[deliverable/linux.git] / drivers / infiniband / hw / cxgb4 / device.c

/*
 * Copyright (c) 2009-2010 Chelsio, Inc. 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
 * General Public License (GPL) Version 2, available from the file
 * COPYING in the main directory of this source tree, or the
 * OpenIB.org BSD license below:
 *
 *     Redistribution and use in source and binary forms, with or
 *     without modification, are permitted provided that the following
 *     conditions are met:
 *
 *      - Redistributions of source code must retain the above
 *	  copyright notice, this list of conditions and the following
 *	  disclaimer.
 *
 *      - Redistributions in binary form must reproduce the above
 *	  copyright notice, this list of conditions and the following
 *	  disclaimer in the documentation and/or other materials
 *	  provided with the distribution.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
 * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
 * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
 * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
 * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
 * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
 * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
 * SOFTWARE.
 */
#include <linux/module.h>
#include <linux/moduleparam.h>
#include <linux/debugfs.h>

#include <rdma/ib_verbs.h>

#include "iw_cxgb4.h"

#define DRV_VERSION "0.1"

MODULE_AUTHOR("Steve Wise");
MODULE_DESCRIPTION("Chelsio T4 RDMA Driver");
MODULE_LICENSE("Dual BSD/GPL");
MODULE_VERSION(DRV_VERSION);

static LIST_HEAD(dev_list);
static DEFINE_MUTEX(dev_mutex);

static struct dentry *c4iw_debugfs_root;

struct c4iw_debugfs_data {
	struct c4iw_dev *devp;
	char *buf;
	int bufsize;
	int pos;
};

static int count_idrs(int id, void *p, void *data)
{
	int *countp = data;

	*countp = *countp + 1;
	return 0;
}

static ssize_t debugfs_read(struct file *file, char __user *buf, size_t count,
			    loff_t *ppos)
{
	struct c4iw_debugfs_data *d = file->private_data;
	loff_t pos = *ppos;
	loff_t avail = d->pos;

	if (pos < 0)
		return -EINVAL;
	if (pos >= avail)
		return 0;
	if (count > avail - pos)
		count = avail - pos;

	while (count) {
		size_t len = 0;

		len = min((int)count, (int)d->pos - (int)pos);
		if (copy_to_user(buf, d->buf + pos, len))
			return -EFAULT;
		if (len == 0)
			return -EINVAL;

		buf += len;
		pos += len;
		count -= len;
	}
	count = pos - *ppos;
	*ppos = pos;
	return count;
}

static int dump_qp(int id, void *p, void *data)
{
	struct c4iw_qp *qp = p;
	struct c4iw_debugfs_data *qpd = data;
	int space;
	int cc;

	if (id != qp->wq.sq.qid)
		return 0;

	space = qpd->bufsize - qpd->pos - 1;
	if (space == 0)
		return 1;

	if (qp->ep)
		cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u "
			     "ep tid %u state %u %pI4:%u->%pI4:%u\n",
			     qp->wq.sq.qid, (int)qp->attr.state,
			     qp->ep->hwtid, (int)qp->ep->com.state,
			     &qp->ep->com.local_addr.sin_addr.s_addr,
			     ntohs(qp->ep->com.local_addr.sin_port),
			     &qp->ep->com.remote_addr.sin_addr.s_addr,
			     ntohs(qp->ep->com.remote_addr.sin_port));
	else
		cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u\n",
			      qp->wq.sq.qid, (int)qp->attr.state);
	if (cc < space)
		qpd->pos += cc;
	return 0;
}

static int qp_release(struct inode *inode, struct file *file)
{
	struct c4iw_debugfs_data *qpd = file->private_data;
	if (!qpd) {
		printk(KERN_INFO "%s null qpd?\n", __func__);
		return 0;
	}
	kfree(qpd->buf);
	kfree(qpd);
	return 0;
}

static int qp_open(struct inode *inode, struct file *file)
{
	struct c4iw_debugfs_data *qpd;
	int ret = 0;
	int count = 1;

	qpd = kmalloc(sizeof *qpd, GFP_KERNEL);
	if (!qpd) {
		ret = -ENOMEM;
		goto out;
	}
	qpd->devp = inode->i_private;
	qpd->pos = 0;

	spin_lock_irq(&qpd->devp->lock);
	idr_for_each(&qpd->devp->qpidr, count_idrs, &count);
	spin_unlock_irq(&qpd->devp->lock);

	qpd->bufsize = count * 128;
	qpd->buf = kmalloc(qpd->bufsize, GFP_KERNEL);
	if (!qpd->buf) {
		ret = -ENOMEM;
		goto err1;
	}

	spin_lock_irq(&qpd->devp->lock);
	idr_for_each(&qpd->devp->qpidr, dump_qp, qpd);
	spin_unlock_irq(&qpd->devp->lock);

	qpd->buf[qpd->pos++] = 0;
	file->private_data = qpd;
	goto out;
err1:
	kfree(qpd);
out:
	return ret;
}

static const struct file_operations qp_debugfs_fops = {
	.owner   = THIS_MODULE,
	.open    = qp_open,
	.release = qp_release,
	.read    = debugfs_read,
};

static int dump_stag(int id, void *p, void *data)
{
	struct c4iw_debugfs_data *stagd = data;
	int space;
	int cc;

	space = stagd->bufsize - stagd->pos - 1;
	if (space == 0)
		return 1;

	cc = snprintf(stagd->buf + stagd->pos, space, "0x%x\n", id<<8);
	if (cc < space)
		stagd->pos += cc;
	return 0;
}

static int stag_release(struct inode *inode, struct file *file)
{
	struct c4iw_debugfs_data *stagd = file->private_data;
	if (!stagd) {
		printk(KERN_INFO "%s null stagd?\n", __func__);
		return 0;
	}
	kfree(stagd->buf);
	kfree(stagd);
	return 0;
}

static int stag_open(struct inode *inode, struct file *file)
{
	struct c4iw_debugfs_data *stagd;
	int ret = 0;
	int count = 1;

	stagd = kmalloc(sizeof *stagd, GFP_KERNEL);
	if (!stagd) {
		ret = -ENOMEM;
		goto out;
	}
	stagd->devp = inode->i_private;
	stagd->pos = 0;

	spin_lock_irq(&stagd->devp->lock);
	idr_for_each(&stagd->devp->mmidr, count_idrs, &count);
	spin_unlock_irq(&stagd->devp->lock);

	stagd->bufsize = count * sizeof("0x12345678\n");
	stagd->buf = kmalloc(stagd->bufsize, GFP_KERNEL);
	if (!stagd->buf) {
		ret = -ENOMEM;
		goto err1;
	}

	spin_lock_irq(&stagd->devp->lock);
	idr_for_each(&stagd->devp->mmidr, dump_stag, stagd);
	spin_unlock_irq(&stagd->devp->lock);

	stagd->buf[stagd->pos++] = 0;
	file->private_data = stagd;
	goto out;
err1:
	kfree(stagd);
out:
	return ret;
}

static const struct file_operations stag_debugfs_fops = {
	.owner   = THIS_MODULE,
	.open    = stag_open,
	.release = stag_release,
	.read    = debugfs_read,
};

static int setup_debugfs(struct c4iw_dev *devp)
{
	struct dentry *de;

	if (!devp->debugfs_root)
		return -1;

	de = debugfs_create_file("qps", S_IWUSR, devp->debugfs_root,
				 (void *)devp, &qp_debugfs_fops);
	if (de && de->d_inode)
		de->d_inode->i_size = 4096;

	de = debugfs_create_file("stags", S_IWUSR, devp->debugfs_root,
				 (void *)devp, &stag_debugfs_fops);
	if (de && de->d_inode)
		de->d_inode->i_size = 4096;
	return 0;
}

void c4iw_release_dev_ucontext(struct c4iw_rdev *rdev,
			       struct c4iw_dev_ucontext *uctx)
{
	struct list_head *pos, *nxt;
	struct c4iw_qid_list *entry;

	mutex_lock(&uctx->lock);
	list_for_each_safe(pos, nxt, &uctx->qpids) {
		entry = list_entry(pos, struct c4iw_qid_list, entry);
		list_del_init(&entry->entry);
		if (!(entry->qid & rdev->qpmask))
			c4iw_put_resource(&rdev->resource.qid_fifo, entry->qid,
					  &rdev->resource.qid_fifo_lock);
		kfree(entry);
	}

	list_for_each_safe(pos, nxt, &uctx->qpids) {
		entry = list_entry(pos, struct c4iw_qid_list, entry);
		list_del_init(&entry->entry);
		kfree(entry);
	}
	mutex_unlock(&uctx->lock);
}

void c4iw_init_dev_ucontext(struct c4iw_rdev *rdev,
			    struct c4iw_dev_ucontext *uctx)
{
	INIT_LIST_HEAD(&uctx->qpids);
	INIT_LIST_HEAD(&uctx->cqids);
	mutex_init(&uctx->lock);
}

/* Caller takes care of locking if needed */
static int c4iw_rdev_open(struct c4iw_rdev *rdev)
{
	int err;

	c4iw_init_dev_ucontext(rdev, &rdev->uctx);

	/*
	 * qpshift is the number of bits to shift the qpid left in order
	 * to get the correct address of the doorbell for that qp.
	 */
	rdev->qpshift = PAGE_SHIFT - ilog2(rdev->lldi.udb_density);
	rdev->qpmask = rdev->lldi.udb_density - 1;
	rdev->cqshift = PAGE_SHIFT - ilog2(rdev->lldi.ucq_density);
	rdev->cqmask = rdev->lldi.ucq_density - 1;
	PDBG("%s dev %s stag start 0x%0x size 0x%0x num stags %d "
	     "pbl start 0x%0x size 0x%0x rq start 0x%0x size 0x%0x "
	     "qp qid start %u size %u cq qid start %u size %u\n",
	     __func__, pci_name(rdev->lldi.pdev), rdev->lldi.vr->stag.start,
	     rdev->lldi.vr->stag.size, c4iw_num_stags(rdev),
	     rdev->lldi.vr->pbl.start,
	     rdev->lldi.vr->pbl.size, rdev->lldi.vr->rq.start,
	     rdev->lldi.vr->rq.size,
	     rdev->lldi.vr->qp.start,
	     rdev->lldi.vr->qp.size,
	     rdev->lldi.vr->cq.start,
	     rdev->lldi.vr->cq.size);
	PDBG("udb len 0x%x udb base %p db_reg %p gts_reg %p qpshift %lu "
	     "qpmask 0x%x cqshift %lu cqmask 0x%x\n",
	     (unsigned)pci_resource_len(rdev->lldi.pdev, 2),
	     (void *)pci_resource_start(rdev->lldi.pdev, 2),
	     rdev->lldi.db_reg,
	     rdev->lldi.gts_reg,
	     rdev->qpshift, rdev->qpmask,
	     rdev->cqshift, rdev->cqmask);

	if (c4iw_num_stags(rdev) == 0) {
		err = -EINVAL;
		goto err1;
	}

	err = c4iw_init_resource(rdev, c4iw_num_stags(rdev), T4_MAX_NUM_PD);
	if (err) {
		printk(KERN_ERR MOD "error %d initializing resources\n", err);
		goto err1;
	}
	err = c4iw_pblpool_create(rdev);
	if (err) {
		printk(KERN_ERR MOD "error %d initializing pbl pool\n", err);
		goto err2;
	}
	err = c4iw_rqtpool_create(rdev);
	if (err) {
		printk(KERN_ERR MOD "error %d initializing rqt pool\n", err);
		goto err3;
	}
	err = c4iw_ocqp_pool_create(rdev);
	if (err) {
		printk(KERN_ERR MOD "error %d initializing ocqp pool\n", err);
		goto err4;
	}
	return 0;
err4:
	c4iw_rqtpool_destroy(rdev);
err3:
	c4iw_pblpool_destroy(rdev);
err2:
	c4iw_destroy_resource(&rdev->resource);
err1:
	return err;
}

