* [PATCH v3 0/5] blk-iocost: random bugfix
@ 2022-12-26 8:58 Yu Kuai
2022-12-26 8:58 ` Yu Kuai
` (4 more replies)
0 siblings, 5 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Yu Kuai <yukuai3@huawei.com>
changes in v3:
- move some patches into separate patchset
- don't return other error number for match_u64() in patch 1
- instead of checking user input separately, set page directly if
'bps + IOC_PAGE_SIZE' will overflow.
Li Nan (2):
blk-iocost: fix divide by 0 error in calc_lcoefs()
blk-iocost: change div64_u64 to DIV64_U64_ROUND_UP in
ioc_refresh_params()
Yu Kuai (3):
blk-iocost: check return value of match_u64()
blk-iocost: don't allow to configure bio based device
blk-iocost: read params inside lock in sysfs apis
block/blk-iocost.c | 31 +++++++++++++++++++++++++------
1 file changed, 25 insertions(+), 6 deletions(-)
--
2.31.1
^ permalink raw reply [flat|nested] 16+ messages in thread
* [PATCH v3 1/5] blk-iocost: check return value of match_u64()
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Yu Kuai <yukuai3@huawei.com>
This patch fixs that the return value of match_u64() from ioc_qos_write()
is not checked,
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
block/blk-iocost.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index 549ddc9e0c6f..93b938517be3 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -3212,7 +3212,8 @@ static ssize_t ioc_qos_write(struct kernfs_open_file *of, char *input,
switch (match_token(p, qos_ctrl_tokens, args)) {
case QOS_ENABLE:
- match_u64(&args[0], &v);
+ if (match_u64(&args[0], &v))
+ goto einval;
enable = v;
continue;
case QOS_CTRL:
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 1/5] blk-iocost: check return value of match_u64()
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai3-hv44wF8Li93QT0dZR+AlfA, yukuai1-XF6JlduFytWkHkcT6e4Xnw,
yi.zhang-hv44wF8Li93QT0dZR+AlfA
From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
This patch fixs that the return value of match_u64() from ioc_qos_write()
is not checked,
Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
---
block/blk-iocost.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index 549ddc9e0c6f..93b938517be3 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -3212,7 +3212,8 @@ static ssize_t ioc_qos_write(struct kernfs_open_file *of, char *input,
switch (match_token(p, qos_ctrl_tokens, args)) {
case QOS_ENABLE:
- match_u64(&args[0], &v);
+ if (match_u64(&args[0], &v))
+ goto einval;
enable = v;
continue;
case QOS_CTRL:
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 2/5] blk-iocost: don't allow to configure bio based device
2022-12-26 8:58 [PATCH v3 0/5] blk-iocost: random bugfix Yu Kuai
2022-12-26 8:58 ` Yu Kuai
@ 2022-12-26 8:58 ` Yu Kuai
2022-12-26 8:58 ` Yu Kuai
` (2 subsequent siblings)
4 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Yu Kuai <yukuai3@huawei.com>
iocost is based on rq_qos, which can only work for request based device,
thus it doesn't make sense to configure iocost for bio based device.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Tejun Heo <tj@kernel.org>
---
block/blk-iocost.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index 93b938517be3..ce4eafbada56 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -3185,6 +3185,11 @@ static ssize_t ioc_qos_write(struct kernfs_open_file *of, char *input,
return PTR_ERR(bdev);
disk = bdev->bd_disk;
+ if (!queue_is_mq(disk->queue)) {
+ ret = -EOPNOTSUPP;
+ goto err;
+ }
+
ioc = q_to_ioc(disk->queue);
if (!ioc) {
ret = blk_iocost_init(disk);
@@ -3365,6 +3370,11 @@ static ssize_t ioc_cost_model_write(struct kernfs_open_file *of, char *input,
return PTR_ERR(bdev);
q = bdev_get_queue(bdev);
+ if (!queue_is_mq(q)) {
+ ret = -EOPNOTSUPP;
+ goto err;
+ }
+
ioc = q_to_ioc(q);
if (!ioc) {
ret = blk_iocost_init(bdev->bd_disk);
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 3/5] blk-iocost: read params inside lock in sysfs apis
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Yu Kuai <yukuai3@huawei.com>
Otherwise, user might get abnormal values if params is updated
concurrently.
