summaryrefslogtreecommitdiffstats
path: root/src/spdk/test/app/fuzz/nvme_fuzz
diff options
context:
space:
mode:
authorDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-21 11:54:28 +0000
committerDaniel Baumann <daniel.baumann@progress-linux.org>2024-04-21 11:54:28 +0000
commite6918187568dbd01842d8d1d2c808ce16a894239 (patch)
tree64f88b554b444a49f656b6c656111a145cbbaa28 /src/spdk/test/app/fuzz/nvme_fuzz
parentInitial commit. (diff)
downloadceph-e6918187568dbd01842d8d1d2c808ce16a894239.tar.xz
ceph-e6918187568dbd01842d8d1d2c808ce16a894239.zip
Adding upstream version 18.2.2.upstream/18.2.2
Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
Diffstat (limited to 'src/spdk/test/app/fuzz/nvme_fuzz')
-rw-r--r--src/spdk/test/app/fuzz/nvme_fuzz/.gitignore1
-rw-r--r--src/spdk/test/app/fuzz/nvme_fuzz/Makefile49
-rw-r--r--src/spdk/test/app/fuzz/nvme_fuzz/README.md52
-rw-r--r--src/spdk/test/app/fuzz/nvme_fuzz/example.json290
-rw-r--r--src/spdk/test/app/fuzz/nvme_fuzz/nvme_fuzz.c931
5 files changed, 1323 insertions, 0 deletions
diff --git a/src/spdk/test/app/fuzz/nvme_fuzz/.gitignore b/src/spdk/test/app/fuzz/nvme_fuzz/.gitignore
new file mode 100644
index 000000000..801146458
--- /dev/null
+++ b/src/spdk/test/app/fuzz/nvme_fuzz/.gitignore
@@ -0,0 +1 @@
+nvme_fuzz
diff --git a/src/spdk/test/app/fuzz/nvme_fuzz/Makefile b/src/spdk/test/app/fuzz/nvme_fuzz/Makefile
new file mode 100644
index 000000000..b7ad5e172
--- /dev/null
+++ b/src/spdk/test/app/fuzz/nvme_fuzz/Makefile
@@ -0,0 +1,49 @@
+#
+# Copyright (c) Intel Corporation.
+# All rights reserved.
+#
+# 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.
+# * Neither the name of Intel Corporation nor the names of its
+# contributors may be used to endorse or promote products derived
+# from this software without specific prior written permission.
+#
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+#
+
+SPDK_ROOT_DIR := $(abspath $(CURDIR)/../../../..)
+include $(SPDK_ROOT_DIR)/mk/spdk.common.mk
+include $(SPDK_ROOT_DIR)/mk/spdk.modules.mk
+
+APP = nvme_fuzz
+
+CFLAGS += -I$(SPDK_ROOT_DIR)/test/app/fuzz/common
+
+C_SRCS := nvme_fuzz.c
+
+SPDK_LIB_LIST += $(SOCK_MODULES_LIST)
+SPDK_LIB_LIST += conf event json jsonrpc log nvme rpc sock thread trace util
+
+ifeq ($(CONFIG_RDMA),y)
+SPDK_LIB_LIST += rdma
+endif
+
+include $(SPDK_ROOT_DIR)/mk/spdk.app.mk
diff --git a/src/spdk/test/app/fuzz/nvme_fuzz/README.md b/src/spdk/test/app/fuzz/nvme_fuzz/README.md
new file mode 100644
index 000000000..2f188b5b8
--- /dev/null
+++ b/src/spdk/test/app/fuzz/nvme_fuzz/README.md
@@ -0,0 +1,52 @@
+# Overview
+
+This application is intended to fuzz test the NVMe-oF target or a physical NVMe drive by
+submitting randomized NVMe commands through the SPDK NVMe initiator. Both local and remote
+drives are configured through a .ini style config file (See the -C option on the application).
+Multiple controllers and namespaces can be exposed to the fuzzer at a time. In order to
+handle multiple namespaces, the fuzzer will round robin assign a thread to each namespace and
+submit commands to that thread at a set queue depth. (currently 128 for I/O, 16 for Admin). The
+application will terminate under three conditions:
+
+1. The user specified run time expires (see the -t flag).
+2. One of the target controllers stops completing I/O operations back to the fuzzer i.e. controller timeout.
+3. The user specified a json file containing operations to run and the fuzzer has received valid completions for all of them.
+
+# Output
+
+By default, the fuzzer will print commands that:
+
+1. Complete successfully back from the target, or
+2. Are outstanding at the time of a controller timeout.
+
+Commands are dumped as named objects in json format which can then be supplied back to the
+script for targeted debugging on a subsequent run. See `Debugging` below.
+By default no output is generated when a specific command is returned with a failed status.
+This can be overridden with the -V flag. if -V is specified, each command will be dumped as
+it is completed in the JSON format specified above.
