summaryrefslogtreecommitdiff
path: root/kernel/trace/blktrace.c
diff options
context:
space:
mode:
Diffstat (limited to 'kernel/trace/blktrace.c')
-rw-r--r--kernel/trace/blktrace.c1692
1 files changed, 1039 insertions, 653 deletions
diff --git a/kernel/trace/blktrace.c b/kernel/trace/blktrace.c
index b8b8560bfb95..d031c8d80be4 100644
--- a/kernel/trace/blktrace.c
+++ b/kernel/trace/blktrace.c
@@ -1,20 +1,11 @@
+// SPDX-License-Identifier: GPL-2.0
/*
* Copyright (C) 2006 Jens Axboe <axboe@kernel.dk>
*
- * This program is free software; you can redistribute it and/or modify
- * it under the terms of the GNU General Public License version 2 as
- * published by the Free Software Foundation.
- *
- * This program is distributed in the hope that it will be useful,
- * but WITHOUT ANY WARRANTY; without even the implied warranty of
- * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
- * GNU General Public License for more details.
- *
- * You should have received a copy of the GNU General Public License
- * along with this program; if not, write to the Free Software
- * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
- *
*/
+
+#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
+
#include <linux/kernel.h>
#include <linux/blkdev.h>
#include <linux/blktrace_api.h>
@@ -26,6 +17,10 @@
#include <linux/export.h>
#include <linux/time.h>
#include <linux/uaccess.h>
+#include <linux/list.h>
+#include <linux/blk-cgroup.h>
+
+#include "../../block/blk.h"
#include <trace/events/block.h>
@@ -38,12 +33,21 @@ static unsigned int blktrace_seq __read_mostly = 1;
static struct trace_array *blk_tr;
static bool blk_tracer_enabled __read_mostly;
+static LIST_HEAD(running_trace_list);
+static __cacheline_aligned_in_smp DEFINE_RAW_SPINLOCK(running_trace_lock);
+
/* Select an alternative, minimalistic output than the original one */
#define TRACE_BLK_OPT_CLASSIC 0x1
+#define TRACE_BLK_OPT_CGROUP 0x2
+#define TRACE_BLK_OPT_CGNAME 0x4
static struct tracer_opt blk_tracer_opts[] = {
/* Default disable the minimalistic output */
{ TRACER_OPT(blk_classic, TRACE_BLK_OPT_CLASSIC) },
+#ifdef CONFIG_BLK_CGROUP
+ { TRACER_OPT(blk_cgroup, TRACE_BLK_OPT_CGROUP) },
+ { TRACER_OPT(blk_cgname, TRACE_BLK_OPT_CGNAME) },
+#endif
{ }
};
@@ -53,87 +57,197 @@ static struct tracer_flags blk_tracer_flags = {
};
/* Global reference count of probes */
-static atomic_t blk_probes_ref = ATOMIC_INIT(0);
+static DEFINE_MUTEX(blk_probe_mutex);
+static int blk_probes_ref;
static void blk_register_tracepoints(void);
static void blk_unregister_tracepoints(void);
+static void record_blktrace_event(struct blk_io_trace *t, pid_t pid, int cpu,
+ sector_t sector, int bytes, u64 what,
+ dev_t dev, int error, u64 cgid,
+ ssize_t cgid_len, void *pdu_data, int pdu_len)
+
+{
+ /*
+ * These two are not needed in ftrace as they are in the
+ * generic trace_entry, filled by tracing_generic_entry_update,
+ * but for the trace_event->bin() synthesizer benefit we do it
+ * here too.
+ */
+ t->cpu = cpu;
+ t->pid = pid;
+
+ t->sector = sector;
+ t->bytes = bytes;
+ t->action = lower_32_bits(what);
+ t->device = dev;
+ t->error = error;
+ t->pdu_len = pdu_len + cgid_len;
+
+ if (cgid_len)
+ memcpy((void *)t + sizeof(*t), &cgid, cgid_len);
+ if (pdu_len)
+ memcpy((void *)t + sizeof(*t) + cgid_len, pdu_data, pdu_len);
+}
+
+static void record_blktrace_event2(struct blk_io_trace2 *t2, pid_t pid, int cpu,
+ sector_t sector, int bytes, u64 what,
+ dev_t dev, int error, u64 cgid,
+ ssize_t cgid_len, void *pdu_data,
+ int pdu_len)
+{
+ t2->pid = pid;
+ t2->cpu = cpu;
+
+ t2->sector = sector;
+ t2->bytes = bytes;
+ t2->action = what;
+ t2->device = dev;
+ t2->error = error;
+ t2->pdu_len = pdu_len + cgid_len;
+
+ if (cgid_len)
+ memcpy((void *)t2 + sizeof(*t2), &cgid, cgid_len);
+ if (pdu_len)
+ memcpy((void *)t2 + sizeof(*t2) + cgid_len, pdu_data, pdu_len);
+}
+
+static void relay_blktrace_event1(struct blk_trace *bt, unsigned long sequence,
+ pid_t pid, int cpu, sector_t sector, int bytes,
+ u64 what, int error, u64 cgid,
+ ssize_t cgid_len, void *pdu_data, int pdu_len)
+{
+ struct blk_io_trace *t;
+ size_t trace_len = sizeof(*t) + pdu_len + cgid_len;
+
+ t = relay_reserve(bt->rchan, trace_len);
+ if (!t)
+ return;
+
+ t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION;
+ t->sequence = sequence;
+ t->time = ktime_to_ns(ktime_get());
+
+ record_blktrace_event(t, pid, cpu, sector, bytes, what, bt->dev, error,
+ cgid, cgid_len, pdu_data, pdu_len);
+}
+
+static void relay_blktrace_event2(struct blk_trace *bt, unsigned long sequence,
+ pid_t pid, int cpu, sector_t sector,
+ int bytes, u64 what, int error, u64 cgid,
+ ssize_t cgid_len, void *pdu_data, int pdu_len)
+{
+ struct blk_io_trace2 *t;
+ size_t trace_len = sizeof(struct blk_io_trace2) + pdu_len + cgid_len;
+
+ t = relay_reserve(bt->rchan, trace_len);
+ if (!t)
+ return;
+
+ t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE2_VERSION;
+ t->sequence = sequence;
+ t->time = ktime_to_ns(ktime_get());
+
+ record_blktrace_event2(t, pid, cpu, sector, bytes, what, bt->dev, error,
+ cgid, cgid_len, pdu_data, pdu_len);
+}
+
+static void relay_blktrace_event(struct blk_trace *bt, unsigned long sequence,
+ pid_t pid, int cpu, sector_t sector, int bytes,
+ u64 what, int error, u64 cgid,
+ ssize_t cgid_len, void *pdu_data, int pdu_len)
+{
+ if (bt->version == 2)
+ return relay_blktrace_event2(bt, sequence, pid, cpu, sector,
+ bytes, what, error, cgid, cgid_len,
+ pdu_data, pdu_len);
+ return relay_blktrace_event1(bt, sequence, pid, cpu, sector, bytes,
+ what, error, cgid, cgid_len, pdu_data,
+ pdu_len);
+}
+
/*
* Send out a notify message.
*/
-static void trace_note(struct blk_trace *bt, pid_t pid, int action,
- const void *data, size_t len)
+static void trace_note(struct blk_trace *bt, pid_t pid, u64 action,
+ const void *data, size_t len, u64 cgid)
{
- struct blk_io_trace *t;
struct ring_buffer_event *event = NULL;
- struct ring_buffer *buffer = NULL;
- int pc = 0;
+ struct trace_buffer *buffer = NULL;
+ unsigned int trace_ctx = 0;
int cpu = smp_processor_id();
bool blk_tracer = blk_tracer_enabled;
+ ssize_t cgid_len = cgid ? sizeof(cgid) : 0;
+ action = lower_32_bits(action | (cgid ? __BLK_TN_CGROUP : 0));
if (blk_tracer) {
- buffer = blk_tr->trace_buffer.buffer;
- pc = preempt_count();
+ struct blk_io_trace2 *t;
+ size_t trace_len = sizeof(*t) + cgid_len + len;
+
+ buffer = blk_tr->array_buffer.buffer;
+ trace_ctx = tracing_gen_ctx_flags(0);
event = trace_buffer_lock_reserve(buffer, TRACE_BLK,
- sizeof(*t) + len,
- 0, pc);
+ trace_len, trace_ctx);
if (!event)
return;
t = ring_buffer_event_data(event);
- goto record_it;
+ record_blktrace_event2(t, pid, cpu, 0, 0,
+ action, bt->dev, 0, cgid, cgid_len,
+ (void *)data, len);
+ trace_buffer_unlock_commit(blk_tr, buffer, event, trace_ctx);
+ return;
}
if (!bt->rchan)
return;
- t = relay_reserve(bt->rchan, sizeof(*t) + len);
- if (t) {
- t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION;
- t->time = ktime_to_ns(ktime_get());
-record_it:
- t->device = bt->dev;
- t->action = action;
- t->pid = pid;
- t->cpu = cpu;
- t->pdu_len = len;
- memcpy((void *) t + sizeof(*t), data, len);
-
- if (blk_tracer)
- trace_buffer_unlock_commit(buffer, event, 0, pc);
- }
+ relay_blktrace_event(bt, 0, pid, cpu, 0, 0, action, 0, cgid,
+ cgid_len, (void *)data, len);
}
/*
* Send out a notify for this process, if we haven't done so since a trace
* started
*/
-static void trace_note_tsk(struct blk_trace *bt, struct task_struct *tsk)
+static void trace_note_tsk(struct task_struct *tsk)
{
+ unsigned long flags;
+ struct blk_trace *bt;
+
tsk->btrace_seq = blktrace_seq;
- trace_note(bt, tsk->pid, BLK_TN_PROCESS, tsk->comm, sizeof(tsk->comm));
+ raw_spin_lock_irqsave(&running_trace_lock, flags);
+ list_for_each_entry(bt, &running_trace_list, running_list) {
+ trace_note(bt, tsk->pid, BLK_TN_PROCESS, tsk->comm,
+ sizeof(tsk->comm), 0);
+ }
+ raw_spin_unlock_irqrestore(&running_trace_lock, flags);
}
static void trace_note_time(struct blk_trace *bt)
{
- struct timespec now;
+ struct timespec64 now;
unsigned long flags;
u32 words[2];
- getnstimeofday(&now);
- words[0] = now.tv_sec;
+ /* need to check user space to see if this breaks in y2038 or y2106 */
+ ktime_get_real_ts64(&now);
+ words[0] = (u32)now.tv_sec;
words[1] = now.tv_nsec;
local_irq_save(flags);
- trace_note(bt, 0, BLK_TN_TIMESTAMP, words, sizeof(words));
+ trace_note(bt, 0, BLK_TN_TIMESTAMP, words, sizeof(words), 0);
local_irq_restore(flags);
}
-void __trace_note_message(struct blk_trace *bt, const char *fmt, ...)
+void __blk_trace_note_message(struct blk_trace *bt,
+ struct cgroup_subsys_state *css, const char *fmt, ...)
