* [PATCH -next 0/4] cleanup and fixes for 'carryover_ios/bytes'
@ 2023-08-15 1:41 Yu Kuai
2023-08-15 1:41 ` [PATCH -next 1/4] blk-throttle: print signed value 'carryover_bytes/ios' for user Yu Kuai
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: Yu Kuai @ 2023-08-15 1:41 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw, yukuai3-hv44wF8Li93QT0dZR+AlfA,
mkoutny-IBi9RG/b67k
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai1-XF6JlduFytWkHkcT6e4Xnw, yi.zhang-hv44wF8Li93QT0dZR+AlfA,
yangerkun-hv44wF8Li93QT0dZR+AlfA
From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Patch 1 print signed value for 'carryover_ios/bytes' to user.
Patch 2 fix that comparation ”unsigned value < negative value“ will
pass, causing that io won't be throttled in the silce.
Patch 3,4 fix that 'carryover_ios/bytes' is not cleared while
'io/bytes_disp' is cleared, causing that throttle is not accurate.
Yu Kuai (4):
blk-throttle: print signed value 'carryover_bytes/ios' for user
blk-throttle: fix wrong comparation while 'carryover_ios/bytes' is
negative
blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice()
blk-throttle: consider 'carryover_ios/bytes' in throtl_trim_slice()
block/blk-throttle.c | 105 +++++++++++++++++++++----------------------
block/blk-throttle.h | 4 +-
2 files changed, 54 insertions(+), 55 deletions(-)
--
2.39.2
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH -next 1/4] blk-throttle: print signed value 'carryover_bytes/ios' for user
2023-08-15 1:41 [PATCH -next 0/4] cleanup and fixes for 'carryover_ios/bytes' Yu Kuai
@ 2023-08-15 1:41 ` Yu Kuai
[not found] ` <20230815014123.368929-1-yukuai1-XF6JlduFytWkHkcT6e4Xnw@public.gmane.org>
2023-08-15 1:41 ` [PATCH -next 4/4] blk-throttle: consider 'carryover_ios/bytes' in throtl_trim_slice() Yu Kuai
2 siblings, 0 replies; 6+ messages in thread
From: Yu Kuai @ 2023-08-15 1:41 UTC (permalink / raw)
To: tj, josef, axboe, yukuai3, mkoutny
Cc: cgroups, linux-block, linux-kernel, yukuai1, yi.zhang, yangerkun
From: Yu Kuai <yukuai3@huawei.com>
'carryover_bytes/ios' can be negative, indicate that some bio is
dispatched in advance within slice while configuration is updated.
Print a huge value is not user-friendly.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
block/blk-throttle.c | 2 +-
block/blk-throttle.h | 4 ++--
2 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index 7397ff199d66..5184f17f5129 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -816,7 +816,7 @@ static void tg_update_carryover(struct throtl_grp *tg)
__tg_update_carryover(tg, WRITE);
/* see comments in struct throtl_grp for meaning of these fields. */
- throtl_log(&tg->service_queue, "%s: %llu %llu %u %u\n", __func__,
+ throtl_log(&tg->service_queue, "%s: %lld %lld %d %d\n", __func__,
tg->carryover_bytes[READ], tg->carryover_bytes[WRITE],
tg->carryover_ios[READ], tg->carryover_ios[WRITE]);
}
diff --git a/block/blk-throttle.h b/block/blk-throttle.h
index d1ccbfe9f797..bffbc9cfc8ab 100644
--- a/block/blk-throttle.h
+++ b/block/blk-throttle.h
@@ -127,8 +127,8 @@ struct throtl_grp {
* bytes/ios are waited already in previous configuration, and they will
* be used to calculate wait time under new configuration.
*/
- uint64_t carryover_bytes[2];
- unsigned int carryover_ios[2];
+ long long carryover_bytes[2];
+ int carryover_ios[2];
unsigned long last_check_time;
--
2.39.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH -next 2/4] blk-throttle: fix wrong comparation while 'carryover_ios/bytes' is negative
[not found] ` <20230815014123.368929-1-yukuai1-XF6JlduFytWkHkcT6e4Xnw@public.gmane.org>
@ 2023-08-15 1:41 ` Yu Kuai
2023-08-15 1:41 ` [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice() Yu Kuai
1 sibling, 0 replies; 6+ messages in thread
From: Yu Kuai @ 2023-08-15 1:41 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw, yukuai3-hv44wF8Li93QT0dZR+AlfA,
mkoutny-IBi9RG/b67k
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai1-XF6JlduFytWkHkcT6e4Xnw, yi.zhang-hv44wF8Li93QT0dZR+AlfA,
yangerkun-hv44wF8Li93QT0dZR+AlfA
From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
carryover_ios/bytes[] can be negative in the case that ios are
dispatched in the slice in advance, and then configuration is updated.
