CVE-2024-26586 sync from PR4690
Amit Cohen (1): mlxsw: spectrum_acl_erp: Fix error flow of pool allocation failure
Christophe JAILLET (1): mlxsw: spectrum: Use 'bitmap_zalloc()' when applicable
Ido Schimmel (4): mlxsw: spectrum_acl_tcam: Add missing mutex_destroy() mlxsw: spectrum_acl_tcam: Make fini symmetric to init mlxsw: spectrum_acl_tcam: Reorder functions to avoid forward declarations mlxsw: spectrum_acl_tcam: Fix stack corruption
.../mellanox/mlxsw/spectrum_acl_atcam.c | 8 +- .../mellanox/mlxsw/spectrum_acl_erp.c | 8 +- .../mellanox/mlxsw/spectrum_acl_tcam.c | 131 +++++++++--------- .../ethernet/mellanox/mlxsw/spectrum_cnt.c | 9 +- .../mellanox/mlxsw/spectrum_switchdev.c | 11 +- .../drivers/net/mlxsw/spectrum-2/tc_flower.sh | 106 +++++++++++++- 6 files changed, 187 insertions(+), 86 deletions(-)
反馈： 您发送到kernel@openeuler.org的补丁/补丁集，已成功转换为PR！ PR链接地址： https://gitee.com/openeuler/kernel/pulls/5109 邮件列表地址：https://mailweb.openeuler.org/hyperkitty/list/kernel@openeuler.org/message/W...
FeedBack: The patch(es) which you have sent to kernel@openeuler.org mailing list has been converted to a pull request successfully! Pull request link: https://gitee.com/openeuler/kernel/pulls/5109 Mailing list address: https://mailweb.openeuler.org/hyperkitty/list/kernel@openeuler.org/message/W...
From: Amit Cohen amcohen@nvidia.com
stable inclusion from stable-v5.10.209 commit e8bfdf6301425de9717827f52b7aa19791e7ce2c category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 6d6eeabcfaba2fcadf5443b575789ea606f9de83 ]
Lately, a bug was found when many TC filters are added - at some point, several bugs are printed to dmesg [1] and the switch is crashed with segmentation fault.
The issue starts when gen_pool_free() fails because of unexpected behavior - a try to free memory which is already freed, this leads to BUG() call which crashes the switch and makes many other bugs.
Trying to track down the unexpected behavior led to a bug in eRP code. The function mlxsw_sp_acl_erp_table_alloc() gets a pointer to the allocated index, sets the value and returns an error code. When gen_pool_alloc() fails it returns address 0, we track it and return -ENOBUFS outside, BUT the call for gen_pool_alloc() already override the index in erp_table structure. This is a problem when such allocation is done as part of table expansion. This is not a new table, which will not be used in case of allocation failure. We try to expand eRP table and override the current index (non-zero) with zero. Then, it leads to an unexpected behavior when address 0 is freed twice. Note that address 0 is valid in erp_table->base_index and indeed other tables use it.
gen_pool_alloc() fails in case that there is no space left in the pre-allocated pool, in our case, the pool is limited to ACL_MAX_ERPT_BANK_SIZE, which is read from hardware. When more than max erp entries are required, we exceed the limit and return an error, this error leads to "Failed to migrate vregion" print.
Fix this by changing erp_table->base_index only in case of a successful allocation.