{
int n;
va_list args;
unsigned long flags;
char *buf;
+ u64 cgid = 0;
if (unlikely(bt->trace_state != Blktrace_running &&
!blk_tracer_enabled))
@@ -152,12 +266,18 @@ void __trace_note_message(struct blk_trace *bt, const char *fmt, ...)
n = vscnprintf(buf, BLK_TN_MAX_MSG, fmt, args);
va_end(args);
- trace_note(bt, 0, BLK_TN_MESSAGE, buf, n);
+#ifdef CONFIG_BLK_CGROUP
+ if (css && (blk_tracer_flags.val & TRACE_BLK_OPT_CGROUP))
+ cgid = cgroup_id(css->cgroup);
+ else
+ cgid = 1;
+#endif
+ trace_note(bt, current->pid, BLK_TN_MESSAGE, buf, n, cgid);
local_irq_restore(flags);
}
-EXPORT_SYMBOL_GPL(__trace_note_message);
+EXPORT_SYMBOL_GPL(__blk_trace_note_message);
-static int act_log_check(struct blk_trace *bt, u32 what, sector_t sector,
+static int act_log_check(struct blk_trace *bt, u64 what, sector_t sector,
pid_t pid)
{
if (((bt->act_mask << BLK_TC_SHIFT) & what) == 0)
@@ -177,9 +297,10 @@ static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ),
BLK_TC_ACT(BLK_TC_WRITE) };
#define BLK_TC_RAHEAD BLK_TC_AHEAD
+#define BLK_TC_PREFLUSH BLK_TC_FLUSH
/* The ilog2() calls fall out because they're constant */
-#define MASK_TC_BIT(rw, __name) ((rw & REQ_ ## __name) << \
+#define MASK_TC_BIT(rw, __name) ((__force u32)(rw & REQ_ ## __name) << \
(ilog2(BLK_TC_ ## __name) + BLK_TC_SHIFT - __REQ_ ## __name))
/*
@@ -187,28 +308,74 @@ static const u32 ddir_act[2] = { BLK_TC_ACT(BLK_TC_READ),
* blk_io_trace structure and places it in a per-cpu subbuffer.
*/
static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
- int rw, u32 what, int error, int pdu_len, void *pdu_data)
+ const blk_opf_t opf, u64 what, int error,
+ int pdu_len, void *pdu_data, u64 cgid)
{
struct task_struct *tsk = current;
struct ring_buffer_event *event = NULL;
- struct ring_buffer *buffer = NULL;
- struct blk_io_trace *t;
+ struct trace_buffer *buffer = NULL;
unsigned long flags = 0;
unsigned long *sequence;
+ unsigned int trace_ctx = 0;
pid_t pid;
- int cpu, pc = 0;
+ int cpu;
bool blk_tracer = blk_tracer_enabled;
+ ssize_t cgid_len = cgid ? sizeof(cgid) : 0;
+ const enum req_op op = opf & REQ_OP_MASK;
+ size_t trace_len;
if (unlikely(bt->trace_state != Blktrace_running && !blk_tracer))
return;
- what |= ddir_act[rw & WRITE];
- what |= MASK_TC_BIT(rw, SYNC);
- what |= MASK_TC_BIT(rw, RAHEAD);
- what |= MASK_TC_BIT(rw, META);
- what |= MASK_TC_BIT(rw, DISCARD);
- what |= MASK_TC_BIT(rw, FLUSH);
- what |= MASK_TC_BIT(rw, FUA);
+ what |= ddir_act[op_is_write(op) ? WRITE : READ];
+ what |= MASK_TC_BIT(opf, SYNC);
+ what |= MASK_TC_BIT(opf, RAHEAD);
+ what |= MASK_TC_BIT(opf, META);
+ what |= MASK_TC_BIT(opf, PREFLUSH);
+ what |= MASK_TC_BIT(opf, FUA);
+
+ switch (op) {
+ case REQ_OP_DISCARD:
+ case REQ_OP_SECURE_ERASE:
+ what |= BLK_TC_ACT(BLK_TC_DISCARD);
+ break;
+ case REQ_OP_FLUSH:
+ what |= BLK_TC_ACT(BLK_TC_FLUSH);
+ break;
+ case REQ_OP_ZONE_APPEND:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_APPEND);
+ break;
+ case REQ_OP_ZONE_RESET:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_RESET);
+ break;
+ case REQ_OP_ZONE_RESET_ALL:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_RESET_ALL);
+ break;
+ case REQ_OP_ZONE_FINISH:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_FINISH);
+ break;
+ case REQ_OP_ZONE_OPEN:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_OPEN);
+ break;
+ case REQ_OP_ZONE_CLOSE:
+ what |= BLK_TC_ACT(BLK_TC_ZONE_CLOSE);
+ break;
+ case REQ_OP_WRITE_ZEROES:
+ what |= BLK_TC_ACT(BLK_TC_WRITE_ZEROES);
+ break;
+ default:
+ break;
+ }
+
+ /* Drop trace events for zone operations with blktrace v1 */
+ if (bt->version == 1 && (what >> BLK_TC_SHIFT) > BLK_TC_END_V1) {
+ pr_debug_ratelimited("blktrace v1 cannot trace zone operation 0x%llx\n",
+ (unsigned long long)what);
+ return;
+ }
+
+ if (cgid)
+ what |= __BLK_TA_CGROUP;
pid = tsk->pid;
if (act_log_check(bt, what, sector, pid))
@@ -218,106 +385,196 @@ static void __blk_add_trace(struct blk_trace *bt, sector_t sector, int bytes,
if (blk_tracer) {
tracing_record_cmdline(current);
- buffer = blk_tr->trace_buffer.buffer;
- pc = preempt_count();
+ buffer = blk_tr->array_buffer.buffer;
+ trace_ctx = tracing_gen_ctx_flags(0);
+ switch (bt->version) {
+ case 1:
+ trace_len = sizeof(struct blk_io_trace);
+ break;
+ case 2:
+ default:
+ /*
+ * ftrace always uses v2 (blk_io_trace2) format.
+ *
+ * For sysfs-enabled tracing path (enabled via
+ * /sys/block/DEV/trace/enable), blk_trace_setup_queue()
+ * never initializes bt->version, leaving it 0 from
+ * kzalloc(). We must handle version==0 safely here.
+ *
+ * Fall through to default to ensure we never hit the
+ * old bug where default set trace_len=0, causing
+ * buffer underflow and memory corruption.
+ *
+ * Always use v2 format for ftrace and normalize
+ * bt->version to 2 when uninitialized.
+ */
+ trace_len = sizeof(struct blk_io_trace2);
+ if (bt->version == 0)
+ bt->version = 2;
+ break;
+ }
+ trace_len += pdu_len + cgid_len;
event = trace_buffer_lock_reserve(buffer, TRACE_BLK,
- sizeof(*t) + pdu_len,
- 0, pc);
+ trace_len, trace_ctx);
if (!event)
return;
- t = ring_buffer_event_data(event);
- goto record_it;
+
+ switch (bt->version) {
+ case 1:
+ record_blktrace_event(ring_buffer_event_data(event),
+ pid, cpu, sector, bytes,
+ what, bt->dev, error, cgid, cgid_len,
+ pdu_data, pdu_len);
+ break;
+ case 2:
+ default:
+ /*
+ * Use v2 recording function (record_blktrace_event2)
+ * which writes blk_io_trace2 structure with correct
+ * field layout:
+ * - 32-bit pid at offset 28
+ * - 64-bit action at offset 32
+ *
+ * Fall through to default handles version==0 case
+ * (from sysfs path), ensuring we always use correct
+ * v2 recording function to match the v2 buffer
+ * allocated above.
+ */
+ record_blktrace_event2(ring_buffer_event_data(event),
+ pid, cpu, sector, bytes,
+ what, bt->dev, error, cgid, cgid_len,
+ pdu_data, pdu_len);
+ break;
+ }
+
+ trace_buffer_unlock_commit(blk_tr, buffer, event, trace_ctx);
+ return;
}
+ if (unlikely(tsk->btrace_seq != blktrace_seq))
+ trace_note_tsk(tsk);
+
/*
* A word about the locking here - we disable interrupts to reserve
* some space in the relay per-cpu buffer, to prevent an irq
* from coming in and stepping on our toes.
*/
local_irq_save(flags);
+ sequence = per_cpu_ptr(bt->sequence, cpu);
+ (*sequence)++;
+ relay_blktrace_event(bt, *sequence, pid, cpu, sector, bytes,
+ what, error, cgid, cgid_len, pdu_data, pdu_len);
+ local_irq_restore(flags);
+}
- if (unlikely(tsk->btrace_seq != blktrace_seq))
- trace_note_tsk(bt, tsk);
+static void blk_trace_free(struct request_queue *q, struct blk_trace *bt)
+{
+ relay_close(bt->rchan);
- t = relay_reserve(bt->rchan, sizeof(*t) + pdu_len);
- if (t) {
- sequence = per_cpu_ptr(bt->sequence, cpu);
+ /*
+ * If 'bt->dir' is not set, then both 'dropped' and 'msg' are created
+ * under 'q->debugfs_dir', thus lookup and remove them.
+ */
+ if (!bt->dir) {
+ debugfs_lookup_and_remove("dropped", q->debugfs_dir);
+ debugfs_lookup_and_remove("msg", q->debugfs_dir);
+ } else {
+ debugfs_remove(bt->dir);
+ }
+ free_percpu(bt->sequence);
+ free_percpu(bt->msg_data);
+ kfree(bt);
+}
- t->magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION;
- t->sequence = ++(*sequence);
- t->time = ktime_to_ns(ktime_get());
-record_it:
- /*
- * These two are not needed in ftrace as they are in the
- * generic trace_entry, filled by tracing_generic_entry_update,
- * but for the trace_event->bin() synthesizer benefit we do it
- * here too.
