mirror of
https://github.com/torvalds/linux.git
synced 2024-12-24 20:01:55 +00:00
85caafe307
The current workqueue implemention has the following performance deficiencies on QDR HCAs: - The CQ call backs tend to run on the CPUs processing the receive queues - The single thread queue isn't optimal for multiple HCAs This patch adds a dedicated per HCA bound thread to process CQ callbacks. Reviewed-by: Ramkrishna Vepa <ramkrishna.vepa@intel.com> Signed-off-by: Mike Marciniszyn <mike.marciniszyn@intel.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
541 lines
13 KiB
C
541 lines
13 KiB
C
/*
|
|
* Copyright (c) 2013 Intel Corporation. All rights reserved.
|
|
* Copyright (c) 2006, 2007, 2008, 2010 QLogic Corporation. All rights reserved.
|
|
* Copyright (c) 2005, 2006 PathScale, 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/err.h>
|
|
#include <linux/slab.h>
|
|
#include <linux/vmalloc.h>
|
|
#include <linux/kthread.h>
|
|
|
|
#include "qib_verbs.h"
|
|
#include "qib.h"
|
|
|
|
/**
|
|
* qib_cq_enter - add a new entry to the completion queue
|
|
* @cq: completion queue
|
|
* @entry: work completion entry to add
|
|
* @sig: true if @entry is a solicitated entry
|
|
*
|
|
* This may be called with qp->s_lock held.
|
|
*/
|
|
void qib_cq_enter(struct qib_cq *cq, struct ib_wc *entry, int solicited)
|
|
{
|
|
struct qib_cq_wc *wc;
|
|
unsigned long flags;
|
|
u32 head;
|
|
u32 next;
|
|
|
|
spin_lock_irqsave(&cq->lock, flags);
|
|
|
|
/*
|
|
* Note that the head pointer might be writable by user processes.
|
|
* Take care to verify it is a sane value.
|
|
*/
|
|
wc = cq->queue;
|
|
head = wc->head;
|
|
if (head >= (unsigned) cq->ibcq.cqe) {
|
|
head = cq->ibcq.cqe;
|
|
next = 0;
|
|
} else
|
|
next = head + 1;
|
|
if (unlikely(next == wc->tail)) {
|
|
spin_unlock_irqrestore(&cq->lock, flags);
|
|
if (cq->ibcq.event_handler) {
|
|
struct ib_event ev;
|
|
|
|
ev.device = cq->ibcq.device;
|
|
ev.element.cq = &cq->ibcq;
|
|
ev.event = IB_EVENT_CQ_ERR;
|
|
cq->ibcq.event_handler(&ev, cq->ibcq.cq_context);
|
|
}
|
|
return;
|
|
}
|
|
if (cq->ip) {
|
|
wc->uqueue[head].wr_id = entry->wr_id;
|
|
wc->uqueue[head].status = entry->status;
|
|
wc->uqueue[head].opcode = entry->opcode;
|
|
wc->uqueue[head].vendor_err = entry->vendor_err;
|
|
wc->uqueue[head].byte_len = entry->byte_len;
|
|
wc->uqueue[head].ex.imm_data =
|
|
(__u32 __force)entry->ex.imm_data;
|
|
wc->uqueue[head].qp_num = entry->qp->qp_num;
|
|
wc->uqueue[head].src_qp = entry->src_qp;
|
|
wc->uqueue[head].wc_flags = entry->wc_flags;
|
|
wc->uqueue[head].pkey_index = entry->pkey_index;
|
|
wc->uqueue[head].slid = entry->slid;
|
|
wc->uqueue[head].sl = entry->sl;
|
|
wc->uqueue[head].dlid_path_bits = entry->dlid_path_bits;
|
|
wc->uqueue[head].port_num = entry->port_num;
|
|
/* Make sure entry is written before the head index. */
|
|
smp_wmb();
|
|
} else
|
|
wc->kqueue[head] = *entry;
|
|
wc->head = next;
|
|
|
|
if (cq->notify == IB_CQ_NEXT_COMP ||
|
|
(cq->notify == IB_CQ_SOLICITED &&
|
|
(solicited || entry->status != IB_WC_SUCCESS))) {
|
|
struct kthread_worker *worker;
|
|
/*
|
|
* This will cause send_complete() to be called in
|
|
* another thread.