static void c4iw_rdev_close(struct c4iw_rdev *rdev)
{
	c4iw_pblpool_destroy(rdev);
	c4iw_rqtpool_destroy(rdev);
	c4iw_destroy_resource(&rdev->resource);
}

static void c4iw_remove(struct c4iw_dev *dev)
{
	PDBG("%s c4iw_dev %p\n", __func__,  dev);
	cancel_delayed_work_sync(&dev->db_drop_task);
	list_del(&dev->entry);
	if (dev->registered)
		c4iw_unregister_device(dev);
	c4iw_rdev_close(&dev->rdev);
	idr_destroy(&dev->cqidr);
	idr_destroy(&dev->qpidr);
	idr_destroy(&dev->mmidr);
	iounmap(dev->rdev.oc_mw_kva);
	ib_dealloc_device(&dev->ibdev);
}

static struct c4iw_dev *c4iw_alloc(const struct cxgb4_lld_info *infop)
{
	struct c4iw_dev *devp;
	int ret;

	devp = (struct c4iw_dev *)ib_alloc_device(sizeof(*devp));
	if (!devp) {
		printk(KERN_ERR MOD "Cannot allocate ib device\n");
		return NULL;
	}
	devp->rdev.lldi = *infop;

	devp->rdev.oc_mw_pa = pci_resource_start(devp->rdev.lldi.pdev, 2) +
		(pci_resource_len(devp->rdev.lldi.pdev, 2) -
		 roundup_pow_of_two(devp->rdev.lldi.vr->ocq.size));
	devp->rdev.oc_mw_kva = ioremap_wc(devp->rdev.oc_mw_pa,
					       devp->rdev.lldi.vr->ocq.size);

	printk(KERN_INFO MOD "ocq memory: "
	       "hw_start 0x%x size %u mw_pa 0x%lx mw_kva %p\n",
	       devp->rdev.lldi.vr->ocq.start, devp->rdev.lldi.vr->ocq.size,
	       devp->rdev.oc_mw_pa, devp->rdev.oc_mw_kva);

	mutex_lock(&dev_mutex);

	ret = c4iw_rdev_open(&devp->rdev);
	if (ret) {
		mutex_unlock(&dev_mutex);
		printk(KERN_ERR MOD "Unable to open CXIO rdev err %d\n", ret);
		ib_dealloc_device(&devp->ibdev);
		return NULL;
	}

	idr_init(&devp->cqidr);
	idr_init(&devp->qpidr);
	idr_init(&devp->mmidr);
	spin_lock_init(&devp->lock);
	list_add_tail(&devp->entry, &dev_list);
	mutex_unlock(&dev_mutex);

	if (c4iw_debugfs_root) {
		devp->debugfs_root = debugfs_create_dir(
					pci_name(devp->rdev.lldi.pdev),
					c4iw_debugfs_root);
		setup_debugfs(devp);
	}
	return devp;
}

static void *c4iw_uld_add(const struct cxgb4_lld_info *infop)
{
	struct c4iw_dev *dev;
	static int vers_printed;
	int i;

	if (!vers_printed++)
		printk(KERN_INFO MOD "Chelsio T4 RDMA Driver - version %s\n",
		       DRV_VERSION);

	dev = c4iw_alloc(infop);
	if (!dev)
		goto out;

	PDBG("%s found device %s nchan %u nrxq %u ntxq %u nports %u\n",
	     __func__, pci_name(dev->rdev.lldi.pdev),
	     dev->rdev.lldi.nchan, dev->rdev.lldi.nrxq,
	     dev->rdev.lldi.ntxq, dev->rdev.lldi.nports);

	for (i = 0; i < dev->rdev.lldi.nrxq; i++)
		PDBG("rxqid[%u] %u\n", i, dev->rdev.lldi.rxq_ids[i]);
out:
	return dev;
}

static struct sk_buff *t4_pktgl_to_skb(const struct pkt_gl *gl,
				       unsigned int skb_len,
				       unsigned int pull_len)
{
	struct sk_buff *skb;
	struct skb_shared_info *ssi;

	if (gl->tot_len <= 512) {
		skb = alloc_skb(gl->tot_len, GFP_ATOMIC);
		if (unlikely(!skb))
			goto out;
		__skb_put(skb, gl->tot_len);
		skb_copy_to_linear_data(skb, gl->va, gl->tot_len);
	} else {
		skb = alloc_skb(skb_len, GFP_ATOMIC);
		if (unlikely(!skb))
			goto out;
		__skb_put(skb, pull_len);
		skb_copy_to_linear_data(skb, gl->va, pull_len);

		ssi = skb_shinfo(skb);
		ssi->frags[0].page = gl->frags[0].page;
		ssi->frags[0].page_offset = gl->frags[0].page_offset + pull_len;
		ssi->frags[0].size = gl->frags[0].size - pull_len;
		if (gl->nfrags > 1)
			memcpy(&ssi->frags[1], &gl->frags[1],
			       (gl->nfrags - 1) * sizeof(skb_frag_t));
		ssi->nr_frags = gl->nfrags;

		skb->len = gl->tot_len;
		skb->data_len = skb->len - pull_len;
		skb->truesize += skb->data_len;

