Thread (10 messages) flat view 10 messages, 2 authors, 2020-08-16
STALE2182d

Revision v1 of 218 in this series.

Revisions (218)
  1. v1 [diff vs current]
  2. v1 [diff vs current]
  3. v1 [diff vs current]
  4. v1 [diff vs current]
  5. v1 [diff vs current]
  6. v1 [diff vs current]
  7. v1 [diff vs current]
  8. v1 [diff vs current]
  9. v1 [diff vs current]
  10. v1 [diff vs current]
  11. v1 [diff vs current]
  12. v1 [diff vs current]
  13. v1 [diff vs current]
  14. v1 [diff vs current]
  15. v1 [diff vs current]
  16. v1 [diff vs current]
  17. v1 [diff vs current]
  18. v1 [diff vs current]
  19. v1 [diff vs current]
  20. v1 [diff vs current]
  21. v1 [diff vs current]
  22. v1 [diff vs current]
  23. v1 [diff vs current]
  24. v1 [diff vs current]
  25. v1 [diff vs current]
  26. v1 [diff vs current]
  27. v1 [diff vs current]
  28. v1 [diff vs current]
  29. v1 [diff vs current]
  30. v1 [diff vs current]
  31. v1 [diff vs current]
  32. v1 [diff vs current]
  33. v1 [diff vs current]
  34. v1 [diff vs current]
  35. v1 [diff vs current]
  36. v1 [diff vs current]
  37. v1 [diff vs current]
  38. v1 [diff vs current]
  39. v1 [diff vs current]
  40. v1 [diff vs current]
  41. v1 [diff vs current]
  42. v1 [diff vs current]
  43. v1 [diff vs current]
  44. v1 [diff vs current]
  45. v1 [diff vs current]
  46. v1 [diff vs current]
  47. v1 [diff vs current]
  48. v1 [diff vs current]
  49. v1 [diff vs current]
  50. v1 [diff vs current]
  51. v1 [diff vs current]
  52. v1 [diff vs current]
  53. v1 [diff vs current]
  54. v1 [diff vs current]
  55. v1 [diff vs current]
  56. v1 [diff vs current]
  57. v1 [diff vs current]
  58. v1 [diff vs current]
  59. v1 [diff vs current]
  60. v1 [diff vs current]
  61. v1 [diff vs current]
  62. v1 [diff vs current]
  63. v1 [diff vs current]
  64. v1 [diff vs current]
  65. v1 [diff vs current]
  66. v1 [diff vs current]
  67. v1 [diff vs current]
  68. v1 [diff vs current]
  69. v1 [diff vs current]
  70. v1 [diff vs current]
  71. v1 [diff vs current]
  72. v1 [diff vs current]
  73. v1 [diff vs current]
  74. v1 [diff vs current]
  75. v1 [diff vs current]
  76. v1 [diff vs current]
  77. v1 [diff vs current]
  78. v1 [diff vs current]
  79. v1 [diff vs current]
  80. v1 [diff vs current]
  81. v1 [diff vs current]
  82. v1 [diff vs current]
  83. v1 [diff vs current]
  84. v1 [diff vs current]
  85. v1 [diff vs current]
  86. v1 [diff vs current]
  87. v1 [diff vs current]
  88. v1 [diff vs current]
  89. v1 [diff vs current]
  90. v1 [diff vs current]
  91. v1 [diff vs current]
  92. v1 [diff vs current]
  93. v1 [diff vs current]
  94. v1 current
  95. v1 [diff vs current]
  96. v1 [diff vs current]
  97. v1 [diff vs current]
  98. v1 [diff vs current]
  99. v1 [diff vs current]
  100. v1 [diff vs current]
  101. v1 [diff vs current]
  102. v1 [diff vs current]
  103. v1 [diff vs current]
  104. v1 [diff vs current]
  105. v1 [diff vs current]
  106. v1 [diff vs current]
  107. v1 [diff vs current]
  108. v1 [diff vs current]
  109. v1 [diff vs current]
  110. v1 [diff vs current]
  111. v1 [diff vs current]
  112. v1 [diff vs current]
  113. v1 [diff vs current]
  114. v1 [diff vs current]
  115. v2 [diff vs current]
  116. v1 [diff vs current]
  117. v1 [diff vs current]
  118. v1 [diff vs current]
  119. v1 [diff vs current]
  120. v1 [diff vs current]
  121. v1 [diff vs current]
  122. v1 [diff vs current]
  123. v1 [diff vs current]
  124. v1 [diff vs current]
  125. v1 [diff vs current]
  126. v2 [diff vs current]
  127. v1 [diff vs current]
  128. v1 [diff vs current]
  129. v1 [diff vs current]
  130. v1 [diff vs current]
  131. v1 [diff vs current]
  132. v1 [diff vs current]
  133. v1 [diff vs current]
  134. v1 [diff vs current]
  135. v1 [diff vs current]
  136. v1 [diff vs current]
  137. v1 [diff vs current]
  138. v1 [diff vs current]
  139. v1 [diff vs current]
  140. v1 [diff vs current]
  141. v1 [diff vs current]
  142. v1 [diff vs current]
  143. v1 [diff vs current]
  144. v1 [diff vs current]
  145. v1 [diff vs current]
  146. v1 [diff vs current]
  147. v1 [diff vs current]
  148. v1 [diff vs current]
  149. v1 [diff vs current]
  150. v1 [diff vs current]
  151. v1 [diff vs current]
  152. v1 [diff vs current]
  153. v1 [diff vs current]
  154. v1 [diff vs current]
  155. v1 [diff vs current]
  156. v1 [diff vs current]
  157. v1 [diff vs current]
  158. v1 [diff vs current]
  159. v1 [diff vs current]
  160. v1 [diff vs current]
  161. v1 [diff vs current]
  162. v1 [diff vs current]
  163. v1 [diff vs current]
  164. v1 [diff vs current]
  165. v1 [diff vs current]
  166. v1 [diff vs current]
  167. v1 [diff vs current]
  168. v1 [diff vs current]
  169. v1 [diff vs current]
  170. v1 [diff vs current]
  171. v1 [diff vs current]
  172. v1 [diff vs current]
  173. v2 [diff vs current]
  174. v1 [diff vs current]
  175. v1 [diff vs current]
  176. v2 [diff vs current]
  177. v1 [diff vs current]
  178. v1 [diff vs current]
  179. v1 [diff vs current]
  180. v1 [diff vs current]
  181. v2 [diff vs current]
  182. v1 [diff vs current]
  183. v1 [diff vs current]
  184. v1 [diff vs current]
  185. v1 [diff vs current]
  186. v1 [diff vs current]
  187. v3 [diff vs current]
  188. v1 [diff vs current]
  189. v1 [diff vs current]
  190. v1 [diff vs current]
  191. v1 [diff vs current]
  192. v1 [diff vs current]
  193. v1 [diff vs current]
  194. v1 [diff vs current]
  195. v1 [diff vs current]
  196. v1 [diff vs current]
  197. v2 [diff vs current]
  198. v1 [diff vs current]
  199. v1 [diff vs current]
  200. v1 [diff vs current]
  201. v1 [diff vs current]
  202. v2 [diff vs current]
  203. v1 [diff vs current]
  204. v1 [diff vs current]
  205. v1 [diff vs current]
  206. v1 [diff vs current]
  207. v1 [diff vs current]
  208. v1 [diff vs current]
  209. v1 [diff vs current]
  210. v2 [diff vs current]
  211. v1 [diff vs current]
  212. v2 [diff vs current]
  213. v1 [diff vs current]
  214. v1 [diff vs current]
  215. v1 [diff vs current]
  216. v1 [diff vs current]
  217. v2 [diff vs current]
  218. v1 [diff vs current]

