/*
* Copyright (c) 2004, 2005 Topspin Communications. All rights reserved.
+ * Copyright (c) 2005 Mellanox Technologies. All rights reserved.
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
enum {
MTHCA_NUM_ASYNC_EQE = 0x80,
MTHCA_NUM_CMD_EQE = 0x80,
+ MTHCA_NUM_SPARE_EQE = 0x80,
MTHCA_EQ_ENTRY_SIZE = 0x20
};
* Must be packed because start is 64 bits but only aligned to 32 bits.
*/
struct mthca_eq_context {
- u32 flags;
- u64 start;
- u32 logsize_usrpage;
- u32 tavor_pd; /* reserved for Arbel */
- u8 reserved1[3];
- u8 intr;
- u32 arbel_pd; /* lost_count for Tavor */
- u32 lkey;
- u32 reserved2[2];
- u32 consumer_index;
- u32 producer_index;
- u32 reserved3[4];
+ __be32 flags;
+ __be64 start;
+ __be32 logsize_usrpage;
+ __be32 tavor_pd; /* reserved for Arbel */
+ u8 reserved1[3];
+ u8 intr;
+ __be32 arbel_pd; /* lost_count for Tavor */
+ __be32 lkey;
+ u32 reserved2[2];
+ __be32 consumer_index;
+ __be32 producer_index;
+ u32 reserved3[4];
} __attribute__((packed));
#define MTHCA_EQ_STATUS_OK ( 0 << 28)
MTHCA_EVENT_TYPE_PATH_MIG = 0x01,
MTHCA_EVENT_TYPE_COMM_EST = 0x02,
MTHCA_EVENT_TYPE_SQ_DRAINED = 0x03,
- MTHCA_EVENT_TYPE_SRQ_LAST_WQE = 0x13,
+ MTHCA_EVENT_TYPE_SRQ_QP_LAST_WQE = 0x13,
+ MTHCA_EVENT_TYPE_SRQ_LIMIT = 0x14,
MTHCA_EVENT_TYPE_CQ_ERROR = 0x04,
MTHCA_EVENT_TYPE_WQ_CATAS_ERROR = 0x05,
MTHCA_EVENT_TYPE_EEC_CATAS_ERROR = 0x06,
(1ULL << MTHCA_EVENT_TYPE_LOCAL_CATAS_ERROR) | \
(1ULL << MTHCA_EVENT_TYPE_PORT_CHANGE) | \
(1ULL << MTHCA_EVENT_TYPE_ECC_DETECT))
-#define MTHCA_SRQ_EVENT_MASK (1ULL << MTHCA_EVENT_TYPE_SRQ_CATAS_ERROR) | \
- (1ULL << MTHCA_EVENT_TYPE_SRQ_LAST_WQE)
+#define MTHCA_SRQ_EVENT_MASK ((1ULL << MTHCA_EVENT_TYPE_SRQ_CATAS_ERROR) | \
+ (1ULL << MTHCA_EVENT_TYPE_SRQ_QP_LAST_WQE) | \
+ (1ULL << MTHCA_EVENT_TYPE_SRQ_LIMIT))
#define MTHCA_CMD_EVENT_MASK (1ULL << MTHCA_EVENT_TYPE_CMD)
#define MTHCA_EQ_DB_INC_CI (1 << 24)
union {
u32 raw[6];
struct {
- u32 cqn;
+ __be32 cqn;
} __attribute__((packed)) comp;
struct {
- u16 reserved1;
- u16 token;
- u32 reserved2;
- u8 reserved3[3];
- u8 status;
- u64 out_param;
+ u16 reserved1;
+ __be16 token;
+ u32 reserved2;
+ u8 reserved3[3];
+ u8 status;
+ __be64 out_param;
} __attribute__((packed)) cmd;
struct {
- u32 qpn;
+ __be32 qpn;
} __attribute__((packed)) qp;
struct {
- u32 cqn;
- u32 reserved1;
- u8 reserved2[3];
- u8 syndrome;
+ __be32 srqn;
+ } __attribute__((packed)) srq;
+ struct {
+ __be32 cqn;
+ u32 reserved1;
+ u8 reserved2[3];
+ u8 syndrome;
} __attribute__((packed)) cq_err;
struct {
- u32 reserved1[2];
- u32 port;
+ u32 reserved1[2];
+ __be32 port;
} __attribute__((packed)) port_change;
} event;
u8 reserved3[3];
static inline void tavor_set_eq_ci(struct mthca_dev *dev, struct mthca_eq *eq, u32 ci)