		/* Get a reference for the last page, we don't own it */
		get_page(gl->frags[gl->nfrags - 1].page);
	}
out:
	return skb;
}

static int c4iw_uld_rx_handler(void *handle, const __be64 *rsp,
			const struct pkt_gl *gl)
{
	struct c4iw_dev *dev = handle;
	struct sk_buff *skb;
	const struct cpl_act_establish *rpl;
	unsigned int opcode;

	if (gl == NULL) {
		/* omit RSS and rsp_ctrl at end of descriptor */
		unsigned int len = 64 - sizeof(struct rsp_ctrl) - 8;

		skb = alloc_skb(256, GFP_ATOMIC);
		if (!skb)
			goto nomem;
		__skb_put(skb, len);
		skb_copy_to_linear_data(skb, &rsp[1], len);
	} else if (gl == CXGB4_MSG_AN) {
		const struct rsp_ctrl *rc = (void *)rsp;

		u32 qid = be32_to_cpu(rc->pldbuflen_qid);
		c4iw_ev_handler(dev, qid);
		return 0;
	} else {
		skb = t4_pktgl_to_skb(gl, 128, 128);
		if (unlikely(!skb))
			goto nomem;
	}

	rpl = cplhdr(skb);
	opcode = rpl->ot.opcode;

	if (c4iw_handlers[opcode])
		c4iw_handlers[opcode](dev, skb);
	else
		printk(KERN_INFO "%s no handler opcode 0x%x...\n", __func__,
		       opcode);

	return 0;
nomem:
	return -1;
}

static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state)
{
	struct c4iw_dev *dev = handle;

	PDBG("%s new_state %u\n", __func__, new_state);
	switch (new_state) {
	case CXGB4_STATE_UP:
		printk(KERN_INFO MOD "%s: Up\n", pci_name(dev->rdev.lldi.pdev));
		if (!dev->registered) {
			int ret;
			ret = c4iw_register_device(dev);
			if (ret)
				printk(KERN_ERR MOD
				       "%s: RDMA registration failed: %d\n",
				       pci_name(dev->rdev.lldi.pdev), ret);
		}
		break;
	case CXGB4_STATE_DOWN:
		printk(KERN_INFO MOD "%s: Down\n",
		       pci_name(dev->rdev.lldi.pdev));
		if (dev->registered)
			c4iw_unregister_device(dev);
		break;
	case CXGB4_STATE_START_RECOVERY:
		printk(KERN_INFO MOD "%s: Fatal Error\n",
		       pci_name(dev->rdev.lldi.pdev));
		if (dev->registered)
			c4iw_unregister_device(dev);
		break;
	case CXGB4_STATE_DETACH:
		printk(KERN_INFO MOD "%s: Detach\n",
		       pci_name(dev->rdev.lldi.pdev));
		mutex_lock(&dev_mutex);
		c4iw_remove(dev);
		mutex_unlock(&dev_mutex);
		break;
	}
	return 0;
}

static struct cxgb4_uld_info c4iw_uld_info = {
	.name = DRV_NAME,
	.add = c4iw_uld_add,
	.rx_handler = c4iw_uld_rx_handler,
	.state_change = c4iw_uld_state_change,
};

static int __init c4iw_init_module(void)
{
	int err;

	err = c4iw_cm_init();
	if (err)
		return err;

	c4iw_debugfs_root = debugfs_create_dir(DRV_NAME, NULL);
	if (!c4iw_debugfs_root)
		printk(KERN_WARNING MOD
		       "could not create debugfs entry, continuing\n");

	cxgb4_register_uld(CXGB4_ULD_RDMA, &c4iw_uld_info);

	return 0;
}

static void __exit c4iw_exit_module(void)
{
	struct c4iw_dev *dev, *tmp;

	mutex_lock(&dev_mutex);
	list_for_each_entry_safe(dev, tmp, &dev_list, entry) {
		c4iw_remove(dev);
	}
	mutex_unlock(&dev_mutex);
	cxgb4_unregister_uld(CXGB4_ULD_RDMA);
	c4iw_cm_term();
	debugfs_remove_recursive(c4iw_debugfs_root);
}