+At the end of each test run, a summary is printed for each namespace in the following format:
+
+~~~
+NS: 0x200079262300 admin qp, Total commands completed: 462459, total successful commands: 1960, random_seed: 4276918833
+~~~
+
+# Debugging
+
+If a controller hangs when processing I/O generated by the fuzzer, the fuzzer will stop
+submitting I/O and dump out all outstanding I/O on the qpair that timed out. The I/O are
+dumped as valid json. You can combine the dumped commands from the fuzzer into a json
+array in a file and then pass them to the fuzzer using the -j option. Please see the
+example.json file in this directory for an example of a properly formed array of command
+structures.
+
+Please note that you can also craft your own custom command values by using the output
+from the fuzzer as a template.
+
+# JSON Format
+
+Most of the variables in the spdk_nvme_cmd structure are represented as numbers in JSON.
+The only exception to this rule is the dptr union. This is a 16 byte union structure that
+is represented as a base64 string. If writing custom commands for input, please note this
+distinction or the application will be unable to load your custom input.
+
+Happy Fuzzing!
diff --git a/src/spdk/test/app/fuzz/nvme_fuzz/example.json b/src/spdk/test/app/fuzz/nvme_fuzz/example.json
new file mode 100644
index 000000000..95540746e
--- /dev/null
+++ b/src/spdk/test/app/fuzz/nvme_fuzz/example.json
@@ -0,0 +1,290 @@
+{
+"struct spdk_nvme_cmd": {
+ "opc": 7,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 24,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 43,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 12,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 7,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 24,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 43,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 12,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 7,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 24,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 43,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 12,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 7,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 24,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 43,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ },
+ "struct spdk_nvme_cmd": {
+ "opc": 12,
+ "fuse": 1,
+ "rsvd1": 13,
+ "psdt": 1,
+ "cid": 56732,
+ "nsid": 1,
+ "rsvd2": 1516848792,
+ "rsvd3": 1233945838,
+ "mptr": 3452736735,
+ "dptr": "FHENPcH+xM0tioD+0SrNrQ==",
+ "cdw10": 3190735246,
+ "cdw11": 2629178873,
+ "cdw12": 138580308,
+ "cdw13": 1603605200,
+ "cdw14": 3031880384,
+ "cdw15": 644909208
+ }
+}
diff --git a/src/spdk/test/app/fuzz/nvme_fuzz/nvme_fuzz.c b/src/spdk/test/app/fuzz/nvme_fuzz/nvme_fuzz.c
new file mode 100644
index 000000000..127bc1bff
--- /dev/null
+++ b/src/spdk/test/app/fuzz/nvme_fuzz/nvme_fuzz.c
@@ -0,0 +1,931 @@
+/*-
+ * BSD LICENSE
+ *
+ * Copyright (c) Intel Corporation. All rights reserved.
+ * Copyright (c) 2019 Mellanox Technologies LTD. All rights reserved.
+ *
+ * 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.
+ * * Neither the name of Intel Corporation nor the names of its
+ * contributors may be used to endorse or promote products derived
+ * from this software without specific prior written permission.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
+ * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
+ * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
+ * A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
+ * OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
+ * SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
+ * LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
+ * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
+ * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+ * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#include "spdk/stdinc.h"
+#include "spdk/conf.h"
+#include "spdk/env.h"
+#include "spdk/event.h"
+#include "spdk/util.h"
+#include "spdk/string.h"
+#include "spdk/nvme_spec.h"
+#include "spdk/nvme.h"
+#include "spdk/likely.h"
+#include "spdk/json.h"
+#include "fuzz_common.h"
+
+#define UNIQUE_OPCODES 256
+
+const char g_nvme_cmd_json_name[] = "struct spdk_nvme_cmd";
+char *g_conf_file;
+char *g_json_file = NULL;
+uint64_t g_runtime_ticks;
+unsigned int g_seed_value = 0;
+int g_runtime;
+
+int g_num_active_threads = 0;
+uint32_t g_admin_depth = 16;
+uint32_t g_io_depth = 128;
+
+bool g_valid_ns_only = false;
+bool g_verbose_mode = false;
+bool g_run_admin_commands = false;
+bool g_run;
+
+struct spdk_poller *g_app_completion_poller;
+bool g_successful_io_opcodes[UNIQUE_OPCODES] = {0};
+bool g_successful_admin_opcodes[UNIQUE_OPCODES] = {0};
+
+struct spdk_nvme_cmd *g_cmd_array;
+size_t g_cmd_array_size;
+
+/* I need context objects here because I need to keep track of all I/O that are in flight. */
+struct nvme_fuzz_request {
+ struct spdk_nvme_cmd cmd;
+ struct nvme_fuzz_qp *qp;
+ TAILQ_ENTRY(nvme_fuzz_request) link;
+};
+
+struct nvme_fuzz_trid {
+ struct spdk_nvme_transport_id trid;
+ TAILQ_ENTRY(nvme_fuzz_trid) tailq;
+};
+
+struct nvme_fuzz_ctrlr {
+ struct spdk_nvme_ctrlr *ctrlr;
+ TAILQ_ENTRY(nvme_fuzz_ctrlr) tailq;
+};
+
+struct nvme_fuzz_qp {
+ struct spdk_nvme_qpair *qpair;
+ /* array of context objects equal in length to the queue depth */
+ struct nvme_fuzz_request *req_ctx;
+ TAILQ_HEAD(, nvme_fuzz_request) free_ctx_objs;
+ TAILQ_HEAD(, nvme_fuzz_request) outstanding_ctx_objs;
+ unsigned int random_seed;
+ uint64_t completed_cmd_counter;
+ uint64_t submitted_cmd_counter;
+ uint64_t successful_completed_cmd_counter;
+ uint64_t timeout_tsc;
+ uint32_t num_cmds_outstanding;
+ bool timed_out;
+ bool is_admin;
+};
+
+struct nvme_fuzz_ns {
+ struct spdk_nvme_ns *ns;
+ struct spdk_nvme_ctrlr *ctrlr;
+ struct spdk_thread *thread;
+ struct spdk_poller *req_poller;
+ struct nvme_fuzz_qp io_qp;
+ struct nvme_fuzz_qp a_qp;
+ uint32_t nsid;
+ TAILQ_ENTRY(nvme_fuzz_ns) tailq;
+};
+
+static TAILQ_HEAD(, nvme_fuzz_ns) g_ns_list = TAILQ_HEAD_INITIALIZER(g_ns_list);
+static TAILQ_HEAD(, nvme_fuzz_ctrlr) g_ctrlr_list = TAILQ_HEAD_INITIALIZER(g_ctrlr_list);
+static TAILQ_HEAD(, nvme_fuzz_trid) g_trid_list = TAILQ_HEAD_INITIALIZER(g_trid_list);
+
+static bool
+parse_nvme_cmd_obj(void *item, struct spdk_json_val *value, size_t num_values)
+{
+ struct spdk_nvme_cmd *cmd = item;
+ struct spdk_json_val *next_val;
+ uint64_t tmp_val;
+ size_t i = 0;
+
+ while (i < num_values) {
+ if (value->type == SPDK_JSON_VAL_NAME) {
+ next_val = value + 1;
+ if (!strncmp(value->start, "opc", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UNSIGNED_8BIT_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->opc = tmp_val;
+ }
+ } else if (!strncmp(value->start, "fuse", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UNSIGNED_2BIT_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->fuse = tmp_val;
+ }
+ } else if (!strncmp(value->start, "rsvd1", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UNSIGNED_4BIT_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->rsvd1 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "psdt", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UNSIGNED_2BIT_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->psdt = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cid", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT16_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cid = tmp_val;
+ }
+ } else if (!strncmp(value->start, "nsid", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->nsid = tmp_val;
+ }
+ } else if (!strncmp(value->start, "rsvd2", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->rsvd2 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "rsvd3", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->rsvd3 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "mptr", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT64_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->mptr = tmp_val;
+ }
+ } else if (!strncmp(value->start, "dptr", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_STRING) {
+ if (fuzz_get_base_64_buffer_value(&cmd->dptr, sizeof(cmd->dptr), (char *)next_val->start,
+ next_val->len)) {
+ goto invalid;
+ }
+ }
+ } else if (!strncmp(value->start, "cdw10", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw10 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cdw11", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw11 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cdw12", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw12 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cdw13", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw13 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cdw14", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw14 = tmp_val;
+ }
+ } else if (!strncmp(value->start, "cdw15", value->len)) {
+ if (next_val->type == SPDK_JSON_VAL_NUMBER) {
+ if (fuzz_parse_json_num(next_val, UINT32_MAX, &tmp_val)) {
+ goto invalid;
+ }
+ cmd->cdw15 = tmp_val;
+ }
+ }
+ }
+ i++;
+ value++;
+ }
+ return true;
+
+invalid:
+ fprintf(stderr, "Invalid value supplied for cmd->%.*s: %.*s\n", value->len, (char *)value->start,
+ next_val->len, (char *)next_val->start);
+ return false;
+}
+
+static void
+report_successful_opcodes(bool *array, int length)
+{
+ int i;
+
+ for (i = 0; i < length; i++) {
