* [PATCH 0/2] Convert copy_from_user() + kstrtouint() to kstrtouint_from_user()
@ 2026-01-16 13:17 Fushuai Wang
2026-01-16 13:17 ` [PATCH 1/2] sched/debug: " Fushuai Wang
2026-01-16 13:17 ` [PATCH 2/2] x86/tlb: " Fushuai Wang
0 siblings, 2 replies; 4+ messages in thread
From: Fushuai Wang @ 2026-01-16 13:17 UTC (permalink / raw)
To: mingo, peterz, juri.lelli, vincent.guittot, dietmar.eggemann,
rostedt, bsegall, mgorman, vschneid, dave.hansen, luto, tglx, bp,
hpa
Cc: linux-kernel, x86, wangfushuai, aliceryhl, ynorov
From: Fushuai Wang <wangfushuai@baidu.com>
Hi,
There are several places use copy_from_user() to copy a string from
userspase, followed by kstrtoXXX() to convert it to a desired type.
This approach requires manually handling NUL-termination and proper
buffer sizing, which is error-prone and unnecessarily complex. The
kernel provides kstrtoXXX_from_user() helpers to copy a string from
userspace and perform the conversion in one step. So use a
kstrtoXXX_from_user() instead of copy_from_user() + kstrtoXXX().
Recent discussion:
https://lore.kernel.org/all/20260112073039.1185-1-fushuai.wang@linux.dev/T/
Fushuai Wang (2):
sched/debug: Convert copy_from_user() + kstrtouint() to
kstrtouint_from_user()
x86/tlb: Convert copy_from_user() + kstrtouint() to
kstrtouint_from_user()
arch/x86/mm/tlb.c | 13 ++++---------
kernel/sched/debug.c | 14 ++++----------
2 files changed, 8 insertions(+), 19 deletions(-)
--
2.36.1
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 1/2] sched/debug: Convert copy_from_user() + kstrtouint() to kstrtouint_from_user()
2026-01-16 13:17 [PATCH 0/2] Convert copy_from_user() + kstrtouint() to kstrtouint_from_user() Fushuai Wang
@ 2026-01-16 13:17 ` Fushuai Wang
2026-01-16 17:04 ` Yury Norov
2026-01-16 13:17 ` [PATCH 2/2] x86/tlb: " Fushuai Wang
1 sibling, 1 reply; 4+ messages in thread
From: Fushuai Wang @ 2026-01-16 13:17 UTC (permalink / raw)
To: mingo, peterz, juri.lelli, vincent.guittot, dietmar.eggemann,
rostedt, bsegall, mgorman, vschneid, dave.hansen, luto, tglx, bp,
hpa
Cc: linux-kernel, x86, wangfushuai, aliceryhl, ynorov
From: Fushuai Wang <wangfushuai@baidu.com>
Using kstrtouint_from_user() instead of copy_from_user() + kstrtouint()
makes the code simpler and less error-prone.
Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
---
kernel/sched/debug.c | 14 ++++----------
1 file changed, 4 insertions(+), 10 deletions(-)
diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c
index 41caa22e0680..1091f9046260 100644
--- a/kernel/sched/debug.c
+++ b/kernel/sched/debug.c
@@ -172,18 +172,12 @@ static const struct file_operations sched_feat_fops = {
static ssize_t sched_scaling_write(struct file *filp, const char __user *ubuf,
size_t cnt, loff_t *ppos)
{
- char buf[16];
unsigned int scaling;
+ int ret;
- if (cnt > 15)
- cnt = 15;
-
- if (copy_from_user(&buf, ubuf, cnt))
- return -EFAULT;
- buf[cnt] = '\0';
-
- if (kstrtouint(buf, 10, &scaling))
- return -EINVAL;
+ ret = kstrtouint_from_user(ubuf, cnt, 10, &scaling);
+ if (ret < 0)
+ return ret;
if (scaling >= SCHED_TUNABLESCALING_END)
return -EINVAL;
--
2.36.1
^ permalink raw reply [flat|nested] 4+ messages in thread
* [PATCH 2/2] x86/tlb: Convert copy_from_user() + kstrtouint() to kstrtouint_from_user()
2026-01-16 13:17 [PATCH 0/2] Convert copy_from_user() + kstrtouint() to kstrtouint_from_user() Fushuai Wang
2026-01-16 13:17 ` [PATCH 1/2] sched/debug: " Fushuai Wang
@ 2026-01-16 13:17 ` Fushuai Wang
1 sibling, 0 replies; 4+ messages in thread
From: Fushuai Wang @ 2026-01-16 13:17 UTC (permalink / raw)
To: mingo, peterz, juri.lelli, vincent.guittot, dietmar.eggemann,
rostedt, bsegall, mgorman, vschneid, dave.hansen, luto, tglx, bp,
hpa
Cc: linux-kernel, x86, wangfushuai, aliceryhl, ynorov
From: Fushuai Wang <wangfushuai@baidu.com>
Using kstrtouint_from_user() instead of copy_from_user() + kstrtouint()
makes the code simpler and less error-prone.
Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
---
arch/x86/mm/tlb.c | 13 ++++---------
1 file changed, 4 insertions(+), 9 deletions(-)
diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c
index f5b93e01e347..5e6a07306214 100644
--- a/arch/x86/mm/tlb.c
+++ b/arch/x86/mm/tlb.c
@@ -1802,17 +1802,12 @@ static ssize_t tlbflush_read_file(struct file *file, char __user *user_buf,
static ssize_t tlbflush_write_file(struct file *file,
const char __user *user_buf, size_t count, loff_t *ppos)
{
- char buf[32];
- ssize_t len;
int ceiling;
+ int err;
- len = min(count, sizeof(buf) - 1);
- if (copy_from_user(buf, user_buf, len))
- return -EFAULT;
-
- buf[len] = '\0';
- if (kstrtoint(buf, 0, &ceiling))
- return -EINVAL;
+ err = kstrtoint_from_user(user_buf, count, 0, &ceiling);
+ if (err)
+ return err;
if (ceiling < 0)
return -EINVAL;
--
2.36.1
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [PATCH 1/2] sched/debug: Convert copy_from_user() + kstrtouint() to kstrtouint_from_user()
2026-01-16 13:17 ` [PATCH 1/2] sched/debug: " Fushuai Wang
@ 2026-01-16 17:04 ` Yury Norov
0 siblings, 0 replies; 4+ messages in thread
From: Yury Norov @ 2026-01-16 17:04 UTC (permalink / raw)
To: Fushuai Wang
Cc: mingo, peterz, juri.lelli, vincent.guittot, dietmar.eggemann,
rostedt, bsegall, mgorman, vschneid, dave.hansen, luto, tglx, bp,
hpa, linux-kernel, x86, wangfushuai, aliceryhl
On Fri, Jan 16, 2026 at 09:17:50PM +0800, Fushuai Wang wrote:
> From: Fushuai Wang <wangfushuai@baidu.com>
>
> Using kstrtouint_from_user() instead of copy_from_user() + kstrtouint()
> makes the code simpler and less error-prone.
>
> Signed-off-by: Fushuai Wang <wangfushuai@baidu.com>
> ---
> kernel/sched/debug.c | 14 ++++----------
> 1 file changed, 4 insertions(+), 10 deletions(-)
>
> diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c
> index 41caa22e0680..1091f9046260 100644
> --- a/kernel/sched/debug.c
> +++ b/kernel/sched/debug.c
> @@ -172,18 +172,12 @@ static const struct file_operations sched_feat_fops = {
> static ssize_t sched_scaling_write(struct file *filp, const char __user *ubuf,
> size_t cnt, loff_t *ppos)
> {
> - char buf[16];
> unsigned int scaling;
> + int ret;
>
> - if (cnt > 15)
> - cnt = 15;
> -
> - if (copy_from_user(&buf, ubuf, cnt))
> - return -EFAULT;
> - buf[cnt] = '\0';
> -
> - if (kstrtouint(buf, 10, &scaling))
> - return -EINVAL;
> + ret = kstrtouint_from_user(ubuf, cnt, 10, &scaling);
> + if (ret < 0)
> + return ret;
This thing returns 0 on success, so I think you could better do:
if (ret)
return ret;
With that, for the series:
Suggested-by: Yury Norov <ynorov@nvidia.com>
Reviewed-by: Yury Norov <ynorov@nvidia.com>
Thanks,
Yury
>
> if (scaling >= SCHED_TUNABLESCALING_END)
> return -EINVAL;
> --
> 2.36.1
>
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2026-01-16 17:06 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2026-01-16 13:17 [PATCH 0/2] Convert copy_from_user() + kstrtouint() to kstrtouint_from_user() Fushuai Wang
2026-01-16 13:17 ` [PATCH 1/2] sched/debug: " Fushuai Wang
2026-01-16 17:04 ` Yury Norov
2026-01-16 13:17 ` [PATCH 2/2] x86/tlb: " Fushuai Wang
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®