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Tejun Heo <tj@kernel.org>
---
block/blk-iocost.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index ce4eafbada56..f8726e20da20 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -3129,6 +3129,7 @@ static u64 ioc_qos_prfill(struct seq_file *sf, struct blkg_policy_data *pd,
if (!dname)
return 0;
+ spin_lock_irq(&ioc->lock);
seq_printf(sf, "%s enable=%d ctrl=%s rpct=%u.%02u rlat=%u wpct=%u.%02u wlat=%u min=%u.%02u max=%u.%02u\n",
dname, ioc->enabled, ioc->user_qos_params ? "user" : "auto",
ioc->params.qos[QOS_RPPM] / 10000,
@@ -3141,6 +3142,7 @@ static u64 ioc_qos_prfill(struct seq_file *sf, struct blkg_policy_data *pd,
ioc->params.qos[QOS_MIN] % 10000 / 100,
ioc->params.qos[QOS_MAX] / 10000,
ioc->params.qos[QOS_MAX] % 10000 / 100);
+ spin_unlock_irq(&ioc->lock);
return 0;
}
@@ -3320,12 +3322,14 @@ static u64 ioc_cost_model_prfill(struct seq_file *sf,
if (!dname)
return 0;
+ spin_lock_irq(&ioc->lock);
seq_printf(sf, "%s ctrl=%s model=linear "
"rbps=%llu rseqiops=%llu rrandiops=%llu "
"wbps=%llu wseqiops=%llu wrandiops=%llu\n",
dname, ioc->user_cost_model ? "user" : "auto",
u[I_LCOEF_RBPS], u[I_LCOEF_RSEQIOPS], u[I_LCOEF_RRANDIOPS],
u[I_LCOEF_WBPS], u[I_LCOEF_WSEQIOPS], u[I_LCOEF_WRANDIOPS]);
+ spin_unlock_irq(&ioc->lock);
return 0;
}
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 3/5] blk-iocost: read params inside lock in sysfs apis
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai3-hv44wF8Li93QT0dZR+AlfA, yukuai1-XF6JlduFytWkHkcT6e4Xnw,
yi.zhang-hv44wF8Li93QT0dZR+AlfA
From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Otherwise, user might get abnormal values if params is updated
concurrently.
Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Acked-by: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
---
block/blk-iocost.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index ce4eafbada56..f8726e20da20 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -3129,6 +3129,7 @@ static u64 ioc_qos_prfill(struct seq_file *sf, struct blkg_policy_data *pd,
if (!dname)
return 0;
+ spin_lock_irq(&ioc->lock);
seq_printf(sf, "%s enable=%d ctrl=%s rpct=%u.%02u rlat=%u wpct=%u.%02u wlat=%u min=%u.%02u max=%u.%02u\n",
dname, ioc->enabled, ioc->user_qos_params ? "user" : "auto",
ioc->params.qos[QOS_RPPM] / 10000,
@@ -3141,6 +3142,7 @@ static u64 ioc_qos_prfill(struct seq_file *sf, struct blkg_policy_data *pd,
ioc->params.qos[QOS_MIN] % 10000 / 100,
ioc->params.qos[QOS_MAX] / 10000,
ioc->params.qos[QOS_MAX] % 10000 / 100);
+ spin_unlock_irq(&ioc->lock);
return 0;
}
@@ -3320,12 +3322,14 @@ static u64 ioc_cost_model_prfill(struct seq_file *sf,
if (!dname)
return 0;
+ spin_lock_irq(&ioc->lock);
seq_printf(sf, "%s ctrl=%s model=linear "
"rbps=%llu rseqiops=%llu rrandiops=%llu "
"wbps=%llu wseqiops=%llu wrandiops=%llu\n",
dname, ioc->user_cost_model ? "user" : "auto",
u[I_LCOEF_RBPS], u[I_LCOEF_RSEQIOPS], u[I_LCOEF_RRANDIOPS],
u[I_LCOEF_WBPS], u[I_LCOEF_WSEQIOPS], u[I_LCOEF_WRANDIOPS]);
+ spin_unlock_irq(&ioc->lock);
return 0;
}
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs()
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Li Nan <linan122@huawei.com>
echo max of u64 to cost.model can cause divide by 0 error.
