* [net-next 0/9][pull request] Intel Wired LAN Driver Updates 2014-10-02
From: Jeff Kirsher @ 2014-10-02 10:16 UTC (permalink / raw)
To: davem; +Cc: Jeff Kirsher, netdev, nhorman, sassmann, jogreene
This series contains updates to fm10k, igb, ixgbe and i40e.
Alex provides two updates to the fm10k driver. First reduces the buffer
size to 2k for all page sizes, since most frames only have a 1500 MTU
so supporting a buffer size larger than this is somewhat wasteful.
Second fixes an issue where the number of transmit queues was not being
updated, so added the lines necessary to update the number of transmit
queues.
Rick Jones provides two patches to convert ixgbe, igb and i40e to use
dev_consume_skb_any().
Emil provides two patches for ixgbe, first cleans up a couple of wait
loops on auto-negotiation that were not needed. Second fixes an issue
reported by Fujitsu/Red Hat, which consolidates the logic behind the
dynamically setting of TXDCTL.WTHRESH depending on interrupt throttle
rate (ITR) setting regardless of BQL.
Ethan Zhao provides a cleanup patch for ixgbe where he noticed a
duplicate define.
Bernhard Kaindl provides a patch for igb to remove a source of latency
spikes by not calling code that uses mdelay() for feeding a PHY stat
while being called with a spinlock held.
Todd bumps the igb version based on the recent changes.
The following are changes since commit d068b02cfdfc27f5962ec82ec5568b706f599edc:
net: phy: add BCM7425 and BCM7429 PHYs
and are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master
Alexander Duyck (2):
fm10k: Reduce buffer size when pages are larger than 4K
fm10k: Correctly set the number of Tx queues
Bernhard Kaindl (1):
igb: remove blocking phy read from inside spinlock
Emil Tantilov (2):
ixgbe: remove wait loop on autoneg for copper devices
ixgbe: fix setting of TXDCTL.WTRHESH when ITR is set to 0 and no BQL
Ethan Zhao (1):
ixgbe: delete one duplicate marcro definition of IXGBE_MAX_L2A_QUEUES
Rick Jones (2):
ixgbe: Convert the normal transmit complete path to
dev_consume_skb_any()
i40e/igb: Convert to dev_consume_skb_any()
Todd Fujinaka (1):
igb: bump version to 5.2.15
drivers/net/ethernet/intel/fm10k/fm10k.h | 8 ++---
drivers/net/ethernet/intel/fm10k/fm10k_netdev.c | 6 +++-
drivers/net/ethernet/intel/i40e/i40e_txrx.c | 2 +-
drivers/net/ethernet/intel/igb/e1000_hw.h | 5 ---
drivers/net/ethernet/intel/igb/igb.h | 1 -
drivers/net/ethernet/intel/igb/igb_main.c | 16 ++-------
drivers/net/ethernet/intel/ixgbe/ixgbe.h | 1 -
drivers/net/ethernet/intel/ixgbe/ixgbe_ethtool.c | 3 +-
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 6 +---
drivers/net/ethernet/intel/ixgbe/ixgbe_phy.c | 41 ------------------------
10 files changed, 12 insertions(+), 77 deletions(-)
--
1.9.3
^ permalink raw reply
* Re: [net-next PATCH V5] qdisc: bulk dequeue support for qdiscs with TCQ_F_ONETXQUEUE
From: Toke Høiland-Jørgensen @ 2014-10-02 10:08 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: Dave Taht, Jamal Hadi Salim, Tom Herbert, David Miller,
Linux Netdev List, Eric Dumazet, Hannes Frederic Sowa,
Florian Westphal, Daniel Borkmann, Alexander Duyck,
John Fastabend, Florian Fainelli
In-Reply-To: <20141002094418.37f6ecb4@redhat.com>
[-- Attachment #1: Type: text/plain, Size: 358 bytes --]
Jesper Dangaard Brouer <brouer@redhat.com> writes:
> Perhaps Toke have an opinion on the output formatting from bqlmon?
Anything that outputs one record per line and is parsable by regular
expression is fine with me. I don't currently have a runner that
captures json output, so doing that would actually be more work for me
to support presently :)
-Toke
[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 472 bytes --]
^ permalink raw reply
* RE: [PATCH V1 net-next 2/4] net/mlx5_core: Use hardware registers description header file
From: Eli Cohen @ 2014-10-02 9:21 UTC (permalink / raw)
To: David Miller, alexei.starovoitov@gmail.com
Cc: netdev@vger.kernel.org, Or Gerlitz, Yevgeny Petrilin,
joe@perches.com
In-Reply-To: <20141001.145615.234381013367053243.davem@davemloft.net>
>Indeed, please don't invent your own facilities for compile-time checks,
> we have BUILD_BUG_ON() so please use it instead of non_existent_function()
> which BTW doesn't work with some non-gcc compilers.
Sent V2 series with all comments addressed.
^ permalink raw reply
* [PATCH V2 net-next 1/5] net/mlx5_core: Update device capabilities handling
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412241586-10108-1-git-send-email-eli@mellanox.com>
Rearrange struct mlx5_caps so it has a "gen" field to represent the current
capabilities configured for the device. Max capabilities can also be queried
from the device. Also update capabilities struct to contain more fields as per
the latest revision if firmware specification.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/infiniband/hw/mlx5/cq.c | 8 +-
drivers/infiniband/hw/mlx5/mad.c | 2 +-
drivers/infiniband/hw/mlx5/main.c | 83 ++++++----
drivers/infiniband/hw/mlx5/qp.c | 72 +++++---
drivers/infiniband/hw/mlx5/srq.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 24 ++-
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/fw.c | 81 +--------
drivers/net/ethernet/mellanox/mlx5/core/main.c | 219 +++++++++++++++++++------
drivers/net/ethernet/mellanox/mlx5/core/uar.c | 4 +-
include/linux/mlx5/device.h | 24 ++-
include/linux/mlx5/driver.h | 28 +++-
12 files changed, 331 insertions(+), 222 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index e4056279166d..10cfce5119a9 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -752,7 +752,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
return ERR_PTR(-EINVAL);
entries = roundup_pow_of_two(entries + 1);
- if (entries > dev->mdev->caps.max_cqes)
+ if (entries > dev->mdev->caps.gen.max_cqes)
return ERR_PTR(-EINVAL);
cq = kzalloc(sizeof(*cq), GFP_KERNEL);
@@ -919,7 +919,7 @@ int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
int err;
u32 fsel;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
+ if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
return -ENOSYS;
in = kzalloc(sizeof(*in), GFP_KERNEL);
@@ -1074,7 +1074,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
int uninitialized_var(cqe_size);
unsigned long flags;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
+ if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
pr_info("Firmware does not support resize CQ\n");
return -ENOSYS;
}
@@ -1083,7 +1083,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
return -EINVAL;
entries = roundup_pow_of_two(entries + 1);
- if (entries > dev->mdev->caps.max_cqes + 1)
+ if (entries > dev->mdev->caps.gen.max_cqes + 1)
return -EINVAL;
if (entries == ibcq->cqe + 1)
diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c
index b514bbb5610f..657af9a1167c 100644
--- a/drivers/infiniband/hw/mlx5/mad.c
+++ b/drivers/infiniband/hw/mlx5/mad.c
@@ -129,7 +129,7 @@ int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port)
packet_error = be16_to_cpu(out_mad->status);
- dev->mdev->caps.ext_port_cap[port - 1] = (!err && !packet_error) ?
+ dev->mdev->caps.gen.ext_port_cap[port - 1] = (!err && !packet_error) ?
MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO : 0;
out:
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index d8907b20522a..f3114d1132fb 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -157,11 +157,13 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
struct mlx5_ib_dev *dev = to_mdev(ibdev);
struct ib_smp *in_mad = NULL;
struct ib_smp *out_mad = NULL;
+ struct mlx5_general_caps *gen;
int err = -ENOMEM;
int max_rq_sg;
int max_sq_sg;
u64 flags;
+ gen = &dev->mdev->caps.gen;
in_mad = kzalloc(sizeof(*in_mad), GFP_KERNEL);
out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
if (!in_mad || !out_mad)
@@ -183,7 +185,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
IB_DEVICE_PORT_ACTIVE_EVENT |
IB_DEVICE_SYS_IMAGE_GUID |
IB_DEVICE_RC_RNR_NAK_GEN;
- flags = dev->mdev->caps.flags;
+ flags = gen->flags;
if (flags & MLX5_DEV_CAP_FLAG_BAD_PKEY_CNTR)
props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
if (flags & MLX5_DEV_CAP_FLAG_BAD_QKEY_CNTR)
@@ -213,30 +215,31 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
memcpy(&props->sys_image_guid, out_mad->data + 4, 8);
props->max_mr_size = ~0ull;
- props->page_size_cap = dev->mdev->caps.min_page_sz;
- props->max_qp = 1 << dev->mdev->caps.log_max_qp;
- props->max_qp_wr = dev->mdev->caps.max_wqes;
- max_rq_sg = dev->mdev->caps.max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
- max_sq_sg = (dev->mdev->caps.max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
+ props->page_size_cap = gen->min_page_sz;
+ props->max_qp = 1 << gen->log_max_qp;
+ props->max_qp_wr = gen->max_wqes;
+ max_rq_sg = gen->max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
+ max_sq_sg = (gen->max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
sizeof(struct mlx5_wqe_data_seg);
props->max_sge = min(max_rq_sg, max_sq_sg);
- props->max_cq = 1 << dev->mdev->caps.log_max_cq;
- props->max_cqe = dev->mdev->caps.max_cqes - 1;
- props->max_mr = 1 << dev->mdev->caps.log_max_mkey;
- props->max_pd = 1 << dev->mdev->caps.log_max_pd;
- props->max_qp_rd_atom = dev->mdev->caps.max_ra_req_qp;
- props->max_qp_init_rd_atom = dev->mdev->caps.max_ra_res_qp;
+ props->max_cq = 1 << gen->log_max_cq;
+ props->max_cqe = gen->max_cqes - 1;
+ props->max_mr = 1 << gen->log_max_mkey;
+ props->max_pd = 1 << gen->log_max_pd;
+ props->max_qp_rd_atom = 1 << gen->log_max_ra_req_qp;
+ props->max_qp_init_rd_atom = 1 << gen->log_max_ra_res_qp;
+ props->max_srq = 1 << gen->log_max_srq;
+ props->max_srq_wr = gen->max_srq_wqes - 1;
+ props->local_ca_ack_delay = gen->local_ca_ack_delay;
props->max_res_rd_atom = props->max_qp_rd_atom * props->max_qp;
- props->max_srq = 1 << dev->mdev->caps.log_max_srq;
- props->max_srq_wr = dev->mdev->caps.max_srq_wqes - 1;
props->max_srq_sge = max_rq_sg - 1;
props->max_fast_reg_page_list_len = (unsigned int)-1;
- props->local_ca_ack_delay = dev->mdev->caps.local_ca_ack_delay;
+ props->local_ca_ack_delay = gen->local_ca_ack_delay;
props->atomic_cap = IB_ATOMIC_NONE;
props->masked_atomic_cap = IB_ATOMIC_NONE;
props->max_pkeys = be16_to_cpup((__be16 *)(out_mad->data + 28));
- props->max_mcast_grp = 1 << dev->mdev->caps.log_max_mcg;
- props->max_mcast_qp_attach = dev->mdev->caps.max_qp_mcg;
+ props->max_mcast_grp = 1 << gen->log_max_mcg;
+ props->max_mcast_qp_attach = gen->max_qp_mcg;
props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
props->max_mcast_grp;
props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */
@@ -254,10 +257,12 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
struct mlx5_ib_dev *dev = to_mdev(ibdev);
struct ib_smp *in_mad = NULL;
struct ib_smp *out_mad = NULL;
+ struct mlx5_general_caps *gen;
int ext_active_speed;
int err = -ENOMEM;
- if (port < 1 || port > dev->mdev->caps.num_ports) {
+ gen = &dev->mdev->caps.gen;
+ if (port < 1 || port > gen->num_ports) {
mlx5_ib_warn(dev, "invalid port number %d\n", port);
return -EINVAL;
}
@@ -288,8 +293,8 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
props->phys_state = out_mad->data[33] >> 4;
props->port_cap_flags = be32_to_cpup((__be32 *)(out_mad->data + 20));
props->gid_tbl_len = out_mad->data[50];
- props->max_msg_sz = 1 << to_mdev(ibdev)->mdev->caps.log_max_msg;
- props->pkey_tbl_len = to_mdev(ibdev)->mdev->caps.port[port - 1].pkey_table_len;
+ props->max_msg_sz = 1 << gen->log_max_msg;
+ props->pkey_tbl_len = gen->port[port - 1].pkey_table_len;
props->bad_pkey_cntr = be16_to_cpup((__be16 *)(out_mad->data + 46));
props->qkey_viol_cntr = be16_to_cpup((__be16 *)(out_mad->data + 48));
props->active_width = out_mad->data[31] & 0xf;
@@ -316,7 +321,7 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
/* If reported active speed is QDR, check if is FDR-10 */
if (props->active_speed == 4) {
- if (dev->mdev->caps.ext_port_cap[port - 1] &
+ if (gen->ext_port_cap[port - 1] &
MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO) {
init_query_mad(in_mad);
in_mad->attr_id = MLX5_ATTR_EXTENDED_PORT_INFO;
@@ -470,6 +475,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
struct mlx5_ib_alloc_ucontext_req_v2 req;
struct mlx5_ib_alloc_ucontext_resp resp;
struct mlx5_ib_ucontext *context;
+ struct mlx5_general_caps *gen;
struct mlx5_uuar_info *uuari;
struct mlx5_uar *uars;
int gross_uuars;
@@ -480,6 +486,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
int i;
size_t reqlen;
+ gen = &dev->mdev->caps.gen;
if (!dev->ib_active)
return ERR_PTR(-EAGAIN);
@@ -512,14 +519,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE;
gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE;
- resp.qp_tab_size = 1 << dev->mdev->caps.log_max_qp;
- resp.bf_reg_size = dev->mdev->caps.bf_reg_size;
+ resp.qp_tab_size = 1 << gen->log_max_qp;
+ resp.bf_reg_size = gen->bf_reg_size;
resp.cache_line_size = L1_CACHE_BYTES;
- resp.max_sq_desc_sz = dev->mdev->caps.max_sq_desc_sz;
- resp.max_rq_desc_sz = dev->mdev->caps.max_rq_desc_sz;
- resp.max_send_wqebb = dev->mdev->caps.max_wqes;
- resp.max_recv_wr = dev->mdev->caps.max_wqes;
- resp.max_srq_recv_wr = dev->mdev->caps.max_srq_wqes;
+ resp.max_sq_desc_sz = gen->max_sq_desc_sz;
+ resp.max_rq_desc_sz = gen->max_rq_desc_sz;
+ resp.max_send_wqebb = gen->max_wqes;
+ resp.max_recv_wr = gen->max_wqes;
+ resp.max_srq_recv_wr = gen->max_srq_wqes;
context = kzalloc(sizeof(*context), GFP_KERNEL);
if (!context)
@@ -565,7 +572,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
mutex_init(&context->db_page_mutex);
resp.tot_uuars = req.total_num_uuars;
- resp.num_ports = dev->mdev->caps.num_ports;
+ resp.num_ports = gen->num_ports;
err = ib_copy_to_udata(udata, &resp,
sizeof(resp) - sizeof(resp.reserved));
if (err)
@@ -967,9 +974,11 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
static void get_ext_port_caps(struct mlx5_ib_dev *dev)
{
+ struct mlx5_general_caps *gen;
int port;
- for (port = 1; port <= dev->mdev->caps.num_ports; port++)
+ gen = &dev->mdev->caps.gen;
+ for (port = 1; port <= gen->num_ports; port++)
mlx5_query_ext_port_caps(dev, port);
}
@@ -977,9 +986,11 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
{
struct ib_device_attr *dprops = NULL;
struct ib_port_attr *pprops = NULL;
+ struct mlx5_general_caps *gen;
int err = 0;
int port;
+ gen = &dev->mdev->caps.gen;
pprops = kmalloc(sizeof(*pprops), GFP_KERNEL);
if (!pprops)
goto out;
@@ -994,14 +1005,14 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
goto out;
}
- for (port = 1; port <= dev->mdev->caps.num_ports; port++) {
+ for (port = 1; port <= gen->num_ports; port++) {
err = mlx5_ib_query_port(&dev->ib_dev, port, pprops);
if (err) {
mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err);
break;
}
- dev->mdev->caps.port[port - 1].pkey_table_len = dprops->max_pkeys;
- dev->mdev->caps.port[port - 1].gid_table_len = pprops->gid_tbl_len;
+ gen->port[port - 1].pkey_table_len = dprops->max_pkeys;
+ gen->port[port - 1].gid_table_len = pprops->gid_tbl_len;
mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n",
dprops->max_pkeys, pprops->gid_tbl_len);
}
@@ -1279,8 +1290,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
strlcpy(dev->ib_dev.name, "mlx5_%d", IB_DEVICE_NAME_MAX);
dev->ib_dev.owner = THIS_MODULE;
