Skip to contentSkip to conversation

RDMA/efa: Use strscpy instead of strlcpy

PATCH Superseded Gal Pressman 2 participants 7 messages Mar 16, 2021 – Mar 29, 2021

Subsystems (2)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 RDMA/efa: Use strscpy instead of strlcpy

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

---
 drivers/infiniband/hw/efa/efa_main.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/infiniband/hw/efa/efa_main.c +3 -3
index 0f578734bddb..a8da96ef2be8 100644
@@ -1,6 +1,6 @@
 // SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause
 /*
- * Copyright 2018-2020 Amazon.com, Inc. or its affiliates. All rights reserved.
+ * Copyright 2018-2021 Amazon.com, Inc. or its affiliates. All rights reserved.
  */
 
 #include <linux/module.h>
@@ -209,10 +209,10 @@ static void efa_set_host_info(struct efa_dev *dev)
 	if (!hinf)
 		return;
 
-	strlcpy(hinf->os_dist_str, utsname()->release,
+	strscpy(hinf->os_dist_str, utsname()->release,
 		min(sizeof(hinf->os_dist_str), sizeof(utsname()->release)));
 	hinf->os_type = EFA_ADMIN_OS_LINUX;
-	strlcpy(hinf->kernel_ver_str, utsname()->version,
+	strscpy(hinf->kernel_ver_str, utsname()->version,
 		min(sizeof(hinf->kernel_ver_str), sizeof(utsname()->version)));
 	hinf->kernel_ver = LINUX_VERSION_CODE;
 	EFA_SET(&hinf->driver_ver, EFA_ADMIN_HOST_INFO_DRIVER_MAJOR, 0);
Signature · 1 line
2.31.0
maintainer
#1 lore raw
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

Jason

#2 lore raw
Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

It's not a bug as is, but it addresses checkpatch's warning:
WARNING: Prefer strscpy over strlcpy - see: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/

maintainer
#3 lore raw
Quoted 11 lines · Gal Pressman
Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

It's not a bug as is, but it addresses checkpatch's warning:
WARNING: Prefer strscpy over strlcpy - see: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/

Okay.. but why is it so weird:

        strscpy(hinf->kernel_ver_str, utsname()->version,
                min(sizeof(hinf->kernel_ver_str), sizeof(utsname()->version)));

?

utsname()->version is null terminated, yes? Why does it need to be min'd?

Jason

#4 lore raw
Quoted 15 lines · Jason Gunthorpe
Quoted 11 lines · Gal Pressman
Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

It's not a bug as is, but it addresses checkpatch's warning:
WARNING: Prefer strscpy over strlcpy - see: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/

Okay.. but why is it so weird:

Quoted 7 lines
        strscpy(hinf->kernel_ver_str, utsname()->version,
                min(sizeof(hinf->kernel_ver_str), sizeof(utsname()->version)));

?

utsname()->version is null terminated, yes? Why does it need to be min'd?

The size of the kernel buffer is different than the device buffer (65B vs 32B), the min() is there to prevent overflow regardless of the NULL termination. A NULL terminated 60 bytes utsname would be truncated to 32 bytes.

maintainer
#5 lore raw
Quoted 19 lines · Gal Pressman
Quoted 18 lines · Jason Gunthorpe
Quoted 11 lines · Gal Pressman
Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

It's not a bug as is, but it addresses checkpatch's warning:
WARNING: Prefer strscpy over strlcpy - see: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/

Okay.. but why is it so weird:

        strscpy(hinf->kernel_ver_str, utsname()->version,
                min(sizeof(hinf->kernel_ver_str), sizeof(utsname()->version)));
Quoted 8 lines
Quoted 4 lines

?

utsname()->version is null terminated, yes? Why does it need to be min'd?

The size of the kernel buffer is different than the device buffer (65B vs 32B), the min() is there to prevent overflow regardless of the NULL termination. A NULL terminated 60 bytes utsname would be truncated to 32 bytes.

I don't understand.

If version is NULL terminated than this:

strscpy(hinf->kernel_ver_str, utsname()->version,

           sizeof(hinf->kernel_ver_str))

Is the only thing needed? The whole point of strscpy is that it truncates the string to fit the output.

Jason

#6 lore raw
Quoted 31 lines · Jason Gunthorpe
Quoted 19 lines · Gal Pressman
Quoted 18 lines · Jason Gunthorpe
Quoted 11 lines · Gal Pressman
Quoted 6 lines · Jason Gunthorpe
Quoted 2 lines · Gal Pressman

The strlcpy function doesn't limit the source length, use the preferred strscpy function instead.

Why do we need to limit the source length here? Either this is a bug because the source string is no NULL terminated or it is OK as is?

It's not a bug as is, but it addresses checkpatch's warning:
WARNING: Prefer strscpy over strlcpy - see: https://lore.kernel.org/r/CAHk-=wgfRnXz0W3D37d01q3JFkr_i_uTL=V6A6G1oUZcprmknw@mail.gmail.com/

Okay.. but why is it so weird:

        strscpy(hinf->kernel_ver_str, utsname()->version,
                min(sizeof(hinf->kernel_ver_str), sizeof(utsname()->version)));
Quoted 8 lines
Quoted 4 lines

?

utsname()->version is null terminated, yes? Why does it need to be min'd?

The size of the kernel buffer is different than the device buffer (65B vs 32B), the min() is there to prevent overflow regardless of the NULL termination. A NULL terminated 60 bytes utsname would be truncated to 32 bytes.

I don't understand.

Quoted 7 lines

If version is NULL terminated than this:

   strscpy(hinf->kernel_ver_str, utsname()->version,
           sizeof(hinf->kernel_ver_str))

Is the only thing needed? The whole point of strscpy is that it truncates the string to fit the output.

You're right, will update the patch.
Thanks Jason