[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20250902100353835xyAecL45pVFk1sbaC16f4@zte.com.cn>
Date: Tue, 2 Sep 2025 10:03:53 +0800 (CST)
From: <fan.yu9@....com.cn>
To: <akpm@...ux-foundation.org>, <wang.yaxin@....com.cn>
Cc: <corbet@....net>, <linux-kernel@...r.kernel.org>,
<linux-doc@...r.kernel.org>, <xu.xin16@....com.cn>,
<yang.yang29@....com.cn>, <fan.yu9@....com.cn>
Subject: [PATCH linux-next 2/3] tools/delaytop: add flexible sorting by delay field
From: Fan Yu <fan.yu9@....com.cn>
The delaytop tool only supported sorting by CPU delay, which limited
its usefulness when users needed to identify bottlenecks in other
subsystems. Users had no way to sort processes by IO, memory, or
other delay types to quickly pinpoint specific performance issues.
Add -s/--sort option to allow sorting by different delay types:
1) Basic modes: cpu, io, irq, mem
2) Detailed modes (-M required): swap, reclaim, thrashing, compact, wpcopy
Users can now quickly identify bottlenecks in specific subsystems
by sorting processes by the relevant delay metric.
Signed-off-by: Fan Yu <fan.yu9@....com.cn>
---
tools/accounting/delaytop.c | 130 +++++++++++++++++++++++++++++++++---
1 file changed, 121 insertions(+), 9 deletions(-)
diff --git a/tools/accounting/delaytop.c b/tools/accounting/delaytop.c
index f1e2e1cca4b8..39852cd70bdf 100644
--- a/tools/accounting/delaytop.c
+++ b/tools/accounting/delaytop.c
@@ -173,7 +173,9 @@ static void usage(void)
" -o, --once Display once and exit\n"
" -p, --pid=PID Monitor only the specified PID\n"
" -C, --container=PATH Monitor the container at specified cgroup path\n"
- " -M, --memverbose Display memory detailed information\n");
+ " -M, --memverbose Display memory detailed information\n"
+ " -s, --sort=FIELD Sort by delay field (default: cpu)\n"
+ " Types: cpu|io|irq|mem|swap|reclaim|thrashing|compact|wpcopy\n");
exit(0);
}
@@ -188,6 +190,7 @@ static void parse_args(int argc, char **argv)
{"pid", required_argument, 0, 'p'},
{"once", no_argument, 0, 'o'},
{"processes", required_argument, 0, 'P'},
+ {"sort", required_argument, 0, 's'},
{"container", required_argument, 0, 'C'},
{"memverbose", no_argument, 0, 'M'},
{0, 0, 0, 0}
@@ -206,7 +209,7 @@ static void parse_args(int argc, char **argv)
while (1) {
int option_index = 0;
- c = getopt_long(argc, argv, "hd:n:p:oP:C:M", long_options, &option_index);
+ c = getopt_long(argc, argv, "hd:n:p:oP:C:Ms:", long_options, &option_index);
if (c == -1)
break;
@@ -256,11 +259,53 @@ static void parse_args(int argc, char **argv)
case 'M':
cfg.mem_verbose_mode = 1;
break;
+ case 's':
+ if (strlen(optarg) == 0) {
+ fprintf(stderr, "Error: empty sort field\n");
+ exit(1);
+ }
+
+ if (strncmp(optarg, "cpu", 3) == 0)
+ cfg.sort_field = 'c';
+ else if (strncmp(optarg, "io", 2) == 0)
+ cfg.sort_field = 'i';
+ else if (strncmp(optarg, "irq", 3) == 0)
+ cfg.sort_field = 'q';
+ else if (strncmp(optarg, "mem", 3) == 0)
+ cfg.sort_field = 'm';
+ else if (strncmp(optarg, "swap", 4) == 0)
+ cfg.sort_field = 's';
+ else if (strncmp(optarg, "reclaim", 7) == 0)
+ cfg.sort_field = 'r';
+ else if (strncmp(optarg, "thrashing", 9) == 0)
+ cfg.sort_field = 't';
+ else if (strncmp(optarg, "compact", 7) == 0)
+ cfg.sort_field = 'p';
+ else if (strncmp(optarg, "wpcopy", 7) == 0)
+ cfg.sort_field = 'w';
+ else {
+ fprintf(stderr, "Error: invalid sort field\n");
+ fprintf(stderr, "Try to use cpu|io|irq|mem|");
+ fprintf(stderr, "swap|reclaim|thrashing|compact|wpcopy\n");
+ exit(1);
+ }
+ break;
default:
fprintf(stderr, "Try 'delaytop --help' for more information.\n");
exit(1);
}
}
+
+ /* Validate sorting field compatibility with memory verbose mode */