Add a test case for such a scenario. Without this fix it causes segmentation fault:
$ TESTS="max_erp_entries_test" ./tc_flower.sh ./tc_flower.sh: line 988: 1560 Segmentation fault tc filter del dev $h2 ingress chain $i protocol ip pref $i handle $j flower &>/dev/null
[1]: kernel BUG at lib/genalloc.c:508! invalid opcode: 0000 [#1] PREEMPT SMP CPU: 6 PID: 3531 Comm: tc Not tainted 6.7.0-rc5-custom-ga6893f479f5e #1 Hardware name: Mellanox Technologies Ltd. MSN4700/VMOD0010, BIOS 5.11 07/12/2021 RIP: 0010:gen_pool_free_owner+0xc9/0xe0 ... Call Trace: <TASK> __mlxsw_sp_acl_erp_table_other_dec+0x70/0xa0 [mlxsw_spectrum] mlxsw_sp_acl_erp_mask_destroy+0xf5/0x110 [mlxsw_spectrum] objagg_obj_root_destroy+0x18/0x80 [objagg] objagg_obj_destroy+0x12c/0x130 [objagg] mlxsw_sp_acl_erp_mask_put+0x37/0x50 [mlxsw_spectrum] mlxsw_sp_acl_ctcam_region_entry_remove+0x74/0xa0 [mlxsw_spectrum] mlxsw_sp_acl_ctcam_entry_del+0x1e/0x40 [mlxsw_spectrum] mlxsw_sp_acl_tcam_ventry_del+0x78/0xd0 [mlxsw_spectrum] mlxsw_sp_flower_destroy+0x4d/0x70 [mlxsw_spectrum] mlxsw_sp_flow_block_cb+0x73/0xb0 [mlxsw_spectrum] tc_setup_cb_destroy+0xc1/0x180 fl_hw_destroy_filter+0x94/0xc0 [cls_flower] __fl_delete+0x1ac/0x1c0 [cls_flower] fl_destroy+0xc2/0x150 [cls_flower] tcf_proto_destroy+0x1a/0xa0 ... mlxsw_spectrum3 0000:07:00.0: Failed to migrate vregion mlxsw_spectrum3 0000:07:00.0: Failed to migrate vregion
Fixes: f465261aa105 ("mlxsw: spectrum_acl: Implement common eRP core") Signed-off-by: Amit Cohen amcohen@nvidia.com Signed-off-by: Ido Schimmel idosch@nvidia.com Signed-off-by: Petr Machata petrm@nvidia.com Acked-by: Paolo Abeni pabeni@redhat.com Link: https://lore.kernel.org/r/4cfca254dfc0e5d283974801a24371c7b6db5989.170550206... Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- .../mellanox/mlxsw/spectrum_acl_erp.c | 8 +-- .../drivers/net/mlxsw/spectrum-2/tc_flower.sh | 52 ++++++++++++++++++- 2 files changed, 56 insertions(+), 4 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c index 4c98950380d5..d231f4d2888b 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_erp.c @@ -301,6 +301,7 @@ mlxsw_sp_acl_erp_table_alloc(struct mlxsw_sp_acl_erp_core *erp_core, unsigned long *p_index) { unsigned int num_rows, entry_size; + unsigned long index;
/* We only allow allocations of entire rows */ if (num_erps % erp_core->num_erp_banks != 0) @@ -309,10 +310,11 @@ mlxsw_sp_acl_erp_table_alloc(struct mlxsw_sp_acl_erp_core *erp_core, entry_size = erp_core->erpt_entries_size[region_type]; num_rows = num_erps / erp_core->num_erp_banks;
- *p_index = gen_pool_alloc(erp_core->erp_tables, num_rows * entry_size); - if (*p_index == 0) + index = gen_pool_alloc(erp_core->erp_tables, num_rows * entry_size); + if (!index) return -ENOBUFS; - *p_index -= MLXSW_SP_ACL_ERP_GENALLOC_OFFSET; + + *p_index = index - MLXSW_SP_ACL_ERP_GENALLOC_OFFSET;
return 0; } diff --git a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh index fb850e0ec837..7bf56ea161e3 100755 --- a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh +++ b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh @@ -10,7 +10,8 @@ lib_dir=$(dirname $0)/../../../../net/forwarding ALL_TESTS="single_mask_test identical_filters_test two_masks_test \ multiple_masks_test ctcam_edge_cases_test delta_simple_test \ delta_two_masks_one_key_test delta_simple_rehash_test \ - bloom_simple_test bloom_complex_test bloom_delta_test" + bloom_simple_test bloom_complex_test bloom_delta_test \ + max_erp_entries_test" NUM_NETIFS=2 source $lib_dir/lib.sh source $lib_dir/tc_common.sh @@ -983,6 +984,55 @@ bloom_delta_test() log_test "bloom delta test ($tcflags)" }