# echo 8:0 rbps=18446744073709551615 > /sys/fs/cgroup/io.cost.model
divide error: 0000 [#1] PREEMPT SMP
RIP: 0010:calc_lcoefs+0x4c/0xc0
Call Trace:
<TASK>
ioc_refresh_params+0x2b3/0x4f0
ioc_cost_model_write+0x3cb/0x4c0
? _copy_from_iter+0x6d/0x6c0
? kernfs_fop_write_iter+0xfc/0x270
cgroup_file_write+0xa0/0x200
kernfs_fop_write_iter+0x17d/0x270
vfs_write+0x414/0x620
ksys_write+0x73/0x160
__x64_sys_write+0x1e/0x30
do_syscall_64+0x35/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
calc_lcoefs() uses the input value of cost.model in DIV_ROUND_UP_ULL,
overflow would happen if bps plus IOC_PAGE_SIZE is greater than
ULLONG_MAX, it can cause divide by 0 error.
Fix the problem by setting basecost
Signed-off-by: Li Nan <linan122@huawei.com>
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
block/blk-iocost.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index f8726e20da20..c6b39024117b 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -866,9 +866,13 @@ static void calc_lcoefs(u64 bps, u64 seqiops, u64 randiops,
*page = *seqio = *randio = 0;
- if (bps)
- *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
- DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
+ if (bps) {
+ if (bps >= U64_MAX - IOC_PAGE_SIZE)
+ *page = 1;
+ else
+ *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
+ DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
+ }
if (seqiops) {
v = DIV64_U64_ROUND_UP(VTIME_PER_SEC, seqiops);
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs()
@ 2022-12-26 8:58 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj-DgEjT+Ai2ygdnm+yROfE0A, josef-DigfWCa+lFGyeJad7bwFQA,
axboe-tSWWG44O7X1aa/9Udqfwiw
Cc: cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai3-hv44wF8Li93QT0dZR+AlfA, yukuai1-XF6JlduFytWkHkcT6e4Xnw,
yi.zhang-hv44wF8Li93QT0dZR+AlfA
From: Li Nan <linan122-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
echo max of u64 to cost.model can cause divide by 0 error.
# echo 8:0 rbps=18446744073709551615 > /sys/fs/cgroup/io.cost.model
divide error: 0000 [#1] PREEMPT SMP
RIP: 0010:calc_lcoefs+0x4c/0xc0
Call Trace:
<TASK>
ioc_refresh_params+0x2b3/0x4f0
ioc_cost_model_write+0x3cb/0x4c0
? _copy_from_iter+0x6d/0x6c0
? kernfs_fop_write_iter+0xfc/0x270
cgroup_file_write+0xa0/0x200
kernfs_fop_write_iter+0x17d/0x270
vfs_write+0x414/0x620
ksys_write+0x73/0x160
__x64_sys_write+0x1e/0x30
do_syscall_64+0x35/0x80
entry_SYSCALL_64_after_hwframe+0x63/0xcd
calc_lcoefs() uses the input value of cost.model in DIV_ROUND_UP_ULL,
overflow would happen if bps plus IOC_PAGE_SIZE is greater than
ULLONG_MAX, it can cause divide by 0 error.
Fix the problem by setting basecost
Signed-off-by: Li Nan <linan122-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
---
block/blk-iocost.c | 10 +++++++---
1 file changed, 7 insertions(+), 3 deletions(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index f8726e20da20..c6b39024117b 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -866,9 +866,13 @@ static void calc_lcoefs(u64 bps, u64 seqiops, u64 randiops,
*page = *seqio = *randio = 0;
- if (bps)
- *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
- DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
+ if (bps) {
+ if (bps >= U64_MAX - IOC_PAGE_SIZE)
+ *page = 1;
+ else
+ *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
+ DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
+ }
if (seqiops) {
v = DIV64_U64_ROUND_UP(VTIME_PER_SEC, seqiops);
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* [PATCH v3 5/5] blk-iocost: change div64_u64 to DIV64_U64_ROUND_UP in ioc_refresh_params()
2022-12-26 8:58 [PATCH v3 0/5] blk-iocost: random bugfix Yu Kuai
` (3 preceding siblings ...)
2022-12-26 8:58 ` Yu Kuai
@ 2022-12-26 8:58 ` Yu Kuai
2023-01-04 21:48 ` Tejun Heo
4 siblings, 1 reply; 16+ messages in thread
From: Yu Kuai @ 2022-12-26 8:58 UTC (permalink / raw)
To: tj, josef, axboe
Cc: cgroups, linux-block, linux-kernel, yukuai3, yukuai1, yi.zhang
From: Li Nan <linan122@huawei.com>
vrate_min is calculated by DIV64_U64_ROUND_UP, but vrate_max is calculated
by div64_u64. Vrate_min may be 1 greater than vrate_max if the input
values min and max of cost.qos are equal.
Signed-off-by: Li Nan <linan122@huawei.com>
Signed-off-by: Yu Kuai <yukuai3@huawei.com>
---
block/blk-iocost.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/block/blk-iocost.c b/block/blk-iocost.c
index c6b39024117b..7a0d754b9eb2 100644
--- a/block/blk-iocost.c
+++ b/block/blk-iocost.c
@@ -930,8 +930,8 @@ static bool ioc_refresh_params(struct ioc *ioc, bool force)
ioc->vrate_min = DIV64_U64_ROUND_UP((u64)ioc->params.qos[QOS_MIN] *
VTIME_PER_USEC, MILLION);
- ioc->vrate_max = div64_u64((u64)ioc->params.qos[QOS_MAX] *
- VTIME_PER_USEC, MILLION);
+ ioc->vrate_max = DIV64_U64_ROUND_UP((u64)ioc->params.qos[QOS_MAX] *
+ VTIME_PER_USEC, MILLION);
return true;
}
--
2.31.1
^ permalink raw reply related [flat|nested] 16+ messages in thread
* Re: [PATCH v3 1/5] blk-iocost: check return value of match_u64()
@ 2023-01-04 21:48 ` Tejun Heo
0 siblings, 0 replies; 16+ messages in thread
From: Tejun Heo @ 2023-01-04 21:48 UTC (permalink / raw)
To: Yu Kuai
Cc: josef, axboe, cgroups, linux-block, linux-kernel, yukuai3, yi.zhang
On Mon, Dec 26, 2022 at 04:58:55PM +0800, Yu Kuai wrote:
> From: Yu Kuai <yukuai3@huawei.com>
>
> This patch fixs that the return value of match_u64() from ioc_qos_write()
> is not checked,
>
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Tejun Heo <tj@kernel.org>
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 1/5] blk-iocost: check return value of match_u64()
@ 2023-01-04 21:48 ` Tejun Heo
0 siblings, 0 replies; 16+ messages in thread
From: Tejun Heo @ 2023-01-04 21:48 UTC (permalink / raw)
To: Yu Kuai
Cc: josef-DigfWCa+lFGyeJad7bwFQA, axboe-tSWWG44O7X1aa/9Udqfwiw,
cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai3-hv44wF8Li93QT0dZR+AlfA, yi.zhang-hv44wF8Li93QT0dZR+AlfA
On Mon, Dec 26, 2022 at 04:58:55PM +0800, Yu Kuai wrote:
> From: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
>
> This patch fixs that the return value of match_u64() from ioc_qos_write()
> is not checked,
>
> Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Acked-by: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 5/5] blk-iocost: change div64_u64 to DIV64_U64_ROUND_UP in ioc_refresh_params()
@ 2023-01-04 21:48 ` Tejun Heo
0 siblings, 0 replies; 16+ messages in thread
From: Tejun Heo @ 2023-01-04 21:48 UTC (permalink / raw)
To: Yu Kuai
Cc: josef, axboe, cgroups, linux-block, linux-kernel, yukuai3, yi.zhang
On Mon, Dec 26, 2022 at 04:58:59PM +0800, Yu Kuai wrote:
> From: Li Nan <linan122@huawei.com>
>
> vrate_min is calculated by DIV64_U64_ROUND_UP, but vrate_max is calculated
> by div64_u64. Vrate_min may be 1 greater than vrate_max if the input
> values min and max of cost.qos are equal.
