Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1424870 > unrolled thread

[PATCH] rxrpc: fix uninitialized variable use

Back to article view | Back to linux.kernel

Contents

  [PATCH] rxrpc: fix uninitialized variable use Arnd Bergmann <arnd@arndb.de> - 2016-06-17 12:00 +0200
    Re: [PATCH] rxrpc: fix uninitialized variable use David Howells <dhowells@redhat.com> - 2016-06-21 11:00 +0200
      Re: [PATCH] rxrpc: fix uninitialized variable use Arnd Bergmann <arnd@arndb.de> - 2016-06-21 11:30 +0200
        Re: [PATCH] rxrpc: fix uninitialized variable use David Howells <dhowells@redhat.com> - 2016-06-21 12:10 +0200

#1424870 — [PATCH] rxrpc: fix uninitialized variable use

Hashing the peer key was introduced for AF_INET, but gcc
warns about the rxrpc_peer_hash_key function returning uninitialized
data for any other value of srx->transport.family:

net/rxrpc/peer_object.c: In function 'rxrpc_peer_hash_key':
net/rxrpc/peer_object.c:57:15: error: 'p' may be used uninitialized in this function [-Werror=maybe-uninitialized]

Assuming that nothing else can be set here, this changes the
function to just return zero in case of an unknown address
family.

Fixes: be6e6707f6ee ("rxrpc: Rework peer object handling to use hash table and RCU")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
This showed up today in linux-next, no idea if my patch is the
right solution for the problem, so please review carefully.

 net/rxrpc/peer_object.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/net/rxrpc/peer_object.c b/net/rxrpc/peer_object.c
index faf222c21698..5ab89295c36c 100644
--- a/net/rxrpc/peer_object.c
+++ b/net/rxrpc/peer_object.c
@@ -50,6 +50,8 @@ static unsigned long rxrpc_peer_hash_key(struct rxrpc_local *local,
 		size = sizeof(srx->transport.sin.sin_addr);
 		p = (u16 *)&srx->transport.sin.sin_addr;
 		break;
+	default:
+		return 0;
 	}
 
 	/* Step through the peer address in 16-bit portions for speed */
-- 
2.9.0

[toc] | [next] | [standalone]

#1427483

Arnd Bergmann <arnd@arndb.de> wrote:

> Hashing the peer key was introduced for AF_INET, but gcc
> warns about the rxrpc_peer_hash_key function returning uninitialized
> data for any other value of srx->transport.family:
> 
> net/rxrpc/peer_object.c: In function 'rxrpc_peer_hash_key':
> net/rxrpc/peer_object.c:57:15: error: 'p' may be used uninitialized in this function [-Werror=maybe-uninitialized]
> 
> Assuming that nothing else can be set here, this changes the
> function to just return zero in case of an unknown address
> family.

I'm actually more tempted to put a BUG() in there because if any new family
support (say AF_INET6) is added, I want to make sure I catch all the places.

David

[toc] | [prev] | [next] | [standalone]

#1427501

On Tuesday, June 21, 2016 9:48:52 AM CEST David Howells wrote:
> Arnd Bergmann <arnd@arndb.de> wrote:
> 
> > Hashing the peer key was introduced for AF_INET, but gcc
> > warns about the rxrpc_peer_hash_key function returning uninitialized
> > data for any other value of srx->transport.family:
> > 
> > net/rxrpc/peer_object.c: In function 'rxrpc_peer_hash_key':
> > net/rxrpc/peer_object.c:57:15: error: 'p' may be used uninitialized in this function [-Werror=maybe-uninitialized]
> > 
> > Assuming that nothing else can be set here, this changes the
> > function to just return zero in case of an unknown address
> > family.
> 
> I'm actually more tempted to put a BUG() in there because if any new family
> support (say AF_INET6) is added, I want to make sure I catch all the places.

Makes sense. Do you want to do the patch yourself, or should I send
a new one doing that?

Maybe WARN() would be better than BUG()? That would still get the attention
it needs but not kill the process.

	Arnd

[toc] | [prev] | [next] | [standalone]

#1427553

Arnd Bergmann <arnd@arndb.de> wrote:

> > I'm actually more tempted to put a BUG() in there because if any new family
> > support (say AF_INET6) is added, I want to make sure I catch all the places.
> 
> Makes sense. Do you want to do the patch yourself, or should I send
> a new one doing that?
> 
> Maybe WARN() would be better than BUG()? That would still get the attention
> it needs but not kill the process.

I can stick a WARN() into your patch.

David

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web