+max_erp_entries_test() +{ + # The number of eRP entries is limited. Once the maximum number of eRPs + # has been reached, filters cannot be added. This test verifies that + # when this limit is reached, inserstion fails without crashing. + + RET=0 + + local num_masks=32 + local num_regions=15 + local chain_failed + local mask_failed + local ret + + if [[ "$tcflags" != "skip_sw" ]]; then + return 0; + fi + + for ((i=1; i < $num_regions; i++)); do + for ((j=$num_masks; j >= 0; j--)); do + tc filter add dev $h2 ingress chain $i protocol ip \ + pref $i handle $j flower $tcflags \ + dst_ip 192.1.0.0/$j &> /dev/null + ret=$? + + if [ $ret -ne 0 ]; then + chain_failed=$i + mask_failed=$j + break 2 + fi + done + done + + # We expect to exceed the maximum number of eRP entries, so that + # insertion eventually fails. Otherwise, the test should be adjusted to + # add more filters. + check_fail $ret "expected to exceed number of eRP entries" + + for ((; i >= 1; i--)); do + for ((j=0; j <= $num_masks; j++)); do + tc filter del dev $h2 ingress chain $i protocol ip \ + pref $i handle $j flower &> /dev/null + done + done + + log_test "max eRP entries test ($tcflags). " \ + "max chain $chain_failed, mask $mask_failed" +} + setup_prepare() { h1=${NETIFS[p1]}
From: Christophe JAILLET christophe.jaillet@wanadoo.fr
stable inclusion from stable-v5.10.209 commit 936d06d699c325dfeb034ee5222f2c88f551fd06 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 2c087dfcc9d5e7e8557d217f01f58ba42d1ddbf1 ]
Use 'bitmap_zalloc()' to simplify code, improve the semantic and avoid some open-coded arithmetic in allocator arguments.
Also change the corresponding 'kfree()' into 'bitmap_free()' to keep consistency.
Signed-off-by: Christophe JAILLET christophe.jaillet@wanadoo.fr Reviewed-by: Ido Schimmel idosch@nvidia.com Signed-off-by: David S. Miller davem@davemloft.net Stable-dep-of: 483ae90d8f97 ("mlxsw: spectrum_acl_tcam: Fix stack corruption") Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- .../ethernet/mellanox/mlxsw/spectrum_acl_atcam.c | 8 +++----- .../ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 15 ++++++--------- .../net/ethernet/mellanox/mlxsw/spectrum_cnt.c | 9 +++------ .../ethernet/mellanox/mlxsw/spectrum_switchdev.c | 11 ++++------- 4 files changed, 16 insertions(+), 27 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c index ded4cf658680..4b713832fdd5 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_atcam.c @@ -119,7 +119,6 @@ mlxsw_sp_acl_atcam_region_12kb_init(struct mlxsw_sp_acl_atcam_region *aregion) { struct mlxsw_sp *mlxsw_sp = aregion->region->mlxsw_sp; struct mlxsw_sp_acl_atcam_region_12kb *region_12kb; - size_t alloc_size; u64 max_lkey_id; int err;
@@ -131,8 +130,7 @@ mlxsw_sp_acl_atcam_region_12kb_init(struct mlxsw_sp_acl_atcam_region *aregion) if (!region_12kb) return -ENOMEM;
- alloc_size = BITS_TO_LONGS(max_lkey_id) * sizeof(unsigned long); - region_12kb->used_lkey_id = kzalloc(alloc_size, GFP_KERNEL); + region_12kb->used_lkey_id = bitmap_zalloc(max_lkey_id, GFP_KERNEL); if (!region_12kb->used_lkey_id) { err = -ENOMEM; goto err_used_lkey_id_alloc; @@ -149,7 +147,7 @@ mlxsw_sp_acl_atcam_region_12kb_init(struct mlxsw_sp_acl_atcam_region *aregion) return 0;
err_rhashtable_init: - kfree(region_12kb->used_lkey_id); + bitmap_free(region_12kb->used_lkey_id); err_used_lkey_id_alloc: kfree(region_12kb); return err; @@ -161,7 +159,7 @@ mlxsw_sp_acl_atcam_region_12kb_fini(struct mlxsw_sp_acl_atcam_region *aregion) struct mlxsw_sp_acl_atcam_region_12kb *region_12kb = aregion->priv;
rhashtable_destroy(®ion_12kb->lkey_ht); - kfree(region_12kb->used_lkey_id); + bitmap_free(region_12kb->used_lkey_id); kfree(region_12kb); }