{
- u32 doorbell[2];
+ __be32 doorbell[2];
doorbell[0] = cpu_to_be32(MTHCA_EQ_DB_SET_CI | eq->eqn);
doorbell[1] = cpu_to_be32(ci & (eq->nent - 1));
{
/* See comment in tavor_set_eq_ci() above. */
wmb();
- __raw_writel(cpu_to_be32(ci), dev->eq_regs.arbel.eq_set_ci_base +
- eq->eqn * 8);
+ __raw_writel((__force u32) cpu_to_be32(ci),
+ dev->eq_regs.arbel.eq_set_ci_base + eq->eqn * 8);
/* We still want ordering, just not swabbing, so add a barrier */
mb();
}
static inline void tavor_eq_req_not(struct mthca_dev *dev, int eqn)
{
- u32 doorbell[2];
+ __be32 doorbell[2];
doorbell[0] = cpu_to_be32(MTHCA_EQ_DB_REQ_NOT | eqn);
doorbell[1] = 0;
static inline void disarm_cq(struct mthca_dev *dev, int eqn, int cqn)
{
if (!mthca_is_memfree(dev)) {
- u32 doorbell[2];
+ __be32 doorbell[2];
doorbell[0] = cpu_to_be32(MTHCA_EQ_DB_DISARM_CQ | eqn);
doorbell[1] = cpu_to_be32(cqn);
{
struct mthca_eqe *eqe;
int disarm_cqn;
- int eqes_found = 0;
+ int eqes_found = 0;
+ int set_ci = 0;
while ((eqe = next_eqe_sw(eq))) {
- int set_ci = 0;
-
/*
* Make sure we read EQ entry contents after we've
* checked the ownership bit.
case MTHCA_EVENT_TYPE_COMP:
disarm_cqn = be32_to_cpu(eqe->event.comp.cqn) & 0xffffff;
disarm_cq(dev, eq->eqn, disarm_cqn);
- mthca_cq_event(dev, disarm_cqn);
+ mthca_cq_completion(dev, disarm_cqn);
break;
case MTHCA_EVENT_TYPE_PATH_MIG:
IB_EVENT_SQ_DRAINED);
break;
+ case MTHCA_EVENT_TYPE_SRQ_QP_LAST_WQE:
+ mthca_qp_event(dev, be32_to_cpu(eqe->event.qp.qpn) & 0xffffff,
+ IB_EVENT_QP_LAST_WQE_REACHED);
+ break;
+
+ case MTHCA_EVENT_TYPE_SRQ_LIMIT:
+ mthca_srq_event(dev, be32_to_cpu(eqe->event.srq.srqn) & 0xffffff,
+ IB_EVENT_SRQ_LIMIT_REACHED);
+ break;
+
case MTHCA_EVENT_TYPE_WQ_CATAS_ERROR:
mthca_qp_event(dev, be32_to_cpu(eqe->event.qp.qpn) & 0xffffff,
IB_EVENT_QP_FATAL);
be16_to_cpu(eqe->event.cmd.token),
eqe->event.cmd.status,
be64_to_cpu(eqe->event.cmd.out_param));
- /*
- * cmd_event() may add more commands.
- * The card will think the queue has overflowed if
- * we don't tell it we've been processing events.
- */
- set_ci = 1;
break;
case MTHCA_EVENT_TYPE_PORT_CHANGE:
eqe->event.cq_err.syndrome == 1 ?
"overrun" : "access violation",
be32_to_cpu(eqe->event.cq_err.cqn) & 0xffffff);
+ mthca_cq_event(dev, be32_to_cpu(eqe->event.cq_err.cqn),
+ IB_EVENT_CQ_ERR);
break;
case MTHCA_EVENT_TYPE_EQ_OVERFLOW:
set_eqe_hw(eqe);
++eq->cons_index;
eqes_found = 1;
+ ++set_ci;
- if (unlikely(set_ci)) {
+ /*
+ * The HCA will think the queue has overflowed if we
+ * don't tell it we've been processing events. We
+ * create our EQs with MTHCA_NUM_SPARE_EQE extra
+ * entries, so we must update our consumer index at
+ * least that often.
