Skip to contentSkip to conversation

mm/damon/dbgfs: add region_stat interface

PATCH In review Xin Hao 3 participants 5 messages Oct 12, 2021

Subsystems (2)
lore
Share & tools
lore

Xin Hao

author
#0 lore raw

PATCH 1/1 mm/damon/dbgfs: add region_stat interface

Using damon-dbgfs has brought great convenience to user-mode operation damon, but sometimes if i want to be able to view the division of task regions, nr_access values etc,but i found that it is impossible to view directly through the dbgfs interface, so there i add a interface "region_stat", it displays like this.

 # cat region_stat
 last_aggregation=120.87s
 target_id=5148
 nr_regions=10
 400000-258c000(34352 KiB): 1
 258c000-4719000(34356 KiB): 0
 4719000-abbf000(103064 KiB): 0
 abbf000-c4d4000(25684 KiB): 11
 c4d4000-ff5c000(59936 KiB): 15
 ff5c000-152f9000(85620 KiB): 20
 152f9000-1599e000(6804 KiB): 10
 1599e000-19573000(61268 KiB): 0
 19573000-1f92c000(102116 KiB): 0
 1f92c000-22a4c000(50304 KiB): 0
---
 mm/damon/dbgfs.c | 60 ++++++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 58 insertions(+), 2 deletions(-)
mm/damon/dbgfs.c +58 -2
index faee070977d8..269a336e92f0 100644
@@ -266,6 +266,57 @@ static ssize_t dbgfs_kdamond_pid_read(struct file *file,
 	return len;
 }
 
+static ssize_t dbgfs_region_stat_read(struct file *file,
+		char __user *buf, size_t count, loff_t *ppos)
+{
+	struct damon_ctx *ctx = file->private_data;
+	struct damon_target *t;
+	char *kbuf;
+	ssize_t len;
+	int id, rc, written = 0;
+
+	kbuf = kmalloc(count, GFP_KERNEL);
+	if (!kbuf)
+		return -ENOMEM;
+
+	mutex_lock(&ctx->kdamond_lock);
+	damon_for_each_target(t, ctx) {
+		struct damon_region *r;
+
+		if (targetid_is_pid(ctx))
+			id = (int)pid_vnr((struct pid *)t->id);
+
+		rc = scnprintf(&kbuf[written], count - written,
+				"last_aggregation=%lld.%lds\ntarget_id=%d\nnr_regions=%u\n",
+				ctx->last_aggregation.tv_sec,
+				ctx->last_aggregation.tv_nsec / 1000000,
+				id, t->nr_regions);
+		if (!rc)
+			goto out;
+
+		written += rc;
+
+		damon_for_each_region(r, t) {
+			rc = scnprintf(&kbuf[written], count - written,
+				       "%lx-%lx(%lu KiB): %u\n",
+				       r->ar.start, r->ar.end,
+					   (r->ar.end - r->ar.start) >> 10,
+					   r->nr_accesses);
+			if (!rc)
+				goto out;
+
+			written += rc;
+		}
+
+		len += simple_read_from_buffer(buf, count, ppos, kbuf, written);
+	}
+
+out:
+	mutex_unlock(&ctx->kdamond_lock);
+	kfree(kbuf);
+	return len;
+}
+
 static int damon_dbgfs_open(struct inode *inode, struct file *file)
 {
 	file->private_data = inode->i_private;
@@ -290,12 +341,17 @@ static const struct file_operations kdamond_pid_fops = {
 	.read = dbgfs_kdamond_pid_read,
 };
 
+static const struct file_operations region_stat_fops = {
+	.open = damon_dbgfs_open,
+	.read = dbgfs_region_stat_read,
+};
+
 static void dbgfs_fill_ctx_dir(struct dentry *dir, struct damon_ctx *ctx)
 {
 	const char * const file_names[] = {"attrs", "target_ids",
-		"kdamond_pid"};
+		"kdamond_pid", "region_stat"};
 	const struct file_operations *fops[] = {&attrs_fops, &target_ids_fops,
-		&kdamond_pid_fops};
+		&kdamond_pid_fops, &region_stat_fops};
 	int i;
 
 	for (i = 0; i < ARRAY_SIZE(file_names); i++)
Signature · 1 line
2.27.0
maintainer
#1 lore raw

Hello Xin, thank you for this patch!

Quoted 6 lines · On Tue, 12 Oct 2021 13:49:48 +0800 Xin Hao <[email protected]> wrote:

Using damon-dbgfs has brought great convenience to user-mode operation damon, but sometimes if i want to be able to view the division of task regions, nr_access values etc,but i found that it is impossible to view directly through the dbgfs interface, so there i add a interface "region_stat", it displays like this.

