From: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
To: linux-kernel@vger.kernel.org
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
stable@vger.kernel.org, Achiad Shochat <achiad@mellanox.com>,
Majd Dibbiny <majd@mellanox.com>, Moni Shoua <monis@mellanox.com>,
Leon Romanovsky <leon@kernel.org>,
Yuval Shaia <yuval.shaia@oracle.com>,
Doug Ledford <dledford@redhat.com>,
Sasha Levin <alexander.levin@verizon.com>
Subject: [PATCH 4.9 21/67] IB/mlx5: Assign DSCP for R-RoCE QPs Address Path
Date: Mon, 6 Nov 2017 10:43:44 +0100 [thread overview]
Message-ID: <20171106091306.231858745@linuxfoundation.org> (raw)
In-Reply-To: <20171106091305.401025609@linuxfoundation.org>
4.9-stable review patch. If anyone has any objections, please let me know.
------------------
From: Majd Dibbiny <majd@mellanox.com>
[ Upstream commit ed88451e1f2d400fd6a743d0a481631cf9f97550 ]
For Routable RoCE QPs, the DSCP should be set in the QP's
address path.
The DSCP's value is derived from the traffic class.
Fixes: 2811ba51b049 ("IB/mlx5: Add RoCE fields to Address Vector")
Cc: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Majd Dibbiny <majd@mellanox.com>
Reviewed-by: Moni Shoua <monis@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Reviewed-by: Yuval Shaia <yuval.shaia@oracle.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
Signed-off-by: Sasha Levin <alexander.levin@verizon.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
drivers/infiniband/hw/mlx5/main.c | 21 +++++++++++++++++++++
drivers/infiniband/hw/mlx5/mlx5_ib.h | 2 ++
drivers/infiniband/hw/mlx5/qp.c | 7 +++++++
3 files changed, 30 insertions(+)
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -325,6 +325,27 @@ __be16 mlx5_get_roce_udp_sport(struct ml
return cpu_to_be16(MLX5_CAP_ROCE(dev->mdev, r_roce_min_src_udp_port));
}
+int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num,
+ int index, enum ib_gid_type *gid_type)
+{
+ struct ib_gid_attr attr;
+ union ib_gid gid;
+ int ret;
+
+ ret = ib_get_cached_gid(&dev->ib_dev, port_num, index, &gid, &attr);
+ if (ret)
+ return ret;
+
+ if (!attr.ndev)
+ return -ENODEV;
+
+ dev_put(attr.ndev);
+
+ *gid_type = attr.gid_type;
+
+ return 0;
+}
+
static int mlx5_use_mad_ifc(struct mlx5_ib_dev *dev)
{
if (MLX5_CAP_GEN(dev->mdev, port_type) == MLX5_CAP_PORT_TYPE_IB)
--- a/drivers/infiniband/hw/mlx5/mlx5_ib.h
+++ b/drivers/infiniband/hw/mlx5/mlx5_ib.h
@@ -892,6 +892,8 @@ int mlx5_ib_set_vf_guid(struct ib_device
__be16 mlx5_get_roce_udp_sport(struct mlx5_ib_dev *dev, u8 port_num,
int index);
+int mlx5_get_roce_gid_type(struct mlx5_ib_dev *dev, u8 port_num,
+ int index, enum ib_gid_type *gid_type);
/* GSI QP helper functions */
struct ib_qp *mlx5_ib_gsi_create_qp(struct ib_pd *pd,
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -2226,6 +2226,7 @@ static int mlx5_set_path(struct mlx5_ib_
{
enum rdma_link_layer ll = rdma_port_get_link_layer(&dev->ib_dev, port);
int err;
+ enum ib_gid_type gid_type;
if (attr_mask & IB_QP_PKEY_INDEX)
path->pkey_index = cpu_to_be16(alt ? attr->alt_pkey_index :
@@ -2244,10 +2245,16 @@ static int mlx5_set_path(struct mlx5_ib_
if (ll == IB_LINK_LAYER_ETHERNET) {
if (!(ah->ah_flags & IB_AH_GRH))
return -EINVAL;
+ err = mlx5_get_roce_gid_type(dev, port, ah->grh.sgid_index,
+ &gid_type);
+ if (err)
+ return err;
memcpy(path->rmac, ah->dmac, sizeof(ah->dmac));
path->udp_sport = mlx5_get_roce_udp_sport(dev, port,
ah->grh.sgid_index);
path->dci_cfi_prio_sl = (ah->sl & 0x7) << 4;
+ if (gid_type == IB_GID_TYPE_ROCE_UDP_ENCAP)
+ path->ecn_dscp = (ah->grh.traffic_class >> 2) & 0x3f;
} else {
path->fl_free_ar = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0;
path->fl_free_ar |=
next prev parent reply other threads:[~2017-11-06 9:46 UTC|newest]
Thread overview: 70+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-06 9:43 [PATCH 4.9 00/67] 4.9.61-stable review Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 01/67] ALSA: timer: Add missing mutex lock for compat ioctls Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 02/67] ALSA: seq: Fix nested rwsem annotation for lockdep splat Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 03/67] cifs: check MaxPathNameComponentLength != 0 before using it Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 04/67] KEYS: return full count in keyring_read() if buffer is too small Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 05/67] KEYS: fix out-of-bounds read during ASN.1 parsing Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 06/67] ASoC: adau17x1: Workaround for noise bug in ADC Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 07/67] arm64: ensure __dump_instr() checks addr_limit Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 08/67] arm/arm64: KVM: set right LR register value for 32 bit guest when inject abort Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 09/67] arm/arm64: kvm: Disable branch profiling in HYP code Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 10/67] ARM: 8715/1: add a private asm/unaligned.h Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 11/67] drm/amdgpu: return -ENOENT from uvd 6.0 early init for harvesting Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 12/67] ocfs2: fstrim: Fix start offset of first cluster group during fstrim Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 14/67] drm/msm: Fix potential buffer overflow issue Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 15/67] drm/msm: fix an integer overflow test Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 16/67] tracing/samples: Fix creation and deletion of simple_thread_fn creation Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 17/67] Fix tracing sample code warning Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 18/67] cpufreq: Do not clear real_cpus mask on policy init Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 19/67] crypto: ccp - Set the AES size field for all modes Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 20/67] staging: fsl-mc: Add missing header Greg Kroah-Hartman
2017-11-06 9:43 ` Greg Kroah-Hartman [this message]
2017-11-06 9:43 ` [PATCH 4.9 22/67] PM / wakeirq: report a wakeup_event on dedicated wekup irq Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 23/67] scsi: megaraid_sas: Do not set fp_possible if TM capable for non-RW syspdIO, change fp_possible to bool Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 24/67] mmc: s3cmci: include linux/interrupt.h for tasklet_struct Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 25/67] mfd: ab8500-sysctrl: Handle probe deferral Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 26/67] mfd: axp20x: Fix axp288 PEK_DBR and PEK_DBF irqs being swapped Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 27/67] bnxt_en: Added PCI IDs for BCM57452 and BCM57454 ASICs Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 28/67] staging: rtl8712u: Fix endian settings for structs describing network packets Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 29/67] PCI/MSI: Return failure when msix_setup_entries() fails Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 30/67] net: mvneta: fix build errors when linux/phy*.h is removed from net/dsa.h Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 31/67] ext4: fix stripe-unaligned allocations Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 32/67] ext4: do not use stripe_width if it is not set Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 33/67] net/ena: change drivers default timeouts Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 34/67] i2c: riic: correctly finish transfers Greg Kroah-Hartman
2017-11-06 9:43 ` [PATCH 4.9 36/67] perf tools: Only increase index if perf_evsel__new_idx() succeeds Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 37/67] iwlwifi: mvm: use the PROBE_RESP_QUEUE to send deauth to unknown station Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 38/67] drm/fsl-dcu: check for clk_prepare_enable() error Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 39/67] clocksource/drivers/arm_arch_timer: Add dt binding for hisilicon-161010101 erratum Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 40/67] net: phy: dp83867: Recover from "port mirroring" N/A MODE4 Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 41/67] cx231xx: Fix I2C on Internal Master 3 Bus Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 42/67] ath10k: fix reading sram contents for QCA4019 Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 43/67] clk: sunxi-ng: Check kzalloc() for errors and cleanup error path Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 44/67] drm/msm/dsi: Set msm_dsi->encoders before initializing bridge Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 45/67] mtd: nand: sunxi: Fix the non-polling case in sunxi_nfc_wait_events() Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 46/67] gpio: mcp23s08: Select REGMAP/REGMAP_I2C to fix build error Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 47/67] xen/manage: correct return value check on xenbus_scanf() Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 48/67] scsi: aacraid: Process Error for response I/O Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 49/67] platform/x86: intel_mid_thermal: Fix module autoload Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 50/67] staging: lustre: llite: dont invoke direct_IO for the EOF case Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 51/67] staging: lustre: hsm: stack overrun in hai_dump_data_field Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 52/67] staging: lustre: ptlrpc: skip lock if export failed Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 53/67] staging: lustre: lmv: Error not handled for lmv_find_target Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 55/67] vfs: open() with O_CREAT should not create inodes with unknown ids Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 56/67] ASoC: Intel: boards: remove .pm_ops in all Atom/DPCM machine drivers Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 57/67] exynos4-is: fimc-is: Unmap region obtained by of_iomap() Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 58/67] mei: return error on notification request to a disconnected client Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 59/67] s390/dasd: check for device error pointer within state change interrupts Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 60/67] s390/prng: Adjust generation of entropy to produce real 256 bits Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 61/67] s390/crypto: Extend key length check for AES-XTS in fips mode Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 62/67] bt8xx: fix memory leak Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 63/67] drm/exynos: g2d: prevent integer overflow in Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 64/67] PCI: Avoid possible deadlock on pci_lock and p->pi_lock Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 65/67] powerpc/64: Dont try to use radix MMU under a hypervisor Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 66/67] xen: dont print error message in case of missing Xenstore entry Greg Kroah-Hartman
2017-11-06 9:44 ` [PATCH 4.9 67/67] staging: r8712u: Fix Sparse warning in rtl871x_xmit.c Greg Kroah-Hartman
2017-11-06 21:17 ` [PATCH 4.9 00/67] 4.9.61-stable review Guenter Roeck
2017-11-06 22:22 ` Shuah Khan
2017-11-06 23:13 ` Shuah Khan
2017-11-07 22:49 ` Tom Gall
2017-11-08 9:21 ` Greg Kroah-Hartman
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20171106091306.231858745@linuxfoundation.org \
--to=gregkh@linuxfoundation.org \
--cc=achiad@mellanox.com \
--cc=alexander.levin@verizon.com \
--cc=dledford@redhat.com \
--cc=leon@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=majd@mellanox.com \
--cc=monis@mellanox.com \
--cc=stable@vger.kernel.org \
--cc=yuval.shaia@oracle.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).