Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Lukasz Odzioba <lukasz.odzioba@intel.com>
To:     herbert@gondor.apana.org.au, linux-kernel@vger.kernel.org
Cc:     lukasz.odzioba@intel.com
Subject: RE: [2/4] crypto: af_alg - Allow arbitrarily long algorithm names" email-alg_bind.txt
Date:   Wed,  8 Nov 2017 17:51:36 +0100
Message-Id: <1510159896-26362-1-git-send-email-lukasz.odzioba@intel.com>
In-Reply-To: <E1cw2aP-00085S-Q4@gondobar>
References: <E1cw2aP-00085S-Q4@gondobar>
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

Hi,
I found this patch by accident and it got my attention.
I think we can't make this name arbitrarily long because bind syscall checks addrlen before feeding protocol with it.
Current limit on my machine is 128 bytes and I can't even reach alg_bind() function if I specify more than that.
We may want to revert that.

net/socket.c:
SYSCALL_DEFINE3(bind, int, fd, struct sockaddr __user *, umyaddr, int, addrlen)
{
	struct socket *sock;
	struct sockaddr_storage address;
	int err, fput_needed;

	sock = sockfd_lookup_light(fd, &err, &fput_needed);
	if (sock) {
		err = move_addr_to_kernel(umyaddr, addrlen, (struct sockaddr *)&address);
(...snip...)

int move_addr_to_kernel(void __user *uaddr, int ulen, struct sockaddr_storage *kaddr)
{
	if (ulen < 0 || ulen > sizeof(struct sockaddr_storage))
		return -EINVAL;
		
Thanks,
Lukas


From 1587900718862404546@xxx Wed Dec 27 02:02:12 +0000 2017
X-GM-THRID: 1585184082520044658
X-Gmail-Labels: Inbox,Category Forums