+ if (array[i] == true) {
+ printf("%d, ", i);
+ }
+ }
+ printf("\n");
+}
+
+static int
+print_nvme_cmd(void *cb_ctx, const void *data, size_t size)
+{
+ fprintf(stderr, "%s\n", (const char *)data);
+ return 0;
+}
+
+static void
+json_dump_nvme_cmd(struct spdk_nvme_cmd *cmd)
+{
+ struct spdk_json_write_ctx *w;
+ char *dptr_value;
+
+ dptr_value = fuzz_get_value_base_64_buffer(&cmd->dptr, sizeof(cmd->dptr));
+ if (dptr_value == NULL) {
+ fprintf(stderr, "Unable to allocate buffer context for printing command.\n");
+ return;
+ }
+
+ w = spdk_json_write_begin(print_nvme_cmd, cmd, SPDK_JSON_WRITE_FLAG_FORMATTED);
+ if (w == NULL) {
+ fprintf(stderr, "Unable to allocate json context for printing command.\n");
+ free(dptr_value);
+ return;
+ }
+
+ spdk_json_write_named_object_begin(w, g_nvme_cmd_json_name);
+ spdk_json_write_named_uint32(w, "opc", cmd->opc);
+ spdk_json_write_named_uint32(w, "fuse", cmd->fuse);
+ spdk_json_write_named_uint32(w, "rsvd1", cmd->rsvd1);
+ spdk_json_write_named_uint32(w, "psdt", cmd->psdt);
+ spdk_json_write_named_uint32(w, "cid", cmd->cid);
+ spdk_json_write_named_uint32(w, "nsid", cmd->nsid);
+ spdk_json_write_named_uint32(w, "rsvd2", cmd->rsvd2);
+ spdk_json_write_named_uint32(w, "rsvd3", cmd->rsvd3);
+ spdk_json_write_named_uint32(w, "mptr", cmd->mptr);
+ spdk_json_write_named_string(w, "dptr", dptr_value);
+ spdk_json_write_named_uint32(w, "cdw10", cmd->cdw10);
+ spdk_json_write_named_uint32(w, "cdw11", cmd->cdw11);
+ spdk_json_write_named_uint32(w, "cdw12", cmd->cdw12);
+ spdk_json_write_named_uint32(w, "cdw13", cmd->cdw13);
+ spdk_json_write_named_uint32(w, "cdw14", cmd->cdw14);
+ spdk_json_write_named_uint32(w, "cdw15", cmd->cdw15);
+ spdk_json_write_object_end(w);
+
+ free(dptr_value);
+ spdk_json_write_end(w);
+}
+
+static void
+json_dump_nvme_cmd_list(struct nvme_fuzz_qp *qp)
+{
+ struct nvme_fuzz_request *ctx;
+
+ TAILQ_FOREACH(ctx, &qp->outstanding_ctx_objs, link) {
+ json_dump_nvme_cmd(&ctx->cmd);
+ }
+}
+
+static void
+handle_timeout(struct nvme_fuzz_qp *qp, bool is_admin)
+{
+ fprintf(stderr, "An %s queue has timed out. Dumping all outstanding commands from that queue\n",
+ is_admin ? "Admin" : "I/O");
+ json_dump_nvme_cmd_list(qp);
+ qp->timed_out = true;
+}
+
+static void submit_ns_cmds(struct nvme_fuzz_ns *ns_entry);
+
+static void
+nvme_fuzz_cpl_cb(void *cb_arg, const struct spdk_nvme_cpl *cpl)
+{
+ struct nvme_fuzz_request *ctx = cb_arg;
+ struct nvme_fuzz_qp *qp = ctx->qp;
+
+ qp->completed_cmd_counter++;
+ if (spdk_unlikely(cpl->status.sc == SPDK_NVME_SC_SUCCESS)) {
+ fprintf(stderr, "The following %s command (command num %lu) completed successfully\n",
+ qp->is_admin ? "Admin" : "I/O", qp->completed_cmd_counter);
+ qp->successful_completed_cmd_counter++;
+ json_dump_nvme_cmd(&ctx->cmd);
+
+ if (qp->is_admin) {
+ __sync_bool_compare_and_swap(&g_successful_admin_opcodes[ctx->cmd.opc], false, true);
+ } else {
+ __sync_bool_compare_and_swap(&g_successful_io_opcodes[ctx->cmd.opc], false, true);
+ }
+ } else if (g_verbose_mode == true) {
+ fprintf(stderr, "The following %s command (command num %lu) failed as expected.\n",
+ qp->is_admin ? "Admin" : "I/O", qp->completed_cmd_counter);
+ json_dump_nvme_cmd(&ctx->cmd);
+ }
+
+ qp->timeout_tsc = fuzz_refresh_timeout();
+ TAILQ_REMOVE(&qp->outstanding_ctx_objs, ctx, link);
+ TAILQ_INSERT_HEAD(&qp->free_ctx_objs, ctx, link);
+ assert(qp->num_cmds_outstanding > 0);
+ qp->num_cmds_outstanding--;
+}
+
+static int
+poll_for_completions(void *arg)
+{
+ struct nvme_fuzz_ns *ns_entry = arg;
+ uint64_t current_ticks = spdk_get_ticks();
+ uint64_t *counter;
+ if (!ns_entry->io_qp.timed_out) {
+ spdk_nvme_qpair_process_completions(ns_entry->io_qp.qpair, 0);
+ /* SAlways have to process admin completions for the purposes of keep alive. */
+ spdk_nvme_ctrlr_process_admin_completions(ns_entry->ctrlr);
+ }
+
+ if (g_cmd_array) {
+ if (g_run_admin_commands) {
+ counter = &ns_entry->a_qp.submitted_cmd_counter;
+ } else {
+ counter = &ns_entry->io_qp.submitted_cmd_counter;
+ }
+
+ if (*counter >= g_cmd_array_size) {
+ g_run = false;
+ }
+ } else {
+ if (current_ticks > g_runtime_ticks) {
+ g_run = false;
+ }
+ }
+
+ if (ns_entry->a_qp.timeout_tsc < current_ticks && !ns_entry->a_qp.timed_out &&
+ ns_entry->a_qp.num_cmds_outstanding > 0) {
+ handle_timeout(&ns_entry->a_qp, true);
+ }
+
+ if (ns_entry->io_qp.timeout_tsc < current_ticks && !ns_entry->io_qp.timed_out &&
+ ns_entry->io_qp.num_cmds_outstanding > 0) {
+ handle_timeout(&ns_entry->io_qp, false);
+ }
+
+ submit_ns_cmds(ns_entry);
+
+ if (g_run) {
+ return 0;
+ }
+ /*
+ * We either processed all I/O properly and can shut down normally, or we
+ * had a qp time out and we need to exit without reducing the values to 0.