>
> Signed-off-by: Li Nan <linan122@huawei.com>
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
Acked-by: Tejun Heo <tj@kernel.org>
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 5/5] blk-iocost: change div64_u64 to DIV64_U64_ROUND_UP in ioc_refresh_params()
@ 2023-01-04 21:48 ` Tejun Heo
0 siblings, 0 replies; 16+ messages in thread
From: Tejun Heo @ 2023-01-04 21:48 UTC (permalink / raw)
To: Yu Kuai
Cc: josef-DigfWCa+lFGyeJad7bwFQA, axboe-tSWWG44O7X1aa/9Udqfwiw,
cgroups-u79uwXL29TY76Z2rM5mHXA,
linux-block-u79uwXL29TY76Z2rM5mHXA,
linux-kernel-u79uwXL29TY76Z2rM5mHXA,
yukuai3-hv44wF8Li93QT0dZR+AlfA, yi.zhang-hv44wF8Li93QT0dZR+AlfA
On Mon, Dec 26, 2022 at 04:58:59PM +0800, Yu Kuai wrote:
> From: Li Nan <linan122-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
>
> vrate_min is calculated by DIV64_U64_ROUND_UP, but vrate_max is calculated
> by div64_u64. Vrate_min may be 1 greater than vrate_max if the input
> values min and max of cost.qos are equal.
>
> Signed-off-by: Li Nan <linan122-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
> Signed-off-by: Yu Kuai <yukuai3-hv44wF8Li93QT0dZR+AlfA@public.gmane.org>
Acked-by: Tejun Heo <tj-DgEjT+Ai2ygdnm+yROfE0A@public.gmane.org>
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs()
2022-12-26 8:58 ` Yu Kuai
(?)
@ 2023-01-04 21:54 ` Tejun Heo
2023-01-05 1:53 ` Yu Kuai
-1 siblings, 1 reply; 16+ messages in thread
From: Tejun Heo @ 2023-01-04 21:54 UTC (permalink / raw)
To: Yu Kuai
Cc: josef, axboe, cgroups, linux-block, linux-kernel, yukuai3, yi.zhang
On Mon, Dec 26, 2022 at 04:58:58PM +0800, Yu Kuai wrote:
> From: Li Nan <linan122@huawei.com>
>
> echo max of u64 to cost.model can cause divide by 0 error.
>
> # echo 8:0 rbps=18446744073709551615 > /sys/fs/cgroup/io.cost.model
>
> divide error: 0000 [#1] PREEMPT SMP
> RIP: 0010:calc_lcoefs+0x4c/0xc0
> Call Trace:
> <TASK>
> ioc_refresh_params+0x2b3/0x4f0
> ioc_cost_model_write+0x3cb/0x4c0
> ? _copy_from_iter+0x6d/0x6c0
> ? kernfs_fop_write_iter+0xfc/0x270
> cgroup_file_write+0xa0/0x200
> kernfs_fop_write_iter+0x17d/0x270
> vfs_write+0x414/0x620
> ksys_write+0x73/0x160
> __x64_sys_write+0x1e/0x30
> do_syscall_64+0x35/0x80
> entry_SYSCALL_64_after_hwframe+0x63/0xcd
>
> calc_lcoefs() uses the input value of cost.model in DIV_ROUND_UP_ULL,
> overflow would happen if bps plus IOC_PAGE_SIZE is greater than
> ULLONG_MAX, it can cause divide by 0 error.
>
> Fix the problem by setting basecost
>
> Signed-off-by: Li Nan <linan122@huawei.com>
> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
> ---
> block/blk-iocost.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)
>
> diff --git a/block/blk-iocost.c b/block/blk-iocost.c
> index f8726e20da20..c6b39024117b 100644
> --- a/block/blk-iocost.c
> +++ b/block/blk-iocost.c
> @@ -866,9 +866,13 @@ static void calc_lcoefs(u64 bps, u64 seqiops, u64 randiops,
>
> *page = *seqio = *randio = 0;
>
> - if (bps)
> - *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
> - DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
> + if (bps) {
> + if (bps >= U64_MAX - IOC_PAGE_SIZE)
> + *page = 1;
> + else
> + *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
> + DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
> + }
This is a nitpick but wouldn't something like the following be easier to
understand?
if (bps) {
u64 bps_pages = DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE);
if (bps_pages)
*pages = DIV64_U64_ROUND_UP(VTIME_PER_SEC, bps_pages);
else
*pages = 1;
}
Out of scope but this seems more like a bug in the DIV macros. The fact that
it returns 0 is an implementation artifact more than anything and a
surprising one at that as it ends up returning 0 for an input that a regular
division would handle just fine and the rounded up result fits well within
the result type.