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index ddf7e03c6b43..f624c849526a 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -36,7 +36,6 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, u64 max_tcam_regions; u64 max_regions; u64 max_groups; - size_t alloc_size; int err;
mutex_init(&tcam->lock); @@ -52,15 +51,13 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, if (max_tcam_regions < max_regions) max_regions = max_tcam_regions;
- alloc_size = sizeof(tcam->used_regions[0]) * BITS_TO_LONGS(max_regions); - tcam->used_regions = kzalloc(alloc_size, GFP_KERNEL); + tcam->used_regions = bitmap_zalloc(max_regions, GFP_KERNEL); if (!tcam->used_regions) return -ENOMEM; tcam->max_regions = max_regions;
max_groups = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUPS); - alloc_size = sizeof(tcam->used_groups[0]) * BITS_TO_LONGS(max_groups); - tcam->used_groups = kzalloc(alloc_size, GFP_KERNEL); + tcam->used_groups = bitmap_zalloc(max_groups, GFP_KERNEL); if (!tcam->used_groups) { err = -ENOMEM; goto err_alloc_used_groups; @@ -76,9 +73,9 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, return 0;
err_tcam_init: - kfree(tcam->used_groups); + bitmap_free(tcam->used_groups); err_alloc_used_groups: - kfree(tcam->used_regions); + bitmap_free(tcam->used_regions); return err; }
@@ -89,8 +86,8 @@ void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp,
mutex_destroy(&tcam->lock); ops->fini(mlxsw_sp, tcam->priv); - kfree(tcam->used_groups); - kfree(tcam->used_regions); + bitmap_free(tcam->used_groups); + bitmap_free(tcam->used_regions); }
int mlxsw_sp_acl_tcam_priority_get(struct mlxsw_sp *mlxsw_sp, diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c index b65b93a2b9bc..fc2257753b9b 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_cnt.c @@ -122,7 +122,6 @@ int mlxsw_sp_counter_pool_init(struct mlxsw_sp *mlxsw_sp) unsigned int sub_pools_count = ARRAY_SIZE(mlxsw_sp_counter_sub_pools); struct devlink *devlink = priv_to_devlink(mlxsw_sp->core); struct mlxsw_sp_counter_pool *pool; - unsigned int map_size; int err;
pool = kzalloc(struct_size(pool, sub_pools, sub_pools_count), @@ -143,9 +142,7 @@ int mlxsw_sp_counter_pool_init(struct mlxsw_sp *mlxsw_sp) devlink_resource_occ_get_register(devlink, MLXSW_SP_RESOURCE_COUNTERS, mlxsw_sp_counter_pool_occ_get, pool);
- map_size = BITS_TO_LONGS(pool->pool_size) * sizeof(unsigned long); - - pool->usage = kzalloc(map_size, GFP_KERNEL); + pool->usage = bitmap_zalloc(pool->pool_size, GFP_KERNEL); if (!pool->usage) { err = -ENOMEM; goto err_usage_alloc; @@ -158,7 +155,7 @@ int mlxsw_sp_counter_pool_init(struct mlxsw_sp *mlxsw_sp) return 0;
err_sub_pools_init: - kfree(pool->usage); + bitmap_free(pool->usage); err_usage_alloc: devlink_resource_occ_get_unregister(devlink, MLXSW_SP_RESOURCE_COUNTERS); @@ -176,7 +173,7 @@ void mlxsw_sp_counter_pool_fini(struct mlxsw_sp *mlxsw_sp) WARN_ON(find_first_bit(pool->usage, pool->pool_size) != pool->pool_size); WARN_ON(atomic_read(&pool->active_entries_count)); - kfree(pool->usage); + bitmap_free(pool->usage); devlink_resource_occ_get_unregister(devlink, MLXSW_SP_RESOURCE_COUNTERS); kfree(pool); diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c index 368fa0e5ad31..ea37f5000caa 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c @@ -1631,16 +1631,13 @@ mlxsw_sp_mid *__mlxsw_sp_mc_alloc(struct mlxsw_sp *mlxsw_sp, u16 fid) { struct mlxsw_sp_mid *mid; - size_t alloc_size;
mid = kzalloc(sizeof(*mid), GFP_KERNEL); if (!mid) return NULL;