+ */
+ if (ns_entry->io_qp.num_cmds_outstanding == 0 &&
+ ns_entry->a_qp.num_cmds_outstanding == 0) {
+ goto exit_handler;
+ } else if (ns_entry->io_qp.timed_out && (!g_run_admin_commands || ns_entry->a_qp.timed_out)) {
+ goto exit_handler;
+ } else {
+ return 0;
+ }
+
+exit_handler:
+ spdk_poller_unregister(&ns_entry->req_poller);
+ __sync_sub_and_fetch(&g_num_active_threads, 1);
+ spdk_thread_exit(ns_entry->thread);
+ return 0;
+}
+
+static void
+prep_nvme_cmd(struct nvme_fuzz_ns *ns_entry, struct nvme_fuzz_qp *qp, struct nvme_fuzz_request *ctx)
+{
+ if (g_cmd_array) {
+ memcpy(&ctx->cmd, &g_cmd_array[qp->submitted_cmd_counter], sizeof(ctx->cmd));
+ } else {
+ fuzz_fill_random_bytes((char *)&ctx->cmd, sizeof(ctx->cmd), &qp->random_seed);
+
+ if (g_valid_ns_only) {
+ ctx->cmd.nsid = ns_entry->nsid;
+ }
+ }
+}
+
+static int
+submit_qp_cmds(struct nvme_fuzz_ns *ns, struct nvme_fuzz_qp *qp)
+{
+ struct nvme_fuzz_request *ctx;
+ int rc;
+
+ if (qp->timed_out) {
+ return 0;
+ }
+ /* If we are reading from an array, we need to stop after the last one. */
+ while ((qp->submitted_cmd_counter < g_cmd_array_size || g_cmd_array_size == 0) &&
+ !TAILQ_EMPTY(&qp->free_ctx_objs)) {
+ ctx = TAILQ_FIRST(&qp->free_ctx_objs);
+ do {
+ prep_nvme_cmd(ns, qp, ctx);
+ } while (qp->is_admin && ctx->cmd.opc == SPDK_NVME_OPC_ASYNC_EVENT_REQUEST);
+
+ TAILQ_REMOVE(&qp->free_ctx_objs, ctx, link);
+ TAILQ_INSERT_HEAD(&qp->outstanding_ctx_objs, ctx, link);
+ qp->num_cmds_outstanding++;
+ qp->submitted_cmd_counter++;
+ if (qp->is_admin) {
+ rc = spdk_nvme_ctrlr_cmd_admin_raw(ns->ctrlr, &ctx->cmd, NULL, 0, nvme_fuzz_cpl_cb, ctx);
+ } else {
+ rc = spdk_nvme_ctrlr_cmd_io_raw(ns->ctrlr, qp->qpair, &ctx->cmd, NULL, 0, nvme_fuzz_cpl_cb, ctx);
+ }
+ if (rc) {
+ return rc;
+ }
+ }
+ return 0;
+}
+
+static void
+submit_ns_cmds(struct nvme_fuzz_ns *ns_entry)
+{
+ int rc;
+
+ if (!g_run) {
+ return;
+ }
+
+ if (g_run_admin_commands) {
+ rc = submit_qp_cmds(ns_entry, &ns_entry->a_qp);
+ if (rc) {
+ goto err_exit;
+ }
+ }
+
+ if (g_cmd_array == NULL || !g_run_admin_commands) {
+ rc = submit_qp_cmds(ns_entry, &ns_entry->io_qp);
+ }
+err_exit:
+ if (rc) {
+ /*
+ * I see the prospect of having a broken qpair on one ns as interesting
+ * enough to recommend stopping the application.
