* [PATCH net] net/tcp_sigpool: Fix some off by one bugs
@ 2023-10-31 9:51 Dan Carpenter
2023-10-31 16:58 ` Dmitry Safonov
` (2 more replies)
0 siblings, 3 replies; 4+ messages in thread
From: Dan Carpenter @ 2023-10-31 9:51 UTC (permalink / raw)
To: Dmitry Safonov
Cc: Eric Dumazet, David S. Miller, David Ahern, Jakub Kicinski,
Paolo Abeni, Steen Hegelund, netdev, kernel-janitors
The "cpool_populated" variable is the number of elements in the cpool[]
array that have been populated. It is incremented in
tcp_sigpool_alloc_ahash() every time we populate a new element.
Unpopulated elements are NULL but if we have populated every element then
this code will read one element beyond the end of the array.
Fixes: 8c73b26315aa ("net/tcp: Prepare tcp_md5sig_pool for TCP-AO")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
---
From static analysis and review.
net/ipv4/tcp_sigpool.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/tcp_sigpool.c b/net/ipv4/tcp_sigpool.c
index 65a8eaae2fec..55b310a722c7 100644
--- a/net/ipv4/tcp_sigpool.c
+++ b/net/ipv4/tcp_sigpool.c
@@ -231,7 +231,7 @@ static void cpool_schedule_cleanup(struct kref *kref)
*/
void tcp_sigpool_release(unsigned int id)
{
- if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
+ if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
return;
/* slow-path */
@@ -245,7 +245,7 @@ EXPORT_SYMBOL_GPL(tcp_sigpool_release);
*/
void tcp_sigpool_get(unsigned int id)
{
- if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
+ if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
return;
kref_get(&cpool[id].kref);
}
@@ -256,7 +256,7 @@ int tcp_sigpool_start(unsigned int id, struct tcp_sigpool *c) __cond_acquires(RC
struct crypto_ahash *hash;
rcu_read_lock_bh();
- if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg)) {
+ if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg)) {
rcu_read_unlock_bh();
return -EINVAL;
}
@@ -301,7 +301,7 @@ EXPORT_SYMBOL_GPL(tcp_sigpool_end);
*/
size_t tcp_sigpool_algo(unsigned int id, char *buf, size_t buf_len)
{
- if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
+ if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
return -EINVAL;
return strscpy(buf, cpool[id].alg, buf_len);
--
2.42.0
^ permalink raw reply related [flat|nested] 4+ messages in thread* Re: [PATCH net] net/tcp_sigpool: Fix some off by one bugs
2023-10-31 9:51 [PATCH net] net/tcp_sigpool: Fix some off by one bugs Dan Carpenter
@ 2023-10-31 16:58 ` Dmitry Safonov
2023-11-01 4:51 ` Eric Dumazet
2023-11-02 5:51 ` patchwork-bot+netdevbpf
2 siblings, 0 replies; 4+ messages in thread
From: Dmitry Safonov @ 2023-10-31 16:58 UTC (permalink / raw)
To: Dan Carpenter
Cc: Eric Dumazet, David S. Miller, David Ahern, Jakub Kicinski,
Paolo Abeni, Steen Hegelund, netdev, kernel-janitors
Hi Dan,
On 10/31/23 09:51, Dan Carpenter wrote:
> The "cpool_populated" variable is the number of elements in the cpool[]
> array that have been populated. It is incremented in
> tcp_sigpool_alloc_ahash() every time we populate a new element.
> Unpopulated elements are NULL but if we have populated every element then
> this code will read one element beyond the end of the array.
>
> Fixes: 8c73b26315aa ("net/tcp: Prepare tcp_md5sig_pool for TCP-AO")
> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Yeah, those are barriers for any issues in the caller-code, so that's
not too critical but nice to have. Thanks for the patch!
Reviewed-by: Dmitry Safonov <dima@arista.com>
> ---
> From static analysis and review.