+ */
+ if (unlikely(set_ci >= MTHCA_NUM_SPARE_EQE)) {
/*
* Conditional on hca_type is OK here because
* this is a rare case, not the fast path.
writel(dev->eq_table.clr_mask, dev->eq_table.clr_int);
ecr = readl(dev->eq_regs.tavor.ecr_base + 4);
- if (ecr) {
- writel(ecr, dev->eq_regs.tavor.ecr_base +
- MTHCA_ECR_CLR_BASE - MTHCA_ECR_BASE + 4);
+ if (!ecr)
+ return IRQ_NONE;
+
+ writel(ecr, dev->eq_regs.tavor.ecr_base +
+ MTHCA_ECR_CLR_BASE - MTHCA_ECR_BASE + 4);
- for (i = 0; i < MTHCA_NUM_EQ; ++i)
- if (ecr & dev->eq_table.eq[i].eqn_mask &&
- mthca_eq_int(dev, &dev->eq_table.eq[i])) {
+ for (i = 0; i < MTHCA_NUM_EQ; ++i)
+ if (ecr & dev->eq_table.eq[i].eqn_mask) {
+ if (mthca_eq_int(dev, &dev->eq_table.eq[i]))
tavor_set_eq_ci(dev, &dev->eq_table.eq[i],
dev->eq_table.eq[i].cons_index);
- tavor_eq_req_not(dev, dev->eq_table.eq[i].eqn);
- }
- }
+ tavor_eq_req_not(dev, dev->eq_table.eq[i].eqn);
+ }
- return IRQ_RETVAL(ecr);
+ return IRQ_HANDLED;
}
static irqreturn_t mthca_tavor_msi_x_interrupt(int irq, void *eq_ptr,
u8 intr,
struct mthca_eq *eq)
{
- int npages = (nent * MTHCA_EQ_ENTRY_SIZE + PAGE_SIZE - 1) /
- PAGE_SIZE;
+ int npages;
u64 *dma_list = NULL;
dma_addr_t t;
- void *mailbox = NULL;
+ struct mthca_mailbox *mailbox;
struct mthca_eq_context *eq_context;
int err = -ENOMEM;
int i;
u8 status;
- /* Make sure EQ size is aligned to a power of 2 size. */
- for (i = 1; i < nent; i <<= 1)
- ; /* nothing */
- nent = i;
-
- eq->dev = dev;
+ eq->dev = dev;
+ eq->nent = roundup_pow_of_two(max(nent, 2));
+ npages = ALIGN(eq->nent * MTHCA_EQ_ENTRY_SIZE, PAGE_SIZE) / PAGE_SIZE;
eq->page_list = kmalloc(npages * sizeof *eq->page_list,
GFP_KERNEL);
if (!dma_list)
goto err_out_free;
- mailbox = kmalloc(sizeof *eq_context + MTHCA_CMD_MAILBOX_EXTRA,
- GFP_KERNEL);
- if (!mailbox)
+ mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL);
+ if (IS_ERR(mailbox))
goto err_out_free;
- eq_context = MAILBOX_ALIGN(mailbox);
+ eq_context = mailbox->buf;
for (i = 0; i < npages; ++i) {
- eq->page_list[i].buf = pci_alloc_consistent(dev->pdev,
- PAGE_SIZE, &t);
+ eq->page_list[i].buf = dma_alloc_coherent(&dev->pdev->dev,
+ PAGE_SIZE, &t, GFP_KERNEL);
if (!eq->page_list[i].buf)
- goto err_out_free;
+ goto err_out_free_pages;
dma_list[i] = t;
pci_unmap_addr_set(&eq->page_list[i], mapping, t);
memset(eq->page_list[i].buf, 0, PAGE_SIZE);
}
- for (i = 0; i < nent; ++i)
+ for (i = 0; i < eq->nent; ++i)
set_eqe_hw(get_eqe(eq, i));
eq->eqn = mthca_alloc(&dev->eq_table.alloc);
if (eq->eqn == -1)
- goto err_out_free;
+ goto err_out_free_pages;
err = mthca_mr_alloc_phys(dev, dev->driver_pd.pd_num,
dma_list, PAGE_SHIFT, npages,
if (err)
goto err_out_free_eq;
- eq->nent = nent;
-
memset(eq_context, 0, sizeof *eq_context);