+ */
+ fprintf(stderr, "Unable to submit command with rc %d\n", rc);
+ g_run = false;
+ }
+}
+
+static void
+free_namespaces(void)
+{
+ struct nvme_fuzz_ns *ns, *tmp;
+
+ TAILQ_FOREACH_SAFE(ns, &g_ns_list, tailq, tmp) {
+ printf("NS: %p I/O qp, Total commands completed: %lu, total successful commands: %lu, random_seed: %u\n",
+ ns->ns,
+ ns->io_qp.completed_cmd_counter, ns->io_qp.successful_completed_cmd_counter, ns->io_qp.random_seed);
+ printf("NS: %p admin qp, Total commands completed: %lu, total successful commands: %lu, random_seed: %u\n",
+ ns->ns,
+ ns->a_qp.completed_cmd_counter, ns->a_qp.successful_completed_cmd_counter, ns->a_qp.random_seed);
+
+ TAILQ_REMOVE(&g_ns_list, ns, tailq);
+ if (ns->io_qp.qpair) {
+ spdk_nvme_ctrlr_free_io_qpair(ns->io_qp.qpair);
+ }
+ if (ns->io_qp.req_ctx) {
+ free(ns->io_qp.req_ctx);
+ }
+ if (ns->a_qp.req_ctx) {
+ free(ns->a_qp.req_ctx);
+ }
+ free(ns);
+ }
+}
+
+static void
+free_controllers(void)
+{
+ struct nvme_fuzz_ctrlr *ctrlr, *tmp;
+
+ TAILQ_FOREACH_SAFE(ctrlr, &g_ctrlr_list, tailq, tmp) {
+ TAILQ_REMOVE(&g_ctrlr_list, ctrlr, tailq);
+ spdk_nvme_detach(ctrlr->ctrlr);
+ free(ctrlr);
+ }
+}
+
+static void
+free_trids(void)
+{
+ struct nvme_fuzz_trid *trid, *tmp;
+
+ TAILQ_FOREACH_SAFE(trid, &g_trid_list, tailq, tmp) {
+ TAILQ_REMOVE(&g_trid_list, trid, tailq);
+ free(trid);
+ }
+}
+
+static void
+register_ns(struct spdk_nvme_ctrlr *ctrlr, struct spdk_nvme_ns *ns, uint32_t nsid)
+{
+ struct nvme_fuzz_ns *ns_entry;
+
+ ns_entry = calloc(1, sizeof(struct nvme_fuzz_ns));
+ if (ns_entry == NULL) {
+ fprintf(stderr, "Unable to allocate an entry for a namespace\n");
+ return;
+ }
+
+ ns_entry->ns = ns;
+ ns_entry->ctrlr = ctrlr;
+ ns_entry->nsid = nsid;
+
+ TAILQ_INIT(&ns_entry->io_qp.free_ctx_objs);
+ TAILQ_INIT(&ns_entry->io_qp.outstanding_ctx_objs);
+ if (g_run_admin_commands) {
+ ns_entry->a_qp.qpair = NULL;
+ TAILQ_INIT(&ns_entry->a_qp.free_ctx_objs);
+ TAILQ_INIT(&ns_entry->a_qp.outstanding_ctx_objs);
+ }
+ TAILQ_INSERT_TAIL(&g_ns_list, ns_entry, tailq);
+}
+
+static void
+register_ctrlr(struct spdk_nvme_ctrlr *ctrlr)
+{
+ struct nvme_fuzz_ctrlr *ctrlr_entry;
+ uint32_t nsid;
+ struct spdk_nvme_ns *ns;
+
+ ctrlr_entry = calloc(1, sizeof(struct nvme_fuzz_ctrlr));
+ if (ctrlr_entry == NULL) {
+ fprintf(stderr, "Unable to allocate an entry for a controller\n");
+ return;
+ }
+
+ ctrlr_entry->ctrlr = ctrlr;
+ TAILQ_INSERT_TAIL(&g_ctrlr_list, ctrlr_entry, tailq);
+
+ for (nsid = spdk_nvme_ctrlr_get_first_active_ns(ctrlr); nsid != 0;
+ nsid = spdk_nvme_ctrlr_get_next_active_ns(ctrlr, nsid)) {
+ ns = spdk_nvme_ctrlr_get_ns(ctrlr, nsid);
+ if (ns == NULL) {
+ continue;
+ }
+ register_ns(ctrlr, ns, nsid);
+ }
+}
+
+static void
+attach_cb(void *cb_ctx, const struct spdk_nvme_transport_id *trid,
+ struct spdk_nvme_ctrlr *ctrlr, const struct spdk_nvme_ctrlr_opts *opts)
+{
+ register_ctrlr(ctrlr);
+}
+
+static bool
+probe_cb(void *cb_ctx, const struct spdk_nvme_transport_id *trid, struct spdk_nvme_ctrlr_opts *opts)
+{
+ printf("Controller trtype %s\ttraddr %s\n", spdk_nvme_transport_id_trtype_str(trid->trtype),
+ trid->traddr);
+
+ return true;
+}
+
+static int
+prep_qpair(struct nvme_fuzz_ns *ns, struct nvme_fuzz_qp *qp, uint32_t max_qdepth)
+{
+ uint32_t i;
+
+ /* ensure that each qpair gets a unique random seed for maximum command dispersion. */
+
+ if (g_seed_value != 0) {
+ qp->random_seed = g_seed_value;
+ } else {
+ /* Take the low 32 bits of spdk_get_ticks. This should be more granular than time(). */
+ qp->random_seed = spdk_get_ticks();
+ }
+
+ qp->timeout_tsc = fuzz_refresh_timeout();
+
+ qp->req_ctx = calloc(max_qdepth, sizeof(struct nvme_fuzz_request));
