diff mbox

[net-next] netlink: Add nla_memdup() to wrap kmemdup() use on nlattr

Message ID 2d0864c85200f1b42b1ebceee7c2dc60fe29f26a.1477471562.git.tgraf@suug.ch
State Not Applicable
Headers show

Commit Message

Thomas Graf Oct. 26, 2016, 8:53 a.m. UTC
Wrap several common instances of:
	kmemdup(nla_data(attr), nla_len(attr), GFP_KERNEL);

Signed-off-by: Thomas Graf <tgraf@suug.ch>
---
 include/net/netlink.h  | 10 ++++++++++
 net/sched/act_bpf.c    |  4 +---
 net/sched/cls_bpf.c    |  4 +---
 net/wireless/nl80211.c |  3 +--
 4 files changed, 13 insertions(+), 8 deletions(-)

Comments

Johannes Berg Oct. 26, 2016, 8:56 a.m. UTC | #1
On Wed, 2016-10-26 at 10:53 +0200, Thomas Graf wrote:
> Wrap several common instances of:
> 	kmemdup(nla_data(attr), nla_len(attr), GFP_KERNEL);

Makes sense

Acked-by: Johannes Berg <johannes@sipsolutions.net>

johannes
Daniel Borkmann Oct. 26, 2016, 8:57 a.m. UTC | #2
On 10/26/2016 10:53 AM, Thomas Graf wrote:
> Wrap several common instances of:
> 	kmemdup(nla_data(attr), nla_len(attr), GFP_KERNEL);
>
> Signed-off-by: Thomas Graf <tgraf@suug.ch>

Thanks!

Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Johannes Berg Oct. 26, 2016, 8:59 a.m. UTC | #3
>  /**
> + * nla_memdup - duplicate attribute memory (kmemdup)
> + * @src: netlink attribute to duplicate from
> + * @gfp: GFP mask

Actually, is there any point in passing a GFP mask? None of the current
users need it, and it seems fairly unlikely to be needed since this is
typically used on the netlink input path, where you surely shouldn't
need GFP_ATOMIC or so?

johannes
Thomas Graf Oct. 26, 2016, 9:52 a.m. UTC | #4
On 10/26/16 at 10:59am, Johannes Berg wrote:
> 
> >  /**
> > + * nla_memdup - duplicate attribute memory (kmemdup)
> > + * @src: netlink attribute to duplicate from
> > + * @gfp: GFP mask
> 
> Actually, is there any point in passing a GFP mask? None of the current
> users need it, and it seems fairly unlikely to be needed since this is
> typically used on the netlink input path, where you surely shouldn't
> need GFP_ATOMIC or so?

I'm fine either way. I didn't want to make assumptions which need
later changes. It's not hurting either and the function prototype
is very small.
Johannes Berg Oct. 26, 2016, 10:18 a.m. UTC | #5
On Wed, 2016-10-26 at 11:52 +0200, Thomas Graf wrote:
> On 10/26/16 at 10:59am, Johannes Berg wrote:
> > 
> > 
> > > 
> > >  /**
> > > + * nla_memdup - duplicate attribute memory (kmemdup)
> > > + * @src: netlink attribute to duplicate from
> > > + * @gfp: GFP mask
> > 
> > Actually, is there any point in passing a GFP mask? None of the
> > current
> > users need it, and it seems fairly unlikely to be needed since this
> > is
> > typically used on the netlink input path, where you surely
> > shouldn't
> > need GFP_ATOMIC or so?
> 
> I'm fine either way. I didn't want to make assumptions which need
> later changes. It's not hurting either and the function prototype
> is very small.

Yeah, I don't really care much - just wondered.

johannes
David Miller Oct. 29, 2016, 6:57 p.m. UTC | #6
From: Thomas Graf <tgraf@suug.ch>
Date: Wed, 26 Oct 2016 10:53:16 +0200

> Wrap several common instances of:
> 	kmemdup(nla_data(attr), nla_len(attr), GFP_KERNEL);
> 
> Signed-off-by: Thomas Graf <tgraf@suug.ch>

Applied.
diff mbox

Patch

diff --git a/include/net/netlink.h b/include/net/netlink.h
index 254a0fc..a34f53a 100644
--- a/include/net/netlink.h
+++ b/include/net/netlink.h
@@ -1191,6 +1191,16 @@  static inline struct in6_addr nla_get_in6_addr(const struct nlattr *nla)
 }
 
 /**
+ * nla_memdup - duplicate attribute memory (kmemdup)
+ * @src: netlink attribute to duplicate from
+ * @gfp: GFP mask
+ */
+static inline void *nla_memdup(const struct nlattr *src, gfp_t gfp)
+{
+	return kmemdup(nla_data(src), nla_len(src), gfp);
+}
+
+/**
  * nla_nest_start - Start a new level of nested attributes
  * @skb: socket buffer to add attributes to
  * @attrtype: attribute type of container
diff --git a/net/sched/act_bpf.c b/net/sched/act_bpf.c
index 1d39600..9ff06cf 100644
--- a/net/sched/act_bpf.c
+++ b/net/sched/act_bpf.c
@@ -226,9 +226,7 @@  static int tcf_bpf_init_from_efd(struct nlattr **tb, struct tcf_bpf_cfg *cfg)
 		return PTR_ERR(fp);
 
 	if (tb[TCA_ACT_BPF_NAME]) {
-		name = kmemdup(nla_data(tb[TCA_ACT_BPF_NAME]),
-			       nla_len(tb[TCA_ACT_BPF_NAME]),
-			       GFP_KERNEL);
+		name = nla_memdup(tb[TCA_ACT_BPF_NAME], GFP_KERNEL);
 		if (!name) {
 			bpf_prog_put(fp);
 			return -ENOMEM;
diff --git a/net/sched/cls_bpf.c b/net/sched/cls_bpf.c
index bb1d5a4..52dc85a 100644
--- a/net/sched/cls_bpf.c
+++ b/net/sched/cls_bpf.c
@@ -369,9 +369,7 @@  static int cls_bpf_prog_from_efd(struct nlattr **tb, struct cls_bpf_prog *prog,
 		return PTR_ERR(fp);
 
 	if (tb[TCA_BPF_NAME]) {
-		name = kmemdup(nla_data(tb[TCA_BPF_NAME]),
-			       nla_len(tb[TCA_BPF_NAME]),
-			       GFP_KERNEL);
+		name = nla_memdup(tb[TCA_BPF_NAME], GFP_KERNEL);
 		if (!name) {
 			bpf_prog_put(fp);
 			return -ENOMEM;
diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c
index c510810..7bfe1e0 100644
--- a/net/wireless/nl80211.c
+++ b/net/wireless/nl80211.c
@@ -10638,8 +10638,7 @@  static int handle_nan_filter(struct nlattr *attr_filter,
 
 	i = 0;
 	nla_for_each_nested(attr, attr_filter, rem) {
-		filter[i].filter = kmemdup(nla_data(attr), nla_len(attr),
-					   GFP_KERNEL);
+		filter[i].filter = nla_memdup(attr, GFP_KERNEL);
 		filter[i].len = nla_len(attr);
 		i++;
 	}