eq_context->flags = cpu_to_be32(MTHCA_EQ_STATUS_OK |
MTHCA_EQ_OWNER_HW |
if (mthca_is_memfree(dev))
eq_context->flags |= cpu_to_be32(MTHCA_EQ_STATE_ARBEL);
- eq_context->logsize_usrpage = cpu_to_be32((ffs(nent) - 1) << 24);
+ eq_context->logsize_usrpage = cpu_to_be32((ffs(eq->nent) - 1) << 24);
if (mthca_is_memfree(dev)) {
eq_context->arbel_pd = cpu_to_be32(dev->driver_pd.pd_num);
} else {
eq_context->intr = intr;
eq_context->lkey = cpu_to_be32(eq->mr.ibmr.lkey);
- err = mthca_SW2HW_EQ(dev, eq_context, eq->eqn, &status);
+ err = mthca_SW2HW_EQ(dev, mailbox, eq->eqn, &status);
if (err) {
mthca_warn(dev, "SW2HW_EQ failed (%d)\n", err);
goto err_out_free_mr;
}
kfree(dma_list);
- kfree(mailbox);
+ mthca_free_mailbox(dev, mailbox);
eq->eqn_mask = swab32(1 << eq->eqn);
eq->cons_index = 0;
dev->eq_table.arm_mask |= eq->eqn_mask;
mthca_dbg(dev, "Allocated EQ %d with %d entries\n",
- eq->eqn, nent);
+ eq->eqn, eq->nent);
return err;
err_out_free_eq:
mthca_free(&dev->eq_table.alloc, eq->eqn);
- err_out_free:
+ err_out_free_pages:
for (i = 0; i < npages; ++i)
if (eq->page_list[i].buf)
- pci_free_consistent(dev->pdev, PAGE_SIZE,
- eq->page_list[i].buf,
- pci_unmap_addr(&eq->page_list[i],
- mapping));
+ dma_free_coherent(&dev->pdev->dev, PAGE_SIZE,
+ eq->page_list[i].buf,
+ pci_unmap_addr(&eq->page_list[i],
+ mapping));
+
+ mthca_free_mailbox(dev, mailbox);
+ err_out_free:
kfree(eq->page_list);
kfree(dma_list);
- kfree(mailbox);
err_out:
return err;
static void mthca_free_eq(struct mthca_dev *dev,
struct mthca_eq *eq)
{
- void *mailbox = NULL;
+ struct mthca_mailbox *mailbox;
int err;
u8 status;
int npages = (eq->nent * MTHCA_EQ_ENTRY_SIZE + PAGE_SIZE - 1) /
PAGE_SIZE;
int i;
- mailbox = kmalloc(sizeof (struct mthca_eq_context) + MTHCA_CMD_MAILBOX_EXTRA,
- GFP_KERNEL);
- if (!mailbox)
+ mailbox = mthca_alloc_mailbox(dev, GFP_KERNEL);
+ if (IS_ERR(mailbox))
return;
- err = mthca_HW2SW_EQ(dev, MAILBOX_ALIGN(mailbox),
- eq->eqn, &status);
+ err = mthca_HW2SW_EQ(dev, mailbox, eq->eqn, &status);
if (err)
mthca_warn(dev, "HW2SW_EQ failed (%d)\n", err);
if (status)
- mthca_warn(dev, "HW2SW_EQ returned status 0x%02x\n",
- status);
+ mthca_warn(dev, "HW2SW_EQ returned status 0x%02x\n", status);
dev->eq_table.arm_mask &= ~eq->eqn_mask;
for (i = 0; i < sizeof (struct mthca_eq_context) / 4; ++i) {
if (i % 4 == 0)
printk("[%02x] ", i * 4);
- printk(" %08x", be32_to_cpup(MAILBOX_ALIGN(mailbox) + i * 4));
+ printk(" %08x", be32_to_cpup(mailbox->buf + i * 4));
if ((i + 1) % 4 == 0)
printk("\n");
}
pci_unmap_addr(&eq->page_list[i], mapping));
kfree(eq->page_list);
- kfree(mailbox);
+ mthca_free_mailbox(dev, mailbox);
}
static void mthca_free_irqs(struct mthca_dev *dev)
if (mthca_map_reg(dev, ((pci_resource_len(dev->pdev, 0) - 1) &
