Messages in this thread Patch in this message | | | From | Abhishek Bist <> | Subject | [PATCH] Staging : wlan-ng: fix memcpy with ether_addr_copy in p80211conv.c | Date | Tue, 5 May 2015 16:25:31 +0530 |
| |
This is a patch which fixes memcpy warning found by a checpatch.pl in p80211conv.c and replaces memcpy with ether_addr_copy. --- drivers/staging/wlan-ng/p80211conv.c | 20 ++++++++++---------- 1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/staging/wlan-ng/p80211conv.c b/drivers/staging/wlan-ng/p80211conv.c index bd69e8c..b3f48074 100644 --- a/drivers/staging/wlan-ng/p80211conv.c +++ b/drivers/staging/wlan-ng/p80211conv.c @@ -178,21 +178,21 @@ int skb_ether_to_p80211(wlandevice_t *wlandev, u32 ethconv, switch (wlandev->macmode) { case WLAN_MACMODE_IBSS_STA: - memcpy(p80211_hdr->a3.a1, &e_hdr.daddr, ETH_ALEN); - memcpy(p80211_hdr->a3.a2, wlandev->netdev->dev_addr, ETH_ALEN); - memcpy(p80211_hdr->a3.a3, wlandev->bssid, ETH_ALEN); + ether_addr_copy(p80211_hdr->a3.a1, &e_hdr.daddr); + ether_addr_copy(p80211_hdr->a3.a2, wlandev->netdev->dev_addr); + ether_addr_copy(p80211_hdr->a3.a3, wlandev->bssid); break; case WLAN_MACMODE_ESS_STA: fc |= cpu_to_le16(WLAN_SET_FC_TODS(1)); - memcpy(p80211_hdr->a3.a1, wlandev->bssid, ETH_ALEN); - memcpy(p80211_hdr->a3.a2, wlandev->netdev->dev_addr, ETH_ALEN); - memcpy(p80211_hdr->a3.a3, &e_hdr.daddr, ETH_ALEN); + ether_addr_copy(p80211_hdr->a3.a1, wlandev->bssid); + ether_addr_copy(p80211_hdr->a3.a2, wlandev->netdev->dev_addr); + ether_addr_copy(p80211_hdr->a3.a3, &e_hdr.daddr); break; case WLAN_MACMODE_ESS_AP: fc |= cpu_to_le16(WLAN_SET_FC_FROMDS(1)); - memcpy(p80211_hdr->a3.a1, &e_hdr.daddr, ETH_ALEN); - memcpy(p80211_hdr->a3.a2, wlandev->bssid, ETH_ALEN); - memcpy(p80211_hdr->a3.a3, &e_hdr.saddr, ETH_ALEN); + ether_addr_copy(p80211_hdr->a3.a1, &e_hdr.daddr); + ether_addr_copy(p80211_hdr->a3.a2, wlandev->bssid); + ether_addr_copy(p80211_hdr->a3.a3, &e_hdr.saddr); break; default: netdev_err(wlandev->netdev, @@ -241,7 +241,7 @@ static void orinoco_spy_gather(wlandevice_t *wlandev, char *mac, for (i = 0; i < wlandev->spy_number; i++) { if (!memcmp(wlandev->spy_address[i], mac, ETH_ALEN)) { - memcpy(wlandev->spy_address[i], mac, ETH_ALEN); + ether_addr_copy(wlandev->spy_address[i], mac); wlandev->spy_stat[i].level = rxmeta->signal; wlandev->spy_stat[i].noise = rxmeta->noise; wlandev->spy_stat[i].qual = -- 1.8.3.1 How do you "know" it is safe to do this? Please prove it.
Yes, it is safe to do this because the basic pre-requisite of ether_addr_copy() over memcpy() is that ether_addr_copy() is strictly designed to copy an ethernet address where source and destination are a 6 byte array which is true in my case as p80211_hdr->a3.a1,p80211_hdr->a3.a2,p80211_hdr->a3.a3 are all a six byte array and the next important thing is that they should be u16 alligned.So, for that sake i checked it by pahole which clearly shows that the source and destination are u16 byte alligned.And on some architecture ether_addr_copy is faster than memcpy(). So, it's seems fair enough to use ether_addr_copy() over memcpy() when the size of data to be copied is ETH_ALEN which is #define ETH_ALEN 6
| |