Thanks.
--
tejun
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs()
2023-01-04 21:54 ` Tejun Heo
@ 2023-01-05 1:53 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2023-01-05 1:53 UTC (permalink / raw)
To: Tejun Heo, Yu Kuai
Cc: josef, axboe, cgroups, linux-block, linux-kernel, yi.zhang, yukuai (C)
Hi,
在 2023/01/05 5:54, Tejun Heo 写道:
> On Mon, Dec 26, 2022 at 04:58:58PM +0800, Yu Kuai wrote:
>> From: Li Nan <linan122@huawei.com>
>>
>> echo max of u64 to cost.model can cause divide by 0 error.
>>
>> # echo 8:0 rbps=18446744073709551615 > /sys/fs/cgroup/io.cost.model
>>
>> divide error: 0000 [#1] PREEMPT SMP
>> RIP: 0010:calc_lcoefs+0x4c/0xc0
>> Call Trace:
>> <TASK>
>> ioc_refresh_params+0x2b3/0x4f0
>> ioc_cost_model_write+0x3cb/0x4c0
>> ? _copy_from_iter+0x6d/0x6c0
>> ? kernfs_fop_write_iter+0xfc/0x270
>> cgroup_file_write+0xa0/0x200
>> kernfs_fop_write_iter+0x17d/0x270
>> vfs_write+0x414/0x620
>> ksys_write+0x73/0x160
>> __x64_sys_write+0x1e/0x30
>> do_syscall_64+0x35/0x80
>> entry_SYSCALL_64_after_hwframe+0x63/0xcd
>>
>> calc_lcoefs() uses the input value of cost.model in DIV_ROUND_UP_ULL,
>> overflow would happen if bps plus IOC_PAGE_SIZE is greater than
>> ULLONG_MAX, it can cause divide by 0 error.
>>
>> Fix the problem by setting basecost
>>
>> Signed-off-by: Li Nan <linan122@huawei.com>
>> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
>> ---
>> block/blk-iocost.c | 10 +++++++---
>> 1 file changed, 7 insertions(+), 3 deletions(-)
>>
>> diff --git a/block/blk-iocost.c b/block/blk-iocost.c
>> index f8726e20da20..c6b39024117b 100644
>> --- a/block/blk-iocost.c
>> +++ b/block/blk-iocost.c
>> @@ -866,9 +866,13 @@ static void calc_lcoefs(u64 bps, u64 seqiops, u64 randiops,
>>
>> *page = *seqio = *randio = 0;
>>
>> - if (bps)
>> - *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
>> - DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
>> + if (bps) {
>> + if (bps >= U64_MAX - IOC_PAGE_SIZE)
>> + *page = 1;
>> + else
>> + *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
>> + DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
>> + }
>
> This is a nitpick but wouldn't something like the following be easier to
> understand?
>
> if (bps) {
> u64 bps_pages = DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE);
>
> if (bps_pages)
> *pages = DIV64_U64_ROUND_UP(VTIME_PER_SEC, bps_pages);
> else
> *pages = 1;
> }
>
Yes, I agree that this is better to understand. I'll send a new version.
Thanks,
Kuai
^ permalink raw reply [flat|nested] 16+ messages in thread
* Re: [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs()
@ 2023-01-05 1:53 ` Yu Kuai
0 siblings, 0 replies; 16+ messages in thread
From: Yu Kuai @ 2023-01-05 1:53 UTC (permalink / raw)
To: Tejun Heo, Yu Kuai
Cc: josef, axboe, cgroups, linux-block, linux-kernel, yi.zhang, yukuai (C)
Hi,
ÔÚ 2023/01/05 5:54, Tejun Heo дµÀ:
> On Mon, Dec 26, 2022 at 04:58:58PM +0800, Yu Kuai wrote:
>> From: Li Nan <linan122@huawei.com>
>>
>> echo max of u64 to cost.model can cause divide by 0 error.
