openkylin
/
linux
mirror of https://gitee.com/openkylin/linux.git
9
0
Fork 0
Code Issues Projects Releases Wiki Activity
linux/drivers/infiniband/hw/cxgb4/provider.c

615 lines
18 KiB
C
Raw Normal View History

RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
/*
* 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/device.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/delay.h>
#include <linux/errno.h>
#include <linux/list.h>
#include <linux/spinlock.h>
#include <linux/ethtool.h>
#include <linux/rtnetlink.h>
#include <linux/inetdevice.h>
#include <linux/io.h>
#include <asm/irq.h>
#include <asm/byteorder.h>
#include <rdma/iw_cm.h>
#include <rdma/ib_verbs.h>
#include <rdma/ib_smi.h>
#include <rdma/ib_umem.h>
#include <rdma/ib_user_verbs.h>
#include "iw_cxgb4.h"
RDMA/cxgb4: Fastreg NSMR fixes - Remove dsgl support - doesn't work in T4. - Wrap the immediate PBL as needed when building it in the wr. - Adjust max pbl depth allowed based on ulptx alignment requirements. - Bump the slots per SQ to 5 to allow up to 128MB fast registers. - Advertise fastreg support by default. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-18 04:40:15 +08:00
static int fastreg_support = 1;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
module_param(fastreg_support, int, 0644);
RDMA/cxgb4: Fastreg NSMR fixes - Remove dsgl support - doesn't work in T4. - Wrap the immediate PBL as needed when building it in the wr. - Adjust max pbl depth allowed based on ulptx alignment requirements. - Bump the slots per SQ to 5 to allow up to 128MB fast registers. - Advertise fastreg support by default. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-18 04:40:15 +08:00
MODULE_PARM_DESC(fastreg_support, "Advertise fastreg support (default=1)");
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
static struct ib_ah *c4iw_ah_create(struct ib_pd *pd,
struct ib_ah_attr *ah_attr)
{
return ERR_PTR(-ENOSYS);
}
static int c4iw_ah_destroy(struct ib_ah *ah)
{
return -ENOSYS;
}
static int c4iw_multicast_attach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
{
return -ENOSYS;
}
static int c4iw_multicast_detach(struct ib_qp *ibqp, union ib_gid *gid, u16 lid)
{
return -ENOSYS;
}
static int c4iw_process_mad(struct ib_device *ibdev, int mad_flags,
IB/core cleanup: Add const on args - device->process_mad The process_mad device function declares some parameters as "in". Make those parameters const and adjust the call tree under process_mad in the various drivers accordingly. Signed-off-by: Ira Weiny <ira.weiny@intel.com> Reviewed-by: Hal Rosenstock <hal@mellanox.com> Reviewed-by: Jason Gunthorpe <jgunthorpe@obsidianresearch.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-06-01 05:15:30 +08:00
u8 port_num, const struct ib_wc *in_wc,
const struct ib_grh *in_grh,
IB/mad: Add support for additional MAD info to/from drivers In order to support alternate sized MADs (and variable sized MADs on OPA devices) add in/out MAD size parameters to the process_mad core call. In addition, add an out_mad_pkey_index to communicate the pkey index the driver wishes the MAD stack to use when sending OPA MAD responses. The out MAD size and the out MAD PKey index are required by the MAD stack to generate responses on OPA devices. Furthermore, the in and out MAD parameters are made generic by specifying them as ib_mad_hdr rather than ib_mad. Drivers are modified as needed and are protected by BUG_ON flags if the MAD sizes passed to them is incorrect. Signed-off-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-06-07 02:38:31 +08:00
const struct ib_mad_hdr *in_mad,
size_t in_mad_size,
struct ib_mad_hdr *out_mad,
size_t *out_mad_size,
u16 *out_mad_pkey_index)
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
{
return -ENOSYS;
}
static int c4iw_dealloc_ucontext(struct ib_ucontext *context)
{
struct c4iw_dev *rhp = to_c4iw_dev(context->device);
struct c4iw_ucontext *ucontext = to_c4iw_ucontext(context);
struct c4iw_mm_entry *mm, *tmp;
PDBG("%s context %p\n", __func__, context);
list_for_each_entry_safe(mm, tmp, &ucontext->mmaps, entry)
kfree(mm);
c4iw_release_dev_ucontext(&rhp->rdev, &ucontext->uctx);
kfree(ucontext);
return 0;
}
static struct ib_ucontext *c4iw_alloc_ucontext(struct ib_device *ibdev,
struct ib_udata *udata)
{
struct c4iw_ucontext *context;
struct c4iw_dev *rhp = to_c4iw_dev(ibdev);
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
static int warned;
struct c4iw_alloc_ucontext_resp uresp;
int ret = 0;
struct c4iw_mm_entry *mm = NULL;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
PDBG("%s ibdev %p\n", __func__, ibdev);
context = kzalloc(sizeof(*context), GFP_KERNEL);
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
if (!context) {
ret = -ENOMEM;
goto err;
}
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
c4iw_init_dev_ucontext(&rhp->rdev, &context->uctx);
INIT_LIST_HEAD(&context->mmaps);
spin_lock_init(&context->mmap_lock);
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
RDMA/cxgb4: add missing padding at end of struct c4iw_alloc_ucontext_resp The i386 ABI disagrees with most other ABIs regarding alignment of data types larger than 4 bytes: on most ABIs a padding must be added at end of the structures, while it is not required on i386. So for most ABI struct c4iw_alloc_ucontext_resp gets implicitly padded to be aligned on a 8 bytes multiple, while for i386, such padding is not added. The tool pahole can be used to find such implicit padding: $ pahole --anon_include \ --nested_anon_include \ --recursive \ --class_name c4iw_alloc_ucontext_resp \ drivers/infiniband/hw/cxgb4/iw_cxgb4.o Then, structure layout can be compared between i386 and x86_64: +++ obj-i386/drivers/infiniband/hw/cxgb4/iw_cxgb4.o.pahole.txt 2014-03-28 11:43:05.547432195 +0100 --- obj-x86_64/drivers/infiniband/hw/cxgb4/iw_cxgb4.o.pahole.txt 2014-03-28 10:55:10.990133017 +0100 @@ -2,9 +2,8 @@ struct c4iw_alloc_ucontext_resp { __u64 status_page_key; /* 0 8 */ __u32 status_page_size; /* 8 4 */ - /* size: 12, cachelines: 1, members: 2 */ - /* last cacheline: 12 bytes */ + /* size: 16, cachelines: 1, members: 2 */ + /* padding: 4 */ + /* last cacheline: 16 bytes */ }; This ABI disagreement will make an x86_64 kernel try to write past the buffer provided by an i386 binary. When boundary check will be implemented, the x86_64 kernel will refuse to write past the i386 userspace provided buffer and the uverbs will fail. If the structure is on a page boundary and the next page is not mapped, ib_copy_to_udata() will fail and the uverb will fail. Additionally, as reported by Dan Carpenter, without the implicit padding being properly cleared, an information leak would take place in most architectures. This patch adds an explicit padding to struct c4iw_alloc_ucontext_resp, and, like 92b0ca7cb149 ("IB/mlx5: Fix stack info leak in mlx5_ib_alloc_ucontext()"), makes function c4iw_alloc_ucontext() not writting this padding field to userspace. This way, x86_64 kernel will be able to write struct c4iw_alloc_ucontext_resp as expected by unpatched and patched i386 libcxgb4. Link: http://marc.info/?i=cover.1399309513.git.ydroneaud@opteya.com Link: http://marc.info/?i=1395848977.3297.15.camel@localhost.localdomain Link: http://marc.info/?i=20140328082428.GH25192@mwanda Cc: <stable@vger.kernel.org> Fixes: 05eb23893c2c ("cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes") Reported-by: Yann Droneaud <ydroneaud@opteya.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-05-06 01:35:26 +08:00
if (udata->outlen < sizeof(uresp) - sizeof(uresp.reserved)) {
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
if (!warned++)
pr_err(MOD "Warning - downlevel libcxgb4 (non-fatal), device status page disabled.");
rhp->rdev.flags |= T4_STATUS_PAGE_DISABLED;
} else {
mm = kmalloc(sizeof(*mm), GFP_KERNEL);
RDMA/cxgb4: set error code on kmalloc() failure If kmalloc() fails in c4iw_alloc_ucontext(), the function leaves but does not set an error code in ret variable: it will return 0 to the caller. This patch set ret to -ENOMEM in such case. Cc: Steve Wise <swise@opengridcomputing.com> Cc: Steve Wise <swise@chelsio.com> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-29 02:55:21 +08:00
if (!mm) {
ret = -ENOMEM;
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
goto err_free;
RDMA/cxgb4: set error code on kmalloc() failure If kmalloc() fails in c4iw_alloc_ucontext(), the function leaves but does not set an error code in ret variable: it will return 0 to the caller. This patch set ret to -ENOMEM in such case. Cc: Steve Wise <swise@opengridcomputing.com> Cc: Steve Wise <swise@chelsio.com> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-29 02:55:21 +08:00
}
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
uresp.status_page_size = PAGE_SIZE;
spin_lock(&context->mmap_lock);
uresp.status_page_key = context->key;
context->key += PAGE_SIZE;
spin_unlock(&context->mmap_lock);
RDMA/cxgb4: add missing padding at end of struct c4iw_alloc_ucontext_resp The i386 ABI disagrees with most other ABIs regarding alignment of data types larger than 4 bytes: on most ABIs a padding must be added at end of the structures, while it is not required on i386. So for most ABI struct c4iw_alloc_ucontext_resp gets implicitly padded to be aligned on a 8 bytes multiple, while for i386, such padding is not added. The tool pahole can be used to find such implicit padding: $ pahole --anon_include \ --nested_anon_include \ --recursive \ --class_name c4iw_alloc_ucontext_resp \ drivers/infiniband/hw/cxgb4/iw_cxgb4.o Then, structure layout can be compared between i386 and x86_64: +++ obj-i386/drivers/infiniband/hw/cxgb4/iw_cxgb4.o.pahole.txt 2014-03-28 11:43:05.547432195 +0100 --- obj-x86_64/drivers/infiniband/hw/cxgb4/iw_cxgb4.o.pahole.txt 2014-03-28 10:55:10.990133017 +0100 @@ -2,9 +2,8 @@ struct c4iw_alloc_ucontext_resp { __u64 status_page_key; /* 0 8 */ __u32 status_page_size; /* 8 4 */ - /* size: 12, cachelines: 1, members: 2 */ - /* last cacheline: 12 bytes */ + /* size: 16, cachelines: 1, members: 2 */ + /* padding: 4 */ + /* last cacheline: 16 bytes */ }; This ABI disagreement will make an x86_64 kernel try to write past the buffer provided by an i386 binary. When boundary check will be implemented, the x86_64 kernel will refuse to write past the i386 userspace provided buffer and the uverbs will fail. If the structure is on a page boundary and the next page is not mapped, ib_copy_to_udata() will fail and the uverb will fail. Additionally, as reported by Dan Carpenter, without the implicit padding being properly cleared, an information leak would take place in most architectures. This patch adds an explicit padding to struct c4iw_alloc_ucontext_resp, and, like 92b0ca7cb149 ("IB/mlx5: Fix stack info leak in mlx5_ib_alloc_ucontext()"), makes function c4iw_alloc_ucontext() not writting this padding field to userspace. This way, x86_64 kernel will be able to write struct c4iw_alloc_ucontext_resp as expected by unpatched and patched i386 libcxgb4. Link: http://marc.info/?i=cover.1399309513.git.ydroneaud@opteya.com Link: http://marc.info/?i=1395848977.3297.15.camel@localhost.localdomain Link: http://marc.info/?i=20140328082428.GH25192@mwanda Cc: <stable@vger.kernel.org> Fixes: 05eb23893c2c ("cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes") Reported-by: Yann Droneaud <ydroneaud@opteya.com> Reported-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Yann Droneaud <ydroneaud@opteya.com> Acked-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-05-06 01:35:26 +08:00