+ if (qp->req_ctx == NULL) {
+ fprintf(stderr, "Unable to allocate I/O contexts for I/O qpair.\n");
+ return -1;
+ }
+
+ for (i = 0; i < max_qdepth; i++) {
+ qp->req_ctx[i].qp = qp;
+ TAILQ_INSERT_HEAD(&qp->free_ctx_objs, &qp->req_ctx[i], link);
+ }
+
+ return 0;
+}
+
+static int
+prepare_qpairs(void)
+{
+ struct spdk_nvme_io_qpair_opts opts;
+ struct nvme_fuzz_ns *ns_entry;
+
+ TAILQ_FOREACH(ns_entry, &g_ns_list, tailq) {
+ spdk_nvme_ctrlr_get_default_io_qpair_opts(ns_entry->ctrlr, &opts, sizeof(opts));
+ ns_entry->io_qp.qpair = spdk_nvme_ctrlr_alloc_io_qpair(ns_entry->ctrlr, &opts, sizeof(opts));
+ if (ns_entry->io_qp.qpair == NULL) {
+ fprintf(stderr, "Unable to create a qpair for a namespace\n");
+ return -1;
+ }
+
+ ns_entry->io_qp.is_admin = false;
+ if (prep_qpair(ns_entry, &ns_entry->io_qp, g_io_depth) != 0) {
+ fprintf(stderr, "Unable to allocate request contexts for I/O qpair.\n");
+ return -1;
+ }
+
+ if (g_run_admin_commands) {
+ ns_entry->a_qp.is_admin = true;
+ if (prep_qpair(ns_entry, &ns_entry->a_qp, g_admin_depth) != 0) {
+ fprintf(stderr, "Unable to allocate request contexts for admin qpair.\n");
+ return -1;
+ }
+ }
+ }
+ return 0;
+}
+
+static void
+start_ns_poller(void *ctx)
+{
+ struct nvme_fuzz_ns *ns_entry = ctx;
+
+ ns_entry->req_poller = SPDK_POLLER_REGISTER(poll_for_completions, ns_entry, 0);
+ submit_ns_cmds(ns_entry);
+}
+
+static int
+check_app_completion(void *ctx)
+{
+
+ if (g_num_active_threads <= 0) {
+ spdk_poller_unregister(&g_app_completion_poller);
+ if (g_cmd_array) {
+ free(g_cmd_array);
+ }
+ printf("Fuzzing completed. Shutting down the fuzz application\n\n");
+ printf("Dumping successful admin opcodes:\n");
+ report_successful_opcodes(g_successful_admin_opcodes, UNIQUE_OPCODES);
+ printf("Dumping successful io opcodes:\n");
+ report_successful_opcodes(g_successful_io_opcodes, UNIQUE_OPCODES);
+ free_namespaces();
+ free_controllers();
+ free_trids();
+ spdk_app_stop(0);
+ }
+ return 0;
+}
+
+static void
+begin_fuzz(void *ctx)
+{
+ struct nvme_fuzz_ns *ns_entry;
+ struct nvme_fuzz_trid *trid;
+ int rc;
+
+ if (!spdk_iommu_is_enabled()) {
+ /* Don't set rc to an error code here. We don't want to fail an automated test based on this. */
+ fprintf(stderr, "The IOMMU must be enabled to run this program to avoid unsafe memory accesses.\n");
+ rc = 0;
+ goto out;
+ }
+
+ TAILQ_FOREACH(trid, &g_trid_list, tailq) {
+ if (spdk_nvme_probe(&trid->trid, trid, probe_cb, attach_cb, NULL) != 0) {
+ fprintf(stderr, "spdk_nvme_probe() failed for transport address '%s'\n",
+ trid->trid.traddr);
+ rc = -1;
+ goto out;
+ }
+ }
+
+ if (TAILQ_EMPTY(&g_ns_list)) {
+ fprintf(stderr, "No valid NVMe Namespaces to fuzz\n");
+ rc = -EINVAL;
+ goto out;
+ }
+
+ rc = prepare_qpairs();
+
+ if (rc < 0) {
+ fprintf(stderr, "Unable to prepare the qpairs\n");
+ goto out;
+ }
+
+ g_runtime_ticks = spdk_get_ticks() + g_runtime * spdk_get_ticks_hz();
+
+ /* Assigning all of the threads and then starting them makes cleanup easier. */
+ TAILQ_FOREACH(ns_entry, &g_ns_list, tailq) {
+ ns_entry->thread = spdk_thread_create(NULL, NULL);
+ if (ns_entry->thread == NULL) {
+ fprintf(stderr, "Failed to allocate thread for namespace.\n");
+ goto out;
+ }
+ }
+
+ TAILQ_FOREACH(ns_entry, &g_ns_list, tailq) {
+ spdk_thread_send_msg(ns_entry->thread, start_ns_poller, ns_entry);
+ __sync_add_and_fetch(&g_num_active_threads, 1);
+ }
+
+ g_app_completion_poller = SPDK_POLLER_REGISTER(check_app_completion, NULL, 1000000);
+ return;
+out:
+ printf("Shutting down the fuzz application\n");
+ free_namespaces();
+ free_controllers();
+ free_trids();
+ spdk_app_stop(rc);
+}
+
+static int
+parse_trids(void)
+{
+ struct spdk_conf *config = NULL;
