Re: [PATCH 7/7] net/utils: Use strlcpy() instead of open-coding it

[Rustad, Mark D]

On Mar 21, 2019, at 3:19 PM, Bart Van Assche <bvanass...@acm.org> wrote:

This patch does not change any functionality but makes the code easier
to read.

Cc: Sagi Grimberg <s...@grimberg.me>
Cc: Christoph Hellwig <h...@lst.de>
Signed-off-by: Bart Van Assche <bvanass...@acm.org>
---
 net/core/utils.c | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/net/core/utils.c b/net/core/utils.c
index 2a597ac7808e..5f11e89f173f 100644
--- a/net/core/utils.c
+++ b/net/core/utils.c

            src + srclen != scope_delim && *scope_delim == '%') {
                struct net_device *dev;
                char scope_id[16];
-               size_t scope_len = min_t(size_t, sizeof(scope_id) - 1,
-                                        src + srclen - scope_delim - 1);
-
-               memcpy(scope_id, scope_delim + 1, scope_len);
-               scope_id[scope_len] = '\0';

+               strlcpy(scope_id, scope_delim + 1, sizeof(scope_id));
                dev = dev_get_by_name(net, scope_id);
                if (dev) {
                        addr6->sin6_scope_id = dev->ifindex;
--
2.21.0.155.ge902e9bcae20

@@ -339,12 +339,8 @@ static int inet6_pton(struct net *net, const char *src, u16 port_num,

--
Mark Rustad, Networking Division, Intel Corporation

This doesn't look right to me. The original code seemingly would stop the copy before getting to a possible '%' character. The new code only stops at eol. So this appears to change the functionality and I assume is broken in any case that uses the %. Take another look at what that min_t is really doing.

signature.asc
Description: Message signed with OpenPGP