot_inuse_add(const struct net *net, + const struct proto *prot, int val) +{ + __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); +} + +static inline void sock_inuse_add(const struct net *net, int val) +{ + this_cpu_add(*net->core.sock_inuse, val); +} + int sock_prot_inuse_get(struct net *net, struct proto *proto); int sock_inuse_get(struct net *net); #else -static inline void sock_prot_inuse_add(struct net *net, struct proto *prot, - int inc) +static inline void sock_prot_inuse_add(const struct net *net, + const struct proto *prot, int val) +{ +} + +static inline void sock_inuse_add(const struct net *net, int val) { } #endif diff --git a/include/uapi/linux/smc.h b/include/uapi/linux/smc.h index 0f7f87c70baf5905ce37eb414880500745f4f1a0..04ae1b11b20fbd554d0eb0f3853ad2bf0349a37e 100644 --- a/include/uapi/linux/smc.h +++ b/include/uapi/linux/smc.h @@ -38,6 +38,9 @@ enum { /* SMC PNET Table commands */ #define SMC_GENL_FAMILY_VERSION 1 #define SMC_PCI_ID_STR_LEN 16 /* Max length of pci id string */ +#define SMC_MAX_HOSTNAME_LEN 32 /* Max length of the hostname */ +#define SMC_MAX_UEID 4 /* Max number of user EIDs */ +#define SMC_MAX_EID_LEN 32 /* Max length of an EID */ /* SMC_GENL_FAMILY commands */ enum { @@ -49,6 +52,13 @@ enum { SMC_NETLINK_GET_DEV_SMCR, SMC_NETLINK_GET_STATS, SMC_NETLINK_GET_FBACK_STATS, + SMC_NETLINK_DUMP_UEID, + SMC_NETLINK_ADD_UEID, + SMC_NETLINK_REMOVE_UEID, + SMC_NETLINK_FLUSH_UEID, + SMC_NETLINK_DUMP_SEID, + SMC_NETLINK_ENABLE_SEID, + SMC_NETLINK_DISABLE_SEID, }; /* SMC_GENL_FAMILY top level attributes */ @@ -74,17 +84,30 @@ enum { SMC_NLA_SYS_IS_ISM_V2, /* u8 */ SMC_NLA_SYS_LOCAL_HOST, /* string */ SMC_NLA_SYS_SEID, /* string */ + SMC_NLA_SYS_IS_SMCR_V2, /* u8 */ __SMC_NLA_SYS_MAX, SMC_NLA_SYS_MAX = __SMC_NLA_SYS_MAX - 1 }; -/* SMC_NLA_LGR_V2 nested attributes */ +/* SMC_NLA_LGR_D_V2_COMMON and SMC_NLA_LGR_R_V2_COMMON nested attributes */ enum { SMC_NLA_LGR_V2_VER, /* u8 */ SMC_NLA_LGR_V2_REL, /* u8 */ SMC_NLA_LGR_V2_OS, /* u8 */ SMC_NLA_LGR_V2_NEG_EID, /* string */ SMC_NLA_LGR_V2_PEER_HOST, /* string */ + __SMC_NLA_LGR_V2_MAX, + SMC_NLA_LGR_V2_MAX = __SMC_NLA_LGR_V2_MAX - 1 +}; + +/* SMC_NLA_LGR_R_V2 nested attributes */ +enum { + SMC_NLA_LGR_R_V2_UNSPEC, + SMC_NLA_LGR_R_V2_DIRECT, /* u8 */ + SMC_NLA_LGR_R_V2_MAX_CONNS, /* u8 */ + SMC_NLA_LGR_R_V2_MAX_LINKS, /* u8 */ + __SMC_NLA_LGR_R_V2_MAX, + SMC_NLA_LGR_R_V2_MAX = __SMC_NLA_LGR_R_V2_MAX - 1 }; /* SMC_GEN_LGR_SMCR attributes */ @@ -96,6 +119,8 @@ enum { SMC_NLA_LGR_R_PNETID, /* string */ SMC_NLA_LGR_R_VLAN_ID, /* u8 */ SMC_NLA_LGR_R_CONNS_NUM, /* u32 */ + SMC_NLA_LGR_R_V2_COMMON, /* nest */ + SMC_NLA_LGR_R_V2, /* nest */ __SMC_NLA_LGR_R_MAX, SMC_NLA_LGR_R_MAX = __SMC_NLA_LGR_R_MAX - 1 }; @@ -128,7 +153,9 @@ enum { SMC_NLA_LGR_D_PNETID, /* string */ SMC_NLA_LGR_D_CHID, /* u16 */ SMC_NLA_LGR_D_PAD, /* flag */ - SMC_NLA_LGR_V2, /* nest */ + SMC_NLA_LGR_D_V2_COMMON, /* nest */ + SMC_NLA_LGR_D_EXT_GID, /* u64 */ + SMC_NLA_LGR_D_PEER_EXT_GID, /* u64 */ __SMC_NLA_LGR_D_MAX, SMC_NLA_LGR_D_MAX = __SMC_NLA_LGR_D_MAX - 1 }; @@ -242,4 +269,21 @@ enum { __SMC_NLA_FBACK_STATS_MAX, SMC_NLA_FBACK_STATS_MAX = __SMC_NLA_FBACK_STATS_MAX - 1 }; + +/* SMC_NETLINK_UEID attributes */ +enum { + SMC_NLA_EID_TABLE_UNSPEC, + SMC_NLA_EID_TABLE_ENTRY, /* string */ + __SMC_NLA_EID_TABLE_MAX, + SMC_NLA_EID_TABLE_MAX = __SMC_NLA_EID_TABLE_MAX - 1 +}; + +/* SMC_NETLINK_SEID attributes */ +enum { + SMC_NLA_SEID_UNSPEC, + SMC_NLA_SEID_ENTRY, /* string */ + SMC_NLA_SEID_ENABLED, /* u8 */ + __SMC_NLA_SEID_TABLE_MAX, + SMC_NLA_SEID_TABLE_MAX = __SMC_NLA_SEID_TABLE_MAX - 1 +}; #endif /* _UAPI_LINUX_SMC_H */ diff --git a/include/uapi/linux/smc_diag.h b/include/uapi/linux/smc_diag.h index 8cb3a6fef55301e8432ea99421b60b7d96ecad45..58eceb7f5df260b878303e84424eb6bf6faddc00 100644 --- a/include/uapi/linux/smc_diag.h +++ b/include/uapi/linux/smc_diag.h @@ -107,6 +107,8 @@ struct smcd_diag_dmbinfo { /* SMC-D Socket internals */ __aligned_u64 my_gid; /* My GID */ __aligned_u64 token; /* Token of DMB */ __aligned_u64 peer_token; /* Token of remote DMBE */ + __aligned_u64 peer_gid_ext; /* Peer GID (extended part) */ + __aligned_u64 my_gid_ext; /* My GID (extended part) */ }; #endif /* _UAPI_SMC_DIAG_H_ */ diff --git a/net/core/sock.c b/net/core/sock.c index 2e39845d094a295f58f8d3b50d7cb5770f4e0fac..0e4ba9d63542ce66a11d3a0fe6720399f670ef3a 100644 --- a/net/core/sock.c +++ b/net/core/sock.c @@ -143,8 +143,6 @@ static DEFINE_MUTEX(proto_list_mutex); static LIST_HEAD(proto_list); -static void sock_inuse_add(struct net *net, int val); - /** * sk_ns_capable - General socket capability test * @sk: Socket to use a capability on or through @@ -3382,19 +3380,8 @@ void sk_get_meminfo(const struct sock *sk, u32 *mem) } #ifdef CONFIG_PROC_FS -#define PROTO_INUSE_NR 64 /* should be enough for the first time */ -struct prot_inuse { - int val[PROTO_INUSE_NR]; -}; - static DECLARE_BITMAP(proto_inuse_idx, PROTO_INUSE_NR); -void sock_prot_inuse_add(struct net *net, struct proto *prot, int val) -{ - __this_cpu_add(net->core.prot_inuse->val[prot->inuse_idx], val); -} -EXPORT_SYMBOL_GPL(sock_prot_inuse_add); - int sock_prot_inuse_get(struct net *net, struct proto *prot) { int cpu, idx = prot->inuse_idx; @@ -3407,11 +3394,6 @@ int sock_prot_inuse_get(struct net *net, struct proto *prot) } EXPORT_SYMBOL_GPL(sock_prot_inuse_get); -static void sock_inuse_add(struct net *net, int val) -{ - this_cpu_add(*net->core.sock_inuse, val); -} - int sock_inuse_get(struct net *net) { int cpu, res = 0; @@ -3490,9 +3472,6 @@ static inline void release_proto_idx(struct proto *prot) { } -static void sock_inuse_add(struct net *net, int val) -{ -} #endif static void tw_prot_cleanup(struct timewait_sock_ops *twsk_prot) diff --git a/net/mptcp/subflow.c b/net/mptcp/subflow.c index 81c4089aa6956213df7906d32336b4ade4956cef..c0fb224987f5ecd886be0eb0a1c52455b3a2671d 100644 --- a/net/mptcp/subflow.c +++ b/net/mptcp/subflow.c @@ -1197,9 +1197,7 @@ int mptcp_subflow_create_socket(struct sock *sk, struct socket **new_sock) */ sf->sk->sk_net_refcnt = 1; get_net(net); -#ifdef CONFIG_PROC_FS - this_cpu_add(*net->core.sock_inuse, 1); -#endif + sock_inuse_add(net, 1); err = tcp_set_ulp(sf->sk, "mptcp"); release_sock(sf->sk); diff --git a/net/smc/Kconfig b/net/smc/Kconfig index 1ab3c5a2c5ada6c36e771b887cfa25515dc18bfc..ba5e6a2dd2fd768f7aaab2b8501e0cabe7fedbec 100644 --- a/net/smc/Kconfig +++ b/net/smc/Kconfig @@ -2,6 +2,7 @@ config SMC tristate "SMC socket protocol family" depends on INET && INFINIBAND + depends on m || ISM != m help SMC-R provides a "sockets over RDMA" solution making use of RDMA over Converged Ethernet (RoCE) technology to upgrade @@ -19,3 +20,16 @@ config SMC_DIAG smcss. if unsure, say Y. + +config SMC_LO + bool "SMC intra-OS shortcut with loopback-ism" + depends on SMC + default n + help + SMC_LO enables the creation of an Emulated-ISM device named + loopback-ism in SMC and makes use of it for transferring data + when communication occurs within the same OS. This helps in + convenient testing of SMC-D since loopback-ism is independent + of architecture or hardware. + + if unsure, say N. diff --git a/net/smc/Makefile b/net/smc/Makefile index 4444145a8ab3dd5c64d3538c03daaf128a9ea906..94049075888d1c209da8477d554cfe222f415bb8 100644 --- a/net/smc/Makefile +++ b/net/smc/Makefile @@ -4,3 +4,4 @@ obj-$(CONFIG_SMC_DIAG) += smc_diag.o smc-y := af_smc.o smc_pnet.o smc_ib.o smc_clc.o smc_core.o smc_wr.o smc_llc.o smc-y += smc_cdc.o smc_tx.o smc_rx.o smc_close.o smc_ism.o smc_netlink.o smc_stats.o smc-$(CONFIG_SYSCTL) += smc_sysctl.o +smc-$(CONFIG_SMC_LO) += smc_loopback.o diff --git a/net/smc/af_smc.c b/net/smc/af_smc.c index a365ea2f5484a3218f008f5a1d804280b6b51ff9..3e350495ab66698e568aea5aaf794d6242aea356 100644 --- a/net/smc/af_smc.c +++ b/net/smc/af_smc.c @@ -51,6 +51,7 @@ #include "smc_close.h" #include "smc_sysctl.h" #include "smc_stats.h" +#include "smc_loopback.h" static DEFINE_MUTEX(smc_server_lgr_pending); /* serialize link group * creation on server @@ -73,6 +74,36 @@ static void smc_set_keepalive(struct sock *sk, int val) smc->clcsock->sk->sk_prot->keepalive(smc->clcsock->sk, val); } +static struct sock *smc_tcp_syn_recv_sock(const struct sock *sk, + struct sk_buff *skb, + struct request_sock *req, + struct dst_entry *dst, + struct request_sock *req_unhash, + bool *own_req) +{ + struct smc_sock *smc; + + smc = smc_clcsock_user_data(sk); + + if (READ_ONCE(sk->sk_ack_backlog) + atomic_read(&smc->queued_smc_hs) > + sk->sk_max_ack_backlog) + goto drop; + + if (sk_acceptq_is_full(&smc->sk)) { + NET_INC_STATS(sock_net(sk), LINUX_MIB_LISTENOVERFLOWS); + goto drop; + } + + /* passthrough to original syn recv sock fct */ + return smc->ori_af_ops->syn_recv_sock(sk, skb, req, dst, req_unhash, + own_req); + +drop: + dst_release(dst); + tcp_listendrop(sk); + return NULL; +} + static struct smc_hashinfo smc_v4_hashinfo = { .lock = __RW_LOCK_UNLOCKED(smc_v4_hashinfo.lock), }; @@ -81,7 +112,7 @@ static struct smc_hashinfo smc_v6_hashinfo = { .lock = __RW_LOCK_UNLOCKED(smc_v6_hashinfo.lock), }; -int smc_hash_sk(struct sock *sk) +static int smc_hash_sk(struct sock *sk) { struct smc_hashinfo *h = sk->sk_prot->h.smc_hash; struct hlist_head *head; @@ -95,9 +126,8 @@ int smc_hash_sk(struct sock *sk) return 0; } -EXPORT_SYMBOL_GPL(smc_hash_sk); -void smc_unhash_sk(struct sock *sk) +static void smc_unhash_sk(struct sock *sk) { struct smc_hashinfo *h = sk->sk_prot->h.smc_hash; @@ -106,7 +136,6 @@ void smc_unhash_sk(struct sock *sk) sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1); write_unlock_bh(&h->lock); } -EXPORT_SYMBOL_GPL(smc_unhash_sk); /* This will be called before user really release sock_lock. So do the * work which we didn't do because of user hold the sock_lock in the @@ -179,7 +208,7 @@ static int __smc_release(struct smc_sock *smc) if (!smc->use_fallback) { rc = smc_close_active(smc); - sock_set_flag(sk, SOCK_DEAD); + smc_sock_set_flag(sk, SOCK_DEAD); sk->sk_shutdown |= SHUTDOWN_MASK; } else { if (sk->sk_state != SMC_CLOSED) { @@ -339,6 +368,7 @@ static int smc_bind(struct socket *sock, struct sockaddr *uaddr, goto out_rel; smc->clcsock->sk->sk_reuse = sk->sk_reuse; + smc->clcsock->sk->sk_reuseport = sk->sk_reuseport; rc = kernel_bind(smc->clcsock, uaddr, addr_len); out_rel: @@ -371,29 +401,11 @@ static int smc_bind(struct socket *sock, struct sockaddr *uaddr, static void smc_adjust_sock_bufsizes(struct sock *nsk, struct sock *osk, unsigned long mask) { - struct net *nnet = sock_net(nsk); - nsk->sk_userlocks = osk->sk_userlocks; - if (osk->sk_userlocks & SOCK_SNDBUF_LOCK) { + if (osk->sk_userlocks & SOCK_SNDBUF_LOCK) nsk->sk_sndbuf = osk->sk_sndbuf; - } else { - if (mask == SK_FLAGS_SMC_TO_CLC) - WRITE_ONCE(nsk->sk_sndbuf, - READ_ONCE(nnet->ipv4.sysctl_tcp_wmem[1])); - else - WRITE_ONCE(nsk->sk_sndbuf, - 2 * sysctl_smcr_wmem(nnet)); - } - if (osk->sk_userlocks & SOCK_RCVBUF_LOCK) { + if (osk->sk_userlocks & SOCK_RCVBUF_LOCK) nsk->sk_rcvbuf = osk->sk_rcvbuf; - } else { - if (mask == SK_FLAGS_SMC_TO_CLC) - WRITE_ONCE(nsk->sk_rcvbuf, - READ_ONCE(nnet->ipv4.sysctl_tcp_rmem[1])); - else - WRITE_ONCE(nsk->sk_rcvbuf, - 2 * sysctl_smcr_rmem(nnet)); - } } static void smc_copy_sock_settings(struct sock *nsk, struct sock *osk, @@ -458,11 +470,26 @@ static int smcr_lgr_reg_rmbs(struct smc_link *link, struct smc_buf_desc *rmb_desc) { struct smc_link_group *lgr = link->lgr; + bool do_slow = false; int i, rc = 0; rc = smc_llc_flow_initiate(lgr, SMC_LLC_FLOW_RKEY); if (rc) return rc; + + down_read(&lgr->llc_conf_lock); + for (i = 0; i < SMC_LINKS_PER_LGR_MAX; i++) { + if (!smc_link_active(&lgr->lnk[i])) + continue; + if (!rmb_desc->is_reg_mr[link->link_idx]) { + up_read(&lgr->llc_conf_lock); + goto slow_path; + } + } + /* mr register already */ + goto fast_path; +slow_path: + do_slow = true; /* protect against parallel smc_llc_cli_rkey_exchange() and * parallel smcr_link_reg_buf() */ @@ -474,7 +501,7 @@ static int smcr_lgr_reg_rmbs(struct smc_link *link, if (rc) goto out; } - +fast_path: /* exchange confirm_rkey msg with peer */ rc = smc_llc_do_confirm_rkey(link, rmb_desc); if (rc) { @@ -483,7 +510,7 @@ static int smcr_lgr_reg_rmbs(struct smc_link *link, } rmb_desc->is_conf_rkey = true; out: - up_write(&lgr->llc_conf_lock); + do_slow ? up_write(&lgr->llc_conf_lock) : up_read(&lgr->llc_conf_lock); smc_llc_flow_stop(lgr, &lgr->llc_flow_lcl); return rc; } @@ -541,23 +568,62 @@ static int smcr_clnt_conf_first_link(struct smc_sock *smc) smc_llc_link_active(link); smcr_lgr_set_type(link->lgr, SMC_LGR_SINGLE); - /* optional 2nd link, receive ADD LINK request from server */ - qentry = smc_llc_wait(link->lgr, NULL, SMC_LLC_WAIT_TIME, - SMC_LLC_ADD_LINK); - if (!qentry) { - struct smc_clc_msg_decline dclc; - - rc = smc_clc_wait_msg(smc, &dclc, sizeof(dclc), - SMC_CLC_DECLINE, CLC_WAIT_TIME_SHORT); - if (rc == -EAGAIN) - rc = 0; /* no DECLINE received, go with one link */ - return rc; + if (link->lgr->max_links > 1) { + /* optional 2nd link, receive ADD LINK request from server */ + qentry = smc_llc_wait(link->lgr, NULL, SMC_LLC_WAIT_TIME, + SMC_LLC_ADD_LINK); + if (!qentry) { + struct smc_clc_msg_decline dclc; + + rc = smc_clc_wait_msg(smc, &dclc, sizeof(dclc), + SMC_CLC_DECLINE, CLC_WAIT_TIME_SHORT); + if (rc == -EAGAIN) + rc = 0; /* no DECLINE received, go with one link */ + return rc; + } + smc_llc_flow_qentry_clr(&link->lgr->llc_flow_lcl); + smc_llc_cli_add_link(link, qentry); } - smc_llc_flow_qentry_clr(&link->lgr->llc_flow_lcl); - smc_llc_cli_add_link(link, qentry); return 0; } +static bool smc_isascii(char *hostname) +{ + int i; + + for (i = 0; i < SMC_MAX_HOSTNAME_LEN; i++) + if (!isascii(hostname[i])) + return false; + return true; +} + +static void smc_conn_save_peer_info_fce(struct smc_sock *smc, + struct smc_clc_msg_accept_confirm *clc) +{ + struct smc_clc_first_contact_ext *fce; + int clc_v2_len; + + if (clc->hdr.version == SMC_V1 || + !(clc->hdr.typev2 & SMC_FIRST_CONTACT_MASK)) + return; + + if (smc->conn.lgr->is_smcd) { + memcpy(smc->conn.lgr->negotiated_eid, clc->d1.eid, + SMC_MAX_EID_LEN); + clc_v2_len = offsetofend(struct smc_clc_msg_accept_confirm, d1); + } else { + memcpy(smc->conn.lgr->negotiated_eid, clc->r1.eid, + SMC_MAX_EID_LEN); + clc_v2_len = offsetofend(struct smc_clc_msg_accept_confirm, r1); + } + fce = (struct smc_clc_first_contact_ext *)(((u8 *)clc) + clc_v2_len); + smc->conn.lgr->peer_os = fce->os_type; + smc->conn.lgr->peer_smc_release = fce->release; + if (smc_isascii(fce->hostname)) + memcpy(smc->conn.lgr->peer_hostname, fce->hostname, + SMC_MAX_HOSTNAME_LEN); +} + static void smcr_conn_save_peer_info(struct smc_sock *smc, struct smc_clc_msg_accept_confirm *clc) { @@ -570,43 +636,17 @@ static void smcr_conn_save_peer_info(struct smc_sock *smc, smc->conn.tx_off = bufsize * (smc->conn.peer_rmbe_idx - 1); } -static bool smc_isascii(char *hostname) -{ - int i; - - for (i = 0; i < SMC_MAX_HOSTNAME_LEN; i++) - if (!isascii(hostname[i])) - return false; - return true; -} - static void smcd_conn_save_peer_info(struct smc_sock *smc, struct smc_clc_msg_accept_confirm *clc) { int bufsize = smc_uncompress_bufsize(clc->d0.dmbe_size); smc->conn.peer_rmbe_idx = clc->d0.dmbe_idx; - smc->conn.peer_token = clc->d0.token; + smc->conn.peer_token = ntohll(clc->d0.token); /* msg header takes up space in the buffer */ smc->conn.peer_rmbe_size = bufsize - sizeof(struct smcd_cdc_msg); atomic_set(&smc->conn.peer_rmbe_space, smc->conn.peer_rmbe_size); smc->conn.tx_off = bufsize * smc->conn.peer_rmbe_idx; - if (clc->hdr.version > SMC_V1 && - (clc->hdr.typev2 & SMC_FIRST_CONTACT_MASK)) { - struct smc_clc_msg_accept_confirm_v2 *clc_v2 = - (struct smc_clc_msg_accept_confirm_v2 *)clc; - struct smc_clc_first_contact_ext *fce = - (struct smc_clc_first_contact_ext *) - (((u8 *)clc_v2) + sizeof(*clc_v2)); - - memcpy(smc->conn.lgr->negotiated_eid, clc_v2->eid, - SMC_MAX_EID_LEN); - smc->conn.lgr->peer_os = fce->os_type; - smc->conn.lgr->peer_smc_release = fce->release; - if (smc_isascii(fce->hostname)) - memcpy(smc->conn.lgr->peer_hostname, fce->hostname, - SMC_MAX_HOSTNAME_LEN); - } } static void smc_conn_save_peer_info(struct smc_sock *smc, @@ -616,14 +656,16 @@ static void smc_conn_save_peer_info(struct smc_sock *smc, smcd_conn_save_peer_info(smc, clc); else smcr_conn_save_peer_info(smc, clc); + smc_conn_save_peer_info_fce(smc, clc); } static void smc_link_save_peer_info(struct smc_link *link, - struct smc_clc_msg_accept_confirm *clc) + struct smc_clc_msg_accept_confirm *clc, + struct smc_init_info *ini) { link->peer_qpn = ntoh24(clc->r0.qpn); - memcpy(link->peer_gid, clc->r0.lcl.gid, SMC_GID_SIZE); - memcpy(link->peer_mac, clc->r0.lcl.mac, sizeof(link->peer_mac)); + memcpy(link->peer_gid, ini->peer_gid, SMC_GID_SIZE); + memcpy(link->peer_mac, ini->peer_mac, sizeof(link->peer_mac)); link->peer_psn = ntoh24(clc->r0.psn); link->peer_mtu = clc->r0.qp_mtu; } @@ -802,6 +844,7 @@ static int smc_switch_to_fallback(struct smc_sock *smc, int reason_code) smc->clcsock->file->private_data = smc->clcsock; smc->clcsock->wq.fasync_list = smc->sk.sk_socket->wq.fasync_list; + smc->sk.sk_socket->wq.fasync_list = NULL; /* There might be some wait entries remaining * in smc sk->sk_wq and they should be woken up @@ -856,13 +899,14 @@ static int smc_connect_decline_fallback(struct smc_sock *smc, int reason_code, return smc_connect_fallback(smc, reason_code); } -/* abort connecting */ -static void smc_connect_abort(struct smc_sock *smc, int local_first) +static void smc_conn_abort(struct smc_sock *smc, int local_first) { + struct smc_connection *conn = &smc->conn; + struct smc_link_group *lgr = conn->lgr; + + smc_conn_free(conn); if (local_first) - smc_lgr_cleanup_early(&smc->conn); - else - smc_conn_free(&smc->conn); + smc_lgr_cleanup_early(lgr); } /* check if there is a rdma device available for this connection. */ @@ -874,7 +918,9 @@ static int smc_find_rdma_device(struct smc_sock *smc, struct smc_init_info *ini) * used for the internal TCP socket */ smc_pnet_find_roce_resource(smc->clcsock->sk, ini); - if (!ini->ib_dev) + if (!ini->check_smcrv2 && !ini->ib_dev) + return SMC_CLC_DECL_NOSMCRDEV; + if (ini->check_smcrv2 && !ini->smcrv2.ib_dev_v2) return SMC_CLC_DECL_NOSMCRDEV; return 0; } @@ -912,7 +958,8 @@ static int smc_find_ism_v2_device_clnt(struct smc_sock *smc, { int rc = SMC_CLC_DECL_NOSMCDDEV; struct smcd_dev *smcd; - int i = 1; + int i = 1, entry = 1; + bool is_emulated; u16 chid; if (smcd_indicated(ini->smc_type_v1)) @@ -924,14 +971,23 @@ static int smc_find_ism_v2_device_clnt(struct smc_sock *smc, chid = smc_ism_get_chid(smcd); if (!smc_find_ism_v2_is_unique_chid(chid, ini, i)) continue; + is_emulated = __smc_ism_is_emulated(chid); if (!smc_pnet_is_pnetid_set(smcd->pnetid) || smc_pnet_is_ndev_pnetid(sock_net(&smc->sk), smcd->pnetid)) { + if (is_emulated && entry == SMCD_CLC_MAX_V2_GID_ENTRIES) + /* It's the last GID-CHID entry left in CLC + * Proposal SMC-Dv2 extension, but an Emulated- + * ISM device will take two entries. So give + * up it and try the next potential ISM device. + */ + continue; ini->ism_dev[i] = smcd; ini->ism_chid[i] = chid; ini->is_smcd = true; rc = 0; i++; - if (i > SMC_MAX_ISM_DEVS) + entry = is_emulated ? entry + 2 : entry + 1; + if (entry > SMCD_CLC_MAX_V2_GID_ENTRIES) break; } } @@ -958,27 +1014,42 @@ static int smc_find_proposal_devices(struct smc_sock *smc, int rc = 0; /* check if there is an ism device available */ - if (ini->smcd_version & SMC_V1) { - if (smc_find_ism_device(smc, ini) || - smc_connect_ism_vlan_setup(smc, ini)) { - if (ini->smc_type_v1 == SMC_TYPE_B) - ini->smc_type_v1 = SMC_TYPE_R; - else - ini->smc_type_v1 = SMC_TYPE_N; - } /* else ISM V1 is supported for this connection */ - if (smc_find_rdma_device(smc, ini)) { - if (ini->smc_type_v1 == SMC_TYPE_B) - ini->smc_type_v1 = SMC_TYPE_D; - else - ini->smc_type_v1 = SMC_TYPE_N; - } /* else RDMA is supported for this connection */ - } - if (smc_ism_is_v2_capable() && smc_find_ism_v2_device_clnt(smc, ini)) - ini->smc_type_v2 = SMC_TYPE_N; + if (!(ini->smcd_version & SMC_V1) || + smc_find_ism_device(smc, ini) || + smc_connect_ism_vlan_setup(smc, ini)) + ini->smcd_version &= ~SMC_V1; + /* else ISM V1 is supported for this connection */ + + /* check if there is an rdma device available */ + if (!(ini->smcr_version & SMC_V1) || + smc_find_rdma_device(smc, ini)) + ini->smcr_version &= ~SMC_V1; + /* else RDMA is supported for this connection */ + + ini->smc_type_v1 = smc_indicated_type(ini->smcd_version & SMC_V1, + ini->smcr_version & SMC_V1); + + /* check if there is an ism v2 device available */ + if (!(ini->smcd_version & SMC_V2) || + !smc_ism_is_v2_capable() || + smc_find_ism_v2_device_clnt(smc, ini)) + ini->smcd_version &= ~SMC_V2; + + /* check if there is an rdma v2 device available */ + ini->check_smcrv2 = true; + ini->smcrv2.saddr = smc->clcsock->sk->sk_rcv_saddr; + if (!(ini->smcr_version & SMC_V2) || + smc->clcsock->sk->sk_family != AF_INET || + !smc_clc_ueid_count() || + smc_find_rdma_device(smc, ini)) + ini->smcr_version &= ~SMC_V2; + ini->check_smcrv2 = false; + + ini->smc_type_v2 = smc_indicated_type(ini->smcd_version & SMC_V2, + ini->smcr_version & SMC_V2); /* if neither ISM nor RDMA are supported, fallback */ - if (!smcr_indicated(ini->smc_type_v1) && - ini->smc_type_v1 == SMC_TYPE_N && ini->smc_type_v2 == SMC_TYPE_N) + if (ini->smc_type_v1 == SMC_TYPE_N && ini->smc_type_v2 == SMC_TYPE_N) rc = SMC_CLC_DECL_NOSMCDEV; return rc; @@ -998,13 +1069,13 @@ static int smc_connect_ism_vlan_cleanup(struct smc_sock *smc, } #define SMC_CLC_MAX_ACCEPT_LEN \ - (sizeof(struct smc_clc_msg_accept_confirm_v2) + \ - sizeof(struct smc_clc_first_contact_ext) + \ + (sizeof(struct smc_clc_msg_accept_confirm) + \ + sizeof(struct smc_clc_first_contact_ext_v2x) + \ sizeof(struct smc_clc_msg_trail)) /* CLC handshake during connect */ static int smc_connect_clc(struct smc_sock *smc, - struct smc_clc_msg_accept_confirm_v2 *aclc2, + struct smc_clc_msg_accept_confirm *aclc, struct smc_init_info *ini) { int rc = 0; @@ -1014,10 +1085,73 @@ static int smc_connect_clc(struct smc_sock *smc, if (rc) return rc; /* receive SMC Accept CLC message */ - return smc_clc_wait_msg(smc, aclc2, SMC_CLC_MAX_ACCEPT_LEN, + return smc_clc_wait_msg(smc, aclc, SMC_CLC_MAX_ACCEPT_LEN, SMC_CLC_ACCEPT, CLC_WAIT_TIME); } +void smc_fill_gid_list(struct smc_link_group *lgr, + struct smc_gidlist *gidlist, + struct smc_ib_device *known_dev, u8 *known_gid) +{ + struct smc_init_info *alt_ini = NULL; + + memset(gidlist, 0, sizeof(*gidlist)); + memcpy(gidlist->list[gidlist->len++], known_gid, SMC_GID_SIZE); + + alt_ini = kzalloc(sizeof(*alt_ini), GFP_KERNEL); + if (!alt_ini) + goto out; + + alt_ini->vlan_id = lgr->vlan_id; + alt_ini->check_smcrv2 = true; + alt_ini->smcrv2.saddr = lgr->saddr; + smc_pnet_find_alt_roce(lgr, alt_ini, known_dev); + + if (!alt_ini->smcrv2.ib_dev_v2) + goto out; + + memcpy(gidlist->list[gidlist->len++], alt_ini->smcrv2.ib_gid_v2, + SMC_GID_SIZE); + +out: + kfree(alt_ini); +} + +static int smc_connect_rdma_v2_prepare(struct smc_sock *smc, + struct smc_clc_msg_accept_confirm *aclc, + struct smc_init_info *ini) +{ + struct smc_clc_first_contact_ext *fce = + smc_get_clc_first_contact_ext(aclc, false); + struct net *net = sock_net(&smc->sk); + int rc; + + if (!ini->first_contact_peer || aclc->hdr.version == SMC_V1) + return 0; + + if (fce->v2_direct) { + memcpy(ini->smcrv2.nexthop_mac, &aclc->r0.lcl.mac, ETH_ALEN); + ini->smcrv2.uses_gateway = false; + } else { + if (smc_ib_find_route(net, smc->clcsock->sk->sk_rcv_saddr, + smc_ib_gid_to_ipv4(aclc->r0.lcl.gid), + ini->smcrv2.nexthop_mac, + &ini->smcrv2.uses_gateway)) + return SMC_CLC_DECL_NOROUTE; + if (!ini->smcrv2.uses_gateway) { + /* mismatch: peer claims indirect, but its direct */ + return SMC_CLC_DECL_NOINDIRECT; + } + } + + ini->release_nr = fce->release; + rc = smc_clc_clnt_v2x_features_validate(fce, ini); + if (rc) + return rc; + + return 0; +} + /* setup for RDMA connection of client */ static int smc_connect_rdma(struct smc_sock *smc, struct smc_clc_msg_accept_confirm *aclc, @@ -1025,11 +1159,20 @@ static int smc_connect_rdma(struct smc_sock *smc, { int i, reason_code = 0; struct smc_link *link; + u8 *eid = NULL; ini->is_smcd = false; - ini->ib_lcl = &aclc->r0.lcl; ini->ib_clcqpn = ntoh24(aclc->r0.qpn); ini->first_contact_peer = aclc->hdr.typev2 & SMC_FIRST_CONTACT_MASK; + memcpy(ini->peer_systemid, aclc->r0.lcl.id_for_peer, SMC_SYSTEMID_LEN); + memcpy(ini->peer_gid, aclc->r0.lcl.gid, SMC_GID_SIZE); + memcpy(ini->peer_mac, aclc->r0.lcl.mac, ETH_ALEN); + ini->max_conns = SMC_CONN_PER_LGR_MAX; + ini->max_links = SMC_LINKS_ADD_LNK_MAX; + + reason_code = smc_connect_rdma_v2_prepare(smc, aclc, ini); + if (reason_code) + return reason_code; mutex_lock(&smc_client_lgr_pending); reason_code = smc_conn_create(smc, ini); @@ -1051,8 +1194,9 @@ static int smc_connect_rdma(struct smc_sock *smc, if (l->peer_qpn == ntoh24(aclc->r0.qpn) && !memcmp(l->peer_gid, &aclc->r0.lcl.gid, SMC_GID_SIZE) && - !memcmp(l->peer_mac, &aclc->r0.lcl.mac, - sizeof(l->peer_mac))) { + (aclc->hdr.version > SMC_V1 || + !memcmp(l->peer_mac, &aclc->r0.lcl.mac, + sizeof(l->peer_mac)))) { link = l; break; } @@ -1071,7 +1215,7 @@ static int smc_connect_rdma(struct smc_sock *smc, } if (ini->first_contact_local) - smc_link_save_peer_info(link, aclc); + smc_link_save_peer_info(link, aclc, ini); if (smc_rmb_rtoken_handling(&smc->conn, link, aclc)) { reason_code = SMC_CLC_DECL_ERR_RTOK; @@ -1100,8 +1244,15 @@ static int smc_connect_rdma(struct smc_sock *smc, } } + if (aclc->hdr.version > SMC_V1) { + eid = aclc->r1.eid; + if (ini->first_contact_local) + smc_fill_gid_list(link->lgr, &ini->smcrv2.gidlist, + link->smcibdev, link->gid); + } + reason_code = smc_clc_send_confirm(smc, ini->first_contact_local, - SMC_V1); + aclc->hdr.version, eid, ini); if (reason_code) goto connect_abort; @@ -1124,7 +1275,7 @@ static int smc_connect_rdma(struct smc_sock *smc, return 0; connect_abort: - smc_connect_abort(smc, ini->first_contact_local); + smc_conn_abort(smc, ini->first_contact_local); mutex_unlock(&smc_client_lgr_pending); smc->connect_nonblock = 0; @@ -1135,13 +1286,13 @@ static int smc_connect_rdma(struct smc_sock *smc, * Determine from the CHID of the received CLC ACCEPT the ISM device chosen. */ static int -smc_v2_determine_accepted_chid(struct smc_clc_msg_accept_confirm_v2 *aclc, +smc_v2_determine_accepted_chid(struct smc_clc_msg_accept_confirm *aclc, struct smc_init_info *ini) { int i; for (i = 0; i < ini->ism_offered_cnt + 1; i++) { - if (ini->ism_chid[i] == ntohs(aclc->chid)) { + if (ini->ism_chid[i] == ntohs(aclc->d1.chid)) { ini->ism_selected = i; return 0; } @@ -1155,20 +1306,33 @@ static int smc_connect_ism(struct smc_sock *smc, struct smc_clc_msg_accept_confirm *aclc, struct smc_init_info *ini) { + u8 *eid = NULL; int rc = 0; ini->is_smcd = true; ini->first_contact_peer = aclc->hdr.typev2 & SMC_FIRST_CONTACT_MASK; if (aclc->hdr.version == SMC_V2) { - struct smc_clc_msg_accept_confirm_v2 *aclc_v2 = - (struct smc_clc_msg_accept_confirm_v2 *)aclc; + if (ini->first_contact_peer) { + struct smc_clc_first_contact_ext *fce = + smc_get_clc_first_contact_ext(aclc, true); + + ini->release_nr = fce->release; + rc = smc_clc_clnt_v2x_features_validate(fce, ini); + if (rc) + return rc; + } - rc = smc_v2_determine_accepted_chid(aclc_v2, ini); + rc = smc_v2_determine_accepted_chid(aclc, ini); if (rc) return rc; + + if (__smc_ism_is_emulated(ini->ism_chid[ini->ism_selected])) + ini->ism_peer_gid[ini->ism_selected].gid_ext = + ntohll(aclc->d1.gid_ext); + /* for non-Emulated-ISM devices, peer gid_ext remains 0. */ } - ini->ism_peer_gid[ini->ism_selected] = aclc->d0.gid; + ini->ism_peer_gid[ini->ism_selected].gid = ntohll(aclc->d0.gid); /* there is only one lgr role for SMC-D; use server lock */ mutex_lock(&smc_server_lgr_pending); @@ -1186,12 +1350,23 @@ static int smc_connect_ism(struct smc_sock *smc, } smc_conn_save_peer_info(smc, aclc); + + if (smc_ism_support_dmb_nocopy(smc->conn.lgr->smcd)) { + rc = smcd_buf_attach(smc); + if (rc) { + rc = SMC_CLC_DECL_MEM; /* try to fallback */ + goto connect_abort; + } + } smc_close_init(smc); smc_rx_init(smc); smc_tx_init(smc); + if (aclc->hdr.version > SMC_V1) + eid = aclc->d1.eid; + rc = smc_clc_send_confirm(smc, ini->first_contact_local, - aclc->hdr.version); + aclc->hdr.version, eid, ini); if (rc) goto connect_abort; mutex_unlock(&smc_server_lgr_pending); @@ -1203,7 +1378,7 @@ static int smc_connect_ism(struct smc_sock *smc, return 0; connect_abort: - smc_connect_abort(smc, ini->first_contact_local); + smc_conn_abort(smc, ini->first_contact_local); mutex_unlock(&smc_server_lgr_pending); smc->connect_nonblock = 0; @@ -1214,16 +1389,19 @@ static int smc_connect_ism(struct smc_sock *smc, static int smc_connect_check_aclc(struct smc_init_info *ini, struct smc_clc_msg_accept_confirm *aclc) { - if ((aclc->hdr.typev1 == SMC_TYPE_R && - !smcr_indicated(ini->smc_type_v1)) || - (aclc->hdr.typev1 == SMC_TYPE_D && - ((!smcd_indicated(ini->smc_type_v1) && - !smcd_indicated(ini->smc_type_v2)) || - (aclc->hdr.version == SMC_V1 && - !smcd_indicated(ini->smc_type_v1)) || - (aclc->hdr.version == SMC_V2 && - !smcd_indicated(ini->smc_type_v2))))) - return SMC_CLC_DECL_MODEUNSUPP; + if (aclc->hdr.version >= SMC_V2) { + if ((aclc->hdr.typev1 == SMC_TYPE_R && + !smcr_indicated(ini->smc_type_v2)) || + (aclc->hdr.typev1 == SMC_TYPE_D && + !smcd_indicated(ini->smc_type_v2))) + return SMC_CLC_DECL_MODEUNSUPP; + } else { + if ((aclc->hdr.typev1 == SMC_TYPE_R && + !smcr_indicated(ini->smc_type_v1)) || + (aclc->hdr.typev1 == SMC_TYPE_D && + !smcd_indicated(ini->smc_type_v1))) + return SMC_CLC_DECL_MODEUNSUPP; + } return 0; } @@ -1232,7 +1410,6 @@ static int smc_connect_check_aclc(struct smc_init_info *ini, static int __smc_connect(struct smc_sock *smc) { u8 version = smc_ism_is_v2_capable() ? SMC_V2 : SMC_V1; - struct smc_clc_msg_accept_confirm_v2 *aclc2; struct smc_clc_msg_accept_confirm *aclc; struct smc_init_info *ini = NULL; u8 *buf = NULL; @@ -1255,19 +1432,16 @@ static int __smc_connect(struct smc_sock *smc) return smc_connect_decline_fallback(smc, SMC_CLC_DECL_MEM, version); - ini->smcd_version = SMC_V1; - ini->smcd_version |= smc_ism_is_v2_capable() ? SMC_V2 : 0; + ini->smcd_version = SMC_V1 | SMC_V2; + ini->smcr_version = SMC_V1 | SMC_V2; ini->smc_type_v1 = SMC_TYPE_B; - ini->smc_type_v2 = smc_ism_is_v2_capable() ? SMC_TYPE_D : SMC_TYPE_N; + ini->smc_type_v2 = SMC_TYPE_B; /* get vlan id from IP device */ if (smc_vlan_by_tcpsk(smc->clcsock, ini)) { ini->smcd_version &= ~SMC_V1; + ini->smcr_version = 0; ini->smc_type_v1 = SMC_TYPE_N; - if (!ini->smcd_version) { - rc = SMC_CLC_DECL_GETVLANERR; - goto fallback; - } } rc = smc_find_proposal_devices(smc, ini); @@ -1279,26 +1453,33 @@ static int __smc_connect(struct smc_sock *smc) rc = SMC_CLC_DECL_MEM; goto fallback; } - aclc2 = (struct smc_clc_msg_accept_confirm_v2 *)buf; - aclc = (struct smc_clc_msg_accept_confirm *)aclc2; + aclc = (struct smc_clc_msg_accept_confirm *)buf; /* perform CLC handshake */ - rc = smc_connect_clc(smc, aclc2, ini); - if (rc) + rc = smc_connect_clc(smc, aclc, ini); + if (rc) { + /* -EAGAIN on timeout, see tcp_recvmsg() */ + if (rc == -EAGAIN) { + rc = -ETIMEDOUT; + smc->sk.sk_err = ETIMEDOUT; + } goto vlan_cleanup; + } /* check if smc modes and versions of CLC proposal and accept match */ rc = smc_connect_check_aclc(ini, aclc); version = aclc->hdr.version == SMC_V1 ? SMC_V1 : SMC_V2; - ini->smcd_version = version; if (rc) goto vlan_cleanup; /* depending on previous steps, connect using rdma or ism */ - if (aclc->hdr.typev1 == SMC_TYPE_R) + if (aclc->hdr.typev1 == SMC_TYPE_R) { + ini->smcr_version = version; rc = smc_connect_rdma(smc, aclc, ini); - else if (aclc->hdr.typev1 == SMC_TYPE_D) + } else if (aclc->hdr.typev1 == SMC_TYPE_D) { + ini->smcd_version = version; rc = smc_connect_ism(smc, aclc, ini); + } if (rc) goto vlan_cleanup; @@ -1409,7 +1590,6 @@ static int smc_connect(struct socket *sock, struct sockaddr *addr, rc = -EISCONN; goto out; case SMC_INIT: - rc = 0; break; } @@ -1484,7 +1664,7 @@ static int smc_clcsock_accept(struct smc_sock *lsmc, struct smc_sock **new_smc) if (new_clcsock) sock_release(new_clcsock); new_sk->sk_state = SMC_CLOSED; - sock_set_flag(new_sk, SOCK_DEAD); + smc_sock_set_flag(new_sk, SOCK_DEAD); mutex_unlock(&lsmc->clcsock_release_lock); sock_put(new_sk); /* final */ *new_smc = NULL; @@ -1638,8 +1818,12 @@ static int smcr_serv_conf_first_link(struct smc_sock *smc) smc_llc_link_active(link); smcr_lgr_set_type(link->lgr, SMC_LGR_SINGLE); - /* initial contact - try to establish second link */ - smc_llc_srv_add_link(link); + if (link->lgr->max_links > 1) { + down_write(&link->lgr->llc_conf_lock); + /* initial contact - try to establish second link */ + smc_llc_srv_add_link(link, NULL); + up_write(&link->lgr->llc_conf_lock); + } return 0; } @@ -1649,6 +1833,10 @@ static void smc_listen_out(struct smc_sock *new_smc) struct smc_sock *lsmc = new_smc->listen_smc; struct sock *newsmcsk = &new_smc->sk; + if (tcp_sk(new_smc->clcsock->sk)->syn_smc) + atomic_dec(&lsmc->queued_smc_hs); + + release_sock(newsmcsk); /* lock in smc_listen_work() */ if (lsmc->sk.sk_state == SMC_LISTEN) { lock_sock_nested(&lsmc->sk, SINGLE_DEPTH_NESTING); smc_accept_enqueue(&lsmc->sk, newsmcsk); @@ -1691,10 +1879,7 @@ static void smc_listen_decline(struct smc_sock *new_smc, int reason_code, int local_first, u8 version) { /* RDMA setup failed, switch back to TCP */ - if (local_first) - smc_lgr_cleanup_early(&new_smc->conn); - else - smc_conn_free(&new_smc->conn); + smc_conn_abort(new_smc, local_first); if (reason_code < 0 || smc_switch_to_fallback(new_smc, reason_code)) { /* error, no fallback possible */ @@ -1717,36 +1902,57 @@ static int smc_listen_v2_check(struct smc_sock *new_smc, { struct smc_clc_smcd_v2_extension *pclc_smcd_v2_ext; struct smc_clc_v2_extension *pclc_v2_ext; + int rc = SMC_CLC_DECL_PEERNOSMC; ini->smc_type_v1 = pclc->hdr.typev1; ini->smc_type_v2 = pclc->hdr.typev2; - ini->smcd_version = ini->smc_type_v1 != SMC_TYPE_N ? SMC_V1 : 0; - if (pclc->hdr.version > SMC_V1) - ini->smcd_version |= - ini->smc_type_v2 != SMC_TYPE_N ? SMC_V2 : 0; - if (!smc_ism_is_v2_capable()) { - ini->smcd_version &= ~SMC_V2; + ini->smcd_version = smcd_indicated(ini->smc_type_v1) ? SMC_V1 : 0; + ini->smcr_version = smcr_indicated(ini->smc_type_v1) ? SMC_V1 : 0; + if (pclc->hdr.version > SMC_V1) { + if (smcd_indicated(ini->smc_type_v2)) + ini->smcd_version |= SMC_V2; + if (smcr_indicated(ini->smc_type_v2)) + ini->smcr_version |= SMC_V2; + } + if (!(ini->smcd_version & SMC_V2) && !(ini->smcr_version & SMC_V2)) { + rc = SMC_CLC_DECL_PEERNOSMC; goto out; } pclc_v2_ext = smc_get_clc_v2_ext(pclc); if (!pclc_v2_ext) { ini->smcd_version &= ~SMC_V2; + ini->smcr_version &= ~SMC_V2; + rc = SMC_CLC_DECL_NOV2EXT; goto out; } pclc_smcd_v2_ext = smc_get_clc_smcd_v2_ext(pclc_v2_ext); - if (!pclc_smcd_v2_ext) - ini->smcd_version &= ~SMC_V2; + if (ini->smcd_version & SMC_V2) { + if (!smc_ism_is_v2_capable()) { + ini->smcd_version &= ~SMC_V2; + rc = SMC_CLC_DECL_NOISM2SUPP; + } else if (!pclc_smcd_v2_ext) { + ini->smcd_version &= ~SMC_V2; + rc = SMC_CLC_DECL_NOV2DEXT; + } else if (!pclc_v2_ext->hdr.eid_cnt && + !pclc_v2_ext->hdr.flag.seid) { + ini->smcd_version &= ~SMC_V2; + rc = SMC_CLC_DECL_NOUEID; + } + } + if (ini->smcr_version & SMC_V2) { + if (!pclc_v2_ext->hdr.eid_cnt) { + ini->smcr_version &= ~SMC_V2; + rc = SMC_CLC_DECL_NOUEID; + } + } + + ini->release_nr = pclc_v2_ext->hdr.flag.release; + if (pclc_v2_ext->hdr.flag.release > SMC_RELEASE) + ini->release_nr = SMC_RELEASE; out: - if (!ini->smcd_version) { - if (pclc->hdr.typev1 == SMC_TYPE_B || - pclc->hdr.typev2 == SMC_TYPE_B) - return SMC_CLC_DECL_NOSMCDEV; - if (pclc->hdr.typev1 == SMC_TYPE_D || - pclc->hdr.typev2 == SMC_TYPE_D) - return SMC_CLC_DECL_NOSMCDDEV; - return SMC_CLC_DECL_NOSMCRDEV; - } + if (!ini->smcd_version && !ini->smcr_version) + return rc; return 0; } @@ -1781,8 +1987,10 @@ static int smc_listen_rdma_init(struct smc_sock *new_smc, return rc; /* create send buffer and rmb */ - if (smc_buf_create(new_smc, false)) + if (smc_buf_create(new_smc, false)) { + smc_conn_abort(new_smc, ini->first_contact_local); return SMC_CLC_DECL_MEM; + } return 0; } @@ -1800,10 +2008,7 @@ static int smc_listen_ism_init(struct smc_sock *new_smc, /* Create send and receive buffers */ rc = smc_buf_create(new_smc, true); if (rc) { - if (ini->first_contact_local) - smc_lgr_cleanup_early(&new_smc->conn); - else - smc_conn_free(&new_smc->conn); + smc_conn_abort(new_smc, ini->first_contact_local); return (rc == -ENOSPC) ? SMC_CLC_DECL_MAX_DMB : SMC_CLC_DECL_MEM; } @@ -1826,7 +2031,8 @@ static bool smc_is_already_selected(struct smcd_dev *smcd, /* check for ISM devices matching proposed ISM devices */ static void smc_check_ism_v2_match(struct smc_init_info *ini, - u16 proposed_chid, u64 proposed_gid, + u16 proposed_chid, + struct smcd_gid *proposed_gid, unsigned int *matches) { struct smcd_dev *smcd; @@ -1838,7 +2044,11 @@ static void smc_check_ism_v2_match(struct smc_init_info *ini, continue; if (smc_ism_get_chid(smcd) == proposed_chid && !smc_ism_cantalk(proposed_gid, ISM_RESERVED_VLANID, smcd)) { - ini->ism_peer_gid[*matches] = proposed_gid; + ini->ism_peer_gid[*matches].gid = proposed_gid->gid; + if (__smc_ism_is_emulated(proposed_chid)) + ini->ism_peer_gid[*matches].gid_ext = + proposed_gid->gid_ext; + /* non-Emulated-ISM's peer gid_ext remains 0. */ ini->ism_dev[*matches] = smcd; (*matches)++; break; @@ -1846,6 +2056,12 @@ static void smc_check_ism_v2_match(struct smc_init_info *ini, } } +static void smc_find_ism_store_rc(u32 rc, struct smc_init_info *ini) +{ + if (!ini->rc) + ini->rc = rc; +} + static void smc_find_ism_v2_device_serv(struct smc_sock *new_smc, struct smc_clc_msg_proposal *pclc, struct smc_init_info *ini) @@ -1854,9 +2070,11 @@ static void smc_find_ism_v2_device_serv(struct smc_sock *new_smc, struct smc_clc_v2_extension *smc_v2_ext; struct smc_clc_msg_smcd *pclc_smcd; unsigned int matches = 0; + struct smcd_gid smcd_gid; u8 smcd_version; u8 *eid = NULL; - int i; + int i, rc; + u16 chid; if (!(ini->smcd_version & SMC_V2) || !smcd_indicated(ini->smc_type_v2)) goto not_found; @@ -1864,47 +2082,69 @@ static void smc_find_ism_v2_device_serv(struct smc_sock *new_smc, pclc_smcd = smc_get_clc_msg_smcd(pclc); smc_v2_ext = smc_get_clc_v2_ext(pclc); smcd_v2_ext = smc_get_clc_smcd_v2_ext(smc_v2_ext); - if (!smcd_v2_ext || - !smc_v2_ext->hdr.flag.seid) /* no system EID support for SMCD */ + if (!pclc_smcd || !smc_v2_ext || !smcd_v2_ext) goto not_found; mutex_lock(&smcd_dev_list.mutex); - if (pclc_smcd->ism.chid) + if (pclc_smcd->ism.chid) { /* check for ISM device matching proposed native ISM device */ + smcd_gid.gid = ntohll(pclc_smcd->ism.gid); + smcd_gid.gid_ext = 0; smc_check_ism_v2_match(ini, ntohs(pclc_smcd->ism.chid), - ntohll(pclc_smcd->ism.gid), &matches); - for (i = 1; i <= smc_v2_ext->hdr.ism_gid_cnt; i++) { + &smcd_gid, &matches); + } + for (i = 0; i < smc_v2_ext->hdr.ism_gid_cnt; i++) { /* check for ISM devices matching proposed non-native ISM * devices */ - smc_check_ism_v2_match(ini, - ntohs(smcd_v2_ext->gidchid[i - 1].chid), - ntohll(smcd_v2_ext->gidchid[i - 1].gid), - &matches); + smcd_gid.gid = ntohll(smcd_v2_ext->gidchid[i].gid); + smcd_gid.gid_ext = 0; + chid = ntohs(smcd_v2_ext->gidchid[i].chid); + if (__smc_ism_is_emulated(chid)) { + if ((i + 1) == smc_v2_ext->hdr.ism_gid_cnt || + chid != ntohs(smcd_v2_ext->gidchid[i + 1].chid)) + /* each Emulated-ISM device takes two GID-CHID + * entries and CHID of the second entry repeats + * that of the first entry. + * + * So check if the next GID-CHID entry exists + * and both two entries' CHIDs are the same. + */ + continue; + smcd_gid.gid_ext = + ntohll(smcd_v2_ext->gidchid[++i].gid); + } + smc_check_ism_v2_match(ini, chid, &smcd_gid, &matches); } mutex_unlock(&smcd_dev_list.mutex); - if (ini->ism_dev[0]) { - smc_ism_get_system_eid(ini->ism_dev[0], &eid); - if (memcmp(eid, smcd_v2_ext->system_eid, SMC_MAX_EID_LEN)) - goto not_found; - } else { + if (!ini->ism_dev[0]) { + smc_find_ism_store_rc(SMC_CLC_DECL_NOSMCD2DEV, ini); goto not_found; } + smc_ism_get_system_eid(&eid); + if (!smc_clc_match_eid(ini->negotiated_eid, smc_v2_ext, + smcd_v2_ext->system_eid, eid)) + goto not_found; + /* separate - outside the smcd_dev_list.lock */ smcd_version = ini->smcd_version; for (i = 0; i < matches; i++) { ini->smcd_version = SMC_V2; ini->is_smcd = true; ini->ism_selected = i; - if (smc_listen_ism_init(new_smc, ini)) + rc = smc_listen_ism_init(new_smc, ini); + if (rc) { + smc_find_ism_store_rc(rc, ini); /* try next active ISM device */ continue; + } return; /* matching and usable V2 ISM device found */ } /* no V2 ISM device could be initialized */ ini->smcd_version = smcd_version; /* restore original value */ + ini->negotiated_eid[0] = 0; not_found: ini->smcd_version &= ~SMC_V2; @@ -1917,6 +2157,7 @@ static void smc_find_ism_v1_device_serv(struct smc_sock *new_smc, struct smc_init_info *ini) { struct smc_clc_msg_smcd *pclc_smcd = smc_get_clc_msg_smcd(pclc); + int rc = 0; /* check if ISM V1 is available */ if (!(ini->smcd_version & SMC_V1) || @@ -1924,14 +2165,19 @@ static void smc_find_ism_v1_device_serv(struct smc_sock *new_smc, !pclc_smcd) goto not_found; ini->is_smcd = true; /* prepare ISM check */ - ini->ism_peer_gid[0] = ntohll(pclc_smcd->ism.gid); - if (smc_find_ism_device(new_smc, ini)) + ini->ism_peer_gid[0].gid = ntohll(pclc_smcd->ism.gid); + ini->ism_peer_gid[0].gid_ext = 0; + rc = smc_find_ism_device(new_smc, ini); + if (rc) goto not_found; ini->ism_selected = 0; - if (!smc_listen_ism_init(new_smc, ini)) + rc = smc_listen_ism_init(new_smc, ini); + if (!rc) return; /* V1 ISM device found */ not_found: + smc_find_ism_store_rc(rc, ini); + ini->smcd_version &= ~SMC_V1; ini->ism_dev[0] = NULL; ini->is_smcd = false; } @@ -1955,24 +2201,74 @@ static int smc_listen_rdma_reg(struct smc_sock *new_smc, bool local_first) return 0; } +static void smc_find_rdma_v2_device_serv(struct smc_sock *new_smc, + struct smc_clc_msg_proposal *pclc, + struct smc_init_info *ini) +{ + struct smc_clc_v2_extension *smc_v2_ext; + u8 smcr_version; + int rc; + + if (!(ini->smcr_version & SMC_V2) || !smcr_indicated(ini->smc_type_v2)) + goto not_found; + + smc_v2_ext = smc_get_clc_v2_ext(pclc); + if (!smc_v2_ext || + !smc_clc_match_eid(ini->negotiated_eid, smc_v2_ext, NULL, NULL)) + goto not_found; + + /* prepare RDMA check */ + memcpy(ini->peer_systemid, pclc->lcl.id_for_peer, SMC_SYSTEMID_LEN); + memcpy(ini->peer_gid, smc_v2_ext->roce, SMC_GID_SIZE); + memcpy(ini->peer_mac, pclc->lcl.mac, ETH_ALEN); + ini->check_smcrv2 = true; + ini->smcrv2.clc_sk = new_smc->clcsock->sk; + ini->smcrv2.saddr = new_smc->clcsock->sk->sk_rcv_saddr; + ini->smcrv2.daddr = smc_ib_gid_to_ipv4(smc_v2_ext->roce); + rc = smc_find_rdma_device(new_smc, ini); + if (rc) { + smc_find_ism_store_rc(rc, ini); + goto not_found; + } + if (!ini->smcrv2.uses_gateway) + memcpy(ini->smcrv2.nexthop_mac, pclc->lcl.mac, ETH_ALEN); + + smcr_version = ini->smcr_version; + ini->smcr_version = SMC_V2; + rc = smc_listen_rdma_init(new_smc, ini); + if (!rc) { + rc = smc_listen_rdma_reg(new_smc, ini->first_contact_local); + if (rc) + smc_conn_abort(new_smc, ini->first_contact_local); + } + if (!rc) + return; + ini->smcr_version = smcr_version; + smc_find_ism_store_rc(rc, ini); + +not_found: + ini->smcr_version &= ~SMC_V2; + ini->smcrv2.ib_dev_v2 = NULL; + ini->check_smcrv2 = false; +} + static int smc_find_rdma_v1_device_serv(struct smc_sock *new_smc, struct smc_clc_msg_proposal *pclc, struct smc_init_info *ini) { int rc; - if (!smcr_indicated(ini->smc_type_v1)) + if (!(ini->smcr_version & SMC_V1) || !smcr_indicated(ini->smc_type_v1)) return SMC_CLC_DECL_NOSMCDEV; /* prepare RDMA check */ - ini->ib_lcl = &pclc->lcl; + memcpy(ini->peer_systemid, pclc->lcl.id_for_peer, SMC_SYSTEMID_LEN); + memcpy(ini->peer_gid, pclc->lcl.gid, SMC_GID_SIZE); + memcpy(ini->peer_mac, pclc->lcl.mac, ETH_ALEN); rc = smc_find_rdma_device(new_smc, ini); if (rc) { /* no RDMA device found */ - if (ini->smc_type_v1 == SMC_TYPE_B) - /* neither ISM nor RDMA device found */ - rc = SMC_CLC_DECL_NOSMCDEV; - return rc; + return SMC_CLC_DECL_NOSMCDEV; } rc = smc_listen_rdma_init(new_smc, ini); if (rc) @@ -1985,47 +2281,60 @@ static int smc_listen_find_device(struct smc_sock *new_smc, struct smc_clc_msg_proposal *pclc, struct smc_init_info *ini) { - int rc; + int prfx_rc; /* check for ISM device matching V2 proposed device */ smc_find_ism_v2_device_serv(new_smc, pclc, ini); if (ini->ism_dev[0]) return 0; - if (!(ini->smcd_version & SMC_V1)) - return SMC_CLC_DECL_NOSMCDEV; - - /* check for matching IP prefix and subnet length */ - rc = smc_listen_prfx_check(new_smc, pclc); - if (rc) - return rc; + /* check for matching IP prefix and subnet length (V1) */ + prfx_rc = smc_listen_prfx_check(new_smc, pclc); + if (prfx_rc) + smc_find_ism_store_rc(prfx_rc, ini); /* get vlan id from IP device */ if (smc_vlan_by_tcpsk(new_smc->clcsock, ini)) - return SMC_CLC_DECL_GETVLANERR; + return ini->rc ?: SMC_CLC_DECL_GETVLANERR; /* check for ISM device matching V1 proposed device */ - smc_find_ism_v1_device_serv(new_smc, pclc, ini); + if (!prfx_rc) + smc_find_ism_v1_device_serv(new_smc, pclc, ini); if (ini->ism_dev[0]) return 0; - if (pclc->hdr.typev1 == SMC_TYPE_D) - return SMC_CLC_DECL_NOSMCDDEV; /* skip RDMA and decline */ + if (!smcr_indicated(pclc->hdr.typev1) && + !smcr_indicated(pclc->hdr.typev2)) + /* skip RDMA and decline */ + return ini->rc ?: SMC_CLC_DECL_NOSMCDDEV; + + /* check if RDMA V2 is available */ + smc_find_rdma_v2_device_serv(new_smc, pclc, ini); + if (ini->smcrv2.ib_dev_v2) + return 0; + + /* check if RDMA V1 is available */ + if (!prfx_rc) { + int rc; - /* check if RDMA is available */ - return smc_find_rdma_v1_device_serv(new_smc, pclc, ini); + rc = smc_find_rdma_v1_device_serv(new_smc, pclc, ini); + smc_find_ism_store_rc(rc, ini); + return (!rc) ? 0 : ini->rc; + } + return prfx_rc; } /* listen worker: finish RDMA setup */ static int smc_listen_rdma_finish(struct smc_sock *new_smc, struct smc_clc_msg_accept_confirm *cclc, - bool local_first) + bool local_first, + struct smc_init_info *ini) { struct smc_link *link = new_smc->conn.lnk; int reason_code = 0; if (local_first) - smc_link_save_peer_info(link, cclc); + smc_link_save_peer_info(link, cclc, ini); if (smc_rmb_rtoken_handling(&new_smc->conn, link, cclc)) return SMC_CLC_DECL_ERR_RTOK; @@ -2046,14 +2355,16 @@ static void smc_listen_work(struct work_struct *work) { struct smc_sock *new_smc = container_of(work, struct smc_sock, smc_listen_work); - u8 version = smc_ism_is_v2_capable() ? SMC_V2 : SMC_V1; struct socket *newclcsock = new_smc->clcsock; struct smc_clc_msg_accept_confirm *cclc; struct smc_clc_msg_proposal_area *buf; struct smc_clc_msg_proposal *pclc; struct smc_init_info *ini = NULL; + u8 proposal_version = SMC_V1; + u8 accept_version; int rc = 0; + lock_sock(&new_smc->sk); /* release in smc_listen_out() */ if (new_smc->listen_smc->sk.sk_state != SMC_LISTEN) return smc_listen_out_err(new_smc); @@ -2085,7 +2396,9 @@ static void smc_listen_work(struct work_struct *work) SMC_CLC_PROPOSAL, CLC_WAIT_TIME); if (rc) goto out_decl; - version = pclc->hdr.version == SMC_V1 ? SMC_V1 : version; + + if (pclc->hdr.version > SMC_V1) + proposal_version = SMC_V2; /* IPSec connections opt out of SMC optimizations */ if (using_ipsec(new_smc)) { @@ -2104,6 +2417,10 @@ static void smc_listen_work(struct work_struct *work) if (rc) goto out_decl; + rc = smc_clc_srv_v2x_features_validate(pclc, ini); + if (rc) + goto out_decl; + mutex_lock(&smc_server_lgr_pending); smc_close_init(new_smc); smc_rx_init(new_smc); @@ -2115,8 +2432,9 @@ static void smc_listen_work(struct work_struct *work) goto out_unlock; /* send SMC Accept CLC message */ + accept_version = ini->is_smcd ? ini->smcd_version : ini->smcr_version; rc = smc_clc_send_accept(new_smc, ini->first_contact_local, - ini->smcd_version == SMC_V2 ? SMC_V2 : SMC_V1); + accept_version, ini->negotiated_eid, ini); if (rc) goto out_unlock; @@ -2135,15 +2453,35 @@ static void smc_listen_work(struct work_struct *work) goto out_decl; } + rc = smc_clc_v2x_features_confirm_check(cclc, ini); + if (rc) { + if (!ini->is_smcd) + goto out_unlock; + goto out_decl; + } + + /* fce smc release version is needed in smc_listen_rdma_finish, + * so save fce info here. + */ + smc_conn_save_peer_info_fce(new_smc, cclc); + /* finish worker */ if (!ini->is_smcd) { rc = smc_listen_rdma_finish(new_smc, cclc, - ini->first_contact_local); + ini->first_contact_local, ini); if (rc) goto out_unlock; mutex_unlock(&smc_server_lgr_pending); } smc_conn_save_peer_info(new_smc, cclc); + + if (ini->is_smcd && + smc_ism_support_dmb_nocopy(new_smc->conn.lgr->smcd)) { + rc = smcd_buf_attach(new_smc); + if (rc) + goto out_decl; + } + smc_listen_out_connected(new_smc); SMC_STAT_SERV_SUCC_INC(ini); goto out_free; @@ -2152,7 +2490,7 @@ static void smc_listen_work(struct work_struct *work) mutex_unlock(&smc_server_lgr_pending); out_decl: smc_listen_decline(new_smc, rc, ini ? ini->first_contact_local : 0, - version); + proposal_version); out_free: kfree(ini); kfree(buf); @@ -2175,6 +2513,9 @@ static void smc_tcp_listen_work(struct work_struct *work) if (!new_smc) continue; + if (tcp_sk(new_smc->clcsock->sk)->syn_smc) + atomic_inc(&lsmc->queued_smc_hs); + new_smc->listen_smc = lsmc; new_smc->use_fallback = lsmc->use_fallback; new_smc->fallback_rsn = lsmc->fallback_rsn; @@ -2247,6 +2588,14 @@ static int smc_listen(struct socket *sock, int backlog) smc_clcsock_data_ready, &smc->clcsk_data_ready); write_unlock_bh(&smc->clcsock->sk->sk_callback_lock); + /* save original ops */ + smc->ori_af_ops = inet_csk(smc->clcsock->sk)->icsk_af_ops; + + smc->af_ops = *smc->ori_af_ops; + smc->af_ops.syn_recv_sock = smc_tcp_syn_recv_sock; + + inet_csk(smc->clcsock->sk)->icsk_af_ops = &smc->af_ops; + rc = kernel_listen(smc->clcsock, backlog); if (rc) { write_lock_bh(&smc->clcsock->sk->sk_callback_lock); @@ -2326,7 +2675,7 @@ static int smc_accept(struct socket *sock, struct socket *new_sock, release_sock(clcsk); } else if (!atomic_read(&smc_sk(nsk)->conn.bytes_to_rcv)) { lock_sock(nsk); - smc_rx_wait(smc_sk(nsk), &timeo, smc_rx_data_available); + smc_rx_wait(smc_sk(nsk), &timeo, 0, smc_rx_data_available); release_sock(nsk); } } @@ -2471,6 +2820,13 @@ static __poll_t smc_poll(struct file *file, struct socket *sock, } else { sk_set_bit(SOCKWQ_ASYNC_NOSPACE, sk); set_bit(SOCK_NOSPACE, &sk->sk_socket->flags); + + if (sk->sk_state != SMC_INIT) { + /* Race breaker the same way as tcp_poll(). */ + smp_mb__after_atomic(); + if (atomic_read(&smc->conn.sndbuf_space)) + mask |= EPOLLOUT | EPOLLWRNORM; + } } if (atomic_read(&smc->conn.bytes_to_rcv)) mask |= EPOLLIN | EPOLLRDNORM; @@ -2750,34 +3106,6 @@ static int smc_ioctl(struct socket *sock, unsigned int cmd, return put_user(answ, (int __user *)arg); } -static ssize_t smc_sendpage(struct socket *sock, struct page *page, - int offset, size_t size, int flags) -{ - struct sock *sk = sock->sk; - struct smc_sock *smc; - int rc = -EPIPE; - - smc = smc_sk(sk); - lock_sock(sk); - if (sk->sk_state != SMC_ACTIVE) { - release_sock(sk); - goto out; - } - release_sock(sk); - if (smc->use_fallback) { - rc = kernel_sendpage(smc->clcsock, page, offset, - size, flags); - } else { - SMC_STAT_INC(!smc->conn.lnk, sendpage_cnt); - lock_sock(sk); - rc = smc_tx_sendpage(smc, page, offset, size, flags); - release_sock(sk); - } - -out: - return rc; -} - /* Map the affected portions of the rmbe into an spd, note the number of bytes * to splice in conn->splice_pending, and press 'go'. Delays consumer cursor * updates till whenever a respective page has been fully processed. @@ -2849,12 +3177,11 @@ static const struct proto_ops smc_sock_ops = { .sendmsg = smc_sendmsg, .recvmsg = smc_recvmsg, .mmap = sock_no_mmap, - .sendpage = smc_sendpage, .splice_read = smc_splice_read, }; -static int smc_create(struct net *net, struct socket *sock, int protocol, - int kern) +static int __smc_create(struct net *net, struct socket *sock, int protocol, + int kern, struct socket *clcsock) { int family = (protocol == SMCPROTO_SMC6) ? PF_INET6 : PF_INET; struct smc_sock *smc; @@ -2880,11 +3207,27 @@ static int smc_create(struct net *net, struct socket *sock, int protocol, smc = smc_sk(sk); smc->use_fallback = false; /* assume rdma capability first */ smc->fallback_rsn = 0; - rc = sock_create_kern(net, family, SOCK_STREAM, IPPROTO_TCP, - &smc->clcsock); - if (rc) { - sk_common_release(sk); - goto out; + + rc = 0; + if (!clcsock) { + rc = sock_create_kern(net, family, SOCK_STREAM, IPPROTO_TCP, + &smc->clcsock); + if (rc) { + sk_common_release(sk); + goto out; + } + + /* smc_clcsock_release() does not wait smc->clcsock->sk's + * destruction; its sk_state might not be TCP_CLOSE after + * smc->sk is close()d, and TCP timers can be fired later, + * which need net ref. + */ + sk = smc->clcsock->sk; + sk->sk_net_refcnt = 1; + get_net(net); + sock_inuse_add(net, 1); + } else { + smc->clcsock = clcsock; } @@ -2892,12 +3235,76 @@ static int smc_create(struct net *net, struct socket *sock, int protocol, return rc; } +static int smc_create(struct net *net, struct socket *sock, int protocol, + int kern) +{ + return __smc_create(net, sock, protocol, kern, NULL); +} + static const struct net_proto_family smc_sock_family_ops = { .family = PF_SMC, .owner = THIS_MODULE, .create = smc_create, }; +static int smc_ulp_init(struct sock *sk) +{ + struct socket *tcp = sk->sk_socket; + struct net *net = sock_net(sk); + struct socket *smcsock; + int protocol, ret; + + /* only TCP can be replaced */ + if (tcp->type != SOCK_STREAM || sk->sk_protocol != IPPROTO_TCP || + (sk->sk_family != AF_INET && sk->sk_family != AF_INET6)) + return -ESOCKTNOSUPPORT; + /* don't handle wq now */ + if (tcp->state != SS_UNCONNECTED || !tcp->file || tcp->wq.fasync_list) + return -ENOTCONN; + + if (sk->sk_family == AF_INET) + protocol = SMCPROTO_SMC; + else + protocol = SMCPROTO_SMC6; + + smcsock = sock_alloc(); + if (!smcsock) + return -ENFILE; + + smcsock->type = SOCK_STREAM; + __module_get(THIS_MODULE); /* tried in __tcp_ulp_find_autoload */ + ret = __smc_create(net, smcsock, protocol, 1, tcp); + if (ret) { + sock_release(smcsock); /* module_put() which ops won't be NULL */ + return ret; + } + + /* replace tcp socket to smc */ + smcsock->file = tcp->file; + smcsock->file->private_data = smcsock; + smcsock->file->f_inode = SOCK_INODE(smcsock); /* replace inode when sock_close */ + smcsock->file->f_path.dentry->d_inode = SOCK_INODE(smcsock); /* dput() in __fput */ + tcp->file = NULL; + + return ret; +} + +static void smc_ulp_clone(const struct request_sock *req, struct sock *newsk, + const gfp_t priority) +{ + struct inet_connection_sock *icsk = inet_csk(newsk); + + /* don't inherit ulp ops to child when listen */ + icsk->icsk_ulp_ops = NULL; +} + +static struct tcp_ulp_ops smc_ulp_ops __read_mostly = { + .name = "smc", + .owner = THIS_MODULE, + .init = smc_ulp_init, + .clone = smc_ulp_clone, +}; + unsigned int smc_net_id; static __net_init int smc_net_init(struct net *net) @@ -2931,12 +3338,14 @@ static int __init smc_init(void) if (rc) return rc; - smc_ism_init(); + rc = smc_ism_init(); + if (rc) + goto out_pernet_subsys; smc_clc_init(); rc = smc_nl_init(); if (rc) - goto out_pernet_subsys; + goto out_ism; rc = smc_pnet_init(); if (rc) @@ -3005,9 +3414,25 @@ static int __init smc_init(void) goto out_sock; } + rc = smc_loopback_init(); + if (rc) { + pr_err("%s: smc_loopback_init fails with %d\n", __func__, rc); + goto out_ib; + } + + rc = tcp_register_ulp(&smc_ulp_ops); + if (rc) { + pr_err("%s: tcp_ulp_register fails with %d\n", __func__, rc); + goto out_lo; + } + static_branch_enable(&tcp_have_smc); return 0; +out_lo: + smc_loopback_exit(); +out_ib: + smc_ib_unregister_client(); out_sock: sock_unregister(PF_SMC); out_proto6: @@ -3028,6 +3453,9 @@ static int __init smc_init(void) smc_pnet_exit(); out_nl: smc_nl_exit(); +out_ism: + smc_clc_exit(); + smc_ism_exit(); out_pernet_subsys: unregister_pernet_subsys(&smc_net_ops); @@ -3037,9 +3465,12 @@ static int __init smc_init(void) static void __exit smc_exit(void) { static_branch_disable(&tcp_have_smc); + tcp_unregister_ulp(&smc_ulp_ops); sock_unregister(PF_SMC); smc_core_exit(); + smc_loopback_exit(); smc_ib_unregister_client(); + smc_ism_exit(); destroy_workqueue(smc_close_wq); destroy_workqueue(smc_tcp_ls_wq); destroy_workqueue(smc_hs_wq); @@ -3048,6 +3479,7 @@ static void __exit smc_exit(void) proto_unregister(&smc_proto); smc_pnet_exit(); smc_nl_exit(); + smc_clc_exit(); unregister_pernet_subsys(&smc_net_ops); rcu_barrier(); } @@ -3059,3 +3491,4 @@ MODULE_AUTHOR("Ursula Braun "); MODULE_DESCRIPTION("smc socket address family"); MODULE_LICENSE("GPL"); MODULE_ALIAS_NETPROTO(PF_SMC); +MODULE_ALIAS_TCP_ULP("smc"); diff --git a/net/smc/smc.h b/net/smc/smc.h index 7c50514aaa7eb33e36faf73809ba8e8dc11f623e..ac5e239b7a412c844d7e05c3c36d9053705178de 100644 --- a/net/smc/smc.h +++ b/net/smc/smc.h @@ -20,14 +20,12 @@ #define SMC_V1 1 /* SMC version V1 */ #define SMC_V2 2 /* SMC version V2 */ -#define SMC_RELEASE 0 -#define SMC_MAX_ISM_DEVS 8 /* max # of proposed non-native ISM - * devices - */ -#define SMC_AUTOCORKING_DEFAULT_SIZE 0x10000 /* 64K by default */ -#define SMC_MAX_HOSTNAME_LEN 32 -#define SMC_MAX_EID_LEN 32 +#define SMC_RELEASE_0 0 +#define SMC_RELEASE_1 1 +#define SMC_RELEASE SMC_RELEASE_1 /* the latest release version */ + +#define SMC_AUTOCORKING_DEFAULT_SIZE 0x10000 /* 64K by default */ extern struct proto smc_proto; extern struct proto smc_proto6; @@ -53,10 +51,30 @@ enum smc_state { /* possible states of an SMC socket */ SMC_PROCESSABORT = 27, }; +enum smc_supplemental_features { + SMC_SPF_EMULATED_ISM_DEV = 0, +}; + +#define SMC_FEATURE_MASK \ + (BIT(SMC_SPF_EMULATED_ISM_DEV)) + struct smc_link_group; struct smc_wr_rx_hdr { /* common prefix part of LLC and CDC to demultiplex */ - u8 type; + union { + u8 type; +#if defined(__BIG_ENDIAN_BITFIELD) + struct { + u8 llc_version:4, + llc_type:4; + }; +#elif defined(__LITTLE_ENDIAN_BITFIELD) + struct { + u8 llc_type:4, + llc_version:4; + }; +#endif + }; } __aligned(1); struct smc_cdc_conn_state_flags { @@ -178,7 +196,6 @@ struct smc_connection { * - dec on polled tx cqe */ wait_queue_head_t cdc_pend_tx_wq; /* wakeup on no cdc_pend_tx_wr*/ - atomic_t tx_pushing; /* nr_threads trying tx push */ struct delayed_work tx_work; /* retry of smc_cdc_msg_send */ u32 tx_off; /* base offset in peer rmb */ @@ -219,7 +236,7 @@ struct smc_connection { */ u64 peer_token; /* SMC-D token of peer */ u8 killed : 1; /* abnormal termination */ - u8 freed : 1; /* normal termiation */ + u8 freed : 1; /* normal termination */ u8 out_of_sync : 1; /* out of sync with peer */ }; @@ -253,6 +270,10 @@ struct smc_sock { /* smc sock container */ bool use_fallback; /* fallback to tcp */ int fallback_rsn; /* reason for fallback */ u32 peer_diagnosis; /* decline reason from peer */ + atomic_t queued_smc_hs; /* queued smc handshakes */ + struct inet_connection_sock_af_ops af_ops; + const struct inet_connection_sock_af_ops *ori_af_ops; + /* original af ops */ int sockopt_defer_accept; /* sockopt TCP_DEFER_ACCEPT * value @@ -285,7 +306,7 @@ static inline void smc_init_saved_callbacks(struct smc_sock *smc) smc->clcsk_error_report = NULL; } -static inline struct smc_sock *smc_clcsock_user_data(struct sock *clcsk) +static inline struct smc_sock *smc_clcsock_user_data(const struct sock *clcsk) { return (struct smc_sock *) ((uintptr_t)clcsk->sk_user_data & ~SK_USER_DATA_NOCOPY); @@ -353,7 +374,17 @@ static inline bool using_ipsec(struct smc_sock *smc) } #endif +struct smc_gidlist; + struct sock *smc_accept_dequeue(struct sock *parent, struct socket *new_sock); void smc_close_non_accepted(struct sock *sk); +void smc_fill_gid_list(struct smc_link_group *lgr, + struct smc_gidlist *gidlist, + struct smc_ib_device *known_dev, u8 *known_gid); + +static inline void smc_sock_set_flag(struct sock *sk, enum sock_flags flag) +{ + set_bit(flag, &sk->sk_flags); +} #endif /* __SMC_H */ diff --git a/net/smc/smc_cdc.c b/net/smc/smc_cdc.c index 6cdfe7abffce4027df3cdd7715bbcad6cc591ead..eac1921836f40d197904abfd4f34131eee7c7663 100644 --- a/net/smc/smc_cdc.c +++ b/net/smc/smc_cdc.c @@ -18,6 +18,7 @@ #include "smc_tx.h" #include "smc_rx.h" #include "smc_close.h" +#include "smc_ism.h" /********************************** send *************************************/ @@ -258,6 +259,14 @@ int smcd_cdc_msg_send(struct smc_connection *conn) return rc; smc_curs_copy(&conn->rx_curs_confirmed, &curs, conn); conn->local_rx_ctrl.prod_flags.cons_curs_upd_req = 0; + + if (smc_ism_support_dmb_nocopy(conn->lgr->smcd)) + /* if local sndbuf shares the same memory region with + * peer DMB, then don't update the tx_curs_fin + * and sndbuf_space until peer has consumed the data. + */ + return 0; + /* Calculate transmitted data and increment free send buffer space */ diff = smc_curs_diff(conn->sndbuf_desc->len, &conn->tx_curs_fin, &conn->tx_curs_sent); @@ -269,7 +278,7 @@ int smcd_cdc_msg_send(struct smc_connection *conn) smc_curs_copy(&conn->tx_curs_fin, &conn->tx_curs_sent, conn); smc_tx_sndbuf_nonfull(smc); - return rc; + return 0; } /********************************* receive ***********************************/ @@ -326,7 +335,7 @@ static void smc_cdc_msg_recv_action(struct smc_sock *smc, { union smc_host_cursor cons_old, prod_old; struct smc_connection *conn = &smc->conn; - int diff_cons, diff_prod; + int diff_cons, diff_prod, diff_tx; smc_curs_copy(&prod_old, &conn->local_rx_ctrl.prod, conn); smc_curs_copy(&cons_old, &conn->local_rx_ctrl.cons, conn); @@ -342,6 +351,29 @@ static void smc_cdc_msg_recv_action(struct smc_sock *smc, atomic_add(diff_cons, &conn->peer_rmbe_space); /* guarantee 0 <= peer_rmbe_space <= peer_rmbe_size */ smp_mb__after_atomic(); + + /* if local sndbuf shares the same memory region with + * peer RMB, then update tx_curs_fin and sndbuf_space + * here since peer has already consumed the data. + */ + if (conn->lgr->is_smcd && + smc_ism_support_dmb_nocopy(conn->lgr->smcd)) { + /* Calculate consumed data and + * increment free send buffer space. + */ + diff_tx = smc_curs_diff(conn->sndbuf_desc->len, + &conn->tx_curs_fin, + &conn->local_rx_ctrl.cons); + /* increase local sndbuf space and fin_curs */ + smp_mb__before_atomic(); + atomic_add(diff_tx, &conn->sndbuf_space); + /* guarantee 0 <= sndbuf_space <= sndbuf_desc->len */ + smp_mb__after_atomic(); + smc_curs_copy(&conn->tx_curs_fin, + &conn->local_rx_ctrl.cons, conn); + + smc_tx_sndbuf_nonfull(smc); + } } diff_prod = smc_curs_diff(conn->rmb_desc->len, &prod_old, @@ -388,7 +420,7 @@ static void smc_cdc_msg_recv_action(struct smc_sock *smc, smc->sk.sk_shutdown |= RCV_SHUTDOWN; if (smc->clcsock && smc->clcsock->sk) smc->clcsock->sk->sk_shutdown |= RCV_SHUTDOWN; - sock_set_flag(&smc->sk, SOCK_DONE); + smc_sock_set_flag(&smc->sk, SOCK_DONE); sock_hold(&smc->sk); /* sock_put in close_work */ if (!queue_work(smc_close_wq, &conn->close_work)) sock_put(&smc->sk); @@ -411,9 +443,9 @@ static void smc_cdc_msg_recv(struct smc_sock *smc, struct smc_cdc_msg *cdc) * Context: * - tasklet context */ -static void smcd_cdc_rx_tsklet(unsigned long data) +static void smcd_cdc_rx_tsklet(struct tasklet_struct *t) { - struct smc_connection *conn = (struct smc_connection *)data; + struct smc_connection *conn = from_tasklet(conn, t, rx_tsklet); struct smcd_cdc_msg *data_cdc; struct smcd_cdc_msg cdc; struct smc_sock *smc; @@ -433,7 +465,7 @@ static void smcd_cdc_rx_tsklet(unsigned long data) */ void smcd_cdc_rx_init(struct smc_connection *conn) { - tasklet_init(&conn->rx_tsklet, smcd_cdc_rx_tsklet, (unsigned long)conn); + tasklet_setup(&conn->rx_tsklet, smcd_cdc_rx_tsklet); } /***************************** init, exit, misc ******************************/ diff --git a/net/smc/smc_clc.c b/net/smc/smc_clc.c index 614b48c490e23dfa95b8cc642ea4cd9754ab3ad1..b72d87d89ca6cc3e79b869d9b2a516262066a43c 100644 --- a/net/smc/smc_clc.c +++ b/net/smc/smc_clc.c @@ -26,10 +26,12 @@ #include "smc_clc.h" #include "smc_ib.h" #include "smc_ism.h" +#include "smc_netlink.h" #define SMCR_CLC_ACCEPT_CONFIRM_LEN 68 #define SMCD_CLC_ACCEPT_CONFIRM_LEN 48 #define SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 78 +#define SMCR_CLC_ACCEPT_CONFIRM_LEN_V2 108 #define SMC_CLC_RECV_BUF_LEN 100 /* eye catcher "SMCR" EBCDIC for CLC messages */ @@ -39,6 +41,307 @@ static const char SMCD_EYECATCHER[4] = {'\xe2', '\xd4', '\xc3', '\xc4'}; static u8 