New Coverity defect in BPF verifier

New Coverity defect in BPF verifier

[Stephen Hemminger]

Date: Mon, 22 May 2017 04:40:24 -0700
From: scan-admin@coverity.com
To: stephen@networkplumber.org
Subject: New Defects reported by Coverity Scan for Linux


Hi,

Please find the latest report on new defect(s) introduced to Linux found with Coverity Scan.

1 new defect(s) introduced to Linux found with Coverity Scan.
17 defect(s), reported by Coverity Scan earlier, were marked fixed in the recent build analyzed by Coverity Scan.

New defect(s) Reported-by: Coverity Scan
Showing 1 of 1 defect(s)


** CID 1409762:  Control flow issues  (DEADCODE)
/kernel/bpf/verifier.c: 815 in check_pkt_ptr_alignment()


________________________________________________________________________________________________________
*** CID 1409762:  Control flow issues  (DEADCODE)
/kernel/bpf/verifier.c: 815 in check_pkt_ptr_alignment()
809     	}
810     
811     	/* skb->data is NET_IP_ALIGN-ed, but for strict alignment checking
812     	 * we force this to 2 which is universally what architectures use
813     	 * when they don't set CONFIG_HAVE_EFFICIENT_UNALIGNED_ACCESS.
814     	 */
>>>     CID 1409762:  Control flow issues  (DEADCODE)
>>>     Execution cannot reach the expression "0" inside this statement: "ip_align = (strict ? 2 : 0);".  
815     	ip_align = strict ? 2 : NET_IP_ALIGN;
816     	if ((ip_align + reg_off + off) % size != 0) {
817     		verbose("misaligned packet access off %d+%d+%d size %d\n",
818     			ip_align, reg_off, off, size);
819     		return -EACCES;
820     	}

Re: New Coverity defect in BPF verifier

[David Miller]

Stephen, if you had even lightly scanned netdev today, you can see
that we are discussing this already and there are at least two
patches proposed.

Thank you.

Re: New Coverity defect in BPF verifier

[Stephen Hemminger]

On Mon, 22 May 2017 12:16:04 -0400 (EDT)
David Miller <davem@davemloft.net> wrote:

> Stephen, if you had even lightly scanned netdev today, you can see
> that we are discussing this already and there are at least two
> patches proposed.
> 
> Thank you.

Sorry hadn't gotten that far. Doing full scan of 100's of messages
a day is not always possible.