dev->ib_dev.node_type = RDMA_NODE_IB_CA;
- dev->ib_dev.local_dma_lkey = mdev->caps.reserved_lkey;
- dev->num_ports = mdev->caps.num_ports;
+ dev->ib_dev.local_dma_lkey = mdev->caps.gen.reserved_lkey;
+ dev->num_ports = mdev->caps.gen.num_ports;
dev->ib_dev.phys_port_cnt = dev->num_ports;
dev->ib_dev.num_comp_vectors = dev->num_comp_vectors;
dev->ib_dev.dma_device = &mdev->pdev->dev;
@@ -1355,7 +1366,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
dev->ib_dev.free_fast_reg_page_list = mlx5_ib_free_fast_reg_page_list;
dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status;
- if (mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC) {
+ if (mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_XRC) {
dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd;
dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd;
dev->ib_dev.uverbs_cmd_mask |=
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 8c574b63d77b..dbfe498870c1 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -158,11 +158,13 @@ static void mlx5_ib_qp_event(struct mlx5_core_qp *qp, int type)
static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
int has_rq, struct mlx5_ib_qp *qp, struct mlx5_ib_create_qp *ucmd)
{
+ struct mlx5_general_caps *gen;
int wqe_size;
int wq_size;
+ gen = &dev->mdev->caps.gen;
/* Sanity check RQ size before proceeding */
- if (cap->max_recv_wr > dev->mdev->caps.max_wqes)
+ if (cap->max_recv_wr > gen->max_wqes)
return -EINVAL;
if (!has_rq) {
@@ -182,10 +184,10 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
wq_size = roundup_pow_of_two(cap->max_recv_wr) * wqe_size;
wq_size = max_t(int, wq_size, MLX5_SEND_WQE_BB);
qp->rq.wqe_cnt = wq_size / wqe_size;
- if (wqe_size > dev->mdev->caps.max_rq_desc_sz) {
+ if (wqe_size > gen->max_rq_desc_sz) {
mlx5_ib_dbg(dev, "wqe_size %d, max %d\n",
wqe_size,
- dev->mdev->caps.max_rq_desc_sz);
+ gen->max_rq_desc_sz);
return -EINVAL;
}
qp->rq.wqe_shift = ilog2(wqe_size);
@@ -266,9 +268,11 @@ static int calc_send_wqe(struct ib_qp_init_attr *attr)
static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
struct mlx5_ib_qp *qp)
{
+ struct mlx5_general_caps *gen;
int wqe_size;
int wq_size;
+ gen = &dev->mdev->caps.gen;
if (!attr->cap.max_send_wr)
return 0;
@@ -277,9 +281,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
if (wqe_size < 0)
return wqe_size;
- if (wqe_size > dev->mdev->caps.max_sq_desc_sz) {
+ if (wqe_size > gen->max_sq_desc_sz) {
mlx5_ib_dbg(dev, "wqe_size(%d) > max_sq_desc_sz(%d)\n",
- wqe_size, dev->mdev->caps.max_sq_desc_sz);
+ wqe_size, gen->max_sq_desc_sz);
return -EINVAL;
}
@@ -292,9 +296,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
wq_size = roundup_pow_of_two(attr->cap.max_send_wr * wqe_size);
qp->sq.wqe_cnt = wq_size / MLX5_SEND_WQE_BB;
- if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+ if (qp->sq.wqe_cnt > gen->max_wqes) {
mlx5_ib_dbg(dev, "wqe count(%d) exceeds limits(%d)\n",
- qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+ qp->sq.wqe_cnt, gen->max_wqes);
return -ENOMEM;
}
qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB);
@@ -309,11 +313,13 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
struct mlx5_ib_qp *qp,
struct mlx5_ib_create_qp *ucmd)
{
+ struct mlx5_general_caps *gen;
int desc_sz = 1 << qp->sq.wqe_shift;
- if (desc_sz > dev->mdev->caps.max_sq_desc_sz) {
+ gen = &dev->mdev->caps.gen;
+ if (desc_sz > gen->max_sq_desc_sz) {
mlx5_ib_warn(dev, "desc_sz %d, max_sq_desc_sz %d\n",
- desc_sz, dev->mdev->caps.max_sq_desc_sz);
+ desc_sz, gen->max_sq_desc_sz);
return -EINVAL;
}
@@ -325,9 +331,9 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
qp->sq.wqe_cnt = ucmd->sq_wqe_count;
- if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+ if (qp->sq.wqe_cnt > gen->max_wqes) {
mlx5_ib_warn(dev, "wqe_cnt %d, max_wqes %d\n",
- qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+ qp->sq.wqe_cnt, gen->max_wqes);
return -EINVAL;
}
@@ -803,16 +809,18 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
struct mlx5_ib_resources *devr = &dev->devr;
struct mlx5_ib_create_qp_resp resp;
struct mlx5_create_qp_mbox_in *in;
+ struct mlx5_general_caps *gen;
struct mlx5_ib_create_qp ucmd;
int inlen = sizeof(*in);
int err;
+ gen = &dev->mdev->caps.gen;
mutex_init(&qp->mutex);
spin_lock_init(&qp->sq.lock);
spin_lock_init(&qp->rq.lock);
if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) {
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
mlx5_ib_dbg(dev, "block multicast loopback isn't supported\n");
return -EINVAL;
} else {
@@ -851,9 +859,9 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
mlx5_ib_dbg(dev, "invalid rq params\n");
return -EINVAL;
}
- if (ucmd.sq_wqe_count > dev->mdev->caps.max_wqes) {
+ if (ucmd.sq_wqe_count > gen->max_wqes) {
mlx5_ib_dbg(dev, "requested sq_wqe_count (%d) > max allowed (%d)\n",
- ucmd.sq_wqe_count, dev->mdev->caps.max_wqes);
+ ucmd.sq_wqe_count, gen->max_wqes);
return -EINVAL;
}
err = create_user_qp(dev, pd, qp, udata, &in, &resp, &inlen);
@@ -1144,6 +1152,7 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
struct ib_qp_init_attr *init_attr,
struct ib_udata *udata)
{
+ struct mlx5_general_caps *gen;
struct mlx5_ib_dev *dev;
struct mlx5_ib_qp *qp;
u16 xrcdn = 0;
@@ -1161,11 +1170,12 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
}
dev = to_mdev(to_mxrcd(init_attr->xrcd)->ibxrcd.device);
}
+ gen = &dev->mdev->caps.gen;
switch (init_attr->qp_type) {
case IB_QPT_XRC_TGT:
case IB_QPT_XRC_INI:
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC)) {
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC)) {
mlx5_ib_dbg(dev, "XRC not supported\n");
return ERR_PTR(-ENOSYS);
}
@@ -1272,6 +1282,9 @@ enum {
static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
{
+ struct mlx5_general_caps *gen;
+
+ gen = &dev->mdev->caps.gen;
if (rate == IB_RATE_PORT_CURRENT) {
return 0;
} else if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS) {
@@ -1279,7 +1292,7 @@ static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
} else {
while (rate != IB_RATE_2_5_GBPS &&
!(1 << (rate + MLX5_STAT_RATE_OFFSET) &
- dev->mdev->caps.stat_rate_support))
+ gen->stat_rate_support))
--rate;
}
@@ -1290,8 +1303,10 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
struct mlx5_qp_path *path, u8 port, int attr_mask,
u32 path_flags, const struct ib_qp_attr *attr)
{
+ struct mlx5_general_caps *gen;
int err;
+ gen = &dev->mdev->caps.gen;
path->fl = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0;
path->free_ar = (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x80 : 0;
@@ -1318,9 +1333,9 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
path->port = port;
if (ah->ah_flags & IB_AH_GRH) {
- if (ah->grh.sgid_index >= dev->mdev->caps.port[port - 1].gid_table_len) {
+ if (ah->grh.sgid_index >= gen->port[port - 1].gid_table_len) {
pr_err(KERN_ERR "sgid_index (%u) too large. max is %d\n",
- ah->grh.sgid_index, dev->mdev->caps.port[port - 1].gid_table_len);
+ ah->grh.sgid_index, gen->port[port - 1].gid_table_len);
return -EINVAL;
}
@@ -1492,6 +1507,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
struct mlx5_ib_qp *qp = to_mqp(ibqp);
struct mlx5_ib_cq *send_cq, *recv_cq;
struct mlx5_qp_context *context;
+ struct mlx5_general_caps *gen;
struct mlx5_modify_qp_mbox_in *in;
struct mlx5_ib_pd *pd;
enum mlx5_qp_state mlx5_cur, mlx5_new;
@@ -1500,6 +1516,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
int mlx5_st;
int err;
+ gen = &dev->mdev->caps.gen;
in = kzalloc(sizeof(*in), GFP_KERNEL);
if (!in)
return -ENOMEM;
@@ -1539,7 +1556,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
err = -EINVAL;
goto out;
}
- context->mtu_msgmax = (attr->path_mtu << 5) | dev->mdev->caps.log_max_msg;
+ context->mtu_msgmax = (attr->path_mtu << 5) | gen->log_max_msg;
}
if (attr_mask & IB_QP_DEST_QPN)
@@ -1685,9 +1702,11 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
struct mlx5_ib_dev *dev = to_mdev(ibqp->device);
struct mlx5_ib_qp *qp = to_mqp(ibqp);
enum ib_qp_state cur_state, new_state;
+ struct mlx5_general_caps *gen;
int err = -EINVAL;
int port;
+ gen = &dev->mdev->caps.gen;
mutex_lock(&qp->mutex);
cur_state = attr_mask & IB_QP_CUR_STATE ? attr->cur_qp_state : qp->state;
@@ -1699,21 +1718,21 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
goto out;
if ((attr_mask & IB_QP_PORT) &&
- (attr->port_num == 0 || attr->port_num > dev->mdev->caps.num_ports))
+ (attr->port_num == 0 || attr->port_num > gen->num_ports))
goto out;
if (attr_mask & IB_QP_PKEY_INDEX) {
port = attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
- if (attr->pkey_index >= dev->mdev->caps.port[port - 1].pkey_table_len)
+ if (attr->pkey_index >= gen->port[port - 1].pkey_table_len)
goto out;
}
if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
- attr->max_rd_atomic > dev->mdev->caps.max_ra_res_qp)
+ attr->max_rd_atomic > (1 << gen->log_max_ra_res_qp))
goto out;
if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
- attr->max_dest_rd_atomic > dev->mdev->caps.max_ra_req_qp)
+ attr->max_dest_rd_atomic > (1 << gen->log_max_ra_req_qp))
goto out;
if (cur_state == new_state && cur_state == IB_QPS_RESET) {
@@ -2893,7 +2912,8 @@ static void to_ib_ah_attr(struct mlx5_ib_dev *ibdev, struct ib_ah_attr *ib_ah_at
memset(ib_ah_attr, 0, sizeof(*ib_ah_attr));
ib_ah_attr->port_num = path->port;
- if (ib_ah_attr->port_num == 0 || ib_ah_attr->port_num > dev->caps.num_ports)
+ if (ib_ah_attr->port_num == 0 ||
+ ib_ah_attr->port_num > dev->caps.gen.num_ports)
return;
ib_ah_attr->sl = path->sl & 0xf;
@@ -3011,10 +3031,12 @@ struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev,
struct ib_udata *udata)
{
struct mlx5_ib_dev *dev = to_mdev(ibdev);
+ struct mlx5_general_caps *gen;
struct mlx5_ib_xrcd *xrcd;
int err;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC))
+ gen = &dev->mdev->caps.gen;
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC))
return ERR_PTR(-ENOSYS);
xrcd = kmalloc(sizeof(*xrcd), GFP_KERNEL);
diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 70bd131ba646..97cc1baaa8e3 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -238,6 +238,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
struct ib_udata *udata)
{
struct mlx5_ib_dev *dev = to_mdev(pd->device);
+ struct mlx5_general_caps *gen;
struct mlx5_ib_srq *srq;
int desc_size;
int buf_size;
@@ -247,11 +248,12 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
int is_xrc;
u32 flgs, xrcdn;
+ gen = &dev->mdev->caps.gen;
/* Sanity check SRQ size before proceeding */
- if (init_attr->attr.max_wr >= dev->mdev->caps.max_srq_wqes) {
+ if (init_attr->attr.max_wr >= gen->max_srq_wqes) {
mlx5_ib_dbg(dev, "max_wr %d, cap %d\n",
init_attr->attr.max_wr,
- dev->mdev->caps.max_srq_wqes);
+ gen->max_srq_wqes);
return ERR_PTR(-EINVAL);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 65a7da69e2ac..6eb0f85cf872 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1538,16 +1538,9 @@ static const char *cmd_status_str(u8 status)
}
}
-int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
+static int cmd_status_to_err(u8 status)
{
- if (!hdr->status)
- return 0;
-
- pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
- cmd_status_str(hdr->status), hdr->status,
- be32_to_cpu(hdr->syndrome));
-
- switch (hdr->status) {
+ switch (status) {
case MLX5_CMD_STAT_OK: return 0;
case MLX5_CMD_STAT_INT_ERR: return -EIO;
case MLX5_CMD_STAT_BAD_OP_ERR: return -EINVAL;
@@ -1567,3 +1560,16 @@ int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
default: return -EIO;
}
}
+
+/* this will be available till all the commands use set/get macros */
+int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
+{
+ if (!hdr->status)
+ return 0;
+
+ pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
+ cmd_status_str(hdr->status), hdr->status,
+ be32_to_cpu(hdr->syndrome));
+
+ return cmd_status_to_err(hdr->status);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 4e8bd0b34bb0..11b9b840ad4d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -468,7 +468,7 @@ int mlx5_start_eqs(struct mlx5_core_dev *dev)
err = mlx5_create_map_eq(dev, &table->pages_eq,
MLX5_EQ_VEC_PAGES,
- dev->caps.max_vf + 1,
+ dev->caps.gen.max_vf + 1,
1 << MLX5_EVENT_TYPE_PAGE_REQUEST, "mlx5_pages_eq",
&dev->priv.uuari.uars[0]);
if (err) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
index f012658b6a92..087c4c797deb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
@@ -64,86 +64,9 @@ out_out:
return err;
}
-int mlx5_cmd_query_hca_cap(struct mlx5_core_dev *dev,
- struct mlx5_caps *caps)
+int mlx5_cmd_query_hca_cap(struct mlx5_core_dev *dev, struct mlx5_caps *caps)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *out;
- struct mlx5_cmd_query_hca_cap_mbox_in in;
- struct mlx5_query_special_ctxs_mbox_out ctx_out;
- struct mlx5_query_special_ctxs_mbox_in ctx_in;
- int err;
- u16 t16;
-
- out = kzalloc(sizeof(*out), GFP_KERNEL);
- if (!out)
- return -ENOMEM;
-
- memset(&in, 0, sizeof(in));
- in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- in.hdr.opmod = cpu_to_be16(0x1);
- err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
- if (err)
- goto out_out;
-
- if (out->hdr.status) {
- err = mlx5_cmd_status_to_err(&out->hdr);
- goto out_out;
- }
-
-
- caps->log_max_eq = out->hca_cap.log_max_eq & 0xf;
- caps->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
- caps->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
- caps->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
- caps->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
- caps->flags = be64_to_cpu(out->hca_cap.flags);
- caps->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
- caps->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
- caps->num_ports = out->hca_cap.num_ports & 0xf;
- caps->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
- if (caps->num_ports > MLX5_MAX_PORTS) {
- mlx5_core_err(dev, "device has %d ports while the driver supports max %d ports\n",
- caps->num_ports, MLX5_MAX_PORTS);
- err = -EINVAL;
- goto out_out;
- }
- caps->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
- caps->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
- caps->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
- caps->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
- caps->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
- caps->log_max_mcg = out->hca_cap.log_max_mcg;
- caps->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
- caps->max_ra_res_qp = 1 << (out->hca_cap.log_max_ra_res_qp & 0x3f);
- caps->max_ra_req_qp = 1 << (out->hca_cap.log_max_ra_req_qp & 0x3f);
- caps->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
- t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
- if (t16 & 0x8000) {
- caps->bf_reg_size = 1 << (t16 & 0x1f);
- caps->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
- } else {
- caps->bf_reg_size = 0;
- caps->bf_regs_per_page = 0;
- }
- caps->min_page_sz = ~(u32)((1 << out->hca_cap.log_pg_sz) - 1);
-
- memset(&ctx_in, 0, sizeof(ctx_in));
- memset(&ctx_out, 0, sizeof(ctx_out));
- ctx_in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS);
- err = mlx5_cmd_exec(dev, &ctx_in, sizeof(ctx_in),
- &ctx_out, sizeof(ctx_out));
- if (err)
- goto out_out;
-
- if (ctx_out.hdr.status)
- err = mlx5_cmd_status_to_err(&ctx_out.hdr);
-
- caps->reserved_lkey = be32_to_cpu(ctx_out.reserved_lkey);