smc_hostname[SMC_MAX_HOSTNAME_LEN]; +struct smc_clc_eid_table { + rwlock_t lock; + struct list_head list; + u8 ueid_cnt; + u8 seid_enabled; +}; + +static struct smc_clc_eid_table smc_clc_eid_table; + +struct smc_clc_eid_entry { + struct list_head list; + u8 eid[SMC_MAX_EID_LEN]; +}; + +/* The size of a user EID is 32 characters. + * Valid characters should be (single-byte character set) A-Z, 0-9, '.' and '-'. + * Blanks should only be used to pad to the expected size. + * First character must be alphanumeric. + */ +static bool smc_clc_ueid_valid(char *ueid) +{ + char *end = ueid + SMC_MAX_EID_LEN; + + while (--end >= ueid && isspace(*end)) + ; + if (end < ueid) + return false; + if (!isalnum(*ueid) || islower(*ueid)) + return false; + while (ueid <= end) { + if ((!isalnum(*ueid) || islower(*ueid)) && *ueid != '.' && + *ueid != '-') + return false; + ueid++; + } + return true; +} + +static int smc_clc_ueid_add(char *ueid) +{ + struct smc_clc_eid_entry *new_ueid, *tmp_ueid; + int rc; + + if (!smc_clc_ueid_valid(ueid)) + return -EINVAL; + + /* add a new ueid entry to the ueid table if there isn't one */ + new_ueid = kzalloc(sizeof(*new_ueid), GFP_KERNEL); + if (!new_ueid) + return -ENOMEM; + memcpy(new_ueid->eid, ueid, SMC_MAX_EID_LEN); + + write_lock(&smc_clc_eid_table.lock); + if (smc_clc_eid_table.ueid_cnt >= SMC_MAX_UEID) { + rc = -ERANGE; + goto err_out; + } + list_for_each_entry(tmp_ueid, &smc_clc_eid_table.list, list) { + if (!memcmp(tmp_ueid->eid, ueid, SMC_MAX_EID_LEN)) { + rc = -EEXIST; + goto err_out; + } + } + list_add_tail(&new_ueid->list, &smc_clc_eid_table.list); + smc_clc_eid_table.ueid_cnt++; + write_unlock(&smc_clc_eid_table.lock); + return 0; + +err_out: + write_unlock(&smc_clc_eid_table.lock); + kfree(new_ueid); + return rc; +} + +int smc_clc_ueid_count(void) +{ + int count; + + read_lock(&smc_clc_eid_table.lock); + count = smc_clc_eid_table.ueid_cnt; + read_unlock(&smc_clc_eid_table.lock); + + return count; +} + +int smc_nl_add_ueid(struct sk_buff *skb, struct genl_info *info) +{ + struct nlattr *nla_ueid = info->attrs[SMC_NLA_EID_TABLE_ENTRY]; + char *ueid; + + if (!nla_ueid || nla_len(nla_ueid) != SMC_MAX_EID_LEN + 1) + return -EINVAL; + ueid = (char *)nla_data(nla_ueid); + + return smc_clc_ueid_add(ueid); +} + +/* remove one or all ueid entries from the table */ +static int smc_clc_ueid_remove(char *ueid) +{ + struct smc_clc_eid_entry *lst_ueid, *tmp_ueid; + int rc = -ENOENT; + + /* remove table entry */ + write_lock(&smc_clc_eid_table.lock); + list_for_each_entry_safe(lst_ueid, tmp_ueid, &smc_clc_eid_table.list, + list) { + if (!ueid || !memcmp(lst_ueid->eid, ueid, SMC_MAX_EID_LEN)) { + list_del(&lst_ueid->list); + smc_clc_eid_table.ueid_cnt--; + kfree(lst_ueid); + rc = 0; + } + } +#if IS_ENABLED(CONFIG_S390) + if (!rc && !smc_clc_eid_table.ueid_cnt) { + smc_clc_eid_table.seid_enabled = 1; + rc = -EAGAIN; /* indicate success and enabling of seid */ + } +#endif + write_unlock(&smc_clc_eid_table.lock); + return rc; +} + +int smc_nl_remove_ueid(struct sk_buff *skb, struct genl_info *info) +{ + struct nlattr *nla_ueid = info->attrs[SMC_NLA_EID_TABLE_ENTRY]; + char *ueid; + + if (!nla_ueid || nla_len(nla_ueid) != SMC_MAX_EID_LEN + 1) + return -EINVAL; + ueid = (char *)nla_data(nla_ueid); + + return smc_clc_ueid_remove(ueid); +} + +int smc_nl_flush_ueid(struct sk_buff *skb, struct genl_info *info) +{ + smc_clc_ueid_remove(NULL); + return 0; +} + +static int smc_nl_ueid_dumpinfo(struct sk_buff *skb, u32 portid, u32 seq, + u32 flags, char *ueid) +{ + char ueid_str[SMC_MAX_EID_LEN + 1]; + void *hdr; + + hdr = genlmsg_put(skb, portid, seq, &smc_gen_nl_family, + flags, SMC_NETLINK_DUMP_UEID); + if (!hdr) + return -ENOMEM; + snprintf(ueid_str, sizeof(ueid_str), "%s", ueid); + if (nla_put_string(skb, SMC_NLA_EID_TABLE_ENTRY, ueid_str)) { + genlmsg_cancel(skb, hdr); + return -EMSGSIZE; + } + genlmsg_end(skb, hdr); + return 0; +} + +static int _smc_nl_ueid_dump(struct sk_buff *skb, u32 portid, u32 seq, + int start_idx) +{ + struct smc_clc_eid_entry *lst_ueid; + int idx = 0; + + read_lock(&smc_clc_eid_table.lock); + list_for_each_entry(lst_ueid, &smc_clc_eid_table.list, list) { + if (idx++ < start_idx) + continue; + if (smc_nl_ueid_dumpinfo(skb, portid, seq, NLM_F_MULTI, + lst_ueid->eid)) { + --idx; + break; + } + } + read_unlock(&smc_clc_eid_table.lock); + return idx; +} + +int smc_nl_dump_ueid(struct sk_buff *skb, struct netlink_callback *cb) +{ + struct smc_nl_dmp_ctx *cb_ctx = smc_nl_dmp_ctx(cb); + int idx; + + idx = _smc_nl_ueid_dump(skb, NETLINK_CB(cb->skb).portid, + cb->nlh->nlmsg_seq, cb_ctx->pos[0]); + + cb_ctx->pos[0] = idx; + return skb->len; +} + +int smc_nl_dump_seid(struct sk_buff *skb, struct netlink_callback *cb) +{ + struct smc_nl_dmp_ctx *cb_ctx = smc_nl_dmp_ctx(cb); + char seid_str[SMC_MAX_EID_LEN + 1]; + u8 seid_enabled; + void *hdr; + u8 *seid; + + if (cb_ctx->pos[0]) + return skb->len; + + hdr = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, + &smc_gen_nl_family, NLM_F_MULTI, + SMC_NETLINK_DUMP_SEID); + if (!hdr) + return -ENOMEM; + if (!smc_ism_is_v2_capable()) + goto end; + + smc_ism_get_system_eid(&seid); + snprintf(seid_str, sizeof(seid_str), "%s", seid); + if (nla_put_string(skb, SMC_NLA_SEID_ENTRY, seid_str)) + goto err; + read_lock(&smc_clc_eid_table.lock); + seid_enabled = smc_clc_eid_table.seid_enabled; + read_unlock(&smc_clc_eid_table.lock); + if (nla_put_u8(skb, SMC_NLA_SEID_ENABLED, seid_enabled)) + goto err; +end: + genlmsg_end(skb, hdr); + cb_ctx->pos[0]++; + return skb->len; +err: + genlmsg_cancel(skb, hdr); + return -EMSGSIZE; +} + +int smc_nl_enable_seid(struct sk_buff *skb, struct genl_info *info) +{ +#if IS_ENABLED(CONFIG_S390) + write_lock(&smc_clc_eid_table.lock); + smc_clc_eid_table.seid_enabled = 1; + write_unlock(&smc_clc_eid_table.lock); + return 0; +#else + return -EOPNOTSUPP; +#endif +} + +int smc_nl_disable_seid(struct sk_buff *skb, struct genl_info *info) +{ + int rc = 0; + +#if IS_ENABLED(CONFIG_S390) + write_lock(&smc_clc_eid_table.lock); + if (!smc_clc_eid_table.ueid_cnt) + rc = -ENOENT; + else + smc_clc_eid_table.seid_enabled = 0; + write_unlock(&smc_clc_eid_table.lock); +#else + rc = -EOPNOTSUPP; +#endif + return rc; +} + +static bool _smc_clc_match_ueid(u8 *peer_ueid) +{ + struct smc_clc_eid_entry *tmp_ueid; + + list_for_each_entry(tmp_ueid, &smc_clc_eid_table.list, list) { + if (!memcmp(tmp_ueid->eid, peer_ueid, SMC_MAX_EID_LEN)) + return true; + } + return false; +} + +bool smc_clc_match_eid(u8 *negotiated_eid, + struct smc_clc_v2_extension *smc_v2_ext, + u8 *peer_eid, u8 *local_eid) +{ + bool match = false; + int i; + + negotiated_eid[0] = 0; + read_lock(&smc_clc_eid_table.lock); + if (peer_eid && local_eid && + smc_clc_eid_table.seid_enabled && + smc_v2_ext->hdr.flag.seid && + !memcmp(peer_eid, local_eid, SMC_MAX_EID_LEN)) { + memcpy(negotiated_eid, peer_eid, SMC_MAX_EID_LEN); + match = true; + goto out; + } + + for (i = 0; i < smc_v2_ext->hdr.eid_cnt; i++) { + if (_smc_clc_match_ueid(smc_v2_ext->user_eids[i])) { + memcpy(negotiated_eid, smc_v2_ext->user_eids[i], + SMC_MAX_EID_LEN); + match = true; + goto out; + } + } +out: + read_unlock(&smc_clc_eid_table.lock); + return match; +} + /* check arriving CLC proposal */ static bool smc_clc_msg_prop_valid(struct smc_clc_msg_proposal *pclc) { @@ -47,7 +350,6 @@ static bool smc_clc_msg_prop_valid(struct smc_clc_msg_proposal *pclc) struct smc_clc_msg_hdr *hdr = &pclc->hdr; struct smc_clc_v2_extension *v2_ext; - v2_ext = smc_get_clc_v2_ext(pclc); pclc_prfx = smc_clc_proposal_get_prefix(pclc); if (!pclc_prfx || pclc_prfx->ipv6_prefixes_cnt > SMC_CLC_MAX_V6_PREFIX) @@ -64,6 +366,13 @@ static bool smc_clc_msg_prop_valid(struct smc_clc_msg_proposal *pclc) sizeof(struct smc_clc_msg_trail)) return false; } else { + v2_ext = smc_get_clc_v2_ext(pclc); + if ((hdr->typev2 != SMC_TYPE_N && + (!v2_ext || v2_ext->hdr.eid_cnt > SMC_CLC_MAX_UEID)) || + (smcd_indicated(hdr->typev2) && + v2_ext->hdr.ism_gid_cnt > SMCD_CLC_MAX_V2_GID_ENTRIES)) + return false; + if (ntohs(hdr->length) != sizeof(*pclc) + sizeof(struct smc_clc_msg_smcd) + @@ -86,9 +395,9 @@ static bool smc_clc_msg_prop_valid(struct smc_clc_msg_proposal *pclc) /* check arriving CLC accept or confirm */ static bool -smc_clc_msg_acc_conf_valid(struct smc_clc_msg_accept_confirm_v2 *clc_v2) +smc_clc_msg_acc_conf_valid(struct smc_clc_msg_accept_confirm *clc) { - struct smc_clc_msg_hdr *hdr = &clc_v2->hdr; + struct smc_clc_msg_hdr *hdr = &clc->hdr; if (hdr->typev1 != SMC_TYPE_R && hdr->typev1 != SMC_TYPE_D) return false; @@ -100,21 +409,58 @@ smc_clc_msg_acc_conf_valid(struct smc_clc_msg_accept_confirm_v2 *clc_v2) return false; } else { if (hdr->typev1 == SMC_TYPE_D && - ntohs(hdr->length) != SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 && - (ntohs(hdr->length) != SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 + - sizeof(struct smc_clc_first_contact_ext))) + ntohs(hdr->length) < SMCD_CLC_ACCEPT_CONFIRM_LEN_V2) + return false; + if (hdr->typev1 == SMC_TYPE_R && + ntohs(hdr->length) < SMCR_CLC_ACCEPT_CONFIRM_LEN_V2) + return false; + } + return true; +} + +/* check arriving CLC decline */ +static bool +smc_clc_msg_decl_valid(struct smc_clc_msg_decline *dclc) +{ + struct smc_clc_msg_hdr *hdr = &dclc->hdr; + + if (hdr->typev1 != SMC_TYPE_R && hdr->typev1 != SMC_TYPE_D) + return false; + if (hdr->version == SMC_V1) { + if (ntohs(hdr->length) != sizeof(struct smc_clc_msg_decline)) + return false; + } else { + if (ntohs(hdr->length) != sizeof(struct smc_clc_msg_decline_v2)) return false; } return true; } -static void smc_clc_fill_fce(struct smc_clc_first_contact_ext *fce, int *len) +static int smc_clc_fill_fce_v2x(struct smc_clc_first_contact_ext_v2x *fce_v2x, + struct smc_init_info *ini) { - memset(fce, 0, sizeof(*fce)); - fce->os_type = SMC_CLC_OS_LINUX; - fce->release = SMC_RELEASE; - memcpy(fce->hostname, smc_hostname, sizeof(smc_hostname)); - (*len) += sizeof(*fce); + int ret = sizeof(*fce_v2x); + + memset(fce_v2x, 0, sizeof(*fce_v2x)); + fce_v2x->fce_v2_base.os_type = SMC_CLC_OS_LINUX; + fce_v2x->fce_v2_base.release = ini->release_nr; + memcpy(fce_v2x->fce_v2_base.hostname, + smc_hostname, sizeof(smc_hostname)); + if (ini->is_smcd && ini->release_nr < SMC_RELEASE_1) { + ret = sizeof(struct smc_clc_first_contact_ext); + goto out; + } + + if (ini->release_nr >= SMC_RELEASE_1) { + if (!ini->is_smcd) { + fce_v2x->max_conns = ini->max_conns; + fce_v2x->max_links = ini->max_links; + } + fce_v2x->feature_mask = htons(ini->feature_mask); + } + +out: + return ret; } /* check if received message has a correct header length and contains valid @@ -122,7 +468,7 @@ static void smc_clc_fill_fce(struct smc_clc_first_contact_ext *fce, int *len) */ static bool smc_clc_msg_hdr_valid(struct smc_clc_msg_hdr *clcm, bool check_trl) { - struct smc_clc_msg_accept_confirm_v2 *clc_v2; + struct smc_clc_msg_accept_confirm *clc; struct smc_clc_msg_proposal *pclc; struct smc_clc_msg_decline *dclc; struct smc_clc_msg_trail *trl; @@ -140,18 +486,17 @@ static bool smc_clc_msg_hdr_valid(struct smc_clc_msg_hdr *clcm, bool check_trl) break; case SMC_CLC_ACCEPT: case SMC_CLC_CONFIRM: - clc_v2 = (struct smc_clc_msg_accept_confirm_v2 *)clcm; - if (!smc_clc_msg_acc_conf_valid(clc_v2)) + clc = (struct smc_clc_msg_accept_confirm *)clcm; + if (!smc_clc_msg_acc_conf_valid(clc)) return false; trl = (struct smc_clc_msg_trail *) - ((u8 *)clc_v2 + ntohs(clc_v2->hdr.length) - - sizeof(*trl)); + ((u8 *)clc + ntohs(clc->hdr.length) - sizeof(*trl)); break; case SMC_CLC_DECLINE: dclc = (struct smc_clc_msg_decline *)clcm; - if (ntohs(dclc->hdr.length) != sizeof(*dclc)) + if (!smc_clc_msg_decl_valid(dclc)) return false; - trl = &dclc->trl; + check_trl = false; break; default: return false; @@ -427,6 +772,11 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen, SMC_CLC_RECV_BUF_LEN : datlen; iov_iter_kvec(&msg.msg_iter, READ, &vec, 1, recvlen); len = sock_recvmsg(smc->clcsock, &msg, krflags); + if (len < recvlen) { + smc->sk.sk_err = EPROTO; + reason_code = -EPROTO; + goto out; + } datlen -= len; } if (clcm->type == SMC_CLC_DECLINE) { @@ -450,15 +800,16 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen, /* send CLC DECLINE message across internal TCP socket */ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info, u8 version) { - struct smc_clc_msg_decline dclc; + struct smc_clc_msg_decline *dclc_v1; + struct smc_clc_msg_decline_v2 dclc; struct msghdr msg; + int len, send_len; struct kvec vec; - int len; + dclc_v1 = (struct smc_clc_msg_decline *)&dclc; memset(&dclc, 0, sizeof(dclc)); memcpy(dclc.hdr.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER)); dclc.hdr.type = SMC_CLC_DECLINE; - dclc.hdr.length = htons(sizeof(struct smc_clc_msg_decline)); dclc.hdr.version = version; dclc.os_type = version == SMC_V1 ? 0 : SMC_CLC_OS_LINUX; dclc.hdr.typev2 = (peer_diag_info == SMC_CLC_DECL_SYNCERR) ? @@ -468,14 +819,22 @@ int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info, u8 version) memcpy(dclc.id_for_peer, local_systemid, sizeof(local_systemid)); dclc.peer_diagnosis = htonl(peer_diag_info); - memcpy(dclc.trl.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER)); + if (version == SMC_V1) { + memcpy(dclc_v1->trl.eyecatcher, SMC_EYECATCHER, + sizeof(SMC_EYECATCHER)); + send_len = sizeof(*dclc_v1); + } else { + memcpy(dclc.trl.eyecatcher, SMC_EYECATCHER, + sizeof(SMC_EYECATCHER)); + send_len = sizeof(dclc); + } + dclc.hdr.length = htons(send_len); memset(&msg, 0, sizeof(msg)); vec.iov_base = &dclc; - vec.iov_len = sizeof(struct smc_clc_msg_decline); - len = kernel_sendmsg(smc->clcsock, &msg, &vec, 1, - sizeof(struct smc_clc_msg_decline)); - if (len < 0 || len < sizeof(struct smc_clc_msg_decline)) + vec.iov_len = send_len; + len = kernel_sendmsg(smc->clcsock, &msg, &vec, 1, send_len); + if (len < 0 || len < send_len) len = -EPROTO; return len > 0 ? 0 : len; } @@ -492,6 +851,7 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) struct smc_clc_v2_extension *v2_ext; struct smc_clc_msg_smcd *pclc_smcd; struct smc_clc_msg_trail *trl; + struct smcd_dev *smcd; int len, i, plen, rc; int reason_code = 0; struct kvec vec[8]; @@ -505,8 +865,10 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) pclc_smcd = &pclc->pclc_smcd; pclc_prfx = &pclc->pclc_prfx; ipv6_prfx = pclc->pclc_prfx_ipv6; - v2_ext = &pclc->pclc_v2_ext; - smcd_v2_ext = &pclc->pclc_smcd_v2_ext; + v2_ext = container_of(&pclc->pclc_v2_ext, + struct smc_clc_v2_extension, fixed); + smcd_v2_ext = container_of(&pclc->pclc_smcd_v2_ext, + struct smc_clc_smcd_v2_extension, fixed); gidchids = pclc->pclc_gidchids; trl = &pclc->pclc_trl; @@ -545,9 +907,13 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) ETH_ALEN); } if (smcd_indicated(ini->smc_type_v1)) { + struct smcd_gid smcd_gid; + /* add SMC-D specifics */ if (ini->ism_dev[0]) { - pclc_smcd->ism.gid = htonll(ini->ism_dev[0]->local_gid); + smcd = ini->ism_dev[0]; + smcd->ops->get_local_gid(smcd, &smcd_gid); + pclc_smcd->ism.gid = htonll(smcd_gid.gid); pclc_smcd->ism.chid = htons(smc_ism_get_chid(ini->ism_dev[0])); } @@ -555,9 +921,10 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) if (ini->smc_type_v2 == SMC_TYPE_N) { pclc_smcd->v2_ext_offset = 0; } else { + struct smc_clc_eid_entry *ueident; u16 v2_ext_offset; - u8 *eid = NULL; + v2_ext->hdr.flag.release = SMC_RELEASE; v2_ext_offset = sizeof(*pclc_smcd) - offsetofend(struct smc_clc_msg_smcd, v2_ext_offset); if (ini->smc_type_v1 != SMC_TYPE_N) @@ -565,32 +932,63 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) pclc_prfx->ipv6_prefixes_cnt * sizeof(ipv6_prfx[0]); pclc_smcd->v2_ext_offset = htons(v2_ext_offset); - v2_ext->hdr.eid_cnt = 0; - v2_ext->hdr.ism_gid_cnt = ini->ism_offered_cnt; - v2_ext->hdr.flag.release = SMC_RELEASE; - v2_ext->hdr.flag.seid = 1; + plen += sizeof(*v2_ext); + + v2_ext->feature_mask = htons(SMC_FEATURE_MASK); + read_lock(&smc_clc_eid_table.lock); + v2_ext->hdr.eid_cnt = smc_clc_eid_table.ueid_cnt; + plen += smc_clc_eid_table.ueid_cnt * SMC_MAX_EID_LEN; + i = 0; + list_for_each_entry(ueident, &smc_clc_eid_table.list, list) { + memcpy(v2_ext->user_eids[i++], ueident->eid, + sizeof(ueident->eid)); + } + read_unlock(&smc_clc_eid_table.lock); + } + if (smcd_indicated(ini->smc_type_v2)) { + struct smcd_gid smcd_gid; + u8 *eid = NULL; + int entry = 0; + + v2_ext->hdr.flag.seid = smc_clc_eid_table.seid_enabled; v2_ext->hdr.smcd_v2_ext_offset = htons(sizeof(*v2_ext) - offsetofend(struct smc_clnt_opts_area_hdr, smcd_v2_ext_offset) + v2_ext->hdr.eid_cnt * SMC_MAX_EID_LEN); - if (ini->ism_dev[0]) - smc_ism_get_system_eid(ini->ism_dev[0], &eid); - else - smc_ism_get_system_eid(ini->ism_dev[1], &eid); - if (eid) + smc_ism_get_system_eid(&eid); + if (eid && v2_ext->hdr.flag.seid) memcpy(smcd_v2_ext->system_eid, eid, SMC_MAX_EID_LEN); - plen += sizeof(*v2_ext) + sizeof(*smcd_v2_ext); + plen += sizeof(*smcd_v2_ext); if (ini->ism_offered_cnt) { for (i = 1; i <= ini->ism_offered_cnt; i++) { - gidchids[i - 1].gid = - htonll(ini->ism_dev[i]->local_gid); - gidchids[i - 1].chid = + smcd = ini->ism_dev[i]; + smcd->ops->get_local_gid(smcd, &smcd_gid); + gidchids[entry].chid = htons(smc_ism_get_chid(ini->ism_dev[i])); + gidchids[entry].gid = htonll(smcd_gid.gid); + if (smc_ism_is_emulated(smcd)) { + /* an Emulated-ISM device takes two + * entries. CHID of the second entry + * repeats that of the first entry. + */ + gidchids[entry + 1].chid = + gidchids[entry].chid; + gidchids[entry + 1].gid = + htonll(smcd_gid.gid_ext); + entry++; + } + entry++; } - plen += ini->ism_offered_cnt * - sizeof(struct smc_clc_smcd_gid_chid); + plen += entry * sizeof(struct smc_clc_smcd_gid_chid); } + v2_ext->hdr.ism_gid_cnt = entry; + } + if (smcr_indicated(ini->smc_type_v2)) { + memcpy(v2_ext->roce, ini->smcrv2.ib_gid_v2, SMC_GID_SIZE); + v2_ext->max_conns = SMC_CONN_PER_LGR_PREFER; + v2_ext->max_links = SMC_LINKS_PER_LGR_MAX_PREFER; } + pclc_base->hdr.length = htons(plen); memcpy(trl->eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER)); @@ -612,13 +1010,16 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) } if (ini->smc_type_v2 != SMC_TYPE_N) { vec[i].iov_base = v2_ext; - vec[i++].iov_len = sizeof(*v2_ext); - vec[i].iov_base = smcd_v2_ext; - vec[i++].iov_len = sizeof(*smcd_v2_ext); - if (ini->ism_offered_cnt) { - vec[i].iov_base = gidchids; - vec[i++].iov_len = ini->ism_offered_cnt * + vec[i++].iov_len = sizeof(*v2_ext) + + (v2_ext->hdr.eid_cnt * SMC_MAX_EID_LEN); + if (smcd_indicated(ini->smc_type_v2)) { + vec[i].iov_base = smcd_v2_ext; + vec[i++].iov_len = sizeof(*smcd_v2_ext); + if (ini->ism_offered_cnt) { + vec[i].iov_base = gidchids; + vec[i++].iov_len = v2_ext->hdr.ism_gid_cnt * sizeof(struct smc_clc_smcd_gid_chid); + } } } vec[i].iov_base = trl; @@ -637,99 +1038,165 @@ int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini) return reason_code; } +static void +smcd_clc_prep_confirm_accept(struct smc_connection *conn, + struct smc_clc_msg_accept_confirm *clc, + int first_contact, u8 version, + u8 *eid, struct smc_init_info *ini, + int *fce_len, + struct smc_clc_first_contact_ext_v2x *fce_v2x, + struct smc_clc_msg_trail *trl) +{ + struct smcd_dev *smcd = conn->lgr->smcd; + struct smcd_gid smcd_gid; + u16 chid; + int len; + + /* SMC-D specific settings */ + memcpy(clc->hdr.eyecatcher, SMCD_EYECATCHER, + sizeof(SMCD_EYECATCHER)); + smcd->ops->get_local_gid(smcd, &smcd_gid); + clc->hdr.typev1 = SMC_TYPE_D; + clc->d0.gid = htonll(smcd_gid.gid); + clc->d0.token = htonll(conn->rmb_desc->token); + clc->d0.dmbe_size = conn->rmbe_size_comp; + clc->d0.dmbe_idx = 0; + memcpy(&clc->d0.linkid, conn->lgr->id, SMC_LGR_ID_SIZE); + if (version == SMC_V1) { + clc->hdr.length = htons(SMCD_CLC_ACCEPT_CONFIRM_LEN); + } else { + chid = smc_ism_get_chid(smcd); + clc->d1.chid = htons(chid); + if (eid && eid[0]) + memcpy(clc->d1.eid, eid, SMC_MAX_EID_LEN); + if (__smc_ism_is_emulated(chid)) + clc->d1.gid_ext = htonll(smcd_gid.gid_ext); + len = SMCD_CLC_ACCEPT_CONFIRM_LEN_V2; + if (first_contact) { + *fce_len = smc_clc_fill_fce_v2x(fce_v2x, ini); + len += *fce_len; + } + clc->hdr.length = htons(len); + } + memcpy(trl->eyecatcher, SMCD_EYECATCHER, + sizeof(SMCD_EYECATCHER)); +} + +static void +smcr_clc_prep_confirm_accept(struct smc_connection *conn, + struct smc_clc_msg_accept_confirm *clc, + int first_contact, u8 version, + u8 *eid, struct smc_init_info *ini, + int *fce_len, + struct smc_clc_first_contact_ext_v2x *fce_v2x, + struct smc_clc_fce_gid_ext *gle, + struct smc_clc_msg_trail *trl) +{ + struct smc_link *link = conn->lnk; + int len; + + /* SMC-R specific settings */ + memcpy(clc->hdr.eyecatcher, SMC_EYECATCHER, + sizeof(SMC_EYECATCHER)); + clc->hdr.typev1 = SMC_TYPE_R; + memcpy(clc->r0.lcl.id_for_peer, local_systemid, + sizeof(local_systemid)); + memcpy(&clc->r0.lcl.gid, link->gid, SMC_GID_SIZE); + memcpy(&clc->r0.lcl.mac, &link->smcibdev->mac[link->ibport - 1], + ETH_ALEN); + hton24(clc->r0.qpn, link->roce_qp->qp_num); + clc->r0.rmb_rkey = + htonl(conn->rmb_desc->mr[link->link_idx]->rkey); + clc->r0.rmbe_idx = 1; /* for now: 1 RMB = 1 RMBE */ + clc->r0.rmbe_alert_token = htonl(conn->alert_token_local); + switch (clc->hdr.type) { + case SMC_CLC_ACCEPT: + clc->r0.qp_mtu = link->path_mtu; + break; + case SMC_CLC_CONFIRM: + clc->r0.qp_mtu = min(link->path_mtu, link->peer_mtu); + break; + } + clc->r0.rmbe_size = conn->rmbe_size_comp; + clc->r0.rmb_dma_addr = conn->rmb_desc->is_vm ? + cpu_to_be64((uintptr_t)conn->rmb_desc->cpu_addr) : + cpu_to_be64((u64)sg_dma_address + (conn->rmb_desc->sgt[link->link_idx].sgl)); + hton24(clc->r0.psn, link->psn_initial); + if (version == SMC_V1) { + clc->hdr.length = htons(SMCR_CLC_ACCEPT_CONFIRM_LEN); + } else { + if (eid && eid[0]) + memcpy(clc->r1.eid, eid, SMC_MAX_EID_LEN); + len = SMCR_CLC_ACCEPT_CONFIRM_LEN_V2; + if (first_contact) { + *fce_len = smc_clc_fill_fce_v2x(fce_v2x, ini); + len += *fce_len; + fce_v2x->fce_v2_base.v2_direct = + !link->lgr->uses_gateway; + if (clc->hdr.type == SMC_CLC_CONFIRM) { + memset(gle, 0, sizeof(*gle)); + gle->gid_cnt = ini->smcrv2.gidlist.len; + len += sizeof(*gle); + len += gle->gid_cnt * sizeof(gle->gid[0]); + } + } + clc->hdr.length = htons(len); + } + memcpy(trl->eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER)); +} + /* build and send CLC CONFIRM / ACCEPT message */ static int smc_clc_send_confirm_accept(struct smc_sock *smc, - struct smc_clc_msg_accept_confirm_v2 *clc_v2, - int first_contact, u8 version) + struct smc_clc_msg_accept_confirm *clc, + int first_contact, u8 version, + u8 *eid, struct smc_init_info *ini) { + struct smc_clc_first_contact_ext_v2x fce_v2x; struct smc_connection *conn = &smc->conn; - struct smc_clc_msg_accept_confirm *clc; - struct smc_clc_first_contact_ext fce; + struct smc_clc_fce_gid_ext gle; struct smc_clc_msg_trail trl; - struct kvec vec[3]; + int i, fce_len; + struct kvec vec[5]; struct msghdr msg; - int i, len; /* send SMC Confirm CLC msg */ - clc = (struct smc_clc_msg_accept_confirm *)clc_v2; clc->hdr.version = version; /* SMC version */ if (first_contact) clc->hdr.typev2 |= SMC_FIRST_CONTACT_MASK; - if (conn->lgr->is_smcd) { - /* SMC-D specific settings */ - memcpy(clc->hdr.eyecatcher, SMCD_EYECATCHER, - sizeof(SMCD_EYECATCHER)); - clc->hdr.typev1 = SMC_TYPE_D; - clc->d0.gid = conn->lgr->smcd->local_gid; - clc->d0.token = conn->rmb_desc->token; - clc->d0.dmbe_size = conn->rmbe_size_comp; - clc->d0.dmbe_idx = 0; - memcpy(&clc->d0.linkid, conn->lgr->id, SMC_LGR_ID_SIZE); - if (version == SMC_V1) { - clc->hdr.length = htons(SMCD_CLC_ACCEPT_CONFIRM_LEN); - } else { - u8 *eid = NULL; - - clc_v2->chid = htons(smc_ism_get_chid(conn->lgr->smcd)); - smc_ism_get_system_eid(conn->lgr->smcd, &eid); - if (eid) - memcpy(clc_v2->eid, eid, SMC_MAX_EID_LEN); - len = SMCD_CLC_ACCEPT_CONFIRM_LEN_V2; - if (first_contact) - smc_clc_fill_fce(&fce, &len); - clc_v2->hdr.length = htons(len); - } - memcpy(trl.eyecatcher, SMCD_EYECATCHER, - sizeof(SMCD_EYECATCHER)); - } else { - struct smc_link *link = conn->lnk; - - /* SMC-R specific settings */ - link = conn->lnk; - memcpy(clc->hdr.eyecatcher, SMC_EYECATCHER, - sizeof(SMC_EYECATCHER)); - clc->hdr.typev1 = SMC_TYPE_R; - clc->hdr.length = htons(SMCR_CLC_ACCEPT_CONFIRM_LEN); - memcpy(clc->r0.lcl.id_for_peer, local_systemid, - sizeof(local_systemid)); - memcpy(&clc->r0.lcl.gid, link->gid, SMC_GID_SIZE); - memcpy(&clc->r0.lcl.mac, &link->smcibdev->mac[link->ibport - 1], - ETH_ALEN); - hton24(clc->r0.qpn, link->roce_qp->qp_num); - clc->r0.rmb_rkey = - htonl(conn->rmb_desc->mr[link->link_idx]->rkey); - clc->r0.rmbe_idx = 1; /* for now: 1 RMB = 1 RMBE */ - clc->r0.rmbe_alert_token = htonl(conn->alert_token_local); - switch (clc->hdr.type) { - case SMC_CLC_ACCEPT: - clc->r0.qp_mtu = link->path_mtu; - break; - case SMC_CLC_CONFIRM: - clc->r0.qp_mtu = min(link->path_mtu, link->peer_mtu); - break; - } - clc->r0.rmbe_size = conn->rmbe_size_comp; - clc->r0.rmb_dma_addr = conn->rmb_desc->is_vm ? - cpu_to_be64((uintptr_t)conn->rmb_desc->cpu_addr) : - cpu_to_be64((u64)sg_dma_address - (conn->rmb_desc->sgt[link->link_idx].sgl)); - hton24(clc->r0.psn, link->psn_initial); - memcpy(trl.eyecatcher, SMC_EYECATCHER, sizeof(SMC_EYECATCHER)); - } - + if (conn->lgr->is_smcd) + smcd_clc_prep_confirm_accept(conn, clc, first_contact, + version, eid, ini, &fce_len, + &fce_v2x, &trl); + else + smcr_clc_prep_confirm_accept(conn, clc, first_contact, + version, eid, ini, &fce_len, + &fce_v2x, &gle, &trl); memset(&msg, 0, sizeof(msg)); i = 0; - vec[i].iov_base = clc_v2; + vec[i].iov_base = clc; if (version > SMC_V1) - vec[i++].iov_len = SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 - sizeof(trl); + vec[i++].iov_len = (clc->hdr.typev1 == SMC_TYPE_D ? + SMCD_CLC_ACCEPT_CONFIRM_LEN_V2 : + SMCR_CLC_ACCEPT_CONFIRM_LEN_V2) - + sizeof(trl); else vec[i++].iov_len = (clc->hdr.typev1 == SMC_TYPE_D ? SMCD_CLC_ACCEPT_CONFIRM_LEN : SMCR_CLC_ACCEPT_CONFIRM_LEN) - sizeof(trl); if (version > SMC_V1 && first_contact) { - vec[i].iov_base = &fce; - vec[i++].iov_len = sizeof(fce); + vec[i].iov_base = &fce_v2x; + vec[i++].iov_len = fce_len; + if (!conn->lgr->is_smcd) { + if (clc->hdr.type == SMC_CLC_CONFIRM) { + vec[i].iov_base = &gle; + vec[i++].iov_len = sizeof(gle); + vec[i].iov_base = &ini->smcrv2.gidlist.list; + vec[i++].iov_len = gle.gid_cnt * + sizeof(gle.gid[0]); + } + } } vec[i].iov_base = &trl; vec[i++].iov_len = sizeof(trl); @@ -739,18 +1206,18 @@ static int smc_clc_send_confirm_accept(struct smc_sock *smc, /* send CLC CONFIRM message across internal TCP socket */ int smc_clc_send_confirm(struct smc_sock *smc, bool clnt_first_contact, - u8 version) + u8 version, u8 *eid, struct smc_init_info *ini) { - struct smc_clc_msg_accept_confirm_v2 cclc_v2; + struct smc_clc_msg_accept_confirm cclc; int reason_code = 0; int len; /* send SMC Confirm CLC msg */ - memset(&cclc_v2, 0, sizeof(cclc_v2)); - cclc_v2.hdr.type = SMC_CLC_CONFIRM; - len = smc_clc_send_confirm_accept(smc, &cclc_v2, clnt_first_contact, - version); - if (len < ntohs(cclc_v2.hdr.length)) { + memset(&cclc, 0, sizeof(cclc)); + cclc.hdr.type = SMC_CLC_CONFIRM; + len = smc_clc_send_confirm_accept(smc, &cclc, clnt_first_contact, + version, eid, ini); + if (len < ntohs(cclc.hdr.length)) { if (len >= 0) { reason_code = -ENETUNREACH; smc->sk.sk_err = -reason_code; @@ -764,21 +1231,106 @@ int smc_clc_send_confirm(struct smc_sock *smc, bool clnt_first_contact, /* send CLC ACCEPT message across internal TCP socket */ int smc_clc_send_accept(struct smc_sock *new_smc, bool srv_first_contact, - u8 version) + u8 version, u8 *negotiated_eid, struct smc_init_info *ini) { - struct smc_clc_msg_accept_confirm_v2 aclc_v2; + struct smc_clc_msg_accept_confirm aclc; int len; - memset(&aclc_v2, 0, sizeof(aclc_v2)); - aclc_v2.hdr.type = SMC_CLC_ACCEPT; - len = smc_clc_send_confirm_accept(new_smc, &aclc_v2, srv_first_contact, - version); - if (len < ntohs(aclc_v2.hdr.length)) + memset(&aclc, 0, sizeof(aclc)); + aclc.hdr.type = SMC_CLC_ACCEPT; + len = smc_clc_send_confirm_accept(new_smc, &aclc, srv_first_contact, + version, negotiated_eid, ini); + if (len < ntohs(aclc.hdr.length)) len = len >= 0 ? -EPROTO : -new_smc->clcsock->sk->sk_err; return len > 0 ? 0 : len; } +int smc_clc_srv_v2x_features_validate(struct smc_clc_msg_proposal *pclc, + struct smc_init_info *ini) +{ + struct smc_clc_v2_extension *pclc_v2_ext; + + ini->max_conns = SMC_CONN_PER_LGR_MAX; + ini->max_links = SMC_LINKS_ADD_LNK_MAX; + ini->feature_mask = SMC_FEATURE_MASK; + + if ((!(ini->smcd_version & SMC_V2) && !(ini->smcr_version & SMC_V2)) || + ini->release_nr < SMC_RELEASE_1) + return 0; + + pclc_v2_ext = smc_get_clc_v2_ext(pclc); + if (!pclc_v2_ext) + return SMC_CLC_DECL_NOV2EXT; + + if (ini->smcr_version & SMC_V2) { + ini->max_conns = min_t(u8, pclc_v2_ext->max_conns, SMC_CONN_PER_LGR_PREFER); + if (ini->max_conns < SMC_CONN_PER_LGR_MIN) + return SMC_CLC_DECL_MAXCONNERR; + + ini->max_links = min_t(u8, pclc_v2_ext->max_links, SMC_LINKS_PER_LGR_MAX_PREFER); + if (ini->max_links < SMC_LINKS_ADD_LNK_MIN) + return SMC_CLC_DECL_MAXLINKERR; + } + + return 0; +} + +int smc_clc_clnt_v2x_features_validate(struct smc_clc_first_contact_ext *fce, + struct smc_init_info *ini) +{ + struct smc_clc_first_contact_ext_v2x *fce_v2x = + (struct smc_clc_first_contact_ext_v2x *)fce; + + if (ini->release_nr < SMC_RELEASE_1) + return 0; + + if (!ini->is_smcd) { + if (fce_v2x->max_conns < SMC_CONN_PER_LGR_MIN) + return SMC_CLC_DECL_MAXCONNERR; + ini->max_conns = fce_v2x->max_conns; + + if (fce_v2x->max_links > SMC_LINKS_ADD_LNK_MAX || + fce_v2x->max_links < SMC_LINKS_ADD_LNK_MIN) + return SMC_CLC_DECL_MAXLINKERR; + ini->max_links = fce_v2x->max_links; + } + /* common supplemental features of server and client */ + ini->feature_mask = ntohs(fce_v2x->feature_mask) & SMC_FEATURE_MASK; + + return 0; +} + +int smc_clc_v2x_features_confirm_check(struct smc_clc_msg_accept_confirm *cclc, + struct smc_init_info *ini) +{ + struct smc_clc_first_contact_ext *fce = + smc_get_clc_first_contact_ext(cclc, ini->is_smcd); + struct smc_clc_first_contact_ext_v2x *fce_v2x = + (struct smc_clc_first_contact_ext_v2x *)fce; + + if (cclc->hdr.version == SMC_V1 || + !(cclc->hdr.typev2 & SMC_FIRST_CONTACT_MASK)) + return 0; + + if (ini->release_nr != fce->release) + return SMC_CLC_DECL_RELEASEERR; + + if (fce->release < SMC_RELEASE_1) + return 0; + + if (!ini->is_smcd) { + if (fce_v2x->max_conns != ini->max_conns) + return SMC_CLC_DECL_MAXCONNERR; + if (fce_v2x->max_links != ini->max_links) + return SMC_CLC_DECL_MAXLINKERR; + } + /* common supplemental features returned by client */ + ini->feature_mask = ntohs(fce_v2x->feature_mask); + + return 0; +} + void smc_clc_get_hostname(u8 **host) { *host = &smc_hostname[0]; @@ -792,4 +1344,18 @@ void __init smc_clc_init(void) u = utsname(); memcpy(smc_hostname, u->nodename, min_t(size_t, strlen(u->nodename), sizeof(smc_hostname))); + + INIT_LIST_HEAD(&smc_clc_eid_table.list); + rwlock_init(&smc_clc_eid_table.lock); + smc_clc_eid_table.ueid_cnt = 0; +#if IS_ENABLED(CONFIG_S390) + smc_clc_eid_table.seid_enabled = 1; +#else + smc_clc_eid_table.seid_enabled = 0; +#endif +} + +void smc_clc_exit(void) +{ + smc_clc_ueid_remove(NULL); } diff --git a/net/smc/smc_clc.h b/net/smc/smc_clc.h index 4b68eb626aa253bd77574b6bd5aec36762eafd88..94cb20b8952858bfa4b19402b5bafdbd6e129764 100644 --- a/net/smc/smc_clc.h +++ b/net/smc/smc_clc.h @@ -14,8 +14,10 @@ #define _SMC_CLC_H #include +#include #include "smc.h" +#include "smc_netlink.h" #define SMC_CLC_PROPOSAL 0x01 #define SMC_CLC_ACCEPT 0x02 @@ -37,6 +39,15 @@ #define SMC_CLC_DECL_NOSMCDEV 0x03030000 /* no SMC device found (R or D) */ #define SMC_CLC_DECL_NOSMCDDEV 0x03030001 /* no SMC-D device found */ #define SMC_CLC_DECL_NOSMCRDEV 0x03030002 /* no SMC-R device found */ +#define SMC_CLC_DECL_NOISM2SUPP 0x03030003 /* hardware has no ISMv2 support */ +#define SMC_CLC_DECL_NOV2EXT 0x03030004 /* peer sent no clc v2 extension */ +#define SMC_CLC_DECL_NOV2DEXT 0x03030005 /* peer sent no clc SMC-Dv2 ext. */ +#define SMC_CLC_DECL_NOSEID 0x03030006 /* peer sent no SEID */ +#define SMC_CLC_DECL_NOSMCD2DEV 0x03030007 /* no SMC-Dv2 device found */ +#define SMC_CLC_DECL_NOUEID 0x03030008 /* peer sent no UEID */ +#define SMC_CLC_DECL_RELEASEERR 0x03030009 /* release version negotiate failed */ +#define SMC_CLC_DECL_MAXCONNERR 0x0303000a /* max connections negotiate failed */ +#define SMC_CLC_DECL_MAXLINKERR 0x0303000b /* max links negotiate failed */ #define SMC_CLC_DECL_MODEUNSUPP 0x03040000 /* smc modes do not match (R or D)*/ #define SMC_CLC_DECL_RMBE_EC 0x03050000 /* peer has eyecatcher in RMBE */ #define SMC_CLC_DECL_OPTUNSUPP 0x03060000 /* fastopen sockopt not supported */ @@ -47,6 +58,8 @@ #define SMC_CLC_DECL_NOSRVLINK 0x030b0000 /* SMC-R link from srv not found */ #define SMC_CLC_DECL_VERSMISMAT 0x030c0000 /* SMC version mismatch */ #define SMC_CLC_DECL_MAX_DMB 0x030d0000 /* SMC-D DMB limit exceeded */ +#define SMC_CLC_DECL_NOROUTE 0x030e0000 /* SMC-Rv2 conn. no route to peer */ +#define SMC_CLC_DECL_NOINDIRECT 0x030f0000 /* SMC-Rv2 conn. indirect mismatch*/ #define SMC_CLC_DECL_SYNCERR 0x04000000 /* synchronization error */ #define SMC_CLC_DECL_PEERDECL 0x05000000 /* peer declined during handshake */ #define SMC_CLC_DECL_INTERR 0x09990000 /* internal error */ @@ -121,9 +134,15 @@ struct smc_clc_smcd_gid_chid { */ struct smc_clc_v2_extension { - struct smc_clnt_opts_area_hdr hdr; - u8 roce[16]; /* RoCEv2 GID */ - u8 reserved[16]; + /* New members must be added within the struct_group() macro below. */ + struct_group_tagged(smc_clc_v2_extension_fixed, fixed, + struct smc_clnt_opts_area_hdr hdr; + u8 roce[16]; /* RoCEv2 GID */ + u8 max_conns; + u8 max_links; + __be16 feature_mask; + u8 reserved[12]; + ); u8 user_eids[][SMC_MAX_EID_LEN]; }; @@ -135,14 +154,19 @@ struct smc_clc_msg_proposal_prefix { /* prefix part of clc proposal message*/ } __aligned(4); struct smc_clc_msg_smcd { /* SMC-D GID information */ - struct smc_clc_smcd_gid_chid ism; /* ISM native GID+CHID of requestor */ + struct smc_clc_smcd_gid_chid ism; /* ISM native GID+CHID of requester */ __be16 v2_ext_offset; /* SMC Version 2 Extension Offset */ - u8 reserved[28]; + u8 vendor_oui[3]; /* vendor organizationally unique identifier */ + u8 vendor_exp_options[5]; + u8 reserved[20]; }; struct smc_clc_smcd_v2_extension { - u8 system_eid[SMC_MAX_EID_LEN]; - u8 reserved[16]; + /* New members must be added within the struct_group() macro below. */ + struct_group_tagged(smc_clc_smcd_v2_extension_fixed, fixed, + u8 system_eid[SMC_MAX_EID_LEN]; + u8 reserved[16]; + ); struct smc_clc_smcd_gid_chid gidchid[]; }; @@ -153,15 +177,23 @@ struct smc_clc_msg_proposal { /* clc proposal message sent by Linux */ } __aligned(4); #define SMC_CLC_MAX_V6_PREFIX 8 +#define SMC_CLC_MAX_UEID 8 +#define SMCD_CLC_MAX_V2_GID_ENTRIES 8 /* max # of CHID-GID entries in CLC + * proposal SMC-Dv2 extension. + * each ISM device takes one entry and + * each Emulated-ISM takes two entries + */ struct smc_clc_msg_proposal_area { struct smc_clc_msg_proposal pclc_base; struct smc_clc_msg_smcd pclc_smcd; struct smc_clc_msg_proposal_prefix pclc_prfx; struct smc_clc_ipv6_prefix pclc_prfx_ipv6[SMC_CLC_MAX_V6_PREFIX]; - struct smc_clc_v2_extension pclc_v2_ext; - struct smc_clc_smcd_v2_extension pclc_smcd_v2_ext; - struct smc_clc_smcd_gid_chid pclc_gidchids[SMC_MAX_ISM_DEVS]; + struct smc_clc_v2_extension_fixed pclc_v2_ext; + u8 user_eids[SMC_CLC_MAX_UEID][SMC_MAX_EID_LEN]; + struct smc_clc_smcd_v2_extension_fixed pclc_smcd_v2_ext; + struct smc_clc_smcd_gid_chid + pclc_gidchids[SMCD_CLC_MAX_V2_GID_ENTRIES]; struct smc_clc_msg_trail pclc_trl; }; @@ -185,8 +217,8 @@ struct smcr_clc_msg_accept_confirm { /* SMCR accept/confirm */ } __packed; struct smcd_clc_msg_accept_confirm_common { /* SMCD accept/confirm */ - u64 gid; /* Sender GID */ - u64 token; /* DMB token */ + __be64 gid; /* Sender GID */ + __be64 token; /* DMB token */ u8 dmbe_idx; /* DMBE index */ #if defined(__BIG_ENDIAN_BITFIELD) u8 dmbe_size : 4, /* buf size (compressed) */ @@ -204,11 +236,14 @@ struct smcd_clc_msg_accept_confirm_common { /* SMCD accept/confirm */ #define SMC_CLC_OS_AIX 3 struct smc_clc_first_contact_ext { - u8 reserved1; #if defined(__BIG_ENDIAN_BITFIELD) + u8 v2_direct : 1, + reserved : 7; u8 os_type : 4, release : 4; #elif defined(__LITTLE_ENDIAN_BITFIELD) + u8 reserved : 7, + v2_direct : 1; u8 release : 4, os_type : 4; #endif @@ -216,26 +251,46 @@ struct smc_clc_first_contact_ext { u8 hostname[SMC_MAX_HOSTNAME_LEN]; }; -struct smc_clc_msg_accept_confirm { /* clc accept / confirm message */ - struct smc_clc_msg_hdr hdr; +struct smc_clc_first_contact_ext_v2x { + struct smc_clc_first_contact_ext fce_v2_base; union { - struct smcr_clc_msg_accept_confirm r0; /* SMC-R */ - struct { /* SMC-D */ - struct smcd_clc_msg_accept_confirm_common d0; - u32 reserved5[3]; + struct { + u8 max_conns; /* for SMC-R only */ + u8 max_links; /* for SMC-R only */ }; + u8 reserved3[2]; /* for SMC-D only */ }; -} __packed; /* format defined in RFC7609 */ + __be16 feature_mask; + __be32 vendor_exp_options; + u8 reserved4[8]; +} __packed; /* format defined in + * IBM Shared Memory Communications Version 2 (Third Edition) + * (https://www.ibm.com/support/pages/node/7009315) + */ + +struct smc_clc_fce_gid_ext { + u8 gid_cnt; + u8 reserved2[3]; + u8 gid[][SMC_GID_SIZE]; +}; -struct smc_clc_msg_accept_confirm_v2 { /* clc accept / confirm message */ +struct smc_clc_msg_accept_confirm { /* clc accept / confirm message */ struct smc_clc_msg_hdr hdr; union { - struct smcr_clc_msg_accept_confirm r0; /* SMC-R */ + struct { /* SMC-R */ + struct smcr_clc_msg_accept_confirm r0; + struct { /* v2 only */ + u8 eid[SMC_MAX_EID_LEN]; + u8 reserved6[8]; + } __packed r1; + }; struct { /* SMC-D */ struct smcd_clc_msg_accept_confirm_common d0; - __be16 chid; - u8 eid[SMC_MAX_EID_LEN]; - u8 reserved5[8]; + struct { /* v2 only, but 12 bytes reserved in v1 */ + __be16 chid; + u8 eid[SMC_MAX_EID_LEN]; + __be64 gid_ext; + } __packed d1; }; }; }; @@ -255,6 +310,24 @@ struct smc_clc_msg_decline { /* clc decline message */ struct smc_clc_msg_trail trl; /* eye catcher "SMCD" or "SMCR" EBCDIC */ } __aligned(4); +#define SMC_DECL_DIAG_COUNT_V2 4 /* no. of additional peer diagnosis codes */ + +struct smc_clc_msg_decline_v2 { /* clc decline message */ + struct smc_clc_msg_hdr hdr; + u8 id_for_peer[SMC_SYSTEMID_LEN]; /* sender peer_id */ + __be32 peer_diagnosis; /* diagnosis information */ +#if defined(__BIG_ENDIAN_BITFIELD) + u8 os_type : 4, + reserved : 4; +#elif defined(__LITTLE_ENDIAN_BITFIELD) + u8 reserved : 4, + os_type : 4; +#endif + u8 reserved2[3]; + __be32 peer_diagnosis_v2[SMC_DECL_DIAG_COUNT_V2]; + struct smc_clc_msg_trail trl; /* eye catcher "SMCD" or "SMCR" EBCDIC */ +} __aligned(4); + /* determine start of the prefix area within the proposal message */ static inline struct smc_clc_msg_proposal_prefix * smc_clc_proposal_get_prefix(struct smc_clc_msg_proposal *pclc) @@ -277,6 +350,17 @@ static inline bool smcd_indicated(int smc_type) return smc_type == SMC_TYPE_D || smc_type == SMC_TYPE_B; } +static inline u8 smc_indicated_type(int is_smcd, int is_smcr) +{ + if (is_smcd && is_smcr) + return SMC_TYPE_B; + if (is_smcd) + return SMC_TYPE_D; + if (is_smcr) + return SMC_TYPE_R; + return SMC_TYPE_N; +} + /* get SMC-D info from proposal message */ static inline struct smc_clc_msg_smcd * smc_get_clc_msg_smcd(struct smc_clc_msg_proposal *prop) @@ -292,8 +376,14 @@ static inline struct smc_clc_v2_extension * smc_get_clc_v2_ext(struct smc_clc_msg_proposal *prop) { struct smc_clc_msg_smcd *prop_smcd = smc_get_clc_msg_smcd(prop); + u16 max_offset; + + max_offset = offsetof(struct smc_clc_msg_proposal_area, pclc_v2_ext) - + offsetof(struct smc_clc_msg_proposal_area, pclc_smcd) - + offsetofend(struct smc_clc_msg_smcd, v2_ext_offset); - if (!prop_smcd || !ntohs(prop_smcd->v2_ext_offset)) + if (!prop_smcd || !ntohs(prop_smcd->v2_ext_offset) || + ntohs(prop_smcd->v2_ext_offset) > max_offset) return NULL; return (struct smc_clc_v2_extension *) @@ -306,9 +396,15 @@ smc_get_clc_v2_ext(struct smc_clc_msg_proposal *prop) static inline struct smc_clc_smcd_v2_extension * smc_get_clc_smcd_v2_ext(struct smc_clc_v2_extension *prop_v2ext) { + u16 max_offset = offsetof(struct smc_clc_msg_proposal_area, pclc_smcd_v2_ext) - + offsetof(struct smc_clc_msg_proposal_area, pclc_v2_ext) - + offsetof(struct smc_clc_v2_extension, hdr) - + offsetofend(struct smc_clnt_opts_area_hdr, smcd_v2_ext_offset); + if (!prop_v2ext) return NULL; - if (!ntohs(prop_v2ext->hdr.smcd_v2_ext_offset)) + if (!ntohs(prop_v2ext->hdr.smcd_v2_ext_offset) || + ntohs(prop_v2ext->hdr.smcd_v2_ext_offset) > max_offset) return NULL; return (struct smc_clc_smcd_v2_extension *) @@ -319,6 +415,26 @@ smc_get_clc_smcd_v2_ext(struct smc_clc_v2_extension *prop_v2ext) ntohs(prop_v2ext->hdr.smcd_v2_ext_offset)); } +static inline struct smc_clc_first_contact_ext * +smc_get_clc_first_contact_ext(struct smc_clc_msg_accept_confirm *clc, + bool is_smcd) +{ + int clc_v2_len; + + if (clc->hdr.version == SMC_V1 || + !(clc->hdr.typev2 & SMC_FIRST_CONTACT_MASK)) + return NULL; + + if (is_smcd) + clc_v2_len = + offsetofend(struct smc_clc_msg_accept_confirm, d1); + else + clc_v2_len = + offsetofend(struct smc_clc_msg_accept_confirm, r1); + + return (struct smc_clc_first_contact_ext *)(((u8 *)clc) + clc_v2_len); +} + struct smcd_dev; struct smc_init_info; @@ -329,10 +445,28 @@ int smc_clc_wait_msg(struct smc_sock *smc, void *buf, int buflen, int smc_clc_send_decline(struct smc_sock *smc, u32 peer_diag_info, u8 version); int smc_clc_send_proposal(struct smc_sock *smc, struct smc_init_info *ini); int smc_clc_send_confirm(struct smc_sock *smc, bool clnt_first_contact, - u8 version); + u8 version, u8 *eid, struct smc_init_info *ini); int smc_clc_send_accept(struct smc_sock *smc, bool srv_first_contact, - u8 version); + u8 version, u8 *negotiated_eid, struct smc_init_info *ini); +int smc_clc_srv_v2x_features_validate(struct smc_clc_msg_proposal *pclc, + struct smc_init_info *ini); +int smc_clc_clnt_v2x_features_validate(struct smc_clc_first_contact_ext *fce, + struct smc_init_info *ini); +int smc_clc_v2x_features_confirm_check(struct smc_clc_msg_accept_confirm *cclc, + struct smc_init_info *ini); void smc_clc_init(void) __init; +void smc_clc_exit(void); void smc_clc_get_hostname(u8 **host); +bool smc_clc_match_eid(u8 *negotiated_eid, + struct smc_clc_v2_extension *smc_v2_ext, + u8 *peer_eid, u8 *local_eid); +int smc_clc_ueid_count(void); +int smc_nl_dump_ueid(struct sk_buff *skb, struct netlink_callback *cb); +int smc_nl_add_ueid(struct sk_buff *skb, struct genl_info *info); +int smc_nl_remove_ueid(struct sk_buff *skb, struct genl_info *info); +int smc_nl_flush_ueid(struct sk_buff *skb, struct genl_info *info); +int smc_nl_dump_seid(struct sk_buff *skb, struct netlink_callback *cb); +int smc_nl_enable_seid(struct sk_buff *skb, struct genl_info *info); +int smc_nl_disable_seid(struct sk_buff *skb, struct genl_info *info); #endif diff --git a/net/smc/smc_close.c b/net/smc/smc_close.c index ed5d582cafd8a0e57b28c541a17e9da0a09a1c81..24679800375b3e97371318649e0d1355a044a568 100644 --- a/net/smc/smc_close.c +++ b/net/smc/smc_close.c @@ -174,7 +174,7 @@ void smc_close_active_abort(struct smc_sock *smc) break; } - sock_set_flag(sk, SOCK_DEAD); + smc_sock_set_flag(sk, SOCK_DEAD); sk->sk_state_change(sk); if (release_clcsock) { @@ -373,9 +373,9 @@ static void smc_close_passive_work(struct work_struct *work) if (rxflags->peer_conn_abort) { /* peer has not received all data */ smc_close_passive_abort_received(smc); - release_sock(&smc->sk); + release_sock(sk); cancel_delayed_work_sync(&conn->tx_work); - lock_sock(&smc->sk); + lock_sock(sk); goto wakeup; } diff --git a/net/smc/smc_core.c b/net/smc/smc_core.c index 5312ca02a3e40166f53f12c3269486cb56930139..8dff012bddf973d2f32401330217cdbcfd8e3a4f 100644 --- a/net/smc/smc_core.c +++ b/net/smc/smc_core.c @@ -127,6 +127,7 @@ static int smcr_lgr_conn_assign_link(struct smc_connection *conn, bool first) int i, j; /* do link balancing */ + conn->lnk = NULL; /* reset conn->lnk first */ for (i = 0; i < SMC_LINKS_PER_LGR_MAX; i++) { struct smc_link *lnk = &conn->lgr->lnk[i]; @@ -172,8 +173,10 @@ static int smc_lgr_register_conn(struct smc_connection *conn, bool first) if (!conn->lgr->is_smcd) { rc = smcr_lgr_conn_assign_link(conn, first); - if (rc) + if (rc) { + conn->lgr = NULL; return rc; + } } /* find a new alert_token_local value not yet used by some connection * in this link group @@ -225,7 +228,6 @@ int smc_nl_get_sys_info(struct sk_buff *skb, struct netlink_callback *cb) struct smc_nl_dmp_ctx *cb_ctx = smc_nl_dmp_ctx(cb); char hostname[SMC_MAX_HOSTNAME_LEN + 1]; char smc_seid[SMC_MAX_EID_LEN + 1]; - struct smcd_dev *smcd_dev; struct nlattr *attrs; u8 *seid = NULL; u8 *host = NULL; @@ -247,6 +249,8 @@ int smc_nl_get_sys_info(struct sk_buff *skb, struct netlink_callback *cb) goto errattr; if (nla_put_u8(skb, SMC_NLA_SYS_IS_ISM_V2, smc_ism_is_v2_capable())) goto errattr; + if (nla_put_u8(skb, SMC_NLA_SYS_IS_SMCR_V2, true)) + goto errattr; smc_clc_get_hostname(&host); if (host) { memcpy(hostname, host, SMC_MAX_HOSTNAME_LEN); @@ -254,13 +258,8 @@ int smc_nl_get_sys_info(struct sk_buff *skb, struct netlink_callback *cb) if (nla_put_string(skb, SMC_NLA_SYS_LOCAL_HOST, hostname)) goto errattr; } - mutex_lock(&smcd_dev_list.mutex); - smcd_dev = list_first_entry_or_null(&smcd_dev_list.list, - struct smcd_dev, list); - if (smcd_dev) - smc_ism_get_system_eid(smcd_dev, &seid); - mutex_unlock(&smcd_dev_list.mutex); - if (seid && smc_ism_is_v2_capable()) { + if (smc_ism_is_v2_capable()) { + smc_ism_get_system_eid(&seid); memcpy(smc_seid, seid, SMC_MAX_EID_LEN); smc_seid[SMC_MAX_EID_LEN] = 0; if (nla_put_string(skb, SMC_NLA_SYS_SEID, smc_seid)) @@ -279,12 +278,69 @@ int smc_nl_get_sys_info(struct sk_buff *skb, struct netlink_callback *cb) return skb->len; } +/* Fill SMC_NLA_LGR_D_V2_COMMON/SMC_NLA_LGR_R_V2_COMMON nested attributes */ +static int smc_nl_fill_lgr_v2_common(struct smc_link_group *lgr, + struct sk_buff *skb, + struct netlink_callback *cb, + struct nlattr *v2_attrs) +{ + char smc_host[SMC_MAX_HOSTNAME_LEN + 1]; + char smc_eid[SMC_MAX_EID_LEN + 1]; + + if (nla_put_u8(skb, SMC_NLA_LGR_V2_VER, lgr->smc_version)) + goto errv2attr; + if (nla_put_u8(skb, SMC_NLA_LGR_V2_REL, lgr->peer_smc_release)) + goto errv2attr; + if (nla_put_u8(skb, SMC_NLA_LGR_V2_OS, lgr->peer_os)) + goto errv2attr; + memcpy(smc_host, lgr->peer_hostname, SMC_MAX_HOSTNAME_LEN); + smc_host[SMC_MAX_HOSTNAME_LEN] = 0; + if (nla_put_string(skb, SMC_NLA_LGR_V2_PEER_HOST, smc_host)) + goto errv2attr; + memcpy(smc_eid, lgr->negotiated_eid, SMC_MAX_EID_LEN); + smc_eid[SMC_MAX_EID_LEN] = 0; + if (nla_put_string(skb, SMC_NLA_LGR_V2_NEG_EID, smc_eid)) + goto errv2attr; + + nla_nest_end(skb, v2_attrs); + return 0; + +errv2attr: + nla_nest_cancel(skb, v2_attrs); + return -EMSGSIZE; +} + +static int smc_nl_fill_smcr_lgr_v2(struct smc_link_group *lgr, + struct sk_buff *skb, + struct netlink_callback *cb) +{ + struct nlattr *v2_attrs; + + v2_attrs = nla_nest_start(skb, SMC_NLA_LGR_R_V2); + if (!v2_attrs) + goto errattr; + if (nla_put_u8(skb, SMC_NLA_LGR_R_V2_DIRECT, !lgr->uses_gateway)) + goto errv2attr; + if (nla_put_u8(skb, SMC_NLA_LGR_R_V2_MAX_CONNS, lgr->max_conns)) + goto errv2attr; + if (nla_put_u8(skb, SMC_NLA_LGR_R_V2_MAX_LINKS, lgr->max_links)) + goto errv2attr; + + nla_nest_end(skb, v2_attrs); + return 0; + +errv2attr: + nla_nest_cancel(skb, v2_attrs); +errattr: + return -EMSGSIZE; +} + static int smc_nl_fill_lgr(struct smc_link_group *lgr, struct sk_buff *skb, struct netlink_callback *cb) { char smc_target[SMC_MAX_PNETID_LEN + 1]; - struct nlattr *attrs; + struct nlattr *attrs, *v2_attrs; attrs = nla_nest_start(skb, SMC_GEN_LGR_SMCR); if (!attrs) @@ -304,6 +360,15 @@ static int smc_nl_fill_lgr(struct smc_link_group *lgr, smc_target[SMC_MAX_PNETID_LEN] = 0; if (nla_put_string(skb, SMC_NLA_LGR_R_PNETID, smc_target)) goto errattr; + if (lgr->smc_version > SMC_V1) { + v2_attrs = nla_nest_start(skb, SMC_NLA_LGR_R_V2_COMMON); + if (!v2_attrs) + goto errattr; + if (smc_nl_fill_lgr_v2_common(lgr, skb, cb, v2_attrs)) + goto errattr; + if (smc_nl_fill_smcr_lgr_v2(lgr, skb, cb)) + goto errattr; + } nla_nest_end(skb, attrs); return 0; @@ -436,10 +501,9 @@ static int smc_nl_fill_smcd_lgr(struct smc_link_group *lgr, struct sk_buff *skb, struct netlink_callback *cb) { - char smc_host[SMC_MAX_HOSTNAME_LEN + 1]; char smc_pnet[SMC_MAX_PNETID_LEN + 1]; - char smc_eid[SMC_MAX_EID_LEN + 1]; - struct nlattr *v2_attrs; + struct smcd_dev *smcd = lgr->smcd; + struct smcd_gid smcd_gid; struct nlattr *attrs; void *nlh; @@ -455,12 +519,19 @@ static int smc_nl_fill_smcd_lgr(struct smc_link_group *lgr, if (nla_put_u32(skb, SMC_NLA_LGR_D_ID, *((u32 *)&lgr->id))) goto errattr; - if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_GID, lgr->smcd->local_gid, - SMC_NLA_LGR_D_PAD)) + smcd->ops->get_local_gid(smcd, &smcd_gid); + if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_GID, + smcd_gid.gid, SMC_NLA_LGR_D_PAD)) + goto errattr; + if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_EXT_GID, + smcd_gid.gid_ext, SMC_NLA_LGR_D_PAD)) goto errattr; - if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_PEER_GID, lgr->peer_gid, + if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_PEER_GID, lgr->peer_gid.gid, SMC_NLA_LGR_D_PAD)) goto errattr; + if (nla_put_u64_64bit(skb, SMC_NLA_LGR_D_PEER_EXT_GID, + lgr->peer_gid.gid_ext, SMC_NLA_LGR_D_PAD)) + goto errattr; if (nla_put_u8(skb, SMC_NLA_LGR_D_VLAN_ID, lgr->vlan_id)) goto errattr; if (nla_put_u32(skb, SMC_NLA_LGR_D_CONNS_NUM, lgr->conns_num)) @@ -471,32 +542,19 @@ static int smc_nl_fill_smcd_lgr(struct smc_link_group *lgr, smc_pnet[SMC_MAX_PNETID_LEN] = 0; if (nla_put_string(skb, SMC_NLA_LGR_D_PNETID, smc_pnet)) goto errattr; + if (lgr->smc_version > SMC_V1) { + struct nlattr *v2_attrs; - v2_attrs = nla_nest_start(skb, SMC_NLA_LGR_V2); - if (!v2_attrs) - goto errattr; - if (nla_put_u8(skb, SMC_NLA_LGR_V2_VER, lgr->smc_version)) - goto errv2attr; - if (nla_put_u8(skb, SMC_NLA_LGR_V2_REL, lgr->peer_smc_release)) - goto errv2attr; - if (nla_put_u8(skb, SMC_NLA_LGR_V2_OS, lgr->peer_os)) - goto errv2attr; - memcpy(smc_host, lgr->peer_hostname, SMC_MAX_HOSTNAME_LEN); - smc_host[SMC_MAX_HOSTNAME_LEN] = 0; - if (nla_put_string(skb, SMC_NLA_LGR_V2_PEER_HOST, smc_host)) - goto errv2attr; - memcpy(smc_eid, lgr->negotiated_eid, SMC_MAX_EID_LEN); - smc_eid[SMC_MAX_EID_LEN] = 0; - if (nla_put_string(skb, SMC_NLA_LGR_V2_NEG_EID, smc_eid)) - goto errv2attr; - - nla_nest_end(skb, v2_attrs); + v2_attrs = nla_nest_start(skb, SMC_NLA_LGR_D_V2_COMMON); + if (!v2_attrs) + goto errattr; + if (smc_nl_fill_lgr_v2_common(lgr, skb, cb, v2_attrs)) + goto errattr; + } nla_nest_end(skb, attrs); genlmsg_end(skb, nlh); return 0; -errv2attr: - nla_nest_cancel(skb, v2_attrs); errattr: nla_nest_cancel(skb, attrs); errout: @@ -581,15 +639,13 @@ int smcd_nl_get_lgr(struct sk_buff *skb, struct netlink_callback *cb) return skb->len; } -void smc_lgr_cleanup_early(struct smc_connection *conn) +void smc_lgr_cleanup_early(struct smc_link_group *lgr) { - struct smc_link_group *lgr = conn->lgr; spinlock_t *lgr_lock; if (!lgr) return; - smc_conn_free(conn); smc_lgr_list_head(lgr, &lgr_lock); spin_lock_bh(lgr_lock); /* do not use this link group for new connections */ @@ -692,28 +748,34 @@ static void smcr_copy_dev_info_to_link(struct smc_link *link) int smcr_link_init(struct smc_link_group *lgr, struct smc_link *lnk, u8 link_idx, struct smc_init_info *ini) { + struct smc_ib_device *smcibdev; u8 rndvec[3]; int rc; - get_device(&ini->ib_dev->ibdev->dev); - atomic_inc(&ini->ib_dev->lnk_cnt); + if (lgr->smc_version == SMC_V2) { + lnk->smcibdev = ini->smcrv2.ib_dev_v2; + lnk->ibport = ini->smcrv2.ib_port_v2; + } else { + lnk->smcibdev = ini->ib_dev; + lnk->ibport = ini->ib_port; + } + get_device(&lnk->smcibdev->ibdev->dev); + atomic_inc(&lnk->smcibdev->lnk_cnt); refcount_set(&lnk->refcnt, 1); /* link refcnt is set to 1 */ lnk->clearing = 0; + lnk->path_mtu = lnk->smcibdev->pattr[lnk->ibport - 1].active_mtu; lnk->link_id = smcr_next_link_id(lgr); lnk->lgr = lgr; smc_lgr_hold(lgr); /* lgr_put in smcr_link_clear() */ lnk->link_idx = link_idx; lnk->wr_rx_id_compl = 0; - lnk->smcibdev = ini->ib_dev; - lnk->ibport = ini->ib_port; smc_ibdev_cnt_inc(lnk); smcr_copy_dev_info_to_link(lnk); - lnk->path_mtu = ini->ib_dev->pattr[ini->ib_port - 1].active_mtu; atomic_set(&lnk->conn_cnt, 0); smc_llc_link_set_uid(lnk); INIT_WORK(&lnk->link_down_wrk, smc_link_down_work); - if (!ini->ib_dev->initialized) { - rc = (int)smc_ib_setup_per_ibdev(ini->ib_dev); + if (!lnk->smcibdev->initialized) { + rc = (int)smc_ib_setup_per_ibdev(lnk->smcibdev); if (rc) goto out; } @@ -721,7 +783,9 @@ int smcr_link_init(struct smc_link_group *lgr, struct smc_link *lnk, lnk->psn_initial = rndvec[0] + (rndvec[1] << 8) + (rndvec[2] << 16); rc = smc_ib_determine_gid(lnk->smcibdev, lnk->ibport, - ini->vlan_id, lnk->gid, &lnk->sgid_index); + ini->vlan_id, lnk->gid, &lnk->sgid_index, + lgr->smc_version == SMC_V2 ? + &ini->smcrv2 : NULL); if (rc) goto out; rc = smc_llc_link_init(lnk); @@ -752,11 +816,12 @@ int smcr_link_init(struct smc_link_group *lgr, struct smc_link *lnk, smc_llc_link_clear(lnk, false); out: smc_ibdev_cnt_dec(lnk); - put_device(&ini->ib_dev->ibdev->dev); + put_device(&lnk->smcibdev->ibdev->dev); + smcibdev = lnk->smcibdev; memset(lnk, 0, sizeof(struct smc_link)); lnk->state = SMC_LNK_UNUSED; - if (!atomic_dec_return(&ini->ib_dev->lnk_cnt)) - wake_up(&ini->ib_dev->lnks_deleted); + if (!atomic_dec_return(&smcibdev->lnk_cnt)) + wake_up(&smcibdev->lnks_deleted); return rc; } @@ -765,6 +830,7 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini) { struct smc_link_group *lgr; struct list_head *lgr_list; + struct smcd_dev *smcd; struct smc_link *lnk; spinlock_t *lgr_lock; u8 link_idx; @@ -811,8 +877,12 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini) lgr->conns_all = RB_ROOT; if (ini->is_smcd) { /* SMC-D specific settings */ - get_device(&ini->ism_dev[ini->ism_selected]->dev); - lgr->peer_gid = ini->ism_peer_gid[ini->ism_selected]; + smcd = ini->ism_dev[ini->ism_selected]; + get_device(smcd->ops->get_dev(smcd)); + lgr->peer_gid.gid = + ini->ism_peer_gid[ini->ism_selected].gid; + lgr->peer_gid.gid_ext = + ini->ism_peer_gid[ini->ism_selected].gid_ext; lgr->smcd = ini->ism_dev[ini->ism_selected]; lgr_list = &ini->ism_dev[ini->ism_selected]->lgr_list; lgr_lock = &lgr->smcd->lgr_lock; @@ -821,18 +891,43 @@ static int smc_lgr_create(struct smc_sock *smc, struct smc_init_info *ini) atomic_inc(&ini->ism_dev[ini->ism_selected]->lgr_cnt); } else { /* SMC-R specific settings */ + struct smc_ib_device *ibdev; + int ibport; + lgr->role = smc->listen_smc ? SMC_SERV : SMC_CLNT; - memcpy(lgr->peer_systemid, ini->ib_lcl->id_for_peer, + lgr->smc_version = ini->smcr_version; + memcpy(lgr->peer_systemid, ini->peer_systemid, SMC_SYSTEMID_LEN); - memcpy(lgr->pnet_id, ini->ib_dev->pnetid[ini->ib_port - 1], + if (lgr->smc_version == SMC_V2) { + ibdev = ini->smcrv2.ib_dev_v2; + ibport = ini->smcrv2.ib_port_v2; + lgr->saddr = ini->smcrv2.saddr; + lgr->uses_gateway = ini->smcrv2.uses_gateway; + memcpy(lgr->nexthop_mac, ini->smcrv2.nexthop_mac, + ETH_ALEN); + lgr->max_conns = ini->max_conns; + lgr->max_links = ini->max_links; + } else { + ibdev = ini->ib_dev; + ibport = ini->ib_port; + lgr->max_conns = SMC_CONN_PER_LGR_MAX; + lgr->max_links = SMC_LINKS_ADD_LNK_MAX; + } + memcpy(lgr->pnet_id, ibdev->pnetid[ibport - 1], SMC_MAX_PNETID_LEN); + rc = smc_wr_alloc_lgr_mem(lgr); + if (rc) + goto free_wq; smc_llc_lgr_init(lgr, smc); link_idx = SMC_SINGLE_LINK; lnk = &lgr->lnk[link_idx]; rc = smcr_link_init(lgr, lnk, link_idx, ini); - if (rc) + if (rc) { + smc_wr_free_lgr_mem(lgr); goto free_wq; + } + lgr->net = smc_ib_net(lnk->smcibdev); lgr_list = &smc_lgr_list.list; lgr_lock = &smc_lgr_list.lock; lgr->buf_type = sysctl_smcr_buf_type(sock_net(&smc->sk)); @@ -1050,6 +1145,20 @@ static void smcr_buf_unuse(struct smc_buf_desc *buf_desc, bool is_rmb, } } +static void smcd_buf_detach(struct smc_connection *conn) +{ + struct smcd_dev *smcd = conn->lgr->smcd; + u64 peer_token = conn->peer_token; + + if (!conn->sndbuf_desc) + return; + + smc_ism_detach_dmb(smcd, peer_token); + + kfree(conn->sndbuf_desc); + conn->sndbuf_desc = NULL; +} + static void smc_buf_unuse(struct smc_connection *conn, struct smc_link_group *lgr) { @@ -1094,6 +1203,8 @@ void smc_conn_free(struct smc_connection *conn) if (lgr->is_smcd) { if (!list_empty(&lgr->list)) smc_ism_unset_conn(conn); + if (smc_ism_support_dmb_nocopy(lgr->smcd)) + smcd_buf_detach(conn); tasklet_kill(&conn->rx_tsklet); } else { smc_cdc_wait_pend_tx_wr(conn); @@ -1300,10 +1411,11 @@ static void smc_lgr_free(struct smc_link_group *lgr) destroy_workqueue(lgr->tx_wq); if (lgr->is_smcd) { smc_ism_put_vlan(lgr->smcd, lgr->vlan_id); - put_device(&lgr->smcd->dev); + put_device(lgr->smcd->ops->get_dev(lgr->smcd)); if (!atomic_dec_return(&lgr->smcd->lgr_cnt)) wake_up(&lgr->smcd->lgrs_deleted); } else { + smc_wr_free_lgr_mem(lgr); if (!atomic_dec_return(&lgr_cnt)) wake_up(&lgrs_deleted); } @@ -1321,20 +1433,6 @@ void smc_lgr_put(struct smc_link_group *lgr) __smc_lgr_free(lgr); } -static void smcd_unregister_all_dmbs(struct smc_link_group *lgr) -{ - int i; - - for (i = 0; i < SMC_RMBE_SIZES; i++) { - struct smc_buf_desc *buf_desc; - - list_for_each_entry(buf_desc, &lgr->rmbs[i], list) { - buf_desc->len += sizeof(struct smcd_cdc_msg); - smc_ism_unregister_dmb(lgr->smcd, buf_desc); - } - } -} - static void smc_sk_wake_ups(struct smc_sock *smc) { smc->sk.sk_write_space(&smc->sk); @@ -1356,6 +1454,8 @@ static void smc_conn_kill(struct smc_connection *conn, bool soft) smc_sk_wake_ups(smc); if (conn->lgr->is_smcd) { smc_ism_unset_conn(conn); + if (smc_ism_support_dmb_nocopy(conn->lgr->smcd)) + smcd_buf_detach(conn); if (soft) tasklet_kill(&conn->rx_tsklet); else @@ -1371,7 +1471,6 @@ static void smc_lgr_cleanup(struct smc_link_group *lgr) { if (lgr->is_smcd) { smc_ism_signal_shutdown(lgr); - smcd_unregister_all_dmbs(lgr); } else { u32 rsn = lgr->llc_termination_rsn; @@ -1440,7 +1539,8 @@ void smc_lgr_terminate_sched(struct smc_link_group *lgr) } /* Called when peer lgr shutdown (regularly or abnormally) is received */ -void smc_smcd_terminate(struct smcd_dev *dev, u64 peer_gid, unsigned short vlan) +void smc_smcd_terminate(struct smcd_dev *dev, struct smcd_gid *peer_gid, + unsigned short vlan) { struct smc_link_group *lgr, *l; LIST_HEAD(lgr_free_list); @@ -1448,9 +1548,12 @@ void smc_smcd_terminate(struct smcd_dev *dev, u64 peer_gid, unsigned short vlan) /* run common cleanup function and build free list */ spin_lock_bh(&dev->lgr_lock); list_for_each_entry_safe(lgr, l, &dev->lgr_list, list) { - if ((!peer_gid || lgr->peer_gid == peer_gid) && + if ((!peer_gid->gid || + (lgr->peer_gid.gid == peer_gid->gid && + !smc_ism_is_emulated(dev) ? 1 : + lgr->peer_gid.gid_ext == peer_gid->gid_ext)) && (vlan == VLAN_VID_MASK || lgr->vlan_id == vlan)) { - if (peer_gid) /* peer triggered termination */ + if (peer_gid->gid) /* peer triggered termination */ lgr->peer_shutdown = 1; list_move(&lgr->list, &lgr_free_list); lgr->freeing = 1; @@ -1557,9 +1660,9 @@ void smcr_lgr_set_type(struct smc_link_group *lgr, enum smc_lgr_type new_type) lgr_type = "ASYMMETRIC_LOCAL"; break; } - pr_warn_ratelimited("smc: SMC-R lg %*phN state changed: " + pr_warn_ratelimited("smc: SMC-R lg %*phN net %llu state changed: " "%s, pnetid %.16s\n", SMC_LGR_ID_SIZE, &lgr->id, - lgr_type, lgr->pnet_id); + lgr->net->net_cookie, lgr_type, lgr->pnet_id); } /* set new lgr type and tag a link as asymmetric */ @@ -1595,7 +1698,11 @@ void smcr_port_add(struct smc_ib_device *smcibdev, u8 ibport) if (strncmp(smcibdev->pnetid[ibport - 1], lgr->pnet_id, SMC_MAX_PNETID_LEN) || lgr->type == SMC_LGR_SYMMETRIC || - lgr->type == SMC_LGR_ASYMMETRIC_PEER) + lgr->type == SMC_LGR_ASYMMETRIC_PEER || + !rdma_dev_access_netns(smcibdev->ibdev, lgr->net)) + continue; + + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) continue; /* trigger local add link processing */ @@ -1659,8 +1766,11 @@ void smcr_link_down_cond(struct smc_link *lnk) /* will get the lgr->llc_conf_mutex lock */ void smcr_link_down_cond_sched(struct smc_link *lnk) { - if (smc_link_downing(&lnk->state)) - schedule_work(&lnk->link_down_wrk); + if (smc_link_downing(&lnk->state)) { + smcr_link_hold(lnk); /* smcr_link_put in link_down_wrk */ + if (!schedule_work(&lnk->link_down_wrk)) + smcr_link_put(lnk); + } } void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport) @@ -1691,11 +1801,14 @@ static void smc_link_down_work(struct work_struct *work) struct smc_link_group *lgr = link->lgr; if (list_empty(&lgr->list)) - return; + goto out; wake_up_all(&lgr->llc_msg_waiter); down_write(&lgr->llc_conf_lock); smcr_link_down(link); up_write(&lgr->llc_conf_lock); + +out: + smcr_link_put(link); /* smcr_link_hold by schedulers of link_down_work */ } static int smc_vlan_by_tcpsk_walk(struct net_device *lower_dev, @@ -1746,37 +1859,57 @@ int smc_vlan_by_tcpsk(struct socket *clcsock, struct smc_init_info *ini) return rc; } -static bool smcr_lgr_match(struct smc_link_group *lgr, - struct smc_clc_msg_local *lcl, - enum smc_lgr_role role, u32 clcqpn) +static bool smcr_lgr_match(struct smc_link_group *lgr, u8 smcr_version, + u8 peer_systemid[], + u8 peer_gid[], + u8 peer_mac_v1[], + enum smc_lgr_role role, u32 clcqpn, + struct net *net) { + struct smc_link *lnk; int i; - if (memcmp(lgr->peer_systemid, lcl->id_for_peer, SMC_SYSTEMID_LEN) || + if (memcmp(lgr->peer_systemid, peer_systemid, SMC_SYSTEMID_LEN) || lgr->role != role) return false; for (i = 0; i < SMC_LINKS_PER_LGR_MAX; i++) { - if (!smc_link_active(&lgr->lnk[i])) + lnk = &lgr->lnk[i]; + + if (!smc_link_active(lnk)) continue; - if ((lgr->role == SMC_SERV || lgr->lnk[i].peer_qpn == clcqpn) && - !memcmp(lgr->lnk[i].peer_gid, &lcl->gid, SMC_GID_SIZE) && - !memcmp(lgr->lnk[i].peer_mac, lcl->mac, sizeof(lcl->mac))) + /* use verbs API to check netns, instead of lgr->net */ + if (!rdma_dev_access_netns(lnk->smcibdev->ibdev, net)) + return false; + if ((lgr->role == SMC_SERV || lnk->peer_qpn == clcqpn) && + !memcmp(lnk->peer_gid, peer_gid, SMC_GID_SIZE) && + (smcr_version == SMC_V2 || + !memcmp(lnk->peer_mac, peer_mac_v1, ETH_ALEN))) return true; } return false; } static bool smcd_lgr_match(struct smc_link_group *lgr, - struct smcd_dev *smcismdev, u64 peer_gid) + struct smcd_dev *smcismdev, + struct smcd_gid *peer_gid) { - return lgr->peer_gid == peer_gid && lgr->smcd == smcismdev; + if (lgr->peer_gid.gid != peer_gid->gid || + lgr->smcd != smcismdev) + return false; + + if (smc_ism_is_emulated(smcismdev) && + lgr->peer_gid.gid_ext != peer_gid->gid_ext) + return false; + + return true; } /* create a new SMC connection (and a new link group if necessary) */ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini) { struct smc_connection *conn = &smc->conn; + struct net *net = sock_net(&smc->sk); struct list_head *lgr_list; struct smc_link_group *lgr; enum smc_lgr_role role; @@ -1799,13 +1932,16 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini) write_lock_bh(&lgr->conns_lock); if ((ini->is_smcd ? smcd_lgr_match(lgr, ini->ism_dev[ini->ism_selected], - ini->ism_peer_gid[ini->ism_selected]) : - smcr_lgr_match(lgr, ini->ib_lcl, role, ini->ib_clcqpn)) && + &ini->ism_peer_gid[ini->ism_selected]) : + smcr_lgr_match(lgr, ini->smcr_version, + ini->peer_systemid, + ini->peer_gid, ini->peer_mac, role, + ini->ib_clcqpn, net)) && !lgr->sync_err && (ini->smcd_version == SMC_V2 || lgr->vlan_id == ini->vlan_id) && (role == SMC_CLNT || ini->is_smcd || - (lgr->conns_num < SMC_RMBS_PER_LGR_MAX && + (lgr->conns_num < lgr->max_conns && !bitmap_full(lgr->rtokens_used_mask, SMC_RMBS_PER_LGR_MAX)))) { /* link group found */ ini->first_contact_local = 0; @@ -1840,8 +1976,10 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini) write_lock_bh(&lgr->conns_lock); rc = smc_lgr_register_conn(conn, true); write_unlock_bh(&lgr->conns_lock); - if (rc) + if (rc) { + smc_lgr_cleanup_early(lgr); goto out; + } } conn->local_tx_ctrl.common.type = SMC_CDC_MSG_TYPE; conn->local_tx_ctrl.len = SMC_WR_TX_SIZE; @@ -1872,8 +2010,10 @@ int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini) */ static u8 smc_compress_bufsize(struct smc_link_group *lgr, int size, bool is_smcd, bool is_rmb) { - const unsigned int max_scat = SG_MAX_SINGLE_ALLOC * PAGE_SIZE; - u8 compressed, max_phy_compressed; + u8 compressed; +#ifdef CONFIG_ARCH_NO_SG_CHAIN + u8 max_phy_compressed; +#endif if (size <= SMC_BUF_MIN_SIZE) return 0; @@ -1882,8 +2022,9 @@ static u8 smc_compress_bufsize(struct smc_link_group *lgr, int size, bool is_smc compressed = min_t(u8, ilog2(size) + 1, is_smcd ? SMCD_DMBE_SIZES : SMCR_RMBE_SIZES); +#ifdef CONFIG_ARCH_NO_SG_CHAIN if (!is_smcd && is_rmb && lgr->buf_type != SMCR_VIRT_CONT_BUFS) { - max_phy_compressed = ilog2(max_scat >> 14); + max_phy_compressed = ilog2((SG_MAX_SINGLE_ALLOC * PAGE_SIZE) >> 14); switch (lgr->buf_type) { case SMCR_MIXED_BUFS: if (compressed > max_phy_compressed) @@ -1897,6 +2038,7 @@ static u8 smc_compress_bufsize(struct smc_link_group *lgr, int size, bool is_smc break; } } +#endif return compressed; } @@ -1967,7 +2109,7 @@ static int smcr_buf_map_link(struct smc_buf_desc *buf_desc, bool is_rmb, for_each_sg(buf_desc->sgt[lnk->link_idx].sgl, sg, nents, i) { size = min_t(int, PAGE_SIZE - offset, buf_size); sg_set_page(sg, vmalloc_to_page(buf), size, offset); - buf += size / sizeof(*buf); + buf += size; buf_size -= size; offset = 0; } @@ -2117,7 +2259,7 @@ int smcr_buf_reg_lgr(struct smc_link *lnk) } static struct smc_buf_desc *smcr_new_buf_create(struct smc_link_group *lgr, - bool is_rmb, int bufsize) + int bufsize) { struct smc_buf_desc *buf_desc; @@ -2143,7 +2285,7 @@ static struct smc_buf_desc *smcr_new_buf_create(struct smc_link_group *lgr, } if (lgr->buf_type == SMCR_PHYS_CONT_BUFS) goto out; - fallthrough; // try virtually continguous buf + fallthrough; // try virtually contiguous buf case SMCR_VIRT_CONT_BUFS: buf_desc->order = get_order(bufsize); buf_desc->cpu_addr = vzalloc(PAGE_SIZE << buf_desc->order); @@ -2264,7 +2406,7 @@ static int __smc_buf_create(struct smc_sock *smc, bool is_smcd, bool is_rmb) if (is_smcd) buf_desc = smcd_new_buf_create(lgr, is_rmb, bufsize); else - buf_desc = smcr_new_buf_create(lgr, is_rmb, bufsize); + buf_desc = smcr_new_buf_create(lgr, bufsize); if (PTR_ERR(buf_desc) == -ENOMEM) break; @@ -2348,12 +2490,18 @@ int smc_buf_create(struct smc_sock *smc, bool is_smcd) int rc; /* create send buffer */ + if (is_smcd && + smc_ism_support_dmb_nocopy(smc->conn.lgr->smcd)) + goto create_rmb; + rc = __smc_buf_create(smc, is_smcd, false); if (rc) return rc; + +create_rmb: /* create rmb */ rc = __smc_buf_create(smc, is_smcd, true); - if (rc) { + if (rc && smc->conn.sndbuf_desc) { down_write(&smc->conn.lgr->sndbufs_lock); list_del(&smc->conn.sndbuf_desc->list); up_write(&smc->conn.lgr->sndbufs_lock); @@ -2363,6 +2511,41 @@ int smc_buf_create(struct smc_sock *smc, bool is_smcd) return rc; } +int smcd_buf_attach(struct smc_sock *smc) +{ + struct smc_connection *conn = &smc->conn; + struct smcd_dev *smcd = conn->lgr->smcd; + u64 peer_token = conn->peer_token; + struct smc_buf_desc *buf_desc; + int rc; + + buf_desc = kzalloc(sizeof(*buf_desc), GFP_KERNEL); + if (!buf_desc) + return -ENOMEM; + + /* The ghost sndbuf_desc describes the same memory region as + * peer RMB. Its lifecycle is consistent with the connection's + * and it will be freed with the connections instead of the + * link group. + */ + rc = smc_ism_attach_dmb(smcd, peer_token, buf_desc); + if (rc) + goto free; + + smc->sk.sk_sndbuf = buf_desc->len; + buf_desc->cpu_addr = + (u8 *)buf_desc->cpu_addr + sizeof(struct smcd_cdc_msg); + buf_desc->len -= sizeof(struct smcd_cdc_msg); + conn->sndbuf_desc = buf_desc; + conn->sndbuf_desc->used = 1; + atomic_set(&conn->sndbuf_space, conn->sndbuf_desc->len); + return 0; + +free: + kfree(buf_desc); + return rc; +} + static inline int smc_rmb_reserve_rtoken_idx(struct smc_link_group *lgr) { int i; @@ -2519,6 +2702,7 @@ static int smc_core_reboot_event(struct notifier_block *this, { smc_lgrs_shutdown(); smc_ib_unregister_client(); + smc_ism_exit(); return 0; } diff --git a/net/smc/smc_core.h b/net/smc/smc_core.h index 16c286360d36685b09cb03e6e7b68135720f34fb..7148e191c14ac37f3d03df09a07d19da5c1db9ff 100644 --- a/net/smc/smc_core.h +++ b/net/smc/smc_core.h @@ -17,11 +17,22 @@ #include #include #include +#include #include "smc.h" #include "smc_ib.h" +#include "smc_clc.h" #define SMC_RMBS_PER_LGR_MAX 255 /* max. # of RMBs per link group */ +#define SMC_CONN_PER_LGR_MIN 16 /* min. # of connections per link group */ +#define SMC_CONN_PER_LGR_MAX 255 /* max. # of connections per link group, + * also is the default value for SMC-R v1 and v2.0 + */ +#define SMC_CONN_PER_LGR_PREFER 255 /* Preferred connections per link group used for + * SMC-R v2.1 and later negotiation, vendors or + * distributions may modify it to a value between + * 16-255 as needed. + */ struct smc_lgr_list { /* list of link group definition */ struct list_head list; @@ -42,11 +53,16 @@ enum smc_link_state { /* possible states of a link */ }; #define SMC_WR_BUF_SIZE 48 /* size of work request buffer */ +#define SMC_WR_BUF_V2_SIZE 8192 /* size of v2 work request buffer */ struct smc_wr_buf { u8 raw[SMC_WR_BUF_SIZE]; }; +struct smc_wr_v2_buf { + u8 raw[SMC_WR_BUF_V2_SIZE]; +}; + #define SMC_WR_REG_MR_WAIT_TIME (5 * HZ)/* wait time for ib_wr_reg_mr result */ enum smc_wr_reg_state { @@ -92,18 +108,26 @@ struct smc_link { struct smc_wr_tx_pend *wr_tx_pends; /* WR send waiting for CQE */ struct completion *wr_tx_compl; /* WR send CQE completion */ /* above four vectors have wr_tx_cnt elements and use the same index */ + struct ib_send_wr *wr_tx_v2_ib; /* WR send v2 meta data */ + struct ib_sge *wr_tx_v2_sge; /* WR send v2 gather meta data*/ + struct smc_wr_tx_pend *wr_tx_v2_pend; /* WR send v2 waiting for CQE */ dma_addr_t wr_tx_dma_addr; /* DMA address of wr_tx_bufs */ + dma_addr_t wr_tx_v2_dma_addr; /* DMA address of v2 tx buf*/ atomic_long_t wr_tx_id; /* seq # of last sent WR */ unsigned long *wr_tx_mask; /* bit mask of used indexes */ u32 wr_tx_cnt; /* number of WR send buffers */ wait_queue_head_t wr_tx_wait; /* wait for free WR send buf */ - atomic_t wr_tx_refcnt; /* tx refs to link */ + struct { + struct percpu_ref wr_tx_refs; + } ____cacheline_aligned_in_smp; + struct completion tx_ref_comp; struct smc_wr_buf *wr_rx_bufs; /* WR recv payload buffers */ struct ib_recv_wr *wr_rx_ibs; /* WR recv meta data */ struct ib_sge *wr_rx_sges; /* WR recv scatter meta data */ /* above three vectors have wr_rx_cnt elements and use the same index */ dma_addr_t wr_rx_dma_addr; /* DMA address of wr_rx_bufs */ + dma_addr_t wr_rx_v2_dma_addr; /* DMA address of v2 rx buf*/ u64 wr_rx_id; /* seq # of last recv WR */ u64 wr_rx_id_compl; /* seq # of last completed WR */ u32 wr_rx_cnt; /* number of WR recv buffers */ @@ -112,7 +136,10 @@ struct smc_link { struct ib_reg_wr wr_reg; /* WR register memory region */ wait_queue_head_t wr_reg_wait; /* wait for wr_reg result */ - atomic_t wr_reg_refcnt; /* reg refs to link */ + struct { + struct percpu_ref wr_reg_refs; + } ____cacheline_aligned_in_smp; + struct completion reg_ref_comp; enum smc_wr_reg_state wr_reg_state; /* state of wr_reg request */ u8 gid[SMC_GID_SIZE];/* gid matching used vlan id*/ @@ -148,6 +175,15 @@ struct smc_link { */ #define SMC_LINKS_PER_LGR_MAX 3 #define SMC_SINGLE_LINK 0 +#define SMC_LINKS_ADD_LNK_MIN 1 /* min. # of links per link group */ +#define SMC_LINKS_ADD_LNK_MAX 2 /* max. # of links per link group, also is the + * default value for smc-r v1.0 and v2.0 + */ +#define SMC_LINKS_PER_LGR_MAX_PREFER 2 /* Preferred max links per link group used for + * SMC-R v2.1 and later negotiation, vendors or + * distributions may modify it to a value between + * 1-2 as needed. + */ /* tx/rx buffer list element for sndbufs list and rmbs list of a lgr */ struct smc_buf_desc { @@ -224,6 +260,7 @@ enum smc_llc_flowtype { SMC_LLC_FLOW_NONE = 0, SMC_LLC_FLOW_ADD_LINK = 2, SMC_LLC_FLOW_DEL_LINK = 4, + SMC_LLC_FLOW_REQ_ADD_LINK = 5, SMC_LLC_FLOW_RKEY = 6, }; @@ -267,6 +304,10 @@ struct smc_link_group { /* client or server */ struct smc_link lnk[SMC_LINKS_PER_LGR_MAX]; /* smc link */ + struct smc_wr_v2_buf *wr_rx_buf_v2; + /* WR v2 recv payload buffer */ + struct smc_wr_v2_buf *wr_tx_buf_v2; + /* WR v2 send payload buffer */ char peer_systemid[SMC_SYSTEMID_LEN]; /* unique system_id of peer */ struct smc_rtoken rtokens[SMC_RMBS_PER_LGR_MAX] @@ -306,9 +347,18 @@ struct smc_link_group { /* link keep alive time */ u32 llc_termination_rsn; /* rsn code for termination */ + u8 nexthop_mac[ETH_ALEN]; + u8 uses_gateway; + __be32 saddr; + /* net namespace */ + struct net *net; + u8 max_conns; + /* max conn can be assigned to lgr */ + u8 max_links; + /* max links can be added in lgr */ }; struct { /* SMC-D */ - u64 peer_gid; + struct smcd_gid peer_gid; /* Peer GID (remote) */ struct smcd_dev *smcd; /* ISM device for VLAN reg. */ @@ -320,23 +370,64 @@ struct smc_link_group { struct smc_clc_msg_local; +#define GID_LIST_SIZE 2 + +struct smc_gidlist { + u8 len; + u8 list[GID_LIST_SIZE][SMC_GID_SIZE]; +}; + +struct smc_init_info_smcrv2 { + /* Input fields */ + __be32 saddr; + struct sock *clc_sk; + __be32 daddr; + + /* Output fields when saddr is set */ + struct smc_ib_device *ib_dev_v2; + u8 ib_port_v2; + u8 ib_gid_v2[SMC_GID_SIZE]; + + /* Additional output fields when clc_sk and daddr is set as well */ + u8 uses_gateway; + u8 nexthop_mac[ETH_ALEN]; + + struct smc_gidlist gidlist; +}; + +#define SMC_MAX_V2_ISM_DEVS SMCD_CLC_MAX_V2_GID_ENTRIES + /* max # of proposed non-native ISM devices, + * which can't exceed the max # of CHID-GID + * entries in CLC proposal SMC-Dv2 extension. + */ struct smc_init_info { u8 is_smcd; u8 smc_type_v1; u8 smc_type_v2; + u8 release_nr; + u8 max_conns; + u8 max_links; u8 first_contact_peer; u8 first_contact_local; + u16 feature_mask; unsigned short vlan_id; + u32 rc; + u8 negotiated_eid[SMC_MAX_EID_LEN]; /* SMC-R */ - struct smc_clc_msg_local *ib_lcl; + u8 smcr_version; + u8 check_smcrv2; + u8 peer_gid[SMC_GID_SIZE]; + u8 peer_mac[ETH_ALEN]; + u8 peer_systemid[SMC_SYSTEMID_LEN]; struct smc_ib_device *ib_dev; u8 ib_gid[SMC_GID_SIZE]; u8 ib_port; u32 ib_clcqpn; + struct smc_init_info_smcrv2 smcrv2; /* SMC-D */ - u64 ism_peer_gid[SMC_MAX_ISM_DEVS + 1]; - struct smcd_dev *ism_dev[SMC_MAX_ISM_DEVS + 1]; - u16 ism_chid[SMC_MAX_ISM_DEVS + 1]; + struct smcd_gid ism_peer_gid[SMC_MAX_V2_ISM_DEVS + 1]; + struct smcd_dev *ism_dev[SMC_MAX_V2_ISM_DEVS + 1]; + u16 ism_chid[SMC_MAX_V2_ISM_DEVS + 1]; u8 ism_offered_cnt; /* # of ISM devices offered */ u8 ism_selected; /* index of selected ISM dev*/ u8 smcd_version; @@ -375,7 +466,13 @@ static inline struct smc_connection *smc_lgr_find_conn( return res; } -/* returns true if the specified link is usable */ +/* + * Returns true if the specified link is usable. + * + * usable means the link is ready to receive RDMA messages, map memory + * on the link, etc. This doesn't ensure we are able to send RDMA messages + * on this link, if sending RDMA messages is needed, use smc_link_sendable() + */ static inline bool smc_link_usable(struct smc_link *lnk) { if (lnk->state == SMC_LNK_UNUSED || lnk->state == SMC_LNK_INACTIVE) @@ -383,6 +480,15 @@ static inline bool smc_link_usable(struct smc_link *lnk) return true; } +/* + * Returns true if the specified link is ready to receive AND send RDMA + * messages. + * + * For the client side in first contact, the underlying QP may still in + * RESET or RTR when the link state is ACTIVATING, checks in smc_link_usable() + * is not strong enough. For those places that need to send any CDC or LLC + * messages, use smc_link_sendable(), otherwise, use smc_link_usable() instead + */ static inline bool smc_link_sendable(struct smc_link *lnk) { return smc_link_usable(lnk) && @@ -435,19 +541,19 @@ static inline void smc_set_pci_values(struct pci_dev *pci_dev, struct smc_sock; struct smc_clc_msg_accept_confirm; -struct smc_clc_msg_local; -void smc_lgr_cleanup_early(struct smc_connection *conn); +void smc_lgr_cleanup_early(struct smc_link_group *lgr); void smc_lgr_terminate_sched(struct smc_link_group *lgr); void smc_lgr_hold(struct smc_link_group *lgr); void smc_lgr_put(struct smc_link_group *lgr); void smcr_port_add(struct smc_ib_device *smcibdev, u8 ibport); void smcr_port_err(struct smc_ib_device *smcibdev, u8 ibport); -void smc_smcd_terminate(struct smcd_dev *dev, u64 peer_gid, +void smc_smcd_terminate(struct smcd_dev *dev, struct smcd_gid *peer_gid, unsigned short vlan); void smc_smcd_terminate_all(struct smcd_dev *dev); void smc_smcr_terminate_all(struct smc_ib_device *smcibdev); int smc_buf_create(struct smc_sock *smc, bool is_smcd); +int smcd_buf_attach(struct smc_sock *smc); int smc_uncompress_bufsize(u8 compressed); int smc_rmb_rtoken_handling(struct smc_connection *conn, struct smc_link *link, struct smc_clc_msg_accept_confirm *clc); @@ -463,7 +569,6 @@ int smc_vlan_by_tcpsk(struct socket *clcsock, struct smc_init_info *ini); void smc_conn_free(struct smc_connection *conn); int smc_conn_create(struct smc_sock *smc, struct smc_init_info *ini); -void smc_lgr_schedule_free_work_fast(struct smc_link_group *lgr); int smc_core_init(void); void smc_core_exit(void); diff --git a/net/smc/smc_diag.c b/net/smc/smc_diag.c index 5fb64b66a31970e2d09901b2e919403188f9c676..eb00035cf9c6703e3668ff040a9e7155aa2b6e42 100644 --- a/net/smc/smc_diag.c +++ b/net/smc/smc_diag.c @@ -21,6 +21,7 @@ #include "smc.h" #include "smc_core.h" +#include "smc_ism.h" struct smc_diag_dump_ctx { int pos[2]; @@ -147,17 +148,17 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb, !list_empty(&smc->conn.lgr->list)) { struct smc_diag_lgrinfo linfo = { .role = smc->conn.lgr->role, - .lnk[0].ibport = smc->conn.lgr->lnk[0].ibport, - .lnk[0].link_id = smc->conn.lgr->lnk[0].link_id, + .lnk[0].ibport = smc->conn.lnk->ibport, + .lnk[0].link_id = smc->conn.lnk->link_id, }; memcpy(linfo.lnk[0].ibname, smc->conn.lgr->lnk[0].smcibdev->ibdev->name, - sizeof(smc->conn.lgr->lnk[0].smcibdev->ibdev->name)); + sizeof(smc->conn.lnk->smcibdev->ibdev->name)); smc_gid_be16_convert(linfo.lnk[0].gid, - smc->conn.lgr->lnk[0].gid); + smc->conn.lnk->gid); smc_gid_be16_convert(linfo.lnk[0].peer_gid, - smc->conn.lgr->lnk[0].peer_gid); + smc->conn.lnk->peer_gid); if (nla_put(skb, SMC_DIAG_LGRINFO, sizeof(linfo), &linfo) < 0) goto errout; @@ -167,12 +168,17 @@ static int __smc_diag_dump(struct sock *sk, struct sk_buff *skb, !list_empty(&smc->conn.lgr->list) && smc->conn.rmb_desc) { struct smc_connection *conn = &smc->conn; struct smcd_diag_dmbinfo dinfo; + struct smcd_dev *smcd = conn->lgr->smcd; + struct smcd_gid smcd_gid; memset(&dinfo, 0, sizeof(dinfo)); dinfo.linkid = *((u32 *)conn->lgr->id); - dinfo.peer_gid = conn->lgr->peer_gid; - dinfo.my_gid = conn->lgr->smcd->local_gid; + dinfo.peer_gid = conn->lgr->peer_gid.gid; + dinfo.peer_gid_ext = conn->lgr->peer_gid.gid_ext; + smcd->ops->get_local_gid(smcd, &smcd_gid); + dinfo.my_gid = smcd_gid.gid; + dinfo.my_gid_ext = smcd_gid.gid_ext; dinfo.token = conn->rmb_desc->token; dinfo.peer_token = conn->peer_token; diff --git a/net/smc/smc_ib.c b/net/smc/smc_ib.c index 3e45bcc51cb423cd5c63f985c19b06fadf71ac04..2be0c0cdb6149da6ef9bcbcd068ae0b78fd9089f 100644 --- a/net/smc/smc_ib.c +++ b/net/smc/smc_ib.c @@ -17,6 +17,7 @@ #include #include #include +#include #include #include @@ -62,16 +63,23 @@ static int smc_ib_modify_qp_rtr(struct smc_link *lnk) IB_QP_STATE | IB_QP_AV | IB_QP_PATH_MTU | IB_QP_DEST_QPN | IB_QP_RQ_PSN | IB_QP_MAX_DEST_RD_ATOMIC | IB_QP_MIN_RNR_TIMER; struct ib_qp_attr qp_attr; + u8 hop_lim = 1; memset(&qp_attr, 0, sizeof(qp_attr)); qp_attr.qp_state = IB_QPS_RTR; qp_attr.path_mtu = min(lnk->path_mtu, lnk->peer_mtu); qp_attr.ah_attr.type = RDMA_AH_ATTR_TYPE_ROCE; rdma_ah_set_port_num(&qp_attr.ah_attr, lnk->ibport); - rdma_ah_set_grh(&qp_attr.ah_attr, NULL, 0, lnk->sgid_index, 1, 0); + if (lnk->lgr->smc_version == SMC_V2 && lnk->lgr->uses_gateway) + hop_lim = IPV6_DEFAULT_HOPLIMIT; + rdma_ah_set_grh(&qp_attr.ah_attr, NULL, 0, lnk->sgid_index, hop_lim, 0); rdma_ah_set_dgid_raw(&qp_attr.ah_attr, lnk->peer_gid); - memcpy(&qp_attr.ah_attr.roce.dmac, lnk->peer_mac, - sizeof(lnk->peer_mac)); + if (lnk->lgr->smc_version == SMC_V2 && lnk->lgr->uses_gateway) + memcpy(&qp_attr.ah_attr.roce.dmac, lnk->lgr->nexthop_mac, + sizeof(lnk->lgr->nexthop_mac)); + else + memcpy(&qp_attr.ah_attr.roce.dmac, lnk->peer_mac, + sizeof(lnk->peer_mac)); qp_attr.dest_qp_num = lnk->peer_qpn; qp_attr.rq_psn = lnk->peer_psn; /* starting receive packet seq # */ qp_attr.max_dest_rd_atomic = 1; /* max # of resources for incoming @@ -183,9 +191,87 @@ bool smc_ib_port_active(struct smc_ib_device *smcibdev, u8 ibport) return smcibdev->pattr[ibport - 1].state == IB_PORT_ACTIVE; } +int smc_ib_find_route(struct net *net, __be32 saddr, __be32 daddr, + u8 nexthop_mac[], u8 *uses_gateway) +{ + struct neighbour *neigh = NULL; + struct rtable *rt = NULL; + struct flowi4 fl4 = { + .saddr = saddr, + .daddr = daddr + }; + + if (daddr == cpu_to_be32(INADDR_NONE)) + goto out; + rt = ip_route_output_flow(net, &fl4, NULL); + if (IS_ERR(rt)) + goto out; + if (rt->rt_uses_gateway && rt->rt_gw_family != AF_INET) + goto out_rt; + neigh = dst_neigh_lookup(&rt->dst, &fl4.daddr); + if (!neigh) + goto out_rt; + memcpy(nexthop_mac, neigh->ha, ETH_ALEN); + *uses_gateway = rt->rt_uses_gateway; + neigh_release(neigh); + ip_rt_put(rt); + return 0; + +out_rt: + ip_rt_put(rt); +out: + return -ENOENT; +} + +static int smc_ib_determine_gid_rcu(const struct net_device *ndev, + const struct ib_gid_attr *attr, + u8 gid[], u8 *sgid_index, + struct smc_init_info_smcrv2 *smcrv2) +{ + if (!smcrv2 && attr->gid_type == IB_GID_TYPE_ROCE) { + if (gid) + memcpy(gid, &attr->gid, SMC_GID_SIZE); + if (sgid_index) + *sgid_index = attr->index; + return 0; + } + if (smcrv2 && attr->gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP && + smc_ib_gid_to_ipv4((u8 *)&attr->gid) != cpu_to_be32(INADDR_NONE)) { + struct in_device *in_dev = __in_dev_get_rcu(ndev); + struct net *net = dev_net(ndev); + const struct in_ifaddr *ifa; + bool subnet_match = false; + + if (!in_dev) + goto out; + in_dev_for_each_ifa_rcu(ifa, in_dev) { + if (!inet_ifa_match(smcrv2->saddr, ifa)) + continue; + subnet_match = true; + break; + } + if (!subnet_match) + goto out; + if (smcrv2->daddr && smc_ib_find_route(net, smcrv2->saddr, + smcrv2->daddr, + smcrv2->nexthop_mac, + &smcrv2->uses_gateway)) + goto out; + + if (gid) + memcpy(gid, &attr->gid, SMC_GID_SIZE); + if (sgid_index) + *sgid_index = attr->index; + return 0; + } +out: + return -ENODEV; +} + /* determine the gid for an ib-device port and vlan id */ int smc_ib_determine_gid(struct smc_ib_device *smcibdev, u8 ibport, - unsigned short vlan_id, u8 gid[], u8 *sgid_index) + unsigned short vlan_id, u8 gid[], u8 *sgid_index, + struct smc_init_info_smcrv2 *smcrv2) { const struct ib_gid_attr *attr; const struct net_device *ndev; @@ -202,13 +288,12 @@ int smc_ib_determine_gid(struct smc_ib_device *smcibdev, u8 ibport, ((!vlan_id && !is_vlan_dev(ndev)) || (vlan_id && is_vlan_dev(ndev) && vlan_dev_vlan_id(ndev) == vlan_id))) { - rcu_read_unlock(); - if (gid) - memcpy(gid, &attr->gid, SMC_GID_SIZE); - if (sgid_index) - *sgid_index = attr->index; - rdma_put_gid_attr(attr); - return 0; + if (!smc_ib_determine_gid_rcu(ndev, attr, gid, + sgid_index, smcrv2)) { + rcu_read_unlock(); + rdma_put_gid_attr(attr); + return 0; + } } rcu_read_unlock(); rdma_put_gid_attr(attr); @@ -216,6 +301,58 @@ int smc_ib_determine_gid(struct smc_ib_device *smcibdev, u8 ibport, return -ENODEV; } +/* check if gid is still defined on smcibdev */ +static bool smc_ib_check_link_gid(u8 gid[SMC_GID_SIZE], bool smcrv2, + struct smc_ib_device *smcibdev, u8 ibport) +{ + const struct ib_gid_attr *attr; + bool rc = false; + int i; + + for (i = 0; !rc && i < smcibdev->pattr[ibport - 1].gid_tbl_len; i++) { + attr = rdma_get_gid_attr(smcibdev->ibdev, ibport, i); + if (IS_ERR(attr)) + continue; + + rcu_read_lock(); + if ((!smcrv2 && attr->gid_type == IB_GID_TYPE_ROCE) || + (smcrv2 && attr->gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP && + !(ipv6_addr_type((const struct in6_addr *)&attr->gid) + & IPV6_ADDR_LINKLOCAL))) + if (!memcmp(gid, &attr->gid, SMC_GID_SIZE)) + rc = true; + rcu_read_unlock(); + rdma_put_gid_attr(attr); + } + return rc; +} + +/* check all links if the gid is still defined on smcibdev */ +static void smc_ib_gid_check(struct smc_ib_device *smcibdev, u8 ibport) +{ + struct smc_link_group *lgr; + int i; + + spin_lock_bh(&smc_lgr_list.lock); + list_for_each_entry(lgr, &smc_lgr_list.list, list) { + if (strncmp(smcibdev->pnetid[ibport - 1], lgr->pnet_id, + SMC_MAX_PNETID_LEN)) + continue; /* lgr is not affected */ + if (list_empty(&lgr->list)) + continue; + for (i = 0; i < SMC_LINKS_PER_LGR_MAX; i++) { + if (lgr->lnk[i].state == SMC_LNK_UNUSED || + lgr->lnk[i].smcibdev != smcibdev) + continue; + if (!smc_ib_check_link_gid(lgr->lnk[i].gid, + lgr->smc_version == SMC_V2, + smcibdev, ibport)) + smcr_port_err(smcibdev, ibport); + } + } + spin_unlock_bh(&smc_lgr_list.lock); +} + static int smc_ib_remember_port_attr(struct smc_ib_device *smcibdev, u8 ibport) { int rc; @@ -254,6 +391,7 @@ static void smc_ib_port_event_work(struct work_struct *work) } else { clear_bit(port_idx, smcibdev->ports_going_away); smcr_port_add(smcibdev, port_idx + 1); + smc_ib_gid_check(smcibdev, port_idx + 1); } } } @@ -522,6 +660,7 @@ void smc_ib_destroy_queue_pair(struct smc_link *lnk) /* create a queue pair within the protection domain for a link */ int smc_ib_create_queue_pair(struct smc_link *lnk) { + int sges_per_buf = (lnk->lgr->smc_version == SMC_V2) ? 2 : 1; struct ib_qp_init_attr qp_attr = { .event_handler = smc_ib_qp_event_handler, .qp_context = lnk, @@ -535,7 +674,7 @@ int smc_ib_create_queue_pair(struct smc_link *lnk) .max_send_wr = SMC_WR_BUF_CNT * 3, .max_recv_wr = SMC_WR_BUF_CNT * 3, .max_send_sge = SMC_IB_MAX_SEND_SGE, - .max_recv_sge = 1, + .max_recv_sge = sges_per_buf, .max_inline_data = 0, }, .sq_sig_type = IB_SIGNAL_REQ_WR, @@ -758,9 +897,7 @@ static void smc_copy_netdev_ifindex(struct smc_ib_device *smcibdev, int port) struct ib_device *ibdev = smcibdev->ibdev; struct net_device *ndev; - if (!ibdev->ops.get_netdev) - return; - ndev = ibdev->ops.get_netdev(ibdev, port + 1); + ndev = ib_device_get_netdev(ibdev, port + 1); if (ndev) { smcibdev->ndev_ifidx[port] = ndev->ifindex; dev_put(ndev); @@ -780,9 +917,7 @@ void smc_ib_ndev_change(struct net_device *ndev, unsigned long event) port_cnt = smcibdev->ibdev->phys_port_cnt; for (i = 0; i < min_t(size_t, port_cnt, SMC_MAX_PORTS); i++) { libdev = smcibdev->ibdev; - if (!libdev->ops.get_netdev) - continue; - lndev = libdev->ops.get_netdev(libdev, i + 1); + lndev = ib_device_get_netdev(libdev, i + 1); if (lndev) dev_put(lndev); if (lndev != ndev) diff --git a/net/smc/smc_ib.h b/net/smc/smc_ib.h index ead57976d354d3f6a2a5345e03e63c6b3488e67c..ef8ac2b7546df57c2acddf193079b29166bd20c6 100644 --- a/net/smc/smc_ib.h +++ b/net/smc/smc_ib.h @@ -59,6 +59,24 @@ struct smc_ib_device { /* ib-device infos for smc */ int ndev_ifidx[SMC_MAX_PORTS]; /* ndev if indexes */ }; +static inline __be32 smc_ib_gid_to_ipv4(u8 gid[SMC_GID_SIZE]) +{ + struct in6_addr *addr6 = (struct in6_addr *)gid; + + if (ipv6_addr_v4mapped(addr6) || + !(addr6->s6_addr32[0] | addr6->s6_addr32[1] | addr6->s6_addr32[2])) + return addr6->s6_addr32[3]; + return cpu_to_be32(INADDR_NONE); +} + +static inline struct net *smc_ib_net(struct smc_ib_device *smcibdev) +{ + if (smcibdev && smcibdev->ibdev) + return read_pnet(&smcibdev->ibdev->coredev.rdma_net); + return NULL; +} + +struct smc_init_info_smcrv2; struct smc_buf_desc; struct smc_link; @@ -78,7 +96,6 @@ void smc_ib_destroy_queue_pair(struct smc_link *lnk); int smc_ib_create_queue_pair(struct smc_link *lnk); int smc_ib_ready_link(struct smc_link *lnk); int smc_ib_modify_qp_rts(struct smc_link *lnk); -int smc_ib_modify_qp_reset(struct smc_link *lnk); int smc_ib_modify_qp_error(struct smc_link *lnk); long smc_ib_setup_per_ibdev(struct smc_ib_device *smcibdev); int smc_ib_get_memory_region(struct ib_pd *pd, int access_flags, @@ -93,7 +110,10 @@ void smc_ib_sync_sg_for_device(struct smc_link *lnk, struct smc_buf_desc *buf_slot, enum dma_data_direction data_direction); int smc_ib_determine_gid(struct smc_ib_device *smcibdev, u8 ibport, - unsigned short vlan_id, u8 gid[], u8 *sgid_index); + unsigned short vlan_id, u8 gid[], u8 *sgid_index, + struct smc_init_info_smcrv2 *smcrv2); +int smc_ib_find_route(struct net *net, __be32 saddr, __be32 daddr, + u8 nexthop_mac[], u8 *uses_gateway); bool smc_ib_is_valid_local_systemid(void); int smcr_nl_get_device(struct sk_buff *skb, struct netlink_callback *cb); #endif diff --git a/net/smc/smc_ism.c b/net/smc/smc_ism.c index 967712ba52a0d36d525ce808ad1d0ffb62bf5837..d22a2b89c0f173b0f4f20cdcf3266d95087c0bb2 100644 --- a/net/smc/smc_ism.c +++ b/net/smc/smc_ism.c @@ -16,6 +16,7 @@ #include "smc_ism.h" #include "smc_pnet.h" #include "smc_netlink.h" +#include "linux/ism.h" struct smcd_dev_list smcd_dev_list = { .list = LIST_HEAD_INIT(smcd_dev_list.list), @@ -23,28 +24,59 @@ struct smcd_dev_list smcd_dev_list = { }; static bool smc_ism_v2_capable; +static u8 smc_ism_v2_system_eid[SMC_MAX_EID_LEN]; + +#if IS_ENABLED(CONFIG_ISM) +static void smcd_register_dev(struct ism_dev *ism); +static void smcd_unregister_dev(struct ism_dev *ism); +static void smcd_handle_event(struct ism_dev *ism, struct ism_event *event); +static void smcd_handle_irq(struct ism_dev *ism, unsigned int dmbno, + u16 dmbemask); + +static struct ism_client smc_ism_client = { + .name = "SMC-D", + .add = smcd_register_dev, + .remove = smcd_unregister_dev, + .handle_event = smcd_handle_event, + .handle_irq = smcd_handle_irq, +}; +#endif -/* Test if an ISM communication is possible - same CPC */ -int smc_ism_cantalk(u64 peer_gid, unsigned short vlan_id, struct smcd_dev *smcd) +static void smc_ism_create_system_eid(void) { - return smcd->ops->query_remote_gid(smcd, peer_gid, vlan_id ? 1 : 0, - vlan_id); + struct smc_ism_seid *seid = + (struct smc_ism_seid *)smc_ism_v2_system_eid; +#if IS_ENABLED(CONFIG_S390) + struct cpuid id; + u16 ident_tail; + char tmp[5]; + + memcpy(seid->seid_string, "IBM-SYSZ-ISMSEID00000000", 24); + get_cpu_id(&id); + ident_tail = (u16)(id.ident & SMC_ISM_IDENT_MASK); + snprintf(tmp, 5, "%04X", ident_tail); + memcpy(seid->serial_number, tmp, 4); + snprintf(tmp, 5, "%04X", id.machine); + memcpy(seid->type, tmp, 4); +#else + memset(seid, 0, SMC_MAX_EID_LEN); +#endif } -int smc_ism_write(struct smcd_dev *smcd, const struct smc_ism_position *pos, - void *data, size_t len) +/* Test if an ISM communication is possible - same CPC */ +int smc_ism_cantalk(struct smcd_gid *peer_gid, unsigned short vlan_id, + struct smcd_dev *smcd) { - int rc; - - rc = smcd->ops->move_data(smcd, pos->token, pos->index, pos->signal, - pos->offset, data, len); - - return rc < 0 ? rc : 0; + return smcd->ops->query_remote_gid(smcd, peer_gid, vlan_id ? 1 : 0, + vlan_id); } -void smc_ism_get_system_eid(struct smcd_dev *smcd, u8 **eid) +void smc_ism_get_system_eid(u8 **eid) { - smcd->ops->get_system_eid(smcd, eid); + if (!smc_ism_v2_capable) + *eid = NULL; + else + *eid = smc_ism_v2_system_eid; } u16 smc_ism_get_chid(struct smcd_dev *smcd) @@ -58,6 +90,11 @@ bool smc_ism_is_v2_capable(void) return smc_ism_v2_capable; } +void smc_ism_set_v2_capable(void) +{ + smc_ism_v2_capable = true; +} + /* Set a connection using this DMBE. */ void smc_ism_set_conn(struct smc_connection *conn) { @@ -93,6 +130,8 @@ int smc_ism_get_vlan(struct smcd_dev *smcd, unsigned short vlanid) if (!vlanid) /* No valid vlan id */ return -EINVAL; + if (!smcd->ops->add_vlan_id) + return -EOPNOTSUPP; /* create new vlan entry, in case we need it */ new_vlan = kzalloc(sizeof(*new_vlan), GFP_KERNEL); @@ -138,6 +177,8 @@ int smc_ism_put_vlan(struct smcd_dev *smcd, unsigned short vlanid) if (!vlanid) /* No valid vlan id */ return -EINVAL; + if (!smcd->ops->del_vlan_id) + return -EOPNOTSUPP; spin_lock_irqsave(&smcd->lock, flags); list_for_each_entry(vlan, &smcd->vlan, list) { @@ -196,8 +237,8 @@ int smc_ism_register_dmb(struct smc_link_group *lgr, int dmb_len, dmb.dmb_len = dmb_len; dmb.sba_idx = dmb_desc->sba_idx; dmb.vlan_id = lgr->vlan_id; - dmb.rgid = lgr->peer_gid; - rc = lgr->smcd->ops->register_dmb(lgr->smcd, &dmb); + dmb.rgid = lgr->peer_gid.gid; + rc = lgr->smcd->ops->register_dmb(lgr->smcd, &dmb, lgr->smcd->client); if (!rc) { dmb_desc->sba_idx = dmb.sba_idx; dmb_desc->token = dmb.dmb_tok; @@ -208,6 +249,46 @@ int smc_ism_register_dmb(struct smc_link_group *lgr, int dmb_len, return rc; } +bool smc_ism_support_dmb_nocopy(struct smcd_dev *smcd) +{ + /* for now only loopback-ism supports + * merging sndbuf with peer DMB to avoid + * data copies between them. + */ + return (smcd->ops->support_dmb_nocopy && + smcd->ops->support_dmb_nocopy(smcd)); +} + +int smc_ism_attach_dmb(struct smcd_dev *dev, u64 token, + struct smc_buf_desc *dmb_desc) +{ + struct smcd_dmb dmb; + int rc = 0; + + if (!dev->ops->attach_dmb) + return -EINVAL; + + memset(&dmb, 0, sizeof(dmb)); + dmb.dmb_tok = token; + rc = dev->ops->attach_dmb(dev, &dmb); + if (!rc) { + dmb_desc->sba_idx = dmb.sba_idx; + dmb_desc->token = dmb.dmb_tok; + dmb_desc->cpu_addr = dmb.cpu_addr; + dmb_desc->dma_addr = dmb.dma_addr; + dmb_desc->len = dmb.dmb_len; + } + return rc; +} + +int smc_ism_detach_dmb(struct smcd_dev *dev, u64 token) +{ + if (!dev->ops->detach_dmb) + return -EINVAL; + + return dev->ops->detach_dmb(dev, token); +} + static int smc_nl_handle_smcd_dev(struct smcd_dev *smcd, struct sk_buff *skb, struct netlink_callback *cb) @@ -216,9 +297,11 @@ static int smc_nl_handle_smcd_dev(struct smcd_dev *smcd, struct smc_pci_dev smc_pci_dev; struct nlattr *port_attrs; struct nlattr *attrs; + struct ism_dev *ism; int use_cnt = 0; void *nlh; + ism = smcd->priv; nlh = genlmsg_put(skb, NETLINK_CB(cb->skb).portid, cb->nlh->nlmsg_seq, &smc_gen_nl_family, NLM_F_MULTI, SMC_NETLINK_GET_DEV_SMCD); @@ -233,7 +316,7 @@ static int smc_nl_handle_smcd_dev(struct smcd_dev *smcd, if (nla_put_u8(skb, SMC_NLA_DEV_IS_CRIT, use_cnt > 0)) goto errattr; memset(&smc_pci_dev, 0, sizeof(smc_pci_dev)); - smc_set_pci_values(to_pci_dev(smcd->dev.parent), &smc_pci_dev); + smc_set_pci_values(to_pci_dev(ism->dev.parent), &smc_pci_dev); if (nla_put_u32(skb, SMC_NLA_DEV_PCI_FID, smc_pci_dev.pci_fid)) goto errattr; if (nla_put_u16(skb, SMC_NLA_DEV_PCI_CHID, smc_pci_dev.pci_pchid)) @@ -283,6 +366,8 @@ static void smc_nl_prep_smcd_dev(struct smcd_dev_list *dev_list, list_for_each_entry(smcd, &dev_list->list, list) { if (num < snum) goto next; + if (smc_ism_is_loopback(smcd)) + goto next; if (smc_nl_handle_smcd_dev(smcd, skb, cb)) goto errout; next: @@ -299,10 +384,11 @@ int smcd_nl_get_device(struct sk_buff *skb, struct netlink_callback *cb) return skb->len; } +#if IS_ENABLED(CONFIG_ISM) struct smc_ism_event_work { struct work_struct work; struct smcd_dev *smcd; - struct smcd_event event; + struct ism_event event; }; #define ISM_EVENT_REQUEST 0x0001 @@ -322,18 +408,21 @@ union smcd_sw_event_info { static void smcd_handle_sw_event(struct smc_ism_event_work *wrk) { + struct smcd_gid peer_gid = { .gid = wrk->event.tok, + .gid_ext = 0 }; union smcd_sw_event_info ev_info; ev_info.info = wrk->event.info; switch (wrk->event.code) { case ISM_EVENT_CODE_SHUTDOWN: /* Peer shut down DMBs */ - smc_smcd_terminate(wrk->smcd, wrk->event.tok, ev_info.vlan_id); + smc_smcd_terminate(wrk->smcd, &peer_gid, ev_info.vlan_id); break; case ISM_EVENT_CODE_TESTLINK: /* Activity timer */ - if (ev_info.code == ISM_EVENT_REQUEST) { + if (ev_info.code == ISM_EVENT_REQUEST && + wrk->smcd->ops->signal_event) { ev_info.code = ISM_EVENT_RESPONSE; wrk->smcd->ops->signal_event(wrk->smcd, - wrk->event.tok, + &peer_gid, ISM_EVENT_REQUEST_IR, ISM_EVENT_CODE_TESTLINK, ev_info.info); @@ -342,33 +431,17 @@ static void smcd_handle_sw_event(struct smc_ism_event_work *wrk) } } -int smc_ism_signal_shutdown(struct smc_link_group *lgr) -{ - int rc; - union smcd_sw_event_info ev_info; - - if (lgr->peer_shutdown) - return 0; - - memcpy(ev_info.uid, lgr->id, SMC_LGR_ID_SIZE); - ev_info.vlan_id = lgr->vlan_id; - ev_info.code = ISM_EVENT_REQUEST; - rc = lgr->smcd->ops->signal_event(lgr->smcd, lgr->peer_gid, - ISM_EVENT_REQUEST_IR, - ISM_EVENT_CODE_SHUTDOWN, - ev_info.info); - return rc; -} - /* worker for SMC-D events */ static void smc_ism_event_work(struct work_struct *work) { struct smc_ism_event_work *wrk = container_of(work, struct smc_ism_event_work, work); + struct smcd_gid smcd_gid = { .gid = wrk->event.tok, + .gid_ext = 0 }; switch (wrk->event.type) { case ISM_EVENT_GID: /* GID event, token is peer GID */ - smc_smcd_terminate(wrk->smcd, wrk->event.tok, VLAN_VID_MASK); + smc_smcd_terminate(wrk->smcd, &smcd_gid, VLAN_VID_MASK); break; case ISM_EVENT_DMB: break; @@ -379,44 +452,25 @@ static void smc_ism_event_work(struct work_struct *work) kfree(wrk); } -static void smcd_release(struct device *dev) -{ - struct smcd_dev *smcd = container_of(dev, struct smcd_dev, dev); - - kfree(smcd->conn); - kfree(smcd); -} - -struct smcd_dev *smcd_alloc_dev(struct device *parent, const char *name, - const struct smcd_ops *ops, int max_dmbs) +static struct smcd_dev *smcd_alloc_dev(struct device *parent, const char *name, + const struct smcd_ops *ops, int max_dmbs) { struct smcd_dev *smcd; - smcd = kzalloc(sizeof(*smcd), GFP_KERNEL); + smcd = devm_kzalloc(parent, sizeof(*smcd), GFP_KERNEL); if (!smcd) return NULL; - smcd->conn = kcalloc(max_dmbs, sizeof(struct smc_connection *), - GFP_KERNEL); - if (!smcd->conn) { - kfree(smcd); + smcd->conn = devm_kcalloc(parent, max_dmbs, + sizeof(struct smc_connection *), GFP_KERNEL); + if (!smcd->conn) return NULL; - } smcd->event_wq = alloc_ordered_workqueue("ism_evt_wq-%s)", WQ_MEM_RECLAIM, name); - if (!smcd->event_wq) { - kfree(smcd->conn); - kfree(smcd); + if (!smcd->event_wq) return NULL; - } - smcd->dev.parent = parent; - smcd->dev.release = smcd_release; - device_initialize(&smcd->dev); - dev_set_name(&smcd->dev, name); smcd->ops = ops; - if (smc_pnetid_by_dev_port(parent, 0, smcd->pnetid)) - smc_pnetid_by_table_smcd(smcd); spin_lock_init(&smcd->lock); spin_lock_init(&smcd->lgr_lock); @@ -425,66 +479,67 @@ struct smcd_dev *smcd_alloc_dev(struct device *parent, const char *name, init_waitqueue_head(&smcd->lgrs_deleted); return smcd; } -EXPORT_SYMBOL_GPL(smcd_alloc_dev); -int smcd_register_dev(struct smcd_dev *smcd) +static void smcd_register_dev(struct ism_dev *ism) { - int rc; + const struct smcd_ops *ops = ism_get_smcd_ops(); + struct smcd_dev *smcd, *fentry; - mutex_lock(&smcd_dev_list.mutex); - if (list_empty(&smcd_dev_list.list)) { - u8 *system_eid = NULL; + if (!ops) + return; - smc_ism_get_system_eid(smcd, &system_eid); - if (system_eid[24] != '0' || system_eid[28] != '0') - smc_ism_v2_capable = true; - } - /* sort list: devices without pnetid before devices with pnetid */ - if (smcd->pnetid[0]) + smcd = smcd_alloc_dev(&ism->pdev->dev, dev_name(&ism->pdev->dev), ops, + ISM_NR_DMBS); + if (!smcd) + return; + smcd->priv = ism; + smcd->client = &smc_ism_client; + ism_set_priv(ism, &smc_ism_client, smcd); + if (smc_pnetid_by_dev_port(&ism->pdev->dev, 0, smcd->pnetid)) + smc_pnetid_by_table_smcd(smcd); + + if (smcd->ops->supports_v2()) + smc_ism_set_v2_capable(); + mutex_lock(&smcd_dev_list.mutex); + /* sort list: + * - devices without pnetid before devices with pnetid; + * - loopback-ism always at the very beginning; + */ + if (!smcd->pnetid[0]) { + fentry = list_first_entry_or_null(&smcd_dev_list.list, + struct smcd_dev, list); + if (fentry && smc_ism_is_loopback(fentry)) + list_add(&smcd->list, &fentry->list); + else + list_add(&smcd->list, &smcd_dev_list.list); + } else { list_add_tail(&smcd->list, &smcd_dev_list.list); - else - list_add(&smcd->list, &smcd_dev_list.list); + } mutex_unlock(&smcd_dev_list.mutex); pr_warn_ratelimited("smc: adding smcd device %s with pnetid %.16s%s\n", - dev_name(&smcd->dev), smcd->pnetid, + dev_name(&ism->dev), smcd->pnetid, smcd->pnetid_by_user ? " (user defined)" : ""); - rc = device_add(&smcd->dev); - if (rc) { - mutex_lock(&smcd_dev_list.mutex); - list_del(&smcd->list); - mutex_unlock(&smcd_dev_list.mutex); - } - - return rc; + return; } -EXPORT_SYMBOL_GPL(smcd_register_dev); -void smcd_unregister_dev(struct smcd_dev *smcd) +static void smcd_unregister_dev(struct ism_dev *ism) { + struct smcd_dev *smcd = ism_get_priv(ism, &smc_ism_client); + pr_warn_ratelimited("smc: removing smcd device %s\n", - dev_name(&smcd->dev)); + dev_name(&ism->dev)); + smcd->going_away = 1; + smc_smcd_terminate_all(smcd); mutex_lock(&smcd_dev_list.mutex); list_del_init(&smcd->list); mutex_unlock(&smcd_dev_list.mutex); - smcd->going_away = 1; - smc_smcd_terminate_all(smcd); - flush_workqueue(smcd->event_wq); destroy_workqueue(smcd->event_wq); - - device_del(&smcd->dev); } -EXPORT_SYMBOL_GPL(smcd_unregister_dev); - -void smcd_free_dev(struct smcd_dev *smcd) -{ - put_device(&smcd->dev); -} -EXPORT_SYMBOL_GPL(smcd_free_dev); /* SMCD Device event handler. Called from ISM device interrupt handler. - * Parameters are smcd device pointer, + * Parameters are ism device pointer, * - event->type (0 --> DMB, 1 --> GID), * - event->code (event code), * - event->tok (either DMB token when event type 0, or GID when event type 1) @@ -494,8 +549,9 @@ EXPORT_SYMBOL_GPL(smcd_free_dev); * Context: * - Function called in IRQ context from ISM device driver event handler. */ -void smcd_handle_event(struct smcd_dev *smcd, struct smcd_event *event) +static void smcd_handle_event(struct ism_dev *ism, struct ism_event *event) { + struct smcd_dev *smcd = ism_get_priv(ism, &smc_ism_client); struct smc_ism_event_work *wrk; if (smcd->going_away) @@ -509,17 +565,18 @@ void smcd_handle_event(struct smcd_dev *smcd, struct smcd_event *event) wrk->event = *event; queue_work(smcd->event_wq, &wrk->work); } -EXPORT_SYMBOL_GPL(smcd_handle_event); /* SMCD Device interrupt handler. Called from ISM device interrupt handler. - * Parameters are smcd device pointer and DMB number. Find the connection and - * schedule the tasklet for this connection. + * Parameters are the ism device pointer, DMB number, and the DMBE bitmask. + * Find the connection and schedule the tasklet for this connection. * * Context: * - Function called in IRQ context from ISM device driver IRQ handler. */ -void smcd_handle_irq(struct smcd_dev *smcd, unsigned int dmbno) +static void smcd_handle_irq(struct ism_dev *ism, unsigned int dmbno, + u16 dmbemask) { + struct smcd_dev *smcd = ism_get_priv(ism, &smc_ism_client); struct smc_connection *conn = NULL; unsigned long flags; @@ -529,9 +586,46 @@ void smcd_handle_irq(struct smcd_dev *smcd, unsigned int dmbno) tasklet_schedule(&conn->rx_tsklet); spin_unlock_irqrestore(&smcd->lock, flags); } -EXPORT_SYMBOL_GPL(smcd_handle_irq); +#endif + +int smc_ism_signal_shutdown(struct smc_link_group *lgr) +{ + int rc = 0; +#if IS_ENABLED(CONFIG_ISM) + union smcd_sw_event_info ev_info; + + if (lgr->peer_shutdown) + return 0; + if (!lgr->smcd->ops->signal_event) + return 0; + + memcpy(ev_info.uid, lgr->id, SMC_LGR_ID_SIZE); + ev_info.vlan_id = lgr->vlan_id; + ev_info.code = ISM_EVENT_REQUEST; + rc = lgr->smcd->ops->signal_event(lgr->smcd, &lgr->peer_gid, + ISM_EVENT_REQUEST_IR, + ISM_EVENT_CODE_SHUTDOWN, + ev_info.info); +#endif + return rc; +} -void __init smc_ism_init(void) +int smc_ism_init(void) { + int rc = 0; + smc_ism_v2_capable = false; + smc_ism_create_system_eid(); + +#if IS_ENABLED(CONFIG_ISM) + rc = ism_register_client(&smc_ism_client); +#endif + return rc; +} + +void smc_ism_exit(void) +{ +#if IS_ENABLED(CONFIG_ISM) + ism_unregister_client(&smc_ism_client); +#endif } diff --git a/net/smc/smc_ism.h b/net/smc/smc_ism.h index 113efc7352ed67b953105a8bf7ee6a95d9313bc8..6763133dd8d047329ed22735689fa0dc9fbfcfcf 100644 --- a/net/smc/smc_ism.h +++ b/net/smc/smc_ism.h @@ -15,6 +15,9 @@ #include "smc.h" +#define SMC_EMULATED_ISM_CHID_MASK 0xFF00 +#define SMC_ISM_IDENT_MASK 0x00FFFF + struct smcd_dev_list { /* List of SMCD devices */ struct list_head list; struct mutex mutex; /* Protects list of devices */ @@ -28,16 +31,16 @@ struct smc_ism_vlanid { /* VLAN id set on ISM device */ refcount_t refcnt; /* Reference count */ }; -struct smc_ism_position { /* ISM device position to write to */ - u64 token; /* Token of DMB */ - u32 offset; /* Offset into DMBE */ - u8 index; /* Index of DMBE */ - u8 signal; /* Generate interrupt on owner side */ +struct smc_ism_seid { + u8 seid_string[24]; + u8 serial_number[4]; + u8 type[4]; }; struct smcd_dev; -int smc_ism_cantalk(u64 peer_gid, unsigned short vlan_id, struct smcd_dev *dev); +int smc_ism_cantalk(struct smcd_gid *peer_gid, unsigned short vlan_id, + struct smcd_dev *dev); void smc_ism_set_conn(struct smc_connection *conn); void smc_ism_unset_conn(struct smc_connection *conn); int smc_ism_get_vlan(struct smcd_dev *dev, unsigned short vlan_id); @@ -45,12 +48,50 @@ int smc_ism_put_vlan(struct smcd_dev *dev, unsigned short vlan_id); int smc_ism_register_dmb(struct smc_link_group *lgr, int buf_size, struct smc_buf_desc *dmb_desc); int smc_ism_unregister_dmb(struct smcd_dev *dev, struct smc_buf_desc *dmb_desc); -int smc_ism_write(struct smcd_dev *dev, const struct smc_ism_position *pos, - void *data, size_t len); +bool smc_ism_support_dmb_nocopy(struct smcd_dev *smcd); +int smc_ism_attach_dmb(struct smcd_dev *dev, u64 token, + struct smc_buf_desc *dmb_desc); +int smc_ism_detach_dmb(struct smcd_dev *dev, u64 token); int smc_ism_signal_shutdown(struct smc_link_group *lgr); -void smc_ism_get_system_eid(struct smcd_dev *dev, u8 **eid); +void smc_ism_get_system_eid(u8 **eid); u16 smc_ism_get_chid(struct smcd_dev *dev); bool smc_ism_is_v2_capable(void); -void smc_ism_init(void); +void smc_ism_set_v2_capable(void); +int smc_ism_init(void); +void smc_ism_exit(void); int smcd_nl_get_device(struct sk_buff *skb, struct netlink_callback *cb); + +static inline int smc_ism_write(struct smcd_dev *smcd, u64 dmb_tok, + unsigned int idx, bool sf, unsigned int offset, + void *data, size_t len) +{ + int rc; + + rc = smcd->ops->move_data(smcd, dmb_tok, idx, sf, offset, data, len); + return rc < 0 ? rc : 0; +} + +static inline bool __smc_ism_is_emulated(u16 chid) +{ + /* CHIDs in range of 0xFF00 to 0xFFFF are reserved + * for Emulated-ISM device. + * + * loopback-ism: 0xFFFF + * virtio-ism: 0xFF00 ~ 0xFFFE + */ + return ((chid & 0xFF00) == 0xFF00); +} + +static inline bool smc_ism_is_emulated(struct smcd_dev *smcd) +{ + u16 chid = smcd->ops->get_chid(smcd); + + return __smc_ism_is_emulated(chid); +} + +static inline bool smc_ism_is_loopback(struct smcd_dev *smcd) +{ + return (smcd->ops->get_chid(smcd) == 0xFFFF); +} + #endif diff --git a/net/smc/smc_llc.c b/net/smc/smc_llc.c index 97d10b933512e9b520597acca449d4a86ff18499..19fe78316ef988933f3999d25604872290cead34 100644 --- a/net/smc/smc_llc.c +++ b/net/smc/smc_llc.c @@ -23,16 +23,24 @@ struct smc_llc_hdr { struct smc_wr_rx_hdr common; - u8 length; /* 44 */ -#if defined(__BIG_ENDIAN_BITFIELD) - u8 reserved:4, - add_link_rej_rsn:4; + union { + struct { + u8 length; /* 44 */ + #if defined(__BIG_ENDIAN_BITFIELD) + u8 reserved:4, + add_link_rej_rsn:4; #elif defined(__LITTLE_ENDIAN_BITFIELD) - u8 add_link_rej_rsn:4, - reserved:4; + u8 add_link_rej_rsn:4, + reserved:4; #endif + }; + u16 length_v2; /* 44 - 8192*/ + }; u8 flags; -}; +} __packed; /* format defined in + * IBM Shared Memory Communications Version 2 + * (https://www.ibm.com/support/pages/node/6326337) + */ #define SMC_LLC_FLAG_NO_RMBE_EYEC 0x03 @@ -44,14 +52,13 @@ struct smc_llc_msg_confirm_link { /* type 0x01 */ u8 link_num; u8 link_uid[SMC_LGR_ID_SIZE]; u8 max_links; - u8 reserved[9]; + u8 max_conns; + u8 reserved[8]; }; #define SMC_LLC_FLAG_ADD_LNK_REJ 0x40 #define SMC_LLC_REJ_RSN_NO_ALT_PATH 1 -#define SMC_LLC_ADD_LNK_MAX_LINKS 2 - struct smc_llc_msg_add_link { /* type 0x02 */ struct smc_llc_hdr hd; u8 sender_mac[ETH_ALEN]; @@ -76,6 +83,32 @@ struct smc_llc_msg_add_link_cont_rt { __be64 rmb_vaddr_new; }; +struct smc_llc_msg_add_link_v2_ext { +#if defined(__BIG_ENDIAN_BITFIELD) + u8 v2_direct : 1, + reserved : 7; +#elif defined(__LITTLE_ENDIAN_BITFIELD) + u8 reserved : 7, + v2_direct : 1; +#endif + u8 reserved2; + u8 client_target_gid[SMC_GID_SIZE]; + u8 reserved3[8]; + u16 num_rkeys; + struct smc_llc_msg_add_link_cont_rt rt[]; +} __packed; /* format defined in + * IBM Shared Memory Communications Version 2 + * (https://www.ibm.com/support/pages/node/6326337) + */ + +struct smc_llc_msg_req_add_link_v2 { + struct smc_llc_hdr hd; + u8 reserved[20]; + u8 gid_cnt; + u8 reserved2[3]; + u8 gid[][SMC_GID_SIZE]; +}; + #define SMC_LLC_RKEYS_PER_CONT_MSG 2 struct smc_llc_msg_add_link_cont { /* type 0x03 */ @@ -114,7 +147,8 @@ struct smc_rmb_rtoken { __be64 rmb_vaddr; } __packed; /* format defined in RFC7609 */ -#define SMC_LLC_RKEYS_PER_MSG 3 +#define SMC_LLC_RKEYS_PER_MSG 3 +#define SMC_LLC_RKEYS_PER_MSG_V2 255 struct smc_llc_msg_confirm_rkey { /* type 0x06 */ struct smc_llc_hdr hd; @@ -135,9 +169,18 @@ struct smc_llc_msg_delete_rkey { /* type 0x09 */ u8 reserved2[4]; }; +struct smc_llc_msg_delete_rkey_v2 { /* type 0x29 */ + struct smc_llc_hdr hd; + u8 num_rkeys; + u8 num_inval_rkeys; + u8 reserved[2]; + __be32 rkey[]; +}; + union smc_llc_msg { struct smc_llc_msg_confirm_link confirm_link; struct smc_llc_msg_add_link add_link; + struct smc_llc_msg_req_add_link_v2 req_add_link; struct smc_llc_msg_add_link_cont add_link_cont; struct smc_llc_msg_del_link delete_link; @@ -189,7 +232,7 @@ static inline void smc_llc_flow_qentry_set(struct smc_llc_flow *flow, static void smc_llc_flow_parallel(struct smc_link_group *lgr, u8 flow_type, struct smc_llc_qentry *qentry) { - u8 msg_type = qentry->msg.raw.hdr.common.type; + u8 msg_type = qentry->msg.raw.hdr.common.llc_type; if ((msg_type == SMC_LLC_ADD_LINK || msg_type == SMC_LLC_DELETE_LINK) && flow_type != msg_type && !lgr->delayed_event) { @@ -198,9 +241,10 @@ static void smc_llc_flow_parallel(struct smc_link_group *lgr, u8 flow_type, } /* drop parallel or already-in-progress llc requests */ if (flow_type != msg_type) - pr_warn_once("smc: SMC-R lg %*phN dropped parallel " + pr_warn_once("smc: SMC-R lg %*phN net %llu dropped parallel " "LLC msg: msg %d flow %d role %d\n", SMC_LGR_ID_SIZE, &lgr->id, + lgr->net->net_cookie, qentry->msg.raw.hdr.common.type, flow_type, lgr->role); kfree(qentry); @@ -219,7 +263,7 @@ static bool smc_llc_flow_start(struct smc_llc_flow *flow, spin_unlock_bh(&lgr->llc_flow_lock); return false; } - switch (qentry->msg.raw.hdr.common.type) { + switch (qentry->msg.raw.hdr.common.llc_type) { case SMC_LLC_ADD_LINK: flow->type = SMC_LLC_FLOW_ADD_LINK; break; @@ -306,7 +350,7 @@ struct smc_llc_qentry *smc_llc_wait(struct smc_link_group *lgr, smc_llc_flow_qentry_del(flow); goto out; } - rcv_msg = flow->qentry->msg.raw.hdr.common.type; + rcv_msg = flow->qentry->msg.raw.hdr.common.llc_type; if (exp_msg && rcv_msg != exp_msg) { if (exp_msg == SMC_LLC_ADD_LINK && rcv_msg == SMC_LLC_DELETE_LINK) { @@ -315,9 +359,10 @@ struct smc_llc_qentry *smc_llc_wait(struct smc_link_group *lgr, smc_llc_flow_qentry_clr(flow)); return NULL; } - pr_warn_once("smc: SMC-R lg %*phN dropped unexpected LLC msg: " + pr_warn_once("smc: SMC-R lg %*phN net %llu dropped unexpected LLC msg: " "msg %d exp %d flow %d role %d flags %x\n", - SMC_LGR_ID_SIZE, &lgr->id, rcv_msg, exp_msg, + SMC_LGR_ID_SIZE, &lgr->id, lgr->net->net_cookie, + rcv_msg, exp_msg, flow->type, lgr->role, flow->qentry->msg.raw.hdr.flags); smc_llc_flow_qentry_del(flow); @@ -374,6 +419,30 @@ static int smc_llc_add_pending_send(struct smc_link *link, return 0; } +static int smc_llc_add_pending_send_v2(struct smc_link *link, + struct smc_wr_v2_buf **wr_buf, + struct smc_wr_tx_pend_priv **pend) +{ + int rc; + + rc = smc_wr_tx_get_v2_slot(link, smc_llc_tx_handler, wr_buf, pend); + if (rc < 0) + return rc; + return 0; +} + +static void smc_llc_init_msg_hdr(struct smc_llc_hdr *hdr, + struct smc_link_group *lgr, size_t len) +{ + if (lgr->smc_version == SMC_V2) { + hdr->common.llc_version = SMC_V2; + hdr->length_v2 = len; + } else { + hdr->common.llc_version = 0; + hdr->length = len; + } +} + /* high-level API to send LLC confirm link */ int smc_llc_send_confirm_link(struct smc_link *link, enum smc_llc_reqresp reqresp) @@ -390,8 +459,8 @@ int smc_llc_send_confirm_link(struct smc_link *link, goto put_out; confllc = (struct smc_llc_msg_confirm_link *)wr_buf; memset(confllc, 0, sizeof(*confllc)); - confllc->hd.common.type = SMC_LLC_CONFIRM_LINK; - confllc->hd.length = sizeof(struct smc_llc_msg_confirm_link); + confllc->hd.common.llc_type = SMC_LLC_CONFIRM_LINK; + smc_llc_init_msg_hdr(&confllc->hd, link->lgr, sizeof(*confllc)); confllc->hd.flags |= SMC_LLC_FLAG_NO_RMBE_EYEC; if (reqresp == SMC_LLC_RESP) confllc->hd.flags |= SMC_LLC_FLAG_RESP; @@ -401,7 +470,12 @@ int smc_llc_send_confirm_link(struct smc_link *link, hton24(confllc->sender_qp_num, link->roce_qp->qp_num); confllc->link_num = link->link_id; memcpy(confllc->link_uid, link->link_uid, SMC_LGR_ID_SIZE); - confllc->max_links = SMC_LLC_ADD_LNK_MAX_LINKS; + confllc->max_links = SMC_LINKS_ADD_LNK_MAX; + if (link->lgr->smc_version == SMC_V2 && + link->lgr->peer_smc_release >= SMC_RELEASE_1) { + confllc->max_conns = link->lgr->max_conns; + confllc->max_links = link->lgr->max_links; + } /* send llc message */ rc = smc_wr_tx_send(link, pend); put_out: @@ -426,8 +500,8 @@ static int smc_llc_send_confirm_rkey(struct smc_link *send_link, goto put_out; rkeyllc = (struct smc_llc_msg_confirm_rkey *)wr_buf; memset(rkeyllc, 0, sizeof(*rkeyllc)); - rkeyllc->hd.common.type = SMC_LLC_CONFIRM_RKEY; - rkeyllc->hd.length = sizeof(struct smc_llc_msg_confirm_rkey); + rkeyllc->hd.common.llc_type = SMC_LLC_CONFIRM_RKEY; + smc_llc_init_msg_hdr(&rkeyllc->hd, send_link->lgr, sizeof(*rkeyllc)); rtok_ix = 1; for (i = 0; i < SMC_LINKS_PER_LGR_MAX; i++) { @@ -474,8 +548,8 @@ static int smc_llc_send_delete_rkey(struct smc_link *link, goto put_out; rkeyllc = (struct smc_llc_msg_delete_rkey *)wr_buf; memset(rkeyllc, 0, sizeof(*rkeyllc)); - rkeyllc->hd.common.type = SMC_LLC_DELETE_RKEY; - rkeyllc->hd.length = sizeof(struct smc_llc_msg_delete_rkey); + rkeyllc->hd.common.llc_type = SMC_LLC_DELETE_RKEY; + smc_llc_init_msg_hdr(&rkeyllc->hd, link->lgr, sizeof(*rkeyllc)); rkeyllc->num_rkeys = 1; rkeyllc->rkey[0] = htonl(rmb_desc->mr[link->link_idx]->rkey); /* send llc message */ @@ -485,26 +559,119 @@ static int smc_llc_send_delete_rkey(struct smc_link *link, return rc; } +/* return first buffer from any of the next buf lists */ +static struct smc_buf_desc *_smc_llc_get_next_rmb(struct smc_link_group *lgr, + int *buf_lst) +{ + struct smc_buf_desc *buf_pos; + + while (*buf_lst < SMC_RMBE_SIZES) { + buf_pos = list_first_entry_or_null(&lgr->rmbs[*buf_lst], + struct smc_buf_desc, list); + if (buf_pos) + return buf_pos; + (*buf_lst)++; + } + return NULL; +} + +/* return next rmb from buffer lists */ +static struct smc_buf_desc *smc_llc_get_next_rmb(struct smc_link_group *lgr, + int *buf_lst, + struct smc_buf_desc *buf_pos) +{ + struct smc_buf_desc *buf_next; + + if (!buf_pos) + return _smc_llc_get_next_rmb(lgr, buf_lst); + + if (list_is_last(&buf_pos->list, &lgr->rmbs[*buf_lst])) { + (*buf_lst)++; + return _smc_llc_get_next_rmb(lgr, buf_lst); + } + buf_next = list_next_entry(buf_pos, list); + return buf_next; +} + +static struct smc_buf_desc *smc_llc_get_first_rmb(struct smc_link_group *lgr, + int *buf_lst) +{ + *buf_lst = 0; + return smc_llc_get_next_rmb(lgr, buf_lst, NULL); +} + +static int smc_llc_fill_ext_v2(struct smc_llc_msg_add_link_v2_ext *ext, + struct smc_link *link, struct smc_link *link_new) +{ + struct smc_link_group *lgr = link->lgr; + struct smc_buf_desc *buf_pos; + int prim_lnk_idx, lnk_idx, i; + struct smc_buf_desc *rmb; + int len = sizeof(*ext); + int buf_lst; + + ext->v2_direct = !lgr->uses_gateway; + memcpy(ext->client_target_gid, link_new->gid, SMC_GID_SIZE); + + prim_lnk_idx = link->link_idx; + lnk_idx = link_new->link_idx; + down_write(&lgr->rmbs_lock); + ext->num_rkeys = lgr->conns_num; + if (!ext->num_rkeys) + goto out; + buf_pos = smc_llc_get_first_rmb(lgr, &buf_lst); + for (i = 0; i < ext->num_rkeys; i++) { + while (buf_pos && !(buf_pos)->used) + buf_pos = smc_llc_get_next_rmb(lgr, &buf_lst, buf_pos); + if (!buf_pos) + break; + rmb = buf_pos; + ext->rt[i].rmb_key = htonl(rmb->mr[prim_lnk_idx]->rkey); + ext->rt[i].rmb_key_new = htonl(rmb->mr[lnk_idx]->rkey); + ext->rt[i].rmb_vaddr_new = + cpu_to_be64((u64)sg_dma_address(rmb->sgt[lnk_idx].sgl)); + buf_pos = smc_llc_get_next_rmb(lgr, &buf_lst, buf_pos); + } + len += i * sizeof(ext->rt[0]); +out: + up_write(&lgr->rmbs_lock); + return len; +} + /* send ADD LINK request or response */ int smc_llc_send_add_link(struct smc_link *link, u8 mac[], u8 gid[], struct smc_link *link_new, enum smc_llc_reqresp reqresp) { + struct smc_llc_msg_add_link_v2_ext *ext = NULL; struct smc_llc_msg_add_link *addllc; struct smc_wr_tx_pend_priv *pend; - struct smc_wr_buf *wr_buf; + int len = sizeof(*addllc); int rc; if (!smc_wr_tx_link_hold(link)) return -ENOLINK; - rc = smc_llc_add_pending_send(link, &wr_buf, &pend); - if (rc) - goto put_out; - addllc = (struct smc_llc_msg_add_link *)wr_buf; + if (link->lgr->smc_version == SMC_V2) { + struct smc_wr_v2_buf *wr_buf; + + rc = smc_llc_add_pending_send_v2(link, &wr_buf, &pend); + if (rc) + goto put_out; + addllc = (struct smc_llc_msg_add_link *)wr_buf; + ext = (struct smc_llc_msg_add_link_v2_ext *) + &wr_buf->raw[sizeof(*addllc)]; + memset(ext, 0, SMC_WR_TX_SIZE); + } else { + struct smc_wr_buf *wr_buf; + + rc = smc_llc_add_pending_send(link, &wr_buf, &pend); + if (rc) + goto put_out; + addllc = (struct smc_llc_msg_add_link *)wr_buf; + } memset(addllc, 0, sizeof(*addllc)); - addllc->hd.common.type = SMC_LLC_ADD_LINK; - addllc->hd.length = sizeof(struct smc_llc_msg_add_link); + addllc->hd.common.llc_type = SMC_LLC_ADD_LINK; if (reqresp == SMC_LLC_RESP) addllc->hd.flags |= SMC_LLC_FLAG_RESP; memcpy(addllc->sender_mac, mac, ETH_ALEN); @@ -519,8 +686,14 @@ int smc_llc_send_add_link(struct smc_link *link, u8 mac[], u8 gid[], addllc->qp_mtu = min(link_new->path_mtu, link_new->peer_mtu); } + if (ext && link_new) + len += smc_llc_fill_ext_v2(ext, link, link_new); + smc_llc_init_msg_hdr(&addllc->hd, link->lgr, len); /* send llc message */ - rc = smc_wr_tx_send(link, pend); + if (link->lgr->smc_version == SMC_V2) + rc = smc_wr_tx_v2_send(link, pend, len); + else + rc = smc_wr_tx_send(link, pend); put_out: smc_wr_tx_link_put(link); return rc; @@ -544,8 +717,8 @@ int smc_llc_send_delete_link(struct smc_link *link, u8 link_del_id, delllc = (struct smc_llc_msg_del_link *)wr_buf; memset(delllc, 0, sizeof(*delllc)); - delllc->hd.common.type = SMC_LLC_DELETE_LINK; - delllc->hd.length = sizeof(struct smc_llc_msg_del_link); + delllc->hd.common.llc_type = SMC_LLC_DELETE_LINK; + smc_llc_init_msg_hdr(&delllc->hd, link->lgr, sizeof(*delllc)); if (reqresp == SMC_LLC_RESP) delllc->hd.flags |= SMC_LLC_FLAG_RESP; if (orderly) @@ -577,8 +750,8 @@ static int smc_llc_send_test_link(struct smc_link *link, u8 user_data[16]) goto put_out; testllc = (struct smc_llc_msg_test_link *)wr_buf; memset(testllc, 0, sizeof(*testllc)); - testllc->hd.common.type = SMC_LLC_TEST_LINK; - testllc->hd.length = sizeof(struct smc_llc_msg_test_link); + testllc->hd.common.llc_type = SMC_LLC_TEST_LINK; + smc_llc_init_msg_hdr(&testllc->hd, link->lgr, sizeof(*testllc)); memcpy(testllc->user_data, user_data, sizeof(testllc->user_data)); /* send llc message */ rc = smc_wr_tx_send(link, pend); @@ -654,44 +827,6 @@ static int smc_llc_alloc_alt_link(struct smc_link_group *lgr, return -EMLINK; } -/* return first buffer from any of the next buf lists */ -static struct smc_buf_desc *_smc_llc_get_next_rmb(struct smc_link_group *lgr, - int *buf_lst) -{ - struct smc_buf_desc *buf_pos; - - while (*buf_lst < SMC_RMBE_SIZES) { - buf_pos = list_first_entry_or_null(&lgr->rmbs[*buf_lst], - struct smc_buf_desc, list); - if (buf_pos) - return buf_pos; - (*buf_lst)++; - } - return NULL; -} - -/* return next rmb from buffer lists */ -static struct smc_buf_desc *smc_llc_get_next_rmb(struct smc_link_group *lgr, - int *buf_lst, - struct smc_buf_desc *buf_pos) -{ - struct smc_buf_desc *buf_next; - - if (!buf_pos || list_is_last(&buf_pos->list, &lgr->rmbs[*buf_lst])) { - (*buf_lst)++; - return _smc_llc_get_next_rmb(lgr, buf_lst); - } - buf_next = list_next_entry(buf_pos, list); - return buf_next; -} - -static struct smc_buf_desc *smc_llc_get_first_rmb(struct smc_link_group *lgr, - int *buf_lst) -{ - *buf_lst = 0; - return smc_llc_get_next_rmb(lgr, buf_lst, NULL); -} - /* send one add_link_continue msg */ static int smc_llc_add_link_cont(struct smc_link *link, struct smc_link *link_new, u8 *num_rkeys_todo, @@ -738,7 +873,7 @@ static int smc_llc_add_link_cont(struct smc_link *link, (*num_rkeys_todo)--; *buf_pos = smc_llc_get_next_rmb(lgr, buf_lst, *buf_pos); } - addc_llc->hd.common.type = SMC_LLC_ADD_LINK_CONT; + addc_llc->hd.common.llc_type = SMC_LLC_ADD_LINK_CONT; addc_llc->hd.length = sizeof(struct smc_llc_msg_add_link_cont); if (lgr->role == SMC_CLNT) addc_llc->hd.flags |= SMC_LLC_FLAG_RESP; @@ -797,6 +932,8 @@ static int smc_llc_cli_add_link_reject(struct smc_llc_qentry *qentry) qentry->msg.raw.hdr.flags |= SMC_LLC_FLAG_RESP; qentry->msg.raw.hdr.flags |= SMC_LLC_FLAG_ADD_LNK_REJ; qentry->msg.raw.hdr.add_link_rej_rsn = SMC_LLC_REJ_RSN_NO_ALT_PATH; + smc_llc_init_msg_hdr(&qentry->msg.raw.hdr, qentry->link->lgr, + sizeof(qentry->msg)); return smc_llc_send_message(qentry->link, &qentry->msg); } @@ -817,7 +954,7 @@ static int smc_llc_cli_conf_link(struct smc_link *link, SMC_LLC_DEL_LOST_PATH); return -ENOLINK; } - if (qentry->msg.raw.hdr.common.type != SMC_LLC_CONFIRM_LINK) { + if (qentry->msg.raw.hdr.common.llc_type != SMC_LLC_CONFIRM_LINK) { /* received DELETE_LINK instead */ qentry->msg.raw.hdr.flags |= SMC_LLC_FLAG_RESP; smc_llc_send_message(link, &qentry->msg); @@ -858,6 +995,26 @@ static int smc_llc_cli_conf_link(struct smc_link *link, return 0; } +static void smc_llc_save_add_link_rkeys(struct smc_link *link, + struct smc_link *link_new) +{ + struct smc_llc_msg_add_link_v2_ext *ext; + struct smc_link_group *lgr = link->lgr; + int max, i; + + ext = (struct smc_llc_msg_add_link_v2_ext *)((u8 *)lgr->wr_rx_buf_v2 + + SMC_WR_TX_SIZE); + max = min_t(u8, ext->num_rkeys, SMC_LLC_RKEYS_PER_MSG_V2); + down_write(&lgr->rmbs_lock); + for (i = 0; i < max; i++) { + smc_rtoken_set(lgr, link->link_idx, link_new->link_idx, + ext->rt[i].rmb_key, + ext->rt[i].rmb_vaddr_new, + ext->rt[i].rmb_key_new); + } + up_write(&lgr->rmbs_lock); +} + static void smc_llc_save_add_link_info(struct smc_link *link, struct smc_llc_msg_add_link *add_llc) { @@ -874,31 +1031,52 @@ int smc_llc_cli_add_link(struct smc_link *link, struct smc_llc_qentry *qentry) struct smc_llc_msg_add_link *llc = &qentry->msg.add_link; enum smc_lgr_type lgr_new_t = SMC_LGR_SYMMETRIC; struct smc_link_group *lgr = smc_get_lgr(link); + struct smc_init_info *ini = NULL; struct smc_link *lnk_new = NULL; - struct smc_init_info ini; int lnk_idx, rc = 0; if (!llc->qp_mtu) goto out_reject; - ini.vlan_id = lgr->vlan_id; - smc_pnet_find_alt_roce(lgr, &ini, link->smcibdev); + ini = kzalloc(sizeof(*ini), GFP_KERNEL); + if (!ini) { + rc = -ENOMEM; + goto out_reject; + } + + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) { + rc = 0; + goto out_reject; + } + + ini->vlan_id = lgr->vlan_id; + if (lgr->smc_version == SMC_V2) { + ini->check_smcrv2 = true; + ini->smcrv2.saddr = lgr->saddr; + ini->smcrv2.daddr = smc_ib_gid_to_ipv4(llc->sender_gid); + } + smc_pnet_find_alt_roce(lgr, ini, link->smcibdev); if (!memcmp(llc->sender_gid, link->peer_gid, SMC_GID_SIZE) && - !memcmp(llc->sender_mac, link->peer_mac, ETH_ALEN)) { - if (!ini.ib_dev) + (lgr->smc_version == SMC_V2 || + !memcmp(llc->sender_mac, link->peer_mac, ETH_ALEN))) { + if (!ini->ib_dev && !ini->smcrv2.ib_dev_v2) goto out_reject; lgr_new_t = SMC_LGR_ASYMMETRIC_PEER; } - if (!ini.ib_dev) { + if (lgr->smc_version == SMC_V2 && !ini->smcrv2.ib_dev_v2) { + lgr_new_t = SMC_LGR_ASYMMETRIC_LOCAL; + ini->smcrv2.ib_dev_v2 = link->smcibdev; + ini->smcrv2.ib_port_v2 = link->ibport; + } else if (lgr->smc_version < SMC_V2 && !ini->ib_dev) { lgr_new_t = SMC_LGR_ASYMMETRIC_LOCAL; - ini.ib_dev = link->smcibdev; - ini.ib_port = link->ibport; + ini->ib_dev = link->smcibdev; + ini->ib_port = link->ibport; } lnk_idx = smc_llc_alloc_alt_link(lgr, lgr_new_t); if (lnk_idx < 0) goto out_reject; lnk_new = &lgr->lnk[lnk_idx]; - rc = smcr_link_init(lgr, lnk_new, lnk_idx, &ini); + rc = smcr_link_init(lgr, lnk_new, lnk_idx, ini); if (rc) goto out_reject; smc_llc_save_add_link_info(lnk_new, llc); @@ -914,16 +1092,20 @@ int smc_llc_cli_add_link(struct smc_link *link, struct smc_llc_qentry *qentry) goto out_clear_lnk; rc = smc_llc_send_add_link(link, - lnk_new->smcibdev->mac[ini.ib_port - 1], + lnk_new->smcibdev->mac[lnk_new->ibport - 1], lnk_new->gid, lnk_new, SMC_LLC_RESP); if (rc) goto out_clear_lnk; - rc = smc_llc_cli_rkey_exchange(link, lnk_new); - if (rc) { - rc = 0; - goto out_clear_lnk; + if (lgr->smc_version == SMC_V2) { + smc_llc_save_add_link_rkeys(link, lnk_new); + } else { + rc = smc_llc_cli_rkey_exchange(link, lnk_new); + if (rc) { + rc = 0; + goto out_clear_lnk; + } } - rc = smc_llc_cli_conf_link(link, &ini, lnk_new, lgr_new_t); + rc = smc_llc_cli_conf_link(link, ini, lnk_new, lgr_new_t); if (!rc) goto out; out_clear_lnk: @@ -932,29 +1114,81 @@ int smc_llc_cli_add_link(struct smc_link *link, struct smc_llc_qentry *qentry) out_reject: smc_llc_cli_add_link_reject(qentry); out: + kfree(ini); kfree(qentry); return rc; } +static void smc_llc_send_request_add_link(struct smc_link *link) +{ + struct smc_llc_msg_req_add_link_v2 *llc; + struct smc_wr_tx_pend_priv *pend; + struct smc_wr_v2_buf *wr_buf; + struct smc_gidlist gidlist; + int rc, len, i; + + if (!smc_wr_tx_link_hold(link)) + return; + if (link->lgr->type == SMC_LGR_SYMMETRIC || + link->lgr->type == SMC_LGR_ASYMMETRIC_PEER) + goto put_out; + + smc_fill_gid_list(link->lgr, &gidlist, link->smcibdev, link->gid); + if (gidlist.len <= 1) + goto put_out; + + rc = smc_llc_add_pending_send_v2(link, &wr_buf, &pend); + if (rc) + goto put_out; + llc = (struct smc_llc_msg_req_add_link_v2 *)wr_buf; + memset(llc, 0, SMC_WR_TX_SIZE); + + llc->hd.common.llc_type = SMC_LLC_REQ_ADD_LINK; + for (i = 0; i < gidlist.len; i++) + memcpy(llc->gid[i], gidlist.list[i], sizeof(gidlist.list[0])); + llc->gid_cnt = gidlist.len; + len = sizeof(*llc) + (gidlist.len * sizeof(gidlist.list[0])); + smc_llc_init_msg_hdr(&llc->hd, link->lgr, len); + rc = smc_wr_tx_v2_send(link, pend, len); + if (!rc) + /* set REQ_ADD_LINK flow and wait for response from peer */ + link->lgr->llc_flow_lcl.type = SMC_LLC_FLOW_REQ_ADD_LINK; +put_out: + smc_wr_tx_link_put(link); +} + /* as an SMC client, invite server to start the add_link processing */ static void smc_llc_cli_add_link_invite(struct smc_link *link, struct smc_llc_qentry *qentry) { struct smc_link_group *lgr = smc_get_lgr(link); - struct smc_init_info ini; + struct smc_init_info *ini = NULL; + + if (lgr->smc_version == SMC_V2) { + smc_llc_send_request_add_link(link); + goto out; + } if (lgr->type == SMC_LGR_SYMMETRIC || lgr->type == SMC_LGR_ASYMMETRIC_PEER) goto out; - ini.vlan_id = lgr->vlan_id; - smc_pnet_find_alt_roce(lgr, &ini, link->smcibdev); - if (!ini.ib_dev) + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) + goto out; + + ini = kzalloc(sizeof(*ini), GFP_KERNEL); + if (!ini) + goto out; + + ini->vlan_id = lgr->vlan_id; + smc_pnet_find_alt_roce(lgr, ini, link->smcibdev); + if (!ini->ib_dev) goto out; - smc_llc_send_add_link(link, ini.ib_dev->mac[ini.ib_port - 1], - ini.ib_gid, NULL, SMC_LLC_REQ); + smc_llc_send_add_link(link, ini->ib_dev->mac[ini->ib_port - 1], + ini->ib_gid, NULL, SMC_LLC_REQ); out: + kfree(ini); kfree(qentry); } @@ -970,7 +1204,7 @@ static bool smc_llc_is_empty_llc_message(union smc_llc_msg *llc) static bool smc_llc_is_local_add_link(union smc_llc_msg *llc) { - if (llc->raw.hdr.common.type == SMC_LLC_ADD_LINK && + if (llc->raw.hdr.common.llc_type == SMC_LLC_ADD_LINK && smc_llc_is_empty_llc_message(llc)) return true; return false; @@ -1137,7 +1371,7 @@ static int smc_llc_srv_conf_link(struct smc_link *link, /* receive CONFIRM LINK response over the RoCE fabric */ qentry = smc_llc_wait(lgr, link, SMC_LLC_WAIT_FIRST_TIME, 0); if (!qentry || - qentry->msg.raw.hdr.common.type != SMC_LLC_CONFIRM_LINK) { + qentry->msg.raw.hdr.common.llc_type != SMC_LLC_CONFIRM_LINK) { /* send DELETE LINK */ smc_llc_send_delete_link(link, link_new->link_id, SMC_LLC_REQ, false, SMC_LLC_DEL_LOST_PATH); @@ -1156,37 +1390,85 @@ static int smc_llc_srv_conf_link(struct smc_link *link, return 0; } -int smc_llc_srv_add_link(struct smc_link *link) +static void smc_llc_send_req_add_link_response(struct smc_llc_qentry *qentry) +{ + qentry->msg.raw.hdr.flags |= SMC_LLC_FLAG_RESP; + smc_llc_init_msg_hdr(&qentry->msg.raw.hdr, qentry->link->lgr, + sizeof(qentry->msg)); + memset(&qentry->msg.raw.data, 0, sizeof(qentry->msg.raw.data)); + smc_llc_send_message(qentry->link, &qentry->msg); +} + +int smc_llc_srv_add_link(struct smc_link *link, + struct smc_llc_qentry *req_qentry) { enum smc_lgr_type lgr_new_t = SMC_LGR_SYMMETRIC; struct smc_link_group *lgr = link->lgr; struct smc_llc_msg_add_link *add_llc; struct smc_llc_qentry *qentry = NULL; - struct smc_link *link_new; - struct smc_init_info ini; + bool send_req_add_link_resp = false; + struct smc_link *link_new = NULL; + struct smc_init_info *ini = NULL; int lnk_idx, rc = 0; + if (req_qentry && + req_qentry->msg.raw.hdr.common.llc_type == SMC_LLC_REQ_ADD_LINK) + send_req_add_link_resp = true; + + ini = kzalloc(sizeof(*ini), GFP_KERNEL); + if (!ini) { + rc = -ENOMEM; + goto out; + } + + if (lgr->type == SMC_LGR_SINGLE && lgr->max_links <= 1) { + rc = 0; + goto out; + } + /* ignore client add link recommendation, start new flow */ - ini.vlan_id = lgr->vlan_id; - smc_pnet_find_alt_roce(lgr, &ini, link->smcibdev); - if (!ini.ib_dev) { + ini->vlan_id = lgr->vlan_id; + if (lgr->smc_version == SMC_V2) { + ini->check_smcrv2 = true; + ini->smcrv2.saddr = lgr->saddr; + if (send_req_add_link_resp) { + struct smc_llc_msg_req_add_link_v2 *req_add = + &req_qentry->msg.req_add_link; + + ini->smcrv2.daddr = smc_ib_gid_to_ipv4(req_add->gid[0]); + } + } + smc_pnet_find_alt_roce(lgr, ini, link->smcibdev); + if (lgr->smc_version == SMC_V2 && !ini->smcrv2.ib_dev_v2) { lgr_new_t = SMC_LGR_ASYMMETRIC_LOCAL; - ini.ib_dev = link->smcibdev; - ini.ib_port = link->ibport; + ini->smcrv2.ib_dev_v2 = link->smcibdev; + ini->smcrv2.ib_port_v2 = link->ibport; + } else if (lgr->smc_version < SMC_V2 && !ini->ib_dev) { + lgr_new_t = SMC_LGR_ASYMMETRIC_LOCAL; + ini->ib_dev = link->smcibdev; + ini->ib_port = link->ibport; } lnk_idx = smc_llc_alloc_alt_link(lgr, lgr_new_t); - if (lnk_idx < 0) - return 0; + if (lnk_idx < 0) { + rc = 0; + goto out; + } - rc = smcr_link_init(lgr, &lgr->lnk[lnk_idx], lnk_idx, &ini); + rc = smcr_link_init(lgr, &lgr->lnk[lnk_idx], lnk_idx, ini); if (rc) - return rc; + goto out; link_new = &lgr->lnk[lnk_idx]; + + rc = smcr_buf_map_lgr(link_new); + if (rc) + goto out_err; + rc = smc_llc_send_add_link(link, - link_new->smcibdev->mac[ini.ib_port - 1], + link_new->smcibdev->mac[link_new->ibport-1], link_new->gid, link_new, SMC_LLC_REQ); if (rc) goto out_err; + send_req_add_link_resp = false; /* receive ADD LINK response over the RoCE fabric */ qentry = smc_llc_wait(lgr, link, SMC_LLC_WAIT_TIME, SMC_LLC_ADD_LINK); if (!qentry) { @@ -1201,48 +1483,59 @@ int smc_llc_srv_add_link(struct smc_link *link) } if (lgr->type == SMC_LGR_SINGLE && (!memcmp(add_llc->sender_gid, link->peer_gid, SMC_GID_SIZE) && - !memcmp(add_llc->sender_mac, link->peer_mac, ETH_ALEN))) { + (lgr->smc_version == SMC_V2 || + !memcmp(add_llc->sender_mac, link->peer_mac, ETH_ALEN)))) { lgr_new_t = SMC_LGR_ASYMMETRIC_PEER; } smc_llc_save_add_link_info(link_new, add_llc); smc_llc_flow_qentry_del(&lgr->llc_flow_lcl); rc = smc_ib_ready_link(link_new); - if (rc) - goto out_err; - rc = smcr_buf_map_lgr(link_new); if (rc) goto out_err; rc = smcr_buf_reg_lgr(link_new); if (rc) goto out_err; - rc = smc_llc_srv_rkey_exchange(link, link_new); - if (rc) - goto out_err; + if (lgr->smc_version == SMC_V2) { + smc_llc_save_add_link_rkeys(link, link_new); + } else { + rc = smc_llc_srv_rkey_exchange(link, link_new); + if (rc) + goto out_err; + } rc = smc_llc_srv_conf_link(link, link_new, lgr_new_t); if (rc) goto out_err; + kfree(ini); return 0; out_err: - link_new->state = SMC_LNK_INACTIVE; - smcr_link_clear(link_new, false); + if (link_new) { + link_new->state = SMC_LNK_INACTIVE; + smcr_link_clear(link_new, false); + } +out: + kfree(ini); + if (send_req_add_link_resp) + smc_llc_send_req_add_link_response(req_qentry); return rc; } static void smc_llc_process_srv_add_link(struct smc_link_group *lgr) { struct smc_link *link = lgr->llc_flow_lcl.qentry->link; + struct smc_llc_qentry *qentry; int rc; - smc_llc_flow_qentry_del(&lgr->llc_flow_lcl); + qentry = smc_llc_flow_qentry_clr(&lgr->llc_flow_lcl); down_write(&lgr->llc_conf_lock); - rc = smc_llc_srv_add_link(link); + rc = smc_llc_srv_add_link(link, qentry); if (!rc && lgr->type == SMC_LGR_SYMMETRIC) { /* delete any asymmetric link */ smc_llc_delete_asym_link(lgr); } up_write(&lgr->llc_conf_lock); + kfree(qentry); } /* enqueue a local add_link req to trigger a new add_link flow */ @@ -1250,8 +1543,8 @@ void smc_llc_add_link_local(struct smc_link *link) { struct smc_llc_msg_add_link add_llc = {}; - add_llc.hd.length = sizeof(add_llc); - add_llc.hd.common.type = SMC_LLC_ADD_LINK; + add_llc.hd.common.llc_type = SMC_LLC_ADD_LINK; + smc_llc_init_msg_hdr(&add_llc.hd, link->lgr, sizeof(add_llc)); /* no dev and port needed */ smc_llc_enqueue(link, (union smc_llc_msg *)&add_llc); } @@ -1273,7 +1566,8 @@ static void smc_llc_add_link_work(struct work_struct *work) else smc_llc_process_srv_add_link(lgr); out: - smc_llc_flow_stop(lgr, &lgr->llc_flow_lcl); + if (lgr->llc_flow_lcl.type != SMC_LLC_FLOW_REQ_ADD_LINK) + smc_llc_flow_stop(lgr, &lgr->llc_flow_lcl); } /* enqueue a local del_link msg to trigger a new del_link flow, @@ -1283,8 +1577,8 @@ void smc_llc_srv_delete_link_local(struct smc_link *link, u8 del_link_id) { struct smc_llc_msg_del_link del_llc = {}; - del_llc.hd.length = sizeof(del_llc); - del_llc.hd.common.type = SMC_LLC_DELETE_LINK; + del_llc.hd.common.llc_type = SMC_LLC_DELETE_LINK; + smc_llc_init_msg_hdr(&del_llc.hd, link->lgr, sizeof(del_llc)); del_llc.link_num = del_link_id; del_llc.reason = htonl(SMC_LLC_DEL_LOST_PATH); del_llc.hd.flags |= SMC_LLC_FLAG_DEL_LINK_ORDERLY; @@ -1354,8 +1648,8 @@ void smc_llc_send_link_delete_all(struct smc_link_group *lgr, bool ord, u32 rsn) struct smc_llc_msg_del_link delllc = {}; int i; - delllc.hd.common.type = SMC_LLC_DELETE_LINK; - delllc.hd.length = sizeof(delllc); + delllc.hd.common.llc_type = SMC_LLC_DELETE_LINK; + smc_llc_init_msg_hdr(&delllc.hd, lgr, sizeof(delllc)); if (ord) delllc.hd.flags |= SMC_LLC_FLAG_DEL_LINK_ORDERLY; delllc.hd.flags |= SMC_LLC_FLAG_DEL_LINK_ALL; @@ -1471,6 +1765,8 @@ static void smc_llc_rmt_conf_rkey(struct smc_link_group *lgr) link = qentry->link; num_entries = llc->rtoken[0].num_rkeys; + if (num_entries > SMC_LLC_RKEYS_PER_MSG) + goto out_err; /* first rkey entry is for receiving link */ rk_idx = smc_rtoken_add(link, llc->rtoken[0].rmb_vaddr, @@ -1489,6 +1785,7 @@ static void smc_llc_rmt_conf_rkey(struct smc_link_group *lgr) llc->hd.flags |= SMC_LLC_FLAG_RKEY_RETRY; out: llc->hd.flags |= SMC_LLC_FLAG_RESP; + smc_llc_init_msg_hdr(&llc->hd, link->lgr, sizeof(*llc)); smc_llc_send_message(link, &qentry->msg); smc_llc_flow_qentry_del(&lgr->llc_flow_rmt); } @@ -1506,6 +1803,28 @@ static void smc_llc_rmt_delete_rkey(struct smc_link_group *lgr) llc = &qentry->msg.delete_rkey; link = qentry->link; + if (lgr->smc_version == SMC_V2) { + struct smc_llc_msg_delete_rkey_v2 *llcv2; + + memcpy(lgr->wr_rx_buf_v2, llc, sizeof(*llc)); + llcv2 = (struct smc_llc_msg_delete_rkey_v2 *)lgr->wr_rx_buf_v2; + llcv2->num_inval_rkeys = 0; + + max = min_t(u8, llcv2->num_rkeys, SMC_LLC_RKEYS_PER_MSG_V2); + for (i = 0; i < max; i++) { + if (smc_rtoken_delete(link, llcv2->rkey[i])) + llcv2->num_inval_rkeys++; + } + memset(&llc->rkey[0], 0, sizeof(llc->rkey)); + memset(&llc->reserved2, 0, sizeof(llc->reserved2)); + smc_llc_init_msg_hdr(&llc->hd, link->lgr, sizeof(*llc)); + if (llcv2->num_inval_rkeys) { + llc->hd.flags |= SMC_LLC_FLAG_RKEY_NEG; + llc->err_mask = llcv2->num_inval_rkeys; + } + goto finish; + } + max = min_t(u8, llc->num_rkeys, SMC_LLC_DEL_RKEY_MAX); for (i = 0; i < max; i++) { if (smc_rtoken_delete(link, llc->rkey[i])) @@ -1515,6 +1834,7 @@ static void smc_llc_rmt_delete_rkey(struct smc_link_group *lgr) llc->hd.flags |= SMC_LLC_FLAG_RKEY_NEG; llc->err_mask = err_mask; } +finish: llc->hd.flags |= SMC_LLC_FLAG_RESP; smc_llc_send_message(link, &qentry->msg); smc_llc_flow_qentry_del(&lgr->llc_flow_rmt); @@ -1522,8 +1842,9 @@ static void smc_llc_rmt_delete_rkey(struct smc_link_group *lgr) static void smc_llc_protocol_violation(struct smc_link_group *lgr, u8 type) { - pr_warn_ratelimited("smc: SMC-R lg %*phN LLC protocol violation: " - "llc_type %d\n", SMC_LGR_ID_SIZE, &lgr->id, type); + pr_warn_ratelimited("smc: SMC-R lg %*phN net %llu LLC protocol violation: " + "llc_type %d\n", SMC_LGR_ID_SIZE, &lgr->id, + lgr->net->net_cookie, type); smc_llc_set_termination_rsn(lgr, SMC_LLC_DEL_PROT_VIOL); smc_lgr_terminate_sched(lgr); } @@ -1550,7 +1871,7 @@ static void smc_llc_event_handler(struct smc_llc_qentry *qentry) if (!smc_link_usable(link)) goto out; - switch (llc->raw.hdr.common.type) { + switch (llc->raw.hdr.common.llc_type) { case SMC_LLC_TEST_LINK: llc->test_link.hd.flags |= SMC_LLC_FLAG_RESP; smc_llc_send_message(link, llc); @@ -1575,8 +1896,18 @@ static void smc_llc_event_handler(struct smc_llc_qentry *qentry) smc_llc_flow_qentry_set(&lgr->llc_flow_lcl, qentry); wake_up(&lgr->llc_msg_waiter); - } else if (smc_llc_flow_start(&lgr->llc_flow_lcl, - qentry)) { + return; + } + if (lgr->llc_flow_lcl.type == + SMC_LLC_FLOW_REQ_ADD_LINK) { + /* server started add_link processing */ + lgr->llc_flow_lcl.type = SMC_LLC_FLOW_ADD_LINK; + smc_llc_flow_qentry_set(&lgr->llc_flow_lcl, + qentry); + schedule_work(&lgr->llc_add_link_work); + return; + } + if (smc_llc_flow_start(&lgr->llc_flow_lcl, qentry)) { schedule_work(&lgr->llc_add_link_work); } } else if (smc_llc_flow_start(&lgr->llc_flow_lcl, qentry)) { @@ -1624,6 +1955,23 @@ static void smc_llc_event_handler(struct smc_llc_qentry *qentry) smc_llc_flow_stop(lgr, &lgr->llc_flow_rmt); } return; + case SMC_LLC_REQ_ADD_LINK: + /* handle response here, smc_llc_flow_stop() cannot be called + * in tasklet context + */ + if (lgr->role == SMC_CLNT && + lgr->llc_flow_lcl.type == SMC_LLC_FLOW_REQ_ADD_LINK && + (llc->raw.hdr.flags & SMC_LLC_FLAG_RESP)) { + smc_llc_flow_stop(link->lgr, &lgr->llc_flow_lcl); + } else if (lgr->role == SMC_SERV) { + if (smc_llc_flow_start(&lgr->llc_flow_lcl, qentry)) { + /* as smc server, handle client suggestion */ + lgr->llc_flow_lcl.type = SMC_LLC_FLOW_ADD_LINK; + schedule_work(&lgr->llc_add_link_work); + } + return; + } + break; default: smc_llc_protocol_violation(lgr, llc->raw.hdr.common.type); break; @@ -1667,7 +2015,7 @@ static void smc_llc_rx_response(struct smc_link *link, { enum smc_llc_flowtype flowtype = link->lgr->llc_flow_lcl.type; struct smc_llc_flow *flow = &link->lgr->llc_flow_lcl; - u8 llc_type = qentry->msg.raw.hdr.common.type; + u8 llc_type = qentry->msg.raw.hdr.common.llc_type; switch (llc_type) { case SMC_LLC_TEST_LINK: @@ -1693,7 +2041,8 @@ static void smc_llc_rx_response(struct smc_link *link, /* not used because max links is 3 */ break; default: - smc_llc_protocol_violation(link->lgr, llc_type); + smc_llc_protocol_violation(link->lgr, + qentry->msg.raw.hdr.common.type); break; } kfree(qentry); @@ -1718,7 +2067,8 @@ static void smc_llc_enqueue(struct smc_link *link, union smc_llc_msg *llc) memcpy(&qentry->msg, llc, sizeof(union smc_llc_msg)); /* process responses immediately */ - if (llc->raw.hdr.flags & SMC_LLC_FLAG_RESP) { + if ((llc->raw.hdr.flags & SMC_LLC_FLAG_RESP) && + llc->raw.hdr.common.llc_type != SMC_LLC_REQ_ADD_LINK) { smc_llc_rx_response(link, qentry); return; } @@ -1738,8 +2088,13 @@ static void smc_llc_rx_handler(struct ib_wc *wc, void *buf) if (wc->byte_len < sizeof(*llc)) return; /* short message */ - if (llc->raw.hdr.length != sizeof(*llc)) - return; /* invalid message */ + if (!llc->raw.hdr.common.llc_version) { + if (llc->raw.hdr.length != sizeof(*llc)) + return; /* invalid message */ + } else { + if (llc->raw.hdr.length_v2 < sizeof(*llc)) + return; /* invalid message */ + } smc_llc_enqueue(link, llc); } @@ -1818,9 +2173,10 @@ int smc_llc_link_init(struct smc_link *link) void smc_llc_link_active(struct smc_link *link) { - pr_warn_ratelimited("smc: SMC-R lg %*phN link added: id %*phN, " + pr_warn_ratelimited("smc: SMC-R lg %*phN net %llu link added: id %*phN, " "peerid %*phN, ibdev %s, ibport %d\n", SMC_LGR_ID_SIZE, &link->lgr->id, + link->lgr->net->net_cookie, SMC_LGR_ID_SIZE, &link->link_uid, SMC_LGR_ID_SIZE, &link->peer_link_uid, link->smcibdev->ibdev->name, link->ibport); @@ -1836,9 +2192,10 @@ void smc_llc_link_active(struct smc_link *link) void smc_llc_link_clear(struct smc_link *link, bool log) { if (log) - pr_warn_ratelimited("smc: SMC-R lg %*phN link removed: id %*phN" + pr_warn_ratelimited("smc: SMC-R lg %*phN net %llu link removed: id %*phN" ", peerid %*phN, ibdev %s, ibport %d\n", SMC_LGR_ID_SIZE, &link->lgr->id, + link->lgr->net->net_cookie, SMC_LGR_ID_SIZE, &link->link_uid, SMC_LGR_ID_SIZE, &link->peer_link_uid, link->smcibdev->ibdev->name, link->ibport); @@ -1958,6 +2315,35 @@ static struct smc_wr_rx_handler smc_llc_rx_handlers[] = { .handler = smc_llc_rx_handler, .type = SMC_LLC_DELETE_RKEY }, + /* V2 types */ + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_CONFIRM_LINK_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_TEST_LINK_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_ADD_LINK_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_DELETE_LINK_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_REQ_ADD_LINK_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_CONFIRM_RKEY_V2 + }, + { + .handler = smc_llc_rx_handler, + .type = SMC_LLC_DELETE_RKEY_V2 + }, { .handler = NULL, } diff --git a/net/smc/smc_llc.h b/net/smc/smc_llc.h index cc00a2ec4e92e8060bce1926169832f7f7d96eb2..4404e52b3346fbc51855667fb4832f32a9987ef9 100644 --- a/net/smc/smc_llc.h +++ b/net/smc/smc_llc.h @@ -30,10 +30,19 @@ enum smc_llc_msg_type { SMC_LLC_ADD_LINK = 0x02, SMC_LLC_ADD_LINK_CONT = 0x03, SMC_LLC_DELETE_LINK = 0x04, + SMC_LLC_REQ_ADD_LINK = 0x05, SMC_LLC_CONFIRM_RKEY = 0x06, SMC_LLC_TEST_LINK = 0x07, SMC_LLC_CONFIRM_RKEY_CONT = 0x08, SMC_LLC_DELETE_RKEY = 0x09, + /* V2 types */ + SMC_LLC_CONFIRM_LINK_V2 = 0x21, + SMC_LLC_ADD_LINK_V2 = 0x22, + SMC_LLC_DELETE_LINK_V2 = 0x24, + SMC_LLC_REQ_ADD_LINK_V2 = 0x25, + SMC_LLC_CONFIRM_RKEY_V2 = 0x26, + SMC_LLC_TEST_LINK_V2 = 0x27, + SMC_LLC_DELETE_RKEY_V2 = 0x29, }; #define smc_link_downing(state) \ @@ -102,7 +111,8 @@ void smc_llc_flow_qentry_del(struct smc_llc_flow *flow); void smc_llc_send_link_delete_all(struct smc_link_group *lgr, bool ord, u32 rsn); int smc_llc_cli_add_link(struct smc_link *link, struct smc_llc_qentry *qentry); -int smc_llc_srv_add_link(struct smc_link *link); +int smc_llc_srv_add_link(struct smc_link *link, + struct smc_llc_qentry *req_qentry); void smc_llc_add_link_local(struct smc_link *link); int smc_llc_init(void) __init; diff --git a/net/smc/smc_loopback.c b/net/smc/smc_loopback.c new file mode 100644 index 0000000000000000000000000000000000000000..3c5f64ca41153f63fdb0c0f134952a5a6afaebf6 --- /dev/null +++ b/net/smc/smc_loopback.c @@ -0,0 +1,427 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Shared Memory Communications Direct over loopback-ism device. + * + * Functions for loopback-ism device. + * + * Copyright (c) 2024, Alibaba Inc. + * + * Author: Wen Gu + * Tony Lu + * + */ + +#include +#include +#include + +#include "smc_cdc.h" +#include "smc_ism.h" +#include "smc_loopback.h" + +#define SMC_LO_V2_CAPABLE 0x1 /* loopback-ism acts as ISMv2 */ +#define SMC_LO_SUPPORT_NOCOPY 0x1 +#define SMC_DMA_ADDR_INVALID (~(dma_addr_t)0) + +static const char smc_lo_dev_name[] = "loopback-ism"; +static struct smc_lo_dev *lo_dev; + +static void smc_lo_generate_ids(struct smc_lo_dev *ldev) +{ + struct smcd_gid *lgid = &ldev->local_gid; + uuid_t uuid; + + uuid_gen(&uuid); + memcpy(&lgid->gid, &uuid, sizeof(lgid->gid)); + memcpy(&lgid->gid_ext, (u8 *)&uuid + sizeof(lgid->gid), + sizeof(lgid->gid_ext)); + + ldev->chid = SMC_LO_RESERVED_CHID; +} + +static int smc_lo_query_rgid(struct smcd_dev *smcd, struct smcd_gid *rgid, + u32 vid_valid, u32 vid) +{ + struct smc_lo_dev *ldev = smcd->priv; + + /* rgid should be the same as lgid */ + if (!ldev || rgid->gid != ldev->local_gid.gid || + rgid->gid_ext != ldev->local_gid.gid_ext) + return -ENETUNREACH; + return 0; +} + +static int smc_lo_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb, + void *client_priv) +{ + struct smc_lo_dmb_node *dmb_node, *tmp_node; + struct smc_lo_dev *ldev = smcd->priv; + int sba_idx, rc; + + /* check space for new dmb */ + for_each_clear_bit(sba_idx, ldev->sba_idx_mask, SMC_LO_MAX_DMBS) { + if (!test_and_set_bit(sba_idx, ldev->sba_idx_mask)) + break; + } + if (sba_idx == SMC_LO_MAX_DMBS) + return -ENOSPC; + + dmb_node = kzalloc(sizeof(*dmb_node), GFP_KERNEL); + if (!dmb_node) { + rc = -ENOMEM; + goto err_bit; + } + + dmb_node->sba_idx = sba_idx; + dmb_node->len = dmb->dmb_len; + dmb_node->cpu_addr = kzalloc(dmb_node->len, GFP_KERNEL | + __GFP_NOWARN | __GFP_NORETRY | + __GFP_NOMEMALLOC); + if (!dmb_node->cpu_addr) { + rc = -ENOMEM; + goto err_node; + } + dmb_node->dma_addr = SMC_DMA_ADDR_INVALID; + refcount_set(&dmb_node->refcnt, 1); + +again: + /* add new dmb into hash table */ + get_random_bytes(&dmb_node->token, sizeof(dmb_node->token)); + write_lock_bh(&ldev->dmb_ht_lock); + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb_node->token) { + if (tmp_node->token == dmb_node->token) { + write_unlock_bh(&ldev->dmb_ht_lock); + goto again; + } + } + hash_add(ldev->dmb_ht, &dmb_node->list, dmb_node->token); + write_unlock_bh(&ldev->dmb_ht_lock); + atomic_inc(&ldev->dmb_cnt); + + dmb->sba_idx = dmb_node->sba_idx; + dmb->dmb_tok = dmb_node->token; + dmb->cpu_addr = dmb_node->cpu_addr; + dmb->dma_addr = dmb_node->dma_addr; + dmb->dmb_len = dmb_node->len; + + return 0; + +err_node: + kfree(dmb_node); +err_bit: + clear_bit(sba_idx, ldev->sba_idx_mask); + return rc; +} + +static void __smc_lo_unregister_dmb(struct smc_lo_dev *ldev, + struct smc_lo_dmb_node *dmb_node) +{ + /* remove dmb from hash table */ + write_lock_bh(&ldev->dmb_ht_lock); + hash_del(&dmb_node->list); + write_unlock_bh(&ldev->dmb_ht_lock); + + clear_bit(dmb_node->sba_idx, ldev->sba_idx_mask); + kvfree(dmb_node->cpu_addr); + kfree(dmb_node); + + if (atomic_dec_and_test(&ldev->dmb_cnt)) + wake_up(&ldev->ldev_release); +} + +static int smc_lo_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) +{ + struct smc_lo_dmb_node *dmb_node = NULL, *tmp_node; + struct smc_lo_dev *ldev = smcd->priv; + + /* find dmb from hash table */ + read_lock_bh(&ldev->dmb_ht_lock); + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb->dmb_tok) { + if (tmp_node->token == dmb->dmb_tok) { + dmb_node = tmp_node; + break; + } + } + if (!dmb_node) { + read_unlock_bh(&ldev->dmb_ht_lock); + return -EINVAL; + } + read_unlock_bh(&ldev->dmb_ht_lock); + + if (refcount_dec_and_test(&dmb_node->refcnt)) + __smc_lo_unregister_dmb(ldev, dmb_node); + return 0; +} + +static int smc_lo_support_dmb_nocopy(struct smcd_dev *smcd) +{ + return SMC_LO_SUPPORT_NOCOPY; +} + +static int smc_lo_attach_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb) +{ + struct smc_lo_dmb_node *dmb_node = NULL, *tmp_node; + struct smc_lo_dev *ldev = smcd->priv; + + /* find dmb_node according to dmb->dmb_tok */ + read_lock_bh(&ldev->dmb_ht_lock); + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb->dmb_tok) { + if (tmp_node->token == dmb->dmb_tok) { + dmb_node = tmp_node; + break; + } + } + if (!dmb_node) { + read_unlock_bh(&ldev->dmb_ht_lock); + return -EINVAL; + } + read_unlock_bh(&ldev->dmb_ht_lock); + + if (!refcount_inc_not_zero(&dmb_node->refcnt)) + /* the dmb is being unregistered, but has + * not been removed from the hash table. + */ + return -EINVAL; + + /* provide dmb information */ + dmb->sba_idx = dmb_node->sba_idx; + dmb->dmb_tok = dmb_node->token; + dmb->cpu_addr = dmb_node->cpu_addr; + dmb->dma_addr = dmb_node->dma_addr; + dmb->dmb_len = dmb_node->len; + return 0; +} + +static int smc_lo_detach_dmb(struct smcd_dev *smcd, u64 token) +{ + struct smc_lo_dmb_node *dmb_node = NULL, *tmp_node; + struct smc_lo_dev *ldev = smcd->priv; + + /* find dmb_node according to dmb->dmb_tok */ + read_lock_bh(&ldev->dmb_ht_lock); + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, token) { + if (tmp_node->token == token) { + dmb_node = tmp_node; + break; + } + } + if (!dmb_node) { + read_unlock_bh(&ldev->dmb_ht_lock); + return -EINVAL; + } + read_unlock_bh(&ldev->dmb_ht_lock); + + if (refcount_dec_and_test(&dmb_node->refcnt)) + __smc_lo_unregister_dmb(ldev, dmb_node); + return 0; +} + +static int smc_lo_move_data(struct smcd_dev *smcd, u64 dmb_tok, + unsigned int idx, bool sf, unsigned int offset, + void *data, unsigned int size) +{ + struct smc_lo_dmb_node *rmb_node = NULL, *tmp_node; + struct smc_lo_dev *ldev = smcd->priv; + struct smc_connection *conn; + + if (!sf) + /* since sndbuf is merged with peer DMB, there is + * no need to copy data from sndbuf to peer DMB. + */ + return 0; + + read_lock_bh(&ldev->dmb_ht_lock); + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb_tok) { + if (tmp_node->token == dmb_tok) { + rmb_node = tmp_node; + break; + } + } + if (!rmb_node) { + read_unlock_bh(&ldev->dmb_ht_lock); + return -EINVAL; + } + memcpy((char *)rmb_node->cpu_addr + offset, data, size); + read_unlock_bh(&ldev->dmb_ht_lock); + + conn = smcd->conn[rmb_node->sba_idx]; + if (!conn || conn->killed) + return -EPIPE; + tasklet_schedule(&conn->rx_tsklet); + return 0; +} + +static int smc_lo_supports_v2(void) +{ + return SMC_LO_V2_CAPABLE; +} + +static void smc_lo_get_local_gid(struct smcd_dev *smcd, + struct smcd_gid *smcd_gid) +{ + struct smc_lo_dev *ldev = smcd->priv; + + smcd_gid->gid = ldev->local_gid.gid; + smcd_gid->gid_ext = ldev->local_gid.gid_ext; +} + +static u16 smc_lo_get_chid(struct smcd_dev *smcd) +{ + return ((struct smc_lo_dev *)smcd->priv)->chid; +} + +static struct device *smc_lo_get_dev(struct smcd_dev *smcd) +{ + return &((struct smc_lo_dev *)smcd->priv)->dev; +} + +static const struct smcd_ops lo_ops = { + .query_remote_gid = smc_lo_query_rgid, + .register_dmb = smc_lo_register_dmb, + .unregister_dmb = smc_lo_unregister_dmb, + .support_dmb_nocopy = smc_lo_support_dmb_nocopy, + .attach_dmb = smc_lo_attach_dmb, + .detach_dmb = smc_lo_detach_dmb, + .add_vlan_id = NULL, + .del_vlan_id = NULL, + .set_vlan_required = NULL, + .reset_vlan_required = NULL, + .signal_event = NULL, + .move_data = smc_lo_move_data, + .supports_v2 = smc_lo_supports_v2, + .get_local_gid = smc_lo_get_local_gid, + .get_chid = smc_lo_get_chid, + .get_dev = smc_lo_get_dev, +}; + +static struct smcd_dev *smcd_lo_alloc_dev(const struct smcd_ops *ops, + int max_dmbs) +{ + struct smcd_dev *smcd; + + smcd = kzalloc(sizeof(*smcd), GFP_KERNEL); + if (!smcd) + return NULL; + + smcd->conn = kcalloc(max_dmbs, sizeof(struct smc_connection *), + GFP_KERNEL); + if (!smcd->conn) + goto out_smcd; + + smcd->ops = ops; + + spin_lock_init(&smcd->lock); + spin_lock_init(&smcd->lgr_lock); + INIT_LIST_HEAD(&smcd->vlan); + INIT_LIST_HEAD(&smcd->lgr_list); + init_waitqueue_head(&smcd->lgrs_deleted); + return smcd; + +out_smcd: + kfree(smcd); + return NULL; +} + +static int smcd_lo_register_dev(struct smc_lo_dev *ldev) +{ + struct smcd_dev *smcd; + + smcd = smcd_lo_alloc_dev(&lo_ops, SMC_LO_MAX_DMBS); + if (!smcd) + return -ENOMEM; + ldev->smcd = smcd; + smcd->priv = ldev; + smc_ism_set_v2_capable(); + mutex_lock(&smcd_dev_list.mutex); + list_add(&smcd->list, &smcd_dev_list.list); + mutex_unlock(&smcd_dev_list.mutex); + pr_warn_ratelimited("smc: adding smcd device %s\n", + dev_name(&ldev->dev)); + return 0; +} + +static void smcd_lo_unregister_dev(struct smc_lo_dev *ldev) +{ + struct smcd_dev *smcd = ldev->smcd; + + pr_warn_ratelimited("smc: removing smcd device %s\n", + dev_name(&ldev->dev)); + smcd->going_away = 1; + smc_smcd_terminate_all(smcd); + mutex_lock(&smcd_dev_list.mutex); + list_del_init(&smcd->list); + mutex_unlock(&smcd_dev_list.mutex); + kfree(smcd->conn); + kfree(smcd); +} + +static int smc_lo_dev_init(struct smc_lo_dev *ldev) +{ + smc_lo_generate_ids(ldev); + rwlock_init(&ldev->dmb_ht_lock); + hash_init(ldev->dmb_ht); + atomic_set(&ldev->dmb_cnt, 0); + init_waitqueue_head(&ldev->ldev_release); + + return smcd_lo_register_dev(ldev); +} + +static void smc_lo_dev_exit(struct smc_lo_dev *ldev) +{ + smcd_lo_unregister_dev(ldev); + if (atomic_read(&ldev->dmb_cnt)) + wait_event(ldev->ldev_release, !atomic_read(&ldev->dmb_cnt)); +} + +static void smc_lo_dev_release(struct device *dev) +{ + struct smc_lo_dev *ldev = + container_of(dev, struct smc_lo_dev, dev); + + kfree(ldev); +} + +static int smc_lo_dev_probe(void) +{ + struct smc_lo_dev *ldev; + int ret; + + ldev = kzalloc(sizeof(*ldev), GFP_KERNEL); + if (!ldev) + return -ENOMEM; + + ldev->dev.parent = NULL; + ldev->dev.release = smc_lo_dev_release; + device_initialize(&ldev->dev); + dev_set_name(&ldev->dev, smc_lo_dev_name); + + ret = smc_lo_dev_init(ldev); + if (ret) + goto free_dev; + + lo_dev = ldev; /* global loopback device */ + return 0; + +free_dev: + put_device(&ldev->dev); + return ret; +} + +static void smc_lo_dev_remove(void) +{ + if (!lo_dev) + return; + + smc_lo_dev_exit(lo_dev); + put_device(&lo_dev->dev); /* device_initialize in smc_lo_dev_probe */ +} + +int smc_loopback_init(void) +{ + return smc_lo_dev_probe(); +} + +void smc_loopback_exit(void) +{ + smc_lo_dev_remove(); +} diff --git a/net/smc/smc_loopback.h b/net/smc/smc_loopback.h new file mode 100644 index 0000000000000000000000000000000000000000..04dc6808d2e1bae009126cd0f66a180ceb2e0ba6 --- /dev/null +++ b/net/smc/smc_loopback.h @@ -0,0 +1,60 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Shared Memory Communications Direct over loopback-ism device. + * + * SMC-D loopback-ism device structure definitions. + * + * Copyright (c) 2024, Alibaba Inc. + * + * Author: Wen Gu + * Tony Lu + * + */ + +#ifndef _SMC_LOOPBACK_H +#define _SMC_LOOPBACK_H + +#include +#include + +#if IS_ENABLED(CONFIG_SMC_LO) +#define SMC_LO_MAX_DMBS 5000 +#define SMC_LO_DMBS_HASH_BITS 12 +#define SMC_LO_RESERVED_CHID 0xFFFF + +struct smc_lo_dmb_node { + struct hlist_node list; + u64 token; + u32 len; + u32 sba_idx; + void *cpu_addr; + dma_addr_t dma_addr; + refcount_t refcnt; +}; + +struct smc_lo_dev { + struct smcd_dev *smcd; + struct device dev; + u16 chid; + struct smcd_gid local_gid; + atomic_t dmb_cnt; + rwlock_t dmb_ht_lock; + DECLARE_BITMAP(sba_idx_mask, SMC_LO_MAX_DMBS); + DECLARE_HASHTABLE(dmb_ht, SMC_LO_DMBS_HASH_BITS); + wait_queue_head_t ldev_release; +}; + +int smc_loopback_init(void); +void smc_loopback_exit(void); +#else +static inline int smc_loopback_init(void) +{ + return 0; +} + +static inline void smc_loopback_exit(void) +{ +} +#endif + +#endif /* _SMC_LOOPBACK_H */ diff --git a/net/smc/smc_netlink.c b/net/smc/smc_netlink.c index 6fb6f96c1d1775a5dbd30aaff1afe64ee40df348..f13ab0661ed5ee615e5f3dca2b87e7a5edcd11f4 100644 --- a/net/smc/smc_netlink.c +++ b/net/smc/smc_netlink.c @@ -19,11 +19,19 @@ #include "smc_core.h" #include "smc_ism.h" #include "smc_ib.h" +#include "smc_clc.h" #include "smc_stats.h" #include "smc_netlink.h" -#define SMC_CMD_MAX_ATTR 1 +const struct nla_policy +smc_gen_ueid_policy[SMC_NLA_EID_TABLE_MAX + 1] = { + [SMC_NLA_EID_TABLE_UNSPEC] = { .type = NLA_UNSPEC }, + [SMC_NLA_EID_TABLE_ENTRY] = { .type = NLA_STRING, + .len = SMC_MAX_EID_LEN, + }, +}; +#define SMC_CMD_MAX_ATTR 1 /* SMC_GENL generic netlink operation definition */ static const struct genl_ops smc_gen_nl_ops[] = { { @@ -66,6 +74,43 @@ static const struct genl_ops smc_gen_nl_ops[] = { /* can be retrieved by unprivileged users */ .dumpit = smc_nl_get_fback_stats, }, + { + .cmd = SMC_NETLINK_DUMP_UEID, + /* can be retrieved by unprivileged users */ + .dumpit = smc_nl_dump_ueid, + }, + { + .cmd = SMC_NETLINK_ADD_UEID, + .flags = GENL_ADMIN_PERM, + .doit = smc_nl_add_ueid, + .policy = smc_gen_ueid_policy, + }, + { + .cmd = SMC_NETLINK_REMOVE_UEID, + .flags = GENL_ADMIN_PERM, + .doit = smc_nl_remove_ueid, + .policy = smc_gen_ueid_policy, + }, + { + .cmd = SMC_NETLINK_FLUSH_UEID, + .flags = GENL_ADMIN_PERM, + .doit = smc_nl_flush_ueid, + }, + { + .cmd = SMC_NETLINK_DUMP_SEID, + /* can be retrieved by unprivileged users */ + .dumpit = smc_nl_dump_seid, + }, + { + .cmd = SMC_NETLINK_ENABLE_SEID, + .flags = GENL_ADMIN_PERM, + .doit = smc_nl_enable_seid, + }, + { + .cmd = SMC_NETLINK_DISABLE_SEID, + .flags = GENL_ADMIN_PERM, + .doit = smc_nl_disable_seid, + }, }; static const struct nla_policy smc_gen_nl_policy[2] = { diff --git a/net/smc/smc_netlink.h b/net/smc/smc_netlink.h index 5ce2c0a89ccda1430d2afcce8184b38911d08f31..e8c6c3f0e98ca1b48fffa4494db92de324a3ace8 100644 --- a/net/smc/smc_netlink.h +++ b/net/smc/smc_netlink.h @@ -17,6 +17,8 @@ extern struct genl_family smc_gen_nl_family; +extern const struct nla_policy smc_gen_ueid_policy[]; + struct smc_nl_dmp_ctx { int pos[3]; }; diff --git a/net/smc/smc_pnet.c b/net/smc/smc_pnet.c index bf8c662a6116a43d92492e0fd8911beb21fc04db..38cbf7b0a86ac76ac23236f31414784024e2c492 100644 --- a/net/smc/smc_pnet.c +++ b/net/smc/smc_pnet.c @@ -102,7 +102,7 @@ static int smc_pnet_remove_by_pnetid(struct net *net, char *pnet_name) struct smc_pnetentry *pnetelem, *tmp_pe; struct smc_pnettable *pnettable; struct smc_ib_device *ibdev; - struct smcd_dev *smcd_dev; + struct smcd_dev *smcd; struct smc_net *sn; int rc = -ENOENT; int ibport; @@ -160,16 +160,17 @@ static int smc_pnet_remove_by_pnetid(struct net *net, char *pnet_name) mutex_unlock(&smc_ib_devices.mutex); /* remove smcd devices */ mutex_lock(&smcd_dev_list.mutex); - list_for_each_entry(smcd_dev, &smcd_dev_list.list, list) { - if (smcd_dev->pnetid_by_user && + list_for_each_entry(smcd, &smcd_dev_list.list, list) { + if (smcd->pnetid_by_user && (!pnet_name || - smc_pnet_match(pnet_name, smcd_dev->pnetid))) { + smc_pnet_match(pnet_name, smcd->pnetid))) { pr_warn_ratelimited("smc: smcd device %s " "erased user defined pnetid " - "%.16s\n", dev_name(&smcd_dev->dev), - smcd_dev->pnetid); - memset(smcd_dev->pnetid, 0, SMC_MAX_PNETID_LEN); - smcd_dev->pnetid_by_user = false; + "%.16s\n", + dev_name(smcd->ops->get_dev(smcd)), + smcd->pnetid); + memset(smcd->pnetid, 0, SMC_MAX_PNETID_LEN); + smcd->pnetid_by_user = false; rc = 0; } } @@ -329,8 +330,8 @@ static struct smcd_dev *smc_pnet_find_smcd(char *smcd_name) mutex_lock(&smcd_dev_list.mutex); list_for_each_entry(smcd_dev, &smcd_dev_list.list, list) { - if (!strncmp(dev_name(&smcd_dev->dev), smcd_name, - IB_DEVICE_NAME_MAX - 1)) + if (!strncmp(dev_name(smcd_dev->ops->get_dev(smcd_dev)), + smcd_name, IB_DEVICE_NAME_MAX - 1)) goto out; } smcd_dev = NULL; @@ -407,7 +408,8 @@ static int smc_pnet_add_ib(struct smc_pnettable *pnettable, char *ib_name, struct smc_ib_device *ib_dev; bool smcddev_applied = true; bool ibdev_applied = true; - struct smcd_dev *smcd_dev; + struct smcd_dev *smcd; + struct device *dev; bool new_ibdev; /* try to apply the pnetid to active devices */ @@ -421,14 +423,16 @@ static int smc_pnet_add_ib(struct smc_pnettable *pnettable, char *ib_name, ib_port, ib_dev->pnetid[ib_port - 1]); } - smcd_dev = smc_pnet_find_smcd(ib_name); - if (smcd_dev) { - smcddev_applied = smc_pnet_apply_smcd(smcd_dev, pnet_name); - if (smcddev_applied) + smcd = smc_pnet_find_smcd(ib_name); + if (smcd) { + smcddev_applied = smc_pnet_apply_smcd(smcd, pnet_name); + if (smcddev_applied) { + dev = smcd->ops->get_dev(smcd); pr_warn_ratelimited("smc: smcd device %s " "applied user defined pnetid " - "%.16s\n", dev_name(&smcd_dev->dev), - smcd_dev->pnetid); + "%.16s\n", dev_name(dev), + smcd->pnetid); + } } /* Apply fails when a device has a hardware-defined pnetid set, do not * add a pnet table entry in that case. @@ -744,7 +748,7 @@ static int smc_pnet_add_pnetid(struct net *net, u8 *pnetid) write_lock(&sn->pnetids_ndev.lock); list_for_each_entry(pi, &sn->pnetids_ndev.list, list) { - if (smc_pnet_match(pnetid, pe->pnetid)) { + if (smc_pnet_match(pnetid, pi->pnetid)) { refcount_inc(&pi->refcnt); kfree(pe); goto unlock; @@ -965,30 +969,48 @@ static int smc_pnet_find_ndev_pnetid_by_table(struct net_device *ndev, return rc; } +static int smc_pnet_determine_gid(struct smc_ib_device *ibdev, int i, + struct smc_init_info *ini) +{ + if (!ini->check_smcrv2 && + !smc_ib_determine_gid(ibdev, i, ini->vlan_id, ini->ib_gid, NULL, + NULL)) { + ini->ib_dev = ibdev; + ini->ib_port = i; + return 0; + } + if (ini->check_smcrv2 && + !smc_ib_determine_gid(ibdev, i, ini->vlan_id, ini->smcrv2.ib_gid_v2, + NULL, &ini->smcrv2)) { + ini->smcrv2.ib_dev_v2 = ibdev; + ini->smcrv2.ib_port_v2 = i; + return 0; + } + return -ENODEV; +} + /* find a roce device for the given pnetid */ static void _smc_pnet_find_roce_by_pnetid(u8 *pnet_id, struct smc_init_info *ini, - struct smc_ib_device *known_dev) + struct smc_ib_device *known_dev, + struct net *net) { struct smc_ib_device *ibdev; int i; - ini->ib_dev = NULL; mutex_lock(&smc_ib_devices.mutex); list_for_each_entry(ibdev, &smc_ib_devices.list, list) { - if (ibdev == known_dev) + if (ibdev == known_dev || + !rdma_dev_access_netns(ibdev->ibdev, net)) continue; for (i = 1; i <= SMC_MAX_PORTS; i++) { if (!rdma_is_port_valid(ibdev->ibdev, i)) continue; if (smc_pnet_match(ibdev->pnetid[i - 1], pnet_id) && smc_ib_port_active(ibdev, i) && - !test_bit(i - 1, ibdev->ports_going_away) && - !smc_ib_determine_gid(ibdev, i, ini->vlan_id, - ini->ib_gid, NULL)) { - ini->ib_dev = ibdev; - ini->ib_port = i; - goto out; + !test_bit(i - 1, ibdev->ports_going_away)) { + if (!smc_pnet_determine_gid(ibdev, i, ini)) + goto out; } } } @@ -996,12 +1018,14 @@ static void _smc_pnet_find_roce_by_pnetid(u8 *pnet_id, mutex_unlock(&smc_ib_devices.mutex); } -/* find alternate roce device with same pnet_id and vlan_id */ +/* find alternate roce device with same pnet_id, vlan_id and net namespace */ void smc_pnet_find_alt_roce(struct smc_link_group *lgr, struct smc_init_info *ini, struct smc_ib_device *known_dev) { - _smc_pnet_find_roce_by_pnetid(lgr->pnet_id, ini, known_dev); + struct net *net = lgr->net; + + _smc_pnet_find_roce_by_pnetid(lgr->pnet_id, ini, known_dev, net); } /* if handshake network device belongs to a roce device, return its @@ -1010,6 +1034,7 @@ void smc_pnet_find_alt_roce(struct smc_link_group *lgr, static void smc_pnet_find_rdma_dev(struct net_device *netdev, struct smc_init_info *ini) { + struct net *net = dev_net(netdev); struct smc_ib_device *ibdev; mutex_lock(&smc_ib_devices.mutex); @@ -1017,23 +1042,22 @@ static void smc_pnet_find_rdma_dev(struct net_device *netdev, struct net_device *ndev; int i; + /* check rdma net namespace */ + if (!rdma_dev_access_netns(ibdev->ibdev, net)) + continue; + for (i = 1; i <= SMC_MAX_PORTS; i++) { if (!rdma_is_port_valid(ibdev->ibdev, i)) continue; - if (!ibdev->ibdev->ops.get_netdev) - continue; - ndev = ibdev->ibdev->ops.get_netdev(ibdev->ibdev, i); + ndev = ib_device_get_netdev(ibdev->ibdev, i); if (!ndev) continue; dev_put(ndev); if (netdev == ndev && smc_ib_port_active(ibdev, i) && - !test_bit(i - 1, ibdev->ports_going_away) && - !smc_ib_determine_gid(ibdev, i, ini->vlan_id, - ini->ib_gid, NULL)) { - ini->ib_dev = ibdev; - ini->ib_port = i; - break; + !test_bit(i - 1, ibdev->ports_going_away)) { + if (!smc_pnet_determine_gid(ibdev, i, ini)) + break; } } } @@ -1050,15 +1074,19 @@ static void smc_pnet_find_roce_by_pnetid(struct net_device *ndev, struct smc_init_info *ini) { u8 ndev_pnetid[SMC_MAX_PNETID_LEN]; + struct net_device *base_ndev; + struct net *net; - ndev = pnet_find_base_ndev(ndev); - if (smc_pnetid_by_dev_port(ndev->dev.parent, ndev->dev_port, + base_ndev = pnet_find_base_ndev(ndev); + net = dev_net(ndev); + if (smc_pnetid_by_dev_port(base_ndev->dev.parent, base_ndev->dev_port, ndev_pnetid) && + smc_pnet_find_ndev_pnetid_by_table(base_ndev, ndev_pnetid) && smc_pnet_find_ndev_pnetid_by_table(ndev, ndev_pnetid)) { - smc_pnet_find_rdma_dev(ndev, ini); + smc_pnet_find_rdma_dev(base_ndev, ini); return; /* pnetid could not be determined */ } - _smc_pnet_find_roce_by_pnetid(ndev_pnetid, ini, NULL); + _smc_pnet_find_roce_by_pnetid(ndev_pnetid, ini, NULL, net); } static void smc_pnet_find_ism_by_pnetid(struct net_device *ndev, @@ -1077,8 +1105,8 @@ static void smc_pnet_find_ism_by_pnetid(struct net_device *ndev, list_for_each_entry(ismdev, &smcd_dev_list.list, list) { if (smc_pnet_match(ismdev->pnetid, ndev_pnetid) && !ismdev->going_away && - (!ini->ism_peer_gid[0] || - !smc_ism_cantalk(ini->ism_peer_gid[0], ini->vlan_id, + (!ini->ism_peer_gid[0].gid || + !smc_ism_cantalk(&ini->ism_peer_gid[0], ini->vlan_id, ismdev))) { ini->ism_dev[0] = ismdev; break; @@ -1095,8 +1123,6 @@ void smc_pnet_find_roce_resource(struct sock *sk, struct smc_init_info *ini) { struct dst_entry *dst = sk_dst_get(sk); - ini->ib_dev = NULL; - ini->ib_port = 0; if (!dst) goto out; if (!dst->dev) @@ -1161,7 +1187,7 @@ int smc_pnetid_by_table_ib(struct smc_ib_device *smcibdev, u8 ib_port) */ int smc_pnetid_by_table_smcd(struct smcd_dev *smcddev) { - const char *ib_name = dev_name(&smcddev->dev); + const char *ib_name = dev_name(smcddev->ops->get_dev(smcddev)); struct smc_pnettable *pnettable; struct smc_pnetentry *tmp_pe; struct smc_net *sn; diff --git a/net/smc/smc_rx.c b/net/smc/smc_rx.c index de6e3ac5244ea98095cfb537c1ddc58d882a7f04..62275335000c331fa8b7b537a245b7570424cb59 100644 --- a/net/smc/smc_rx.c +++ b/net/smc/smc_rx.c @@ -192,7 +192,7 @@ static int smc_rx_splice(struct pipe_inode_info *pipe, char *src, size_t len, partial[i].offset = offset; partial[i].len = size; partial[i].private = (unsigned long)priv[i]; - buf += size / sizeof(*buf); + buf += size; left -= size; offset = 0; } @@ -233,22 +233,23 @@ static int smc_rx_splice(struct pipe_inode_info *pipe, char *src, size_t len, return -ENOMEM; } -static int smc_rx_data_available_and_no_splice_pend(struct smc_connection *conn) +static int smc_rx_data_available_and_no_splice_pend(struct smc_connection *conn, size_t peeked) { - return atomic_read(&conn->bytes_to_rcv) && + return smc_rx_data_available(conn, peeked) && !atomic_read(&conn->splice_pending); } /* blocks rcvbuf consumer until >=len bytes available or timeout or interrupted * @smc smc socket * @timeo pointer to max seconds to wait, pointer to value 0 for no timeout + * @peeked number of bytes already peeked * @fcrit add'l criterion to evaluate as function pointer * Returns: * 1 if at least 1 byte available in rcvbuf or if socket error/shutdown. * 0 otherwise (nothing in rcvbuf nor timeout, e.g. interrupted). */ -int smc_rx_wait(struct smc_sock *smc, long *timeo, - int (*fcrit)(struct smc_connection *conn)) +int smc_rx_wait(struct smc_sock *smc, long *timeo, size_t peeked, + int (*fcrit)(struct smc_connection *conn, size_t baseline)) { DEFINE_WAIT_FUNC(wait, woken_wake_function); struct smc_connection *conn = &smc->conn; @@ -257,7 +258,7 @@ int smc_rx_wait(struct smc_sock *smc, long *timeo, struct sock *sk = &smc->sk; int rc; - if (fcrit(conn)) + if (fcrit(conn, peeked)) return 1; sk_set_bit(SOCKWQ_ASYNC_WAITDATA, sk); add_wait_queue(sk_sleep(sk), &wait); @@ -266,7 +267,7 @@ int smc_rx_wait(struct smc_sock *smc, long *timeo, cflags->peer_conn_abort || READ_ONCE(sk->sk_shutdown) & RCV_SHUTDOWN || conn->killed || - fcrit(conn), + fcrit(conn, peeked), &wait); remove_wait_queue(sk_sleep(sk), &wait); sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk); @@ -317,11 +318,11 @@ static int smc_rx_recv_urg(struct smc_sock *smc, struct msghdr *msg, int len, return -EAGAIN; } -static bool smc_rx_recvmsg_data_available(struct smc_sock *smc) +static bool smc_rx_recvmsg_data_available(struct smc_sock *smc, size_t peeked) { struct smc_connection *conn = &smc->conn; - if (smc_rx_data_available(conn)) + if (smc_rx_data_available(conn, peeked)) return true; else if (conn->urg_state == SMC_URG_VALID) /* we received a single urgent Byte - skip */ @@ -339,10 +340,10 @@ static bool smc_rx_recvmsg_data_available(struct smc_sock *smc) int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, struct pipe_inode_info *pipe, size_t len, int flags) { - size_t copylen, read_done = 0, read_remaining = len; + size_t copylen, read_done = 0, read_remaining = len, peeked_bytes = 0; size_t chunk_len, chunk_off, chunk_len_sum; struct smc_connection *conn = &smc->conn; - int (*func)(struct smc_connection *conn); + int (*func)(struct smc_connection *conn, size_t baseline); union smc_host_cursor cons; int readable, chunk; char *rcvbuf_base; @@ -379,14 +380,14 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, if (conn->killed) break; - if (smc_rx_recvmsg_data_available(smc)) + if (smc_rx_recvmsg_data_available(smc, peeked_bytes)) goto copy; if (sk->sk_shutdown & RCV_SHUTDOWN) { /* smc_cdc_msg_recv_action() could have run after * above smc_rx_recvmsg_data_available() */ - if (smc_rx_recvmsg_data_available(smc)) + if (smc_rx_recvmsg_data_available(smc, peeked_bytes)) goto copy; break; } @@ -420,26 +421,28 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, } } - if (!smc_rx_data_available(conn)) { - smc_rx_wait(smc, &timeo, smc_rx_data_available); + if (!smc_rx_data_available(conn, peeked_bytes)) { + smc_rx_wait(smc, &timeo, peeked_bytes, smc_rx_data_available); continue; } copy: /* initialize variables for 1st iteration of subsequent loop */ /* could be just 1 byte, even after waiting on data above */ - readable = atomic_read(&conn->bytes_to_rcv); + readable = smc_rx_data_available(conn, peeked_bytes); splbytes = atomic_read(&conn->splice_pending); if (!readable || (msg && splbytes)) { if (splbytes) func = smc_rx_data_available_and_no_splice_pend; else func = smc_rx_data_available; - smc_rx_wait(smc, &timeo, func); + smc_rx_wait(smc, &timeo, peeked_bytes, func); continue; } smc_curs_copy(&cons, &conn->local_tx_ctrl.cons, conn); + if ((flags & MSG_PEEK) && peeked_bytes) + smc_curs_add(conn->rmb_desc->len, &cons, peeked_bytes); /* subsequent splice() calls pick up where previous left */ if (splbytes) smc_curs_add(conn->rmb_desc->len, &cons, splbytes); @@ -475,6 +478,8 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, } read_remaining -= chunk_len; read_done += chunk_len; + if (flags & MSG_PEEK) + peeked_bytes += chunk_len; if (chunk_len_sum == copylen) break; /* either on 1st or 2nd iteration */ diff --git a/net/smc/smc_rx.h b/net/smc/smc_rx.h index db823c97d824ea6ebd66ccd2f0e38e82528b8e96..994f5e42d1ba2674c1b3a86e7c7b2af40f41839a 100644 --- a/net/smc/smc_rx.h +++ b/net/smc/smc_rx.h @@ -21,11 +21,11 @@ void smc_rx_init(struct smc_sock *smc); int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg, struct pipe_inode_info *pipe, size_t len, int flags); -int smc_rx_wait(struct smc_sock *smc, long *timeo, - int (*fcrit)(struct smc_connection *conn)); -static inline int smc_rx_data_available(struct smc_connection *conn) +int smc_rx_wait(struct smc_sock *smc, long *timeo, size_t peeked, + int (*fcrit)(struct smc_connection *conn, size_t baseline)); +static inline int smc_rx_data_available(struct smc_connection *conn, size_t peeked) { - return atomic_read(&conn->bytes_to_rcv); + return atomic_read(&conn->bytes_to_rcv) - peeked; } #endif /* SMC_RX_H */ diff --git a/net/smc/smc_stats.c b/net/smc/smc_stats.c index 44491016707d04093fbb0ae156ee3b438d99b4f5..be1c479f280f30aa25fa842d178db3f31a0a0426 100644 --- a/net/smc/smc_stats.c +++ b/net/smc/smc_stats.c @@ -223,7 +223,7 @@ static int smc_nl_fill_stats_tech_data(struct sk_buff *skb, SMC_NLA_STATS_PAD)) goto errattr; if (nla_put_u64_64bit(skb, SMC_NLA_STATS_T_SENDPAGE_CNT, - smc_tech->sendpage_cnt, + 0, SMC_NLA_STATS_PAD)) goto errattr; if (nla_put_u64_64bit(skb, SMC_NLA_STATS_T_CORK_CNT, diff --git a/net/smc/smc_stats.h b/net/smc/smc_stats.h index c159acd932c5226bad601fb8bb2708ab401b1661..607df9d2b8b00068e5ba4939202ac38a3b591079 100644 --- a/net/smc/smc_stats.h +++ b/net/smc/smc_stats.h @@ -19,7 +19,7 @@ #include "smc_clc.h" -#define SMC_MAX_FBACK_RSN_CNT 30 +#define SMC_MAX_FBACK_RSN_CNT 36 extern struct smc_stats __percpu *smc_stats; /* per cpu counters for SMC */ extern struct smc_stats_reason fback_rsn; @@ -75,7 +75,6 @@ struct smc_stats_tech { u64 clnt_v2_succ_cnt; u64 srv_v1_succ_cnt; u64 srv_v2_succ_cnt; - u64 sendpage_cnt; u64 urg_data_cnt; u64 splice_cnt; u64 cork_cnt; diff --git a/net/smc/smc_tx.c b/net/smc/smc_tx.c index d6b9ccb88e65a61a63e129d1f7929a0ecc9a0d83..c9b477f5af3d079da7856b11a79cd658f5cdc0be 100644 --- a/net/smc/smc_tx.c +++ b/net/smc/smc_tx.c @@ -295,37 +295,17 @@ int smc_tx_sendmsg(struct smc_sock *smc, struct msghdr *msg, size_t len) return rc; } -int smc_tx_sendpage(struct smc_sock *smc, struct page *page, int offset, - size_t size, int flags) -{ - struct msghdr msg = {.msg_flags = flags}; - char *kaddr = kmap(page); - struct kvec iov; - int rc; - - iov.iov_base = kaddr + offset; - iov.iov_len = size; - iov_iter_kvec(&msg.msg_iter, WRITE, &iov, 1, size); - rc = smc_tx_sendmsg(smc, &msg, size); - kunmap(page); - return rc; -} - /***************************** sndbuf consumer *******************************/ /* sndbuf consumer: actual data transfer of one target chunk with ISM write */ int smcd_tx_ism_write(struct smc_connection *conn, void *data, size_t len, u32 offset, int signal) { - struct smc_ism_position pos; int rc; - memset(&pos, 0, sizeof(pos)); - pos.token = conn->peer_token; - pos.index = conn->peer_rmbe_idx; - pos.offset = conn->tx_off + offset; - pos.signal = signal; - rc = smc_ism_write(conn->lgr->smcd, &pos, data, len); + rc = smc_ism_write(conn->lgr->smcd, conn->peer_token, + conn->peer_rmbe_idx, signal, conn->tx_off + offset, + data, len); if (rc) conn->local_tx_ctrl.conn_state_flags.peer_conn_abort = 1; return rc; @@ -637,7 +617,7 @@ static int smcd_tx_sndbuf_nonempty(struct smc_connection *conn) return rc; } -static int __smc_tx_sndbuf_nonempty(struct smc_connection *conn) +int smc_tx_sndbuf_nonempty(struct smc_connection *conn) { struct smc_sock *smc = container_of(conn, struct smc_sock, conn); int rc = 0; @@ -669,34 +649,6 @@ static int __smc_tx_sndbuf_nonempty(struct smc_connection *conn) return rc; } -int smc_tx_sndbuf_nonempty(struct smc_connection *conn) -{ - int rc; - - /* This make sure only one can send simultaneously to prevent wasting - * of CPU and CDC slot. - * Record whether someone has tried to push while we are pushing. - */ - if (atomic_inc_return(&conn->tx_pushing) > 1) - return 0; - -again: - atomic_set(&conn->tx_pushing, 1); - smp_wmb(); /* Make sure tx_pushing is 1 before real send */ - rc = __smc_tx_sndbuf_nonempty(conn); - - /* We need to check whether someone else have added some data into - * the send queue and tried to push but failed after the atomic_set() - * when we are pushing. - * If so, we need to push again to prevent those data hang in the send - * queue. - */ - if (unlikely(!atomic_dec_and_test(&conn->tx_pushing))) - goto again; - - return rc; -} - /* Wakeup sndbuf consumers from process context * since there is more data to transmit. The caller * must hold sock lock. diff --git a/net/smc/smc_tx.h b/net/smc/smc_tx.h index 34b578498b1f1cd78a75fc9bed698e9bd1080dae..a59f370b8b43279adc4437c64938ff2fc419623a 100644 --- a/net/smc/smc_tx.h +++ b/net/smc/smc_tx.h @@ -31,8 +31,6 @@ void smc_tx_pending(struct smc_connection *conn); void smc_tx_work(struct work_struct *work); void smc_tx_init(struct smc_sock *smc); int smc_tx_sendmsg(struct smc_sock *smc, struct msghdr *msg, size_t len); -int smc_tx_sendpage(struct smc_sock *smc, struct page *page, int offset, - size_t size, int flags); int smc_tx_sndbuf_nonempty(struct smc_connection *conn); void smc_tx_sndbuf_nonfull(struct smc_sock *smc); void smc_tx_consumer_update(struct smc_connection *conn, bool force); diff --git a/net/smc/smc_wr.c b/net/smc/smc_wr.c index 64c455abf8618ee93a8c5c3fb2892d175f1da501..2d4a4060363d3ed47043209b0280a952f00d6325 100644 --- a/net/smc/smc_wr.c +++ b/net/smc/smc_wr.c @@ -54,11 +54,7 @@ struct smc_wr_tx_pend { /* control data for a pending send request */ /* returns true if at least one tx work request is pending on the given link */ static inline bool smc_wr_is_tx_pend(struct smc_link *link) { - if (find_first_bit(link->wr_tx_mask, link->wr_tx_cnt) != - link->wr_tx_cnt) { - return true; - } - return false; + return !bitmap_empty(link->wr_tx_mask, link->wr_tx_cnt); } /* wait till all pending tx work requests on the given link are completed */ @@ -96,20 +92,39 @@ static inline void smc_wr_tx_process_cqe(struct ib_wc *wc) } pnd_snd_idx = smc_wr_tx_find_pending_index(link, wc->wr_id); - if (pnd_snd_idx == link->wr_tx_cnt) - return; - link->wr_tx_pends[pnd_snd_idx].wc_status = wc->status; - if (link->wr_tx_pends[pnd_snd_idx].compl_requested) - complete(&link->wr_tx_compl[pnd_snd_idx]); - memcpy(&pnd_snd, &link->wr_tx_pends[pnd_snd_idx], sizeof(pnd_snd)); - /* clear the full struct smc_wr_tx_pend including .priv */ - memset(&link->wr_tx_pends[pnd_snd_idx], 0, - sizeof(link->wr_tx_pends[pnd_snd_idx])); - memset(&link->wr_tx_bufs[pnd_snd_idx], 0, - sizeof(link->wr_tx_bufs[pnd_snd_idx])); - if (!test_and_clear_bit(pnd_snd_idx, link->wr_tx_mask)) - return; + if (pnd_snd_idx == link->wr_tx_cnt) { + if (link->lgr->smc_version != SMC_V2 || + link->wr_tx_v2_pend->wr_id != wc->wr_id) + return; + link->wr_tx_v2_pend->wc_status = wc->status; + memcpy(&pnd_snd, link->wr_tx_v2_pend, sizeof(pnd_snd)); + /* clear the full struct smc_wr_tx_pend including .priv */ + memset(link->wr_tx_v2_pend, 0, + sizeof(*link->wr_tx_v2_pend)); + memset(link->lgr->wr_tx_buf_v2, 0, + sizeof(*link->lgr->wr_tx_buf_v2)); + } else { + link->wr_tx_pends[pnd_snd_idx].wc_status = wc->status; + if (link->wr_tx_pends[pnd_snd_idx].compl_requested) + complete(&link->wr_tx_compl[pnd_snd_idx]); + memcpy(&pnd_snd, &link->wr_tx_pends[pnd_snd_idx], + sizeof(pnd_snd)); + /* clear the full struct smc_wr_tx_pend including .priv */ + memset(&link->wr_tx_pends[pnd_snd_idx], 0, + sizeof(link->wr_tx_pends[pnd_snd_idx])); + memset(&link->wr_tx_bufs[pnd_snd_idx], 0, + sizeof(link->wr_tx_bufs[pnd_snd_idx])); + if (!test_and_clear_bit(pnd_snd_idx, link->wr_tx_mask)) + return; + } + if (wc->status) { + if (link->lgr->smc_version == SMC_V2) { + memset(link->wr_tx_v2_pend, 0, + sizeof(*link->wr_tx_v2_pend)); + memset(link->lgr->wr_tx_buf_v2, 0, + sizeof(*link->lgr->wr_tx_buf_v2)); + } /* terminate link */ smcr_link_down_cond_sched(link); } @@ -118,9 +133,9 @@ static inline void smc_wr_tx_process_cqe(struct ib_wc *wc) wake_up(&link->wr_tx_wait); } -static void smc_wr_tx_tasklet_fn(unsigned long data) +static void smc_wr_tx_tasklet_fn(struct tasklet_struct *t) { - struct smc_ib_device *dev = (struct smc_ib_device *)data; + struct smc_ib_device *dev = from_tasklet(dev, t, send_tasklet); struct ib_wc wc[SMC_WR_MAX_POLL_CQE]; int i = 0, rc; int polled = 0; @@ -226,6 +241,33 @@ int smc_wr_tx_get_free_slot(struct smc_link *link, return 0; } +int smc_wr_tx_get_v2_slot(struct smc_link *link, + smc_wr_tx_handler handler, + struct smc_wr_v2_buf **wr_buf, + struct smc_wr_tx_pend_priv **wr_pend_priv) +{ + struct smc_wr_tx_pend *wr_pend; + struct ib_send_wr *wr_ib; + u64 wr_id; + + if (link->wr_tx_v2_pend->idx == link->wr_tx_cnt) + return -EBUSY; + + *wr_buf = NULL; + *wr_pend_priv = NULL; + wr_id = smc_wr_tx_get_next_wr_id(link); + wr_pend = link->wr_tx_v2_pend; + wr_pend->wr_id = wr_id; + wr_pend->handler = handler; + wr_pend->link = link; + wr_pend->idx = link->wr_tx_cnt; + wr_ib = link->wr_tx_v2_ib; + wr_ib->wr_id = wr_id; + *wr_buf = link->lgr->wr_tx_buf_v2; + *wr_pend_priv = &wr_pend->priv; + return 0; +} + int smc_wr_tx_put_slot(struct smc_link *link, struct smc_wr_tx_pend_priv *wr_pend_priv) { @@ -243,6 +285,14 @@ int smc_wr_tx_put_slot(struct smc_link *link, test_and_clear_bit(idx, link->wr_tx_mask); wake_up(&link->wr_tx_wait); return 1; + } else if (link->lgr->smc_version == SMC_V2 && + pend->idx == link->wr_tx_cnt) { + /* Large v2 buffer */ + memset(&link->wr_tx_v2_pend, 0, + sizeof(link->wr_tx_v2_pend)); + memset(&link->lgr->wr_tx_buf_v2, 0, + sizeof(link->lgr->wr_tx_buf_v2)); + return 1; } return 0; @@ -267,6 +317,22 @@ int smc_wr_tx_send(struct smc_link *link, struct smc_wr_tx_pend_priv *priv) return rc; } +int smc_wr_tx_v2_send(struct smc_link *link, struct smc_wr_tx_pend_priv *priv, + int len) +{ + int rc; + + link->wr_tx_v2_ib->sg_list[0].length = len; + ib_req_notify_cq(link->smcibdev->roce_cq_send, + IB_CQ_NEXT_COMP | IB_CQ_REPORT_MISSED_EVENTS); + rc = ib_post_send(link->roce_qp, link->wr_tx_v2_ib, NULL); + if (rc) { + smc_wr_tx_put_slot(link, priv); + smcr_link_down_cond_sched(link); + } + return rc; +} + /* Send prepared WR slot via ib_post_send and wait for send completion * notification. * @priv: pointer to smc_wr_tx_pend_priv identifying prepared message buffer @@ -311,12 +377,11 @@ int smc_wr_reg_send(struct smc_link *link, struct ib_mr *mr) if (rc) return rc; - atomic_inc(&link->wr_reg_refcnt); + percpu_ref_get(&link->wr_reg_refs); rc = wait_event_interruptible_timeout(link->wr_reg_wait, (link->wr_reg_state != POSTED), SMC_WR_REG_MR_WAIT_TIME); - if (atomic_dec_and_test(&link->wr_reg_refcnt)) - wake_up_all(&link->wr_reg_wait); + percpu_ref_put(&link->wr_reg_refs); if (!rc) { /* timeout - terminate link */ smcr_link_down_cond_sched(link); @@ -411,9 +476,9 @@ static inline void smc_wr_rx_process_cqes(struct ib_wc wc[], int num) } } -static void smc_wr_rx_tasklet_fn(unsigned long data) +static void smc_wr_rx_tasklet_fn(struct tasklet_struct *t) { - struct smc_ib_device *dev = (struct smc_ib_device *)data; + struct smc_ib_device *dev = from_tasklet(dev, t, recv_tasklet); struct ib_wc wc[SMC_WR_MAX_POLL_CQE]; int polled = 0; int rc; @@ -490,6 +555,7 @@ void smc_wr_remember_qp_attr(struct smc_link *lnk) static void smc_wr_init_sge(struct smc_link *lnk) { + int sges_per_buf = (lnk->lgr->smc_version == SMC_V2) ? 2 : 1; u32 i; bool send_inline = (lnk->qp_attr.cap.max_inline_data > SMC_WR_TX_SIZE); @@ -521,14 +587,44 @@ static void smc_wr_init_sge(struct smc_link *lnk) lnk->wr_tx_rdmas[i].wr_tx_rdma[1].wr.sg_list = lnk->wr_tx_rdma_sges[i].tx_rdma_sge[1].wr_tx_rdma_sge; } + + if (lnk->lgr->smc_version == SMC_V2) { + lnk->wr_tx_v2_sge->addr = lnk->wr_tx_v2_dma_addr; + lnk->wr_tx_v2_sge->length = SMC_WR_BUF_V2_SIZE; + lnk->wr_tx_v2_sge->lkey = lnk->roce_pd->local_dma_lkey; + + lnk->wr_tx_v2_ib->next = NULL; + lnk->wr_tx_v2_ib->sg_list = lnk->wr_tx_v2_sge; + lnk->wr_tx_v2_ib->num_sge = 1; + lnk->wr_tx_v2_ib->opcode = IB_WR_SEND; + lnk->wr_tx_v2_ib->send_flags = + IB_SEND_SIGNALED | IB_SEND_SOLICITED; + } + + /* With SMC-Rv2 there can be messages larger than SMC_WR_TX_SIZE. + * Each ib_recv_wr gets 2 sges, the second one is a spillover buffer + * and the same buffer for all sges. When a larger message arrived then + * the content of the first small sge is copied to the beginning of + * the larger spillover buffer, allowing easy data mapping. + */ for (i = 0; i < lnk->wr_rx_cnt; i++) { - lnk->wr_rx_sges[i].addr = + int x = i * sges_per_buf; + + lnk->wr_rx_sges[x].addr = lnk->wr_rx_dma_addr + i * SMC_WR_BUF_SIZE; - lnk->wr_rx_sges[i].length = SMC_WR_BUF_SIZE; - lnk->wr_rx_sges[i].lkey = lnk->roce_pd->local_dma_lkey; + lnk->wr_rx_sges[x].length = SMC_WR_TX_SIZE; + lnk->wr_rx_sges[x].lkey = lnk->roce_pd->local_dma_lkey; + if (lnk->lgr->smc_version == SMC_V2) { + lnk->wr_rx_sges[x + 1].addr = + lnk->wr_rx_v2_dma_addr + SMC_WR_TX_SIZE; + lnk->wr_rx_sges[x + 1].length = + SMC_WR_BUF_V2_SIZE - SMC_WR_TX_SIZE; + lnk->wr_rx_sges[x + 1].lkey = + lnk->roce_pd->local_dma_lkey; + } lnk->wr_rx_ibs[i].next = NULL; - lnk->wr_rx_ibs[i].sg_list = &lnk->wr_rx_sges[i]; - lnk->wr_rx_ibs[i].num_sge = 1; + lnk->wr_rx_ibs[i].sg_list = &lnk->wr_rx_sges[x]; + lnk->wr_rx_ibs[i].num_sge = sges_per_buf; } lnk->wr_reg.wr.next = NULL; lnk->wr_reg.wr.num_sge = 0; @@ -550,8 +646,12 @@ void smc_wr_free_link(struct smc_link *lnk) smc_wr_wakeup_tx_wait(lnk); smc_wr_tx_wait_no_pending_sends(lnk); - wait_event(lnk->wr_reg_wait, (!atomic_read(&lnk->wr_reg_refcnt))); - wait_event(lnk->wr_tx_wait, (!atomic_read(&lnk->wr_tx_refcnt))); + percpu_ref_kill(&lnk->wr_reg_refs); + wait_for_completion(&lnk->reg_ref_comp); + percpu_ref_exit(&lnk->wr_reg_refs); + percpu_ref_kill(&lnk->wr_tx_refs); + wait_for_completion(&lnk->tx_ref_comp); + percpu_ref_exit(&lnk->wr_tx_refs); if (lnk->wr_rx_dma_addr) { ib_dma_unmap_single(ibdev, lnk->wr_rx_dma_addr, @@ -559,21 +659,50 @@ void smc_wr_free_link(struct smc_link *lnk) DMA_FROM_DEVICE); lnk->wr_rx_dma_addr = 0; } + if (lnk->wr_rx_v2_dma_addr) { + ib_dma_unmap_single(ibdev, lnk->wr_rx_v2_dma_addr, + SMC_WR_BUF_V2_SIZE, + DMA_FROM_DEVICE); + lnk->wr_rx_v2_dma_addr = 0; + } if (lnk->wr_tx_dma_addr) { ib_dma_unmap_single(ibdev, lnk->wr_tx_dma_addr, SMC_WR_BUF_SIZE * lnk->wr_tx_cnt, DMA_TO_DEVICE); lnk->wr_tx_dma_addr = 0; } + if (lnk->wr_tx_v2_dma_addr) { + ib_dma_unmap_single(ibdev, lnk->wr_tx_v2_dma_addr, + SMC_WR_BUF_V2_SIZE, + DMA_TO_DEVICE); + lnk->wr_tx_v2_dma_addr = 0; + } +} + +void smc_wr_free_lgr_mem(struct smc_link_group *lgr) +{ + if (lgr->smc_version < SMC_V2) + return; + + kfree(lgr->wr_rx_buf_v2); + lgr->wr_rx_buf_v2 = NULL; + kfree(lgr->wr_tx_buf_v2); + lgr->wr_tx_buf_v2 = NULL; } void smc_wr_free_link_mem(struct smc_link *lnk) { + kfree(lnk->wr_tx_v2_ib); + lnk->wr_tx_v2_ib = NULL; + kfree(lnk->wr_tx_v2_sge); + lnk->wr_tx_v2_sge = NULL; + kfree(lnk->wr_tx_v2_pend); + lnk->wr_tx_v2_pend = NULL; kfree(lnk->wr_tx_compl); lnk->wr_tx_compl = NULL; kfree(lnk->wr_tx_pends); lnk->wr_tx_pends = NULL; - kfree(lnk->wr_tx_mask); + bitmap_free(lnk->wr_tx_mask); lnk->wr_tx_mask = NULL; kfree(lnk->wr_tx_sges); lnk->wr_tx_sges = NULL; @@ -593,8 +722,26 @@ void smc_wr_free_link_mem(struct smc_link *lnk) lnk->wr_rx_bufs = NULL; } +int smc_wr_alloc_lgr_mem(struct smc_link_group *lgr) +{ + if (lgr->smc_version < SMC_V2) + return 0; + + lgr->wr_rx_buf_v2 = kzalloc(SMC_WR_BUF_V2_SIZE, GFP_KERNEL); + if (!lgr->wr_rx_buf_v2) + return -ENOMEM; + lgr->wr_tx_buf_v2 = kzalloc(SMC_WR_BUF_V2_SIZE, GFP_KERNEL); + if (!lgr->wr_tx_buf_v2) { + kfree(lgr->wr_rx_buf_v2); + return -ENOMEM; + } + return 0; +} + int smc_wr_alloc_link_mem(struct smc_link *link) { + int sges_per_buf = link->lgr->smc_version == SMC_V2 ? 2 : 1; + /* allocate link related memory */ link->wr_tx_bufs = kcalloc(SMC_WR_BUF_CNT, SMC_WR_BUF_SIZE, GFP_KERNEL); if (!link->wr_tx_bufs) @@ -627,13 +774,11 @@ int smc_wr_alloc_link_mem(struct smc_link *link) if (!link->wr_tx_sges) goto no_mem_wr_tx_rdma_sges; link->wr_rx_sges = kcalloc(SMC_WR_BUF_CNT * 3, - sizeof(link->wr_rx_sges[0]), + sizeof(link->wr_rx_sges[0]) * sges_per_buf, GFP_KERNEL); if (!link->wr_rx_sges) goto no_mem_wr_tx_sges; - link->wr_tx_mask = kcalloc(BITS_TO_LONGS(SMC_WR_BUF_CNT), - sizeof(*link->wr_tx_mask), - GFP_KERNEL); + link->wr_tx_mask = bitmap_zalloc(SMC_WR_BUF_CNT, GFP_KERNEL); if (!link->wr_tx_mask) goto no_mem_wr_rx_sges; link->wr_tx_pends = kcalloc(SMC_WR_BUF_CNT, @@ -646,8 +791,29 @@ int smc_wr_alloc_link_mem(struct smc_link *link) GFP_KERNEL); if (!link->wr_tx_compl) goto no_mem_wr_tx_pends; + + if (link->lgr->smc_version == SMC_V2) { + link->wr_tx_v2_ib = kzalloc(sizeof(*link->wr_tx_v2_ib), + GFP_KERNEL); + if (!link->wr_tx_v2_ib) + goto no_mem_tx_compl; + link->wr_tx_v2_sge = kzalloc(sizeof(*link->wr_tx_v2_sge), + GFP_KERNEL); + if (!link->wr_tx_v2_sge) + goto no_mem_v2_ib; + link->wr_tx_v2_pend = kzalloc(sizeof(*link->wr_tx_v2_pend), + GFP_KERNEL); + if (!link->wr_tx_v2_pend) + goto no_mem_v2_sge; + } return 0; +no_mem_v2_sge: + kfree(link->wr_tx_v2_sge); +no_mem_v2_ib: + kfree(link->wr_tx_v2_ib); +no_mem_tx_compl: + kfree(link->wr_tx_compl); no_mem_wr_tx_pends: kfree(link->wr_tx_pends); no_mem_wr_tx_mask: @@ -680,10 +846,22 @@ void smc_wr_remove_dev(struct smc_ib_device *smcibdev) void smc_wr_add_dev(struct smc_ib_device *smcibdev) { - tasklet_init(&smcibdev->recv_tasklet, smc_wr_rx_tasklet_fn, - (unsigned long)smcibdev); - tasklet_init(&smcibdev->send_tasklet, smc_wr_tx_tasklet_fn, - (unsigned long)smcibdev); + tasklet_setup(&smcibdev->recv_tasklet, smc_wr_rx_tasklet_fn); + tasklet_setup(&smcibdev->send_tasklet, smc_wr_tx_tasklet_fn); +} + +static void smcr_wr_tx_refs_free(struct percpu_ref *ref) +{ + struct smc_link *lnk = container_of(ref, struct smc_link, wr_tx_refs); + + complete(&lnk->tx_ref_comp); +} + +static void smcr_wr_reg_refs_free(struct percpu_ref *ref) +{ + struct smc_link *lnk = container_of(ref, struct smc_link, wr_reg_refs); + + complete(&lnk->reg_ref_comp); } int smc_wr_create_link(struct smc_link *lnk) @@ -701,6 +879,24 @@ int smc_wr_create_link(struct smc_link *lnk) rc = -EIO; goto out; } + if (lnk->lgr->smc_version == SMC_V2) { + lnk->wr_rx_v2_dma_addr = ib_dma_map_single(ibdev, + lnk->lgr->wr_rx_buf_v2, SMC_WR_BUF_V2_SIZE, + DMA_FROM_DEVICE); + if (ib_dma_mapping_error(ibdev, lnk->wr_rx_v2_dma_addr)) { + lnk->wr_rx_v2_dma_addr = 0; + rc = -EIO; + goto dma_unmap; + } + lnk->wr_tx_v2_dma_addr = ib_dma_map_single(ibdev, + lnk->lgr->wr_tx_buf_v2, SMC_WR_BUF_V2_SIZE, + DMA_TO_DEVICE); + if (ib_dma_mapping_error(ibdev, lnk->wr_tx_v2_dma_addr)) { + lnk->wr_tx_v2_dma_addr = 0; + rc = -EIO; + goto dma_unmap; + } + } lnk->wr_tx_dma_addr = ib_dma_map_single( ibdev, lnk->wr_tx_bufs, SMC_WR_BUF_SIZE * lnk->wr_tx_cnt, DMA_TO_DEVICE); @@ -709,16 +905,35 @@ int smc_wr_create_link(struct smc_link *lnk) goto dma_unmap; } smc_wr_init_sge(lnk); - memset(lnk->wr_tx_mask, 0, - BITS_TO_LONGS(SMC_WR_BUF_CNT) * sizeof(*lnk->wr_tx_mask)); + bitmap_zero(lnk->wr_tx_mask, SMC_WR_BUF_CNT); init_waitqueue_head(&lnk->wr_tx_wait); - atomic_set(&lnk->wr_tx_refcnt, 0); + rc = percpu_ref_init(&lnk->wr_tx_refs, smcr_wr_tx_refs_free, 0, GFP_KERNEL); + if (rc) + goto dma_unmap; + init_completion(&lnk->tx_ref_comp); init_waitqueue_head(&lnk->wr_reg_wait); - atomic_set(&lnk->wr_reg_refcnt, 0); + rc = percpu_ref_init(&lnk->wr_reg_refs, smcr_wr_reg_refs_free, 0, GFP_KERNEL); + if (rc) + goto cancel_ref; + init_completion(&lnk->reg_ref_comp); init_waitqueue_head(&lnk->wr_rx_empty_wait); return rc; +cancel_ref: + percpu_ref_exit(&lnk->wr_tx_refs); dma_unmap: + if (lnk->wr_rx_v2_dma_addr) { + ib_dma_unmap_single(ibdev, lnk->wr_rx_v2_dma_addr, + SMC_WR_BUF_V2_SIZE, + DMA_FROM_DEVICE); + lnk->wr_rx_v2_dma_addr = 0; + } + if (lnk->wr_tx_v2_dma_addr) { + ib_dma_unmap_single(ibdev, lnk->wr_tx_v2_dma_addr, + SMC_WR_BUF_V2_SIZE, + DMA_TO_DEVICE); + lnk->wr_tx_v2_dma_addr = 0; + } ib_dma_unmap_single(ibdev, lnk->wr_rx_dma_addr, SMC_WR_BUF_SIZE * lnk->wr_rx_cnt, DMA_FROM_DEVICE); diff --git a/net/smc/smc_wr.h b/net/smc/smc_wr.h index 4fbd0ef26f000c6d9f5fffcc92ef5124aa73c8fb..f3008dda222a6acf60737aa4ab118062b4597171 100644 --- a/net/smc/smc_wr.h +++ b/net/smc/smc_wr.h @@ -63,14 +63,13 @@ static inline bool smc_wr_tx_link_hold(struct smc_link *link) { if (!smc_link_sendable(link)) return false; - atomic_inc(&link->wr_tx_refcnt); + percpu_ref_get(&link->wr_tx_refs); return true; } static inline void smc_wr_tx_link_put(struct smc_link *link) { - if (atomic_dec_and_test(&link->wr_tx_refcnt)) - wake_up_all(&link->wr_tx_wait); + percpu_ref_put(&link->wr_tx_refs); } static inline void smc_wr_drain_cq(struct smc_link *lnk) @@ -105,8 +104,10 @@ static inline int smc_wr_rx_post(struct smc_link *link) int smc_wr_create_link(struct smc_link *lnk); int smc_wr_alloc_link_mem(struct smc_link *lnk); +int smc_wr_alloc_lgr_mem(struct smc_link_group *lgr); void smc_wr_free_link(struct smc_link *lnk); void smc_wr_free_link_mem(struct smc_link *lnk); +void smc_wr_free_lgr_mem(struct smc_link_group *lgr); void smc_wr_remember_qp_attr(struct smc_link *lnk); void smc_wr_remove_dev(struct smc_ib_device *smcibdev); void smc_wr_add_dev(struct smc_ib_device *smcibdev); @@ -115,17 +116,19 @@ int smc_wr_tx_get_free_slot(struct smc_link *link, smc_wr_tx_handler handler, struct smc_wr_buf **wr_buf, struct smc_rdma_wr **wrs, struct smc_wr_tx_pend_priv **wr_pend_priv); +int smc_wr_tx_get_v2_slot(struct smc_link *link, + smc_wr_tx_handler handler, + struct smc_wr_v2_buf **wr_buf, + struct smc_wr_tx_pend_priv **wr_pend_priv); int smc_wr_tx_put_slot(struct smc_link *link, struct smc_wr_tx_pend_priv *wr_pend_priv); int smc_wr_tx_send(struct smc_link *link, struct smc_wr_tx_pend_priv *wr_pend_priv); +int smc_wr_tx_v2_send(struct smc_link *link, + struct smc_wr_tx_pend_priv *priv, int len); int smc_wr_tx_send_wait(struct smc_link *link, struct smc_wr_tx_pend_priv *priv, unsigned long timeout); void smc_wr_tx_cq_handler(struct ib_cq *ib_cq, void *cq_context); -void smc_wr_tx_dismiss_slots(struct smc_link *lnk, u8 wr_rx_hdr_type, - smc_wr_tx_filter filter, - smc_wr_tx_dismisser dismisser, - unsigned long data); void smc_wr_tx_wait_no_pending_sends(struct smc_link *link); int smc_wr_rx_register_handler(struct smc_wr_rx_handler *handler);