Skip to contentSkip to conversation

net/unix: use consistent error code in SO_PEERPIDFD

PATCH Applied David Rheinsberg 6 participants 6 messages Aug 7, 2023 – Aug 8, 2023

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net/unix: use consistent error code in SO_PEERPIDFD

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

This changes SO_PEERPIDFD to return ENODATA instead. This is also what SO_PEERGROUPS returns, and thus keeps a consistent behavior across sockopts.

Note that this code is returned in 2 cases: First, if the socket type is not AF_UNIX, and secondly if the socket was not yet connected. In both cases ENODATA seems suitable.

---
Hi!

The SO_PEERPIDFD sockopt has been queued for 6.5, so hopefully we can get that in before the release?

Thanks
David

 net/core/sock.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/sock.c +1 -1
index 6d4f28efe29a..732fc37a4771 100644
@@ -1778,7 +1778,7 @@ int sk_getsockopt(struct sock *sk, int level, int optname,
 		spin_unlock(&sk->sk_peer_lock);
 
 		if (!peer_pid)
-			return -ESRCH;
+			return -ENODATA;
 
 		pidfd = pidfd_prepare(peer_pid, 0, &pidfd_file);
 		put_pid(peer_pid);
Signature · 1 line
2.41.0
#1 lore raw
Quoted 16 lines · David Rheinsberg

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

This changes SO_PEERPIDFD to return ENODATA instead. This is also what SO_PEERGROUPS returns, and thus keeps a consistent behavior across sockopts.

Quoted 5 lines

Note that this code is returned in 2 cases: First, if the socket type is not AF_UNIX, and secondly if the socket was not yet connected. In both cases ENODATA seems suitable.

Hi David!

I generally agree with this. But just to be more precise, l2cap sockets (AF_BLUETOOTH) also properly
set ->sk_peer_pid and SO_PEERPIDFD will work just fine. This thing is not limited
only to AF_UNIX sockets.

Kind regards,
Alex

Quoted 28 lines

---
Hi!

The SO_PEERPIDFD sockopt has been queued for 6.5, so hopefully we can get that in before the release?

Thanks
David

 net/core/sock.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/sock.c +1 -2
index 6d4f28efe29a..732fc37a4771 100644
@@ -1778,7 +1778,7 @@ int sk_getsockopt(struct sock *sk, int level, int optname,
                spin_unlock(&sk->sk_peer_lock);
 
                if (!peer_pid)
-                       return -ESRCH;
+                       return -ENODATA;
 
                pidfd = pidfd_prepare(peer_pid, 0, &pidfd_file);
                put_pid(peer_pid);
--

2.41.0

Reviewed-by on v1 1/1
#2 lore raw
In reply to David Rheinsberg
Quoted 19 lines · David Rheinsberg

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

This changes SO_PEERPIDFD to return ENODATA instead. This is also what SO_PEERGROUPS returns, and thus keeps a consistent behavior across sockopts.

Note that this code is returned in 2 cases: First, if the socket type is not AF_UNIX, and secondly if the socket was not yet connected. In both cases ENODATA seems suitable.

Quoted 6 lines

---
Hi!

The SO_PEERPIDFD sockopt has been queued for 6.5, so hopefully we can get that in before the release?

Shouldn't be an issue afaict.

Looks good to me,

Acked-by on v1 1/1
#3 lore raw
In reply to David Rheinsberg
Quoted 24 lines · David Rheinsberg

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

This changes SO_PEERPIDFD to return ENODATA instead. This is also what SO_PEERGROUPS returns, and thus keeps a consistent behavior across sockopts.

Note that this code is returned in 2 cases: First, if the socket type is not AF_UNIX, and secondly if the socket was not yet connected. In both cases ENODATA seems suitable.

---
Hi!

Quoted 8 lines

The SO_PEERPIDFD sockopt has been queued for 6.5, so hopefully we can get that in before the release?

Thanks
David

 net/core/sock.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

As one of the first users of this new API, I am fine with this for 6.5, thanks.

reviewer
#4 lore raw
In reply to David Rheinsberg
Quoted 20 lines · David Rheinsberg

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

This changes SO_PEERPIDFD to return ENODATA instead. This is also what SO_PEERGROUPS returns, and thus keeps a consistent behavior across sockopts.

Note that this code is returned in 2 cases: First, if the socket type is not AF_UNIX, and secondly if the socket was not yet connected. In both cases ENODATA seems suitable.

Quoted 8 lines

---
Hi!

The SO_PEERPIDFD sockopt has been queued for 6.5, so hopefully we can get that in before the release?

Thanks
David

As a fix, it should probably have a fixes tag.
This one seems appropriate.

Fixes 7b26952a91cf ("net: core: add getsockopt SO_PEERPIDFD")

And the patch should be targeted at net

Subject: [PATCH net] ...

It's probably not necessary to repost just to address these minor points. But please consider them if you need to post a v2 for some other reason.

#5 lore raw
In reply to David Rheinsberg
This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:

Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:

Quoted 4 lines · On Mon, 7 Aug 2023 10:12:25 +0200 you wrote:

Change the new (unreleased) SO_PEERPIDFD sockopt to return ENODATA rather than ESRCH if a socket type does not support remote peer-PID queries.

Quoted 8 lines

Currently, SO_PEERPIDFD returns ESRCH when the socket in question is not an AF_UNIX socket. This is quite unexpected, given that one would assume ESRCH means the peer process already exited and thus cannot be found. However, in that case the sockopt actually returns EINVAL (via pidfd_prepare()). This is rather inconsistent with other syscalls, which usually return ESRCH if a given PID refers to a non-existant process.

[...]

Here is the summary with links:
- net/unix: use consistent error code in SO_PEERPIDFD

    https://git.kernel.org/netdev/net/c/b6f79e826fbd

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html