- alloc_size = sizeof(unsigned long) * - BITS_TO_LONGS(mlxsw_core_max_ports(mlxsw_sp->core)); - - mid->ports_in_mid = kzalloc(alloc_size, GFP_KERNEL); + mid->ports_in_mid = bitmap_zalloc(mlxsw_core_max_ports(mlxsw_sp->core), + GFP_KERNEL); if (!mid->ports_in_mid) goto err_ports_in_mid_alloc;
@@ -1659,7 +1656,7 @@ mlxsw_sp_mid *__mlxsw_sp_mc_alloc(struct mlxsw_sp *mlxsw_sp, return mid;
err_write_mdb_entry: - kfree(mid->ports_in_mid); + bitmap_free(mid->ports_in_mid); err_ports_in_mid_alloc: kfree(mid); return NULL; @@ -1676,7 +1673,7 @@ static int mlxsw_sp_port_remove_from_mid(struct mlxsw_sp_port *mlxsw_sp_port, mlxsw_core_max_ports(mlxsw_sp->core))) { err = mlxsw_sp_mc_remove_mdb_entry(mlxsw_sp, mid); list_del(&mid->list); - kfree(mid->ports_in_mid); + bitmap_free(mid->ports_in_mid); kfree(mid); } return err;
From: Ido Schimmel idosch@nvidia.com
stable inclusion from stable-v5.10.209 commit a557dbda6407b3818a0f7b681fc7dcf7540dd0d9 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 65823e07b1e4055b6278725fd92f4d7e6f8d53fd ]
Pair mutex_init() with a mutex_destroy() in the error path. Found during code review. No functional changes.
Signed-off-by: Ido Schimmel idosch@nvidia.com Reviewed-by: Jiri Pirko jiri@nvidia.com Signed-off-by: Petr Machata petrm@nvidia.com Reviewed-by: Jacob Keller jacob.e.keller@intel.com Signed-off-by: Jakub Kicinski kuba@kernel.org Stable-dep-of: 483ae90d8f97 ("mlxsw: spectrum_acl_tcam: Fix stack corruption") Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index f624c849526a..4469488c2d7e 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -52,8 +52,10 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, max_regions = max_tcam_regions;
tcam->used_regions = bitmap_zalloc(max_regions, GFP_KERNEL); - if (!tcam->used_regions) - return -ENOMEM; + if (!tcam->used_regions) { + err = -ENOMEM; + goto err_alloc_used_regions; + } tcam->max_regions = max_regions;
max_groups = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUPS); @@ -76,6 +78,8 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, bitmap_free(tcam->used_groups); err_alloc_used_groups: bitmap_free(tcam->used_regions); +err_alloc_used_regions: + mutex_destroy(&tcam->lock); return err; }
From: Ido Schimmel idosch@nvidia.com
stable inclusion from stable-v5.10.209 commit d3669ebc4c2a7ede3d9901ba5bee8b806eea98de category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 61fe3b9102ac84ba479ab84d8f5454af2e21e468 ]
Move mutex_destroy() to the end to make the function symmetric with mlxsw_sp_acl_tcam_init(). No functional changes.
Signed-off-by: Ido Schimmel idosch@nvidia.com Reviewed-by: Jiri Pirko jiri@nvidia.com Signed-off-by: Petr Machata petrm@nvidia.com Reviewed-by: Jacob Keller jacob.e.keller@intel.com Signed-off-by: Jakub Kicinski kuba@kernel.org Stable-dep-of: 483ae90d8f97 ("mlxsw: spectrum_acl_tcam: Fix stack corruption") Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 4469488c2d7e..042c85c7a96d 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -88,10 +88,10 @@ void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, { const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops;
- mutex_destroy(&tcam->lock); ops->fini(mlxsw_sp, tcam->priv); bitmap_free(tcam->used_groups); bitmap_free(tcam->used_regions); + mutex_destroy(&tcam->lock); }
int mlxsw_sp_acl_tcam_priority_get(struct mlxsw_sp *mlxsw_sp,
From: Ido Schimmel idosch@nvidia.com
stable inclusion from stable-v5.10.209 commit a9b233b287aa44be00e7278b7c38d098bc0f1604 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 194ab9476089bbfc021073214e071a404e375ee6 ]
Move the initialization and de-initialization code further below in order to avoid forward declarations in the next patch. No functional changes.
