All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH net 0/2] Couple of skb memory leak fixes
@ 2022-01-17  9:27 Gal Pressman
  2022-01-17  9:27 ` [PATCH net 1/2] net/tls: Fix another skb memory leak when running kTLS traffic Gal Pressman
                   ` (2 more replies)
  0 siblings, 3 replies; 6+ messages in thread
From: Gal Pressman @ 2022-01-17  9:27 UTC (permalink / raw)
  To: Eric Dumazet, Jakub Kicinski, David S. Miller
  Cc: netdev, Tariq Toukan, Saeed Mahameed, Gal Pressman

As discussed in:
https://lore.kernel.org/netdev/20220102081253.9123-1-gal@nvidia.com/

These are the two followup suggestions from Eric and Jakub.
Patch #1 adds a sk_defer_free_flush() call to the kTLS splice_read
handler.
Patch #2 verifies the defer list is empty on socket destroy, and calls a
defer free flush as well.

Thanks!

Gal Pressman (2):
  net/tls: Fix another skb memory leak when running kTLS traffic
  net: Flush deferred skb free on socket destroy

 net/core/sock.c  | 3 +++
 net/tls/tls_sw.c | 1 +
 2 files changed, 4 insertions(+)

-- 
2.25.1


^ permalink raw reply	[flat|nested] 6+ messages in thread
* [PATCH net 1/2] net/tls: Fix another skb memory leak when running kTLS traffic
  2022-01-17  9:27 [PATCH net 0/2] Couple of skb memory leak fixes Gal Pressman
@ 2022-01-17  9:27 ` Gal Pressman
  2022-01-17  9:27 ` [PATCH net 2/2] net: Flush deferred skb free on socket destroy Gal Pressman
  2022-01-17 13:10 ` [PATCH net 0/2] Couple of skb memory leak fixes patchwork-bot+netdevbpf
  2 siblings, 0 replies; 6+ messages in thread
From: Gal Pressman @ 2022-01-17  9:27 UTC (permalink / raw)
  To: Eric Dumazet, Jakub Kicinski, David S. Miller
  Cc: netdev, Tariq Toukan, Saeed Mahameed, Gal Pressman

This patch is a followup to
commit ffef737fd037 ("net/tls: Fix skb memory leak when running kTLS traffic")

Which was missing another sk_defer_free_flush() call in
tls_sw_splice_read().

Fixes: f35f821935d8 ("tcp: defer skb freeing after socket lock is released")
Signed-off-by: Gal Pressman <gal@nvidia.com>
---
 net/tls/tls_sw.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index 95e774f1b91f..efc84845bb6b 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -2059,6 +2059,7 @@ ssize_t tls_sw_splice_read(struct socket *sock,  loff_t *ppos,
 
 splice_read_end:
 	release_sock(sk);
+	sk_defer_free_flush(sk);
 	return copied ? : err;
 }
 
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 6+ messages in thread
* [PATCH net 2/2] net: Flush deferred skb free on socket destroy
  2022-01-17  9:27 [PATCH net 0/2] Couple of skb memory leak fixes Gal Pressman
  2022-01-17  9:27 ` [PATCH net 1/2] net/tls: Fix another skb memory leak when running kTLS traffic Gal Pressman
@ 2022-01-17  9:27 ` Gal Pressman
  2022-01-23  9:50   ` Heiko Carstens
  2022-01-17 13:10 ` [PATCH net 0/2] Couple of skb memory leak fixes patchwork-bot+netdevbpf
  2 siblings, 1 reply; 6+ messages in thread
From: Gal Pressman @ 2022-01-17  9:27 UTC (permalink / raw)
  To: Eric Dumazet, Jakub Kicinski, David S. Miller
  Cc: netdev, Tariq Toukan, Saeed Mahameed, Gal Pressman

The cited Fixes patch moved to a deferred skb approach where the skbs
are not freed immediately under the socket lock.  Add a WARN_ON_ONCE()
to verify the deferred list is empty on socket destroy, and empty it to
prevent potential memory leaks.

Fixes: f35f821935d8 ("tcp: defer skb freeing after socket lock is released")
Signed-off-by: Gal Pressman <gal@nvidia.com>
---
 net/core/sock.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/core/sock.c b/net/core/sock.c
index f32ec08a0c37..4ff806d71921 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2049,6 +2049,9 @@ void sk_destruct(struct sock *sk)
 {
 	bool use_call_rcu = sock_flag(sk, SOCK_RCU_FREE);
 
+	WARN_ON_ONCE(!llist_empty(&sk->defer_list));
+	sk_defer_free_flush(sk);
+
 	if (rcu_access_pointer(sk->sk_reuseport_cb)) {
 		reuseport_detach_sock(sk);
 		use_call_rcu = true;
-- 
2.25.1


^ permalink raw reply related	[flat|nested] 6+ messages in thread
* Re: [PATCH net 0/2] Couple of skb memory leak fixes
  2022-01-17  9:27 [PATCH net 0/2] Couple of skb memory leak fixes Gal Pressman
  2022-01-17  9:27 ` [PATCH net 1/2] net/tls: Fix another skb memory leak when running kTLS traffic Gal Pressman
  2022-01-17  9:27 ` [PATCH net 2/2] net: Flush deferred skb free on socket destroy Gal Pressman
@ 2022-01-17 13:10 ` patchwork-bot+netdevbpf
  2 siblings, 0 replies; 6+ messages in thread
From: patchwork-bot+netdevbpf @ 2022-01-17 13:10 UTC (permalink / raw)
  To: Gal Pressman; +Cc: edumazet, kuba, davem, netdev, tariqt, saeedm

Hello:

This series was applied to netdev/net.git (master)
by David S. Miller <davem@davemloft.net>:

On Mon, 17 Jan 2022 11:27:31 +0200 you wrote:
> As discussed in:
> https://lore.kernel.org/netdev/20220102081253.9123-1-gal@nvidia.com/
> 
> These are the two followup suggestions from Eric and Jakub.
> Patch #1 adds a sk_defer_free_flush() call to the kTLS splice_read
> handler.
> Patch #2 verifies the defer list is empty on socket destroy, and calls a
> defer free flush as well.
> 
> [...]

Here is the summary with links:
  - [net,1/2] net/tls: Fix another skb memory leak when running kTLS traffic
    https://git.kernel.org/netdev/net/c/db094aa8140e
  - [net,2/2] net: Flush deferred skb free on socket destroy
    https://git.kernel.org/netdev/net/c/79074a72d335

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] 6+ messages in thread
* Re: [PATCH net 2/2] net: Flush deferred skb free on socket destroy
  2022-01-17  9:27 ` [PATCH net 2/2] net: Flush deferred skb free on socket destroy Gal Pressman
@ 2022-01-23  9:50   ` Heiko Carstens
  2022-01-23  9:53     ` Gal Pressman
  0 siblings, 1 reply; 6+ messages in thread
From: Heiko Carstens @ 2022-01-23  9:50 UTC (permalink / raw)
  To: Gal Pressman
  Cc: Eric Dumazet, Jakub Kicinski, David S. Miller, netdev,
	Tariq Toukan, Saeed Mahameed

On Mon, Jan 17, 2022 at 11:27:33AM +0200, Gal Pressman wrote:
> The cited Fixes patch moved to a deferred skb approach where the skbs
> are not freed immediately under the socket lock.  Add a WARN_ON_ONCE()
> to verify the deferred list is empty on socket destroy, and empty it to
> prevent potential memory leaks.
> 
> Fixes: f35f821935d8 ("tcp: defer skb freeing after socket lock is released")
> Signed-off-by: Gal Pressman <gal@nvidia.com>
> ---
>  net/core/sock.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/net/core/sock.c b/net/core/sock.c
> index f32ec08a0c37..4ff806d71921 100644
> --- a/net/core/sock.c
> +++ b/net/core/sock.c
> @@ -2049,6 +2049,9 @@ void sk_destruct(struct sock *sk)
>  {
>  	bool use_call_rcu = sock_flag(sk, SOCK_RCU_FREE);
>  
> +	WARN_ON_ONCE(!llist_empty(&sk->defer_list));
> +	sk_defer_free_flush(sk);
> +

This leads to a link error if CONFIG_INET is not set:

s390x-11.2.0-ld: net/core/sock.o: in function `sk_defer_free_flush':
linux/./include/net/tcp.h:1378: undefined reference to `__sk_defer_free_flush'
make: *** [Makefile:1155: vmlinux] Error 1

^ permalink raw reply	[flat|nested] 6+ messages in thread
* Re: [PATCH net 2/2] net: Flush deferred skb free on socket destroy
  2022-01-23  9:50   ` Heiko Carstens
@ 2022-01-23  9:53     ` Gal Pressman
  0 siblings, 0 replies; 6+ messages in thread
From: Gal Pressman @ 2022-01-23  9:53 UTC (permalink / raw)
  To: Heiko Carstens
  Cc: Eric Dumazet, Jakub Kicinski, David S. Miller, netdev,
	Tariq Toukan, Saeed Mahameed

On 23/01/2022 11:50, Heiko Carstens wrote:
> On Mon, Jan 17, 2022 at 11:27:33AM +0200, Gal Pressman wrote:
>> The cited Fixes patch moved to a deferred skb approach where the skbs
>> are not freed immediately under the socket lock.  Add a WARN_ON_ONCE()
>> to verify the deferred list is empty on socket destroy, and empty it to
>> prevent potential memory leaks.
>>
>> Fixes: f35f821935d8 ("tcp: defer skb freeing after socket lock is released")
>> Signed-off-by: Gal Pressman <gal@nvidia.com>
>> ---
>>  net/core/sock.c | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/net/core/sock.c b/net/core/sock.c
>> index f32ec08a0c37..4ff806d71921 100644
>> --- a/net/core/sock.c
>> +++ b/net/core/sock.c
>> @@ -2049,6 +2049,9 @@ void sk_destruct(struct sock *sk)
>>  {
>>  	bool use_call_rcu = sock_flag(sk, SOCK_RCU_FREE);
>>  
>> +	WARN_ON_ONCE(!llist_empty(&sk->defer_list));
>> +	sk_defer_free_flush(sk);
>> +
> This leads to a link error if CONFIG_INET is not set:
>
> s390x-11.2.0-ld: net/core/sock.o: in function `sk_defer_free_flush':
> linux/./include/net/tcp.h:1378: undefined reference to `__sk_defer_free_flush'
> make: *** [Makefile:1155: vmlinux] Error 1

Thanks, it is fixed in:

https://lore.kernel.org/netdev/20220120123440.9088-1-gal@nvidia.com/


^ permalink raw reply	[flat|nested] 6+ messages in thread
end of thread, other threads:[~2022-01-23  9:54 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-01-17  9:27 [PATCH net 0/2] Couple of skb memory leak fixes Gal Pressman
2022-01-17  9:27 ` [PATCH net 1/2] net/tls: Fix another skb memory leak when running kTLS traffic Gal Pressman
2022-01-17  9:27 ` [PATCH net 2/2] net: Flush deferred skb free on socket destroy Gal Pressman
2022-01-23  9:50   ` Heiko Carstens
2022-01-23  9:53     ` Gal Pressman
2022-01-17 13:10 ` [PATCH net 0/2] Couple of skb memory leak fixes patchwork-bot+netdevbpf

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.