- */
- t->cpu = cpu;
- t->pid = pid;
+static void get_probe_ref(void)
+{
+ mutex_lock(&blk_probe_mutex);
+ if (++blk_probes_ref == 1)
+ blk_register_tracepoints();
+ mutex_unlock(&blk_probe_mutex);
+}
- t->sector = sector;
- t->bytes = bytes;
- t->action = what;
- t->device = bt->dev;
- t->error = error;
- t->pdu_len = pdu_len;
+static void put_probe_ref(void)
+{
+ mutex_lock(&blk_probe_mutex);
+ if (!--blk_probes_ref)
+ blk_unregister_tracepoints();
+ mutex_unlock(&blk_probe_mutex);
+}
- if (pdu_len)
- memcpy((void *) t + sizeof(*t), pdu_data, pdu_len);
+static int blk_trace_start(struct blk_trace *bt)
+{
+ if (bt->trace_state != Blktrace_setup &&
+ bt->trace_state != Blktrace_stopped)
+ return -EINVAL;
- if (blk_tracer) {
- trace_buffer_unlock_commit(buffer, event, 0, pc);
- return;
- }
- }
+ blktrace_seq++;
+ smp_mb();
+ bt->trace_state = Blktrace_running;
+ raw_spin_lock_irq(&running_trace_lock);
+ list_add(&bt->running_list, &running_trace_list);
+ raw_spin_unlock_irq(&running_trace_lock);
+ trace_note_time(bt);
- local_irq_restore(flags);
+ return 0;
}
-static struct dentry *blk_tree_root;
-static DEFINE_MUTEX(blk_tree_mutex);
-
-static void blk_trace_free(struct blk_trace *bt)
+static int blk_trace_stop(struct blk_trace *bt)
{
- debugfs_remove(bt->msg_file);
- debugfs_remove(bt->dropped_file);
- relay_close(bt->rchan);
- debugfs_remove(bt->dir);
- free_percpu(bt->sequence);
- free_percpu(bt->msg_data);
- kfree(bt);
+ if (bt->trace_state != Blktrace_running)
+ return -EINVAL;
+
+ bt->trace_state = Blktrace_stopped;
+ raw_spin_lock_irq(&running_trace_lock);
+ list_del_init(&bt->running_list);
+ raw_spin_unlock_irq(&running_trace_lock);
+ relay_flush(bt->rchan);
+
+ return 0;
}
-static void blk_trace_cleanup(struct blk_trace *bt)
+static void blk_trace_cleanup(struct request_queue *q, struct blk_trace *bt)
{
- blk_trace_free(bt);
- if (atomic_dec_and_test(&blk_probes_ref))
- blk_unregister_tracepoints();
+ blk_trace_stop(bt);
+ synchronize_rcu();
+ blk_trace_free(q, bt);
+ put_probe_ref();
}
-int blk_trace_remove(struct request_queue *q)
+static int __blk_trace_remove(struct request_queue *q)
{
struct blk_trace *bt;
- bt = xchg(&q->blk_trace, NULL);
+ bt = rcu_replace_pointer(q->blk_trace, NULL,
+ lockdep_is_held(&q->debugfs_mutex));
if (!bt)
return -EINVAL;
- if (bt->trace_state != Blktrace_running)
- blk_trace_cleanup(bt);
+ blk_trace_cleanup(q, bt);
return 0;
}
+
+int blk_trace_remove(struct request_queue *q)
+{
+ int ret;
+
+ mutex_lock(&q->debugfs_mutex);
+ ret = __blk_trace_remove(q);
+ mutex_unlock(&q->debugfs_mutex);
+
+ return ret;
+}
EXPORT_SYMBOL_GPL(blk_trace_remove);
static ssize_t blk_dropped_read(struct file *filp, char __user *buffer,
size_t count, loff_t *ppos)
{
struct blk_trace *bt = filp->private_data;
+ size_t dropped = relay_stats(bt->rchan, RELAY_STATS_BUF_FULL);
char buf[16];
- snprintf(buf, sizeof(buf), "%u\n", atomic_read(&bt->dropped));
+ snprintf(buf, sizeof(buf), "%zu\n", dropped);
return simple_read_from_buffer(buffer, count, ppos, buf, strlen(buf));
}
@@ -338,18 +595,12 @@ static ssize_t blk_msg_write(struct file *filp, const char __user *buffer,
if (count >= BLK_TN_MAX_MSG)
return -EINVAL;
- msg = kmalloc(count + 1, GFP_KERNEL);
- if (msg == NULL)
- return -ENOMEM;
+ msg = memdup_user_nul(buffer, count);
+ if (IS_ERR(msg))
+ return PTR_ERR(msg);
- if (copy_from_user(msg, buffer, count)) {
- kfree(msg);
- return -EFAULT;
- }
-
- msg[count] = '\0';
bt = filp->private_data;
- __trace_note_message(bt, "%s", msg);
+ __blk_trace_note_message(bt, NULL, "%s", msg);
kfree(msg);
return count;
@@ -362,23 +613,6 @@ static const struct file_operations blk_msg_fops = {
.llseek = noop_llseek,
};
-/*
- * Keep track of how many times we encountered a full subbuffer, to aid
- * the user space app in telling how many lost events there were.
- */
-static int blk_subbuf_start_callback(struct rchan_buf *buf, void *subbuf,
- void *prev_subbuf, size_t prev_padding)
-{
- struct blk_trace *bt;
-
- if (!relay_buf_full(buf))
- return 1;
-
- bt = buf->chan->private_data;
- atomic_inc(&bt->dropped);
- return 0;
-}
-
static int blk_remove_buf_file_callback(struct dentry *dentry)
{
debugfs_remove(dentry);
@@ -396,8 +630,7 @@ static struct dentry *blk_create_buf_file_callback(const char *filename,
&relay_file_operations);
}
-static struct rchan_callbacks blk_relay_callbacks = {
- .subbuf_start = blk_subbuf_start_callback,
+static const struct rchan_callbacks blk_relay_callbacks = {
.create_buf_file = blk_create_buf_file_callback,
.remove_buf_file = blk_remove_buf_file_callback,
};
@@ -405,14 +638,9 @@ static struct rchan_callbacks blk_relay_callbacks = {
static void blk_trace_setup_lba(struct blk_trace *bt,
struct block_device *bdev)
{
- struct hd_struct *part = NULL;
-
- if (bdev)
- part = bdev->bd_part;
-
- if (part) {
- bt->start_lba = part->start_sect;
- bt->end_lba = part->start_sect + part->nr_sects;
+ if (bdev) {
+ bt->start_lba = bdev->bd_start_sect;
+ bt->end_lba = bdev->bd_start_sect + bdev_nr_sectors(bdev);
} else {
bt->start_lba = 0;
bt->end_lba = -1ULL;
@@ -422,31 +650,30 @@ static void blk_trace_setup_lba(struct blk_trace *bt,
/*
* Setup everything required to start tracing
*/
-int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
- struct block_device *bdev,
- struct blk_user_trace_setup *buts)
+static struct blk_trace *blk_trace_setup_prepare(struct request_queue *q,
+ char *name, dev_t dev,
+ u32 buf_size, u32 buf_nr,
+ struct block_device *bdev)
{
- struct blk_trace *old_bt, *bt = NULL;
+ struct blk_trace *bt = NULL;
struct dentry *dir = NULL;
- int ret, i;
-
- if (!buts->buf_size || !buts->buf_nr)
- return -EINVAL;
+ int ret;
- strncpy(buts->name, name, BLKTRACE_BDEV_SIZE);
- buts->name[BLKTRACE_BDEV_SIZE - 1] = '\0';
+ lockdep_assert_held(&q->debugfs_mutex);
/*
- * some device names have larger paths - convert the slashes
- * to underscores for this to work as expected
+ * bdev can be NULL, as with scsi-generic, this is a helpful as
+ * we can be.
*/
- for (i = 0; i < strlen(buts->name); i++)
- if (buts->name[i] == '/')
- buts->name[i] = '_';
+ if (rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex))) {
+ pr_warn("Concurrent blktraces are not allowed on %s\n", name);
+ return ERR_PTR(-EBUSY);
+ }
bt = kzalloc(sizeof(*bt), GFP_KERNEL);
if (!bt)
- return -ENOMEM;
+ return ERR_PTR(-ENOMEM);
ret = -ENOMEM;
bt->sequence = alloc_percpu(unsigned long);
@@ -457,48 +684,69 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
if (!bt->msg_data)
goto err;
- ret = -ENOENT;
-
- mutex_lock(&blk_tree_mutex);
- if (!blk_tree_root) {
- blk_tree_root = debugfs_create_dir("block", NULL);
- if (!blk_tree_root) {
- mutex_unlock(&blk_tree_mutex);
- goto err;
- }
- }
- mutex_unlock(&blk_tree_mutex);
-
- dir = debugfs_create_dir(buts->name, blk_tree_root);
+ /*
+ * When tracing the whole disk reuse the existing debugfs directory
+ * created by the block layer on init. For partitions block devices,
+ * and scsi-generic block devices we create a temporary new debugfs
+ * directory that will be removed once the trace ends.
+ */
+ if (bdev && !bdev_is_partition(bdev))
+ dir = q->debugfs_dir;
+ else
+ bt->dir = dir = debugfs_create_dir(name, blk_debugfs_root);
- if (!dir)
+ /*
+ * As blktrace relies on debugfs for its interface the debugfs directory
+ * is required, contrary to the usual mantra of not checking for debugfs
+ * files or directories.
+ */
+ if (IS_ERR_OR_NULL(dir)) {
+ pr_warn("debugfs_dir not present for %s so skipping\n", name);
+ ret = -ENOENT;
goto err;
+ }
- bt->dir = dir;
bt->dev = dev;
- atomic_set(&bt->dropped, 0);
+ INIT_LIST_HEAD(&bt->running_list);
ret = -EIO;
- bt->dropped_file = debugfs_create_file("dropped", 0444, dir, bt,
- &blk_dropped_fops);
- if (!bt->dropped_file)
- goto err;
-
- bt->msg_file = debugfs_create_file("msg", 0222, dir, bt, &blk_msg_fops);
- if (!bt->msg_file)
- goto err;
+ debugfs_create_file("dropped", 0444, dir, bt, &blk_dropped_fops);
+ debugfs_create_file("msg", 0222, dir, bt, &blk_msg_fops);
- bt->rchan = relay_open("trace", dir, buts->buf_size,
- buts->buf_nr, &blk_relay_callbacks, bt);
+ bt->rchan = relay_open("trace", dir, buf_size, buf_nr,
+ &blk_relay_callbacks, bt);
if (!bt->rchan)
goto err;
+ blk_trace_setup_lba(bt, bdev);
+
+ return bt;
+
+err:
+ blk_trace_free(q, bt);
+
+ return ERR_PTR(ret);
+}
+
+static void blk_trace_setup_finalize(struct request_queue *q,
+ char *name, int version,
+ struct blk_trace *bt,
+ struct blk_user_trace_setup2 *buts)
+
+{
+ strscpy_pad(buts->name, name, BLKTRACE_BDEV_SIZE2);
+
+ /*
+ * some device names have larger paths - convert the slashes
+ * to underscores for this to work as expected
+ */
+ strreplace(buts->name, '/', '_');
+
+ bt->version = version;
bt->act_mask = buts->act_mask;
if (!bt->act_mask)
bt->act_mask = (u16) -1;
- blk_trace_setup_lba(bt, bdev);
-
/* overwrite with user settings */
if (buts->start_lba)
bt->start_lba = buts->start_lba;
@@ -508,36 +756,45 @@ int do_blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
bt->pid = buts->pid;
bt->trace_state = Blktrace_setup;
- ret = -EBUSY;
- old_bt = xchg(&q->blk_trace, bt);
- if (old_bt) {
- (void) xchg(&q->blk_trace, old_bt);
- goto err;
- }
-
- if (atomic_inc_return(&blk_probes_ref) == 1)
- blk_register_tracepoints();
-
- return 0;
-err:
- blk_trace_free(bt);
- return ret;
+ rcu_assign_pointer(q->blk_trace, bt);
+ get_probe_ref();
}
int blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
struct block_device *bdev,
char __user *arg)
{
+ struct blk_user_trace_setup2 buts2;
struct blk_user_trace_setup buts;
+ struct blk_trace *bt;
int ret;
ret = copy_from_user(&buts, arg, sizeof(buts));
if (ret)
return -EFAULT;
- ret = do_blk_trace_setup(q, name, dev, bdev, &buts);
- if (ret)
- return ret;
+ if (!buts.buf_size || !buts.buf_nr)
+ return -EINVAL;
+
+ buts2 = (struct blk_user_trace_setup2) {
+ .act_mask = buts.act_mask,
+ .buf_size = buts.buf_size,
+ .buf_nr = buts.buf_nr,
+ .start_lba = buts.start_lba,
+ .end_lba = buts.end_lba,
+ .pid = buts.pid,
+ };
+
+ mutex_lock(&q->debugfs_mutex);
+ bt = blk_trace_setup_prepare(q, name, dev, buts.buf_size, buts.buf_nr,
+ bdev);
+ if (IS_ERR(bt)) {
+ mutex_unlock(&q->debugfs_mutex);
+ return PTR_ERR(bt);
+ }
+ blk_trace_setup_finalize(q, name, 1, bt, &buts2);
+ strcpy(buts.name, buts2.name);
+ mutex_unlock(&q->debugfs_mutex);
if (copy_to_user(arg, &buts, sizeof(buts))) {
blk_trace_remove(q);
@@ -547,19 +804,54 @@ int blk_trace_setup(struct request_queue *q, char *name, dev_t dev,
}
EXPORT_SYMBOL_GPL(blk_trace_setup);
+static int blk_trace_setup2(struct request_queue *q, char *name, dev_t dev,
+ struct block_device *bdev, char __user *arg)
+{
+ struct blk_user_trace_setup2 buts2;
+ struct blk_trace *bt;
+
+ if (copy_from_user(&buts2, arg, sizeof(buts2)))
+ return -EFAULT;
+
+ if (!buts2.buf_size || !buts2.buf_nr)
+ return -EINVAL;
+
+ if (buts2.flags != 0)
+ return -EINVAL;
+
+ mutex_lock(&q->debugfs_mutex);
+ bt = blk_trace_setup_prepare(q, name, dev, buts2.buf_size, buts2.buf_nr,
+ bdev);
+ if (IS_ERR(bt)) {
+ mutex_unlock(&q->debugfs_mutex);
+ return PTR_ERR(bt);
+ }
+ blk_trace_setup_finalize(q, name, 2, bt, &buts2);
+ mutex_unlock(&q->debugfs_mutex);
+
+ if (copy_to_user(arg, &buts2, sizeof(buts2))) {
+ blk_trace_remove(q);
+ return -EFAULT;
+ }
+ return 0;
+}
+
#if defined(CONFIG_COMPAT) && defined(CONFIG_X86_64)
static int compat_blk_trace_setup(struct request_queue *q, char *name,
dev_t dev, struct block_device *bdev,
char __user *arg)
{
- struct blk_user_trace_setup buts;
+ struct blk_user_trace_setup2 buts2;
struct compat_blk_user_trace_setup cbuts;
- int ret;
+ struct blk_trace *bt;
if (copy_from_user(&cbuts, arg, sizeof(cbuts)))
return -EFAULT;
- buts = (struct blk_user_trace_setup) {
+ if (!cbuts.buf_size || !cbuts.buf_nr)
+ return -EINVAL;
+
+ buts2 = (struct blk_user_trace_setup2) {
.act_mask = cbuts.act_mask,
.buf_size = cbuts.buf_size,
.buf_nr = cbuts.buf_nr,
@@ -567,13 +859,18 @@ static int compat_blk_trace_setup(struct request_queue *q, char *name,
.end_lba = cbuts.end_lba,
.pid = cbuts.pid,
};
- memcpy(&buts.name, &cbuts.name, 32);
- ret = do_blk_trace_setup(q, name, dev, bdev, &buts);
- if (ret)
- return ret;
+ mutex_lock(&q->debugfs_mutex);
+ bt = blk_trace_setup_prepare(q, name, dev, buts2.buf_size, buts2.buf_nr,
+ bdev);
+ if (IS_ERR(bt)) {
+ mutex_unlock(&q->debugfs_mutex);
+ return PTR_ERR(bt);
+ }
+ blk_trace_setup_finalize(q, name, 1, bt, &buts2);
+ mutex_unlock(&q->debugfs_mutex);
- if (copy_to_user(arg, &buts.name, 32)) {
+ if (copy_to_user(arg, &buts2.name, ARRAY_SIZE(buts2.name))) {
blk_trace_remove(q);
return -EFAULT;
}
@@ -582,43 +879,41 @@ static int compat_blk_trace_setup(struct request_queue *q, char *name,
}
#endif
-int blk_trace_startstop(struct request_queue *q, int start)
+static int __blk_trace_startstop(struct request_queue *q, int start)
{
- int ret;
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
+ bt = rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex));
if (bt == NULL)
return -EINVAL;
- /*
- * For starting a trace, we can transition from a setup or stopped
- * trace. For stopping a trace, the state must be running
- */
- ret = -EINVAL;
- if (start) {
- if (bt->trace_state == Blktrace_setup ||
- bt->trace_state == Blktrace_stopped) {
- blktrace_seq++;
- smp_mb();
- bt->trace_state = Blktrace_running;
-
- trace_note_time(bt);
- ret = 0;
- }
- } else {
- if (bt->trace_state == Blktrace_running) {
- bt->trace_state = Blktrace_stopped;
- relay_flush(bt->rchan);
- ret = 0;
- }
- }
+ if (start)
+ return blk_trace_start(bt);
+ else
+ return blk_trace_stop(bt);
+}
+
+int blk_trace_startstop(struct request_queue *q, int start)
+{
+ int ret;
+
+ mutex_lock(&q->debugfs_mutex);
+ ret = __blk_trace_startstop(q, start);
+ mutex_unlock(&q->debugfs_mutex);
return ret;
}
EXPORT_SYMBOL_GPL(blk_trace_startstop);
+/*
+ * When reading or writing the blktrace sysfs files, the references to the
+ * opened sysfs or device files should prevent the underlying block device
+ * from being removed. So no further delete protection is really needed.
+ */
+
/**
- * blk_trace_ioctl: - handle the ioctls associated with tracing
+ * blk_trace_ioctl - handle the ioctls associated with tracing
* @bdev: the block device
* @cmd: the ioctl cmd
* @arg: the argument data, if any
@@ -626,29 +921,28 @@ EXPORT_SYMBOL_GPL(blk_trace_startstop);
**/
int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg)
{
- struct request_queue *q;
+ struct request_queue *q = bdev_get_queue(bdev);
int ret, start = 0;
char b[BDEVNAME_SIZE];
- q = bdev_get_queue(bdev);
- if (!q)
- return -ENXIO;
-
- mutex_lock(&bdev->bd_mutex);
-
switch (cmd) {
+ case BLKTRACESETUP2:
+ snprintf(b, sizeof(b), "%pg", bdev);
+ ret = blk_trace_setup2(q, b, bdev->bd_dev, bdev, arg);
+ break;
case BLKTRACESETUP:
- bdevname(bdev, b);
+ snprintf(b, sizeof(b), "%pg", bdev);
ret = blk_trace_setup(q, b, bdev->bd_dev, bdev, arg);
break;
#if defined(CONFIG_COMPAT) && defined(CONFIG_X86_64)
case BLKTRACESETUP32:
- bdevname(bdev, b);
+ snprintf(b, sizeof(b), "%pg", bdev);
ret = compat_blk_trace_setup(q, b, bdev->bd_dev, bdev, arg);
break;
#endif
case BLKTRACESTART:
start = 1;
+ fallthrough;
case BLKTRACESTOP:
ret = blk_trace_startstop(q, start);
break;
@@ -659,22 +953,51 @@ int blk_trace_ioctl(struct block_device *bdev, unsigned cmd, char __user *arg)
ret = -ENOTTY;
break;
}
-
- mutex_unlock(&bdev->bd_mutex);
return ret;
}
/**
- * blk_trace_shutdown: - stop and cleanup trace structures
+ * blk_trace_shutdown - stop and cleanup trace structures
* @q: the request queue associated with the device
*
**/
void blk_trace_shutdown(struct request_queue *q)
{
- if (q->blk_trace) {
- blk_trace_startstop(q, 0);
- blk_trace_remove(q);
- }
+ if (rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex)))
+ __blk_trace_remove(q);
+}
+
+#ifdef CONFIG_BLK_CGROUP
+static u64 blk_trace_bio_get_cgid(struct request_queue *q, struct bio *bio)
+{
+ struct cgroup_subsys_state *blkcg_css;
+ struct blk_trace *bt;
+
+ /* We don't use the 'bt' value here except as an optimization... */
+ bt = rcu_dereference_protected(q->blk_trace, 1);
+ if (!bt || !(blk_tracer_flags.val & TRACE_BLK_OPT_CGROUP))
+ return 0;
+
+ blkcg_css = bio_blkcg_css(bio);
+ if (!blkcg_css)
+ return 0;
+ return cgroup_id(blkcg_css->cgroup);
+}
+#else
+static u64 blk_trace_bio_get_cgid(struct request_queue *q, struct bio *bio)
+{
+ return 0;
+}
+#endif
+
+static u64
+blk_trace_request_get_cgid(struct request *rq)
+{
+ if (!rq->bio)
+ return 0;
+ /* Use the first bio */
+ return blk_trace_bio_get_cgid(rq->q, rq->bio);
}
/*
@@ -683,63 +1006,83 @@ void blk_trace_shutdown(struct request_queue *q)
/**
* blk_add_trace_rq - Add a trace for a request oriented action
- * @q: queue the io is for
* @rq: the source request
+ * @error: return status to log
+ * @nr_bytes: number of completed bytes
* @what: the action
+ * @cgid: the cgroup info
*
* Description:
* Records an action against a request. Will log the bio offset + size.
*
**/
-static void blk_add_trace_rq(struct request_queue *q, struct request *rq,
- u32 what)
+static void blk_add_trace_rq(struct request *rq, blk_status_t error,
+ unsigned int nr_bytes, u64 what, u64 cgid)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
- if (likely(!bt))
+ rcu_read_lock();
+ bt = rcu_dereference(rq->q->blk_trace);
+ if (likely(!bt)) {
+ rcu_read_unlock();
return;
+ }
- if (rq->cmd_type == REQ_TYPE_BLOCK_PC) {
+ if (blk_rq_is_passthrough(rq))
what |= BLK_TC_ACT(BLK_TC_PC);
- __blk_add_trace(bt, 0, blk_rq_bytes(rq), rq->cmd_flags,
- what, rq->errors, rq->cmd_len, rq->cmd);
- } else {
+ else
what |= BLK_TC_ACT(BLK_TC_FS);
- __blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq),
- rq->cmd_flags, what, rq->errors, 0, NULL);
- }
+
+ __blk_add_trace(bt, blk_rq_trace_sector(rq), nr_bytes, rq->cmd_flags,
+ what, blk_status_to_errno(error), 0, NULL, cgid);
+ rcu_read_unlock();
+}
+
+static void blk_add_trace_rq_insert(void *ignore, struct request *rq)
+{
+ blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_INSERT,
+ blk_trace_request_get_cgid(rq));
}
-static void blk_add_trace_rq_abort(void *ignore,
- struct request_queue *q, struct request *rq)
+static void blk_add_trace_rq_issue(void *ignore, struct request *rq)
{
- blk_add_trace_rq(q, rq, BLK_TA_ABORT);
+ blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_ISSUE,
+ blk_trace_request_get_cgid(rq));
}
-static void blk_add_trace_rq_insert(void *ignore,
- struct request_queue *q, struct request *rq)
+static void blk_add_trace_rq_merge(void *ignore, struct request *rq)
{
- blk_add_trace_rq(q, rq, BLK_TA_INSERT);
+ blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_BACKMERGE,