-
-out_out:
- kfree(out);
-
- return err;
+ return mlx5_core_get_caps(dev, caps, HCA_CAP_OPMOD_GET_CUR);
}
int mlx5_cmd_init_hca(struct mlx5_core_dev *dev)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index f2716cc1f51d..d9f74618befa 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -207,11 +207,11 @@ static void release_bar(struct pci_dev *pdev)
static int mlx5_enable_msix(struct mlx5_core_dev *dev)
{
struct mlx5_eq_table *table = &dev->priv.eq_table;
- int num_eqs = 1 << dev->caps.log_max_eq;
+ int num_eqs = 1 << dev->caps.gen.log_max_eq;
int nvec;
int i;
- nvec = dev->caps.num_ports * num_online_cpus() + MLX5_EQ_VEC_COMP_BASE;
+ nvec = dev->caps.gen.num_ports * num_online_cpus() + MLX5_EQ_VEC_COMP_BASE;
nvec = min_t(int, nvec, num_eqs);
if (nvec <= MLX5_EQ_VEC_COMP_BASE)
return -ENOMEM;
@@ -250,13 +250,34 @@ struct mlx5_reg_host_endianess {
#define CAP_MASK(pos, size) ((u64)((1 << (size)) - 1) << (pos))
enum {
- MLX5_CAP_BITS_RW_MASK = CAP_MASK(MLX5_CAP_OFF_CMDIF_CSUM, 2) |
- CAP_MASK(MLX5_CAP_OFF_DCT, 1),
+ MLX5_CAP_BITS_RW_MASK = CAP_MASK(MLX5_CAP_OFF_CMDIF_CSUM, 2) |
+ MLX5_DEV_CAP_FLAG_DCT,
};
+static u16 to_fw_pkey_sz(u32 size)
+{
+ switch (size) {
+ case 128:
+ return 0;
+ case 256:
+ return 1;
+ case 512:
+ return 2;
+ case 1024:
+ return 3;
+ case 2048:
+ return 4;
+ case 4096:
+ return 5;
+ default:
+ pr_warn("invalid pkey table size %d\n", size);
+ return 0;
+ }
+}
+
/* selectively copy writable fields clearing any reserved area
*/
-static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_hca_cap *from)
+static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_general_caps *from)
{
u64 v64;
@@ -265,76 +286,172 @@ static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_hca_cap *from)
to->log_max_ra_res_dc = from->log_max_ra_res_dc & 0x3f;
to->log_max_ra_req_qp = from->log_max_ra_req_qp & 0x3f;
to->log_max_ra_res_qp = from->log_max_ra_res_qp & 0x3f;
- to->log_max_atomic_size_qp = from->log_max_atomic_size_qp;
- to->log_max_atomic_size_dc = from->log_max_atomic_size_dc;
- v64 = be64_to_cpu(from->flags) & MLX5_CAP_BITS_RW_MASK;
+ to->pkey_table_size = cpu_to_be16(to_fw_pkey_sz(from->pkey_table_size));
+ v64 = from->flags & MLX5_CAP_BITS_RW_MASK;
to->flags = cpu_to_be64(v64);
}
-enum {
- HCA_CAP_OPMOD_GET_MAX = 0,
- HCA_CAP_OPMOD_GET_CUR = 1,
-};
+static u16 get_pkey_table_size(int pkey)
+{
+ if (pkey > MLX5_MAX_LOG_PKEY_TABLE)
+ return 0;
-static int handle_hca_cap(struct mlx5_core_dev *dev)
+ return MLX5_MIN_PKEY_TABLE_SIZE << pkey;
+}
+
+static void fw2drv_caps(struct mlx5_caps *caps,
+ struct mlx5_cmd_query_hca_cap_mbox_out *out)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *query_out = NULL;
- struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
- struct mlx5_cmd_query_hca_cap_mbox_in query_ctx;
- struct mlx5_cmd_set_hca_cap_mbox_out set_out;
- u64 flags;
+ struct mlx5_general_caps *gen = &caps->gen;
+ u16 t16;
+
+ gen->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
+ gen->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
+ gen->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
+ gen->log_max_strq = out->hca_cap.log_max_strq_sz;
+ gen->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
+ gen->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
+ gen->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
+ gen->max_eqes = out->hca_cap.log_max_eq_sz;
+ gen->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
+ gen->log_max_eq = out->hca_cap.log_max_eq & 0xf;
+ gen->max_indirection = out->hca_cap.max_indirection;
+ gen->log_max_mrw_sz = out->hca_cap.log_max_mrw_sz;
+ gen->log_max_bsf_list_size = 0;
+ gen->log_max_klm_list_size = 0;
+ gen->log_max_ra_req_dc = out->hca_cap.log_max_ra_req_dc;
+ gen->log_max_ra_res_dc = out->hca_cap.log_max_ra_res_dc;
+ gen->log_max_ra_req_qp = out->hca_cap.log_max_ra_req_qp;
+ gen->log_max_ra_res_qp = out->hca_cap.log_max_ra_res_qp;
+ gen->max_qp_counters = be16_to_cpu(out->hca_cap.max_qp_count);
+ gen->pkey_table_size = get_pkey_table_size(be16_to_cpu(out->hca_cap.pkey_table_size));
+ gen->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
+ gen->num_ports = out->hca_cap.num_ports & 0xf;
+ gen->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
+ gen->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
+ gen->flags = be64_to_cpu(out->hca_cap.flags);
+ pr_debug("flags = 0x%llx\n", gen->flags);
+ gen->uar_sz = out->hca_cap.uar_sz;
+ gen->min_log_pg_sz = out->hca_cap.log_pg_sz;
+
+ t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
+ if (t16 & 0x8000) {
+ gen->bf_reg_size = 1 << (t16 & 0x1f);
+ gen->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
+ } else {
+ gen->bf_reg_size = 0;
+ gen->bf_regs_per_page = 0;
+ }
+ gen->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
+ gen->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
+ gen->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
+ gen->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
+ gen->log_max_xrcd = out->hca_cap.log_max_xrcd;
+ gen->log_uar_page_sz = be16_to_cpu(out->hca_cap.log_uar_page_sz);
+}
+
+static const char *caps_opmod_str(u16 opmod)
+{
+ switch (opmod) {
+ case HCA_CAP_OPMOD_GET_MAX:
+ return "GET_MAX";
+ case HCA_CAP_OPMOD_GET_CUR:
+ return "GET_CUR";
+ default:
+ return "Invalid";
+ }
+}
+
+int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
+ u16 opmod)
+{
+ struct mlx5_cmd_query_hca_cap_mbox_out *out;
+ struct mlx5_cmd_query_hca_cap_mbox_in in;
int err;
- memset(&query_ctx, 0, sizeof(query_ctx));
- query_out = kzalloc(sizeof(*query_out), GFP_KERNEL);
- if (!query_out)
+ memset(&in, 0, sizeof(in));
+ out = kzalloc(sizeof(*out), GFP_KERNEL);
+ if (!out)
return -ENOMEM;
- set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
- if (!set_ctx) {
- err = -ENOMEM;
+ in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
+ in.hdr.opmod = cpu_to_be16(opmod);
+ err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
+
+ err = mlx5_cmd_status_to_err(&out->hdr);
+ if (err) {
+ mlx5_core_warn(dev, "query max hca cap failed, %d\n", err);
goto query_ex;
}
+ mlx5_core_dbg(dev, "%s\n", caps_opmod_str(opmod));
+ fw2drv_caps(caps, out);
- query_ctx.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- query_ctx.hdr.opmod = cpu_to_be16(HCA_CAP_OPMOD_GET_CUR);
- err = mlx5_cmd_exec(dev, &query_ctx, sizeof(query_ctx),
- query_out, sizeof(*query_out));
+query_ex:
+ kfree(out);
+ return err;
+}
+
+static int set_caps(struct mlx5_core_dev *dev,
+ struct mlx5_cmd_set_hca_cap_mbox_in *in)
+{
+ struct mlx5_cmd_set_hca_cap_mbox_out out;
+ int err;
+
+ memset(&out, 0, sizeof(out));
+
+ in->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
+ err = mlx5_cmd_exec(dev, in, sizeof(*in), &out, sizeof(out));
if (err)
- goto query_ex;
+ return err;
- err = mlx5_cmd_status_to_err(&query_out->hdr);
- if (err) {
- mlx5_core_warn(dev, "query hca cap failed, %d\n", err);
+ err = mlx5_cmd_status_to_err(&out.hdr);
+
+ return err;
+}
+
+static int handle_hca_cap(struct mlx5_core_dev *dev)
+{
+ struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
+ struct mlx5_profile *prof = dev->profile;
+ struct mlx5_caps *cur_caps = NULL;
+ struct mlx5_caps *max_caps = NULL;
+ int err = -ENOMEM;
+
+ set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
+ if (!set_ctx)
goto query_ex;
- }
- copy_rw_fields(&set_ctx->hca_cap, &query_out->hca_cap);
+ max_caps = kzalloc(sizeof(*max_caps), GFP_KERNEL);
+ if (!max_caps)
+ goto query_ex;
- if (dev->profile && dev->profile->mask & MLX5_PROF_MASK_QP_SIZE)
- set_ctx->hca_cap.log_max_qp = dev->profile->log_max_qp;
+ cur_caps = kzalloc(sizeof(*cur_caps), GFP_KERNEL);
+ if (!cur_caps)
+ goto query_ex;
- flags = be64_to_cpu(query_out->hca_cap.flags);
- /* disable checksum */
- flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
-
- set_ctx->hca_cap.flags = cpu_to_be64(flags);
- memset(&set_out, 0, sizeof(set_out));
- set_ctx->hca_cap.log_uar_page_sz = cpu_to_be16(PAGE_SHIFT - 12);
- set_ctx->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
- err = mlx5_cmd_exec(dev, set_ctx, sizeof(*set_ctx),
- &set_out, sizeof(set_out));
- if (err) {
- mlx5_core_warn(dev, "set hca cap failed, %d\n", err);
+ err = mlx5_core_get_caps(dev, max_caps, HCA_CAP_OPMOD_GET_MAX);
+ if (err)
goto query_ex;
- }
- err = mlx5_cmd_status_to_err(&set_out.hdr);
+ err = mlx5_core_get_caps(dev, cur_caps, HCA_CAP_OPMOD_GET_CUR);
if (err)
goto query_ex;
+ /* we limit the size of the pkey table to 128 entries for now */
+ cur_caps->gen.pkey_table_size = 128;
+
+ if (prof->mask & MLX5_PROF_MASK_QP_SIZE)
+ cur_caps->gen.log_max_qp = prof->log_max_qp;
+
+ /* disable checksum */
+ cur_caps->gen.flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
+
+ copy_rw_fields(&set_ctx->hca_cap, &cur_caps->gen);
+ err = set_caps(dev, set_ctx);
+
query_ex:
- kfree(query_out);
+ kfree(cur_caps);
+ kfree(max_caps);
kfree(set_ctx);
return err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/uar.c b/drivers/net/ethernet/mellanox/mlx5/core/uar.c
index 68f5d9c77c7b..0a6348cefc01 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/uar.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/uar.c
@@ -174,11 +174,11 @@ int mlx5_alloc_uuars(struct mlx5_core_dev *dev, struct mlx5_uuar_info *uuari)
for (i = 0; i < tot_uuars; i++) {
bf = &uuari->bfs[i];
- bf->buf_size = dev->caps.bf_reg_size / 2;
+ bf->buf_size = dev->caps.gen.bf_reg_size / 2;
bf->uar = &uuari->uars[i / MLX5_BF_REGS_PER_PAGE];
bf->regreg = uuari->uars[i / MLX5_BF_REGS_PER_PAGE].map;
bf->reg = NULL; /* Add WC support */
- bf->offset = (i % MLX5_BF_REGS_PER_PAGE) * dev->caps.bf_reg_size +
+ bf->offset = (i % MLX5_BF_REGS_PER_PAGE) * dev->caps.gen.bf_reg_size +
MLX5_BF_OFFSET;
bf->need_lock = need_uuar_lock(i);
spin_lock_init(&bf->lock);
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 334947151dfc..dce01fd854a8 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -71,6 +71,11 @@ enum {
};
enum {
+ MLX5_MIN_PKEY_TABLE_SIZE = 128,
+ MLX5_MAX_LOG_PKEY_TABLE = 5,
+};
+
+enum {
MLX5_PERM_LOCAL_READ = 1 << 2,
MLX5_PERM_LOCAL_WRITE = 1 << 3,
MLX5_PERM_REMOTE_READ = 1 << 4,
@@ -184,10 +189,10 @@ enum {
MLX5_DEV_CAP_FLAG_CQ_MODER = 1LL << 29,
MLX5_DEV_CAP_FLAG_RESIZE_CQ = 1LL << 30,
MLX5_DEV_CAP_FLAG_RESIZE_SRQ = 1LL << 32,
+ MLX5_DEV_CAP_FLAG_DCT = 1LL << 37,
MLX5_DEV_CAP_FLAG_REMOTE_FENCE = 1LL << 38,
MLX5_DEV_CAP_FLAG_TLP_HINTS = 1LL << 39,
MLX5_DEV_CAP_FLAG_SIG_HAND_OVER = 1LL << 40,
- MLX5_DEV_CAP_FLAG_DCT = 1LL << 41,
MLX5_DEV_CAP_FLAG_CMDIF_CSUM = 3LL << 46,
};
@@ -243,10 +248,14 @@ enum {
};
enum {
- MLX5_CAP_OFF_DCT = 41,
MLX5_CAP_OFF_CMDIF_CSUM = 46,
};
+enum {
+ HCA_CAP_OPMOD_GET_MAX = 0,
+ HCA_CAP_OPMOD_GET_CUR = 1,
+};
+
struct mlx5_inbox_hdr {
__be16 opcode;
u8 rsvd[4];
@@ -303,9 +312,10 @@ struct mlx5_hca_cap {
u8 log_max_ra_req_qp;
u8 rsvd10;
u8 log_max_ra_res_qp;
- u8 rsvd11[4];
+ u8 pad_cap;
+ u8 rsvd11[3];
__be16 max_qp_count;
- __be16 rsvd12;
+ __be16 pkey_table_size;
u8 rsvd13;
u8 local_ca_ack_delay;
u8 rsvd14;
@@ -335,11 +345,7 @@ struct mlx5_hca_cap {
u8 log_max_xrcd;
u8 rsvd25[42];
__be16 log_uar_page_sz;
- u8 rsvd26[28];
- u8 log_max_atomic_size_qp;
- u8 rsvd27[2];
- u8 log_max_atomic_size_dc;
- u8 rsvd28[76];
+ u8 rsvd26[108];
};
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index b88e9b46d957..45a2add747e0 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -335,23 +335,30 @@ struct mlx5_port_caps {
int pkey_table_len;
};
-struct mlx5_caps {
+struct mlx5_general_caps {
u8 log_max_eq;
u8 log_max_cq;
u8 log_max_qp;
u8 log_max_mkey;
u8 log_max_pd;
u8 log_max_srq;
+ u8 log_max_strq;
+ u8 log_max_mrw_sz;
+ u8 log_max_bsf_list_size;
+ u8 log_max_klm_list_size;
u32 max_cqes;
int max_wqes;
+ u32 max_eqes;
+ u32 max_indirection;
int max_sq_desc_sz;
int max_rq_desc_sz;
+ int max_dc_sq_desc_sz;
u64 flags;
u16 stat_rate_support;
int log_max_msg;
int num_ports;
- int max_ra_res_qp;
- int max_ra_req_qp;
+ u8 log_max_ra_res_qp;
+ u8 log_max_ra_req_qp;
int max_srq_wqes;
int bf_reg_size;
int bf_regs_per_page;
@@ -363,6 +370,19 @@ struct mlx5_caps {
u8 log_max_mcg;
u32 max_qp_mcg;
int min_page_sz;
+ int pd_cap;
+ u32 max_qp_counters;
+ u32 pkey_table_size;
+ u8 log_max_ra_req_dc;
+ u8 log_max_ra_res_dc;
+ u32 uar_sz;
+ u8 min_log_pg_sz;
+ u8 log_max_xrcd;
+ u16 log_uar_page_sz;
+};
+
+struct mlx5_caps {
+ struct mlx5_general_caps gen;
};
struct mlx5_cmd_mailbox {
@@ -695,6 +715,8 @@ void mlx5_cmd_cleanup(struct mlx5_core_dev *dev);
void mlx5_cmd_use_events(struct mlx5_core_dev *dev);
void mlx5_cmd_use_polling(struct mlx5_core_dev *dev);
int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr);
+int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
+ u16 opmod);
int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
int out_size);
int mlx5_cmd_exec_cb(struct mlx5_core_dev *dev, void *in, int in_size,
--
2.1.1
^ permalink raw reply related
* [PATCH V2 net-next 3/5] net/mlx5_core: use set/get macros in device caps
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412241586-10108-1-git-send-email-eli@mellanox.com>
Transform device capabilities related commands to use set/get macros to
manipulate command mailboxes.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 17 ++
drivers/net/ethernet/mellanox/mlx5/core/main.c | 150 +++++++++---------
include/linux/mlx5/device.h | 92 -----------
include/linux/mlx5/driver.h | 1 +
include/linux/mlx5/mlx5_ifc.h | 206 +++++++++++++++++++++++++
5 files changed, 298 insertions(+), 168 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 3ecef1310bae..368c6c5ea014 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1537,3 +1537,20 @@ int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
return cmd_status_to_err(hdr->status);
}
+
+int mlx5_cmd_status_to_err_v2(void *ptr)
+{
+ u32 syndrome;
+ u8 status;
+
+ status = be32_to_cpu(*(__be32 *)ptr) >> 24;
+ if (!status)
+ return 0;
+
+ syndrome = be32_to_cpu(*(__be32 *)(ptr + 4));
+
+ pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
+ cmd_status_str(status), status, syndrome);
+
+ return cmd_status_to_err(status);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index d9f74618befa..b9e3259e415f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -43,6 +43,7 @@
#include <linux/mlx5/qp.h>
#include <linux/mlx5/srq.h>
#include <linux/debugfs.h>
+#include <linux/mlx5/mlx5_ifc.h>
#include "mlx5_core.h"
#define DRIVER_NAME "mlx5_core"
@@ -277,18 +278,20 @@ static u16 to_fw_pkey_sz(u32 size)
/* selectively copy writable fields clearing any reserved area
*/
-static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_general_caps *from)
+static void copy_rw_fields(void *to, struct mlx5_caps *from)
{
+ __be64 *flags_off = (__be64 *)MLX5_ADDR_OF(cmd_hca_cap, to, reserved_22);
u64 v64;
- to->log_max_qp = from->log_max_qp & 0x1f;
- to->log_max_ra_req_dc = from->log_max_ra_req_dc & 0x3f;
- to->log_max_ra_res_dc = from->log_max_ra_res_dc & 0x3f;
- to->log_max_ra_req_qp = from->log_max_ra_req_qp & 0x3f;
- to->log_max_ra_res_qp = from->log_max_ra_res_qp & 0x3f;
- to->pkey_table_size = cpu_to_be16(to_fw_pkey_sz(from->pkey_table_size));
- v64 = from->flags & MLX5_CAP_BITS_RW_MASK;
- to->flags = cpu_to_be64(v64);
+ MLX5_SET(cmd_hca_cap, to, log_max_qp, from->gen.log_max_qp);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_req_qp, from->gen.log_max_ra_req_qp);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_res_qp, from->gen.log_max_ra_res_qp);
+ MLX5_SET(cmd_hca_cap, to, pkey_table_size, from->gen.pkey_table_size);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_req_dc, from->gen.log_max_ra_req_dc);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_res_dc, from->gen.log_max_ra_res_dc);
+ MLX5_SET(cmd_hca_cap, to, pkey_table_size, to_fw_pkey_sz(from->gen.pkey_table_size));
+ v64 = from->gen.flags & MLX5_CAP_BITS_RW_MASK;
+ *flags_off = cpu_to_be64(v64);
}
static u16 get_pkey_table_size(int pkey)
@@ -299,55 +302,47 @@ static u16 get_pkey_table_size(int pkey)
return MLX5_MIN_PKEY_TABLE_SIZE << pkey;
}
-static void fw2drv_caps(struct mlx5_caps *caps,
- struct mlx5_cmd_query_hca_cap_mbox_out *out)
+static void fw2drv_caps(struct mlx5_caps *caps, void *out)
{
struct mlx5_general_caps *gen = &caps->gen;
- u16 t16;
-
- gen->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
- gen->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
- gen->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
- gen->log_max_strq = out->hca_cap.log_max_strq_sz;
- gen->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
- gen->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
- gen->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
- gen->max_eqes = out->hca_cap.log_max_eq_sz;
- gen->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
- gen->log_max_eq = out->hca_cap.log_max_eq & 0xf;
- gen->max_indirection = out->hca_cap.max_indirection;
- gen->log_max_mrw_sz = out->hca_cap.log_max_mrw_sz;
- gen->log_max_bsf_list_size = 0;
- gen->log_max_klm_list_size = 0;
- gen->log_max_ra_req_dc = out->hca_cap.log_max_ra_req_dc;
- gen->log_max_ra_res_dc = out->hca_cap.log_max_ra_res_dc;
- gen->log_max_ra_req_qp = out->hca_cap.log_max_ra_req_qp;
- gen->log_max_ra_res_qp = out->hca_cap.log_max_ra_res_qp;
- gen->max_qp_counters = be16_to_cpu(out->hca_cap.max_qp_count);
- gen->pkey_table_size = get_pkey_table_size(be16_to_cpu(out->hca_cap.pkey_table_size));
- gen->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
- gen->num_ports = out->hca_cap.num_ports & 0xf;
- gen->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
- gen->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
- gen->flags = be64_to_cpu(out->hca_cap.flags);
- pr_debug("flags = 0x%llx\n", gen->flags);
- gen->uar_sz = out->hca_cap.uar_sz;
- gen->min_log_pg_sz = out->hca_cap.log_pg_sz;
- t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
- if (t16 & 0x8000) {
- gen->bf_reg_size = 1 << (t16 & 0x1f);
- gen->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
- } else {
- gen->bf_reg_size = 0;
- gen->bf_regs_per_page = 0;
- }
- gen->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
- gen->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
- gen->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
- gen->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
- gen->log_max_xrcd = out->hca_cap.log_max_xrcd;
- gen->log_uar_page_sz = be16_to_cpu(out->hca_cap.log_uar_page_sz);
+ gen->max_srq_wqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_srq_sz);
+ gen->max_wqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_qp_sz);
+ gen->log_max_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_qp);
+ gen->log_max_strq = MLX5_GET_PR(cmd_hca_cap, out, log_max_strq_sz);
+ gen->log_max_srq = MLX5_GET_PR(cmd_hca_cap, out, log_max_srqs);
+ gen->max_cqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_cq_sz);
+ gen->log_max_cq = MLX5_GET_PR(cmd_hca_cap, out, log_max_cq);
+ gen->max_eqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_eq_sz);
+ gen->log_max_mkey = MLX5_GET_PR(cmd_hca_cap, out, log_max_mkey);
+ gen->log_max_eq = MLX5_GET_PR(cmd_hca_cap, out, log_max_eq);
+ gen->max_indirection = MLX5_GET_PR(cmd_hca_cap, out, max_indirection);
+ gen->log_max_mrw_sz = MLX5_GET_PR(cmd_hca_cap, out, log_max_mrw_sz);
+ gen->log_max_bsf_list_size = MLX5_GET_PR(cmd_hca_cap, out, log_max_bsf_list_size);
+ gen->log_max_klm_list_size = MLX5_GET_PR(cmd_hca_cap, out, log_max_klm_list_size);
+ gen->log_max_ra_req_dc = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_req_dc);
+ gen->log_max_ra_res_dc = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_res_dc);
+ gen->log_max_ra_req_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_req_qp);
+ gen->log_max_ra_res_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_res_qp);
+ gen->max_qp_counters = MLX5_GET_PR(cmd_hca_cap, out, max_qp_cnt);
+ gen->pkey_table_size = get_pkey_table_size(MLX5_GET_PR(cmd_hca_cap, out, pkey_table_size));
+ gen->local_ca_ack_delay = MLX5_GET_PR(cmd_hca_cap, out, local_ca_ack_delay);
+ gen->num_ports = MLX5_GET_PR(cmd_hca_cap, out, num_ports);
+ gen->log_max_msg = MLX5_GET_PR(cmd_hca_cap, out, log_max_msg);
+ gen->stat_rate_support = MLX5_GET_PR(cmd_hca_cap, out, stat_rate_support);
+ gen->flags = be64_to_cpu(*(__be64 *)MLX5_ADDR_OF(cmd_hca_cap, out, reserved_22));
+ pr_debug("flags = 0x%llx\n", gen->flags);
+ gen->uar_sz = MLX5_GET_PR(cmd_hca_cap, out, uar_sz);
+ gen->min_log_pg_sz = MLX5_GET_PR(cmd_hca_cap, out, log_pg_sz);
+ gen->bf_reg_size = MLX5_GET_PR(cmd_hca_cap, out, bf);
+ gen->bf_reg_size = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_bf_reg_size);
+ gen->max_sq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_sq);
+ gen->max_rq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_rq);
+ gen->max_dc_sq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_sq_dc);
+ gen->max_qp_mcg = MLX5_GET_PR(cmd_hca_cap, out, max_qp_mcg);
+ gen->log_max_pd = MLX5_GET_PR(cmd_hca_cap, out, log_max_pd);
+ gen->log_max_xrcd = MLX5_GET_PR(cmd_hca_cap, out, log_max_xrcd);
+ gen->log_uar_page_sz = MLX5_GET_PR(cmd_hca_cap, out, log_uar_page_sz);
}
static const char *caps_opmod_str(u16 opmod)
@@ -365,59 +360,61 @@ static const char *caps_opmod_str(u16 opmod)
int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
u16 opmod)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *out;
- struct mlx5_cmd_query_hca_cap_mbox_in in;
+ u8 in[MLX5_ST_SZ_BYTES(query_hca_cap_in)];
+ int out_sz = MLX5_ST_SZ_BYTES(query_hca_cap_out);
+ void *out;
int err;
- memset(&in, 0, sizeof(in));
- out = kzalloc(sizeof(*out), GFP_KERNEL);
+ memset(in, 0, sizeof(in));
+ out = kzalloc(out_sz, GFP_KERNEL);
if (!out)
return -ENOMEM;
+ MLX5_SET(query_hca_cap_in, in, opcode, MLX5_CMD_OP_QUERY_HCA_CAP);
+ MLX5_SET(query_hca_cap_in, in, op_mod, opmod);
+ err = mlx5_cmd_exec(dev, in, sizeof(in), out, out_sz);
+ if (err)
+ goto query_ex;
- in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- in.hdr.opmod = cpu_to_be16(opmod);
- err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
-
- err = mlx5_cmd_status_to_err(&out->hdr);
+ err = mlx5_cmd_status_to_err_v2(out);
if (err) {
mlx5_core_warn(dev, "query max hca cap failed, %d\n", err);
goto query_ex;
}
mlx5_core_dbg(dev, "%s\n", caps_opmod_str(opmod));
- fw2drv_caps(caps, out);
+ fw2drv_caps(caps, MLX5_ADDR_OF(query_hca_cap_out, out, capability_struct));
query_ex:
kfree(out);
return err;
}
-static int set_caps(struct mlx5_core_dev *dev,
- struct mlx5_cmd_set_hca_cap_mbox_in *in)
+static int set_caps(struct mlx5_core_dev *dev, void *in, int in_sz)
{
- struct mlx5_cmd_set_hca_cap_mbox_out out;
+ u32 out[MLX5_ST_SZ_DW(set_hca_cap_out)];
int err;
- memset(&out, 0, sizeof(out));
+ memset(out, 0, sizeof(out));
- in->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
- err = mlx5_cmd_exec(dev, in, sizeof(*in), &out, sizeof(out));
+ MLX5_SET(set_hca_cap_in, in, opcode, MLX5_CMD_OP_SET_HCA_CAP);
+ err = mlx5_cmd_exec(dev, in, in_sz, out, sizeof(out));
if (err)
return err;
- err = mlx5_cmd_status_to_err(&out.hdr);
+ err = mlx5_cmd_status_to_err_v2(out);
return err;
}
static int handle_hca_cap(struct mlx5_core_dev *dev)
{
- struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
+ void *set_ctx = NULL;
struct mlx5_profile *prof = dev->profile;
struct mlx5_caps *cur_caps = NULL;
struct mlx5_caps *max_caps = NULL;
int err = -ENOMEM;
+ int set_sz = MLX5_ST_SZ_BYTES(set_hca_cap_in);
- set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
+ set_ctx = kzalloc(set_sz, GFP_KERNEL);
if (!set_ctx)
goto query_ex;
@@ -446,8 +443,9 @@ static int handle_hca_cap(struct mlx5_core_dev *dev)
/* disable checksum */
cur_caps->gen.flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
- copy_rw_fields(&set_ctx->hca_cap, &cur_caps->gen);
- err = set_caps(dev, set_ctx);
+ copy_rw_fields(MLX5_ADDR_OF(set_hca_cap_in, set_ctx, hca_capability_struct),
+ cur_caps);
+ err = set_caps(dev, set_ctx, set_sz);
query_ex:
kfree(cur_caps);
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 0032687f58c7..1d67fd32e71c 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -327,98 +327,6 @@ struct mlx5_cmd_query_adapter_mbox_out {
u8 vsd_psid[16];
};
-struct mlx5_hca_cap {
- u8 rsvd1[16];
- u8 log_max_srq_sz;
- u8 log_max_qp_sz;
- u8 rsvd2;
- u8 log_max_qp;
- u8 log_max_strq_sz;
- u8 log_max_srqs;
- u8 rsvd4[2];
- u8 rsvd5;
- u8 log_max_cq_sz;
- u8 rsvd6;
- u8 log_max_cq;
- u8 log_max_eq_sz;
- u8 log_max_mkey;
- u8 rsvd7;
- u8 log_max_eq;
- u8 max_indirection;
- u8 log_max_mrw_sz;
- u8 log_max_bsf_list_sz;
- u8 log_max_klm_list_sz;
- u8 rsvd_8_0;
- u8 log_max_ra_req_dc;
- u8 rsvd_8_1;
- u8 log_max_ra_res_dc;
- u8 rsvd9;
- u8 log_max_ra_req_qp;
- u8 rsvd10;
- u8 log_max_ra_res_qp;
- u8 pad_cap;
- u8 rsvd11[3];
- __be16 max_qp_count;
- __be16 pkey_table_size;
- u8 rsvd13;
- u8 local_ca_ack_delay;
- u8 rsvd14;
- u8 num_ports;
- u8 log_max_msg;
- u8 rsvd15[3];
- __be16 stat_rate_support;
- u8 rsvd16[2];
- __be64 flags;
- u8 rsvd17;
- u8 uar_sz;
- u8 rsvd18;
- u8 log_pg_sz;
- __be16 bf_log_bf_reg_size;
- u8 rsvd19[4];
- __be16 max_desc_sz_sq;
- u8 rsvd20[2];
- __be16 max_desc_sz_rq;
- u8 rsvd21[2];
- __be16 max_desc_sz_sq_dc;
- __be32 max_qp_mcg;
- u8 rsvd22[3];
- u8 log_max_mcg;
- u8 rsvd23;
- u8 log_max_pd;
- u8 rsvd24;
- u8 log_max_xrcd;
- u8 rsvd25[42];
- __be16 log_uar_page_sz;
- u8 rsvd26[108];
-};
-
-
-struct mlx5_cmd_query_hca_cap_mbox_in {
- struct mlx5_inbox_hdr hdr;
- u8 rsvd[8];
-};
-
-
-struct mlx5_cmd_query_hca_cap_mbox_out {
- struct mlx5_outbox_hdr hdr;
- u8 rsvd0[8];
- struct mlx5_hca_cap hca_cap;
-};
-
-
-struct mlx5_cmd_set_hca_cap_mbox_in {
- struct mlx5_inbox_hdr hdr;
- u8 rsvd[8];
- struct mlx5_hca_cap hca_cap;
-};
-
-
-struct mlx5_cmd_set_hca_cap_mbox_out {
- struct mlx5_outbox_hdr hdr;
- u8 rsvd0[8];
-};
-
-
struct mlx5_cmd_init_hca_mbox_in {
struct mlx5_inbox_hdr hdr;
u8 rsvd0[2];
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 6f48dc793b9f..c439f9c59b93 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -641,6 +641,7 @@ void mlx5_cmd_cleanup(struct mlx5_core_dev *dev);
void mlx5_cmd_use_events(struct mlx5_core_dev *dev);
void mlx5_cmd_use_polling(struct mlx5_core_dev *dev);
int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr);
+int mlx5_cmd_status_to_err_v2(void *ptr);
int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
u16 opmod);
int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
index df3bd9b5fbcf..5f48b8f592c5 100644
--- a/include/linux/mlx5/mlx5_ifc.h
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -140,4 +140,210 @@ enum {
MLX5_CMD_OP_MAX = 0x911
};
+struct mlx5_ifc_cmd_hca_cap_bits {
+ u8 reserved_0[0x80];
+
+ u8 log_max_srq_sz[0x8];
+ u8 log_max_qp_sz[0x8];
+ u8 reserved_1[0xb];
+ u8 log_max_qp[0x5];
+
+ u8 log_max_strq_sz[0x8];
+ u8 reserved_2[0x3];
+ u8 log_max_srqs[0x5];
+ u8 reserved_3[0x10];
+
+ u8 reserved_4[0x8];
+ u8 log_max_cq_sz[0x8];
+ u8 reserved_5[0xb];
+ u8 log_max_cq[0x5];
+
+ u8 log_max_eq_sz[0x8];
+ u8 reserved_6[0x2];
+ u8 log_max_mkey[0x6];
+ u8 reserved_7[0xc];
+ u8 log_max_eq[0x4];
+
+ u8 max_indirection[0x8];
+ u8 reserved_8[0x1];
+ u8 log_max_mrw_sz[0x7];
+ u8 reserved_9[0x2];
+ u8 log_max_bsf_list_size[0x6];
+ u8 reserved_10[0x2];
+ u8 log_max_klm_list_size[0x6];
+
+ u8 reserved_11[0xa];
+ u8 log_max_ra_req_dc[0x6];
+ u8 reserved_12[0xa];
+ u8 log_max_ra_res_dc[0x6];
+
+ u8 reserved_13[0xa];
+ u8 log_max_ra_req_qp[0x6];
+ u8 reserved_14[0xa];
+ u8 log_max_ra_res_qp[0x6];
+
+ u8 pad_cap[0x1];
+ u8 cc_query_allowed[0x1];
+ u8 cc_modify_allowed[0x1];
+ u8 reserved_15[0x1d];
+
+ u8 reserved_16[0x6];
+ u8 max_qp_cnt[0xa];
+ u8 pkey_table_size[0x10];
+
+ u8 eswitch_owner[0x1];
+ u8 reserved_17[0xa];
+ u8 local_ca_ack_delay[0x5];
+ u8 reserved_18[0x8];
+ u8 num_ports[0x8];
+
+ u8 reserved_19[0x3];
+ u8 log_max_msg[0x5];
+ u8 reserved_20[0x18];
+
+ u8 stat_rate_support[0x10];
+ u8 reserved_21[0x10];
+
+ u8 reserved_22[0x10];
+ u8 cmdif_checksum[0x2];
+ u8 sigerr_cqe[0x1];
+ u8 reserved_23[0x1];
+ u8 wq_signature[0x1];
+ u8 sctr_data_cqe[0x1];
+ u8 reserved_24[0x1];
+ u8 sho[0x1];
+ u8 tph[0x1];
+ u8 rf[0x1];
+ u8 dc[0x1];
+ u8 reserved_25[0x2];
+ u8 roce[0x1];
+ u8 atomic[0x1];
+ u8 rsz_srq[0x1];
+
+ u8 cq_oi[0x1];
+ u8 cq_resize[0x1];
+ u8 cq_moderation[0x1];
+ u8 sniffer_rule_flow[0x1];
+ u8 sniffer_rule_vport[0x1];
+ u8 sniffer_rule_phy[0x1];
+ u8 reserved_26[0x1];
+ u8 pg[0x1];
+ u8 block_lb_mc[0x1];
+ u8 reserved_27[0x3];
+ u8 cd[0x1];
+ u8 reserved_28[0x1];
+ u8 apm[0x1];
+ u8 reserved_29[0x7];
+ u8 qkv[0x1];
+ u8 pkv[0x1];
+ u8 reserved_30[0x4];
+ u8 xrc[0x1];
+ u8 ud[0x1];
+ u8 uc[0x1];
+ u8 rc[0x1];
+
+ u8 reserved_31[0xa];
+ u8 uar_sz[0x6];
+ u8 reserved_32[0x8];
+ u8 log_pg_sz[0x8];
+
+ u8 bf[0x1];
+ u8 reserved_33[0xa];
+ u8 log_bf_reg_size[0x5];
+ u8 reserved_34[0x10];
+
+ u8 reserved_35[0x10];
+ u8 max_wqe_sz_sq[0x10];
+
+ u8 reserved_36[0x10];
+ u8 max_wqe_sz_rq[0x10];
+
+ u8 reserved_37[0x10];
+ u8 max_wqe_sz_sq_dc[0x10];
+
+ u8 reserved_38[0x7];
+ u8 max_qp_mcg[0x19];
+
+ u8 reserved_39[0x18];
+ u8 log_max_mcg[0x8];
+
+ u8 reserved_40[0xb];
+ u8 log_max_pd[0x5];
+ u8 reserved_41[0xb];
+ u8 log_max_xrcd[0x5];
+
+ u8 reserved_42[0x20];
+
+ u8 reserved_43[0x3];
+ u8 log_max_rq[0x5];
+ u8 reserved_44[0x3];
+ u8 log_max_sq[0x5];
+ u8 reserved_45[0x3];
+ u8 log_max_tir[0x5];
+ u8 reserved_46[0x3];
+ u8 log_max_tis[0x5];
+
+ u8 reserved_47[0x13];
+ u8 log_max_rq_per_tir[0x5];
+ u8 reserved_48[0x3];
+ u8 log_max_tis_per_sq[0x5];
+
+ u8 reserved_49[0xe0];
+
+ u8 reserved_50[0x10];
+ u8 log_uar_page_sz[0x10];
+
+ u8 reserved_51[0x100];
+
+ u8 reserved_52[0x1f];
+ u8 cqe_zip[0x1];
+
+ u8 cqe_zip_timeout[0x10];
+ u8 cqe_zip_max_num[0x10];
+
+ u8 reserved_53[0x220];
+};
+
+struct mlx5_ifc_set_hca_cap_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_cmd_hca_cap_bits hca_capability_struct;
+};
+
+struct mlx5_ifc_query_hca_cap_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_hca_cap_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 capability_struct[256][0x8];
+};
+
+struct mlx5_ifc_set_hca_cap_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
#endif /* MLX5_IFC_H */
--
2.1.1
^ permalink raw reply related
* [PATCH V2 net-next 2/5] net/mlx5_core: Use hardware registers description header file
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412241586-10108-1-git-send-email-eli@mellanox.com>
Add an auto generated header file that describes hardware registers along with
set of macros that set/get values. The macros do static checks to avoid
overflow, handle endianess, and overall provide a clean way to code commands.