[PATCH 2/8] netfilter: nft_compat: remove flush counter optimization

From: Pablo Neira Ayuso <pablo@netfilter.org>
Date: 2020-08-15 22:19:24
Also in: netfilter-devel
Subsystem: netfilter, networking [general], the rest · Maintainers: Pablo Neira Ayuso, Florian Westphal, "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Linus Torvalds

From: Florian Westphal <fw@strlen.de>

WARNING: CPU: 1 PID: 16059 at lib/refcount.c:31 refcount_warn_saturate+0xdf/0xf
[..]
 __nft_mt_tg_destroy+0x42/0x50 [nft_compat]
 nft_target_destroy+0x63/0x80 [nft_compat]
 nf_tables_expr_destroy+0x1b/0x30 [nf_tables]
 nf_tables_rule_destroy+0x3a/0x70 [nf_tables]
 nf_tables_exit_net+0x186/0x3d0 [nf_tables]

Happens when a compat expr is destoyed from abort path.
There is no functional impact; after this work queue is flushed
unconditionally if its pending.

This removes the waitcount optimization.  Test of repeated
iptables-restore of a ~60k kubernetes ruleset doesn't indicate
a slowdown.  In case the counter is needed after all for some workloads
we can revert this and increment the refcount for the
!= NFT_PREPARE_TRANS case to avoid the increment/decrement imbalance.

While at it, also flush for match case, this was an oversight
in the original patch.