For example:
1) set iops limit to 1000, and slice start is 0, slice end is 100ms;
2) current time is 0, and 100 ios are dispatched, those ios will not be
throttled, hence io_disp is 100;
3) still at current time 0, update iops limit to 100, then carryover_ios
is (0 - 100) = -100;
4) then, dispatch a new io at time 0, the expected result is that this
io will wait for 1s. The calculation in tg_within_iops_limit:
io_disp = 0;
io_allowed = calculate_io_allowed + carryover_ios
= 10 + (-100) = -90;
io won't be throttled if (io_disp + 1 < io_allowed) passed.
Before this patch, in step 4) (io_disp + 1 < io_allowed) is passed,
because -90 for unsigned value is very huge, and such io won't be
throttled.
Fix this problem by checking if 'io/bytes_allowed' is negative first.
Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
---
block/blk-throttle.c | 11 +++++------
1 file changed, 5 insertions(+), 6 deletions(-)
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index 5184f17f5129..7c93144d03da 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -825,7 +825,7 @@ static unsigned long tg_within_iops_limit(struct throtl_grp *tg, struct bio *bio
u32 iops_limit)
{
bool rw = bio_data_dir(bio);
- unsigned int io_allowed;
+ int io_allowed;
unsigned long jiffy_elapsed, jiffy_wait, jiffy_elapsed_rnd;
if (iops_limit == UINT_MAX) {
@@ -838,9 +838,8 @@ static unsigned long tg_within_iops_limit(struct throtl_grp *tg, struct bio *bio
jiffy_elapsed_rnd = roundup(jiffy_elapsed + 1, tg->td->throtl_slice);
io_allowed = calculate_io_allowed(iops_limit, jiffy_elapsed_rnd) +
tg->carryover_ios[rw];
- if (tg->io_disp[rw] + 1 <= io_allowed) {
+ if (io_allowed > 0 && tg->io_disp[rw] + 1 <= io_allowed)
return 0;
- }
/* Calc approx time to dispatch */
jiffy_wait = jiffy_elapsed_rnd - jiffy_elapsed;
@@ -851,7 +850,8 @@ static unsigned long tg_within_bps_limit(struct throtl_grp *tg, struct bio *bio,
u64 bps_limit)
{
bool rw = bio_data_dir(bio);
- u64 bytes_allowed, extra_bytes;
+ long long bytes_allowed;
+ u64 extra_bytes;
unsigned long jiffy_elapsed, jiffy_wait, jiffy_elapsed_rnd;
unsigned int bio_size = throtl_bio_data_size(bio);
@@ -869,9 +869,8 @@ static unsigned long tg_within_bps_limit(struct throtl_grp *tg, struct bio *bio,
jiffy_elapsed_rnd = roundup(jiffy_elapsed_rnd, tg->td->throtl_slice);
bytes_allowed = calculate_bytes_allowed(bps_limit, jiffy_elapsed_rnd) +
tg->carryover_bytes[rw];
- if (tg->bytes_disp[rw] + bio_size <= bytes_allowed) {
+ if (bytes_allowed > 0 && tg->bytes_disp[rw] + bio_size <= bytes_allowed)
return 0;
- }
/* Calc approx time to dispatch */
extra_bytes = tg->bytes_disp[rw] + bio_size - bytes_allowed;
--
2.39.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice()
[not found] ` <20230815014123.368929-1-yukuai1-XF6JlduFytWkHkcT6e4Xnw@public.gmane.org>
2023-08-15 1:41 ` [PATCH -next 2/4] blk-throttle: fix wrong comparation while 'carryover_ios/bytes' is negative Yu Kuai
@ 2023-08-15 1:41 ` Yu Kuai
2023-08-15 15:48 ` kernel test robot
1 sibling, 1 reply; 6+ messages in thread
From: Yu Kuai @ 2023-08-15 1:41 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw, yukuai3-hv44wF8Li93QT0dZR+AlfA,
mkoutny-IBi9RG/b67k
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai1-XF6JlduFytWkHkcT6e4Xnw, yi.zhang-hv44wF8Li93QT0dZR+AlfA,
yangerkun-hv44wF8Li93QT0dZR+AlfA
From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
There are no functional changes, just make the code cleaner.
Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
---
block/blk-throttle.c | 81 +++++++++++++++++++++-----------------------
1 file changed, 38 insertions(+), 43 deletions(-)
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index 7c93144d03da..e5296960c799 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -697,11 +697,40 @@ static bool throtl_slice_used(struct throtl_grp *tg, bool rw)
return true;
}
+static unsigned int calculate_io_allowed(u32 iops_limit,
+ unsigned long jiffy_elapsed)
+{
+ unsigned int io_allowed;
+ u64 tmp;
+
+ /*
+ * jiffy_elapsed should not be a big value as minimum iops can be
+ * 1 then at max jiffy elapsed should be equivalent of 1 second as we
+ * will allow dispatch after 1 second and after that slice should
+ * have been trimmed.
+ */
+
+ tmp = (u64)iops_limit * jiffy_elapsed;
+ do_div(tmp, HZ);
+
+ if (tmp > UINT_MAX)
+ io_allowed = UINT_MAX;
+ else
+ io_allowed = tmp;
+
+ return io_allowed;
+}
+
+static u64 calculate_bytes_allowed(u64 bps_limit, unsigned long jiffy_elapsed)
+{
+ return mul_u64_u64_div_u64(bps_limit, (u64)jiffy_elapsed, (u64)HZ);
+}
+
/* Trim the used slices and adjust slice start accordingly */
static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
{
- unsigned long nr_slices, time_elapsed, io_trim;
- u64 bytes_trim, tmp;
+ unsigned long time_elapsed, io_trim;
+ u64 bytes_trim;
BUG_ON(time_before(tg->slice_end[rw], tg->slice_start[rw]));
@@ -723,19 +752,14 @@ static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
throtl_set_slice_end(tg, rw, jiffies + tg->td->throtl_slice);
- time_elapsed = jiffies - tg->slice_start[rw];
-
- nr_slices = time_elapsed / tg->td->throtl_slice;
-
- if (!nr_slices)
+ time_elapsed = rounddown(jiffies - tg->slice_start[rw],
+ tg->td->throtl_slice);
+ if (!time_elapsed)
return;
- tmp = tg_bps_limit(tg, rw) * tg->td->throtl_slice * nr_slices;
- do_div(tmp, HZ);
- bytes_trim = tmp;
-
- io_trim = (tg_iops_limit(tg, rw) * tg->td->throtl_slice * nr_slices) /
- HZ;
+ bytes_trim = calculate_bytes_allowed(tg_bps_limit(tg, rw),
+ time_elapsed);
+ io_trim = calculate_io_allowed(tg_iops_limit(tg, rw), time_elapsed);
if (!bytes_trim && !io_trim)
return;
@@ -749,7 +773,7 @@ static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
else
tg->io_disp[rw] = 0;
- tg->slice_start[rw] += nr_slices * tg->td->throtl_slice;
+ tg->slice_start[rw] += time_elapsed;
throtl_log(&tg->service_queue,
"[%c] trim slice nr=%lu bytes=%llu io=%lu start=%lu end=%lu jiffies=%lu",
@@ -757,35 +781,6 @@ static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
tg->slice_start[rw], tg->slice_end[rw], jiffies);
}
-static unsigned int calculate_io_allowed(u32 iops_limit,
- unsigned long jiffy_elapsed)
-{
- unsigned int io_allowed;
- u64 tmp;
-
- /*
- * jiffy_elapsed should not be a big value as minimum iops can be
- * 1 then at max jiffy elapsed should be equivalent of 1 second as we
- * will allow dispatch after 1 second and after that slice should
- * have been trimmed.