module_init(c4iw_init_module);
module_exit(c4iw_exit_module);
Commit	Line	Data
cfdda9d7 SW	1	/*
	2	* Copyright (c) 2009-2010 Chelsio, Inc. All rights reserved.
	3	*
	4	* This software is available to you under a choice of one of two
	5	* licenses. You may choose to be licensed under the terms of the GNU
	6	* General Public License (GPL) Version 2, available from the file
	7	* COPYING in the main directory of this source tree, or the
	8	* OpenIB.org BSD license below:
	9	*
	10	* Redistribution and use in source and binary forms, with or
	11	* without modification, are permitted provided that the following
	12	* conditions are met:
	13	*
	14	* - Redistributions of source code must retain the above
	15	* copyright notice, this list of conditions and the following
	16	* disclaimer.
	17	*
	18	* - Redistributions in binary form must reproduce the above
	19	* copyright notice, this list of conditions and the following
	20	* disclaimer in the documentation and/or other materials
	21	* provided with the distribution.
	22	*
	23	* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
	24	* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
	25	* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
	26	* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
	27	* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
	28	* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
	29	* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
	30	* SOFTWARE.
	31	*/
	32	#include <linux/module.h>
	33	#include <linux/moduleparam.h>
	34	#include <linux/debugfs.h>
	35
	36	#include <rdma/ib_verbs.h>
	37
	38	#include "iw_cxgb4.h"
	39
	40	#define DRV_VERSION "0.1"
	41
	42	MODULE_AUTHOR("Steve Wise");
	43	MODULE_DESCRIPTION("Chelsio T4 RDMA Driver");
	44	MODULE_LICENSE("Dual BSD/GPL");
	45	MODULE_VERSION(DRV_VERSION);
	46
	47	static LIST_HEAD(dev_list);
	48	static DEFINE_MUTEX(dev_mutex);
	49
	50	static struct dentry *c4iw_debugfs_root;
	51
9e8d1fa3	52	struct c4iw_debugfs_data {
cfdda9d7 SW	53	struct c4iw_dev *devp;
	54	char *buf;
	55	int bufsize;
	56	int pos;
	57	};
	58
9e8d1fa3	59	static int count_idrs(int id, void p, void data)
cfdda9d7	60	{
cfdda9d7 SW	61	int *countp = data;
cfdda9d7 SW	62
cfdda9d7 SW	63	countp = countp + 1;
	64	return 0;
	65	}
	66
9e8d1fa3 SW	67	static ssize_t debugfs_read(struct file file, char __user buf, size_t count,
	68	loff_t *ppos)
	69	{
	70	struct c4iw_debugfs_data *d = file->private_data;
	71	loff_t pos = *ppos;
	72	loff_t avail = d->pos;
	73
	74	if (pos < 0)
	75	return -EINVAL;
	76	if (pos >= avail)
	77	return 0;
	78	if (count > avail - pos)
	79	count = avail - pos;
	80
	81	while (count) {
	82	size_t len = 0;
	83
	84	len = min((int)count, (int)d->pos - (int)pos);
	85	if (copy_to_user(buf, d->buf + pos, len))
	86	return -EFAULT;
	87	if (len == 0)
	88	return -EINVAL;
	89
	90	buf += len;
	91	pos += len;
	92	count -= len;
	93	}
	94	count = pos - *ppos;
	95	*ppos = pos;
	96	return count;
	97	}
	98
	99	static int dump_qp(int id, void p, void data)
cfdda9d7 SW	100	{
cfdda9d7 SW	101	struct c4iw_qp *qp = p;
9e8d1fa3	102	struct c4iw_debugfs_data *qpd = data;
cfdda9d7 SW	103	int space;
	104	int cc;
	105
	106	if (id != qp->wq.sq.qid)
	107	return 0;
	108
	109	space = qpd->bufsize - qpd->pos - 1;
	110	if (space == 0)
	111	return 1;
	112
	113	if (qp->ep)
	114	cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u "
	115	"ep tid %u state %u %pI4:%u->%pI4:%u\n",
	116	qp->wq.sq.qid, (int)qp->attr.state,
	117	qp->ep->hwtid, (int)qp->ep->com.state,
	118	&qp->ep->com.local_addr.sin_addr.s_addr,
	119	ntohs(qp->ep->com.local_addr.sin_port),
	120	&qp->ep->com.remote_addr.sin_addr.s_addr,
	121	ntohs(qp->ep->com.remote_addr.sin_port));
	122	else
	123	cc = snprintf(qpd->buf + qpd->pos, space, "qp id %u state %u\n",
	124	qp->wq.sq.qid, (int)qp->attr.state);
	125	if (cc < space)
	126	qpd->pos += cc;
	127	return 0;
	128	}
	129
	130	static int qp_release(struct inode inode, struct file file)
	131	{
9e8d1fa3	132	struct c4iw_debugfs_data *qpd = file->private_data;
cfdda9d7 SW	133	if (!qpd) {
	134	printk(KERN_INFO "%s null qpd?\n", __func__);
	135	return 0;
	136	}
	137	kfree(qpd->buf);
	138	kfree(qpd);
	139	return 0;
	140	}
	141
	142	static int qp_open(struct inode inode, struct file file)
	143	{
9e8d1fa3	144	struct c4iw_debugfs_data *qpd;
cfdda9d7 SW	145	int ret = 0;
	146	int count = 1;
	147
	148	qpd = kmalloc(sizeof *qpd, GFP_KERNEL);
	149	if (!qpd) {
	150	ret = -ENOMEM;
	151	goto out;
	152	}
	153	qpd->devp = inode->i_private;
	154	qpd->pos = 0;
	155
	156	spin_lock_irq(&qpd->devp->lock);
9e8d1fa3	157	idr_for_each(&qpd->devp->qpidr, count_idrs, &count);
cfdda9d7 SW	158	spin_unlock_irq(&qpd->devp->lock);
	159