+ if (cfg.mem_verbose_mode == 0 &&
+ cfg.sort_field == 's' ||
+ cfg.sort_field == 'r' ||
+ cfg.sort_field == 't' ||
+ cfg.sort_field == 'p' ||
+ cfg.sort_field == 'w') {
+ fprintf(stderr, "Error: mem verbose mode is off, try to use -M\n");
+ exit(1);
+ }
}
/* Create a raw netlink socket and bind */
@@ -621,12 +666,77 @@ static int compare_tasks(const void *a, const void *b)
case 'c': /* CPU */
avg1 = average_ms(t1->cpu_delay_total, t1->cpu_count);
avg2 = average_ms(t2->cpu_delay_total, t2->cpu_count);
- if (avg1 != avg2)
- return avg2 > avg1 ? 1 : -1;
- return t2->cpu_delay_total > t1->cpu_delay_total ? 1 : -1;
+ break;
+ case 'i': /* IO */
+ avg1 = average_ms(t1->blkio_delay_total, t1->blkio_count);
+ avg2 = average_ms(t2->blkio_delay_total, t2->blkio_count);
+ break;
+ case 'q': /* IRQ */
+ avg1 = average_ms(t1->irq_delay_total, t1->irq_count);
+ avg2 = average_ms(t2->irq_delay_total, t2->irq_count);
+ break;
+ case 'm': /* MEM(total) */
+ avg1 = average_ms(task_total_mem_delay(t1), task_total_mem_count(t1));
+ avg2 = average_ms(task_total_mem_delay(t2), task_total_mem_count(t2));
+ break;
+ /* Memory detailed display mode */
+ case 's': /* swapin (SWAP) */
+ avg1 = average_ms(t1->swapin_delay_total, t1->swapin_count);
+ avg2 = average_ms(t2->swapin_delay_total, t2->swapin_count);
+ break;
+ case 'r': /* freepages (RCL) */
+ avg1 = average_ms(t1->freepages_delay_total, t1->freepages_count);
+ avg2 = average_ms(t2->freepages_delay_total, t2->freepages_count);
+ break;
+ case 't': /* thrashing (THR) */
+ avg1 = average_ms(t1->thrashing_delay_total, t1->thrashing_count);
+ avg2 = average_ms(t2->thrashing_delay_total, t2->thrashing_count);
+ break;
+ case 'p': /* compact (CMP) */
+ avg1 = average_ms(t1->compact_delay_total, t1->compact_count);
+ avg2 = average_ms(t2->compact_delay_total, t2->compact_count);
+ break;
+ case 'w': /* wpcopy (WP) */
+ avg1 = average_ms(t1->wpcopy_delay_total, t1->wpcopy_count);
+ avg2 = average_ms(t2->wpcopy_delay_total, t2->wpcopy_count);
+ break;
+ default:
+ avg1 = average_ms(t1->cpu_delay_total, t1->cpu_count);
+ avg2 = average_ms(t2->cpu_delay_total, t2->cpu_count);
+ break;
+ }
+
+ if (avg1 != avg2)
+ return avg2 > avg1 ? 1 : -1;
+
+ return 0;
+}
+static const char *get_sort_field(char sort_field)
+{
+ switch (sort_field) {
+ case 'c':
+ return "CPU";
+ case 'i':
+ return "IO";
+ case 'q':
+ return "IRQ";
+ /* MEM(total) */
+ case 'm':
+ return "MEM";
+ /* Memory detailed display mode */
+ case 's':
+ return "SWAP";
+ case 'r':
+ return "RCL";
+ case 't':
+ return "THR";
+ case 'p':
+ return "CMP";
+ case 'w':
+ return "WP";
default:
- return t2->cpu_delay_total > t1->cpu_delay_total ? 1 : -1;
+ return "UNKNOWN"; /* handle error */
}
}
@@ -705,6 +815,7 @@ static void display_results(void)
{
time_t now = time(NULL);
struct tm *tm_now = localtime(&now);
+ const char *sort_field;
FILE *out = stdout;
char timestamp[32];
bool suc = true;
@@ -766,8 +877,10 @@ static void display_results(void)
container_stats.nr_stopped, container_stats.nr_uninterruptible,
container_stats.nr_io_wait);
}
- suc &= BOOL_FPRINT(out, "Top %d processes (sorted by CPU delay):\n",
- cfg.max_processes);
+
+ /* Task delay output */
+ suc &= BOOL_FPRINT(out, "Top %d processes (sorted by %s delay):\n",
+ cfg.max_processes, get_sort_field(cfg.sort_field));
suc &= BOOL_FPRINT(out, "%8s %8s %-17s", "PID", "TGID", "COMMAND");
if (!cfg.mem_verbose_mode) {
@@ -787,7 +900,6 @@ static void display_results(void)
suc &= BOOL_FPRINT(out, "-------------------------\n");
}
-
count = task_count < cfg.max_processes ? task_count : cfg.max_processes;
for (i = 0; i < count; i++) {
--
2.25.1
Powered by blists - more mailing lists