- */
-
- tmp = (u64)iops_limit * jiffy_elapsed;
- do_div(tmp, HZ);
-
- if (tmp > UINT_MAX)
- io_allowed = UINT_MAX;
- else
- io_allowed = tmp;
-
- return io_allowed;
-}
-
-static u64 calculate_bytes_allowed(u64 bps_limit, unsigned long jiffy_elapsed)
-{
- return mul_u64_u64_div_u64(bps_limit, (u64)jiffy_elapsed, (u64)HZ);
-}
-
static void __tg_update_carryover(struct throtl_grp *tg, bool rw)
{
unsigned long jiffy_elapsed = jiffies - tg->slice_start[rw];
--
2.39.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH -next 4/4] blk-throttle: consider 'carryover_ios/bytes' in throtl_trim_slice()
2023-08-15 1:41 [PATCH -next 0/4] cleanup and fixes for 'carryover_ios/bytes' Yu Kuai
2023-08-15 1:41 ` [PATCH -next 1/4] blk-throttle: print signed value 'carryover_bytes/ios' for user Yu Kuai
[not found] ` <20230815014123.368929-1-yukuai1-XF6JlduFytWkHkcT6e4Xnw@public.gmane.org>
@ 2023-08-15 1:41 ` Yu Kuai
2 siblings, 0 replies; 6+ messages in thread
From: Yu Kuai @ 2023-08-15 1:41 UTC (permalink / raw)
To: tj, josef, axboe, yukuai3, mkoutny
Cc: cgroups, linux-block, linux-kernel, yukuai1, yi.zhang, yangerkun
From: Yu Kuai <yukuai3@huawei.com>
Currently, 'carryover_ios/bytes' is not handled in throtl_trim_slice(),
for consequence, 'carryover_ios/bytes' will be used to throttle bio
multiple times, for example:
1) set iops limit to 100, and slice start is 0, slice end is 100ms;
2) current time is 0, and 10 ios are dispatched, those io won't be
throttled and io_disp is 10;
3) still at current time 0, update iops limit to 1000, carryover_ios is
updated to (0 - 10) = -10;
4) in this slice(0 - 100ms), io_allowed = 100 + (-10) = 90, which means
only 90 ios can be dispatched without waiting;
5) assume that io is throttled in slice(0 - 100ms), and
throtl_trim_slice() update silce to (100ms - 200ms). In this case,
'carryover_ios/bytes' is not cleared and still only 90 ios can be
dispatched between 100ms - 200ms.
Fix this problem by updating 'carryover_ios/bytes' in
throtl_trim_slice().
Fixes: a880ae93e5b5 ("blk-throttle: fix io hung due to configuration updates")
Reported-by: zhuxiaohui <zhuxiaohui.400@bytedance.com>
Link: https://lore.kernel.org/all/20230812072116.42321-1-zhuxiaohui.400@bytedance.com/
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
block/blk-throttle.c | 19 ++++++++++++-------
1 file changed, 12 insertions(+), 7 deletions(-)
diff --git a/block/blk-throttle.c b/block/blk-throttle.c
index e5296960c799..75b4a78c1e45 100644
--- a/block/blk-throttle.c
+++ b/block/blk-throttle.c
@@ -729,8 +729,9 @@ static u64 calculate_bytes_allowed(u64 bps_limit, unsigned long jiffy_elapsed)
/* Trim the used slices and adjust slice start accordingly */
static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
{
- unsigned long time_elapsed, io_trim;
- u64 bytes_trim;
+ unsigned long time_elapsed;
+ long long bytes_trim;
+ int io_trim;
BUG_ON(time_before(tg->slice_end[rw], tg->slice_start[rw]));
@@ -758,17 +759,21 @@ static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
return;
bytes_trim = calculate_bytes_allowed(tg_bps_limit(tg, rw),
- time_elapsed);
- io_trim = calculate_io_allowed(tg_iops_limit(tg, rw), time_elapsed);
- if (!bytes_trim && !io_trim)
+ time_elapsed) +
+ tg->carryover_bytes[rw];
+ io_trim = calculate_io_allowed(tg_iops_limit(tg, rw), time_elapsed) +
+ tg->carryover_ios[rw];
+ if (bytes_trim <= 0 && io_trim <= 0)
return;
- if (tg->bytes_disp[rw] >= bytes_trim)
+ tg->carryover_bytes[rw] = 0;
+ if ((long long)tg->bytes_disp[rw] >= bytes_trim)
tg->bytes_disp[rw] -= bytes_trim;
else
tg->bytes_disp[rw] = 0;
- if (tg->io_disp[rw] >= io_trim)
+ tg->carryover_ios[rw] = 0;
+ if ((int)tg->io_disp[rw] >= io_trim)
tg->io_disp[rw] -= io_trim;
else
tg->io_disp[rw] = 0;
--
2.39.2
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice()
2023-08-15 1:41 ` [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice() Yu Kuai
@ 2023-08-15 15:48 ` kernel test robot