Currently the header file is small and we will add structs as we make use of
the macros.
A few commands were removed from the commands enum since they are not supported
currently and will be added when support is available.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 36 -------
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 3 -
include/linux/mlx5/device.h | 44 ++++++++
include/linux/mlx5/driver.h | 76 +-------------
include/linux/mlx5/mlx5_ifc.h | 143 ++++++++++++++++++++++++++
5 files changed, 188 insertions(+), 114 deletions(-)
create mode 100644 include/linux/mlx5/mlx5_ifc.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 6eb0f85cf872..3ecef1310bae 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -357,60 +357,24 @@ const char *mlx5_command_str(int command)
case MLX5_CMD_OP_2ERR_QP:
return "2ERR_QP";
- case MLX5_CMD_OP_RTS2SQD_QP:
- return "RTS2SQD_QP";
-
- case MLX5_CMD_OP_SQD2RTS_QP:
- return "SQD2RTS_QP";
-
case MLX5_CMD_OP_2RST_QP:
return "2RST_QP";
case MLX5_CMD_OP_QUERY_QP:
return "QUERY_QP";
- case MLX5_CMD_OP_CONF_SQP:
- return "CONF_SQP";
-
case MLX5_CMD_OP_MAD_IFC:
return "MAD_IFC";
case MLX5_CMD_OP_INIT2INIT_QP:
return "INIT2INIT_QP";
- case MLX5_CMD_OP_SUSPEND_QP:
- return "SUSPEND_QP";
-
- case MLX5_CMD_OP_UNSUSPEND_QP:
- return "UNSUSPEND_QP";
-
- case MLX5_CMD_OP_SQD2SQD_QP:
- return "SQD2SQD_QP";
-
- case MLX5_CMD_OP_ALLOC_QP_COUNTER_SET:
- return "ALLOC_QP_COUNTER_SET";
-
- case MLX5_CMD_OP_DEALLOC_QP_COUNTER_SET:
- return "DEALLOC_QP_COUNTER_SET";
-
- case MLX5_CMD_OP_QUERY_QP_COUNTER_SET:
- return "QUERY_QP_COUNTER_SET";
-
case MLX5_CMD_OP_CREATE_PSV:
return "CREATE_PSV";
case MLX5_CMD_OP_DESTROY_PSV:
return "DESTROY_PSV";
- case MLX5_CMD_OP_QUERY_PSV:
- return "QUERY_PSV";
-
- case MLX5_CMD_OP_QUERY_SIG_RULE_TABLE:
- return "QUERY_SIG_RULE_TABLE";
-
- case MLX5_CMD_OP_QUERY_BLOCK_SIZE_TABLE:
- return "QUERY_BLOCK_SIZE_TABLE";
-
case MLX5_CMD_OP_CREATE_SRQ:
return "CREATE_SRQ";
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/qp.c b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
index 8145b4668229..415b67ce379e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
@@ -184,13 +184,10 @@ int mlx5_core_qp_modify(struct mlx5_core_dev *dev, enum mlx5_qp_state cur_state,
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
[MLX5_QP_STATE_ERR] = MLX5_CMD_OP_2ERR_QP,
[MLX5_QP_STATE_RTS] = MLX5_CMD_OP_RTS2RTS_QP,
- [MLX5_QP_STATE_SQD] = MLX5_CMD_OP_RTS2SQD_QP,
},
[MLX5_QP_STATE_SQD] = {
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
[MLX5_QP_STATE_ERR] = MLX5_CMD_OP_2ERR_QP,
- [MLX5_QP_STATE_RTS] = MLX5_CMD_OP_SQD2RTS_QP,
- [MLX5_QP_STATE_SQD] = MLX5_CMD_OP_SQD2SQD_QP,
},
[MLX5_QP_STATE_SQER] = {
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index dce01fd854a8..0032687f58c7 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -44,6 +44,50 @@
#error Host endianness not defined
#endif
+/* helper macros */
+#define __mlx5_nullp(typ) ((struct mlx5_ifc_##typ##_bits *)0)
+#define __mlx5_bit_sz(typ, fld) sizeof(__mlx5_nullp(typ)->fld)
+#define __mlx5_bit_off(typ, fld) ((unsigned)(unsigned long)(&(__mlx5_nullp(typ)->fld)))
+#define __mlx5_dw_off(typ, fld) (__mlx5_bit_off(typ, fld) / 32)
+#define __mlx5_64_off(typ, fld) (__mlx5_bit_off(typ, fld) / 64)
+#define __mlx5_dw_bit_off(typ, fld) (32 - __mlx5_bit_sz(typ, fld) - (__mlx5_bit_off(typ, fld) & 0x1f))
+#define __mlx5_mask(typ, fld) ((u32)((1ull << __mlx5_bit_sz(typ, fld)) - 1))
+#define __mlx5_dw_mask(typ, fld) (__mlx5_mask(typ, fld) << __mlx5_dw_bit_off(typ, fld))
+#define __mlx5_st_sz_bits(typ) sizeof(struct mlx5_ifc_##typ##_bits)
+
+#define MLX5_FLD_SZ_BYTES(typ, fld) (__mlx5_bit_sz(typ, fld) / 8)
+#define MLX5_ST_SZ_BYTES(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 8)
+#define MLX5_ST_SZ_DW(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 32)
+#define MLX5_BYTE_OFF(typ, fld) (__mlx5_bit_off(typ, fld) / 8)
+#define MLX5_ADDR_OF(typ, p, fld) ((char *)(p) + MLX5_BYTE_OFF(typ, fld))
+
+/* insert a value to a struct */
+#define MLX5_SET(typ, p, fld, v) do { \
+ BUILD_BUG_ON(__mlx5_st_sz_bits(typ) % 32); \
+ *((__be32 *)(p) + __mlx5_dw_off(typ, fld)) = \
+ cpu_to_be32((be32_to_cpu(*((__be32 *)(p) + __mlx5_dw_off(typ, fld))) & \
+ (~__mlx5_dw_mask(typ, fld))) | (((v) & __mlx5_mask(typ, fld)) \
+ << __mlx5_dw_bit_off(typ, fld))); \
+} while (0)
+
+#define MLX5_GET(typ, p, fld) ((be32_to_cpu(*((__be32 *)(p) +\
+__mlx5_dw_off(typ, fld))) >> __mlx5_dw_bit_off(typ, fld)) & \
+__mlx5_mask(typ, fld))
+
+#define MLX5_GET_PR(typ, p, fld) ({ \
+ u32 ___t = MLX5_GET(typ, p, fld); \
+ pr_debug(#fld " = 0x%x\n", ___t); \
+ ___t; \
+})
+
+#define MLX5_SET64(typ, p, fld, v) do { \
+ BUILD_BUG_ON(__mlx5_bit_sz(typ, fld) != 64); \
+ BUILD_BUG_ON(__mlx5_bit_off(typ, fld) % 64); \
+ *((__be64 *)(p) + __mlx5_64_off(typ, fld)) = cpu_to_be64(v); \
+} while (0)
+
+#define MLX5_GET64(typ, p, fld) be64_to_cpu(*((__be64 *)(p) + __mlx5_64_off(typ, fld)))
+
enum {
MLX5_MAX_COMMANDS = 32,
MLX5_CMD_DATA_BLOCK_SIZE = 512,
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 45a2add747e0..6f48dc793b9f 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -44,6 +44,7 @@
#include <linux/mlx5/device.h>
#include <linux/mlx5/doorbell.h>
+#include <linux/mlx5/mlx5_ifc.h>
enum {
MLX5_BOARD_ID_LEN = 64,
@@ -99,81 +100,6 @@ enum {
};
enum {
- MLX5_CMD_OP_QUERY_HCA_CAP = 0x100,
- MLX5_CMD_OP_QUERY_ADAPTER = 0x101,
- MLX5_CMD_OP_INIT_HCA = 0x102,
- MLX5_CMD_OP_TEARDOWN_HCA = 0x103,
- MLX5_CMD_OP_ENABLE_HCA = 0x104,
- MLX5_CMD_OP_DISABLE_HCA = 0x105,
- MLX5_CMD_OP_QUERY_PAGES = 0x107,
- MLX5_CMD_OP_MANAGE_PAGES = 0x108,
- MLX5_CMD_OP_SET_HCA_CAP = 0x109,
-
- MLX5_CMD_OP_CREATE_MKEY = 0x200,
- MLX5_CMD_OP_QUERY_MKEY = 0x201,
- MLX5_CMD_OP_DESTROY_MKEY = 0x202,
- MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS = 0x203,
-
- MLX5_CMD_OP_CREATE_EQ = 0x301,
- MLX5_CMD_OP_DESTROY_EQ = 0x302,
- MLX5_CMD_OP_QUERY_EQ = 0x303,
-
- MLX5_CMD_OP_CREATE_CQ = 0x400,
- MLX5_CMD_OP_DESTROY_CQ = 0x401,
- MLX5_CMD_OP_QUERY_CQ = 0x402,
- MLX5_CMD_OP_MODIFY_CQ = 0x403,
-
- MLX5_CMD_OP_CREATE_QP = 0x500,
- MLX5_CMD_OP_DESTROY_QP = 0x501,
- MLX5_CMD_OP_RST2INIT_QP = 0x502,
- MLX5_CMD_OP_INIT2RTR_QP = 0x503,
- MLX5_CMD_OP_RTR2RTS_QP = 0x504,
- MLX5_CMD_OP_RTS2RTS_QP = 0x505,
- MLX5_CMD_OP_SQERR2RTS_QP = 0x506,
- MLX5_CMD_OP_2ERR_QP = 0x507,
- MLX5_CMD_OP_RTS2SQD_QP = 0x508,
- MLX5_CMD_OP_SQD2RTS_QP = 0x509,
- MLX5_CMD_OP_2RST_QP = 0x50a,
- MLX5_CMD_OP_QUERY_QP = 0x50b,
- MLX5_CMD_OP_CONF_SQP = 0x50c,
- MLX5_CMD_OP_MAD_IFC = 0x50d,
- MLX5_CMD_OP_INIT2INIT_QP = 0x50e,
- MLX5_CMD_OP_SUSPEND_QP = 0x50f,
- MLX5_CMD_OP_UNSUSPEND_QP = 0x510,
- MLX5_CMD_OP_SQD2SQD_QP = 0x511,
- MLX5_CMD_OP_ALLOC_QP_COUNTER_SET = 0x512,
- MLX5_CMD_OP_DEALLOC_QP_COUNTER_SET = 0x513,
- MLX5_CMD_OP_QUERY_QP_COUNTER_SET = 0x514,
-
- MLX5_CMD_OP_CREATE_PSV = 0x600,
- MLX5_CMD_OP_DESTROY_PSV = 0x601,
- MLX5_CMD_OP_QUERY_PSV = 0x602,
- MLX5_CMD_OP_QUERY_SIG_RULE_TABLE = 0x603,
- MLX5_CMD_OP_QUERY_BLOCK_SIZE_TABLE = 0x604,
-
- MLX5_CMD_OP_CREATE_SRQ = 0x700,
- MLX5_CMD_OP_DESTROY_SRQ = 0x701,
- MLX5_CMD_OP_QUERY_SRQ = 0x702,
- MLX5_CMD_OP_ARM_RQ = 0x703,
- MLX5_CMD_OP_RESIZE_SRQ = 0x704,
-
- MLX5_CMD_OP_ALLOC_PD = 0x800,
- MLX5_CMD_OP_DEALLOC_PD = 0x801,
- MLX5_CMD_OP_ALLOC_UAR = 0x802,
- MLX5_CMD_OP_DEALLOC_UAR = 0x803,
-
- MLX5_CMD_OP_ATTACH_TO_MCG = 0x806,
- MLX5_CMD_OP_DETACH_FROM_MCG = 0x807,
-
-
- MLX5_CMD_OP_ALLOC_XRCD = 0x80e,
- MLX5_CMD_OP_DEALLOC_XRCD = 0x80f,
-
- MLX5_CMD_OP_ACCESS_REG = 0x805,
- MLX5_CMD_OP_MAX = 0x810,
-};
-
-enum {
MLX5_REG_PCAP = 0x5001,
MLX5_REG_PMTU = 0x5003,
MLX5_REG_PTYS = 0x5004,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
new file mode 100644
index 000000000000..df3bd9b5fbcf
--- /dev/null
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -0,0 +1,143 @@
+/*
+ * Copyright (c) 2014, Mellanox Technologies 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.
+ */
+
+#ifndef MLX5_IFC_H
+#define MLX5_IFC_H
+
+enum {
+ MLX5_CMD_OP_QUERY_HCA_CAP = 0x100,
+ MLX5_CMD_OP_QUERY_ADAPTER = 0x101,
+ MLX5_CMD_OP_INIT_HCA = 0x102,
+ MLX5_CMD_OP_TEARDOWN_HCA = 0x103,
+ MLX5_CMD_OP_ENABLE_HCA = 0x104,
+ MLX5_CMD_OP_DISABLE_HCA = 0x105,
+ MLX5_CMD_OP_QUERY_PAGES = 0x107,
+ MLX5_CMD_OP_MANAGE_PAGES = 0x108,
+ MLX5_CMD_OP_SET_HCA_CAP = 0x109,
+ MLX5_CMD_OP_CREATE_MKEY = 0x200,
+ MLX5_CMD_OP_QUERY_MKEY = 0x201,
+ MLX5_CMD_OP_DESTROY_MKEY = 0x202,
+ MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS = 0x203,
+ MLX5_CMD_OP_PAGE_FAULT_RESUME = 0x204,
+ MLX5_CMD_OP_CREATE_EQ = 0x301,
+ MLX5_CMD_OP_DESTROY_EQ = 0x302,
+ MLX5_CMD_OP_QUERY_EQ = 0x303,
+ MLX5_CMD_OP_GEN_EQE = 0x304,
+ MLX5_CMD_OP_CREATE_CQ = 0x400,
+ MLX5_CMD_OP_DESTROY_CQ = 0x401,
+ MLX5_CMD_OP_QUERY_CQ = 0x402,
+ MLX5_CMD_OP_MODIFY_CQ = 0x403,
+ MLX5_CMD_OP_CREATE_QP = 0x500,
+ MLX5_CMD_OP_DESTROY_QP = 0x501,
+ MLX5_CMD_OP_RST2INIT_QP = 0x502,
+ MLX5_CMD_OP_INIT2RTR_QP = 0x503,
+ MLX5_CMD_OP_RTR2RTS_QP = 0x504,
+ MLX5_CMD_OP_RTS2RTS_QP = 0x505,
+ MLX5_CMD_OP_SQERR2RTS_QP = 0x506,
+ MLX5_CMD_OP_2ERR_QP = 0x507,
+ MLX5_CMD_OP_2RST_QP = 0x50a,
+ MLX5_CMD_OP_QUERY_QP = 0x50b,
+ MLX5_CMD_OP_INIT2INIT_QP = 0x50e,
+ MLX5_CMD_OP_CREATE_PSV = 0x600,
+ MLX5_CMD_OP_DESTROY_PSV = 0x601,
+ MLX5_CMD_OP_CREATE_SRQ = 0x700,
+ MLX5_CMD_OP_DESTROY_SRQ = 0x701,
+ MLX5_CMD_OP_QUERY_SRQ = 0x702,
+ MLX5_CMD_OP_ARM_RQ = 0x703,
+ MLX5_CMD_OP_RESIZE_SRQ = 0x704,
+ MLX5_CMD_OP_CREATE_DCT = 0x710,
+ MLX5_CMD_OP_DESTROY_DCT = 0x711,
+ MLX5_CMD_OP_DRAIN_DCT = 0x712,
+ MLX5_CMD_OP_QUERY_DCT = 0x713,
+ MLX5_CMD_OP_ARM_DCT_FOR_KEY_VIOLATION = 0x714,
+ MLX5_CMD_OP_QUERY_VPORT_STATE = 0x750,
+ MLX5_CMD_OP_MODIFY_VPORT_STATE = 0x751,
+ MLX5_CMD_OP_QUERY_ESW_VPORT_CONTEXT = 0x752,
+ MLX5_CMD_OP_MODIFY_ESW_VPORT_CONTEXT = 0x753,
+ MLX5_CMD_OP_QUERY_NIC_VPORT_CONTEXT = 0x754,
+ MLX5_CMD_OP_MODIFY_NIC_VPORT_CONTEXT = 0x755,
+ MLX5_CMD_OP_QUERY_RCOE_ADDRESS = 0x760,
+ MLX5_CMD_OP_SET_ROCE_ADDRESS = 0x761,
+ MLX5_CMD_OP_QUERY_VPORT_COUNTER = 0x770,
+ MLX5_CMD_OP_ALLOC_Q_COUNTER = 0x771,
+ MLX5_CMD_OP_DEALLOC_Q_COUNTER = 0x772,
+ MLX5_CMD_OP_QUERY_Q_COUNTER = 0x773,
+ MLX5_CMD_OP_ALLOC_PD = 0x800,
+ MLX5_CMD_OP_DEALLOC_PD = 0x801,
+ MLX5_CMD_OP_ALLOC_UAR = 0x802,
+ MLX5_CMD_OP_DEALLOC_UAR = 0x803,
+ MLX5_CMD_OP_CONFIG_INT_MODERATION = 0x804,
+ MLX5_CMD_OP_ACCESS_REG = 0x805,
+ MLX5_CMD_OP_ATTACH_TO_MCG = 0x806,
+ MLX5_CMD_OP_DETACH_FROM_MCG = 0x807,
+ MLX5_CMD_OP_GET_DROPPED_PACKET_LOG = 0x80a,
+ MLX5_CMD_OP_MAD_IFC = 0x50d,
+ MLX5_CMD_OP_QUERY_MAD_DEMUX = 0x80b,
+ MLX5_CMD_OP_SET_MAD_DEMUX = 0x80c,
+ MLX5_CMD_OP_NOP = 0x80d,
+ MLX5_CMD_OP_ALLOC_XRCD = 0x80e,
+ MLX5_CMD_OP_DEALLOC_XRCD = 0x80f,
+ MLX5_CMD_OP_SET_BURST_SIZE = 0x812,
+ MLX5_CMD_OP_QUERY_BURST_SZIE = 0x813,
+ MLX5_CMD_OP_ACTIVATE_TRACER = 0x814,
+ MLX5_CMD_OP_DEACTIVATE_TRACER = 0x815,
+ MLX5_CMD_OP_CREATE_SNIFFER_RULE = 0x820,
+ MLX5_CMD_OP_DESTROY_SNIFFER_RULE = 0x821,
+ MLX5_CMD_OP_QUERY_CONG_PARAMS = 0x822,
+ MLX5_CMD_OP_MODIFY_CONG_PARAMS = 0x823,
+ MLX5_CMD_OP_QUERY_CONG_STATISTICS = 0x824,
+ MLX5_CMD_OP_CREATE_TIR = 0x900,
+ MLX5_CMD_OP_MODIFY_TIR = 0x901,
+ MLX5_CMD_OP_DESTROY_TIR = 0x902,
+ MLX5_CMD_OP_QUERY_TIR = 0x903,
+ MLX5_CMD_OP_CREATE_TIS = 0x912,
+ MLX5_CMD_OP_MODIFY_TIS = 0x913,
+ MLX5_CMD_OP_DESTROY_TIS = 0x914,
+ MLX5_CMD_OP_QUERY_TIS = 0x915,
+ MLX5_CMD_OP_CREATE_SQ = 0x904,
+ MLX5_CMD_OP_MODIFY_SQ = 0x905,
+ MLX5_CMD_OP_DESTROY_SQ = 0x906,
+ MLX5_CMD_OP_QUERY_SQ = 0x907,
+ MLX5_CMD_OP_CREATE_RQ = 0x908,
+ MLX5_CMD_OP_MODIFY_RQ = 0x909,
+ MLX5_CMD_OP_DESTROY_RQ = 0x90a,
+ MLX5_CMD_OP_QUERY_RQ = 0x90b,
+ MLX5_CMD_OP_CREATE_RMP = 0x90c,
+ MLX5_CMD_OP_MODIFY_RMP = 0x90d,
+ MLX5_CMD_OP_DESTROY_RMP = 0x90e,
+ MLX5_CMD_OP_QUERY_RMP = 0x90f,
+ MLX5_CMD_OP_SET_FLOW_TABLE_ENTRY = 0x910,
+ MLX5_CMD_OP_QUERY_FLOW_TABLE_ENTRY = 0x911,
+ MLX5_CMD_OP_MAX = 0x911
+};
+
+#endif /* MLX5_IFC_H */
--
2.1.1
^ permalink raw reply related
* [PATCH V2 net-next 5/5] net/mlx5_core: Add ConnectX-4 to list of supported devices
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412241586-10108-1-git-send-email-eli@mellanox.com>
Add the upcoming ConnectX-4 device to the list of supported devices by then
mlx5 driver.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/main.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index b9e3259e415f..3d8e8e489b2d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -897,6 +897,7 @@ static void remove_one(struct pci_dev *pdev)
static const struct pci_device_id mlx5_core_pci_table[] = {
{ PCI_VDEVICE(MELLANOX, 4113) }, /* MT4113 Connect-IB */
+ { PCI_VDEVICE(MELLANOX, 4115) }, /* ConnectX-4 */
{ 0, }
};
--
2.1.1
^ permalink raw reply related
* [PATCH V2 net-next 4/5] net/mlx5_core: Identify resources by their type
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412241586-10108-1-git-send-email-eli@mellanox.com>
This patch puts a common part as the first field of mlx5_core_qp. This field is
used to identify which resource generated an event. This is required since upcoming
new resource types such as DC targets are allocated for the same numerical space
as regular QPs and may generate the same events. By searching the resource in the
same table we can then look at the common field to identify the resource.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 12 +++---
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 57 ++++++++++++++++++++--------
include/linux/mlx5/driver.h | 13 ++++++-
include/linux/mlx5/qp.h | 3 +-
4 files changed, 60 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 11b9b840ad4d..ed53291468f3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -198,7 +198,7 @@ static int mlx5_eq_int(struct mlx5_core_dev *dev, struct mlx5_eq *eq)
int eqes_found = 0;
int set_ci = 0;
u32 cqn;
- u32 srqn;
+ u32 rsn;
u8 port;
while ((eqe = next_eqe_sw(eq))) {
@@ -224,18 +224,18 @@ static int mlx5_eq_int(struct mlx5_core_dev *dev, struct mlx5_eq *eq)
case MLX5_EVENT_TYPE_PATH_MIG_FAILED:
case MLX5_EVENT_TYPE_WQ_INVAL_REQ_ERROR:
case MLX5_EVENT_TYPE_WQ_ACCESS_ERROR:
+ rsn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
mlx5_core_dbg(dev, "event %s(%d) arrived\n",
eqe_type_str(eqe->type), eqe->type);
- mlx5_qp_event(dev, be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff,
- eqe->type);
+ mlx5_rsc_event(dev, rsn, eqe->type);
break;
case MLX5_EVENT_TYPE_SRQ_RQ_LIMIT:
case MLX5_EVENT_TYPE_SRQ_CATAS_ERROR:
- srqn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
+ rsn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
mlx5_core_dbg(dev, "SRQ event %s(%d): srqn 0x%x\n",
- eqe_type_str(eqe->type), eqe->type, srqn);
- mlx5_srq_event(dev, srqn, eqe->type);
+ eqe_type_str(eqe->type), eqe->type, rsn);
+ mlx5_srq_event(dev, rsn, eqe->type);
break;
case MLX5_EVENT_TYPE_CMD:
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/qp.c b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
index 415b67ce379e..5261a2b0da43 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
@@ -39,28 +39,53 @@
#include "mlx5_core.h"
-void mlx5_qp_event(struct mlx5_core_dev *dev, u32 qpn, int event_type)
+static struct mlx5_core_rsc_common *mlx5_get_rsc(struct mlx5_core_dev *dev,
+ u32 rsn)
{
struct mlx5_qp_table *table = &dev->priv.qp_table;
- struct mlx5_core_qp *qp;
+ struct mlx5_core_rsc_common *common;
spin_lock(&table->lock);
- qp = radix_tree_lookup(&table->tree, qpn);
- if (qp)
- atomic_inc(&qp->refcount);
+ common = radix_tree_lookup(&table->tree, rsn);
+ if (common)
+ atomic_inc(&common->refcount);
spin_unlock(&table->lock);
- if (!qp) {
- mlx5_core_warn(dev, "Async event for bogus QP 0x%x\n", qpn);
- return;
+ if (!common) {
+ mlx5_core_warn(dev, "Async event for bogus resource 0x%x\n",
+ rsn);
+ return NULL;
}
+ return common;
+}
- qp->event(qp, event_type);
+void mlx5_core_put_rsc(struct mlx5_core_rsc_common *common)
+{
+ if (atomic_dec_and_test(&common->refcount))
+ complete(&common->free);
+}
+
+void mlx5_rsc_event(struct mlx5_core_dev *dev, u32 rsn, int event_type)
+{
+ struct mlx5_core_rsc_common *common = mlx5_get_rsc(dev, rsn);
+ struct mlx5_core_qp *qp;
+
+ if (!common)
+ return;
+
+ switch (common->res) {
+ case MLX5_RES_QP:
+ qp = (struct mlx5_core_qp *)common;
+ qp->event(qp, event_type);
+ break;
+
+ default:
+ mlx5_core_warn(dev, "invalid resource type for 0x%x\n", rsn);
+ }
- if (atomic_dec_and_test(&qp->refcount))
- complete(&qp->free);
+ mlx5_core_put_rsc(common);
}
int mlx5_core_create_qp(struct mlx5_core_dev *dev,
@@ -92,6 +117,7 @@ int mlx5_core_create_qp(struct mlx5_core_dev *dev,
qp->qpn = be32_to_cpu(out.qpn) & 0xffffff;
mlx5_core_dbg(dev, "qpn = 0x%x\n", qp->qpn);
+ qp->common.res = MLX5_RES_QP;
spin_lock_irq(&table->lock);
err = radix_tree_insert(&table->tree, qp->qpn, qp);
spin_unlock_irq(&table->lock);
@@ -106,9 +132,9 @@ int mlx5_core_create_qp(struct mlx5_core_dev *dev,
qp->qpn);
qp->pid = current->pid;
- atomic_set(&qp->refcount, 1);
+ atomic_set(&qp->common.refcount, 1);
atomic_inc(&dev->num_qps);
- init_completion(&qp->free);
+ init_completion(&qp->common.free);
return 0;
@@ -138,9 +164,8 @@ int mlx5_core_destroy_qp(struct mlx5_core_dev *dev,
radix_tree_delete(&table->tree, qp->qpn);
spin_unlock_irqrestore(&table->lock, flags);
- if (atomic_dec_and_test(&qp->refcount))
- complete(&qp->free);
- wait_for_completion(&qp->free);
+ mlx5_core_put_rsc((struct mlx5_core_rsc_common *)qp);
+ wait_for_completion(&qp->common.free);
memset(&in, 0, sizeof(in));
memset(&out, 0, sizeof(out));
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index c439f9c59b93..246310dc8bef 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -375,6 +375,16 @@ struct mlx5_core_mr {
u32 pd;
};
+enum mlx5_res_type {
+ MLX5_RES_QP,
+};
+
+struct mlx5_core_rsc_common {
+ enum mlx5_res_type res;
+ atomic_t refcount;
+ struct completion free;
+};
+
struct mlx5_core_srq {
u32 srqn;
int max;
@@ -700,7 +710,7 @@ int mlx5_eq_init(struct mlx5_core_dev *dev);
void mlx5_eq_cleanup(struct mlx5_core_dev *dev);
void mlx5_fill_page_array(struct mlx5_buf *buf, __be64 *pas);
void mlx5_cq_completion(struct mlx5_core_dev *dev, u32 cqn);
-void mlx5_qp_event(struct mlx5_core_dev *dev, u32 qpn, int event_type);
+void mlx5_rsc_event(struct mlx5_core_dev *dev, u32 rsn, int event_type);
void mlx5_srq_event(struct mlx5_core_dev *dev, u32 srqn, int event_type);
struct mlx5_core_srq *mlx5_core_get_srq(struct mlx5_core_dev *dev, u32 srqn);
void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, unsigned long vector);
@@ -737,6 +747,7 @@ void mlx5_cmdif_debugfs_cleanup(struct mlx5_core_dev *dev);
int mlx5_core_create_psv(struct mlx5_core_dev *dev, u32 pdn,
int npsvs, u32 *sig_index);
int mlx5_core_destroy_psv(struct mlx5_core_dev *dev, int psv_num);
+void mlx5_core_put_rsc(struct mlx5_core_rsc_common *common);
static inline u32 mlx5_mkey_to_idx(u32 mkey)
{
diff --git a/include/linux/mlx5/qp.h b/include/linux/mlx5/qp.h
index 9709b30e2d69..7c4c0f1f5805 100644
--- a/include/linux/mlx5/qp.h
+++ b/include/linux/mlx5/qp.h
@@ -342,10 +342,9 @@ struct mlx5_stride_block_ctrl_seg {
};
struct mlx5_core_qp {
+ struct mlx5_core_rsc_common common; /* must be first */
void (*event) (struct mlx5_core_qp *, int);
int qpn;
- atomic_t refcount;
- struct completion free;
struct mlx5_rsc_debug *dbg;
int pid;
};
--
2.1.1
^ permalink raw reply related
* [PATCH V2 net-next 0/5] mlx5 update for 3.18
From: Eli Cohen @ 2014-10-02 9:19 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
Hi Dave,
This series integrates a new mechanism for populating and extracting field values
used in the driver/firmware interaction around command mailboxes.
Changes from V1:
- Remove unused definition of memcpy_cpu_to_be32()
- Remove definitions of non_existent_*() and use BUILD_BUG_ON() instead.
- Added a patch one line patch to add support for ConnectX-4 devices.
Changes from V0:
- trimmed the auto-generated file to a minimum, as required by the reviewers.
Eli
Eli Cohen (5):
net/mlx5_core: Update device capabilities handling
net/mlx5_core: Use hardware registers description header file
net/mlx5_core: use set/get macros in device caps
net/mlx5_core: Identify resources by their type
net/mlx5_core: Add ConnectX-4 to list of supported devices
drivers/infiniband/hw/mlx5/cq.c | 8 +-
drivers/infiniband/hw/mlx5/mad.c | 2 +-
drivers/infiniband/hw/mlx5/main.c | 83 +++---
drivers/infiniband/hw/mlx5/qp.c | 72 +++--
drivers/infiniband/hw/mlx5/srq.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 77 +++---
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 14 +-
drivers/net/ethernet/mellanox/mlx5/core/fw.c | 81 +-----
drivers/net/ethernet/mellanox/mlx5/core/main.c | 230 ++++++++++++----
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 60 +++--
drivers/net/ethernet/mellanox/mlx5/core/uar.c | 4 +-
include/linux/mlx5/device.h | 152 ++++-------
include/linux/mlx5/driver.h | 118 +++------
include/linux/mlx5/mlx5_ifc.h | 349 +++++++++++++++++++++++++
include/linux/mlx5/qp.h | 3 +-
15 files changed, 804 insertions(+), 455 deletions(-)
create mode 100644 include/linux/mlx5/mlx5_ifc.h
--
2.1.1
^ permalink raw reply
* RE: [PATCH net-next v6 2/2] bonding: Simplify the xmit function for modes that use xmit_hash
From: David Laight @ 2014-10-02 9:12 UTC (permalink / raw)
To: 'Mahesh Bandewar', Jay Vosburgh
Cc: Veaceslav Falico, Andy Gospodarek, David Miller, netdev,
Eric Dumazet, Maciej Zenczykowski
In-Reply-To: <CAF2d9jioDfj+qmvrQBQcaHnvXbL+Fnsh3-jxwn2KU+zvV=xgJQ@mail.gmail.com>
From: Mahesh Bandewar
> On Wed, Oct 1, 2014 at 9:49 PM, Jay Vosburgh <jay.vosburgh@canonical.com> wrote:
> > Mahesh Bandewar <maheshb@google.com> wrote:
...
> >> * Select aggregation groups, and assign each port for it's aggregetor. The
> >> * selection logic is called in the inititalization (after all the handshkes),
> >> * and after every lacpdu receive (if selected is off).
> >> */
> >>-static void ad_port_selection_logic(struct port *port)
> >>+static void ad_port_selection_logic(struct port *port, bool *update_slave_arr)
> >
> > Since this function is void, why not have it return a value
> > instead of the bool *update_slave_arr? That would eliminate the need
> > for some call sites to pass a "dummy" to the function. This comment
> > applies to ad_agg_selection_logic and ad_enable_collecting_distributing
> > as well.
> >
> Yes, I had similar discussion with Nik earlier and overloading the
> return value did not feel clean and future-proof and hence decided to
> take this approach.
What overload?
Returning values by reference parameters isn't really a good idea.
It kills performance and optimisations.
If you ever need a second return value then solve the problem then.
David
^ permalink raw reply
* Re: [PATCH 1/2] iproute2: Add support for IPv6 VTI tunnels to ip6tunnel
From: Jiri Pirko @ 2014-10-02 9:11 UTC (permalink / raw)
To: Steffen Klassert; +Cc: Stephen Hemminger, netdev
In-Reply-To: <20141002084820.GF6390@secunet.com>
Thu, Oct 02, 2014 at 10:48:20AM CEST, steffen.klassert@secunet.com wrote:
>On Thu, Oct 02, 2014 at 10:41:09AM +0200, Jiri Pirko wrote:
>> Fri, Sep 26, 2014 at 09:10:56AM CEST, steffen.klassert@secunet.com wrote:
>> >
>> >@@ -459,11 +462,14 @@ static int do_add(int cmd, int argc, char **argv)
>> > switch (p.proto) {
>> > case IPPROTO_IPIP:
>> > case IPPROTO_IPV6:
>> >+ if (p.i_flags != VTI_ISVTI)
>> >+ return tnl_add_ioctl(cmd, "ip6_vti0", p.name, &p);
>> ^ Wouldn't it be more
>> consistent to not to use
>> the underscore?
>
>The ipv4 version of vti uses ip_vti0, so I tried to be consistent
>with that.
Okay, fine with me.
^ permalink raw reply
* RE: [PATCHv9 net-next 2/4] sunvnet: make transmit path zero-copy in the kernel
From: David Laight @ 2014-10-02 9:06 UTC (permalink / raw)
To: 'Raghuram Kothakota', David L Stevens
Cc: David Miller, netdev@vger.kernel.org, Sowmini Varadhan
In-Reply-To: <77E6695E-58F8-4D10-8ED7-EEC4A487339E@oracle.com>
From: Raghuram Kothakota
> Sorry I am late in providing my comments, but I feel it is important
> to share this comment.
A comment on the original patch...
...