	160	qpd->bufsize = count * 128;
	161	qpd->buf = kmalloc(qpd->bufsize, GFP_KERNEL);
	162	if (!qpd->buf) {
	163	ret = -ENOMEM;
	164	goto err1;
	165	}
	166
	167	spin_lock_irq(&qpd->devp->lock);
9e8d1fa3	168	idr_for_each(&qpd->devp->qpidr, dump_qp, qpd);
cfdda9d7 SW	169	spin_unlock_irq(&qpd->devp->lock);
	170
	171	qpd->buf[qpd->pos++] = 0;
	172	file->private_data = qpd;
	173	goto out;
	174	err1:
	175	kfree(qpd);
	176	out:
	177	return ret;
	178	}
	179
9e8d1fa3 SW	180	static const struct file_operations qp_debugfs_fops = {
	181	.owner = THIS_MODULE,
	182	.open = qp_open,
	183	.release = qp_release,
	184	.read = debugfs_read,
	185	};
	186
	187	static int dump_stag(int id, void p, void data)
cfdda9d7	188	{
9e8d1fa3 SW	189	struct c4iw_debugfs_data *stagd = data;
	190	int space;
	191	int cc;
cfdda9d7	192
9e8d1fa3 SW	193	space = stagd->bufsize - stagd->pos - 1;
	194	if (space == 0)
	195	return 1;
	196
	197	cc = snprintf(stagd->buf + stagd->pos, space, "0x%x\n", id<<8);
	198	if (cc < space)
	199	stagd->pos += cc;
	200	return 0;
	201	}
	202
	203	static int stag_release(struct inode inode, struct file file)
	204	{
	205	struct c4iw_debugfs_data *stagd = file->private_data;
	206	if (!stagd) {
	207	printk(KERN_INFO "%s null stagd?\n", __func__);
cfdda9d7	208	return 0;
9e8d1fa3 SW	209	}
	210	kfree(stagd->buf);
	211	kfree(stagd);
	212	return 0;
	213	}
cfdda9d7	214
9e8d1fa3 SW	215	static int stag_open(struct inode inode, struct file file)
	216	{
	217	struct c4iw_debugfs_data *stagd;
	218	int ret = 0;
	219	int count = 1;
cfdda9d7	220
9e8d1fa3 SW	221	stagd = kmalloc(sizeof *stagd, GFP_KERNEL);
	222	if (!stagd) {
	223	ret = -ENOMEM;
	224	goto out;
	225	}
	226	stagd->devp = inode->i_private;
	227	stagd->pos = 0;
cfdda9d7	228
9e8d1fa3 SW	229	spin_lock_irq(&stagd->devp->lock);
	230	idr_for_each(&stagd->devp->mmidr, count_idrs, &count);
	231	spin_unlock_irq(&stagd->devp->lock);
	232
	233	stagd->bufsize = count * sizeof("0x12345678\n");
	234	stagd->buf = kmalloc(stagd->bufsize, GFP_KERNEL);
	235	if (!stagd->buf) {
	236	ret = -ENOMEM;
	237	goto err1;
cfdda9d7	238	}
9e8d1fa3 SW	239
	240	spin_lock_irq(&stagd->devp->lock);
	241	idr_for_each(&stagd->devp->mmidr, dump_stag, stagd);
	242	spin_unlock_irq(&stagd->devp->lock);
	243
	244	stagd->buf[stagd->pos++] = 0;
	245	file->private_data = stagd;
	246	goto out;
	247	err1:
	248	kfree(stagd);
	249	out:
	250	return ret;
cfdda9d7 SW	251	}
cfdda9d7 SW	252
9e8d1fa3	253	static const struct file_operations stag_debugfs_fops = {
cfdda9d7	254	.owner = THIS_MODULE,
9e8d1fa3 SW	255	.open = stag_open,
	256	.release = stag_release,
	257	.read = debugfs_read,
cfdda9d7 SW	258	};
	259
	260	static int setup_debugfs(struct c4iw_dev *devp)
	261	{
	262	struct dentry *de;
	263
	264	if (!devp->debugfs_root)
	265	return -1;
	266
	267	de = debugfs_create_file("qps", S_IWUSR, devp->debugfs_root,
	268	(void *)devp, &qp_debugfs_fops);
	269	if (de && de->d_inode)
	270	de->d_inode->i_size = 4096;
9e8d1fa3 SW	271
	272	de = debugfs_create_file("stags", S_IWUSR, devp->debugfs_root,
	273	(void *)devp, &stag_debugfs_fops);
	274	if (de && de->d_inode)
	275	de->d_inode->i_size = 4096;
cfdda9d7 SW	276	return 0;
	277	}
	278
	279	void c4iw_release_dev_ucontext(struct c4iw_rdev *rdev,
	280	struct c4iw_dev_ucontext *uctx)
	281	{
	282	struct list_head pos, nxt;
	283	struct c4iw_qid_list *entry;
	284
	285	mutex_lock(&uctx->lock);
	286	list_for_each_safe(pos, nxt, &uctx->qpids) {
	287	entry = list_entry(pos, struct c4iw_qid_list, entry);
	288	list_del_init(&entry->entry);
	289	if (!(entry->qid & rdev->qpmask))
	290	c4iw_put_resource(&rdev->resource.qid_fifo, entry->qid,
	291	&rdev->resource.qid_fifo_lock);
	292	kfree(entry);
	293	}
	294
	295	list_for_each_safe(pos, nxt, &uctx->qpids) {
	296	entry = list_entry(pos, struct c4iw_qid_list, entry);
	297	list_del_init(&entry->entry);
	298	kfree(entry);
	299	}
	300	mutex_unlock(&uctx->lock);
	301	}
	302
	303	void c4iw_init_dev_ucontext(struct c4iw_rdev *rdev,
	304	struct c4iw_dev_ucontext *uctx)
	305	{
	306	INIT_LIST_HEAD(&uctx->qpids);
	307	INIT_LIST_HEAD(&uctx->cqids);
	308	mutex_init(&uctx->lock);
	309	}
	310
	311	/* Caller takes care of locking if needed */
	312	static int c4iw_rdev_open(struct c4iw_rdev *rdev)
	313	{
	314	int err;
	315
	316	c4iw_init_dev_ucontext(rdev, &rdev->uctx);
	317
	318	/*
	319	* qpshift is the number of bits to shift the qpid left in order
	320	* to get the correct address of the doorbell for that qp.
	321	*/
	322	rdev->qpshift = PAGE_SHIFT - ilog2(rdev->lldi.udb_density);
	323	rdev->qpmask = rdev->lldi.udb_density - 1;
	324	rdev->cqshift = PAGE_SHIFT - ilog2(rdev->lldi.ucq_density);
	325	rdev->cqmask = rdev->lldi.ucq_density - 1;