+ blk_trace_request_get_cgid(rq));
}
-static void blk_add_trace_rq_issue(void *ignore,
- struct request_queue *q, struct request *rq)
+static void blk_add_trace_rq_requeue(void *ignore, struct request *rq)
{
- blk_add_trace_rq(q, rq, BLK_TA_ISSUE);
+ blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_REQUEUE,
+ blk_trace_request_get_cgid(rq));
}
-static void blk_add_trace_rq_requeue(void *ignore,
- struct request_queue *q,
- struct request *rq)
+static void blk_add_trace_rq_complete(void *ignore, struct request *rq,
+ blk_status_t error, unsigned int nr_bytes)
{
- blk_add_trace_rq(q, rq, BLK_TA_REQUEUE);
+ blk_add_trace_rq(rq, error, nr_bytes, BLK_TA_COMPLETE,
+ blk_trace_request_get_cgid(rq));
}
-static void blk_add_trace_rq_complete(void *ignore,
- struct request_queue *q,
- struct request *rq)
+static void blk_add_trace_zone_update_request(void *ignore, struct request *rq)
{
- blk_add_trace_rq(q, rq, BLK_TA_COMPLETE);
+ struct blk_trace *bt;
+
+ rcu_read_lock();
+ bt = rcu_dereference(rq->q->blk_trace);
+ if (likely(!bt) || bt->version < 2) {
+ rcu_read_unlock();
+ return;
+ }
+ rcu_read_unlock();
+
+ blk_add_trace_rq(rq, 0, blk_rq_bytes(rq), BLK_TA_ZONE_APPEND,
+ blk_trace_request_get_cgid(rq));
}
/**
@@ -754,162 +1097,171 @@ static void blk_add_trace_rq_complete(void *ignore,
*
**/
static void blk_add_trace_bio(struct request_queue *q, struct bio *bio,
- u32 what, int error)
+ u64 what, int error)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
- if (likely(!bt))
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
+ if (likely(!bt)) {
+ rcu_read_unlock();
return;
+ }
- if (!error && !bio_flagged(bio, BIO_UPTODATE))
- error = EIO;
-
- __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw, what,
- error, 0, NULL);
-}
-
-static void blk_add_trace_bio_bounce(void *ignore,
- struct request_queue *q, struct bio *bio)
-{
- blk_add_trace_bio(q, bio, BLK_TA_BOUNCE, 0);
+ __blk_add_trace(bt, bio->bi_iter.bi_sector, bio->bi_iter.bi_size,
+ bio->bi_opf, what, error, 0, NULL,
+ blk_trace_bio_get_cgid(q, bio));
+ rcu_read_unlock();
}
static void blk_add_trace_bio_complete(void *ignore,
- struct request_queue *q, struct bio *bio,
- int error)
+ struct request_queue *q, struct bio *bio)
{
- blk_add_trace_bio(q, bio, BLK_TA_COMPLETE, error);
+ blk_add_trace_bio(q, bio, BLK_TA_COMPLETE,
+ blk_status_to_errno(bio->bi_status));
}
-static void blk_add_trace_bio_backmerge(void *ignore,
- struct request_queue *q,
- struct request *rq,
- struct bio *bio)
+static void blk_add_trace_bio_backmerge(void *ignore, struct bio *bio)
{
- blk_add_trace_bio(q, bio, BLK_TA_BACKMERGE, 0);
+ blk_add_trace_bio(bio->bi_bdev->bd_disk->queue, bio, BLK_TA_BACKMERGE,
+ 0);
}
-static void blk_add_trace_bio_frontmerge(void *ignore,
- struct request_queue *q,
- struct request *rq,
- struct bio *bio)
+static void blk_add_trace_bio_frontmerge(void *ignore, struct bio *bio)
{
- blk_add_trace_bio(q, bio, BLK_TA_FRONTMERGE, 0);
+ blk_add_trace_bio(bio->bi_bdev->bd_disk->queue, bio, BLK_TA_FRONTMERGE,
+ 0);
}
-static void blk_add_trace_bio_queue(void *ignore,
- struct request_queue *q, struct bio *bio)
+static void blk_add_trace_bio_queue(void *ignore, struct bio *bio)
{
- blk_add_trace_bio(q, bio, BLK_TA_QUEUE, 0);
+ blk_add_trace_bio(bio->bi_bdev->bd_disk->queue, bio, BLK_TA_QUEUE, 0);
}
-static void blk_add_trace_getrq(void *ignore,
- struct request_queue *q,
- struct bio *bio, int rw)
+static void blk_add_trace_getrq(void *ignore, struct bio *bio)
{
- if (bio)
- blk_add_trace_bio(q, bio, BLK_TA_GETRQ, 0);
- else {
- struct blk_trace *bt = q->blk_trace;
-
- if (bt)
- __blk_add_trace(bt, 0, 0, rw, BLK_TA_GETRQ, 0, 0, NULL);
- }
-}
-
-
-static void blk_add_trace_sleeprq(void *ignore,
- struct request_queue *q,
- struct bio *bio, int rw)
-{
- if (bio)
- blk_add_trace_bio(q, bio, BLK_TA_SLEEPRQ, 0);
- else {
- struct blk_trace *bt = q->blk_trace;
-
- if (bt)
- __blk_add_trace(bt, 0, 0, rw, BLK_TA_SLEEPRQ,
- 0, 0, NULL);
- }
+ blk_add_trace_bio(bio->bi_bdev->bd_disk->queue, bio, BLK_TA_GETRQ, 0);
}
static void blk_add_trace_plug(void *ignore, struct request_queue *q)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
if (bt)
- __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL);
+ __blk_add_trace(bt, 0, 0, 0, BLK_TA_PLUG, 0, 0, NULL, 0);
+ rcu_read_unlock();
}
static void blk_add_trace_unplug(void *ignore, struct request_queue *q,
unsigned int depth, bool explicit)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
if (bt) {
__be64 rpdu = cpu_to_be64(depth);
- u32 what;
+ u64 what;
if (explicit)
what = BLK_TA_UNPLUG_IO;
else
what = BLK_TA_UNPLUG_TIMER;
- __blk_add_trace(bt, 0, 0, 0, what, 0, sizeof(rpdu), &rpdu);
+ __blk_add_trace(bt, 0, 0, 0, what, 0, sizeof(rpdu), &rpdu, 0);
}
+ rcu_read_unlock();
+}
+
+static void blk_add_trace_zone_plug(void *ignore, struct request_queue *q,
+ unsigned int zno, sector_t sector,
+ unsigned int sectors)
+{
+ struct blk_trace *bt;
+
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
+ if (bt && bt->version >= 2)
+ __blk_add_trace(bt, sector, sectors << SECTOR_SHIFT, 0,
+ BLK_TA_ZONE_PLUG, 0, 0, NULL, 0);
+ rcu_read_unlock();
+
+ return;
+}
+
+static void blk_add_trace_zone_unplug(void *ignore, struct request_queue *q,
+ unsigned int zno, sector_t sector,
+ unsigned int sectors)
+{
+ struct blk_trace *bt;
+
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
+ if (bt && bt->version >= 2)
+ __blk_add_trace(bt, sector, sectors << SECTOR_SHIFT, 0,
+ BLK_TA_ZONE_UNPLUG, 0, 0, NULL, 0);
+ rcu_read_unlock();
+ return;
}
-static void blk_add_trace_split(void *ignore,
- struct request_queue *q, struct bio *bio,
- unsigned int pdu)
+static void blk_add_trace_split(void *ignore, struct bio *bio, unsigned int pdu)
{
- struct blk_trace *bt = q->blk_trace;
+ struct request_queue *q = bio->bi_bdev->bd_disk->queue;
+ struct blk_trace *bt;
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
if (bt) {
__be64 rpdu = cpu_to_be64(pdu);
- __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw,
- BLK_TA_SPLIT, !bio_flagged(bio, BIO_UPTODATE),
- sizeof(rpdu), &rpdu);
+ __blk_add_trace(bt, bio->bi_iter.bi_sector,
+ bio->bi_iter.bi_size, bio->bi_opf, BLK_TA_SPLIT,
+ blk_status_to_errno(bio->bi_status),
+ sizeof(rpdu), &rpdu,
+ blk_trace_bio_get_cgid(q, bio));
}
+ rcu_read_unlock();
}
/**
* blk_add_trace_bio_remap - Add a trace for a bio-remap operation
* @ignore: trace callback data parameter (not used)
- * @q: queue the io is for
* @bio: the source bio
- * @dev: target device
+ * @dev: source device
* @from: source sector
*
- * Description:
- * Device mapper or raid target sometimes need to split a bio because
- * it spans a stripe (or similar). Add a trace for that action.
- *
+ * Called after a bio is remapped to a different device and/or sector.
**/
-static void blk_add_trace_bio_remap(void *ignore,
- struct request_queue *q, struct bio *bio,
- dev_t dev, sector_t from)
+static void blk_add_trace_bio_remap(void *ignore, struct bio *bio, dev_t dev,
+ sector_t from)
{
- struct blk_trace *bt = q->blk_trace;
+ struct request_queue *q = bio->bi_bdev->bd_disk->queue;
+ struct blk_trace *bt;
struct blk_io_trace_remap r;
- if (likely(!bt))
+ rcu_read_lock();
+ bt = rcu_dereference(q->blk_trace);
+ if (likely(!bt)) {
+ rcu_read_unlock();
return;
+ }
r.device_from = cpu_to_be32(dev);
- r.device_to = cpu_to_be32(bio->bi_bdev->bd_dev);
+ r.device_to = cpu_to_be32(bio_dev(bio));
r.sector_from = cpu_to_be64(from);
- __blk_add_trace(bt, bio->bi_sector, bio->bi_size, bio->bi_rw,
- BLK_TA_REMAP, !bio_flagged(bio, BIO_UPTODATE),
- sizeof(r), &r);
+ __blk_add_trace(bt, bio->bi_iter.bi_sector, bio->bi_iter.bi_size,
+ bio->bi_opf, BLK_TA_REMAP,
+ blk_status_to_errno(bio->bi_status),
+ sizeof(r), &r, blk_trace_bio_get_cgid(q, bio));
+ rcu_read_unlock();
}
/**
* blk_add_trace_rq_remap - Add a trace for a request-remap operation
* @ignore: trace callback data parameter (not used)
- * @q: queue the io is for
* @rq: the source request
* @dev: target device
* @from: source sector
@@ -919,29 +1271,31 @@ static void blk_add_trace_bio_remap(void *ignore,
* Add a trace for that action.
*
**/
-static void blk_add_trace_rq_remap(void *ignore,
- struct request_queue *q,
- struct request *rq, dev_t dev,
+static void blk_add_trace_rq_remap(void *ignore, struct request *rq, dev_t dev,
sector_t from)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
struct blk_io_trace_remap r;
- if (likely(!bt))
+ rcu_read_lock();
+ bt = rcu_dereference(rq->q->blk_trace);
+ if (likely(!bt)) {
+ rcu_read_unlock();
return;
+ }
r.device_from = cpu_to_be32(dev);
- r.device_to = cpu_to_be32(disk_devt(rq->rq_disk));
+ r.device_to = cpu_to_be32(disk_devt(rq->q->disk));
r.sector_from = cpu_to_be64(from);
__blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq),
- rq_data_dir(rq), BLK_TA_REMAP, !!rq->errors,
- sizeof(r), &r);
+ rq->cmd_flags, BLK_TA_REMAP, 0,
+ sizeof(r), &r, blk_trace_request_get_cgid(rq));
+ rcu_read_unlock();
}
/**
* blk_add_driver_data - Add binary message with driver-specific data
- * @q: queue the io is for
* @rq: io request
* @data: driver-specific data
* @len: length of driver-specific data
@@ -950,21 +1304,21 @@ static void blk_add_trace_rq_remap(void *ignore,
* Some drivers might want to write driver-specific data per request.
*
**/
-void blk_add_driver_data(struct request_queue *q,
- struct request *rq,
- void *data, size_t len)
+void blk_add_driver_data(struct request *rq, void *data, size_t len)
{
- struct blk_trace *bt = q->blk_trace;
+ struct blk_trace *bt;
- if (likely(!bt))
+ rcu_read_lock();
+ bt = rcu_dereference(rq->q->blk_trace);
+ if (likely(!bt)) {
+ rcu_read_unlock();
return;
+ }
- if (rq->cmd_type == REQ_TYPE_BLOCK_PC)
- __blk_add_trace(bt, 0, blk_rq_bytes(rq), 0,
- BLK_TA_DRV_DATA, rq->errors, len, data);
- else
- __blk_add_trace(bt, blk_rq_pos(rq), blk_rq_bytes(rq), 0,
- BLK_TA_DRV_DATA, rq->errors, len, data);
+ __blk_add_trace(bt, blk_rq_trace_sector(rq), blk_rq_bytes(rq), 0,
+ BLK_TA_DRV_DATA, 0, len, data,
+ blk_trace_request_get_cgid(rq));
+ rcu_read_unlock();
}
EXPORT_SYMBOL_GPL(blk_add_driver_data);
@@ -972,18 +1326,16 @@ static void blk_register_tracepoints(void)
{
int ret;
- ret = register_trace_block_rq_abort(blk_add_trace_rq_abort, NULL);
- WARN_ON(ret);
ret = register_trace_block_rq_insert(blk_add_trace_rq_insert, NULL);
WARN_ON(ret);
ret = register_trace_block_rq_issue(blk_add_trace_rq_issue, NULL);
WARN_ON(ret);
+ ret = register_trace_block_rq_merge(blk_add_trace_rq_merge, NULL);
+ WARN_ON(ret);
ret = register_trace_block_rq_requeue(blk_add_trace_rq_requeue, NULL);
WARN_ON(ret);
ret = register_trace_block_rq_complete(blk_add_trace_rq_complete, NULL);
WARN_ON(ret);
- ret = register_trace_block_bio_bounce(blk_add_trace_bio_bounce, NULL);
- WARN_ON(ret);
ret = register_trace_block_bio_complete(blk_add_trace_bio_complete, NULL);
WARN_ON(ret);
ret = register_trace_block_bio_backmerge(blk_add_trace_bio_backmerge, NULL);
@@ -994,7 +1346,14 @@ static void blk_register_tracepoints(void)
WARN_ON(ret);
ret = register_trace_block_getrq(blk_add_trace_getrq, NULL);
WARN_ON(ret);
- ret = register_trace_block_sleeprq(blk_add_trace_sleeprq, NULL);
+ ret = register_trace_blk_zone_append_update_request_bio(
+ blk_add_trace_zone_update_request, NULL);
+ WARN_ON(ret);
+ ret = register_trace_disk_zone_wplug_add_bio(blk_add_trace_zone_plug,
+ NULL);
+ WARN_ON(ret);
+ ret = register_trace_blk_zone_wplug_bio(blk_add_trace_zone_unplug,
+ NULL);
WARN_ON(ret);
ret = register_trace_block_plug(blk_add_trace_plug, NULL);
WARN_ON(ret);
@@ -1015,18 +1374,20 @@ static void blk_unregister_tracepoints(void)
unregister_trace_block_split(blk_add_trace_split, NULL);
unregister_trace_block_unplug(blk_add_trace_unplug, NULL);
unregister_trace_block_plug(blk_add_trace_plug, NULL);
- unregister_trace_block_sleeprq(blk_add_trace_sleeprq, NULL);
+ unregister_trace_blk_zone_wplug_bio(blk_add_trace_zone_unplug, NULL);
+ unregister_trace_disk_zone_wplug_add_bio(blk_add_trace_zone_plug, NULL);
+ unregister_trace_blk_zone_append_update_request_bio(
+ blk_add_trace_zone_update_request, NULL);
unregister_trace_block_getrq(blk_add_trace_getrq, NULL);
unregister_trace_block_bio_queue(blk_add_trace_bio_queue, NULL);
unregister_trace_block_bio_frontmerge(blk_add_trace_bio_frontmerge, NULL);
unregister_trace_block_bio_backmerge(blk_add_trace_bio_backmerge, NULL);
unregister_trace_block_bio_complete(blk_add_trace_bio_complete, NULL);
- unregister_trace_block_bio_bounce(blk_add_trace_bio_bounce, NULL);
unregister_trace_block_rq_complete(blk_add_trace_rq_complete, NULL);
unregister_trace_block_rq_requeue(blk_add_trace_rq_requeue, NULL);
+ unregister_trace_block_rq_merge(blk_add_trace_rq_merge, NULL);
unregister_trace_block_rq_issue(blk_add_trace_rq_issue, NULL);
unregister_trace_block_rq_insert(blk_add_trace_rq_insert, NULL);
- unregister_trace_block_rq_abort(blk_add_trace_rq_abort, NULL);
tracepoint_synchronize_unregister();
}
@@ -1035,12 +1396,12 @@ static void blk_unregister_tracepoints(void)
* struct blk_io_tracer formatting routines
*/
-static void fill_rwbs(char *rwbs, const struct blk_io_trace *t)
+static void fill_rwbs(char *rwbs, const struct blk_io_trace2 *t)
{
int i = 0;
int tc = t->action >> BLK_TC_SHIFT;
- if (t->action == BLK_TN_MESSAGE) {
+ if ((t->action & ~__BLK_TN_CGROUP) == BLK_TN_MESSAGE) {
rwbs[i++] = 'N';
goto out;
}
@@ -1050,7 +1411,10 @@ static void fill_rwbs(char *rwbs, const struct blk_io_trace *t)
if (tc & BLK_TC_DISCARD)
rwbs[i++] = 'D';
- else if (tc & BLK_TC_WRITE)
+ else if (tc & BLK_TC_WRITE_ZEROES) {
+ rwbs[i++] = 'W';
+ rwbs[i++] = 'Z';
+ } else if (tc & BLK_TC_WRITE)
rwbs[i++] = 'W';
else if (t->bytes)
rwbs[i++] = 'R';
@@ -1070,14 +1434,24 @@ out:
}
static inline
-const struct blk_io_trace *te_blk_io_trace(const struct trace_entry *ent)
+const struct blk_io_trace2 *te_blk_io_trace(const struct trace_entry *ent)
+{
+ return (const struct blk_io_trace2 *)ent;
+}
+
+static inline const void *pdu_start(const struct trace_entry *ent, bool has_cg)
+{
+ return (void *)(te_blk_io_trace(ent) + 1) + (has_cg ? sizeof(u64) : 0);
+}
+
+static inline u64 t_cgid(const struct trace_entry *ent)
{
- return (const struct blk_io_trace *)ent;
+ return *(u64 *)(te_blk_io_trace(ent) + 1);
}
-static inline const void *pdu_start(const struct trace_entry *ent)
+static inline int pdu_real_len(const struct trace_entry *ent, bool has_cg)
{
- return te_blk_io_trace(ent) + 1;
+ return te_blk_io_trace(ent)->pdu_len - (has_cg ? sizeof(u64) : 0);
}
static inline u32 t_action(const struct trace_entry *ent)
@@ -1105,62 +1479,86 @@ static inline __u16 t_error(const struct trace_entry *ent)
return te_blk_io_trace(ent)->error;
}
-static __u64 get_pdu_int(const struct trace_entry *ent)
+static __u64 get_pdu_int(const struct trace_entry *ent, bool has_cg)
{
- const __u64 *val = pdu_start(ent);
+ const __be64 *val = pdu_start(ent, has_cg);
return be64_to_cpu(*val);
}
-static void get_pdu_remap(const struct trace_entry *ent,
- struct blk_io_trace_remap *r)
-{
- const struct blk_io_trace_remap *__r = pdu_start(ent);
- __u64 sector_from = __r->sector_from;
-
- r->device_from = be32_to_cpu(__r->device_from);
- r->device_to = be32_to_cpu(__r->device_to);
- r->sector_from = be64_to_cpu(sector_from);
-}
-
-typedef int (blk_log_action_t) (struct trace_iterator *iter, const char *act);
+typedef void (blk_log_action_t) (struct trace_iterator *iter, const char *act,
+ bool has_cg);
-static int blk_log_action_classic(struct trace_iterator *iter, const char *act)
+static void blk_log_action_classic(struct trace_iterator *iter, const char *act,
+ bool has_cg)
{
char rwbs[RWBS_LEN];
unsigned long long ts = iter->ts;
unsigned long nsec_rem = do_div(ts, NSEC_PER_SEC);
unsigned secs = (unsigned long)ts;
- const struct blk_io_trace *t = te_blk_io_trace(iter->ent);
+ const struct blk_io_trace2 *t = te_blk_io_trace(iter->ent);
fill_rwbs(rwbs, t);
- return trace_seq_printf(&iter->seq,
- "%3d,%-3d %2d %5d.%09lu %5u %2s %3s ",
- MAJOR(t->device), MINOR(t->device), iter->cpu,
- secs, nsec_rem, iter->ent->pid, act, rwbs);
+ trace_seq_printf(&iter->seq,
+ "%3d,%-3d %2d %5d.%09lu %5u %2s %3s ",
+ MAJOR(t->device), MINOR(t->device), iter->cpu,
+ secs, nsec_rem, iter->ent->pid, act, rwbs);
}
-static int blk_log_action(struct trace_iterator *iter, const char *act)
+static void blk_log_action(struct trace_iterator *iter, const char *act,
+ bool has_cg)
{
char rwbs[RWBS_LEN];
- const struct blk_io_trace *t = te_blk_io_trace(iter->ent);
+ const struct blk_io_trace2 *t = te_blk_io_trace(iter->ent);
fill_rwbs(rwbs, t);
- return trace_seq_printf(&iter->seq, "%3d,%-3d %2s %3s ",
- MAJOR(t->device), MINOR(t->device), act, rwbs);
+ if (has_cg) {
+ u64 id = t_cgid(iter->ent);
+
+ if (blk_tracer_flags.val & TRACE_BLK_OPT_CGNAME) {
+ char blkcg_name_buf[NAME_MAX + 1] = "<...>";
+
+ cgroup_path_from_kernfs_id(id, blkcg_name_buf,
+ sizeof(blkcg_name_buf));
+ trace_seq_printf(&iter->seq, "%3d,%-3d %s %2s %3s ",
+ MAJOR(t->device), MINOR(t->device),
+ blkcg_name_buf, act, rwbs);
+ } else {
+ /*
+ * The cgid portion used to be "INO,GEN". Userland
+ * builds a FILEID_INO32_GEN fid out of them and
+ * opens the cgroup using open_by_handle_at(2).
+ * While 32bit ino setups are still the same, 64bit
+ * ones now use the 64bit ino as the whole ID and
+ * no longer use generation.
+ *
+ * Regardless of the content, always output
+ * "LOW32,HIGH32" so that FILEID_INO32_GEN fid can
+ * be mapped back to @id on both 64 and 32bit ino
+ * setups. See __kernfs_fh_to_dentry().
