Re: genetlink misinterprets NEW as GET

On Thursday 2011-01-06 15:55, Pablo Neira Ayuso wrote:

On 06/01/11 15:25, Jan Engelhardt wrote:

quoted

On Thursday 2011-01-06 14:48, Pablo Neira Ayuso wrote:

quoted

quoted

	/* Modifiers to GET request */
	#define NLM_F_ROOT      0x100
	#define NLM_F_MATCH     0x200
	#define NLM_F_ATOMIC    0x400
	#define NLM_F_DUMP      (NLM_F_ROOT|NLM_F_MATCH)
	
	/* Modifiers to NEW request */
	#define NLM_F_REPLACE   0x100
	#define NLM_F_EXCL      0x200
	#define NLM_F_CREATE    0x400
	#define NLM_F_APPEND    0x800

i getting confused, so ipset is also setting NLM_F_REPLACE to match the
NLM_F_DUMP bitmask?

Any userspace program sending a (ge)netlink message with
NLM_F_CREATE|NLM_F_EXCL -- with the intent of creating an entry with
excl semantics --, will be misunderstood by genetlink.c to be a dump
request.

The problem is of general nature and not limited to ipset. I only
noticed it while making the ipset-genl patch, because ipset sends all
IPSET_CMD_CREATE requests with
NLM_F_REQUEST|NLM_F_ACK|NLM_F_CREATE|NLM_F_EXCL (see
ipset/lib/mnl.c).