|
|
*/
|
|
smp_rmb();
|
|
worker = cq->dd->worker;
|
|
if (likely(worker)) {
|
|
cq->notify = IB_CQ_NONE;
|
|
cq->triggered++;
|
|
queue_kthread_work(worker, &cq->comptask);
|
|
}
|
|
}
|
|
|
|
spin_unlock_irqrestore(&cq->lock, flags);
|
|
}
|
|
|
|
/**
|
|
* qib_poll_cq - poll for work completion entries
|
|
* @ibcq: the completion queue to poll
|
|
* @num_entries: the maximum number of entries to return
|
|
* @entry: pointer to array where work completions are placed
|
|
*
|
|
* Returns the number of completion entries polled.
|
|
*
|
|
* This may be called from interrupt context. Also called by ib_poll_cq()
|
|
* in the generic verbs code.
|
|
*/
|
|
int qib_poll_cq(struct ib_cq *ibcq, int num_entries, struct ib_wc *entry)
|
|
{
|
|
struct qib_cq *cq = to_icq(ibcq);
|
|
struct qib_cq_wc *wc;
|
|
unsigned long flags;
|
|
int npolled;
|
|
u32 tail;
|
|
|
|
/* The kernel can only poll a kernel completion queue */
|
|
if (cq->ip) {
|
|
npolled = -EINVAL;
|
|
goto bail;
|
|
}
|
|
|
|
spin_lock_irqsave(&cq->lock, flags);
|
|
|
|
wc = cq->queue;
|
|
tail = wc->tail;
|
|
if (tail > (u32) cq->ibcq.cqe)
|
|
tail = (u32) cq->ibcq.cqe;
|
|
for (npolled = 0; npolled < num_entries; ++npolled, ++entry) {
|
|
if (tail == wc->head)
|
|
break;
|
|
/* The kernel doesn't need a RMB since it has the lock. */
|
|
*entry = wc->kqueue[tail];
|
|
if (tail >= cq->ibcq.cqe)
|
|
tail = 0;
|
|
else
|
|
tail++;
|
|
}
|
|
wc->tail = tail;
|
|
|
|
spin_unlock_irqrestore(&cq->lock, flags);
|
|
|
|
bail:
|
|
return npolled;
|
|
}
|
|
|
|
static void send_complete(struct kthread_work *work)
|
|
{
|
|
struct qib_cq *cq = container_of(work, struct qib_cq, comptask);
|
|
|
|
/*
|
|
* The completion handler will most likely rearm the notification
|
|
* and poll for all pending entries. If a new completion entry
|
|
* is added while we are in this routine, queue_work()
|
|
* won't call us again until we return so we check triggered to
|
|
* see if we need to call the handler again.
|
|
*/
|
|
for (;;) {
|
|
u8 triggered = cq->triggered;
|
|
|
|
/*
|
|
* IPoIB connected mode assumes the callback is from a
|
|
* soft IRQ. We simulate this by blocking "bottom halves".
|
|
* See the implementation for ipoib_cm_handle_tx_wc(),
|
|
* netif_tx_lock_bh() and netif_tx_lock().
|
|
*/
|
|
local_bh_disable();
|
|
cq->ibcq.comp_handler(&cq->ibcq, cq->ibcq.cq_context);
|
|
local_bh_enable();
|
|
|
|
if (cq->triggered == triggered)
|
|
return;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* qib_create_cq - create a completion queue
|
|
* @ibdev: the device this completion queue is attached to
|
|
* @entries: the minimum size of the completion queue
|
|
* @context: unused by the QLogic_IB driver
|
|
* @udata: user data for libibverbs.so
|
|
*
|
|
* Returns a pointer to the completion queue or negative errno values
|
|
* for failure.
|
|
*
|
|
* Called by ib_create_cq() in the generic verbs code.