Quoted 14 lines
 # cat region_stat
 last_aggregation=120.87s
 target_id=5148
 nr_regions=10
 400000-258c000(34352 KiB): 1
 258c000-4719000(34356 KiB): 0
 4719000-abbf000(103064 KiB): 0
 abbf000-c4d4000(25684 KiB): 11
 c4d4000-ff5c000(59936 KiB): 15
 ff5c000-152f9000(85620 KiB): 20
 152f9000-1599e000(6804 KiB): 10
 1599e000-19573000(61268 KiB): 0
 19573000-1f92c000(102116 KiB): 0
 1f92c000-22a4c000(50304 KiB): 0

I think similar information could also be collected via the 'damon_aggregated' tracepoint[1], which is merged in the mainline, or 'DAMOS_STAT'[2], which is merged in -mm. The recording feature[3] could also be used, though it would take some time before it is merged in the mainline. Have you considered using those but found some problem?

[1] https://git.kernel.org/torvalds/c/2fcb93629ad8
[2] https://lore.kernel.org/linux-mm/[email protected]/
[3] https://lore.kernel.org/linux-mm/[email protected]/

Thanks,
SJ

[...]

author
#2 lore raw
Quoted 24 lines · SeongJae Park

Hello Xin, thank you for this patch!

Quoted 6 lines · On Tue, 12 Oct 2021 13:49:48 +0800 Xin Hao <[email protected]> wrote:

Using damon-dbgfs has brought great convenience to user-mode operation damon, but sometimes if i want to be able to view the division of task regions, nr_access values etc,but i found that it is impossible to view directly through the dbgfs interface, so there i add a interface "region_stat", it displays like this.

Quoted 14 lines
  # cat region_stat
  last_aggregation=120.87s
  target_id=5148
  nr_regions=10
  400000-258c000(34352 KiB): 1
  258c000-4719000(34356 KiB): 0
  4719000-abbf000(103064 KiB): 0
  abbf000-c4d4000(25684 KiB): 11
  c4d4000-ff5c000(59936 KiB): 15
  ff5c000-152f9000(85620 KiB): 20
  152f9000-1599e000(6804 KiB): 10
  1599e000-19573000(61268 KiB): 0
  19573000-1f92c000(102116 KiB): 0
  1f92c000-22a4c000(50304 KiB): 0
Quoted 5 lines

I think similar information could also be collected via the 'damon_aggregated' tracepoint[1], which is merged in the mainline, or 'DAMOS_STAT'[2], which is merged in -mm. The recording feature[3] could also be used, though it would take some time before it is merged in the mainline. Have you considered using those but found some problem?

Yes, i know we can use damon_aggregated tracepoint, but i think, add a "region_stat" will be more

convenient and intuitive, especially when we use damon-dbgfs interface.

Quoted 10 lines

[1] https://git.kernel.org/torvalds/c/2fcb93629ad8
[2] https://lore.kernel.org/linux-mm/[email protected]/
[3] https://lore.kernel.org/linux-mm/[email protected]/

Thanks,
SJ

[...]

Signature · 2 lines
Best Regards!
Xin Hao
maintainer
#3 lore raw
Quoted 32 lines · On Tue, 12 Oct 2021 15:57:17 +0800 Xin Hao <[email protected]> wrote:
Quoted 24 lines · SeongJae Park

Hello Xin, thank you for this patch!

Quoted 6 lines · On Tue, 12 Oct 2021 13:49:48 +0800 Xin Hao <[email protected]> wrote:

Using damon-dbgfs has brought great convenience to user-mode operation damon, but sometimes if i want to be able to view the division of task regions, nr_access values etc,but i found that it is impossible to view directly through the dbgfs interface, so there i add a interface "region_stat", it displays like this.

Quoted 14 lines
  # cat region_stat
  last_aggregation=120.87s
  target_id=5148
  nr_regions=10
  400000-258c000(34352 KiB): 1
  258c000-4719000(34356 KiB): 0
  4719000-abbf000(103064 KiB): 0
  abbf000-c4d4000(25684 KiB): 11
  c4d4000-ff5c000(59936 KiB): 15
  ff5c000-152f9000(85620 KiB): 20
  152f9000-1599e000(6804 KiB): 10
  1599e000-19573000(61268 KiB): 0
  19573000-1f92c000(102116 KiB): 0
  1f92c000-22a4c000(50304 KiB): 0
Quoted 5 lines

I think similar information could also be collected via the 'damon_aggregated' tracepoint[1], which is merged in the mainline, or 'DAMOS_STAT'[2], which is merged in -mm. The recording feature[3] could also be used, though it would take some time before it is merged in the mainline. Have you considered using those but found some problem?

Quoted 4 lines

Yes, i know we can use damon_aggregated tracepoint, but i think, add a "region_stat" will be more

convenient and intuitive, especially when we use damon-dbgfs interface.

Thanks for the answer.

For more conveniend and intuitive interfaces, I recommend DAMON user-sapce tool[1] rather than the debugfs interface. You can also implement your own user-space tool on top of the debugfs interface. For example, the reference implementation[1] implements 'record' feature on top of the tracepoint. The feature can provide below information, which is quite similar to what you want. For more information on this, please refer to the document[2].