+ struct spdk_conf_section *sp;
+ const char *trid_char;
+ struct nvme_fuzz_trid *current_trid;
+ int num_subsystems = 0;
+ int rc = 0;
+
+ if (g_conf_file) {
+ config = spdk_conf_allocate();
+ if (!config) {
+ fprintf(stderr, "Unable to allocate an spdk_conf object\n");
+ return -1;
+ }
+
+ rc = spdk_conf_read(config, g_conf_file);
+ if (rc) {
+ fprintf(stderr, "Unable to convert the conf file into a readable system\n");
+ rc = -1;
+ goto exit;
+ }
+
+ sp = spdk_conf_find_section(config, "Nvme");
+
+ if (sp == NULL) {
+ fprintf(stderr, "No Nvme configuration in conf file\n");
+ goto exit;
+ }
+
+ while ((trid_char = spdk_conf_section_get_nmval(sp, "TransportID", num_subsystems, 0)) != NULL) {
+ current_trid = malloc(sizeof(struct nvme_fuzz_trid));
+ if (!current_trid) {
+ fprintf(stderr, "Unable to allocate memory for transport ID\n");
+ rc = -1;
+ goto exit;
+ }
+ rc = spdk_nvme_transport_id_parse(&current_trid->trid, trid_char);
+
+ if (rc < 0) {
+ fprintf(stderr, "failed to parse transport ID: %s\n", trid_char);
+ free(current_trid);
+ rc = -1;
+ goto exit;
+ }
+ TAILQ_INSERT_TAIL(&g_trid_list, current_trid, tailq);
+ num_subsystems++;
+ }
+ }
+
+exit:
+ if (config != NULL) {
+ spdk_conf_free(config);
+ }
+ return rc;
+}
+
+static void
+nvme_fuzz_usage(void)
+{
+ fprintf(stderr, " -a Perform admin commands. if -j is specified, \
+only admin commands will run. Otherwise they will be run in tandem with I/O commands.\n");
+ fprintf(stderr, " -C <path> Path to a configuration file.\n");
+ fprintf(stderr,
+ " -j <path> Path to a json file containing named objects of type spdk_nvme_cmd. If this option is specified, -t will be ignored.\n");
+ fprintf(stderr, " -N Target only valid namespace with commands. \
+This helps dig deeper into other errors besides invalid namespace.\n");
+ fprintf(stderr, " -S <integer> Seed value for test.\n");
+ fprintf(stderr,
+ " -t <integer> Time in seconds to run the fuzz test. Only valid if -j is not specified.\n");
+ fprintf(stderr, " -V Enable logging of each submitted command.\n");
+}
+
+static int
+nvme_fuzz_parse(int ch, char *arg)
+{
+ int64_t error_test;
+
+ switch (ch) {
+ case 'a':
+ g_run_admin_commands = true;
+ break;
+ case 'C':
+ g_conf_file = optarg;
+ break;
+ case 'j':
+ g_json_file = optarg;
+ break;
+ case 'N':
+ g_valid_ns_only = true;
+ break;
+ case 'S':
+ error_test = spdk_strtol(arg, 10);
+ if (error_test < 0) {
+ fprintf(stderr, "Invalid value supplied for the random seed.\n");
+ return -1;
+ } else {
+ g_seed_value = error_test;
+ }
+ break;
+ case 't':
+ g_runtime = spdk_strtol(optarg, 10);
+ if (g_runtime < 0 || g_runtime > MAX_RUNTIME_S) {
+ fprintf(stderr, "You must supply a positive runtime value less than 86401.\n");
+ return -1;
+ }
+ break;
+ case 'V':
+ g_verbose_mode = true;
+ break;
+ case '?':
+ default:
+ return -EINVAL;
+ }
+ return 0;
+}
+
+int
+main(int argc, char **argv)
+{
+ struct spdk_app_opts opts = {};
+ int rc;
+
+ spdk_app_opts_init(&opts);
+ opts.name = "nvme_fuzz";
+
+ g_runtime = DEFAULT_RUNTIME;
+ g_run = true;
+
+ if ((rc = spdk_app_parse_args(argc, argv, &opts, "aC:j:NS:t:V", NULL, nvme_fuzz_parse,
+ nvme_fuzz_usage) != SPDK_APP_PARSE_ARGS_SUCCESS)) {
+ return rc;
+ }
+
+ if (g_conf_file) {
+ parse_trids();
+ }
+
+ if (g_json_file != NULL) {
+ g_cmd_array_size = fuzz_parse_args_into_array(g_json_file, (void **)&g_cmd_array,
+ sizeof(struct spdk_nvme_cmd), g_nvme_cmd_json_name, parse_nvme_cmd_obj);
+ if (g_cmd_array_size == 0) {
+ fprintf(stderr, "The provided json file did not contain any valid commands. Exiting.");
+ return -EINVAL;
+ }
+ }
+
+ rc = spdk_app_start(&opts, begin_fuzz, NULL);
+
+ return rc;
+}