+ */
+ trace_seq_printf(&iter->seq,
+ "%3d,%-3d %llx,%-llx %2s %3s ",
+ MAJOR(t->device), MINOR(t->device),
+ id & U32_MAX, id >> 32, act, rwbs);
+ }
+ } else
+ trace_seq_printf(&iter->seq, "%3d,%-3d %2s %3s ",
+ MAJOR(t->device), MINOR(t->device), act, rwbs);
}
-static int blk_log_dump_pdu(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_dump_pdu(struct trace_seq *s,
+ const struct trace_entry *ent, bool has_cg)
{
const unsigned char *pdu_buf;
int pdu_len;
- int i, end, ret;
+ int i, end;
- pdu_buf = pdu_start(ent);
- pdu_len = te_blk_io_trace(ent)->pdu_len;
+ pdu_buf = pdu_start(ent, has_cg);
+ pdu_len = pdu_real_len(ent, has_cg);
if (!pdu_len)
- return 1;
+ return;
/* find the last zero that needs to be printed */
for (end = pdu_len - 1; end >= 0; end--)
@@ -1168,119 +1566,108 @@ static int blk_log_dump_pdu(struct trace_seq *s, const struct trace_entry *ent)
break;
end++;
- if (!trace_seq_putc(s, '('))
- return 0;
+ trace_seq_putc(s, '(');
for (i = 0; i < pdu_len; i++) {
- ret = trace_seq_printf(s, "%s%02x",
- i == 0 ? "" : " ", pdu_buf[i]);
- if (!ret)
- return ret;
+ trace_seq_printf(s, "%s%02x",
+ i == 0 ? "" : " ", pdu_buf[i]);
/*
- * stop when the rest is just zeroes and indicate so
+ * stop when the rest is just zeros and indicate so
* with a ".." appended
*/
- if (i == end && end != pdu_len - 1)
- return trace_seq_puts(s, " ..) ");
+ if (i == end && end != pdu_len - 1) {
+ trace_seq_puts(s, " ..) ");
+ return;
+ }
}
- return trace_seq_puts(s, ") ");
+ trace_seq_puts(s, ") ");
}
-static int blk_log_generic(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_generic(struct trace_seq *s, const struct trace_entry *ent, bool has_cg)
{
char cmd[TASK_COMM_LEN];
trace_find_cmdline(ent->pid, cmd);
if (t_action(ent) & BLK_TC_ACT(BLK_TC_PC)) {
- int ret;
-
- ret = trace_seq_printf(s, "%u ", t_bytes(ent));
- if (!ret)
- return 0;
- ret = blk_log_dump_pdu(s, ent);
- if (!ret)
- return 0;
- return trace_seq_printf(s, "[%s]\n", cmd);
+ trace_seq_printf(s, "%u ", t_bytes(ent));
+ blk_log_dump_pdu(s, ent, has_cg);
+ trace_seq_printf(s, "[%s]\n", cmd);
} else {
if (t_sec(ent))
- return trace_seq_printf(s, "%llu + %u [%s]\n",
+ trace_seq_printf(s, "%llu + %u [%s]\n",
t_sector(ent), t_sec(ent), cmd);
- return trace_seq_printf(s, "[%s]\n", cmd);
+ else
+ trace_seq_printf(s, "[%s]\n", cmd);
}
}
-static int blk_log_with_error(struct trace_seq *s,
- const struct trace_entry *ent)
+static void blk_log_with_error(struct trace_seq *s,
+ const struct trace_entry *ent, bool has_cg)
{
if (t_action(ent) & BLK_TC_ACT(BLK_TC_PC)) {
- int ret;
-
- ret = blk_log_dump_pdu(s, ent);
- if (ret)
- return trace_seq_printf(s, "[%d]\n", t_error(ent));
- return 0;
+ blk_log_dump_pdu(s, ent, has_cg);
+ trace_seq_printf(s, "[%d]\n", t_error(ent));
} else {
if (t_sec(ent))
- return trace_seq_printf(s, "%llu + %u [%d]\n",
- t_sector(ent),
- t_sec(ent), t_error(ent));
- return trace_seq_printf(s, "%llu [%d]\n",
- t_sector(ent), t_error(ent));
+ trace_seq_printf(s, "%llu + %u [%d]\n",
+ t_sector(ent),
+ t_sec(ent), t_error(ent));
+ else
+ trace_seq_printf(s, "%llu [%d]\n",
+ t_sector(ent), t_error(ent));
}
}
-static int blk_log_remap(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_remap(struct trace_seq *s, const struct trace_entry *ent, bool has_cg)
{
- struct blk_io_trace_remap r = { .device_from = 0, };
+ const struct blk_io_trace_remap *__r = pdu_start(ent, has_cg);
- get_pdu_remap(ent, &r);
- return trace_seq_printf(s, "%llu + %u <- (%d,%d) %llu\n",
- t_sector(ent), t_sec(ent),
- MAJOR(r.device_from), MINOR(r.device_from),
- (unsigned long long)r.sector_from);
+ trace_seq_printf(s, "%llu + %u <- (%d,%d) %llu\n",
+ t_sector(ent), t_sec(ent),
+ MAJOR(be32_to_cpu(__r->device_from)),
+ MINOR(be32_to_cpu(__r->device_from)),
+ be64_to_cpu(__r->sector_from));
}
-static int blk_log_plug(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_plug(struct trace_seq *s, const struct trace_entry *ent, bool has_cg)
{
char cmd[TASK_COMM_LEN];
trace_find_cmdline(ent->pid, cmd);
- return trace_seq_printf(s, "[%s]\n", cmd);
+ trace_seq_printf(s, "[%s]\n", cmd);
}
-static int blk_log_unplug(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_unplug(struct trace_seq *s, const struct trace_entry *ent, bool has_cg)
{
char cmd[TASK_COMM_LEN];
trace_find_cmdline(ent->pid, cmd);
- return trace_seq_printf(s, "[%s] %llu\n", cmd, get_pdu_int(ent));
+ trace_seq_printf(s, "[%s] %llu\n", cmd, get_pdu_int(ent, has_cg));
}
-static int blk_log_split(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_split(struct trace_seq *s, const struct trace_entry *ent, bool has_cg)
{
char cmd[TASK_COMM_LEN];
trace_find_cmdline(ent->pid, cmd);
- return trace_seq_printf(s, "%llu / %llu [%s]\n", t_sector(ent),
- get_pdu_int(ent), cmd);
+ trace_seq_printf(s, "%llu / %llu [%s]\n", t_sector(ent),
+ get_pdu_int(ent, has_cg), cmd);
}
-static int blk_log_msg(struct trace_seq *s, const struct trace_entry *ent)
+static void blk_log_msg(struct trace_seq *s, const struct trace_entry *ent,
+ bool has_cg)
{
- int ret;
- const struct blk_io_trace *t = te_blk_io_trace(ent);
- ret = trace_seq_putmem(s, t + 1, t->pdu_len);
- if (ret)
- return trace_seq_putc(s, '\n');
- return ret;
+ trace_seq_putmem(s, pdu_start(ent, has_cg),
+ pdu_real_len(ent, has_cg));
+ trace_seq_putc(s, '\n');
}
/*
@@ -1319,7 +1706,8 @@ static void blk_tracer_reset(struct trace_array *tr)
static const struct {
const char *act[2];
- int (*print)(struct trace_seq *s, const struct trace_entry *ent);
+ void (*print)(struct trace_seq *s, const struct trace_entry *ent,
+ bool has_cg);
} what2act[] = {
[__BLK_TA_QUEUE] = {{ "Q", "queue" }, blk_log_generic },
[__BLK_TA_BACKMERGE] = {{ "M", "backmerge" }, blk_log_generic },
@@ -1334,41 +1722,40 @@ static const struct {
[__BLK_TA_UNPLUG_TIMER] = {{ "UT", "unplug_timer" }, blk_log_unplug },
[__BLK_TA_INSERT] = {{ "I", "insert" }, blk_log_generic },
[__BLK_TA_SPLIT] = {{ "X", "split" }, blk_log_split },
- [__BLK_TA_BOUNCE] = {{ "B", "bounce" }, blk_log_generic },
[__BLK_TA_REMAP] = {{ "A", "remap" }, blk_log_remap },
};
static enum print_line_t print_one_line(struct trace_iterator *iter,
bool classic)
{
+ struct trace_array *tr = iter->tr;
struct trace_seq *s = &iter->seq;
- const struct blk_io_trace *t;
+ const struct blk_io_trace2 *t;
u16 what;
- int ret;
bool long_act;
blk_log_action_t *log_action;
+ bool has_cg;
t = te_blk_io_trace(iter->ent);
- what = t->action & ((1 << BLK_TC_SHIFT) - 1);
- long_act = !!(trace_flags & TRACE_ITER_VERBOSE);
+ what = (t->action & ((1 << BLK_TC_SHIFT) - 1)) & ~__BLK_TA_CGROUP;
+ long_act = !!(tr->trace_flags & TRACE_ITER(VERBOSE));
log_action = classic ? &blk_log_action_classic : &blk_log_action;
+ has_cg = t->action & __BLK_TA_CGROUP;
- if (t->action == BLK_TN_MESSAGE) {
- ret = log_action(iter, long_act ? "message" : "m");
- if (ret)
- ret = blk_log_msg(s, iter->ent);
- goto out;
+ if ((t->action & ~__BLK_TN_CGROUP) == BLK_TN_MESSAGE) {
+ log_action(iter, long_act ? "message" : "m", has_cg);
+ blk_log_msg(s, iter->ent, has_cg);
+ return trace_handle_return(s);
}
if (unlikely(what == 0 || what >= ARRAY_SIZE(what2act)))
- ret = trace_seq_printf(s, "Unknown action %x\n", what);
+ trace_seq_printf(s, "Unknown action %x\n", what);
else {
- ret = log_action(iter, what2act[what].act[long_act]);
- if (ret)
- ret = what2act[what].print(s, iter->ent);
+ log_action(iter, what2act[what].act[long_act], has_cg);
+ what2act[what].print(s, iter->ent, has_cg);
}
-out:
- return ret ? TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE;
+
+ return trace_handle_return(s);
}
static enum print_line_t blk_trace_event_print(struct trace_iterator *iter,
@@ -1377,46 +1764,48 @@ static enum print_line_t blk_trace_event_print(struct trace_iterator *iter,
return print_one_line(iter, false);
}
-static int blk_trace_synthesize_old_trace(struct trace_iterator *iter)
+static void blk_trace_synthesize_old_trace(struct trace_iterator *iter)
{
struct trace_seq *s = &iter->seq;
- struct blk_io_trace *t = (struct blk_io_trace *)iter->ent;
- const int offset = offsetof(struct blk_io_trace, sector);
+ struct blk_io_trace2 *t = (struct blk_io_trace2 *)iter->ent;
+ const int offset = offsetof(struct blk_io_trace2, sector);
struct blk_io_trace old = {
.magic = BLK_IO_TRACE_MAGIC | BLK_IO_TRACE_VERSION,
.time = iter->ts,
};
- if (!trace_seq_putmem(s, &old, offset))
- return 0;
- return trace_seq_putmem(s, &t->sector,
- sizeof(old) - offset + t->pdu_len);
+ trace_seq_putmem(s, &old, offset);
+ trace_seq_putmem(s, &t->sector,
+ sizeof(old) - offset + t->pdu_len);
}
static enum print_line_t
blk_trace_event_print_binary(struct trace_iterator *iter, int flags,
struct trace_event *event)
{
- return blk_trace_synthesize_old_trace(iter) ?
- TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE;
+ blk_trace_synthesize_old_trace(iter);
+
+ return trace_handle_return(&iter->seq);
}
static enum print_line_t blk_tracer_print_line(struct trace_iterator *iter)
{
- if (!(blk_tracer_flags.val & TRACE_BLK_OPT_CLASSIC))
+ if ((iter->ent->type != TRACE_BLK) ||
+ !(blk_tracer_flags.val & TRACE_BLK_OPT_CLASSIC))
return TRACE_TYPE_UNHANDLED;
return print_one_line(iter, true);
}
-static int blk_tracer_set_flag(u32 old_flags, u32 bit, int set)
+static int
+blk_tracer_set_flag(struct trace_array *tr, u32 old_flags, u32 bit, int set)
{
/* don't output context-info for blk_classic output */
if (bit == TRACE_BLK_OPT_CLASSIC) {
if (set)
- trace_flags &= ~TRACE_ITER_CONTEXT_INFO;
+ tr->trace_flags &= ~TRACE_ITER(CONTEXT_INFO);
else
- trace_flags |= TRACE_ITER_CONTEXT_INFO;
+ tr->trace_flags |= TRACE_ITER(CONTEXT_INFO);
}
return 0;
}
@@ -1445,17 +1834,21 @@ static struct trace_event trace_blk_event = {
static int __init init_blk_tracer(void)
{
- if (!register_ftrace_event(&trace_blk_event)) {
- pr_warning("Warning: could not register block events\n");
+ if (!register_trace_event(&trace_blk_event)) {
+ pr_warn("Warning: could not register block events\n");
return 1;
}
if (register_tracer(&blk_tracer) != 0) {
- pr_warning("Warning: could not register the block tracer\n");
- unregister_ftrace_event(&trace_blk_event);
+ pr_warn("Warning: could not register the block tracer\n");
+ unregister_trace_event(&trace_blk_event);
return 1;
}
+ BUILD_BUG_ON(__alignof__(struct blk_user_trace_setup2) %
+ __alignof__(long));
+ BUILD_BUG_ON(__alignof__(struct blk_io_trace2) % __alignof__(long));
+
return 0;
}
@@ -1465,14 +1858,16 @@ static int blk_trace_remove_queue(struct request_queue *q)
{
struct blk_trace *bt;
- bt = xchg(&q->blk_trace, NULL);
+ bt = rcu_replace_pointer(q->blk_trace, NULL,
+ lockdep_is_held(&q->debugfs_mutex));
if (bt == NULL)
return -EINVAL;
- if (atomic_dec_and_test(&blk_probes_ref))
- blk_unregister_tracepoints();
+ blk_trace_stop(bt);
- blk_trace_free(bt);
+ put_probe_ref();
+ synchronize_rcu();
+ blk_trace_free(q, bt);
return 0;
}
@@ -1482,7 +1877,7 @@ static int blk_trace_remove_queue(struct request_queue *q)
static int blk_trace_setup_queue(struct request_queue *q,
struct block_device *bdev)
{
- struct blk_trace *old_bt, *bt = NULL;
+ struct blk_trace *bt = NULL;
int ret = -ENOMEM;
bt = kzalloc(sizeof(*bt), GFP_KERNEL);
@@ -1498,19 +1893,12 @@ static int blk_trace_setup_queue(struct request_queue *q,
blk_trace_setup_lba(bt, bdev);
- old_bt = xchg(&q->blk_trace, bt);
- if (old_bt != NULL) {
- (void)xchg(&q->blk_trace, old_bt);
- ret = -EBUSY;
- goto free_bt;
- }
-
- if (atomic_inc_return(&blk_probes_ref) == 1)
- blk_register_tracepoints();
+ rcu_assign_pointer(q->blk_trace, bt);
+ get_probe_ref();
return 0;
free_bt:
- blk_trace_free(bt);
+ blk_trace_free(q, bt);
return ret;
}
@@ -1563,11 +1951,13 @@ static const struct {
{ BLK_TC_COMPLETE, "complete" },
{ BLK_TC_FS, "fs" },
{ BLK_TC_PC, "pc" },
+ { BLK_TC_NOTIFY, "notify" },
{ BLK_TC_AHEAD, "ahead" },
{ BLK_TC_META, "meta" },
{ BLK_TC_DISCARD, "discard" },
{ BLK_TC_DRV_DATA, "drv_data" },
{ BLK_TC_FUA, "fua" },
+ { BLK_TC_WRITE_ZEROES, "write-zeroes" },
};
static int blk_trace_str2mask(const char *str)
@@ -1621,54 +2011,37 @@ static ssize_t blk_trace_mask2str(char *buf, int mask)
return p - buf;
}
-static struct request_queue *blk_trace_get_queue(struct block_device *bdev)
-{
- if (bdev->bd_disk == NULL)
- return NULL;
-
- return bdev_get_queue(bdev);
-}
-
static ssize_t sysfs_blk_trace_attr_show(struct device *dev,
struct device_attribute *attr,
char *buf)
{
- struct hd_struct *p = dev_to_part(dev);
- struct request_queue *q;
- struct block_device *bdev;
+ struct block_device *bdev = dev_to_bdev(dev);
+ struct request_queue *q = bdev_get_queue(bdev);
+ struct blk_trace *bt;
ssize_t ret = -ENXIO;
- bdev = bdget(part_devt(p));
- if (bdev == NULL)
- goto out;
-
- q = blk_trace_get_queue(bdev);
- if (q == NULL)
- goto out_bdput;
-
- mutex_lock(&bdev->bd_mutex);
+ mutex_lock(&q->debugfs_mutex);
+ bt = rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex));
if (attr == &dev_attr_enable) {
- ret = sprintf(buf, "%u\n", !!q->blk_trace);
+ ret = sprintf(buf, "%u\n", !!bt);
goto out_unlock_bdev;
}
- if (q->blk_trace == NULL)
+ if (bt == NULL)
ret = sprintf(buf, "disabled\n");
else if (attr == &dev_attr_act_mask)
- ret = blk_trace_mask2str(buf, q->blk_trace->act_mask);
+ ret = blk_trace_mask2str(buf, bt->act_mask);
else if (attr == &dev_attr_pid)
- ret = sprintf(buf, "%u\n", q->blk_trace->pid);
+ ret = sprintf(buf, "%u\n", bt->pid);
else if (attr == &dev_attr_start_lba)
- ret = sprintf(buf, "%llu\n", q->blk_trace->start_lba);
+ ret = sprintf(buf, "%llu\n", bt->start_lba);
else if (attr == &dev_attr_end_lba)
- ret = sprintf(buf, "%llu\n", q->blk_trace->end_lba);
+ ret = sprintf(buf, "%llu\n", bt->end_lba);
out_unlock_bdev:
- mutex_unlock(&bdev->bd_mutex);
-out_bdput:
- bdput(bdev);
-out:
+ mutex_unlock(&q->debugfs_mutex);
return ret;
}
@@ -1676,9 +2049,9 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev,
struct device_attribute *attr,
const char *buf, size_t count)
{
- struct block_device *bdev;
- struct request_queue *q;
- struct hd_struct *p;
+ struct block_device *bdev = dev_to_bdev(dev);
+ struct request_queue *q = bdev_get_queue(bdev);
+ struct blk_trace *bt;
u64 value;
ssize_t ret = -EINVAL;
@@ -1686,30 +2059,27 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev,
goto out;
if (attr == &dev_attr_act_mask) {
- if (sscanf(buf, "%llx", &value) != 1) {
+ if (kstrtoull(buf, 0, &value)) {
/* Assume it is a list of trace category names */
ret = blk_trace_str2mask(buf);
if (ret < 0)
goto out;
value = ret;
}
- } else if (sscanf(buf, "%llu", &value) != 1)
- goto out;
-
- ret = -ENXIO;
-
- p = dev_to_part(dev);
- bdev = bdget(part_devt(p));
- if (bdev == NULL)
- goto out;
-
- q = blk_trace_get_queue(bdev);
- if (q == NULL)
- goto out_bdput;
+ } else {
+ if (kstrtoull(buf, 0, &value))
+ goto out;
+ }
- mutex_lock(&bdev->bd_mutex);
+ mutex_lock(&q->debugfs_mutex);
+ bt = rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex));
if (attr == &dev_attr_enable) {
+ if (!!value == !!bt) {
+ ret = 0;
+ goto out_unlock_bdev;
+ }
if (value)
ret = blk_trace_setup_queue(q, bdev);
else
@@ -1718,93 +2088,109 @@ static ssize_t sysfs_blk_trace_attr_store(struct device *dev,
}
ret = 0;
- if (q->blk_trace == NULL)
+ if (bt == NULL) {
ret = blk_trace_setup_queue(q, bdev);
+ bt = rcu_dereference_protected(q->blk_trace,
+ lockdep_is_held(&q->debugfs_mutex));
+ }
if (ret == 0) {
if (attr == &dev_attr_act_mask)
- q->blk_trace->act_mask = value;
+ bt->act_mask = value;
else if (attr == &dev_attr_pid)
- q->blk_trace->pid = value;
+ bt->pid = value;
else if (attr == &dev_attr_start_lba)
- q->blk_trace->start_lba = value;
+ bt->start_lba = value;
else if (attr == &dev_attr_end_lba)
- q->blk_trace->end_lba = value;
+ bt->end_lba = value;
}
out_unlock_bdev:
- mutex_unlock(&bdev->bd_mutex);
-out_bdput:
- bdput(bdev);
+ mutex_unlock(&q->debugfs_mutex);
out:
return ret ? ret : count;
}
-
-int blk_trace_init_sysfs(struct device *dev)
-{
- return sysfs_create_group(&dev->kobj, &blk_trace_attr_group);
-}
-
-void blk_trace_remove_sysfs(struct device *dev)
-{
- sysfs_remove_group(&dev->kobj, &blk_trace_attr_group);
-}
-
#endif /* CONFIG_BLK_DEV_IO_TRACE */
#ifdef CONFIG_EVENT_TRACING
-void blk_dump_cmd(char *buf, struct request *rq)
-{
- int i, end;
- int len = rq->cmd_len;
- unsigned char *cmd = rq->cmd;
-
- if (rq->cmd_type != REQ_TYPE_BLOCK_PC) {
- buf[0] = '\0';
- return;
- }
-
- for (end = len - 1; end >= 0; end--)
- if (cmd[end])
- break;
- end++;
-
- for (i = 0; i < len; i++) {
- buf += sprintf(buf, "%s%02x", i == 0 ? "" : " ", cmd[i]);
- if (i == end && end != len - 1) {
- sprintf(buf, " ..");
- break;
- }
- }
-}
-
-void blk_fill_rwbs(char *rwbs, u32 rw, int bytes)
+/**
+ * blk_fill_rwbs - Fill the buffer rwbs by mapping op to character string.
+ * @rwbs: buffer to be filled
+ * @opf: request operation type (REQ_OP_XXX) and flags for the tracepoint
+ *
+ * Description:
+ * Maps each request operation and flag to a single character and fills the
+ * buffer provided by the caller with resulting string.
+ *
+ **/
+void blk_fill_rwbs(char *rwbs, blk_opf_t opf)
{
int i = 0;
- if (rw & REQ_FLUSH)
+ if (opf & REQ_PREFLUSH)
rwbs[i++] = 'F';
- if (rw & WRITE)
+ switch (opf & REQ_OP_MASK) {
+ case REQ_OP_WRITE:
rwbs[i++] = 'W';
- else if (rw & REQ_DISCARD)
+ break;
+ case REQ_OP_DISCARD:
+ rwbs[i++] = 'D';
+ break;
+ case REQ_OP_SECURE_ERASE:
rwbs[i++] = 'D';
- else if (bytes)
+ rwbs[i++] = 'E';
+ break;
+ case REQ_OP_FLUSH:
+ rwbs[i++] = 'F';
+ break;
+ case REQ_OP_READ:
rwbs[i++] = 'R';
- else
+ break;
+ case REQ_OP_ZONE_APPEND:
+ rwbs[i++] = 'Z';
+ rwbs[i++] = 'A';
+ break;
+ case REQ_OP_ZONE_RESET:
+ case REQ_OP_ZONE_RESET_ALL:
+ rwbs[i++] = 'Z';
+ rwbs[i++] = 'R';
+ if ((opf & REQ_OP_MASK) == REQ_OP_ZONE_RESET_ALL)
+ rwbs[i++] = 'A';
+ break;
+ case REQ_OP_ZONE_FINISH:
+ rwbs[i++] = 'Z';
+ rwbs[i++] = 'F';
+ break;
+ case REQ_OP_ZONE_OPEN:
+ rwbs[i++] = 'Z';
+ rwbs[i++] = 'O';
+ break;
+ case REQ_OP_ZONE_CLOSE:
+ rwbs[i++] = 'Z';
+ rwbs[i++] = 'C';
+ break;
+ case REQ_OP_WRITE_ZEROES:
+ rwbs[i++] = 'W';
+ rwbs[i++] = 'Z';
+ break;
+ default:
rwbs[i++] = 'N';
+ }
- if (rw & REQ_FUA)
+ if (opf & REQ_FUA)
rwbs[i++] = 'F';
- if (rw & REQ_RAHEAD)
+ if (opf & REQ_RAHEAD)
rwbs[i++] = 'A';
- if (rw & REQ_SYNC)
+ if (opf & REQ_SYNC)
rwbs[i++] = 'S';
- if (rw & REQ_META)
+ if (opf & REQ_META)
rwbs[i++] = 'M';
- if (rw & REQ_SECURE)
- rwbs[i++] = 'E';
+ if (opf & REQ_ATOMIC)
+ rwbs[i++] = 'U';
+
+ WARN_ON_ONCE(i >= RWBS_LEN);
rwbs[i] = '\0';
}
generated by cgit (git 2.34.1) at 2025-12-21 21:12:28 +0000