Fixes: ffe8923f109b7e ("netfilter: nft_compat: make sure xtables destructors have run")
Reported-by: kernel test robot <redacted>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
---
 net/netfilter/nft_compat.c | 37 ++++++++++++++-----------------------
 1 file changed, 14 insertions(+), 23 deletions(-)
diff --git a/net/netfilter/nft_compat.c b/net/netfilter/nft_compat.c
index 6428856ccbec..8e56f353ff35 100644
--- a/net/netfilter/nft_compat.c
+++ b/net/netfilter/nft_compat.c
@@ -27,8 +27,6 @@ struct nft_xt_match_priv {
 	void *info;
 };
 
-static refcount_t nft_compat_pending_destroy = REFCOUNT_INIT(1);
-
 static int nft_compat_chain_validate_dependency(const struct nft_ctx *ctx,
 						const char *tablename)
 {
@@ -215,6 +213,17 @@ static int nft_parse_compat(const struct nlattr *attr, u16 *proto, bool *inv)
 	return 0;
 }
 
+static void nft_compat_wait_for_destructors(void)
+{
+	/* xtables matches or targets can have side effects, e.g.
+	 * creation/destruction of /proc files.
+	 * The xt ->destroy functions are run asynchronously from
+	 * work queue.  If we have pending invocations we thus
+	 * need to wait for those to finish.
+	 */
+	nf_tables_trans_destroy_flush_work();
+}
+
 static int
 nft_target_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
 		const struct nlattr * const tb[])
@@ -238,14 +247,7 @@ nft_target_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
 
 	nft_target_set_tgchk_param(&par, ctx, target, info, &e, proto, inv);
 
-	/* xtables matches or targets can have side effects, e.g.
-	 * creation/destruction of /proc files.
-	 * The xt ->destroy functions are run asynchronously from
-	 * work queue.  If we have pending invocations we thus
-	 * need to wait for those to finish.
-	 */
-	if (refcount_read(&nft_compat_pending_destroy) > 1)
-		nf_tables_trans_destroy_flush_work();
+	nft_compat_wait_for_destructors();
 
 	ret = xt_check_target(&par, size, proto, inv);
 	if (ret < 0)
@@ -260,7 +262,6 @@ nft_target_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
 
 static void __nft_mt_tg_destroy(struct module *me, const struct nft_expr *expr)
 {
-	refcount_dec(&nft_compat_pending_destroy);
 	module_put(me);
 	kfree(expr->ops);
 }
@@ -468,6 +469,8 @@ __nft_match_init(const struct nft_ctx *ctx, const struct nft_expr *expr,
 
 	nft_match_set_mtchk_param(&par, ctx, match, info, &e, proto, inv);
 
+	nft_compat_wait_for_destructors();
+
 	return xt_check_match(&par, size, proto, inv);
 }
 
@@ -716,14 +719,6 @@ static const struct nfnetlink_subsystem nfnl_compat_subsys = {
 
 static struct nft_expr_type nft_match_type;
 
-static void nft_mt_tg_deactivate(const struct nft_ctx *ctx,
-				 const struct nft_expr *expr,
-				 enum nft_trans_phase phase)
-{
-	if (phase == NFT_TRANS_COMMIT)
-		refcount_inc(&nft_compat_pending_destroy);
-}
-
 static const struct nft_expr_ops *
 nft_match_select_ops(const struct nft_ctx *ctx,
 		     const struct nlattr * const tb[])
@@ -762,7 +757,6 @@ nft_match_select_ops(const struct nft_ctx *ctx,
 	ops->type = &nft_match_type;
 	ops->eval = nft_match_eval;
 	ops->init = nft_match_init;
-	ops->deactivate = nft_mt_tg_deactivate,
 	ops->destroy = nft_match_destroy;
 	ops->dump = nft_match_dump;
 	ops->validate = nft_match_validate;
@@ -853,7 +847,6 @@ nft_target_select_ops(const struct nft_ctx *ctx,
 	ops->size = NFT_EXPR_SIZE(XT_ALIGN(target->targetsize));
 	ops->init = nft_target_init;
 	ops->destroy = nft_target_destroy;
-	ops->deactivate = nft_mt_tg_deactivate,
 	ops->dump = nft_target_dump;
 	ops->validate = nft_target_validate;
 	ops->data = target;
@@ -917,8 +910,6 @@ static void __exit nft_compat_module_exit(void)
 	nfnetlink_subsys_unregister(&nfnl_compat_subsys);
 	nft_unregister_expr(&nft_target_type);
 	nft_unregister_expr(&nft_match_type);
-
-	WARN_ON_ONCE(refcount_read(&nft_compat_pending_destroy) != 1);
 }
 
 MODULE_ALIAS_NFNL_SUBSYS(NFNL_SUBSYS_NFT_COMPAT);
-- 
2.20.1
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help