ret = ib_copy_to_udata(udata, &uresp,
sizeof(uresp) - sizeof(uresp.reserved));
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
if (ret)
goto err_mm;
mm->key = uresp.status_page_key;
mm->addr = virt_to_phys(rhp->rdev.status_page);
mm->len = PAGE_SIZE;
insert_mmap(context, mm);
}
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
return &context->ibucontext;
cxgb4/iw_cxgb4: Doorbell Drop Avoidance Bug Fixes The current logic suffers from a slow response time to disable user DB usage, and also fails to avoid DB FIFO drops under heavy load. This commit fixes these deficiencies and makes the avoidance logic more optimal. This is done by more efficiently notifying the ULDs of potential DB problems, and implements a smoother flow control algorithm in iw_cxgb4, which is the ULD that puts the most load on the DB fifo. Design: cxgb4: Direct ULD callback from the DB FULL/DROP interrupt handler. This allows the ULD to stop doing user DB writes as quickly as possible. While user DB usage is disabled, the LLD will accumulate DB write events for its queues. Then once DB usage is reenabled, a single DB write is done for each queue with its accumulated write count. This reduces the load put on the DB fifo when reenabling. iw_cxgb4: Instead of marking each qp to indicate DB writes are disabled, we create a device-global status page that each user process maps. This allows iw_cxgb4 to only set this single bit to disable all DB writes for all user QPs vs traversing the idr of all the active QPs. If the libcxgb4 doesn't support this, then we fall back to the old approach of marking each QP. Thus we allow the new driver to work with an older libcxgb4. When the LLD upcalls iw_cxgb4 indicating DB FULL, we disable all DB writes via the status page and transition the DB state to STOPPED. As user processes see that DB writes are disabled, they call into iw_cxgb4 to submit their DB write events. Since the DB state is in STOPPED, the QP trying to write gets enqueued on a new DB "flow control" list. As subsequent DB writes are submitted for this flow controlled QP, the amount of writes are accumulated for each QP on the flow control list. So all the user QPs that are actively ringing the DB get put on this list and the number of writes they request are accumulated. When the LLD upcalls iw_cxgb4 indicating DB EMPTY, which is in a workq context, we change the DB state to FLOW_CONTROL, and begin resuming all the QPs that are on the flow control list. This logic runs on until the flow control list is empty or we exit FLOW_CONTROL mode (due to a DB DROP upcall, for example). QPs are removed from this list, and their accumulated DB write counts written to the DB FIFO. Sets of QPs, called chunks in the code, are removed at one time. The chunk size is 64. So 64 QPs are resumed at a time, and before the next chunk is resumed, the logic waits (blocks) for the DB FIFO to drain. This prevents resuming to quickly and overflowing the FIFO. Once the flow control list is empty, the db state transitions back to NORMAL and user QPs are again allowed to write directly to the user DB register. The algorithm is designed such that if the DB write load is high enough, then all the DB writes get submitted by the kernel using this flow controlled approach to avoid DB drops. As the load lightens though, we resume to normal DB writes directly by user applications. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-03-15 00:22:08 +08:00
err_mm:
kfree(mm);
err_free:
kfree(context);
err:
return ERR_PTR(ret);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
}
static int c4iw_mmap(struct ib_ucontext *context, struct vm_area_struct *vma)
{
int len = vma->vm_end - vma->vm_start;
u32 key = vma->vm_pgoff << PAGE_SHIFT;
struct c4iw_rdev *rdev;
int ret = 0;
struct c4iw_mm_entry *mm;
struct c4iw_ucontext *ucontext;
u64 addr;
PDBG("%s pgoff 0x%lx key 0x%x len %d\n", __func__, vma->vm_pgoff,
key, len);
if (vma->vm_start & (PAGE_SIZE-1))
return -EINVAL;
rdev = &(to_c4iw_dev(context->device)->rdev);
ucontext = to_c4iw_ucontext(context);
mm = remove_mmap(ucontext, key, len);
if (!mm)
return -EINVAL;
addr = mm->addr;
kfree(mm);
RDMA/cxgb4: Support on-chip SQs T4 support on-chip SQs to reduce latency. This patch adds support for this in iw_cxgb4: - Manage ocqp memory like other adapter mem resources. - Allocate user mode SQs from ocqp mem if available. - Map ocqp mem to user process using write combining. - Map PCIE_MA_SYNC reg to user process. Bump uverbs ABI. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-14 00:23:57 +08:00
if ((addr >= pci_resource_start(rdev->lldi.pdev, 0)) &&
(addr < (pci_resource_start(rdev->lldi.pdev, 0) +
pci_resource_len(rdev->lldi.pdev, 0)))) {
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
/*
RDMA/cxgb4: Support on-chip SQs T4 support on-chip SQs to reduce latency. This patch adds support for this in iw_cxgb4: - Manage ocqp memory like other adapter mem resources. - Allocate user mode SQs from ocqp mem if available. - Map ocqp mem to user process using write combining. - Map PCIE_MA_SYNC reg to user process. Bump uverbs ABI. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-14 00:23:57 +08:00
* MA_SYNC register...
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
*/
vma->vm_page_prot = pgprot_noncached(vma->vm_page_prot);
RDMA/cxgb4: Support on-chip SQs T4 support on-chip SQs to reduce latency. This patch adds support for this in iw_cxgb4: - Manage ocqp memory like other adapter mem resources. - Allocate user mode SQs from ocqp mem if available. - Map ocqp mem to user process using write combining. - Map PCIE_MA_SYNC reg to user process. Bump uverbs ABI. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-14 00:23:57 +08:00
ret = io_remap_pfn_range(vma, vma->vm_start,
addr >> PAGE_SHIFT,
len, vma->vm_page_prot);
} else if ((addr >= pci_resource_start(rdev->lldi.pdev, 2)) &&
(addr < (pci_resource_start(rdev->lldi.pdev, 2) +
pci_resource_len(rdev->lldi.pdev, 2)))) {
/*
* Map user DB or OCQP memory...
*/
if (addr >= rdev->oc_mw_pa)
vma->vm_page_prot = t4_pgprot_wc(vma->vm_page_prot);
RDMA/cxgb4: Add Support for Chelsio T5 adapter Adds support for Chelsio T5 adapter. Enables T5's Write Combining feature. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-14 13:08:58 +08:00
else {
if (is_t5(rdev->lldi.adapter_type))
vma->vm_page_prot =
t4_pgprot_wc(vma->vm_page_prot);
else
vma->vm_page_prot =
pgprot_noncached(vma->vm_page_prot);
}
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
ret = io_remap_pfn_range(vma, vma->vm_start,
addr >> PAGE_SHIFT,
len, vma->vm_page_prot);
} else {
/*
* Map WQ or CQ contig dma memory...
*/
ret = remap_pfn_range(vma, vma->vm_start,
addr >> PAGE_SHIFT,
len, vma->vm_page_prot);
}
return ret;
}
static int c4iw_deallocate_pd(struct ib_pd *pd)
{
struct c4iw_dev *rhp;
struct c4iw_pd *php;
php = to_c4iw_pd(pd);
rhp = php->rhp;
PDBG("%s ibpd %p pdid 0x%x\n", __func__, pd, php->pdid);
RDMA/cxgb4: Remove kfifo usage Using kfifos for ID management was limiting the number of QPs and preventing NP384 MPI jobs. So replace it with a simple bitmap allocator. Remove IDs from the IDR tables before deallocating them. This bug was causing the BUG_ON() in insert_handle() to fire because the ID was getting reused before being removed from the IDR table. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:32 +08:00
c4iw_put_resource(&rhp->rdev.resource.pdid_table, php->pdid);
RDMA/cxgb4: Add debugfs RDMA memory stats Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:27 +08:00
mutex_lock(&rhp->rdev.stats.lock);
rhp->rdev.stats.pd.cur--;
mutex_unlock(&rhp->rdev.stats.lock);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
kfree(php);
return 0;
}
static struct ib_pd *c4iw_allocate_pd(struct ib_device *ibdev,
struct ib_ucontext *context,
struct ib_udata *udata)
{
struct c4iw_pd *php;
u32 pdid;
struct c4iw_dev *rhp;
PDBG("%s ibdev %p\n", __func__, ibdev);
rhp = (struct c4iw_dev *) ibdev;
RDMA/cxgb4: Remove kfifo usage Using kfifos for ID management was limiting the number of QPs and preventing NP384 MPI jobs. So replace it with a simple bitmap allocator. Remove IDs from the IDR tables before deallocating them. This bug was causing the BUG_ON() in insert_handle() to fire because the ID was getting reused before being removed from the IDR table. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:32 +08:00
pdid = c4iw_get_resource(&rhp->rdev.resource.pdid_table);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
if (!pdid)
return ERR_PTR(-EINVAL);
php = kzalloc(sizeof(*php), GFP_KERNEL);
if (!php) {
RDMA/cxgb4: Remove kfifo usage Using kfifos for ID management was limiting the number of QPs and preventing NP384 MPI jobs. So replace it with a simple bitmap allocator. Remove IDs from the IDR tables before deallocating them. This bug was causing the BUG_ON() in insert_handle() to fire because the ID was getting reused before being removed from the IDR table. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:32 +08:00
c4iw_put_resource(&rhp->rdev.resource.pdid_table, pdid);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
return ERR_PTR(-ENOMEM);
}
php->pdid = pdid;
php->rhp = rhp;
if (context) {
if (ib_copy_to_udata(udata, &php->pdid, sizeof(u32))) {
c4iw_deallocate_pd(&php->ibpd);
return ERR_PTR(-EFAULT);
}
}
RDMA/cxgb4: Add debugfs RDMA memory stats Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:27 +08:00
mutex_lock(&rhp->rdev.stats.lock);
rhp->rdev.stats.pd.cur++;
if (rhp->rdev.stats.pd.cur > rhp->rdev.stats.pd.max)
rhp->rdev.stats.pd.max = rhp->rdev.stats.pd.cur;
mutex_unlock(&rhp->rdev.stats.lock);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
PDBG("%s pdid 0x%0x ptr 0x%p\n", __func__, pdid, php);
return &php->ibpd;
}
static int c4iw_query_pkey(struct ib_device *ibdev, u8 port, u16 index,
u16 *pkey)
{
PDBG("%s ibdev %p\n", __func__, ibdev);
*pkey = 0;
return 0;
}
static int c4iw_query_gid(struct ib_device *ibdev, u8 port, int index,
union ib_gid *gid)
{
struct c4iw_dev *dev;
PDBG("%s ibdev %p, port %d, index %d, gid %p\n",
__func__, ibdev, port, index, gid);
dev = to_c4iw_dev(ibdev);
BUG_ON(port == 0);
memset(&(gid->raw[0]), 0, sizeof(gid->raw));
memcpy(&(gid->raw[0]), dev->rdev.lldi.ports[port-1]->dev_addr, 6);
return 0;
}
