From: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
To: dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org
Cc: linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Harish Chegondi
<harish.chegondi-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>,
Jubin John <jubin.john-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Subject: [PATCH 1/7] IB/qib: Remove srq functionality
Date: Wed, 03 Feb 2016 14:20:02 -0800 [thread overview]
Message-ID: <20160203221957.4030.77015.stgit@scvm10.sc.intel.com> (raw)
In-Reply-To: <20160203221808.4030.95183.stgit-9QXIwq+3FY+1XWohqUldA0EOCMrvLtNR@public.gmane.org>
From: Jubin John <jubin.john-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
srq functionality is now in rdmavt. Remove it from the qib driver.
Reviewed-by: Dennis Dalessandro <dennis.dalessandro-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Reviewed-by: Harish Chegondi <harish.chegondi-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
Signed-off-by: Jubin John <jubin.john-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
---
drivers/infiniband/hw/qib/Makefile | 2
drivers/infiniband/hw/qib/qib_srq.c | 380 ---------------------------------
drivers/infiniband/hw/qib/qib_verbs.c | 6 -
drivers/infiniband/hw/qib/qib_verbs.h | 17 -
4 files changed, 1 insertions(+), 404 deletions(-)
delete mode 100644 drivers/infiniband/hw/qib/qib_srq.c
diff --git a/drivers/infiniband/hw/qib/Makefile b/drivers/infiniband/hw/qib/Makefile
index 8d5e36b..79ebd79 100644
--- a/drivers/infiniband/hw/qib/Makefile
+++ b/drivers/infiniband/hw/qib/Makefile
@@ -3,7 +3,7 @@ obj-$(CONFIG_INFINIBAND_QIB) += ib_qib.o
ib_qib-y := qib_diag.o qib_driver.o qib_eeprom.o \
qib_file_ops.o qib_fs.o qib_init.o qib_intr.o \
qib_mad.o qib_pcie.o qib_pio_copy.o \
- qib_qp.o qib_qsfp.o qib_rc.o qib_ruc.o qib_sdma.o qib_srq.o \
+ qib_qp.o qib_qsfp.o qib_rc.o qib_ruc.o qib_sdma.o \
qib_sysfs.o qib_twsi.o qib_tx.o qib_uc.o qib_ud.o \
qib_user_pages.o qib_user_sdma.o qib_iba7220.o \
qib_sd7220.o qib_iba7322.o qib_verbs.o
diff --git a/drivers/infiniband/hw/qib/qib_srq.c b/drivers/infiniband/hw/qib/qib_srq.c
deleted file mode 100644
index dff8808..0000000
--- a/drivers/infiniband/hw/qib/qib_srq.c
+++ /dev/null
@@ -1,380 +0,0 @@
-/*
- * Copyright (c) 2006, 2007, 2008, 2009 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 "qib_verbs.h"
-
-/**
- * qib_post_srq_receive - post a receive on a shared receive queue
- * @ibsrq: the SRQ to post the receive on
- * @wr: the list of work requests to post
- * @bad_wr: A pointer to the first WR to cause a problem is put here
- *
- * This may be called from interrupt context.
- */
-int qib_post_srq_receive(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
- struct ib_recv_wr **bad_wr)
-{
- struct rvt_srq *srq = ibsrq_to_rvtsrq(ibsrq);
- struct rvt_rwq *wq;
- unsigned long flags;
- int ret;
-
- for (; wr; wr = wr->next) {
- struct rvt_rwqe *wqe;
- u32 next;
- int i;
-
- if ((unsigned) wr->num_sge > srq->rq.max_sge) {
- *bad_wr = wr;
- ret = -EINVAL;
- goto bail;
- }
-
- spin_lock_irqsave(&srq->rq.lock, flags);
- wq = srq->rq.wq;
- next = wq->head + 1;
- if (next >= srq->rq.size)
- next = 0;
- if (next == wq->tail) {
- spin_unlock_irqrestore(&srq->rq.lock, flags);
- *bad_wr = wr;
- ret = -ENOMEM;
- goto bail;
- }
-
- wqe = get_rwqe_ptr(&srq->rq, wq->head);
- wqe->wr_id = wr->wr_id;
- wqe->num_sge = wr->num_sge;
- for (i = 0; i < wr->num_sge; i++)
- wqe->sg_list[i] = wr->sg_list[i];
- /* Make sure queue entry is written before the head index. */
- smp_wmb();
- wq->head = next;
- spin_unlock_irqrestore(&srq->rq.lock, flags);
- }
- ret = 0;
-
-bail:
- return ret;
-}
-
-/**
- * qib_create_srq - create a shared receive queue
- * @ibpd: the protection domain of the SRQ to create
- * @srq_init_attr: the attributes of the SRQ
- * @udata: data from libibverbs when creating a user SRQ
- */
-struct ib_srq *qib_create_srq(struct ib_pd *ibpd,
- struct ib_srq_init_attr *srq_init_attr,
- struct ib_udata *udata)
-{
- struct qib_ibdev *dev = to_idev(ibpd->device);
- struct rvt_srq *srq;
- u32 sz;
- struct ib_srq *ret;
-
- if (srq_init_attr->srq_type != IB_SRQT_BASIC) {
- ret = ERR_PTR(-ENOSYS);
- goto done;
- }
-
- if (srq_init_attr->attr.max_sge == 0 ||
- srq_init_attr->attr.max_sge > ib_qib_max_srq_sges ||
- srq_init_attr->attr.max_wr == 0 ||
- srq_init_attr->attr.max_wr > ib_qib_max_srq_wrs) {
- ret = ERR_PTR(-EINVAL);
- goto done;
- }
-
- srq = kmalloc(sizeof(*srq), GFP_KERNEL);
- if (!srq) {
- ret = ERR_PTR(-ENOMEM);
- goto done;
- }
-
- /*
- * Need to use vmalloc() if we want to support large #s of entries.
- */
- srq->rq.size = srq_init_attr->attr.max_wr + 1;
- srq->rq.max_sge = srq_init_attr->attr.max_sge;
- sz = sizeof(struct ib_sge) * srq->rq.max_sge +
- sizeof(struct rvt_rwqe);
- srq->rq.wq = vmalloc_user(sizeof(struct rvt_rwq) + srq->rq.size * sz);
- if (!srq->rq.wq) {
- ret = ERR_PTR(-ENOMEM);
- goto bail_srq;
- }
-
- /*
- * Return the address of the RWQ as the offset to mmap.
- * See qib_mmap() for details.
- */
- if (udata && udata->outlen >= sizeof(__u64)) {
- int err;
- u32 s = sizeof(struct rvt_rwq) + srq->rq.size * sz;
-
- srq->ip =
- rvt_create_mmap_info(&dev->rdi, s, ibpd->uobject->context,
- srq->rq.wq);
- if (!srq->ip) {
- ret = ERR_PTR(-ENOMEM);
- goto bail_wq;
- }
-
- err = ib_copy_to_udata(udata, &srq->ip->offset,
- sizeof(srq->ip->offset));
- if (err) {
- ret = ERR_PTR(err);
- goto bail_ip;
- }
- } else
- srq->ip = NULL;
-
- /*
- * ib_create_srq() will initialize srq->ibsrq.
- */
- spin_lock_init(&srq->rq.lock);
- srq->rq.wq->head = 0;
- srq->rq.wq->tail = 0;
- srq->limit = srq_init_attr->attr.srq_limit;
-
- spin_lock(&dev->n_srqs_lock);
- if (dev->n_srqs_allocated == ib_qib_max_srqs) {
- spin_unlock(&dev->n_srqs_lock);
- ret = ERR_PTR(-ENOMEM);
- goto bail_ip;
- }
-
- dev->n_srqs_allocated++;
- spin_unlock(&dev->n_srqs_lock);
-
- if (srq->ip) {
- spin_lock_irq(&dev->rdi.pending_lock);
- list_add(&srq->ip->pending_mmaps, &dev->rdi.pending_mmaps);
- spin_unlock_irq(&dev->rdi.pending_lock);
- }
-
- ret = &srq->ibsrq;
- goto done;
-
-bail_ip:
- kfree(srq->ip);
-bail_wq:
- vfree(srq->rq.wq);
-bail_srq:
- kfree(srq);
-done:
- return ret;
-}
-
-/**
- * qib_modify_srq - modify a shared receive queue
- * @ibsrq: the SRQ to modify
- * @attr: the new attributes of the SRQ
- * @attr_mask: indicates which attributes to modify
- * @udata: user data for libibverbs.so
- */
-int qib_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
- enum ib_srq_attr_mask attr_mask,
- struct ib_udata *udata)
-{
- struct rvt_srq *srq = ibsrq_to_rvtsrq(ibsrq);
- struct rvt_rwq *wq;
- int ret = 0;
-
- if (attr_mask & IB_SRQ_MAX_WR) {
- struct rvt_rwq *owq;
- struct rvt_rwqe *p;
- u32 sz, size, n, head, tail;
-
- /* Check that the requested sizes are below the limits. */
- if ((attr->max_wr > ib_qib_max_srq_wrs) ||
- ((attr_mask & IB_SRQ_LIMIT) ?
- attr->srq_limit : srq->limit) > attr->max_wr) {
- ret = -EINVAL;
- goto bail;
- }
-
- sz = sizeof(struct rvt_rwqe) +
- srq->rq.max_sge * sizeof(struct ib_sge);
- size = attr->max_wr + 1;
- wq = vmalloc_user(sizeof(struct rvt_rwq) + size * sz);
- if (!wq) {
- ret = -ENOMEM;
- goto bail;
- }
-
- /* Check that we can write the offset to mmap. */
- if (udata && udata->inlen >= sizeof(__u64)) {
- __u64 offset_addr;
- __u64 offset = 0;
-
- ret = ib_copy_from_udata(&offset_addr, udata,
- sizeof(offset_addr));
- if (ret)
- goto bail_free;
- udata->outbuf =
- (void __user *) (unsigned long) offset_addr;
- ret = ib_copy_to_udata(udata, &offset,
- sizeof(offset));
- if (ret)
- goto bail_free;
- }
-
- spin_lock_irq(&srq->rq.lock);
- /*
- * validate head and tail pointer values and compute
- * the number of remaining WQEs.
- */
- owq = srq->rq.wq;
- head = owq->head;
- tail = owq->tail;
- if (head >= srq->rq.size || tail >= srq->rq.size) {
- ret = -EINVAL;
- goto bail_unlock;
- }
- n = head;
- if (n < tail)
- n += srq->rq.size - tail;
- else
- n -= tail;
- if (size <= n) {
- ret = -EINVAL;
- goto bail_unlock;
- }
- n = 0;
- p = wq->wq;
- while (tail != head) {
- struct rvt_rwqe *wqe;
- int i;
-
- wqe = get_rwqe_ptr(&srq->rq, tail);
- p->wr_id = wqe->wr_id;
- p->num_sge = wqe->num_sge;
- for (i = 0; i < wqe->num_sge; i++)
- p->sg_list[i] = wqe->sg_list[i];
- n++;
- p = (struct rvt_rwqe *)((char *)p + sz);
- if (++tail >= srq->rq.size)
- tail = 0;
- }
- srq->rq.wq = wq;
- srq->rq.size = size;
- wq->head = n;
- wq->tail = 0;
- if (attr_mask & IB_SRQ_LIMIT)
- srq->limit = attr->srq_limit;
- spin_unlock_irq(&srq->rq.lock);
-
- vfree(owq);
-
- if (srq->ip) {
- struct rvt_mmap_info *ip = srq->ip;
- struct qib_ibdev *dev = to_idev(srq->ibsrq.device);
- u32 s = sizeof(struct rvt_rwq) + size * sz;
-
- rvt_update_mmap_info(&dev->rdi, ip, s, wq);
-
- /*
- * Return the offset to mmap.
- * See qib_mmap() for details.
- */
- if (udata && udata->inlen >= sizeof(__u64)) {
- ret = ib_copy_to_udata(udata, &ip->offset,
- sizeof(ip->offset));
- if (ret)
- goto bail;
- }
-
- /*
- * Put user mapping info onto the pending list
- * unless it already is on the list.
- */
- spin_lock_irq(&dev->rdi.pending_lock);
- if (list_empty(&ip->pending_mmaps))
- list_add(&ip->pending_mmaps,
- &dev->rdi.pending_mmaps);
- spin_unlock_irq(&dev->rdi.pending_lock);
- }
- } else if (attr_mask & IB_SRQ_LIMIT) {
- spin_lock_irq(&srq->rq.lock);
- if (attr->srq_limit >= srq->rq.size)
- ret = -EINVAL;
- else
- srq->limit = attr->srq_limit;
- spin_unlock_irq(&srq->rq.lock);
- }
- goto bail;
-
-bail_unlock:
- spin_unlock_irq(&srq->rq.lock);
-bail_free:
- vfree(wq);
-bail:
- return ret;
-}
-
-int qib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr)
-{
- struct rvt_srq *srq = ibsrq_to_rvtsrq(ibsrq);
-
- attr->max_wr = srq->rq.size - 1;
- attr->max_sge = srq->rq.max_sge;
- attr->srq_limit = srq->limit;
- return 0;
-}
-
-/**
- * qib_destroy_srq - destroy a shared receive queue
- * @ibsrq: the SRQ to destroy
- */
-int qib_destroy_srq(struct ib_srq *ibsrq)
-{
- struct rvt_srq *srq = ibsrq_to_rvtsrq(ibsrq);
- struct qib_ibdev *dev = to_idev(ibsrq->device);
-
- spin_lock(&dev->n_srqs_lock);
- dev->n_srqs_allocated--;
- spin_unlock(&dev->n_srqs_lock);
- if (srq->ip)
- kref_put(&srq->ip->ref, rvt_release_mmap_info);
- else
- vfree(srq->rq.wq);
- kfree(srq);
-
- return 0;
-}
diff --git a/drivers/infiniband/hw/qib/qib_verbs.c b/drivers/infiniband/hw/qib/qib_verbs.c
index 8b97ca1..3785a52 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.c
+++ b/drivers/infiniband/hw/qib/qib_verbs.c
@@ -1656,7 +1656,6 @@ int qib_register_ib_device(struct qib_devdata *dd)
/* Only need to initialize non-zero fields. */
spin_lock_init(&dev->n_qps_lock);
- spin_lock_init(&dev->n_srqs_lock);
init_timer(&dev->mem_timer);
dev->mem_timer.function = mem_timer;
dev->mem_timer.data = (unsigned long) dev;
@@ -1754,17 +1753,12 @@ int qib_register_ib_device(struct qib_devdata *dd)
ibdev->destroy_ah = NULL;
ibdev->modify_ah = NULL;
ibdev->query_ah = NULL;
- ibdev->create_srq = qib_create_srq;
- ibdev->modify_srq = qib_modify_srq;
- ibdev->query_srq = qib_query_srq;
- ibdev->destroy_srq = qib_destroy_srq;
ibdev->create_qp = NULL;
ibdev->modify_qp = qib_modify_qp;
ibdev->query_qp = NULL;
ibdev->destroy_qp = qib_destroy_qp;
ibdev->post_send = NULL;
ibdev->post_recv = NULL;
- ibdev->post_srq_recv = qib_post_srq_receive;
ibdev->create_cq = NULL;
ibdev->destroy_cq = NULL;
ibdev->resize_cq = NULL;
diff --git a/drivers/infiniband/hw/qib/qib_verbs.h b/drivers/infiniband/hw/qib/qib_verbs.h
index 34f7784..a7e3c71 100644
--- a/drivers/infiniband/hw/qib/qib_verbs.h
+++ b/drivers/infiniband/hw/qib/qib_verbs.h
@@ -270,8 +270,6 @@ struct qib_ibdev {
u32 n_qps_allocated; /* number of QPs allocated for device */
spinlock_t n_qps_lock;
- u32 n_srqs_allocated; /* number of SRQs allocated for device */
- spinlock_t n_srqs_lock;
#ifdef CONFIG_DEBUG_FS
/* per HCA debugfs */
struct dentry *qib_ibdev_dbg;
@@ -428,21 +426,6 @@ int qib_post_ud_send(struct rvt_qp *qp, struct ib_send_wr *wr);
void qib_ud_rcv(struct qib_ibport *ibp, struct qib_ib_header *hdr,
int has_grh, void *data, u32 tlen, struct rvt_qp *qp);
-int qib_post_srq_receive(struct ib_srq *ibsrq, struct ib_recv_wr *wr,
- struct ib_recv_wr **bad_wr);
-
-struct ib_srq *qib_create_srq(struct ib_pd *ibpd,
- struct ib_srq_init_attr *srq_init_attr,
- struct ib_udata *udata);
-
-int qib_modify_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr,
- enum ib_srq_attr_mask attr_mask,
- struct ib_udata *udata);
-
-int qib_query_srq(struct ib_srq *ibsrq, struct ib_srq_attr *attr);
-
-int qib_destroy_srq(struct ib_srq *ibsrq);
-
void mr_rcu_callback(struct rcu_head *list);
static inline void qib_put_ss(struct rvt_sge_state *ss)
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
next prev parent reply other threads:[~2016-02-03 22:20 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-02-03 22:19 [PATCH 0/7] IB/qib: Finish adding core rdmavt verbs support Dennis Dalessandro
[not found] ` <20160203221808.4030.95183.stgit-9QXIwq+3FY+1XWohqUldA0EOCMrvLtNR@public.gmane.org>
2016-02-03 22:20 ` Dennis Dalessandro [this message]
2016-02-03 22:20 ` [PATCH 2/7] IB/qib: Clean up register_ib_device Dennis Dalessandro
2016-02-03 22:20 ` [PATCH 3/7] IB/qib: Remove qib_lookup_qpn and use rvt_lookup_qpn instead Dennis Dalessandro
2016-02-03 22:20 ` [PATCH 4/7] IB/qib: Remove modify queue pair code Dennis Dalessandro
2016-02-03 22:20 ` [PATCH 5/7] IB/qib: Remove destroy " Dennis Dalessandro
2016-02-03 22:20 ` [PATCH 6/7] IB/qib: Support query gid in rdmavt Dennis Dalessandro
2016-02-03 22:20 ` [PATCH 7/7] IB/qib: Remove modify_port and port_immutable functions Dennis Dalessandro
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20160203221957.4030.77015.stgit@scvm10.sc.intel.com \
--to=dennis.dalessandro-ral2jqcrhueavxtiumwx3w@public.gmane.org \
--cc=dledford-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org \
--cc=harish.chegondi-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org \
--cc=jubin.john-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org \
--cc=linux-rdma-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.