    # damo record $(pidof $workload)
    # damo report raw
    base_time_absolute: 8 m 59.809 s
    monitoring_start:                0 ns
    monitoring_end:            104.599 ms
    monitoring_duration:       104.599 ms
    target_id: 18446623438842320000
    nr_regions: 3
    563ebaa00000-563ebc99e000(  31.617 MiB):        1
    7f938d7e1000-7f938ddfc000(   6.105 MiB):        0
    7fff66b0a000-7fff66bb2000( 672.000 KiB):        0

    monitoring_start:          104.599 ms
    monitoring_end:            208.590 ms
    monitoring_duration:       103.991 ms
    target_id: 18446623438842320000
    nr_regions: 4
    563ebaa00000-563ebc99e000(  31.617 MiB):        1
    7f938d7e1000-7f938d9b5000(   1.828 MiB):        0
    7f938d9b5000-7f938ddfc000(   4.277 MiB):        0
    7fff66b0a000-7fff66bb2000( 672.000 KiB):        5

If this is still not so useful for you, could you please elaborate more on your detailed requirements and use cases?

[1] https://github.com/awslabs/damo
[2] https://github.com/awslabs/damo/blob/next/USAGE.md#raw

Thanks,
SJ

Quoted 15 lines
Quoted 11 lines

[1] https://git.kernel.org/torvalds/c/2fcb93629ad8
[2] https://lore.kernel.org/linux-mm/[email protected]/
[3] https://lore.kernel.org/linux-mm/[email protected]/

Thanks,
SJ

[...]

Signature · 2 lines
Best Regards!
Xin Hao
#4 lore raw
In reply to Xin Hao
Thank you for the patch! Perhaps something to improve:

Hi Xin,

Thank you for the patch! Perhaps something to improve:

[auto build test WARNING on v5.15-rc5]
[cannot apply to hnaz-mm/master next-20211012]
[If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Xin-Hao/mm-damon-dbgfs-add-region_stat-interface/20211012-135023
base:    64570fbc14f8d7cb3fe3995f20e26bc25ce4b2cc
config: arm64-randconfig-r014-20211012 (attached as .config)
compiler: clang version 14.0.0 (https://github.com/llvm/llvm-project c3dcf39554dbea780d6cb7e12239451ba47a2668)

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install arm64 cross compiling tool for clang build
# apt-get install binutils-aarch64-linux-gnu
# https://github.com/0day-ci/linux/commit/256bdd48bd7719684cd8adbbda8a9df3dc9c1008
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Xin-Hao/mm-damon-dbgfs-add-region_stat-interface/20211012-135023 git checkout 256bdd48bd7719684cd8adbbda8a9df3dc9c1008
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 ARCH=arm64

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

>> mm/damon/dbgfs.c:311:3: warning: variable 'len' is uninitialized when used here [-Wuninitialized]
                   len += simple_read_from_buffer(buf, count, ppos, kbuf, written);
                   ^~~
   mm/damon/dbgfs.c:275:13: note: initialize the variable 'len' to silence this warning
           ssize_t len;
                      ^
                       = 0

1 warning generated.

vim +/len +311 mm/damon/dbgfs.c

268
269 static ssize_t dbgfs_region_stat_read(struct file *file,
270 char __user *buf, size_t count, loff_t *ppos)
271 {
272 struct damon_ctx *ctx = file->private_data;
273 struct damon_target *t;
274 char *kbuf;
275 ssize_t len;
276 int id, rc, written = 0;
277
278 kbuf = kmalloc(count, GFP_KERNEL);
279 if (!kbuf)
280 return -ENOMEM;
281
282 mutex_lock(&ctx->kdamond_lock);
283 damon_for_each_target(t, ctx) {
284 struct damon_region *r;
285
286 if (targetid_is_pid(ctx))
287 id = (int)pid_vnr((struct pid *)t->id);
288
289 rc = scnprintf(&kbuf[written], count - written,
290 "last_aggregation=%lld.%lds\ntarget_id=%d\nnr_regions=%u\n",
291 ctx->last_aggregation.tv_sec,
292 ctx->last_aggregation.tv_nsec / 1000000,
293 id, t->nr_regions);
294 if (!rc)
295 goto out;
296
297 written += rc;
298
299 damon_for_each_region(r, t) {
300 rc = scnprintf(&kbuf[written], count - written,
301 "%lx-%lx(%lu KiB): %u\n",
302 r->ar.start, r->ar.end,
303 (r->ar.end - r->ar.start) >> 10,
304 r->nr_accesses);
305 if (!rc)
306 goto out;
307
308 written += rc;
309 }
310

 > 311			len += simple_read_from_buffer(buf, count, ppos, kbuf, written);

312 }
313
314 out:
315 mutex_unlock(&ctx->kdamond_lock);
316 kfree(kbuf);
317 return len;
318 }
319

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org