dev->fw.arbel.eq_arm_base) + 4, 4,
&dev->eq_regs.arbel.eq_arm)) {
- mthca_err(dev, "Couldn't map interrupt clear register, "
- "aborting.\n");
+ mthca_err(dev, "Couldn't map EQ arm register, aborting.\n");
mthca_unmap_reg(dev, (pci_resource_len(dev->pdev, 0) - 1) &
dev->fw.arbel.clr_int_base, MTHCA_CLR_INT_SIZE,
dev->clr_base);
dev->fw.arbel.eq_set_ci_base,
MTHCA_EQ_SET_CI_SIZE,
&dev->eq_regs.arbel.eq_set_ci_base)) {
- mthca_err(dev, "Couldn't map interrupt clear register, "
- "aborting.\n");
+ mthca_err(dev, "Couldn't map EQ CI register, aborting.\n");
mthca_unmap_reg(dev, ((pci_resource_len(dev->pdev, 0) - 1) &
dev->fw.arbel.eq_arm_base) + 4, 4,
dev->eq_regs.arbel.eq_arm);
}
-static void __devexit mthca_unmap_eq_regs(struct mthca_dev *dev)
+static void mthca_unmap_eq_regs(struct mthca_dev *dev)
{
if (mthca_is_memfree(dev)) {
mthca_unmap_reg(dev, (pci_resource_len(dev->pdev, 0) - 1) &
return ret;
}
-void __devexit mthca_unmap_eq_icm(struct mthca_dev *dev)
+void mthca_unmap_eq_icm(struct mthca_dev *dev)
{
u8 status;
- mthca_UNMAP_ICM(dev, dev->eq_table.icm_virt, PAGE_SIZE / 4096, &status);
+ mthca_UNMAP_ICM(dev, dev->eq_table.icm_virt, 1, &status);
pci_unmap_page(dev->pdev, dev->eq_table.icm_dma, PAGE_SIZE,
PCI_DMA_BIDIRECTIONAL);
__free_page(dev->eq_table.icm_page);
dev->eq_table.clr_mask =
swab32(1 << (dev->eq_table.inta_pin & 31));
dev->eq_table.clr_int = dev->clr_base +
- (dev->eq_table.inta_pin < 31 ? 4 : 0);
+ (dev->eq_table.inta_pin < 32 ? 4 : 0);
}
dev->eq_table.arm_mask = 0;
intr = (dev->mthca_flags & MTHCA_FLAG_MSI) ?
128 : dev->eq_table.inta_pin;
- err = mthca_create_eq(dev, dev->limits.num_cqs,
+ err = mthca_create_eq(dev, dev->limits.num_cqs + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 128 : intr,
&dev->eq_table.eq[MTHCA_EQ_COMP]);
if (err)
goto err_out_unmap;
- err = mthca_create_eq(dev, MTHCA_NUM_ASYNC_EQE,
+ err = mthca_create_eq(dev, MTHCA_NUM_ASYNC_EQE + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 129 : intr,
&dev->eq_table.eq[MTHCA_EQ_ASYNC]);
if (err)
goto err_out_comp;
- err = mthca_create_eq(dev, MTHCA_NUM_CMD_EQE,
+ err = mthca_create_eq(dev, MTHCA_NUM_CMD_EQE + MTHCA_NUM_SPARE_EQE,
(dev->mthca_flags & MTHCA_FLAG_MSI_X) ? 130 : intr,
&dev->eq_table.eq[MTHCA_EQ_CMD]);
if (err)
mthca_warn(dev, "MAP_EQ for cmd EQ %d returned status 0x%02x\n",
dev->eq_table.eq[MTHCA_EQ_CMD].eqn, status);
- for (i = 0; i < MTHCA_EQ_CMD; ++i)
+ for (i = 0; i < MTHCA_NUM_EQ; ++i)
if (mthca_is_memfree(dev))
arbel_eq_req_not(dev, dev->eq_table.eq[i].eqn_mask);
else
return err;
}
-void __devexit mthca_cleanup_eq_table(struct mthca_dev *dev)
+void mthca_cleanup_eq_table(struct mthca_dev *dev)
{
u8 status;
int i;
git://git.onelab.eu / linux-2.6.git / blobdiff