IB/core: Pass hardware specific data in query_device Vendors should be able to pass vendor specific data to/from user-space via query_device uverb. In order to do this, we need to pass the vendors' specific udata. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-06-11 21:35:25 +08:00
static int c4iw_query_device(struct ib_device *ibdev, struct ib_device_attr *props,
struct ib_udata *uhw)
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
{
struct c4iw_dev *dev;
IB/core: Pass hardware specific data in query_device Vendors should be able to pass vendor specific data to/from user-space via query_device uverb. In order to do this, we need to pass the vendors' specific udata. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-06-11 21:35:25 +08:00
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
PDBG("%s ibdev %p\n", __func__, ibdev);
IB/core: Pass hardware specific data in query_device Vendors should be able to pass vendor specific data to/from user-space via query_device uverb. In order to do this, we need to pass the vendors' specific udata. Signed-off-by: Matan Barak <matanb@mellanox.com> Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-06-11 21:35:25 +08:00
if (uhw->inlen || uhw->outlen)
return -EINVAL;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
dev = to_c4iw_dev(ibdev);
memset(props, 0, sizeof *props);
memcpy(&props->sys_image_guid, dev->rdev.lldi.ports[0]->dev_addr, 6);
RDMA/cxgb4: Add Support for Chelsio T5 adapter Adds support for Chelsio T5 adapter. Enables T5's Write Combining feature. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-14 13:08:58 +08:00
props->hw_ver = CHELSIO_CHIP_RELEASE(dev->rdev.lldi.adapter_type);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
props->fw_ver = dev->rdev.lldi.fw_vers;
props->device_cap_flags = dev->device_cap_flags;
props->page_size_cap = T4_PAGESIZE_MASK;
props->vendor_id = (u32)dev->rdev.lldi.pdev->vendor;
props->vendor_part_id = (u32)dev->rdev.lldi.pdev->device;
props->max_mr_size = T4_MAX_MR_SIZE;
iw_cxgb4: advertise the correct device max attributes Advertise the actual max limits for things like qp depths, number of qps, cqs, etc. Clean up the queue allocation for qps and cqs. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-21 23:25:15 +08:00
props->max_qp = dev->rdev.lldi.vr->qp.size / 2;
iw_cxgb4: Detect Ing. Padding Boundary at run-time Updates iw_cxgb4 to determine the Ingress Padding Boundary from cxgb4_lld_info, and take subsequent actions. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-15 00:04:51 +08:00
props->max_qp_wr = dev->rdev.hw_queue.t4_max_qp_depth;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
props->max_sge = T4_MAX_RECV_SGE;
props->max_sge_rd = 1;
cxgb4/iw_cxgb4: use firmware ord/ird resource limits Advertise a larger max read queue depth for qps, and gather the resource limits from fw and use them to avoid exhaustinq all the resources. Design: cxgb4: Obtain the max_ordird_qp and max_ird_adapter device params from FW at init time and pass them up to the ULDs when they attach. If these parameters are not available, due to older firmware, then hard-code the values based on the known values for older firmware. iw_cxgb4: Fix the c4iw_query_device() to report these correct values based on adapter parameters. ibv_query_device() will always return: max_qp_rd_atom = max_qp_init_rd_atom = min(module_max, max_ordird_qp) max_res_rd_atom = max_ird_adapter Bump up the per qp max module option to 32, allowing it to be increased by the user up to the device max of max_ordird_qp. 32 seems to be sufficient to maximize throughput for streaming read benchmarks. Fail connection setup if the negotiated IRD exhausts the available adapter ird resources. So the driver will track the amount of ird resource in use and not send an RI_WR/INIT to FW that would reduce the available ird resources below zero. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-15 00:04:52 +08:00
props->max_res_rd_atom = dev->rdev.lldi.max_ird_adapter;
props->max_qp_rd_atom = min(dev->rdev.lldi.max_ordird_qp,
c4iw_max_read_depth);
props->max_qp_init_rd_atom = props->max_qp_rd_atom;
iw_cxgb4: advertise the correct device max attributes Advertise the actual max limits for things like qp depths, number of qps, cqs, etc. Clean up the queue allocation for qps and cqs. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-21 23:25:15 +08:00
props->max_cq = dev->rdev.lldi.vr->qp.size;
iw_cxgb4: Detect Ing. Padding Boundary at run-time Updates iw_cxgb4 to determine the Ingress Padding Boundary from cxgb4_lld_info, and take subsequent actions. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-07-15 00:04:51 +08:00
props->max_cqe = dev->rdev.hw_queue.t4_max_cq_depth;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
props->max_mr = c4iw_num_stags(&dev->rdev);
props->max_pd = T4_MAX_NUM_PD;
props->local_ca_ack_delay = 0;
RDMA/cxgb4: Max fastreg depth depends on DSGL support The max depth of a fastreg mr depends on whether the device supports DSGL or not. So compute it dynamically based on the device support and the module use_dsgl option. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2014-04-09 22:38:27 +08:00