|
|
*/
|
|
struct ib_cq *qib_create_cq(struct ib_device *ibdev, int entries,
|
|
int comp_vector, struct ib_ucontext *context,
|
|
struct ib_udata *udata)
|
|
{
|
|
struct qib_ibdev *dev = to_idev(ibdev);
|
|
struct qib_cq *cq;
|
|
struct qib_cq_wc *wc;
|
|
struct ib_cq *ret;
|
|
u32 sz;
|
|
|
|
if (entries < 1 || entries > ib_qib_max_cqes) {
|
|
ret = ERR_PTR(-EINVAL);
|
|
goto done;
|
|
}
|
|
|
|
/* Allocate the completion queue structure. */
|
|
cq = kmalloc(sizeof(*cq), GFP_KERNEL);
|
|
if (!cq) {
|
|
ret = ERR_PTR(-ENOMEM);
|
|
goto done;
|
|
}
|
|
|
|
/*
|
|
* Allocate the completion queue entries and head/tail pointers.
|
|
* This is allocated separately so that it can be resized and
|
|
* also mapped into user space.
|
|
* We need to use vmalloc() in order to support mmap and large
|
|
* numbers of entries.
|
|
*/
|
|
sz = sizeof(*wc);
|
|
if (udata && udata->outlen >= sizeof(__u64))
|
|
sz += sizeof(struct ib_uverbs_wc) * (entries + 1);
|
|
else
|
|
sz += sizeof(struct ib_wc) * (entries + 1);
|
|
wc = vmalloc_user(sz);
|
|
if (!wc) {
|
|
ret = ERR_PTR(-ENOMEM);
|
|
goto bail_cq;
|
|
}
|
|
|
|
/*
|
|
* Return the address of the WC as the offset to mmap.
|
|
* See qib_mmap() for details.
|
|
*/
|
|
if (udata && udata->outlen >= sizeof(__u64)) {
|
|
int err;
|
|
|
|
cq->ip = qib_create_mmap_info(dev, sz, context, wc);
|
|
if (!cq->ip) {
|
|
ret = ERR_PTR(-ENOMEM);
|
|
goto bail_wc;
|
|
}
|
|
|
|
err = ib_copy_to_udata(udata, &cq->ip->offset,
|
|
sizeof(cq->ip->offset));
|
|
if (err) {
|
|
ret = ERR_PTR(err);
|
|
goto bail_ip;
|
|
}
|
|
} else
|
|
cq->ip = NULL;
|
|
|
|
spin_lock(&dev->n_cqs_lock);
|
|
if (dev->n_cqs_allocated == ib_qib_max_cqs) {
|
|
spin_unlock(&dev->n_cqs_lock);
|
|
ret = ERR_PTR(-ENOMEM);
|
|
goto bail_ip;
|
|
}
|
|
|
|
dev->n_cqs_allocated++;
|
|
spin_unlock(&dev->n_cqs_lock);
|
|
|
|
if (cq->ip) {
|
|
spin_lock_irq(&dev->pending_lock);
|
|
list_add(&cq->ip->pending_mmaps, &dev->pending_mmaps);
|
|
spin_unlock_irq(&dev->pending_lock);
|
|
}
|
|
|
|
/*
|
|
* ib_create_cq() will initialize cq->ibcq except for cq->ibcq.cqe.
|
|
* The number of entries should be >= the number requested or return
|
|
* an error.
|
|
*/
|
|
cq->dd = dd_from_dev(dev);
|
|
cq->ibcq.cqe = entries;
|
|
cq->notify = IB_CQ_NONE;
|
|
cq->triggered = 0;
|
|
spin_lock_init(&cq->lock);
|
|
init_kthread_work(&cq->comptask, send_complete);
|
|
wc->head = 0;
|
|
wc->tail = 0;
|
|
cq->queue = wc;
|
|
|
|
ret = &cq->ibcq;
|
|
|
|
goto done;
|
|
|
|
bail_ip:
|
|
kfree(cq->ip);
|
|
bail_wc:
|
|
vfree(wc);
|
|
bail_cq:
|
|
kfree(cq);
|
|
done:
|
|
return ret;
|
|
}
|
|
|
|
/**
|
|
* qib_destroy_cq - destroy a completion queue
|
|
* @ibcq: the completion queue to destroy.