Signed-off-by: Ido Schimmel idosch@nvidia.com Reviewed-by: Jiri Pirko jiri@nvidia.com Signed-off-by: Petr Machata petrm@nvidia.com Reviewed-by: Jacob Keller jacob.e.keller@intel.com Signed-off-by: Jakub Kicinski kuba@kernel.org Stable-dep-of: 483ae90d8f97 ("mlxsw: spectrum_acl_tcam: Fix stack corruption") Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- .../mellanox/mlxsw/spectrum_acl_tcam.c | 130 +++++++++--------- 1 file changed, 65 insertions(+), 65 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 042c85c7a96d..01b542f11fa7 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -29,71 +29,6 @@ size_t mlxsw_sp_acl_tcam_priv_size(struct mlxsw_sp *mlxsw_sp) #define MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_MIN 3000 /* ms */ #define MLXSW_SP_ACL_TCAM_VREGION_REHASH_CREDITS 100 /* number of entries */
-int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_tcam *tcam) -{ - const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; - u64 max_tcam_regions; - u64 max_regions; - u64 max_groups; - int err; - - mutex_init(&tcam->lock); - tcam->vregion_rehash_intrvl = - MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_DFLT; - INIT_LIST_HEAD(&tcam->vregion_list); - - max_tcam_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, - ACL_MAX_TCAM_REGIONS); - max_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_REGIONS); - - /* Use 1:1 mapping between ACL region and TCAM region */ - if (max_tcam_regions < max_regions) - max_regions = max_tcam_regions; - - tcam->used_regions = bitmap_zalloc(max_regions, GFP_KERNEL); - if (!tcam->used_regions) { - err = -ENOMEM; - goto err_alloc_used_regions; - } - tcam->max_regions = max_regions; - - max_groups = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUPS); - tcam->used_groups = bitmap_zalloc(max_groups, GFP_KERNEL); - if (!tcam->used_groups) { - err = -ENOMEM; - goto err_alloc_used_groups; - } - tcam->max_groups = max_groups; - tcam->max_group_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, - ACL_MAX_GROUP_SIZE); - - err = ops->init(mlxsw_sp, tcam->priv, tcam); - if (err) - goto err_tcam_init; - - return 0; - -err_tcam_init: - bitmap_free(tcam->used_groups); -err_alloc_used_groups: - bitmap_free(tcam->used_regions); -err_alloc_used_regions: - mutex_destroy(&tcam->lock); - return err; -} - -void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, - struct mlxsw_sp_acl_tcam *tcam) -{ - const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; - - ops->fini(mlxsw_sp, tcam->priv); - bitmap_free(tcam->used_groups); - bitmap_free(tcam->used_regions); - mutex_destroy(&tcam->lock); -} - int mlxsw_sp_acl_tcam_priority_get(struct mlxsw_sp *mlxsw_sp, struct mlxsw_sp_acl_rule_info *rulei, u32 *priority, bool fillup_priority) @@ -1546,6 +1481,71 @@ mlxsw_sp_acl_tcam_vregion_rehash(struct mlxsw_sp *mlxsw_sp, mlxsw_sp_acl_tcam_vregion_rehash_end(mlxsw_sp, vregion, ctx); }
+int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + u64 max_tcam_regions; + u64 max_regions; + u64 max_groups; + int err; + + mutex_init(&tcam->lock); + tcam->vregion_rehash_intrvl = + MLXSW_SP_ACL_TCAM_VREGION_REHASH_INTRVL_DFLT; + INIT_LIST_HEAD(&tcam->vregion_list); + + max_tcam_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_TCAM_REGIONS); + max_regions = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_REGIONS); + + /* Use 1:1 mapping between ACL region and TCAM region */ + if (max_tcam_regions < max_regions) + max_regions = max_tcam_regions; + + tcam->used_regions = bitmap_zalloc(max_regions, GFP_KERNEL); + if (!tcam->used_regions) { + err = -ENOMEM; + goto err_alloc_used_regions; + } + tcam->max_regions = max_regions; + + max_groups = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUPS); + tcam->used_groups = bitmap_zalloc(max_groups, GFP_KERNEL); + if (!tcam->used_groups) { + err = -ENOMEM; + goto err_alloc_used_groups; + } + tcam->max_groups = max_groups; + tcam->max_group_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, + ACL_MAX_GROUP_SIZE); + + err = ops->init(mlxsw_sp, tcam->priv, tcam); + if (err) + goto err_tcam_init; + + return 0; + +err_tcam_init: + bitmap_free(tcam->used_groups); +err_alloc_used_groups: + bitmap_free(tcam->used_regions); +err_alloc_used_regions: + mutex_destroy(&tcam->lock); + return err; +} + +void mlxsw_sp_acl_tcam_fini(struct mlxsw_sp *mlxsw_sp, + struct mlxsw_sp_acl_tcam *tcam) +{ + const struct mlxsw_sp_acl_tcam_ops *ops = mlxsw_sp->acl_tcam_ops; + + ops->fini(mlxsw_sp, tcam->priv); + bitmap_free(tcam->used_groups); + bitmap_free(tcam->used_regions); + mutex_destroy(&tcam->lock); +} + static const enum mlxsw_afk_element mlxsw_sp_acl_tcam_pattern_ipv4[] = { MLXSW_AFK_ELEMENT_SRC_SYS_PORT, MLXSW_AFK_ELEMENT_DMAC_32_47,
From: Ido Schimmel idosch@nvidia.com
stable inclusion from stable-v5.10.209 commit 56750ea5d15426b5f307554e7699e8b5f76c3182 category: bugfix bugzilla: https://gitee.com/src-openeuler/kernel/issues/I932WD CVE: CVE-2024-26586
Reference: https://git.kernel.org/pub/scm/linux/kernel/git/stable/linux.git/commit/?id=...
---------------------------
[ Upstream commit 483ae90d8f976f8339cf81066312e1329f2d3706 ]
When tc filters are first added to a net device, the corresponding local port gets bound to an ACL group in the device. The group contains a list of ACLs. In turn, each ACL points to a different TCAM region where the filters are stored. During forwarding, the ACLs are sequentially evaluated until a match is found.
One reason to place filters in different regions is when they are added with decreasing priorities and in an alternating order so that two consecutive filters can never fit in the same region because of their key usage.
In Spectrum-2 and newer ASICs the firmware started to report that the maximum number of ACLs in a group is more than 16, but the layout of the register that configures ACL groups (PAGT) was not updated to account for that. It is therefore possible to hit stack corruption [1] in the rare case where more than 16 ACLs in a group are required.
Fix by limiting the maximum ACL group size to the minimum between what the firmware reports and the maximum ACLs that fit in the PAGT register.
Add a test case to make sure the machine does not crash when this condition is hit.
[1] Kernel panic - not syncing: stack-protector: Kernel stack is corrupted in: mlxsw_sp_acl_tcam_group_update+0x116/0x120 [...] dump_stack_lvl+0x36/0x50 panic+0x305/0x330 __stack_chk_fail+0x15/0x20 mlxsw_sp_acl_tcam_group_update+0x116/0x120 mlxsw_sp_acl_tcam_group_region_attach+0x69/0x110 mlxsw_sp_acl_tcam_vchunk_get+0x492/0xa20 mlxsw_sp_acl_tcam_ventry_add+0x25/0xe0 mlxsw_sp_acl_rule_add+0x47/0x240 mlxsw_sp_flower_replace+0x1a9/0x1d0 tc_setup_cb_add+0xdc/0x1c0 fl_hw_replace_filter+0x146/0x1f0 fl_change+0xc17/0x1360 tc_new_tfilter+0x472/0xb90 rtnetlink_rcv_msg+0x313/0x3b0 netlink_rcv_skb+0x58/0x100 netlink_unicast+0x244/0x390 netlink_sendmsg+0x1e4/0x440 ____sys_sendmsg+0x164/0x260 ___sys_sendmsg+0x9a/0xe0 __sys_sendmsg+0x7a/0xc0 do_syscall_64+0x40/0xe0 entry_SYSCALL_64_after_hwframe+0x63/0x6b