0 siblings, 0 replies; 6+ messages in thread
From: kernel test robot @ 2023-08-15 15:48 UTC (permalink / raw)
To: tj, josef, axboe, yukuai3, mkoutny
Cc: oe-kbuild-all, cgroups, linux-block, linux-kernel, yukuai1,
yi.zhang, yangerkun
Hi Yu,
kernel test robot noticed the following build errors:
[auto build test ERROR on next-20230809]
url: https://github.com/intel-lab-lkp/linux/commits/Yu-Kuai/blk-throttle-print-signed-value-carryover_bytes-ios-for-user/20230815-095025
base: next-20230809
patch link: https://lore.kernel.org/r/20230815014123.368929-4-yukuai1%40huaweicloud.com
patch subject: [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice()
config: x86_64-allyesconfig (https://download.01.org/0day-ci/archive/20230815/202308152351.JdDlpV4Q-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce: (https://download.01.org/0day-ci/archive/20230815/202308152351.JdDlpV4Q-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@intel.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202308152351.JdDlpV4Q-lkp@intel.com/
All errors (new ones prefixed by >>):
In file included from block/blk-throttle.c:12:
block/blk-throttle.c: In function 'throtl_trim_slice':
>> block/blk-throttle.c:780:44: error: 'nr_slices' undeclared (first use in this function)
780 | rw == READ ? 'R' : 'W', nr_slices, bytes_trim, io_trim,
| ^~~~~~~~~
include/linux/blktrace_api.h:56:66: note: in definition of macro 'blk_add_cgroup_trace_msg'
56 | __blk_trace_note_message(bt, css, fmt, ##__VA_ARGS__);\
| ^~~~~~~~~~~
block/blk-throttle.c:778:9: note: in expansion of macro 'throtl_log'
778 | throtl_log(&tg->service_queue,
| ^~~~~~~~~~
block/blk-throttle.c:780:44: note: each undeclared identifier is reported only once for each function it appears in
780 | rw == READ ? 'R' : 'W', nr_slices, bytes_trim, io_trim,
| ^~~~~~~~~
include/linux/blktrace_api.h:56:66: note: in definition of macro 'blk_add_cgroup_trace_msg'
56 | __blk_trace_note_message(bt, css, fmt, ##__VA_ARGS__);\
| ^~~~~~~~~~~
block/blk-throttle.c:778:9: note: in expansion of macro 'throtl_log'
778 | throtl_log(&tg->service_queue,
| ^~~~~~~~~~
vim +/nr_slices +780 block/blk-throttle.c
1a46ae7773b475 Yu Kuai 2023-08-15 728
e43473b7f223ec Vivek Goyal 2010-09-15 729 /* Trim the used slices and adjust slice start accordingly */
0f3457f60edc57 Tejun Heo 2013-05-14 730 static inline void throtl_trim_slice(struct throtl_grp *tg, bool rw)
e43473b7f223ec Vivek Goyal 2010-09-15 731 {
1a46ae7773b475 Yu Kuai 2023-08-15 732 unsigned long time_elapsed, io_trim;
1a46ae7773b475 Yu Kuai 2023-08-15 733 u64 bytes_trim;
e43473b7f223ec Vivek Goyal 2010-09-15 734
e43473b7f223ec Vivek Goyal 2010-09-15 735 BUG_ON(time_before(tg->slice_end[rw], tg->slice_start[rw]));
e43473b7f223ec Vivek Goyal 2010-09-15 736
e43473b7f223ec Vivek Goyal 2010-09-15 737 /*
e43473b7f223ec Vivek Goyal 2010-09-15 738 * If bps are unlimited (-1), then time slice don't get
e43473b7f223ec Vivek Goyal 2010-09-15 739 * renewed. Don't try to trim the slice if slice is used. A new
e43473b7f223ec Vivek Goyal 2010-09-15 740 * slice will start when appropriate.
e43473b7f223ec Vivek Goyal 2010-09-15 741 */
0f3457f60edc57 Tejun Heo 2013-05-14 742 if (throtl_slice_used(tg, rw))
e43473b7f223ec Vivek Goyal 2010-09-15 743 return;
e43473b7f223ec Vivek Goyal 2010-09-15 744
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 745 /*
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 746 * A bio has been dispatched. Also adjust slice_end. It might happen
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 747 * that initially cgroup limit was very low resulting in high
b53b072c4bb579 Baolin Wang 2020-09-07 748 * slice_end, but later limit was bumped up and bio was dispatched
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 749 * sooner, then we need to reduce slice_end. A high bogus slice_end
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 750 * is bad because it does not allow new slice to start.
