* [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
@ 2024-07-14 4:11 Kees Cook
2024-07-15 9:41 ` Simon Horman
2024-07-16 15:00 ` patchwork-bot+netdevbpf
0 siblings, 2 replies; 5+ messages in thread
From: Kees Cook @ 2024-07-14 4:11 UTC (permalink / raw)
To: Eric Dumazet
Cc: Kees Cook, David S. Miller, David Ahern, Jakub Kicinski,
Paolo Abeni, netdev, linux-kernel, linux-hardening
Replace the deprecated[1] uses of strncpy() in tcp_ca_get_name_by_key()
and tcp_get_default_congestion_control(). The callers use the results as
standard C strings (via nla_put_string() and proc handlers respectively),
so trailing padding is not needed.
Since passing the destination buffer arguments decays it to a pointer,
the size can't be trivially determined by the compiler. ca->name is
the same length in both cases, so strscpy() won't fail (when ca->name
is NUL-terminated). Include the length explicitly instead of using the
2-argument strscpy().
Link: https://github.com/KSPP/linux/issues/90 [1]
Signed-off-by: Kees Cook <kees@kernel.org>
---
v2: add tcp_get_default_congestion_control() conversion
v1: https://lore.kernel.org/lkml/20240711171652.work.887-kees@kernel.org/
Cc: Eric Dumazet <edumazet@google.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: David Ahern <dsahern@kernel.org>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: netdev@vger.kernel.org
---
net/ipv4/tcp_cong.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/tcp_cong.c b/net/ipv4/tcp_cong.c
index 28ffcfbeef14..874531c7c08b 100644
--- a/net/ipv4/tcp_cong.c
+++ b/net/ipv4/tcp_cong.c
@@ -203,9 +203,10 @@ char *tcp_ca_get_name_by_key(u32 key, char *buffer)
rcu_read_lock();
ca = tcp_ca_find_key(key);
- if (ca)
- ret = strncpy(buffer, ca->name,
- TCP_CA_NAME_MAX);
+ if (ca) {
+ strscpy(buffer, ca->name, TCP_CA_NAME_MAX);
+ ret = buffer;
+ }
rcu_read_unlock();
return ret;
@@ -338,7 +339,7 @@ void tcp_get_default_congestion_control(struct net *net, char *name)
rcu_read_lock();
ca = rcu_dereference(net->ipv4.tcp_congestion_control);
- strncpy(name, ca->name, TCP_CA_NAME_MAX);
+ strscpy(name, ca->name, TCP_CA_NAME_MAX);
rcu_read_unlock();
}
--
2.34.1
^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
2024-07-14 4:11 [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy() Kees Cook
@ 2024-07-15 9:41 ` Simon Horman
2024-07-16 11:31 ` Paolo Abeni
2024-07-16 15:00 ` patchwork-bot+netdevbpf
1 sibling, 1 reply; 5+ messages in thread
From: Simon Horman @ 2024-07-15 9:41 UTC (permalink / raw)
To: Kees Cook
Cc: Eric Dumazet, David S. Miller, David Ahern, Jakub Kicinski,
Paolo Abeni, netdev, linux-kernel, linux-hardening
On Sat, Jul 13, 2024 at 09:11:15PM -0700, Kees Cook wrote:
> Replace the deprecated[1] uses of strncpy() in tcp_ca_get_name_by_key()
> and tcp_get_default_congestion_control(). The callers use the results as
> standard C strings (via nla_put_string() and proc handlers respectively),
> so trailing padding is not needed.
>
> Since passing the destination buffer arguments decays it to a pointer,
> the size can't be trivially determined by the compiler. ca->name is
> the same length in both cases, so strscpy() won't fail (when ca->name
> is NUL-terminated). Include the length explicitly instead of using the
> 2-argument strscpy().
>
> Link: https://github.com/KSPP/linux/issues/90 [1]
> Signed-off-by: Kees Cook <kees@kernel.org>
nit: Looking at git history, the subject prefix should probably be 'tcp'.
And it would be best to explicitly target the patch against net-next.
Subject: [PATCH net-next v2] tcp: ...
That notwithstanding, this looks good to me.
Reviewed-by: Simon Horman <horms@kernel.org>
...
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
2024-07-15 9:41 ` Simon Horman
@ 2024-07-16 11:31 ` Paolo Abeni
2024-07-16 14:39 ` Eric Dumazet
0 siblings, 1 reply; 5+ messages in thread
From: Paolo Abeni @ 2024-07-16 11:31 UTC (permalink / raw)
To: Simon Horman, Kees Cook, Eric Dumazet
Cc: David S. Miller, David Ahern, Jakub Kicinski, netdev,
linux-kernel, linux-hardening
On 7/15/24 11:41, Simon Horman wrote:
> On Sat, Jul 13, 2024 at 09:11:15PM -0700, Kees Cook wrote:
>> Replace the deprecated[1] uses of strncpy() in tcp_ca_get_name_by_key()
>> and tcp_get_default_congestion_control(). The callers use the results as
>> standard C strings (via nla_put_string() and proc handlers respectively),
>> so trailing padding is not needed.
>>
>> Since passing the destination buffer arguments decays it to a pointer,
>> the size can't be trivially determined by the compiler. ca->name is
>> the same length in both cases, so strscpy() won't fail (when ca->name
>> is NUL-terminated). Include the length explicitly instead of using the
>> 2-argument strscpy().
>>
>> Link: https://github.com/KSPP/linux/issues/90 [1]
>> Signed-off-by: Kees Cook <kees@kernel.org>
>
> nit: Looking at git history, the subject prefix should probably be 'tcp'.
> And it would be best to explicitly target the patch against net-next.
>
> Subject: [PATCH net-next v2] tcp: ...
>
> That notwithstanding, this looks good to me.
>
> Reviewed-by: Simon Horman <horms@kernel.org>
@Eric: I can fix the prefix when applying the patch. Please LMK if you
prefer otherwise.
Thanks,
Paolo
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
2024-07-16 11:31 ` Paolo Abeni
@ 2024-07-16 14:39 ` Eric Dumazet
0 siblings, 0 replies; 5+ messages in thread
From: Eric Dumazet @ 2024-07-16 14:39 UTC (permalink / raw)
To: Paolo Abeni
Cc: Simon Horman, Kees Cook, David S. Miller, David Ahern,
Jakub Kicinski, netdev, linux-kernel, linux-hardening
On Tue, Jul 16, 2024 at 4:32 AM Paolo Abeni <pabeni@redhat.com> wrote:
>
> On 7/15/24 11:41, Simon Horman wrote:
> > On Sat, Jul 13, 2024 at 09:11:15PM -0700, Kees Cook wrote:
> >> Replace the deprecated[1] uses of strncpy() in tcp_ca_get_name_by_key()
> >> and tcp_get_default_congestion_control(). The callers use the results as
> >> standard C strings (via nla_put_string() and proc handlers respectively),
> >> so trailing padding is not needed.
> >>
> >> Since passing the destination buffer arguments decays it to a pointer,
> >> the size can't be trivially determined by the compiler. ca->name is
> >> the same length in both cases, so strscpy() won't fail (when ca->name
> >> is NUL-terminated). Include the length explicitly instead of using the
> >> 2-argument strscpy().
> >>
> >> Link: https://github.com/KSPP/linux/issues/90 [1]
> >> Signed-off-by: Kees Cook <kees@kernel.org>
> >
> > nit: Looking at git history, the subject prefix should probably be 'tcp'.
> > And it would be best to explicitly target the patch against net-next.
> >
> > Subject: [PATCH net-next v2] tcp: ...
> >
> > That notwithstanding, this looks good to me.
> >
> > Reviewed-by: Simon Horman <horms@kernel.org>
>
> @Eric: I can fix the prefix when applying the patch. Please LMK if you
> prefer otherwise.
Sure thing, thanks for taking care of this Paolo, Simon, and Kees.
Reviewed-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
2024-07-14 4:11 [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy() Kees Cook
2024-07-15 9:41 ` Simon Horman
@ 2024-07-16 15:00 ` patchwork-bot+netdevbpf
1 sibling, 0 replies; 5+ messages in thread
From: patchwork-bot+netdevbpf @ 2024-07-16 15:00 UTC (permalink / raw)
To: Kees Cook
Cc: edumazet, davem, dsahern, kuba, pabeni, netdev, linux-kernel,
linux-hardening
Hello:
This patch was applied to netdev/net-next.git (main)
by Jakub Kicinski <kuba@kernel.org>:
On Sat, 13 Jul 2024 21:11:15 -0700 you wrote:
> Replace the deprecated[1] uses of strncpy() in tcp_ca_get_name_by_key()
> and tcp_get_default_congestion_control(). The callers use the results as
> standard C strings (via nla_put_string() and proc handlers respectively),
> so trailing padding is not needed.
>
> Since passing the destination buffer arguments decays it to a pointer,
> the size can't be trivially determined by the compiler. ca->name is
> the same length in both cases, so strscpy() won't fail (when ca->name
> is NUL-terminated). Include the length explicitly instead of using the
> 2-argument strscpy().
>
> [...]
Here is the summary with links:
- [v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy()
https://git.kernel.org/netdev/net-next/c/a3bfc095060b
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2024-07-16 15:00 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-07-14 4:11 [PATCH v2] net/ipv4/tcp_cong: Replace strncpy() with strscpy() Kees Cook
2024-07-15 9:41 ` Simon Horman
2024-07-16 11:31 ` Paolo Abeni
2024-07-16 14:39 ` Eric Dumazet
2024-07-16 15:00 ` patchwork-bot+netdevbpf
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).