>
> net/ipv4/tcp_sigpool.c | 8 ++++----
> 1 file changed, 4 insertions(+), 4 deletions(-)
>
> diff --git a/net/ipv4/tcp_sigpool.c b/net/ipv4/tcp_sigpool.c
> index 65a8eaae2fec..55b310a722c7 100644
> --- a/net/ipv4/tcp_sigpool.c
> +++ b/net/ipv4/tcp_sigpool.c
> @@ -231,7 +231,7 @@ static void cpool_schedule_cleanup(struct kref *kref)
> */
> void tcp_sigpool_release(unsigned int id)
> {
> - if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
> + if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
> return;
>
> /* slow-path */
> @@ -245,7 +245,7 @@ EXPORT_SYMBOL_GPL(tcp_sigpool_release);
> */
> void tcp_sigpool_get(unsigned int id)
> {
> - if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
> + if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
> return;
> kref_get(&cpool[id].kref);
> }
> @@ -256,7 +256,7 @@ int tcp_sigpool_start(unsigned int id, struct tcp_sigpool *c) __cond_acquires(RC
> struct crypto_ahash *hash;
>
> rcu_read_lock_bh();
> - if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg)) {
> + if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg)) {
> rcu_read_unlock_bh();
> return -EINVAL;
> }
> @@ -301,7 +301,7 @@ EXPORT_SYMBOL_GPL(tcp_sigpool_end);
> */
> size_t tcp_sigpool_algo(unsigned int id, char *buf, size_t buf_len)
> {
> - if (WARN_ON_ONCE(id > cpool_populated || !cpool[id].alg))
> + if (WARN_ON_ONCE(id >= cpool_populated || !cpool[id].alg))
> return -EINVAL;
>
> return strscpy(buf, cpool[id].alg, buf_len);
Thanks,
Dmitry
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH net] net/tcp_sigpool: Fix some off by one bugs
2023-10-31 9:51 [PATCH net] net/tcp_sigpool: Fix some off by one bugs Dan Carpenter
2023-10-31 16:58 ` Dmitry Safonov
@ 2023-11-01 4:51 ` Eric Dumazet
2023-11-02 5:51 ` patchwork-bot+netdevbpf
2 siblings, 0 replies; 4+ messages in thread
From: Eric Dumazet @ 2023-11-01 4:51 UTC (permalink / raw)
To: Dan Carpenter
Cc: Dmitry Safonov, David S. Miller, David Ahern, Jakub Kicinski,
Paolo Abeni, Steen Hegelund, netdev, kernel-janitors
On Tue, Oct 31, 2023 at 10:51 AM Dan Carpenter <dan.carpenter@linaro.org> wrote:
>
> The "cpool_populated" variable is the number of elements in the cpool[]
> array that have been populated. It is incremented in
> tcp_sigpool_alloc_ahash() every time we populate a new element.
> Unpopulated elements are NULL but if we have populated every element then
> this code will read one element beyond the end of the array.
>
> Fixes: 8c73b26315aa ("net/tcp: Prepare tcp_md5sig_pool for TCP-AO")
> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
> ---
> From static analysis and review.
>
Reviewed-by: Eric Dumazet <edumazet@google.com>
^ permalink raw reply [flat|nested] 4+ messages in thread* Re: [PATCH net] net/tcp_sigpool: Fix some off by one bugs
2023-10-31 9:51 [PATCH net] net/tcp_sigpool: Fix some off by one bugs Dan Carpenter
2023-10-31 16:58 ` Dmitry Safonov
2023-11-01 4:51 ` Eric Dumazet
@ 2023-11-02 5:51 ` patchwork-bot+netdevbpf
2 siblings, 0 replies; 4+ messages in thread
From: patchwork-bot+netdevbpf @ 2023-11-02 5:51 UTC (permalink / raw)
To: Dan Carpenter
Cc: 0x7f454c46, edumazet, davem, dsahern, kuba, pabeni,
Steen.Hegelund, netdev, kernel-janitors
Hello:
This patch was applied to netdev/net.git (main)
by Jakub Kicinski <kuba@kernel.org>:
On Tue, 31 Oct 2023 12:51:09 +0300 you wrote:
> The "cpool_populated" variable is the number of elements in the cpool[]
> array that have been populated. It is incremented in
> tcp_sigpool_alloc_ahash() every time we populate a new element.
> Unpopulated elements are NULL but if we have populated every element then
> this code will read one element beyond the end of the array.
>
> Fixes: 8c73b26315aa ("net/tcp: Prepare tcp_md5sig_pool for TCP-AO")
> Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
>
> [...]
Here is the summary with links:
- [net] net/tcp_sigpool: Fix some off by one bugs
https://git.kernel.org/netdev/net/c/74da77921333
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] 4+ messages in thread
end of thread, other threads:[~2023-11-02 5:51 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-10-31 9:51 [PATCH net] net/tcp_sigpool: Fix some off by one bugs Dan Carpenter
2023-10-31 16:58 ` Dmitry Safonov
2023-11-01 4:51 ` Eric Dumazet
2023-11-02 5:51 ` 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).