|
|
*
|
|
* Returns 0 for success.
|
|
*
|
|
* Called by ib_destroy_cq() in the generic verbs code.
|
|
*/
|
|
int qib_destroy_cq(struct ib_cq *ibcq)
|
|
{
|
|
struct qib_ibdev *dev = to_idev(ibcq->device);
|
|
struct qib_cq *cq = to_icq(ibcq);
|
|
|
|
flush_kthread_work(&cq->comptask);
|
|
spin_lock(&dev->n_cqs_lock);
|
|
dev->n_cqs_allocated--;
|
|
spin_unlock(&dev->n_cqs_lock);
|
|
if (cq->ip)
|
|
kref_put(&cq->ip->ref, qib_release_mmap_info);
|
|
else
|
|
vfree(cq->queue);
|
|
kfree(cq);
|
|
|
|
return 0;
|
|
}
|
|
|
|
/**
|
|
* qib_req_notify_cq - change the notification type for a completion queue
|
|
* @ibcq: the completion queue
|
|
* @notify_flags: the type of notification to request
|
|
*
|
|
* Returns 0 for success.
|
|
*
|
|
* This may be called from interrupt context. Also called by
|
|
* ib_req_notify_cq() in the generic verbs code.
|
|
*/
|
|
int qib_req_notify_cq(struct ib_cq *ibcq, enum ib_cq_notify_flags notify_flags)
|
|
{
|
|
struct qib_cq *cq = to_icq(ibcq);
|
|
unsigned long flags;
|
|
int ret = 0;
|
|
|
|
spin_lock_irqsave(&cq->lock, flags);
|
|
/*
|
|
* Don't change IB_CQ_NEXT_COMP to IB_CQ_SOLICITED but allow
|
|
* any other transitions (see C11-31 and C11-32 in ch. 11.4.2.2).
|
|
*/
|
|
if (cq->notify != IB_CQ_NEXT_COMP)
|
|
cq->notify = notify_flags & IB_CQ_SOLICITED_MASK;
|
|
|
|
if ((notify_flags & IB_CQ_REPORT_MISSED_EVENTS) &&
|
|
cq->queue->head != cq->queue->tail)
|
|
ret = 1;
|
|
|
|
spin_unlock_irqrestore(&cq->lock, flags);
|
|
|
|
return ret;
|
|
}
|
|
|
|
/**
|
|
* qib_resize_cq - change the size of the CQ
|
|
* @ibcq: the completion queue
|
|
*
|
|
* Returns 0 for success.
|
|
*/
|
|
int qib_resize_cq(struct ib_cq *ibcq, int cqe, struct ib_udata *udata)
|
|
{
|
|
struct qib_cq *cq = to_icq(ibcq);
|
|
struct qib_cq_wc *old_wc;
|
|
struct qib_cq_wc *wc;
|
|
u32 head, tail, n;
|
|
int ret;
|
|
u32 sz;
|
|
|
|
if (cqe < 1 || cqe > ib_qib_max_cqes) {
|
|
ret = -EINVAL;
|
|
goto bail;
|
|
}
|
|
|
|
/*
|
|
* Need to use vmalloc() if we want to support large #s of entries.
|
|
*/
|
|
sz = sizeof(*wc);
|
|
if (udata && udata->outlen >= sizeof(__u64))
|
|
sz += sizeof(struct ib_uverbs_wc) * (cqe + 1);
|
|
else
|
|
sz += sizeof(struct ib_wc) * (cqe + 1);
|
|
wc = vmalloc_user(sz);
|
|
if (!wc) {
|
|
ret = -ENOMEM;
|
|
goto bail;
|
|
}
|
|
|
|
/* Check that we can write the offset to mmap. */
|
|
if (udata && udata->outlen >= sizeof(__u64)) {
|
|
__u64 offset = 0;
|
|
|
|
ret = ib_copy_to_udata(udata, &offset, sizeof(offset));
|
|
if (ret)
|
|
goto bail_free;
|
|
}
|
|
|
|
spin_lock_irq(&cq->lock);
|
|
/*
|
|
* Make sure head and tail are sane since they
|
|
* might be user writable.
|
|
*/
|
|
old_wc = cq->queue;
|
|
head = old_wc->head;
|
|
if (head > (u32) cq->ibcq.cqe)
|
|
head = (u32) cq->ibcq.cqe;
|
|
tail = old_wc->tail;
|
|
if (tail > (u32) cq->ibcq.cqe)
|
|
tail = (u32) cq->ibcq.cqe;
|
|
if (head < tail)
|
|
n = cq->ibcq.cqe + 1 + head - tail;
|
|
else
|
|
n = head - tail;
|
|
if (unlikely((u32)cqe < n)) {
|
|
ret = -EINVAL;
|
|
goto bail_unlock;
|
|
}
|
|
for (n = 0; tail != head; n++) {
|
|
if (cq->ip)
|
|
wc->uqueue[n] = old_wc->uqueue[tail];
|
|
else
|
|
wc->kqueue[n] = old_wc->kqueue[tail];
|
|
if (tail == (u32) cq->ibcq.cqe)
|
|
tail = 0;
|
|
else
|
|
tail++;
|
|
}
|
|
cq->ibcq.cqe = cqe;
|
|
wc->head = n;
|
|
wc->tail = 0;
|
|
cq->queue = wc;
|
|
spin_unlock_irq(&cq->lock);
|
|
|
|
vfree(old_wc);
|
|
|
|
if (cq->ip) {
|
|
struct qib_ibdev *dev = to_idev(ibcq->device);
|
|
struct qib_mmap_info *ip = cq->ip;
|
|
|
|
qib_update_mmap_info(dev, ip, sz, wc);
|
|
|
|
/*
|
|
* Return the offset to mmap.
|
|
* See qib_mmap() for details.
|
|
*/
|
|
if (udata && udata->outlen >= sizeof(__u64)) {
|
|
ret = ib_copy_to_udata(udata, &ip->offset,
|
|
sizeof(ip->offset));
|
|
if (ret)
|
|
goto bail;
|
|
}
|
|
|
|
spin_lock_irq(&dev->pending_lock);
|
|
if (list_empty(&ip->pending_mmaps))
|
|
list_add(&ip->pending_mmaps, &dev->pending_mmaps);
|
|
spin_unlock_irq(&dev->pending_lock);
|
|
}
|
|
|
|
ret = 0;
|
|
goto bail;
|
|
|
|
bail_unlock:
|
|
spin_unlock_irq(&cq->lock);
|
|
bail_free:
|
|
vfree(wc);
|
|
bail:
|
|
return ret;
|
|
}
|
|
|
|
int qib_cq_init(struct qib_devdata *dd)
|
|
{
|
|
int ret = 0;
|
|
int cpu;
|
|
struct task_struct *task;
|
|
|
|
if (dd->worker)
|
|
return 0;
|
|
dd->worker = kzalloc(sizeof(*dd->worker), GFP_KERNEL);
|
|
if (!dd->worker)
|
|
return -ENOMEM;
|
|
init_kthread_worker(dd->worker);
|
|
task = kthread_create_on_node(
|
|
kthread_worker_fn,
|
|
dd->worker,
|
|
dd->assigned_node_id,
|
|
"qib_cq%d", dd->unit);
|
|
if (IS_ERR(task))
|
|
goto task_fail;
|
|
cpu = cpumask_first(cpumask_of_node(dd->assigned_node_id));
|
|
kthread_bind(task, cpu);
|
|
wake_up_process(task);
|
|
out:
|
|
return ret;
|
|
task_fail:
|
|
ret = PTR_ERR(task);
|
|
kfree(dd->worker);
|
|
dd->worker = NULL;
|
|
goto out;
|
|
}
|
|
|
|
void qib_cq_exit(struct qib_devdata *dd)
|
|
{
|
|
struct kthread_worker *worker;
|
|
|
|
worker = dd->worker;
|
|
if (!worker)
|
|
return;
|
|
/* blocks future queuing from send_complete() */
|
|
dd->worker = NULL;
|
|
smp_wmb();
|
|
flush_kthread_worker(worker);
|
|
kthread_stop(worker->task);
|
|
kfree(worker);
|
|
}
|