>>
>> # echo 8:0 rbps=18446744073709551615 > /sys/fs/cgroup/io.cost.model
>>
>> divide error: 0000 [#1] PREEMPT SMP
>> RIP: 0010:calc_lcoefs+0x4c/0xc0
>> Call Trace:
>> <TASK>
>> ioc_refresh_params+0x2b3/0x4f0
>> ioc_cost_model_write+0x3cb/0x4c0
>> ? _copy_from_iter+0x6d/0x6c0
>> ? kernfs_fop_write_iter+0xfc/0x270
>> cgroup_file_write+0xa0/0x200
>> kernfs_fop_write_iter+0x17d/0x270
>> vfs_write+0x414/0x620
>> ksys_write+0x73/0x160
>> __x64_sys_write+0x1e/0x30
>> do_syscall_64+0x35/0x80
>> entry_SYSCALL_64_after_hwframe+0x63/0xcd
>>
>> calc_lcoefs() uses the input value of cost.model in DIV_ROUND_UP_ULL,
>> overflow would happen if bps plus IOC_PAGE_SIZE is greater than
>> ULLONG_MAX, it can cause divide by 0 error.
>>
>> Fix the problem by setting basecost
>>
>> Signed-off-by: Li Nan <linan122@huawei.com>
>> Signed-off-by: Yu Kuai <yukuai3@huawei.com>
>> ---
>> block/blk-iocost.c | 10 +++++++---
>> 1 file changed, 7 insertions(+), 3 deletions(-)
>>
>> diff --git a/block/blk-iocost.c b/block/blk-iocost.c
>> index f8726e20da20..c6b39024117b 100644
>> --- a/block/blk-iocost.c
>> +++ b/block/blk-iocost.c
>> @@ -866,9 +866,13 @@ static void calc_lcoefs(u64 bps, u64 seqiops, u64 randiops,
>>
>> *page = *seqio = *randio = 0;
>>
>> - if (bps)
>> - *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
>> - DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
>> + if (bps) {
>> + if (bps >= U64_MAX - IOC_PAGE_SIZE)
>> + *page = 1;
>> + else
>> + *page = DIV64_U64_ROUND_UP(VTIME_PER_SEC,
>> + DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE));
>> + }
>
> This is a nitpick but wouldn't something like the following be easier to
> understand?
>
> if (bps) {
> u64 bps_pages = DIV_ROUND_UP_ULL(bps, IOC_PAGE_SIZE);
>
> if (bps_pages)
> *pages = DIV64_U64_ROUND_UP(VTIME_PER_SEC, bps_pages);
> else
> *pages = 1;
> }
>
Yes, I agree that this is better to understand. I'll send a new version.
Thanks,
Kuai
^ permalink raw reply [flat|nested] 16+ messages in thread
end of thread, other threads:[~2023-01-05 1:53 UTC | newest]
Thread overview: 16+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-12-26 8:58 [PATCH v3 0/5] blk-iocost: random bugfix Yu Kuai
2022-12-26 8:58 ` [PATCH v3 1/5] blk-iocost: check return value of match_u64() Yu Kuai
2022-12-26 8:58 ` Yu Kuai
2023-01-04 21:48 ` Tejun Heo
2023-01-04 21:48 ` Tejun Heo
2022-12-26 8:58 ` [PATCH v3 2/5] blk-iocost: don't allow to configure bio based device Yu Kuai
2022-12-26 8:58 ` [PATCH v3 3/5] blk-iocost: read params inside lock in sysfs apis Yu Kuai
2022-12-26 8:58 ` Yu Kuai
2022-12-26 8:58 ` [PATCH v3 4/5] blk-iocost: fix divide by 0 error in calc_lcoefs() Yu Kuai
2022-12-26 8:58 ` Yu Kuai
2023-01-04 21:54 ` Tejun Heo
2023-01-05 1:53 ` Yu Kuai
2023-01-05 1:53 ` Yu Kuai
2022-12-26 8:58 ` [PATCH v3 5/5] blk-iocost: change div64_u64 to DIV64_U64_ROUND_UP in ioc_refresh_params() Yu Kuai
2023-01-04 21:48 ` Tejun Heo
2023-01-04 21:48 ` Tejun Heo
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.