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 751 */
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 752
297e3d85478482 Shaohua Li 2017-03-27 753 throtl_set_slice_end(tg, rw, jiffies + tg->td->throtl_slice);
d1ae8ffdfaa16b Vivek Goyal 2010-12-01 754
1a46ae7773b475 Yu Kuai 2023-08-15 755 time_elapsed = rounddown(jiffies - tg->slice_start[rw],
1a46ae7773b475 Yu Kuai 2023-08-15 756 tg->td->throtl_slice);
1a46ae7773b475 Yu Kuai 2023-08-15 757 if (!time_elapsed)
e43473b7f223ec Vivek Goyal 2010-09-15 758 return;
e43473b7f223ec Vivek Goyal 2010-09-15 759
1a46ae7773b475 Yu Kuai 2023-08-15 760 bytes_trim = calculate_bytes_allowed(tg_bps_limit(tg, rw),
1a46ae7773b475 Yu Kuai 2023-08-15 761 time_elapsed);
1a46ae7773b475 Yu Kuai 2023-08-15 762 io_trim = calculate_io_allowed(tg_iops_limit(tg, rw), time_elapsed);
8e89d13f4ede24 Vivek Goyal 2010-09-15 763 if (!bytes_trim && !io_trim)
e43473b7f223ec Vivek Goyal 2010-09-15 764 return;
e43473b7f223ec Vivek Goyal 2010-09-15 765
e43473b7f223ec Vivek Goyal 2010-09-15 766 if (tg->bytes_disp[rw] >= bytes_trim)
e43473b7f223ec Vivek Goyal 2010-09-15 767 tg->bytes_disp[rw] -= bytes_trim;
e43473b7f223ec Vivek Goyal 2010-09-15 768 else
e43473b7f223ec Vivek Goyal 2010-09-15 769 tg->bytes_disp[rw] = 0;
e43473b7f223ec Vivek Goyal 2010-09-15 770
8e89d13f4ede24 Vivek Goyal 2010-09-15 771 if (tg->io_disp[rw] >= io_trim)
8e89d13f4ede24 Vivek Goyal 2010-09-15 772 tg->io_disp[rw] -= io_trim;
8e89d13f4ede24 Vivek Goyal 2010-09-15 773 else
8e89d13f4ede24 Vivek Goyal 2010-09-15 774 tg->io_disp[rw] = 0;
8e89d13f4ede24 Vivek Goyal 2010-09-15 775
1a46ae7773b475 Yu Kuai 2023-08-15 776 tg->slice_start[rw] += time_elapsed;
e43473b7f223ec Vivek Goyal 2010-09-15 777
fda6f272c77a7a Tejun Heo 2013-05-14 778 throtl_log(&tg->service_queue,
fda6f272c77a7a Tejun Heo 2013-05-14 779 "[%c] trim slice nr=%lu bytes=%llu io=%lu start=%lu end=%lu jiffies=%lu",
8e89d13f4ede24 Vivek Goyal 2010-09-15 @780 rw == READ ? 'R' : 'W', nr_slices, bytes_trim, io_trim,
e43473b7f223ec Vivek Goyal 2010-09-15 781 tg->slice_start[rw], tg->slice_end[rw], jiffies);
e43473b7f223ec Vivek Goyal 2010-09-15 782 }
e43473b7f223ec Vivek Goyal 2010-09-15 783
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2023-08-15 15:48 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-08-15 1:41 [PATCH -next 0/4] cleanup and fixes for 'carryover_ios/bytes' Yu Kuai
2023-08-15 1:41 ` [PATCH -next 1/4] blk-throttle: print signed value 'carryover_bytes/ios' for user Yu Kuai
[not found] ` <20230815014123.368929-1-yukuai1-XF6JlduFytWkHkcT6e4Xnw@public.gmane.org>
2023-08-15 1:41 ` [PATCH -next 2/4] blk-throttle: fix wrong comparation while 'carryover_ios/bytes' is negative Yu Kuai
2023-08-15 1:41 ` [PATCH -next 3/4] blk-throttle: use calculate_io/bytes_allowed() for throtl_trim_slice() Yu Kuai
2023-08-15 15:48 ` kernel test robot
2023-08-15 1:41 ` [PATCH -next 4/4] blk-throttle: consider 'carryover_ios/bytes' in throtl_trim_slice() Yu Kuai
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).