Fixes: c3ab435466d5 ("mlxsw: spectrum: Extend to support Spectrum-2 ASIC") Reported-by: Orel Hagag orelh@nvidia.com Signed-off-by: Ido Schimmel idosch@nvidia.com Reviewed-by: Amit Cohen amcohen@nvidia.com Signed-off-by: Petr Machata petrm@nvidia.com Acked-by: Paolo Abeni pabeni@redhat.com Link: https://lore.kernel.org/r/2d91c89afba59c22587b444994ae419dbea8d876.170550206... Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Sasha Levin sashal@kernel.org Signed-off-by: Liu Jian liujian56@huawei.com --- .../mellanox/mlxsw/spectrum_acl_tcam.c | 2 + .../drivers/net/mlxsw/spectrum-2/tc_flower.sh | 56 ++++++++++++++++++- 2 files changed, 57 insertions(+), 1 deletion(-)
diff --git a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c index 01b542f11fa7..d3f99c6b3cae 100644 --- a/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c +++ b/drivers/net/ethernet/mellanox/mlxsw/spectrum_acl_tcam.c @@ -1519,6 +1519,8 @@ int mlxsw_sp_acl_tcam_init(struct mlxsw_sp *mlxsw_sp, tcam->max_groups = max_groups; tcam->max_group_size = MLXSW_CORE_RES_GET(mlxsw_sp->core, ACL_MAX_GROUP_SIZE); + tcam->max_group_size = min_t(unsigned int, tcam->max_group_size, + MLXSW_REG_PAGT_ACL_MAX_NUM);
err = ops->init(mlxsw_sp, tcam->priv, tcam); if (err) diff --git a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh index 7bf56ea161e3..616d3581419c 100755 --- a/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh +++ b/tools/testing/selftests/drivers/net/mlxsw/spectrum-2/tc_flower.sh @@ -11,7 +11,7 @@ ALL_TESTS="single_mask_test identical_filters_test two_masks_test \ multiple_masks_test ctcam_edge_cases_test delta_simple_test \ delta_two_masks_one_key_test delta_simple_rehash_test \ bloom_simple_test bloom_complex_test bloom_delta_test \ - max_erp_entries_test" + max_erp_entries_test max_group_size_test" NUM_NETIFS=2 source $lib_dir/lib.sh source $lib_dir/tc_common.sh @@ -1033,6 +1033,60 @@ max_erp_entries_test() "max chain $chain_failed, mask $mask_failed" }
+max_group_size_test() +{ + # The number of ACLs in an ACL group is limited. Once the maximum + # number of ACLs has been reached, filters cannot be added. This test + # verifies that when this limit is reached, insertion fails without + # crashing. + + RET=0 + + local num_acls=32 + local max_size + local ret + + if [[ "$tcflags" != "skip_sw" ]]; then + return 0; + fi + + for ((i=1; i < $num_acls; i++)); do + if [[ $(( i % 2 )) == 1 ]]; then + tc filter add dev $h2 ingress pref $i proto ipv4 \ + flower $tcflags dst_ip 198.51.100.1/32 \ + ip_proto tcp tcp_flags 0x01/0x01 \ + action drop &> /dev/null + else + tc filter add dev $h2 ingress pref $i proto ipv6 \ + flower $tcflags dst_ip 2001:db8:1::1/128 \ + action drop &> /dev/null + fi + + ret=$? + [[ $ret -ne 0 ]] && max_size=$((i - 1)) && break + done + + # We expect to exceed the maximum number of ACLs in a group, so that + # insertion eventually fails. Otherwise, the test should be adjusted to + # add more filters. + check_fail $ret "expected to exceed number of ACLs in a group" + + for ((; i >= 1; i--)); do + if [[ $(( i % 2 )) == 1 ]]; then + tc filter del dev $h2 ingress pref $i proto ipv4 \ + flower $tcflags dst_ip 198.51.100.1/32 \ + ip_proto tcp tcp_flags 0x01/0x01 \ + action drop &> /dev/null + else + tc filter del dev $h2 ingress pref $i proto ipv6 \ + flower $tcflags dst_ip 2001:db8:1::1/128 \ + action drop &> /dev/null + fi + done + + log_test "max ACL group size test ($tcflags). max size $max_size" +} + setup_prepare() { h1=${NETIFS[p1]}
-
Liu Jian -
patchwork bot