Netdev List
 help / color / mirror / Atom feed
From: John Fastabend <john.fastabend@gmail.com>
To: alexei.starovoitov@gmail.com, daniel@iogearbox.net, davem@davemloft.net
Cc: netdev@vger.kernel.org
Subject: [RFC PATCH 04/16] bpf: cfg: detect loop use domination information
Date: Fri, 01 Jun 2018 02:32:38 -0700	[thread overview]
Message-ID: <20180601093238.15353.12809.stgit@john-Precision-Tower-5810> (raw)
In-Reply-To: <20180601092646.15353.28269.stgit@john-Precision-Tower-5810>

From: Jiong Wang <jiong.wang@netronome.com>

If one bb is dominating its predecessor, then there is loop.

Signed-off-by: Jiong Wang <jiong.wang@netronome.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
---
 kernel/bpf/cfg.c      |   22 ++++++++++++++++++++++
 kernel/bpf/cfg.h      |    1 +
 kernel/bpf/verifier.c |    8 ++++++++
 3 files changed, 31 insertions(+)

diff --git a/kernel/bpf/cfg.c b/kernel/bpf/cfg.c
index b50937a..90692e4 100644
--- a/kernel/bpf/cfg.c
+++ b/kernel/bpf/cfg.c
@@ -568,6 +568,28 @@ int subprog_build_dom_info(struct bpf_subprog_info *subprog)
 	return ret;
 }
 
+bool subprog_has_loop(struct bpf_subprog_info *subprog)
+{
+	int lane_len = BITS_TO_LONGS(subprog->bb_num - 2);
+	struct list_head *bb_list = &subprog->bbs;
+	struct bb_node *bb, *entry_bb;
+	struct edge_node *e;
+
+	entry_bb = entry_bb(bb_list);
+	bb = bb_next(entry_bb);
+	list_for_each_entry_from(bb, &exit_bb(bb_list)->l, l)
+		list_for_each_entry(e, &bb->e_prevs, l) {
+			struct bb_node *latch = e->src;
+
+			if (latch != entry_bb &&
+			    test_bit(bb->idx,
+				     subprog->dtree + latch->idx * lane_len))
+				return true;
+		}
+
+	return false;
+}
+
 static void subprog_free_edge(struct bb_node *bb)
 {
 	struct list_head *succs = &bb->e_succs;
diff --git a/kernel/bpf/cfg.h b/kernel/bpf/cfg.h
index cbb44f2..c02c4cf 100644
--- a/kernel/bpf/cfg.h
+++ b/kernel/bpf/cfg.h
@@ -12,6 +12,7 @@
 int subprog_append_bb(struct list_head *bb_list, int head);
 int subprog_build_dom_info(struct bpf_subprog_info *subprog);
 int subprog_fini_bb(struct list_head *bb_list, int subprog_end);
+bool subprog_has_loop(struct bpf_subprog_info *subprog);
 int subprog_init_bb(struct list_head *bb_list, int subprog_start);
 void subprog_free(struct bpf_subprog_info *subprog, int end_idx);
 
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index eccaee4..c349c45 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -904,6 +904,14 @@ static int check_subprogs(struct bpf_verifier_env *env)
 			if (ret < 0)
 				goto free_nodes;
 			subprog[cur_subprog].bb_num = ret;
+			ret = subprog_build_dom_info(&subprog[cur_subprog]);
+			if (ret < 0)
+				goto free_nodes;
+			if (subprog_has_loop(&subprog[cur_subprog])) {
+				verbose(env, "cfg - loop detected");
+				ret = -EINVAL;
+				goto free_nodes;
+			}
 			subprog_start = subprog_end;
 			cur_subprog++;
 			if (cur_subprog < env->subprog_cnt) {

  parent reply	other threads:[~2018-06-01  9:32 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-01  9:32 [RFC PATCH 00/16] bpf, bounded loop support work in progress John Fastabend
2018-06-01  9:32 ` [RFC PATCH 01/16] bpf: cfg: partition basic blocks for each subprog John Fastabend
2018-06-01  9:32 ` [RFC PATCH 02/16] bpf: cfg: add edges between basic blocks to form CFG John Fastabend
2018-06-01  9:32 ` [RFC PATCH 03/16] bpf: cfg: build domination tree using Tarjan algorithm John Fastabend
2018-06-01  9:32 ` John Fastabend [this message]
2018-06-01  9:32 ` [RFC PATCH 05/16] bpf: cfg: detect unreachable basic blocks John Fastabend
2018-06-01  9:32 ` [RFC PATCH 06/16] bpf: cfg: move find_subprog/add_subprog to cfg.c John Fastabend
2018-06-01  9:32 ` [RFC PATCH 07/16] bpf: cfg: build call graph and detect unreachable/recursive call John Fastabend
2018-06-01  9:32 ` [RFC PATCH 08/16] bpf: cfg: remove push_insn and check_cfg John Fastabend
2018-06-01  9:33 ` [RFC PATCH 09/16] bpf: cfg: reduce k*alloc/free call by using memory pool for allocating nodes John Fastabend
2018-06-01  9:33 ` [RFC PATCH 10/16] bpf: cfg: reduce memory usage by using singly list + compat pointer John Fastabend
2018-06-01  9:33 ` [RFC PATCH 11/16] bpf: cfg: detect irreducible loop using Eric Stoltz algorithm John Fastabend
2018-06-01  9:33 ` [RFC PATCH 12/16] bpf: cfg: pretty print CFG and DOM John Fastabend
2018-06-01  9:33 ` [RFC PATCH 13/16] bpf: verifier, can ptr range be calculated with scalar ALU op John Fastabend
2018-06-01  9:33 ` [RFC PATCH 14/16] bpf: verifier, add initial support to allow bounded loops John Fastabend
2018-06-01  9:33 ` [RFC PATCH 15/16] bpf: verifier, simple loop examples John Fastabend
2018-06-01  9:33 ` [RFC PATCH 16/16] bpf: tools: dbg patch to turn on debugging and add primitive examples John Fastabend

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20180601093238.15353.12809.stgit@john-Precision-Tower-5810 \
    --to=john.fastabend@gmail.com \
    --cc=alexei.starovoitov@gmail.com \
    --cc=daniel@iogearbox.net \
    --cc=davem@davemloft.net \
    --cc=netdev@vger.kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox