Merge tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma

Pull rdma updates from Jason Gunthorpe:
 "This has been a smaller cycle than normal. One new driver was
  accepted, which is unusual, and at least one more driver remains in
  review on the list.

  Summary:

   - Driver fixes for hns, hfi1, nes, rxe, i40iw, mlx5, cxgb4,
     vmw_pvrdma

   - Many patches from MatthewW converting radix tree and IDR users to
     use xarray

   - Introduction of tracepoints to the MAD layer

   - Build large SGLs at the start for DMA mapping and get the driver to
     split them

   - Generally clean SGL handling code throughout the subsystem

   - Support for restricting RDMA devices to net namespaces for
     containers

   - Progress to remove object allocation boilerplate code from drivers

   - Change in how the mlx5 driver shows representor ports linked to VFs

   - mlx5 uapi feature to access the on chip SW ICM memory

   - Add a new driver for 'EFA'. This is HW that supports user space
     packet processing through QPs in Amazon's cloud"

* tag 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/rdma/rdma: (186 commits)
  RDMA/ipoib: Allow user space differentiate between valid dev_port
  IB/core, ipoib: Do not overreact to SM LID change event
  RDMA/device: Don't fire uevent before device is fully initialized
  lib/scatterlist: Remove leftover from sg_page_iter comment
  RDMA/efa: Add driver to Kconfig/Makefile
  RDMA/efa: Add the efa module
  RDMA/efa: Add EFA verbs implementation
  RDMA/efa: Add common command handlers
  RDMA/efa: Implement functions that submit and complete admin commands
  RDMA/efa: Add the ABI definitions
  RDMA/efa: Add the com service API definitions
  RDMA/efa: Add the efa_com.h file
  RDMA/efa: Add the efa.h header file
  RDMA/efa: Add EFA device definitions
  RDMA: Add EFA related definitions
  RDMA/umem: Remove hugetlb flag
  RDMA/bnxt_re: Use core helpers to get aligned DMA address
  RDMA/i40iw: Use core helpers to get aligned DMA address within a supported page size
  RDMA/verbs: Add a DMA iterator to return aligned contiguous memory blocks
  RDMA/umem: Add API to find best driver supported page size in an MR
  ...

3 files changed, 269 insertions, 0 deletions

diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile
index 65e667bdf979..4f0a1cdbfe7c 100644
--- a/samples/bpf/Makefile
+++ b/samples/bpf/Makefile
@@ -52,6 +52,7 @@ hostprogs-y += xdpsock
 hostprogs-y += xdp_fwd
 hostprogs-y += task_fd_query
 hostprogs-y += xdp_sample_pkts
+hostprogs-y += ibumad
 hostprogs-y += hbm
 
 # Libbpf dependencies
@@ -108,6 +109,7 @@ xdpsock-objs := xdpsock_user.o
 xdp_fwd-objs := xdp_fwd_user.o
 task_fd_query-objs := bpf_load.o task_fd_query_user.o $(TRACE_HELPERS)
 xdp_sample_pkts-objs := xdp_sample_pkts_user.o $(TRACE_HELPERS)
+ibumad-objs := bpf_load.o ibumad_user.o $(TRACE_HELPERS)
 hbm-objs := bpf_load.o hbm.o $(CGROUP_HELPERS)
 
 # Tell kbuild to always build the programs
@@ -166,6 +168,7 @@ always += xdp_adjust_tail_kern.o
 always += xdp_fwd_kern.o
 always += task_fd_query_kern.o
 always += xdp_sample_pkts_kern.o
+always += ibumad_kern.o
 always += hbm_out_kern.o
 
 KBUILD_HOSTCFLAGS += -I$(objtree)/usr/include
diff --git a/samples/bpf/ibumad_kern.c b/samples/bpf/ibumad_kern.c
new file mode 100644
index 000000000000..38b2b3f22049
--- /dev/null
+++ b/samples/bpf/ibumad_kern.c
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+
+/**
+ * ibumad BPF sample kernel side
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * Copyright(c) 2018 Ira Weiny, Intel Corporation
+ */
+
+#define KBUILD_MODNAME "ibumad_count_pkts_by_class"
+#include <uapi/linux/bpf.h>
+
+#include "bpf_helpers.h"
+
+
+struct bpf_map_def SEC("maps") read_count = {
+        .type        = BPF_MAP_TYPE_ARRAY,
+        .key_size    = sizeof(u32), /* class; u32 required */
+        .value_size  = sizeof(u64), /* count of mads read */
+        .max_entries = 256, /* Room for all Classes */
+};
+
+struct bpf_map_def SEC("maps") write_count = {
+        .type        = BPF_MAP_TYPE_ARRAY,
+        .key_size    = sizeof(u32), /* class; u32 required */
+        .value_size  = sizeof(u64), /* count of mads written */
+        .max_entries = 256, /* Room for all Classes */
+};
+
+#undef DEBUG
+#ifdef DEBUG
+#define bpf_debug(fmt, ...)                         \
+({                                                  \
+        char ____fmt[] = fmt;                       \
+        bpf_trace_printk(____fmt, sizeof(____fmt),  \
+                         ##__VA_ARGS__);            \
+})
+#else
+#define bpf_debug(fmt, ...)
+#endif
+
+/* Taken from the current format defined in
+ * include/trace/events/ib_umad.h
+ * and
+ * /sys/kernel/debug/tracing/events/ib_umad/ib_umad_read/format
+ * /sys/kernel/debug/tracing/events/ib_umad/ib_umad_write/format
+ */
+struct ib_umad_rw_args {
+        u64 pad;
+        u8 port_num;
+        u8 sl;
+        u8 path_bits;
+        u8 grh_present;
+        u32 id;
+        u32 status;
+        u32 timeout_ms;
+        u32 retires;
+        u32 length;
+        u32 qpn;
+        u32 qkey;
+        u8 gid_index;
+        u8 hop_limit;
+        u16 lid;
+        u16 attr_id;
+        u16 pkey_index;
+        u8 base_version;
+        u8 mgmt_class;
+        u8 class_version;
+        u8 method;
+        u32 flow_label;
+        u16 mad_status;
+        u16 class_specific;
+        u32 attr_mod;
+        u64 tid;
+        u8 gid[16];
+        u32 dev_index;
+        u8 traffic_class;
+};
+
+SEC("tracepoint/ib_umad/ib_umad_read_recv")
+int on_ib_umad_read_recv(struct ib_umad_rw_args *ctx)
+{
+        u64 zero = 0, *val;
+        u8 class = ctx->mgmt_class;
+
+        bpf_debug("ib_umad read recv : class 0x%x\n", class);
+
+        val = bpf_map_lookup_elem(&read_count, &class);
+        if (!val) {
+                bpf_map_update_elem(&read_count, &class, &zero, BPF_NOEXIST);
+                val = bpf_map_lookup_elem(&read_count, &class);
+                if (!val)
+                        return 0;
+        }
+
+        (*val) += 1;
+
+        return 0;
+}
+SEC("tracepoint/ib_umad/ib_umad_read_send")
+int on_ib_umad_read_send(struct ib_umad_rw_args *ctx)
+{
+        u64 zero = 0, *val;
+        u8 class = ctx->mgmt_class;
+
+        bpf_debug("ib_umad read send : class 0x%x\n", class);
+
+        val = bpf_map_lookup_elem(&read_count, &class);
+        if (!val) {
+                bpf_map_update_elem(&read_count, &class, &zero, BPF_NOEXIST);
+                val = bpf_map_lookup_elem(&read_count, &class);
+                if (!val)
+                        return 0;
+        }
+
+        (*val) += 1;
+
+        return 0;
+}
+SEC("tracepoint/ib_umad/ib_umad_write")
+int on_ib_umad_write(struct ib_umad_rw_args *ctx)
+{
+        u64 zero = 0, *val;
+        u8 class = ctx->mgmt_class;
+
+        bpf_debug("ib_umad write : class 0x%x\n", class);
+
+        val = bpf_map_lookup_elem(&write_count, &class);
+        if (!val) {
+                bpf_map_update_elem(&write_count, &class, &zero, BPF_NOEXIST);
+                val = bpf_map_lookup_elem(&write_count, &class);
+                if (!val)
+                        return 0;
+        }
+
+        (*val) += 1;
+
+        return 0;
+}
+
+char _license[] SEC("license") = "GPL";
diff --git a/samples/bpf/ibumad_user.c b/samples/bpf/ibumad_user.c
new file mode 100644
index 000000000000..097d76143363
--- /dev/null
+++ b/samples/bpf/ibumad_user.c
@@ -0,0 +1,122 @@
+// SPDX-License-Identifier: GPL-2.0 OR Linux-OpenIB
+
+/**
+ * ibumad BPF sample user side
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * Copyright(c) 2018 Ira Weiny, Intel Corporation
+ */
+
+#include <linux/bpf.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <unistd.h>
+#include <sys/types.h>
+#include <limits.h>
+
+#include <sys/resource.h>
+#include <getopt.h>
+#include <net/if.h>
+
+#include "bpf_load.h"
+#include "bpf_util.h"
+#include "bpf/libbpf.h"
+
+static void dump_counts(int fd)
+{
+        __u32 key;
+        __u64 value;
+
+        for (key = 0; key < 256; key++) {
+                if (bpf_map_lookup_elem(fd, &key, &value)) {
+                        printf("failed to read key %u\n", key);
+                        continue;
+                }
+                if (value)
+                        printf("0x%02x : %llu\n", key, value);
+        }
+}
+
+static void dump_all_counts(void)
+{
+        printf("Read 'Class : count'\n");
+        dump_counts(map_fd[0]);
+        printf("Write 'Class : count'\n");
+        dump_counts(map_fd[1]);
+}
+
+static void dump_exit(int sig)
+{
+        dump_all_counts();
+        exit(0);
+}
+
+static const struct option long_options[] = {
+        {"help",      no_argument,       NULL, 'h'},
+        {"delay",     required_argument, NULL, 'd'},
+};
+
+static void usage(char *cmd)
+{
+        printf("eBPF test program to count packets from various IP addresses\n"
+                "Usage: %s <options>\n"
+                "       --help,   -h  this menu\n"
+                "       --delay,  -d  <delay>  wait <delay> sec between prints [1 - 1000000]\n"
+                , cmd
+                );
+}
+
+int main(int argc, char **argv)
+{
+        unsigned long delay = 5;
+        int longindex = 0;
+        int opt;
+        char bpf_file[256];
+
+        /* Create the eBPF kernel code path name.
+         * This follows the pattern of all of the other bpf samples
+         */
+        snprintf(bpf_file, sizeof(bpf_file), "%s_kern.o", argv[0]);
+
+        /* Do one final dump when exiting */
+        signal(SIGINT, dump_exit);
+        signal(SIGTERM, dump_exit);
+
+        while ((opt = getopt_long(argc, argv, "hd:rSw",
+                                  long_options, &longindex)) != -1) {
+                switch (opt) {
+                case 'd':
+                        delay = strtoul(optarg, NULL, 0);
+                        if (delay == ULONG_MAX || delay < 0 ||
+                            delay > 1000000) {
+                                fprintf(stderr, "ERROR: invalid delay : %s\n",
+                                        optarg);
+                                usage(argv[0]);
+                                return 1;
+                        }
+                        break;
+                default:
+                case 'h':
+                        usage(argv[0]);
+                        return 1;
+                }
+        }
+
+        if (load_bpf_file(bpf_file)) {
+                fprintf(stderr, "ERROR: failed to load eBPF from file : %s\n",
+                        bpf_file);
+                return 1;
+        }
+
+        while (1) {
+                sleep(delay);
+                dump_all_counts();
+        }
+
+        return 0;
+}