props->max_fast_reg_page_list_len = t4_max_fr_depth(use_dsgl);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
return 0;
}
static int c4iw_query_port(struct ib_device *ibdev, u8 port,
struct ib_port_attr *props)
{
struct c4iw_dev *dev;
struct net_device *netdev;
struct in_device *inetdev;
PDBG("%s ibdev %p\n", __func__, ibdev);
dev = to_c4iw_dev(ibdev);
netdev = dev->rdev.lldi.ports[port-1];
memset(props, 0, sizeof(struct ib_port_attr));
props->max_mtu = IB_MTU_4096;
if (netdev->mtu >= 4096)
props->active_mtu = IB_MTU_4096;
else if (netdev->mtu >= 2048)
props->active_mtu = IB_MTU_2048;
else if (netdev->mtu >= 1024)
props->active_mtu = IB_MTU_1024;
else if (netdev->mtu >= 512)
props->active_mtu = IB_MTU_512;
else
props->active_mtu = IB_MTU_256;
if (!netif_carrier_ok(netdev))
props->state = IB_PORT_DOWN;
else {
inetdev = in_dev_get(netdev);
if (inetdev) {
if (inetdev->ifa_list)
props->state = IB_PORT_ACTIVE;
else
props->state = IB_PORT_INIT;
in_dev_put(inetdev);
} else
props->state = IB_PORT_INIT;
}
props->port_cap_flags =
IB_PORT_CM_SUP |
IB_PORT_SNMP_TUNNEL_SUP |
IB_PORT_REINIT_SUP |
IB_PORT_DEVICE_MGMT_SUP |
IB_PORT_VENDOR_CLASS_SUP | IB_PORT_BOOT_MGMT_SUP;
props->gid_tbl_len = 1;
props->pkey_tbl_len = 1;
props->active_width = 2;
IB: Use central enum for speed instead of hard-coded values The kernel IB stack uses one enumeration for IB speed, which wasn't explicitly specified in the verbs header file. Add that enum, and use it all over the code. The IB speed/width notation is also used by iWARP and IBoE HW drivers, which use the convention of rate = speed * width to advertise their port link rate. Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-02-29 00:49:50 +08:00
props->active_speed = IB_SPEED_DDR;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
props->max_msg_sz = -1;
return 0;
}
static ssize_t show_rev(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
ibdev.dev);
PDBG("%s dev 0x%p\n", __func__, dev);
RDMA/cxgb4: Add Support for Chelsio T5 adapter Adds support for Chelsio T5 adapter. Enables T5's Write Combining feature. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2013-03-14 13:08:58 +08:00
return sprintf(buf, "%d\n",
CHELSIO_CHIP_RELEASE(c4iw_dev->rdev.lldi.adapter_type));
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
}
static ssize_t show_fw_ver(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
ibdev.dev);
PDBG("%s dev 0x%p\n", __func__, dev);
return sprintf(buf, "%u.%u.%u.%u\n",
RDMA/cxgb4/cxgb4vf/csiostor: Cleanup macros/register defines related to PCIE, RSS and FW This patch cleanups all PCIE, RSS & FW related macros/register defines that are defined in t4fw_api.h and the affected files. Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-11-21 15:22:05 +08:00
FW_HDR_FW_VER_MAJOR_G(c4iw_dev->rdev.lldi.fw_vers),
FW_HDR_FW_VER_MINOR_G(c4iw_dev->rdev.lldi.fw_vers),
FW_HDR_FW_VER_MICRO_G(c4iw_dev->rdev.lldi.fw_vers),
FW_HDR_FW_VER_BUILD_G(c4iw_dev->rdev.lldi.fw_vers));
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
}
static ssize_t show_hca(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
ibdev.dev);
struct ethtool_drvinfo info;
struct net_device *lldev = c4iw_dev->rdev.lldi.ports[0];
PDBG("%s dev 0x%p\n", __func__, dev);
lldev->ethtool_ops->get_drvinfo(lldev, &info);
return sprintf(buf, "%s\n", info.driver);
}
static ssize_t show_board(struct device *dev, struct device_attribute *attr,
char *buf)
{
struct c4iw_dev *c4iw_dev = container_of(dev, struct c4iw_dev,
ibdev.dev);
PDBG("%s dev 0x%p\n", __func__, dev);
return sprintf(buf, "%x.%x\n", c4iw_dev->rdev.lldi.pdev->vendor,
c4iw_dev->rdev.lldi.pdev->device);
}
static int c4iw_get_mib(struct ib_device *ibdev,
union rdma_protocol_stats *stats)
{
RDMA/cxgb4: Export T4 TCP MIB Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-10-18 23:16:40 +08:00
struct tp_tcp_stats v4, v6;
struct c4iw_dev *c4iw_dev = to_c4iw_dev(ibdev);
cxgb4_get_tcp_stats(c4iw_dev->rdev.lldi.pdev, &v4, &v6);
memset(stats, 0, sizeof *stats);
cxgb4: Add ethtool support to get adapter stats Add ethtool support to get adapter specific hardware statistics Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2015-06-03 23:34:39 +08:00
stats->iw.tcpInSegs = v4.tcp_in_segs + v6.tcp_in_segs;
stats->iw.tcpOutSegs = v4.tcp_out_segs + v6.tcp_out_segs;
stats->iw.tcpRetransSegs = v4.tcp_retrans_segs + v6.tcp_retrans_segs;
stats->iw.tcpOutRsts = v4.tcp_out_rsts + v6.tcp_out_rsts;
RDMA/cxgb4: Export T4 TCP MIB Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-10-18 23:16:40 +08:00
return 0;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
}
static DEVICE_ATTR(hw_rev, S_IRUGO, show_rev, NULL);
static DEVICE_ATTR(fw_ver, S_IRUGO, show_fw_ver, NULL);
static DEVICE_ATTR(hca_type, S_IRUGO, show_hca, NULL);
static DEVICE_ATTR(board_id, S_IRUGO, show_board, NULL);
static struct device_attribute *c4iw_class_attributes[] = {
&dev_attr_hw_rev,
&dev_attr_fw_ver,
&dev_attr_hca_type,
&dev_attr_board_id,
};