> > @@ -811,14 +930,27 @@ static int vnet_start_xmit(struct sk_buff *skb, struct net_device *dev)
> >
> > d = vio_dring_cur(dr);
> >
> > - tx_buf = port->tx_bufs[dr->prod].buf;
> > - skb_copy_from_linear_data(skb, tx_buf + VNET_PACKET_SKIP, skb->len);
> > + txi = dr->prod;
> > +
> > + freeskbs = vnet_clean_tx_ring(port, &pending);
> > +
> > + BUG_ON(port->tx_bufs[txi].skb);
> >
> > len = skb->len;
> > - if (len < ETH_ZLEN) {
> > + if (len < ETH_ZLEN)
> > len = ETH_ZLEN;
> > - memset(tx_buf+VNET_PACKET_SKIP+skb->len, 0, len - skb->len);
> > +
Aren't you transmitting 'random' bytes from the end of the data?
Plausibly they might not even be mapped.
Also, for short frames the copy could well be faster - especially on systems
with non-trivial iommu.
It is also worth checking whether the original copy was aligned.
David
^ permalink raw reply
* [PATCH net v2] r8152: autoresume before setting MAC address
From: Hayes Wang @ 2014-10-02 9:03 UTC (permalink / raw)
To: netdev; +Cc: nic_swsd, linux-kernel, linux-usb, Hayes Wang
In-Reply-To: <1394712342-15778-55-Taiwan-albertk@realtek.com>
Resume the device before setting the MAC address.
Signed-off-by: Hayes Wang <hayeswang@realtek.com>
---
drivers/net/usb/r8152.c | 11 +++++++++--
1 file changed, 9 insertions(+), 2 deletions(-)
diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c
index 604ef21..5ee7582 100644
--- a/drivers/net/usb/r8152.c
+++ b/drivers/net/usb/r8152.c
@@ -1007,9 +1007,14 @@ static int rtl8152_set_mac_address(struct net_device *netdev, void *p)
{
struct r8152 *tp = netdev_priv(netdev);
struct sockaddr *addr = p;
+ int ret = -EADDRNOTAVAIL;
if (!is_valid_ether_addr(addr->sa_data))
- return -EADDRNOTAVAIL;
+ goto out1;
+
+ ret = usb_autopm_get_interface(tp->intf);
+ if (ret < 0)
+ goto out1;
memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
@@ -1017,7 +1022,9 @@ static int rtl8152_set_mac_address(struct net_device *netdev, void *p)
pla_ocp_write(tp, PLA_IDR, BYTE_EN_SIX_BYTES, 8, addr->sa_data);
ocp_write_byte(tp, MCU_TYPE_PLA, PLA_CRWECR, CRWECR_NORAML);
- return 0;
+ usb_autopm_put_interface(tp->intf);
+out1:
+ return ret;
}
static void read_bulk_callback(struct urb *urb)
--
1.9.3
^ permalink raw reply related
* Re: [PATCH] net: ethernet: Remove superfluous ether_setup after alloc_etherdev
From: Nicolas Ferre @ 2014-10-02 8:59 UTC (permalink / raw)
To: Tobias Klauser, David S. Miller, netdev
Cc: Maxime Ripard, Vince Bridgers, Wan ZongShun, Nicolas Pitre,
Steve Glendinning, Giuseppe Cavallaro
In-Reply-To: <1412237730-5811-1-git-send-email-tklauser@distanz.ch>
On 02/10/2014 10:15, Tobias Klauser :
> There is no need to call ether_setup after alloc_ethdev since it was
> already called there.
>
> Follow commits c706471b2601 ("net: axienet: remove unnecessary
> ether_setup after alloc_etherdev") and 3c87dcbfb36c ("net: ll_temac:
> Remove unnecessary ether_setup after alloc_etherdev") and fix the
> pattern in all remaining ethernet drivers.
>
> Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
> ---
> drivers/net/ethernet/adi/bfin_mac.c | 3 ---
> drivers/net/ethernet/allwinner/sun4i-emac.c | 2 --
> drivers/net/ethernet/altera/altera_tse_main.c | 1 -
> drivers/net/ethernet/cadence/at91_ether.c | 1 -
If needed for this driver ^^
Acked-by: Nicolas Ferre <nicolas.ferre@atmel.com>
Thanks.
> drivers/net/ethernet/calxeda/xgmac.c | 1 -
> drivers/net/ethernet/davicom/dm9000.c | 3 ---
> drivers/net/ethernet/ethoc.c | 2 --
> drivers/net/ethernet/lantiq_etop.c | 1 -
> drivers/net/ethernet/moxa/moxart_ether.c | 1 -
> drivers/net/ethernet/netx-eth.c | 2 --
> drivers/net/ethernet/nuvoton/w90p910_ether.c | 1 -
> drivers/net/ethernet/nxp/lpc_eth.c | 3 ---
> drivers/net/ethernet/qualcomm/qca_spi.c | 2 --
> drivers/net/ethernet/smsc/smc911x.c | 3 ---
> drivers/net/ethernet/smsc/smc91x.c | 3 ---
> drivers/net/ethernet/smsc/smsc911x.c | 1 -
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 2 --
> drivers/net/ethernet/wiznet/w5100.c | 1 -
> drivers/net/ethernet/wiznet/w5300.c | 1 -
> 19 files changed, 34 deletions(-)
>
> diff --git a/drivers/net/ethernet/adi/bfin_mac.c b/drivers/net/ethernet/adi/bfin_mac.c
> index afa6684..8ed4d340 100644
> --- a/drivers/net/ethernet/adi/bfin_mac.c
> +++ b/drivers/net/ethernet/adi/bfin_mac.c
> @@ -1692,9 +1692,6 @@ static int bfin_mac_probe(struct platform_device *pdev)
> lp->vlan1_mask = ETH_P_8021Q | mii_bus_data->vlan1_mask;
> lp->vlan2_mask = ETH_P_8021Q | mii_bus_data->vlan2_mask;
>
> - /* Fill in the fields of the device structure with ethernet values. */
> - ether_setup(ndev);
> -
> ndev->netdev_ops = &bfin_mac_netdev_ops;
> ndev->ethtool_ops = &bfin_mac_ethtool_ops;
>
> diff --git a/drivers/net/ethernet/allwinner/sun4i-emac.c b/drivers/net/ethernet/allwinner/sun4i-emac.c
> index 29b9f08..1fcd556 100644
> --- a/drivers/net/ethernet/allwinner/sun4i-emac.c
> +++ b/drivers/net/ethernet/allwinner/sun4i-emac.c
> @@ -878,8 +878,6 @@ static int emac_probe(struct platform_device *pdev)
> emac_powerup(ndev);
> emac_reset(db);
>
> - ether_setup(ndev);
> -
> ndev->netdev_ops = &emac_netdev_ops;
> ndev->watchdog_timeo = msecs_to_jiffies(watchdog);
> ndev->ethtool_ops = &emac_ethtool_ops;
> diff --git a/drivers/net/ethernet/altera/altera_tse_main.c b/drivers/net/ethernet/altera/altera_tse_main.c
> index 7330681..fc2d555 100644
> --- a/drivers/net/ethernet/altera/altera_tse_main.c
> +++ b/drivers/net/ethernet/altera/altera_tse_main.c
> @@ -1433,7 +1433,6 @@ static int altera_tse_probe(struct platform_device *pdev)
> goto err_free_netdev;
>
> /* initialize netdev */
> - ether_setup(ndev);
> ndev->mem_start = control_port->start;
> ndev->mem_end = control_port->end;
> ndev->netdev_ops = &altera_tse_netdev_ops;
> diff --git a/drivers/net/ethernet/cadence/at91_ether.c b/drivers/net/ethernet/cadence/at91_ether.c
> index 4a79eda..4a24b9a 100644
> --- a/drivers/net/ethernet/cadence/at91_ether.c
> +++ b/drivers/net/ethernet/cadence/at91_ether.c
> @@ -351,7 +351,6 @@ static int __init at91ether_probe(struct platform_device *pdev)
> if (res)
> goto err_disable_clock;
>
> - ether_setup(dev);
> dev->netdev_ops = &at91ether_netdev_ops;
> dev->ethtool_ops = &macb_ethtool_ops;
> platform_set_drvdata(pdev, dev);
> diff --git a/drivers/net/ethernet/calxeda/xgmac.c b/drivers/net/ethernet/calxeda/xgmac.c
> index 25d6b2a..47bfea2 100644
> --- a/drivers/net/ethernet/calxeda/xgmac.c
> +++ b/drivers/net/ethernet/calxeda/xgmac.c
> @@ -1735,7 +1735,6 @@ static int xgmac_probe(struct platform_device *pdev)
> SET_NETDEV_DEV(ndev, &pdev->dev);
> priv = netdev_priv(ndev);
> platform_set_drvdata(pdev, ndev);
> - ether_setup(ndev);
> ndev->netdev_ops = &xgmac_netdev_ops;
> ndev->ethtool_ops = &xgmac_ethtool_ops;
> spin_lock_init(&priv->stats_lock);
> diff --git a/drivers/net/ethernet/davicom/dm9000.c b/drivers/net/ethernet/davicom/dm9000.c
> index 70089c2..f3ba840 100644
> --- a/drivers/net/ethernet/davicom/dm9000.c
> +++ b/drivers/net/ethernet/davicom/dm9000.c
> @@ -1613,9 +1613,6 @@ dm9000_probe(struct platform_device *pdev)
>
> /* from this point we assume that we have found a DM9000 */
>
> - /* driver system function */
> - ether_setup(ndev);
> -
> ndev->netdev_ops = &dm9000_netdev_ops;
> ndev->watchdog_timeo = msecs_to_jiffies(watchdog);
> ndev->ethtool_ops = &dm9000_ethtool_ops;
> diff --git a/drivers/net/ethernet/ethoc.c b/drivers/net/ethernet/ethoc.c
> index f3658bd..0bc6c10 100644
> --- a/drivers/net/ethernet/ethoc.c
> +++ b/drivers/net/ethernet/ethoc.c
> @@ -1222,8 +1222,6 @@ static int ethoc_probe(struct platform_device *pdev)
> goto error;
> }
>
> - ether_setup(netdev);
> -
> /* setup the net_device structure */
> netdev->netdev_ops = ðoc_netdev_ops;
> netdev->watchdog_timeo = ETHOC_TIMEOUT;
> diff --git a/drivers/net/ethernet/lantiq_etop.c b/drivers/net/ethernet/lantiq_etop.c
> index fd4b6ae..2dad4d5 100644
> --- a/drivers/net/ethernet/lantiq_etop.c
> +++ b/drivers/net/ethernet/lantiq_etop.c
> @@ -633,7 +633,6 @@ ltq_etop_init(struct net_device *dev)
> int err;
> bool random_mac = false;
>
> - ether_setup(dev);
> dev->watchdog_timeo = 10 * HZ;
> err = ltq_etop_hw_init(dev);
> if (err)
> diff --git a/drivers/net/ethernet/moxa/moxart_ether.c b/drivers/net/ethernet/moxa/moxart_ether.c
> index 2f12c88..bde1b70 100644
> --- a/drivers/net/ethernet/moxa/moxart_ether.c
> +++ b/drivers/net/ethernet/moxa/moxart_ether.c
> @@ -511,7 +511,6 @@ static int moxart_mac_probe(struct platform_device *pdev)
> goto init_fail;
> }
>
> - ether_setup(ndev);
> ndev->netdev_ops = &moxart_netdev_ops;
> netif_napi_add(ndev, &priv->napi, moxart_rx_poll, RX_DESC_NUM);
> ndev->priv_flags |= IFF_UNICAST_FLT;
> diff --git a/drivers/net/ethernet/netx-eth.c b/drivers/net/ethernet/netx-eth.c
> index 31eb911..8176c8a 100644
> --- a/drivers/net/ethernet/netx-eth.c
> +++ b/drivers/net/ethernet/netx-eth.c
> @@ -315,8 +315,6 @@ static int netx_eth_enable(struct net_device *ndev)
> unsigned int mac4321, mac65;
> int running, i;
>
> - ether_setup(ndev);
> -
> ndev->netdev_ops = &netx_eth_netdev_ops;
> ndev->watchdog_timeo = msecs_to_jiffies(5000);
>
> diff --git a/drivers/net/ethernet/nuvoton/w90p910_ether.c b/drivers/net/ethernet/nuvoton/w90p910_ether.c
> index 79645f7..379b7fb 100644
> --- a/drivers/net/ethernet/nuvoton/w90p910_ether.c
> +++ b/drivers/net/ethernet/nuvoton/w90p910_ether.c
> @@ -943,7 +943,6 @@ static int w90p910_ether_setup(struct net_device *dev)
> {
> struct w90p910_ether *ether = netdev_priv(dev);
>
> - ether_setup(dev);
> dev->netdev_ops = &w90p910_ether_netdev_ops;
> dev->ethtool_ops = &w90p910_ether_ethtool_ops;
>
> diff --git a/drivers/net/ethernet/nxp/lpc_eth.c b/drivers/net/ethernet/nxp/lpc_eth.c
> index a44a03c..66fd868 100644
> --- a/drivers/net/ethernet/nxp/lpc_eth.c
> +++ b/drivers/net/ethernet/nxp/lpc_eth.c
> @@ -1377,9 +1377,6 @@ static int lpc_eth_drv_probe(struct platform_device *pdev)
> goto err_out_iounmap;
> }
>
> - /* Fill in the fields of the device structure with ethernet values. */
> - ether_setup(ndev);
> -
> /* Setup driver functions */
> ndev->netdev_ops = &lpc_netdev_ops;
> ndev->ethtool_ops = &lpc_eth_ethtool_ops;
> diff --git a/drivers/net/ethernet/qualcomm/qca_spi.c b/drivers/net/ethernet/qualcomm/qca_spi.c
> index 74eb520..2c811f6 100644
> --- a/drivers/net/ethernet/qualcomm/qca_spi.c
> +++ b/drivers/net/ethernet/qualcomm/qca_spi.c
> @@ -810,8 +810,6 @@ qcaspi_netdev_setup(struct net_device *dev)
> {
> struct qcaspi *qca = NULL;
>
> - ether_setup(dev);
> -
> dev->netdev_ops = &qcaspi_netdev_ops;
> qcaspi_set_ethtool_ops(dev);
> dev->watchdog_timeo = QCASPI_TX_TIMEOUT;
> diff --git a/drivers/net/ethernet/smsc/smc911x.c b/drivers/net/ethernet/smsc/smc911x.c
> index 9778cba..e88df9c 100644
> --- a/drivers/net/ethernet/smsc/smc911x.c
> +++ b/drivers/net/ethernet/smsc/smc911x.c
> @@ -1927,9 +1927,6 @@ static int smc911x_probe(struct net_device *dev)
> }
> dev->irq = irq_canonicalize(dev->irq);
>
> - /* Fill in the fields of the device structure with ethernet values. */
> - ether_setup(dev);
> -
> dev->netdev_ops = &smc911x_netdev_ops;
> dev->watchdog_timeo = msecs_to_jiffies(watchdog);
> dev->ethtool_ops = &smc911x_ethtool_ops;
> diff --git a/drivers/net/ethernet/smsc/smc91x.c b/drivers/net/ethernet/smsc/smc91x.c
> index bcaa41a..5e94d00 100644
> --- a/drivers/net/ethernet/smsc/smc91x.c
> +++ b/drivers/net/ethernet/smsc/smc91x.c
> @@ -1967,9 +1967,6 @@ static int smc_probe(struct net_device *dev, void __iomem *ioaddr,
> }
> dev->irq = irq_canonicalize(dev->irq);
>
> - /* Fill in the fields of the device structure with ethernet values. */
> - ether_setup(dev);
> -
> dev->watchdog_timeo = msecs_to_jiffies(watchdog);
> dev->netdev_ops = &smc_netdev_ops;
> dev->ethtool_ops = &smc_ethtool_ops;
> diff --git a/drivers/net/ethernet/smsc/smsc911x.c b/drivers/net/ethernet/smsc/smsc911x.c
> index 5e13fa5..affb29d 100644
> --- a/drivers/net/ethernet/smsc/smsc911x.c
> +++ b/drivers/net/ethernet/smsc/smsc911x.c
> @@ -2255,7 +2255,6 @@ static int smsc911x_init(struct net_device *dev)
> if (smsc911x_soft_reset(pdata))
> return -ENODEV;
>
> - ether_setup(dev);
> dev->flags |= IFF_MULTICAST;
> netif_napi_add(dev, &pdata->napi, smsc911x_poll, SMSC_NAPI_WEIGHT);
> dev->netdev_ops = &smsc911x_netdev_ops;
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index 9979f67..6f77a46 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -2765,8 +2765,6 @@ struct stmmac_priv *stmmac_dvr_probe(struct device *device,
> priv->device = device;
> priv->dev = ndev;
>
> - ether_setup(ndev);
> -
> stmmac_set_ethtool_ops(ndev);
> priv->pause = pause;
> priv->plat = plat_dat;
> diff --git a/drivers/net/ethernet/wiznet/w5100.c b/drivers/net/ethernet/wiznet/w5100.c
> index 104d46f..0f56b1c 100644
> --- a/drivers/net/ethernet/wiznet/w5100.c
> +++ b/drivers/net/ethernet/wiznet/w5100.c
> @@ -708,7 +708,6 @@ static int w5100_probe(struct platform_device *pdev)
> priv = netdev_priv(ndev);
> priv->ndev = ndev;
>
> - ether_setup(ndev);
> ndev->netdev_ops = &w5100_netdev_ops;
> ndev->ethtool_ops = &w5100_ethtool_ops;
> ndev->watchdog_timeo = HZ;
> diff --git a/drivers/net/ethernet/wiznet/w5300.c b/drivers/net/ethernet/wiznet/w5300.c
> index 1f33c4c..f961f14 100644
> --- a/drivers/net/ethernet/wiznet/w5300.c
> +++ b/drivers/net/ethernet/wiznet/w5300.c
> @@ -620,7 +620,6 @@ static int w5300_probe(struct platform_device *pdev)
> priv = netdev_priv(ndev);
> priv->ndev = ndev;
>
> - ether_setup(ndev);
> ndev->netdev_ops = &w5300_netdev_ops;
> ndev->ethtool_ops = &w5300_ethtool_ops;
> ndev->watchdog_timeo = HZ;
>
--
Nicolas Ferre
^ permalink raw reply
* Re: [PATCH v2 net-next] mlx4: optimize xmit path
From: Amir Vadai @ 2014-10-02 8:57 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: Eric Dumazet, Or Gerlitz, Alexei Starovoitov, David S. Miller,
Eric Dumazet, John Fastabend, Linux Netdev List, Or Gerlitz,
amira, idos, Yevgeny Petrilin, eyalpe
In-Reply-To: <20141002102950.2b209373@redhat.com>
On 10/2/2014 11:29 AM, Jesper Dangaard Brouer wrote:
> On Thu, 2 Oct 2014 11:03:13 +0300
> Amir Vadai <amirv@mellanox.com> wrote:
>
>> I'm still working on it, but I can't reproduce the numbers that you
>> show. On my development machine, I get ~5.5Mpps with burst=8 and ~2Mpps
>> with burst=1.
>
> For pktgen it sounds like you either missing:
>
> 1) flag QUEUE_MAP_CPU
> and/or
> 2) clone_skb $BIGNUMBER
Thanks,
This improved the numbers to ~7Mpps with burst=8 and ~4Mpps with burst=1
- still far away from 15Mpps, but in the direction.
>
> My usage of pktgen:
> http://netoptimizer.blogspot.dk/2014/06/pktgen-for-network-overload-testing.html
>
^ permalink raw reply
* Re: [PATCH 1/2] iproute2: Add support for IPv6 VTI tunnels to ip6tunnel
From: Steffen Klassert @ 2014-10-02 8:48 UTC (permalink / raw)
To: Jiri Pirko; +Cc: Stephen Hemminger, netdev
In-Reply-To: <20141002084109.GI1844@nanopsycho.orion>
On Thu, Oct 02, 2014 at 10:41:09AM +0200, Jiri Pirko wrote:
> Fri, Sep 26, 2014 at 09:10:56AM CEST, steffen.klassert@secunet.com wrote:
> >
> >@@ -459,11 +462,14 @@ static int do_add(int cmd, int argc, char **argv)
> > switch (p.proto) {
> > case IPPROTO_IPIP:
> > case IPPROTO_IPV6:
> >+ if (p.i_flags != VTI_ISVTI)
> >+ return tnl_add_ioctl(cmd, "ip6_vti0", p.name, &p);
> ^ Wouldn't it be more
> consistent to not to use
> the underscore?
The ipv4 version of vti uses ip_vti0, so I tried to be consistent
with that.
^ permalink raw reply
* Re: [PATCH net] r8152: autoresume before setting MAC address
From: Oliver Neukum @ 2014-10-02 8:44 UTC (permalink / raw)
To: Hayes Wang
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, nic_swsd-Rasf1IRRPZFBDgjK7y7TUQ,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-usb-u79uwXL29TY76Z2rM5mHXA
In-Reply-To: <1394712342-15778-55-Taiwan-albertk-Rasf1IRRPZFBDgjK7y7TUQ@public.gmane.org>
On Thu, 2014-10-02 at 16:26 +0800, Hayes Wang wrote:
> Resume the device before setting the MAC address.
>
> Signed-off-by: Hayes Wang <hayeswang-Rasf1IRRPZFBDgjK7y7TUQ@public.gmane.org>
> ---
> drivers/net/usb/r8152.c | 16 +++++++++++++---
> 1 file changed, 13 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c
> index 604ef21..e823358 100644
> --- a/drivers/net/usb/r8152.c
> +++ b/drivers/net/usb/r8152.c
> @@ -1007,9 +1007,16 @@ static int rtl8152_set_mac_address(struct net_device *netdev, void *p)
> {
> struct r8152 *tp = netdev_priv(netdev);
> struct sockaddr *addr = p;
> + int ret;
> +
> + ret = usb_autopm_get_interface(tp->intf);
> + if (ret < 0)
> + goto out1;
>
> - if (!is_valid_ether_addr(addr->sa_data))
> - return -EADDRNOTAVAIL;
It would be more legant to check validity before you do
the pm operation. Error handling would be simpler.
Regards
Oliver
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH 1/2] iproute2: Add support for IPv6 VTI tunnels to ip6tunnel
From: Jiri Pirko @ 2014-10-02 8:41 UTC (permalink / raw)
To: Steffen Klassert; +Cc: Stephen Hemminger, netdev
In-Reply-To: <20140926071056.GB6390@secunet.com>
Fri, Sep 26, 2014 at 09:10:56AM CEST, steffen.klassert@secunet.com wrote:
>
>Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
>---
> ip/ip6tunnel.c | 15 ++++++++++++---
> 1 file changed, 12 insertions(+), 3 deletions(-)
>
>diff --git a/ip/ip6tunnel.c b/ip/ip6tunnel.c
>index 66f9f2a..4b73ec6 100644
>--- a/ip/ip6tunnel.c
>+++ b/ip/ip6tunnel.c
>@@ -47,7 +47,7 @@ static void usage(void) __attribute__((noreturn));
> static void usage(void)
> {
> fprintf(stderr, "Usage: ip -f inet6 tunnel { add | change | del | show } [ NAME ]\n");
>- fprintf(stderr, " [ mode { ip6ip6 | ipip6 | ip6gre | any } ]\n");
>+ fprintf(stderr, " [ mode { ip6ip6 | ipip6 | ip6gre | vti6 | any } ]\n");
> fprintf(stderr, " [ remote ADDR local ADDR ] [ dev PHYS_DEV ]\n");
> fprintf(stderr, " [ encaplimit ELIM ]\n");
> fprintf(stderr ," [ hoplimit TTL ] [ tclass TCLASS ] [ flowlabel FLOWLABEL ]\n");
>@@ -140,7 +140,10 @@ static int parse_args(int argc, char **argv, int cmd, struct ip6_tnl_parm2 *p)
> if (strcmp(*argv, "ipv6/ipv6") == 0 ||
> strcmp(*argv, "ip6ip6") == 0)
> p->proto = IPPROTO_IPV6;
>- else if (strcmp(*argv, "ip/ipv6") == 0 ||
>+ else if (strcmp(*argv, "vti6") == 0) {
>+ p->proto = IPPROTO_IPV6;
>+ p->i_flags |= VTI_ISVTI;
>+ } else if (strcmp(*argv, "ip/ipv6") == 0 ||
> strcmp(*argv, "ipv4/ipv6") == 0 ||
> strcmp(*argv, "ipip6") == 0 ||
> strcmp(*argv, "ip4ip6") == 0)
>@@ -459,11 +462,14 @@ static int do_add(int cmd, int argc, char **argv)
> switch (p.proto) {
> case IPPROTO_IPIP:
> case IPPROTO_IPV6:
>+ if (p.i_flags != VTI_ISVTI)
>+ return tnl_add_ioctl(cmd, "ip6_vti0", p.name, &p);
^ Wouldn't it be more
consistent to not to use
the underscore?
>+ else
> return tnl_add_ioctl(cmd, "ip6tnl0", p.name, &p);
> case IPPROTO_GRE:
> return tnl_add_ioctl(cmd, "ip6gre0", p.name, &p);
> default:
>- fprintf(stderr, "cannot determine tunnel mode (ip6ip6, ipip6 or gre)\n");
>+ fprintf(stderr, "cannot determine tunnel mode (ip6ip6, ipip6, vti6 or gre)\n");
> }
> return -1;
> }
>@@ -480,6 +486,9 @@ static int do_del(int argc, char **argv)
> switch (p.proto) {
> case IPPROTO_IPIP:
> case IPPROTO_IPV6:
>+ if (p.i_flags != VTI_ISVTI)
>+ return tnl_del_ioctl("ip6_vti0", p.name, &p);
>+ else
> return tnl_del_ioctl("ip6tnl0", p.name, &p);
> case IPPROTO_GRE:
> return tnl_del_ioctl("ip6gre0", p.name, &p);
>--
>1.9.1
>
>--
>To unsubscribe from this list: send the line "unsubscribe netdev" in
>the body of a message to majordomo@vger.kernel.org
>More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH iproute2] tests: Dont fail cls-testbed.t if tests/cls dir is empty
From: Vadim Kochan @ 2014-10-02 8:35 UTC (permalink / raw)
To: netdev@vger.kernel.org, Stephen Hemminger; +Cc: Vadim Kochan
In-Reply-To: <1412031436-14492-1-git-send-email-vadim4j@gmail.com>
Hi,
Please ignore this patch, I will send modified one with another sibject.
On Tue, Sep 30, 2014 at 1:57 AM, Vadim Kochan <vadim4j@gmail.com> wrote:
> Curently tests/cls-testbed.t tries to run any *.t in
> tests/cls/ folder but such folder does not exist.
>
> The better solution can be like having a generic mechanism which
> allows to SKIP some tests (by return code?).
>
> Signed-off-by: Vadim Kochan <vadim4j@gmail.com>
> ---
> testsuite/tests/cls-testbed.t | 5 +++++
> 1 file changed, 5 insertions(+)
>
> diff --git a/testsuite/tests/cls-testbed.t b/testsuite/tests/cls-testbed.t
> index efae2a5..be32d79 100755
> --- a/testsuite/tests/cls-testbed.t
> +++ b/testsuite/tests/cls-testbed.t
> @@ -5,6 +5,11 @@ source lib/generic.sh
>
> QDISCS="cbq htb dsmark"
>
> +if [ ! -d tests/cls ]; then
> + ts_log "tests/cls folder is empty"
> + exit 0
> +fi
> +
> for q in ${QDISCS}; do
> ts_log "Preparing classifier testbed with qdisc $q"
>
> --
> 2.1.0
>
^ permalink raw reply
* Re: [PATCH] net: dsa: Fix build warning for !PM_SLEEP
From: Thierry Reding @ 2014-10-02 8:35 UTC (permalink / raw)
To: David Miller; +Cc: f.fainelli, netdev, linux-kernel
In-Reply-To: <20141001.152452.1973501659924138094.davem@davemloft.net>
[-- Attachment #1: Type: text/plain, Size: 992 bytes --]
On Wed, Oct 01, 2014 at 03:24:52PM -0400, David Miller wrote:
> From: Thierry Reding <thierry.reding@gmail.com>
> Date: Wed, 1 Oct 2014 13:59:00 +0200
>
> > From: Thierry Reding <treding@nvidia.com>
> >
> > The dsa_switch_suspend() and dsa_switch_resume() functions are only used
> > when PM_SLEEP is enabled, so they need #ifdef CONFIG_PM_SLEEP protection
> > to avoid a compiler warning.
> >
> > Signed-off-by: Thierry Reding <treding@nvidia.com>
>
> Applied.
>
> Please, in the future, explicitly indicate what tree your changes
> are against. Through trial and error I figured out that this could
> only apply to net-next, but that's not a good use of my time.
I had assumed that working on top of linux-next is what most people were
doing anyway therefore net-next would be the default to apply patches to
unless otherwise stated.
But if this inconveniences you I'll try to remember to explicitly state
which tree it's based on in the future.
Thierry
[-- Attachment #2: Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH v2 net-next] mlx4: optimize xmit path
From: Jesper Dangaard Brouer @ 2014-10-02 8:29 UTC (permalink / raw)
To: Amir Vadai
Cc: brouer, Eric Dumazet, Or Gerlitz, Alexei Starovoitov,
David S. Miller, Eric Dumazet, John Fastabend, Linux Netdev List,
Or Gerlitz, amira, idos, Yevgeny Petrilin, eyalpe
In-Reply-To: <542D06C1.6090802@mellanox.com>
On Thu, 2 Oct 2014 11:03:13 +0300
Amir Vadai <amirv@mellanox.com> wrote:
> I'm still working on it, but I can't reproduce the numbers that you
> show. On my development machine, I get ~5.5Mpps with burst=8 and ~2Mpps
> with burst=1.
For pktgen it sounds like you either missing:
1) flag QUEUE_MAP_CPU
and/or
2) clone_skb $BIGNUMBER
My usage of pktgen:
http://netoptimizer.blogspot.dk/2014/06/pktgen-for-network-overload-testing.html
--
Best regards,
Jesper Dangaard Brouer
MSc.CS, Sr. Network Kernel Developer at Red Hat
Author of http://www.iptv-analyzer.org
LinkedIn: http://www.linkedin.com/in/brouer
^ permalink raw reply
* [PATCH] Don't reset PHY on if_up for ASIX 88772
From: Michel Stam @ 2014-10-02 8:22 UTC (permalink / raw)
To: netdev; +Cc: freddy, grundler, Michel Stam
I've noticed every time the interface is set to 'up,', the kernel
reports that the link speed is set to 100 Mbps/Full Duplex, even
when ethtool is used to set autonegotiation to 'off', half
duplex, 10 Mbps.
It can be tested by:
ifconfig eth0 down
ethtool -s eth0 autoneg off speed 10 duplex half
ifconfig eth0 up
Then checking 'dmesg' for the link speed.
Signed-off-by: Michel Stam <m.stam@fugro.nl>
---
drivers/net/usb/asix_devices.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/usb/asix_devices.c b/drivers/net/usb/asix_devices.c
index 5d19409..2c05f6c 100644
--- a/drivers/net/usb/asix_devices.c
+++ b/drivers/net/usb/asix_devices.c
@@ -890,7 +890,7 @@ static const struct driver_info ax88772_info = {
.unbind = ax88772_unbind,
.status = asix_status,
.link_reset = ax88772_link_reset,
- .reset = ax88772_reset,
+ .reset = ax88772_link_reset,
.flags = FLAG_ETHER | FLAG_FRAMING_AX | FLAG_LINK_INTR | FLAG_MULTI_PACKET,
.rx_fixup = asix_rx_fixup_common,
.tx_fixup = asix_tx_fixup,
--
1.7.12.1
^ permalink raw reply related
* [PATCH net] r8152: autoresume before setting MAC address
From: Hayes Wang @ 2014-10-02 8:26 UTC (permalink / raw)
To: netdev-u79uwXL29TY76Z2rM5mHXA
Cc: nic_swsd-Rasf1IRRPZFBDgjK7y7TUQ,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
linux-usb-u79uwXL29TY76Z2rM5mHXA, Hayes Wang
Resume the device before setting the MAC address.
Signed-off-by: Hayes Wang <hayeswang-Rasf1IRRPZFBDgjK7y7TUQ@public.gmane.org>
---
drivers/net/usb/r8152.c | 16 +++++++++++++---
1 file changed, 13 insertions(+), 3 deletions(-)
diff --git a/drivers/net/usb/r8152.c b/drivers/net/usb/r8152.c
index 604ef21..e823358 100644
--- a/drivers/net/usb/r8152.c
+++ b/drivers/net/usb/r8152.c
@@ -1007,9 +1007,16 @@ static int rtl8152_set_mac_address(struct net_device *netdev, void *p)
{
struct r8152 *tp = netdev_priv(netdev);
struct sockaddr *addr = p;
+ int ret;
+
+ ret = usb_autopm_get_interface(tp->intf);
+ if (ret < 0)
+ goto out1;
- if (!is_valid_ether_addr(addr->sa_data))
- return -EADDRNOTAVAIL;
+ if (!is_valid_ether_addr(addr->sa_data)) {
+ ret = -EADDRNOTAVAIL;
+ goto out2;
+ }
memcpy(netdev->dev_addr, addr->sa_data, netdev->addr_len);
@@ -1017,7 +1024,10 @@ static int rtl8152_set_mac_address(struct net_device *netdev, void *p)
pla_ocp_write(tp, PLA_IDR, BYTE_EN_SIX_BYTES, 8, addr->sa_data);
ocp_write_byte(tp, MCU_TYPE_PLA, PLA_CRWECR, CRWECR_NORAML);
- return 0;
+out2:
+ usb_autopm_put_interface(tp->intf);
+out1:
+ return ret;
}
static void read_bulk_callback(struct urb *urb)
--
1.9.3
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply related
* Re: [PATCH] net: ethernet: Remove superfluous ether_setup after alloc_etherdev
From: Tobias Klauser @ 2014-10-02 8:20 UTC (permalink / raw)
To: David S. Miller, netdev
Cc: Maxime Ripard, Vince Bridgers, Nicolas Ferre, Wan ZongShun,
Nicolas Pitre, Steve Glendinning, Giuseppe Cavallaro
In-Reply-To: <1412237730-5811-1-git-send-email-tklauser@distanz.ch>
On 2014-10-02 at 10:15:30 +0200, Tobias Klauser <tklauser@distanz.ch> wrote:
> There is no need to call ether_setup after alloc_ethdev since it was
> already called there.
>
> Follow commits c706471b2601 ("net: axienet: remove unnecessary
> ether_setup after alloc_etherdev") and 3c87dcbfb36c ("net: ll_temac:
> Remove unnecessary ether_setup after alloc_etherdev") and fix the
> pattern in all remaining ethernet drivers.
>
> Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
This patch is for net-next. Sorry for not mentioning this in the subject
right away.
^ permalink raw reply
* Re: ASIX 88772
From: Christian Riesch @ 2014-10-02 8:16 UTC (permalink / raw)
To: Stam, Michel [FINT]
Cc: netdev@vger.kernel.org, Grant Grundler, Freddy Xin, olofj
In-Reply-To: <C89EFD3CD56F64468D3D206D683A8D2203864153@ldam-msx2.fugro-nl.local>
Hi Michel,
On Thu, Oct 2, 2014 at 9:50 AM, Stam, Michel [FINT] <M.Stam@fugro.nl> wrote:
> LS,
>
> Is it possible for someone to apply this patch?
Patches must be sent in the email body, not as an attachment. Please
use git send-email to send the patch, other mail clients will most
likely break the patch as you already noted. Please send the patch
again.
You should also take the authors of the patch that made the change
(4ad1438f) that you would like to revert on cc, Freddy Xin
<freddy@asix.com.tw> and Grant Grundler <grundler@chromium.org> in
this case.
Regards,
Christian
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox