netfilter: nf_set_pipapo: fix initial map fill
Subsystems (2)
Florian Westphal
maintainerPATCH 1/1 netfilter: nf_set_pipapo: fix initial map fill
The initial buffer has to be inited to all-ones, but it must restrict it to the size of the first field, not the total field size.
After each round in the map search step, the result and the fill map are swapped, so if we have a set where f->bsize of the first element is smaller than m->bsize_max, those one-bits are leaked into future rounds result map.
This makes pipapo find an incorrect matching results for sets where first field size is not the largest.
Followup patch adds a test case to nft_concat_range.sh selftest script.
Thanks to Stefano Brivio for pointing out that we need to zero out the remainder explicitly, only correcting memset() argument isn't enough.
--- net/netfilter/nft_set_pipapo.c | 4 ++-- net/netfilter/nft_set_pipapo.h | 21 +++++++++++++++++++++ net/netfilter/nft_set_pipapo_avx2.c | 10 ++++++---- 3 files changed, 29 insertions(+), 6 deletions(-)
net/netfilter/nft_set_pipapo.c +2 -2
@@ -434,7 +434,7 @@ bool nft_pipapo_lookup(const struct net *net, const struct nft_set *set, res_map = scratch->map + (map_index ? m->bsize_max :0 ); fill_map = scratch->map + (map_index ?0 : m->bsize_max); -memset (res_map,0xff , m->bsize_max *sizeof (*res_map)); +pipapo_resmap_init (m, res_map);nft_pipapo_for_each_field (f, i, m) {bool last = i == m->field_count -1 ;
@@ -542,7 +542,7 @@ static struct nft_pipapo_elem *pipapo_get(const struct net *net, goto out; } -memset (res_map,0xff , m->bsize_max *sizeof (*res_map)); +pipapo_resmap_init (m, res_map);nft_pipapo_for_each_field (f, i, m) {bool last = i == m->field_count -1 ;
net/netfilter/nft_set_pipapo.h +21 -0
@@ -278,4 +278,25 @@ static u64 pipapo_estimate_size(const struct nft_set_desc *desc) return size; } +/** +* pipapo_resmap_init() - Initialise result map before first use +* @m: Matching data, including mapping table +* @res_map: Result map +* +* Initialize all bits covered by the first field to one, so that after +* the first step, only the matching bits of the first bit group remain. +* +* If other fields have a large bitmap, set remainder of res_map to 0. +*/ +static inline void pipapo_resmap_init (const struct nft_pipapo_match *m,unsigned long *res_map) +{ +const struct nft_pipapo_field *f = m->f; +int i; + +for (i =0 ; i < f->bsize; i++) + res_map[i] =ULONG_MAX ; + +for (i = f->bsize; i < m->bsize_max; i++) + res_map[i] =0ul ; +}#endif /* _NFT_SET_PIPAPO_H */
net/netfilter/nft_set_pipapo_avx2.c +6 -4
@@ -1036,6 +1036,7 @@ static int nft_pipapo_avx2_lookup_8b_16(unsigned long *map, unsigned long *fill,/** * nft_pipapo_avx2_lookup_slow() - Fallback function for uncommon field sizes +* @mdata: Matching data, including mapping table * @map: Previous match result, used as initial bitmap * @fill: Destination bitmap to be filled with current match result * @f: Field, containing lookup and mapping tables
@@ -1051,7 +1052,8 @@ static int nft_pipapo_avx2_lookup_8b_16(unsigned long *map, unsigned long *fill,* Return: -1 on no match, rule index of match if @last, otherwise first long * word index to be checked next (i.e. first filled word). */ -static int nft_pipapo_avx2_lookup_slow (unsigned long *map,unsigned long *fill, +static int nft_pipapo_avx2_lookup_slow (const struct nft_pipapo_match *mdata, +unsigned long *map,unsigned long *fill,const struct nft_pipapo_field *f,int offset,const u8 *pkt,bool first,bool last)
@@ -1060,7 +1062,7 @@ static int nft_pipapo_avx2_lookup_slow(unsigned long *map, unsigned long *fill,int i, ret =-1 , b;if (first) -memset (map,0xff , bsize *sizeof (*map)); +pipapo_resmap_init (mdata, map);for (i = offset; i < bsize; i++) {if (f->bb ==8 )
@@ -1186,7 +1188,7 @@ bool nft_pipapo_avx2_lookup(const struct net *net, const struct nft_set *set, }else if (f->groups ==16 ) {NFT_SET_PIPAPO_AVX2_LOOKUP (8 ,16 ); }else { - ret =nft_pipapo_avx2_lookup_slow (res, fill, f, + ret =nft_pipapo_avx2_lookup_slow (m, res, fill, f, ret, rp, first, last); }
@@ -1202,7 +1204,7 @@ bool nft_pipapo_avx2_lookup(const struct net *net, const struct nft_set *set, }else if (f->groups ==32 ) {NFT_SET_PIPAPO_AVX2_LOOKUP (4 ,32 ); }else { - ret =nft_pipapo_avx2_lookup_slow (res, fill, f, + ret =nft_pipapo_avx2_lookup_slow (m, res, fill, f, ret, rp, first, last); }
Signature · 1 line
2.44.2
Quoted 8 lines · On Mon, 15 Jul 2024 13:54:03 +0200 Florian Westphal <[email protected]> wrote:
The initial buffer has to be inited to all-ones, but it must restrict it to the size of the first field, not the total field size.
After each round in the map search step, the result and the fill map are swapped, so if we have a set where
f->bsizeof the first element is smaller thanm->bsize_max, those one-bits are leaked into future rounds result map.
Quoted 7 lines
This makes pipapo find an incorrect matching results for sets where first field size is not the largest.
Followup patch adds a test case to
nft_concat_range.shselftest script.Thanks to Stefano Brivio for pointing out that we need to zero out the remainder explicitly, only correcting
memset()argument isn't enough.
Thanks for fixing this!
Quoted 4 lines
Fixes 3c4287f62044 ("nf_tables: Add set type for arbitrary concatenation of ranges")Reported-by Yi Chen <[email protected]>Cc Stefano Brivio <[email protected]>Signed-off-by Florian Westphal <[email protected]>
Signature · 1 line
Stefano