IB/core: Add per port immutable struct to ib_device As of commit 5eb620c81ce3 "IB/core: Add helpers for uncached GID and P_Key searches"; pkey_tbl_len and gid_tbl_len are immutable data which are stored in the ib_device. The per port core capability flags to be added later are also immutable data to be stored in the ib_device object. In preparation for this create a structure for per port immutable data and place the pkey and gid table lengths within this structure. "get_port_immutable" is added as a mandatory device function to allow the drivers to fill in this data. Signed-off-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-05-14 08:02:58 +08:00
static int c4iw_port_immutable(struct ib_device *ibdev, u8 port_num,
struct ib_port_immutable *immutable)
{
struct ib_port_attr attr;
int err;
err = c4iw_query_port(ibdev, port_num, &attr);
if (err)
return err;
immutable->pkey_tbl_len = attr.pkey_tbl_len;
immutable->gid_tbl_len = attr.gid_tbl_len;
IB/core: Convert core to use bitfield for caps Remove query_protocol callback Use the new Core Capability bits for: rdma_protocol_* rdma_cap_ib_mad rdma_cap_ib_smi rdma_cap_ib_cm rdma_cap_iw_cm rdma_cap_ib_sa rdma_cap_ib_mcast rdma_cap_af_ib rdma_cap_eth_ah Signed-off-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-05-14 08:02:59 +08:00
immutable->core_cap_flags = RDMA_CORE_PORT_IWARP;
IB/core: Add per port immutable struct to ib_device As of commit 5eb620c81ce3 "IB/core: Add helpers for uncached GID and P_Key searches"; pkey_tbl_len and gid_tbl_len are immutable data which are stored in the ib_device. The per port core capability flags to be added later are also immutable data to be stored in the ib_device object. In preparation for this create a structure for per port immutable data and place the pkey and gid table lengths within this structure. "get_port_immutable" is added as a mandatory device function to allow the drivers to fill in this data. Signed-off-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-05-14 08:02:58 +08:00
return 0;
}
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
int c4iw_register_device(struct c4iw_dev *dev)
{
int ret;
int i;
PDBG("%s c4iw_dev %p\n", __func__, dev);
BUG_ON(!dev->rdev.lldi.ports[0]);
strlcpy(dev->ibdev.name, "cxgb4_%d", IB_DEVICE_NAME_MAX);
memset(&dev->ibdev.node_guid, 0, sizeof(dev->ibdev.node_guid));
memcpy(&dev->ibdev.node_guid, dev->rdev.lldi.ports[0]->dev_addr, 6);
dev->ibdev.owner = THIS_MODULE;
dev->device_cap_flags = IB_DEVICE_LOCAL_DMA_LKEY | IB_DEVICE_MEM_WINDOW;
if (fastreg_support)
dev->device_cap_flags |= IB_DEVICE_MEM_MGT_EXTENSIONS;
dev->ibdev.local_dma_lkey = 0;
dev->ibdev.uverbs_cmd_mask =
(1ull << IB_USER_VERBS_CMD_GET_CONTEXT) |
(1ull << IB_USER_VERBS_CMD_QUERY_DEVICE) |
(1ull << IB_USER_VERBS_CMD_QUERY_PORT) |
(1ull << IB_USER_VERBS_CMD_ALLOC_PD) |
(1ull << IB_USER_VERBS_CMD_DEALLOC_PD) |
(1ull << IB_USER_VERBS_CMD_REG_MR) |
(1ull << IB_USER_VERBS_CMD_DEREG_MR) |
(1ull << IB_USER_VERBS_CMD_CREATE_COMP_CHANNEL) |
(1ull << IB_USER_VERBS_CMD_CREATE_CQ) |
(1ull << IB_USER_VERBS_CMD_DESTROY_CQ) |
(1ull << IB_USER_VERBS_CMD_REQ_NOTIFY_CQ) |
(1ull << IB_USER_VERBS_CMD_CREATE_QP) |
(1ull << IB_USER_VERBS_CMD_MODIFY_QP) |
RDMA/cxgb4: Add query_qp support This allows querying the QP state before flushing. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:33 +08:00
(1ull << IB_USER_VERBS_CMD_QUERY_QP) |
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
(1ull << IB_USER_VERBS_CMD_POLL_CQ) |
(1ull << IB_USER_VERBS_CMD_DESTROY_QP) |
(1ull << IB_USER_VERBS_CMD_POST_SEND) |
(1ull << IB_USER_VERBS_CMD_POST_RECV);
dev->ibdev.node_type = RDMA_NODE_RNIC;
memcpy(dev->ibdev.node_desc, C4IW_NODE_DESC, sizeof(C4IW_NODE_DESC));
dev->ibdev.phys_port_cnt = dev->rdev.lldi.nports;
iw_cxgb4: Allocate and use IQs specifically for indirect interrupts Currently indirect interrupts for RDMA CQs funnel through the LLD's RDMA RXQs, which also handle direct interrupts for offload CPLs during RDMA connection setup/teardown. The intended T4 usage model, however, is to have indirect interrupts flow through dedicated IQs. IE not to mix indirect interrupts with CPL messages in an IQ. This patch adds the concept of RDMA concentrator IQs, or CIQs, setup and maintained by the LLD and exported to iw_cxgb4 for use when creating CQs. RDMA CPLs will flow through the LLD's RDMA RXQs, and CQ interrupts flow through the CIQs. Design: cxgb4 creates and exports an array of CIQs for the RDMA ULD. These IQs are sized according to the max available CQs available at adapter init. In addition, these IQs don't need FL buffers since they only service indirect interrupts. One CIQ is setup per RX channel similar to the RDMA RXQs. iw_cxgb4 will utilize these CIQs based on the vector value passed into create_cq(). The num_comp_vectors advertised by iw_cxgb4 will be the number of CIQs configured, and thus the vector value will be the index into the array of CIQs. Based on original work by Steve Wise <swise@opengridcomputing.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-06-07 00:10:42 +08:00
dev->ibdev.num_comp_vectors = dev->rdev.lldi.nciq;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
dev->ibdev.dma_device = &(dev->rdev.lldi.pdev->dev);
dev->ibdev.query_device = c4iw_query_device;
dev->ibdev.query_port = c4iw_query_port;
dev->ibdev.query_pkey = c4iw_query_pkey;
dev->ibdev.query_gid = c4iw_query_gid;
dev->ibdev.alloc_ucontext = c4iw_alloc_ucontext;
dev->ibdev.dealloc_ucontext = c4iw_dealloc_ucontext;
dev->ibdev.mmap = c4iw_mmap;
dev->ibdev.alloc_pd = c4iw_allocate_pd;
dev->ibdev.dealloc_pd = c4iw_deallocate_pd;
dev->ibdev.create_ah = c4iw_ah_create;
dev->ibdev.destroy_ah = c4iw_ah_destroy;
dev->ibdev.create_qp = c4iw_create_qp;
dev->ibdev.modify_qp = c4iw_ib_modify_qp;
RDMA/cxgb4: Add query_qp support This allows querying the QP state before flushing. Signed-off-by: Vipul Pandya <vipul@chelsio.com> Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <roland@purestorage.com>
2012-05-18 17:59:33 +08:00
dev->ibdev.query_qp = c4iw_ib_query_qp;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
dev->ibdev.destroy_qp = c4iw_destroy_qp;
dev->ibdev.create_cq = c4iw_create_cq;
dev->ibdev.destroy_cq = c4iw_destroy_cq;
dev->ibdev.resize_cq = c4iw_resize_cq;
dev->ibdev.poll_cq = c4iw_poll_cq;
dev->ibdev.get_dma_mr = c4iw_get_dma_mr;
dev->ibdev.reg_phys_mr = c4iw_register_phys_mem;
dev->ibdev.rereg_phys_mr = c4iw_reregister_phys_mem;
dev->ibdev.reg_user_mr = c4iw_reg_user_mr;
dev->ibdev.dereg_mr = c4iw_dereg_mr;
dev->ibdev.alloc_mw = c4iw_alloc_mw;
dev->ibdev.bind_mw = c4iw_bind_mw;
dev->ibdev.dealloc_mw = c4iw_dealloc_mw;
dev->ibdev.alloc_fast_reg_mr = c4iw_alloc_fast_reg_mr;
dev->ibdev.alloc_fast_reg_page_list = c4iw_alloc_fastreg_pbl;
dev->ibdev.free_fast_reg_page_list = c4iw_free_fastreg_pbl;
dev->ibdev.attach_mcast = c4iw_multicast_attach;
dev->ibdev.detach_mcast = c4iw_multicast_detach;
dev->ibdev.process_mad = c4iw_process_mad;
dev->ibdev.req_notify_cq = c4iw_arm_cq;
dev->ibdev.post_send = c4iw_post_send;
dev->ibdev.post_recv = c4iw_post_receive;
dev->ibdev.get_protocol_stats = c4iw_get_mib;
RDMA/cxgb4: Support on-chip SQs T4 support on-chip SQs to reduce latency. This patch adds support for this in iw_cxgb4: - Manage ocqp memory like other adapter mem resources. - Allocate user mode SQs from ocqp mem if available. - Map ocqp mem to user process using write combining. - Map PCIE_MA_SYNC reg to user process. Bump uverbs ABI. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-09-14 00:23:57 +08:00
dev->ibdev.uverbs_abi_ver = C4IW_UVERBS_ABI_VERSION;
IB/core: Add per port immutable struct to ib_device As of commit 5eb620c81ce3 "IB/core: Add helpers for uncached GID and P_Key searches"; pkey_tbl_len and gid_tbl_len are immutable data which are stored in the ib_device. The per port core capability flags to be added later are also immutable data to be stored in the ib_device object. In preparation for this create a structure for per port immutable data and place the pkey and gid table lengths within this structure. "get_port_immutable" is added as a mandatory device function to allow the drivers to fill in this data. Signed-off-by: Ira Weiny <ira.weiny@intel.com> Signed-off-by: Doug Ledford <dledford@redhat.com>
2015-05-14 08:02:58 +08:00
dev->ibdev.get_port_immutable = c4iw_port_immutable;
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
dev->ibdev.iwcm = kmalloc(sizeof(struct iw_cm_verbs), GFP_KERNEL);
if (!dev->ibdev.iwcm)
return -ENOMEM;
dev->ibdev.iwcm->connect = c4iw_connect;
dev->ibdev.iwcm->accept = c4iw_accept_cr;
dev->ibdev.iwcm->reject = c4iw_reject_cr;
dev->ibdev.iwcm->create_listen = c4iw_create_listen;
dev->ibdev.iwcm->destroy_listen = c4iw_destroy_listen;
dev->ibdev.iwcm->add_ref = c4iw_qp_add_ref;
dev->ibdev.iwcm->rem_ref = c4iw_qp_rem_ref;
dev->ibdev.iwcm->get_qp = c4iw_get_qp;
IB/core: Allow device-specific per-port sysfs files Add a new parameter to ib_register_device() so that low-level device drivers can pass in a pointer to a callback function that will be called for each port that is registered in sysfs. This allows low-level device drivers to create files in /sys/class/infiniband/<hca>/ports/<N>/ without having to poke through the internals of the RDMA sysfs handling. There is no need for an unregister function since the kobject reference will go to zero when ib_unregister_device() is called. Signed-off-by: Ralph Campbell <ralph.campbell@qlogic.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-05-07 08:03:25 +08:00
ret = ib_register_device(&dev->ibdev, NULL);
RDMA/cxgb4: Add driver for Chelsio T4 RNIC Add an RDMA/iWARP driver for Chelsio T4 Ethernet adapters. Signed-off-by: Steve Wise <swise@opengridcomputing.com> Signed-off-by: Roland Dreier <rolandd@cisco.com>
2010-04-22 06:30:06 +08:00
if (ret)
goto bail1;
for (i = 0; i < ARRAY_SIZE(c4iw_class_attributes); ++i) {
ret = device_create_file(&dev->ibdev.dev,
c4iw_class_attributes[i]);
if (ret)
goto bail2;
}
return 0;
bail2:
ib_unregister_device(&dev->ibdev);
bail1:
kfree(dev->ibdev.iwcm);
return ret;
}
void c4iw_unregister_device(struct c4iw_dev *dev)
{
int i;
PDBG("%s c4iw_dev %p\n", __func__, dev);
for (i = 0; i < ARRAY_SIZE(c4iw_class_attributes); ++i)
device_remove_file(&dev->ibdev.dev,
c4iw_class_attributes[i]);
ib_unregister_device(&dev->ibdev);
kfree(dev->ibdev.iwcm);
return;
}