	326	PDBG("%s dev %s stag start 0x%0x size 0x%0x num stags %d "
93fb72e4 SW	327	"pbl start 0x%0x size 0x%0x rq start 0x%0x size 0x%0x "
93fb72e4 SW	328	"qp qid start %u size %u cq qid start %u size %u\n",
cfdda9d7 SW	329	__func__, pci_name(rdev->lldi.pdev), rdev->lldi.vr->stag.start,
	330	rdev->lldi.vr->stag.size, c4iw_num_stags(rdev),
	331	rdev->lldi.vr->pbl.start,
	332	rdev->lldi.vr->pbl.size, rdev->lldi.vr->rq.start,
93fb72e4 SW	333	rdev->lldi.vr->rq.size,
	334	rdev->lldi.vr->qp.start,
	335	rdev->lldi.vr->qp.size,
	336	rdev->lldi.vr->cq.start,
	337	rdev->lldi.vr->cq.size);
cfdda9d7 SW	338	PDBG("udb len 0x%x udb base %p db_reg %p gts_reg %p qpshift %lu "
	339	"qpmask 0x%x cqshift %lu cqmask 0x%x\n",
	340	(unsigned)pci_resource_len(rdev->lldi.pdev, 2),
	341	(void *)pci_resource_start(rdev->lldi.pdev, 2),
	342	rdev->lldi.db_reg,
	343	rdev->lldi.gts_reg,
	344	rdev->qpshift, rdev->qpmask,
	345	rdev->cqshift, rdev->cqmask);
	346
	347	if (c4iw_num_stags(rdev) == 0) {
	348	err = -EINVAL;
	349	goto err1;
	350	}
	351
	352	err = c4iw_init_resource(rdev, c4iw_num_stags(rdev), T4_MAX_NUM_PD);
	353	if (err) {
	354	printk(KERN_ERR MOD "error %d initializing resources\n", err);
	355	goto err1;
	356	}
	357	err = c4iw_pblpool_create(rdev);
	358	if (err) {
	359	printk(KERN_ERR MOD "error %d initializing pbl pool\n", err);
	360	goto err2;
	361	}
	362	err = c4iw_rqtpool_create(rdev);
	363	if (err) {
	364	printk(KERN_ERR MOD "error %d initializing rqt pool\n", err);
	365	goto err3;
	366	}
c6d7b267 SW	367	err = c4iw_ocqp_pool_create(rdev);
	368	if (err) {
	369	printk(KERN_ERR MOD "error %d initializing ocqp pool\n", err);
	370	goto err4;
	371	}
cfdda9d7	372	return 0;
c6d7b267 SW	373	err4:
c6d7b267 SW	374	c4iw_rqtpool_destroy(rdev);
cfdda9d7 SW	375	err3:
	376	c4iw_pblpool_destroy(rdev);
	377	err2:
	378	c4iw_destroy_resource(&rdev->resource);
	379	err1:
	380	return err;
	381	}
	382
	383	static void c4iw_rdev_close(struct c4iw_rdev *rdev)
	384	{
	385	c4iw_pblpool_destroy(rdev);
	386	c4iw_rqtpool_destroy(rdev);
	387	c4iw_destroy_resource(&rdev->resource);
	388	}
	389
	390	static void c4iw_remove(struct c4iw_dev *dev)
	391	{
	392	PDBG("%s c4iw_dev %p\n", __func__, dev);
	393	cancel_delayed_work_sync(&dev->db_drop_task);
	394	list_del(&dev->entry);
1c01c538 SW	395	if (dev->registered)
1c01c538 SW	396	c4iw_unregister_device(dev);
cfdda9d7 SW	397	c4iw_rdev_close(&dev->rdev);
	398	idr_destroy(&dev->cqidr);
	399	idr_destroy(&dev->qpidr);
	400	idr_destroy(&dev->mmidr);
c6d7b267	401	iounmap(dev->rdev.oc_mw_kva);
cfdda9d7 SW	402	ib_dealloc_device(&dev->ibdev);
	403	}
	404
	405	static struct c4iw_dev c4iw_alloc(const struct cxgb4_lld_info infop)
	406	{
	407	struct c4iw_dev *devp;
	408	int ret;
	409
	410	devp = (struct c4iw_dev )ib_alloc_device(sizeof(devp));
	411	if (!devp) {
	412	printk(KERN_ERR MOD "Cannot allocate ib device\n");
	413	return NULL;
	414	}
	415	devp->rdev.lldi = *infop;
	416
c6d7b267 SW	417	devp->rdev.oc_mw_pa = pci_resource_start(devp->rdev.lldi.pdev, 2) +
	418	(pci_resource_len(devp->rdev.lldi.pdev, 2) -
	419	roundup_pow_of_two(devp->rdev.lldi.vr->ocq.size));
	420	devp->rdev.oc_mw_kva = ioremap_wc(devp->rdev.oc_mw_pa,
	421	devp->rdev.lldi.vr->ocq.size);
	422
	423	printk(KERN_INFO MOD "ocq memory: "
	424	"hw_start 0x%x size %u mw_pa 0x%lx mw_kva %p\n",
	425	devp->rdev.lldi.vr->ocq.start, devp->rdev.lldi.vr->ocq.size,
	426	devp->rdev.oc_mw_pa, devp->rdev.oc_mw_kva);
	427
cfdda9d7 SW	428	mutex_lock(&dev_mutex);
	429
	430	ret = c4iw_rdev_open(&devp->rdev);
	431	if (ret) {
	432	mutex_unlock(&dev_mutex);
	433	printk(KERN_ERR MOD "Unable to open CXIO rdev err %d\n", ret);
	434	ib_dealloc_device(&devp->ibdev);
	435	return NULL;
	436	}
	437
	438	idr_init(&devp->cqidr);
	439	idr_init(&devp->qpidr);
	440	idr_init(&devp->mmidr);
	441	spin_lock_init(&devp->lock);
	442	list_add_tail(&devp->entry, &dev_list);
	443	mutex_unlock(&dev_mutex);
	444
cfdda9d7 SW	445	if (c4iw_debugfs_root) {
	446	devp->debugfs_root = debugfs_create_dir(
	447	pci_name(devp->rdev.lldi.pdev),
	448	c4iw_debugfs_root);
	449	setup_debugfs(devp);
	450	}
	451	return devp;
	452	}
	453
	454	static void c4iw_uld_add(const struct cxgb4_lld_info infop)
	455	{
	456	struct c4iw_dev *dev;
	457	static int vers_printed;
	458	int i;
	459
	460	if (!vers_printed++)
	461	printk(KERN_INFO MOD "Chelsio T4 RDMA Driver - version %s\n",
	462	DRV_VERSION);
	463
	464	dev = c4iw_alloc(infop);
	465	if (!dev)
	466	goto out;
	467
	468	PDBG("%s found device %s nchan %u nrxq %u ntxq %u nports %u\n",
	469	__func__, pci_name(dev->rdev.lldi.pdev),
	470	dev->rdev.lldi.nchan, dev->rdev.lldi.nrxq,
	471	dev->rdev.lldi.ntxq, dev->rdev.lldi.nports);
	472
	473	for (i = 0; i < dev->rdev.lldi.nrxq; i++)
	474	PDBG("rxqid[%u] %u\n", i, dev->rdev.lldi.rxq_ids[i]);
cfdda9d7 SW	475	out:
	476	return dev;
	477	}
	478
	479	static struct sk_buff t4_pktgl_to_skb(const struct pkt_gl gl,
	480	unsigned int skb_len,
	481	unsigned int pull_len)
	482	{
	483	struct sk_buff *skb;
	484	struct skb_shared_info *ssi;
	485
	486	if (gl->tot_len <= 512) {
	487	skb = alloc_skb(gl->tot_len, GFP_ATOMIC);
	488	if (unlikely(!skb))
	489	goto out;
	490	__skb_put(skb, gl->tot_len);
	491	skb_copy_to_linear_data(skb, gl->va, gl->tot_len);
	492	} else {
	493	skb = alloc_skb(skb_len, GFP_ATOMIC);
	494	if (unlikely(!skb))
	495	goto out;
	496	__skb_put(skb, pull_len);
	497	skb_copy_to_linear_data(skb, gl->va, pull_len);
	498
	499	ssi = skb_shinfo(skb);
	500	ssi->frags[0].page = gl->frags[0].page;
	501	ssi->frags[0].page_offset = gl->frags[0].page_offset + pull_len;
	502	ssi->frags[0].size = gl->frags[0].size - pull_len;
	503	if (gl->nfrags > 1)
	504	memcpy(&ssi->frags[1], &gl->frags[1],
	505	(gl->nfrags - 1) * sizeof(skb_frag_t));
	506	ssi->nr_frags = gl->nfrags;
	507
	508	skb->len = gl->tot_len;
	509	skb->data_len = skb->len - pull_len;
	510	skb->truesize += skb->data_len;
	511
	512	/* Get a reference for the last page, we don't own it */
	513	get_page(gl->frags[gl->nfrags - 1].page);
	514	}
	515	out:
	516	return skb;
	517	}
	518
	519	static int c4iw_uld_rx_handler(void handle, const __be64 rsp,
	520	const struct pkt_gl *gl)
	521	{
	522	struct c4iw_dev *dev = handle;
	523	struct sk_buff *skb;
	524	const struct cpl_act_establish *rpl;
	525	unsigned int opcode;
	526
	527	if (gl == NULL) {
	528	/* omit RSS and rsp_ctrl at end of descriptor */
	529	unsigned int len = 64 - sizeof(struct rsp_ctrl) - 8;
	530
	531	skb = alloc_skb(256, GFP_ATOMIC);
	532	if (!skb)
	533	goto nomem;
	534	__skb_put(skb, len);
	535	skb_copy_to_linear_data(skb, &rsp[1], len);
	536	} else if (gl == CXGB4_MSG_AN) {
	537	const struct rsp_ctrl rc = (void )rsp;
	538
539	u32 qid = be32_to_cpu(rc->pldbuflen_qid);
540	c4iw_ev_handler(dev, qid);
541	return 0;
542	} else {
543	skb = t4_pktgl_to_skb(gl, 128, 128);
544	if (unlikely(!skb))
545	goto nomem;
546	}
547
548	rpl = cplhdr(skb);
549	opcode = rpl->ot.opcode;
550
551	if (c4iw_handlers[opcode])
552	c4iw_handlers[opcode](dev, skb);
553	else
554	printk(KERN_INFO "%s no handler opcode 0x%x...\n", __func__,
555	opcode);
556
557	return 0;
558	nomem:
559	return -1;
560	}
561
562	static int c4iw_uld_state_change(void *handle, enum cxgb4_state new_state)
563	{
1c01c538 SW	564	struct c4iw_dev *dev = handle;
1c01c538 SW	565
cfdda9d7	566	PDBG("%s new_state %u\n", __func__, new_state);
1c01c538 SW	567	switch (new_state) {
	568	case CXGB4_STATE_UP:
	569	printk(KERN_INFO MOD "%s: Up\n", pci_name(dev->rdev.lldi.pdev));
	570	if (!dev->registered) {
	571	int ret;
	572	ret = c4iw_register_device(dev);
	573	if (ret)
	574	printk(KERN_ERR MOD
	575	"%s: RDMA registration failed: %d\n",
	576	pci_name(dev->rdev.lldi.pdev), ret);
	577	}
	578	break;
	579	case CXGB4_STATE_DOWN:
	580	printk(KERN_INFO MOD "%s: Down\n",
	581	pci_name(dev->rdev.lldi.pdev));
	582	if (dev->registered)
	583	c4iw_unregister_device(dev);
	584	break;
	585	case CXGB4_STATE_START_RECOVERY:
	586	printk(KERN_INFO MOD "%s: Fatal Error\n",
	587	pci_name(dev->rdev.lldi.pdev));
	588	if (dev->registered)
	589	c4iw_unregister_device(dev);
	590	break;
	591	case CXGB4_STATE_DETACH:
	592	printk(KERN_INFO MOD "%s: Detach\n",
	593	pci_name(dev->rdev.lldi.pdev));
	594	mutex_lock(&dev_mutex);
	595	c4iw_remove(dev);
	596	mutex_unlock(&dev_mutex);
	597	break;
	598	}
cfdda9d7 SW	599	return 0;
	600	}
	601
	602	static struct cxgb4_uld_info c4iw_uld_info = {
	603	.name = DRV_NAME,
	604	.add = c4iw_uld_add,
	605	.rx_handler = c4iw_uld_rx_handler,
	606	.state_change = c4iw_uld_state_change,
	607	};
	608
	609	static int __init c4iw_init_module(void)
	610	{
	611	int err;
	612
	613	err = c4iw_cm_init();
	614	if (err)
	615	return err;
	616
	617	c4iw_debugfs_root = debugfs_create_dir(DRV_NAME, NULL);
	618	if (!c4iw_debugfs_root)
	619	printk(KERN_WARNING MOD
	620	"could not create debugfs entry, continuing\n");
	621
	622	cxgb4_register_uld(CXGB4_ULD_RDMA, &c4iw_uld_info);
	623
	624	return 0;
	625	}
	626
	627	static void __exit c4iw_exit_module(void)
	628	{
	629	struct c4iw_dev dev, tmp;
	630
cfdda9d7 SW	631	mutex_lock(&dev_mutex);
	632	list_for_each_entry_safe(dev, tmp, &dev_list, entry) {
	633	c4iw_remove(dev);
	634	}
	635	mutex_unlock(&dev_mutex);
fd388ce6	636	cxgb4_unregister_uld(CXGB4_ULD_RDMA);
cfdda9d7 SW	637	c4iw_cm_term();
	638	debugfs_remove_recursive(c4iw_debugfs_root);
	639	}
	640
	641	module_init(c4iw_init_module);
	642	module_exit(c4iw_exit_module);