[PATCH v2 bpf] bpf: fix leak in LINK_UPDATE and enforce empty old_prog_fd

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Fix bug of not putting bpf_link in LINK_UPDATE command.
Also enforce zeroed old_prog_fd if no BPF_F_REPLACE flag is specified.

Signed-off-by: Andrii Nakryiko <andriin@xxxxxx>
---
This version will merge with no conflicts with the upcoming LINK_UPDATE
refactoring patch (part of bpf_link observability patch set).

 kernel/bpf/syscall.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index d85f37239540..bca58c235ac0 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -3628,8 +3628,10 @@ static int link_update(union bpf_attr *attr)
 		return PTR_ERR(link);
 
 	new_prog = bpf_prog_get(attr->link_update.new_prog_fd);
-	if (IS_ERR(new_prog))
-		return PTR_ERR(new_prog);
+	if (IS_ERR(new_prog)) {
+		ret = PTR_ERR(new_prog);
+		goto out_put_link;
+	}
 
 	if (flags & BPF_F_REPLACE) {
 		old_prog = bpf_prog_get(attr->link_update.old_prog_fd);
@@ -3638,6 +3640,9 @@ static int link_update(union bpf_attr *attr)
 			old_prog = NULL;
 			goto out_put_progs;
 		}
+	} else if (attr->link_update.old_prog_fd) {
+		ret = -EINVAL;
+		goto out_put_progs;
 	}
 
 #ifdef CONFIG_CGROUP_BPF
@@ -3653,6 +3658,8 @@ static int link_update(union bpf_attr *attr)
 		bpf_prog_put(old_prog);
 	if (ret)
 		bpf_prog_put(new_prog);
+out_put_link:
+	bpf_link_put(link);
 	return ret;
 }
 
-- 
2.24.1





[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux