| Message ID | 1288778342-10056-1-git-send-email-tklauser@distanz.ch |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
On Wednesday 03 November 2010 10:59:02 you wrote: > Replace custom ethernet address check functions by calls to the helpers > in linux/etherdevice.h Have you proven that the addresses are always two bytes aligned in memory? Afaik this function needs it. > In one case where the address was tested for broadcast and multicast > address, the broadcast address check can be omitted as broadcast is also > a multicast address. We need to distinguish between these two types for different optimizations (research is currently done for multicast over mesh - but currently not part of batman-adv). Best regards, Sven
On 2010-11-03 at 11:12:12 +0100, Sven Eckelmann <sven.eckelmann@gmx.de> wrote: > On Wednesday 03 November 2010 10:59:02 you wrote: > > Replace custom ethernet address check functions by calls to the helpers > > in linux/etherdevice.h > > Have you proven that the addresses are always two bytes aligned in memory? > Afaik this function needs it. I don't think they need to be two bytes aligned, but I might be wrong. > > In one case where the address was tested for broadcast and multicast > > address, the broadcast address check can be omitted as broadcast is also > > a multicast address. > > We need to distinguish between these two types for different optimizations > (research is currently done for multicast over mesh - but currently not part > of batman-adv). I was refering to the following part of the patch: - if (is_bcast(ethhdr->h_dest) || is_mcast(ethhdr->h_dest)) { + if (is_multicast_ether_addr(ethhdr->h_dest)) { I think this change should be legitimate as the same branch is done for multicast and broadcast addresses. Thanks Tobias
On Wednesday 03 November 2010 11:25:12 Tobias Klauser wrote: > On 2010-11-03 at 11:12:12 +0100, Sven Eckelmann <sven.eckelmann@gmx.de> wrote: > > On Wednesday 03 November 2010 10:59:02 you wrote: > > > Replace custom ethernet address check functions by calls to the helpers > > > in linux/etherdevice.h > > > > Have you proven that the addresses are always two bytes aligned in > > memory? Afaik this function needs it. > > I don't think they need to be two bytes aligned, but I might be wrong. compare_ether_addr uses a two byte pointer to access 3x two bytes. This makes it necessary to have all those 3 bytes aligned to 2 byte boundaries. Otherwise the compiler has to generate special instructions on architectures which don't support loads on non-aligned addresses. Usually he doesn't do it unless he has some indications that it is necessary (__attribute__ ((packed)) for example). There is also documentation available on that topic in Documentation/unaligned-memory-access.txt And maybe it is good to use is_broadcast_ether_addr, but leave compare_ether_addr part open (or prove that we always have those two operands correctly aligned). > > > In one case where the address was tested for broadcast and multicast > > > address, the broadcast address check can be omitted as broadcast is > > > also a multicast address. > > > > We need to distinguish between these two types for different > > optimizations (research is currently done for multicast over mesh - but > > currently not part of batman-adv). > > I was refering to the following part of the patch: > > - if (is_bcast(ethhdr->h_dest) || is_mcast(ethhdr->h_dest)) { > + if (is_multicast_ether_addr(ethhdr->h_dest)) { > > I think this change should be legitimate as the same branch is done for > multicast and broadcast addresses. Yes, in that situation it is correct. I was talking more about the complete bcast vs. multicast in batman-adv and didn't look at your patch. Best regards, Sven
On Wednesday 03 November 2010 11:56:19 Sven Eckelmann wrote: > On Wednesday 03 November 2010 11:25:12 Tobias Klauser wrote: > > On 2010-11-03 at 11:12:12 +0100, Sven Eckelmann <sven.eckelmann@gmx.de> > > wrote: > > > On Wednesday 03 November 2010 10:59:02 you wrote: > > > > Replace custom ethernet address check functions by calls to the > > > > helpers in linux/etherdevice.h > > > > > > Have you proven that the addresses are always two bytes aligned in > > > memory? Afaik this function needs it. > > > > I don't think they need to be two bytes aligned, but I might be wrong. > > compare_ether_addr uses a two byte pointer to access 3x two bytes. This > makes it necessary to have all those 3 bytes aligned to 2 byte boundaries. 3x two bytes for each ethernet address == 6x two bytes and not 3 bytes... I don't know why I wrote that. Best regards, Sven
On Wednesday 03 November 2010 11:25:12 Tobias Klauser wrote: > I was refering to the following part of the patch: > > - if (is_bcast(ethhdr->h_dest) || is_mcast(ethhdr->h_dest)) { > + if (is_multicast_ether_addr(ethhdr->h_dest)) { > > I think this change should be legitimate as the same branch is done for > multicast and broadcast addresses. Soon we will have different code paths for broadcast and multicast right at this spot as multicast can be optimized further. But I think we can merge the change for the time being. Regards, Marek
diff --git a/drivers/staging/batman-adv/main.c b/drivers/staging/batman-adv/main.c index 0587940..6ea6420 100644 --- a/drivers/staging/batman-adv/main.c +++ b/drivers/staging/batman-adv/main.c @@ -149,7 +149,7 @@ void dec_module_count(void) int compare_orig(void *data1, void *data2) { - return (memcmp(data1, data2, ETH_ALEN) == 0 ? 1 : 0); + return (compare_ether_addr(data1, data2) == 0 ? 1 : 0); } /* hashfunction to choose an entry in a hash table of given size */ @@ -192,16 +192,6 @@ int is_my_mac(uint8_t *addr) } -int is_bcast(uint8_t *addr) -{ - return (addr[0] == (uint8_t)0xff) && (addr[1] == (uint8_t)0xff); -} - -int is_mcast(uint8_t *addr) -{ - return *addr & 0x01; -} - module_init(batman_init); module_exit(batman_exit); diff --git a/drivers/staging/batman-adv/main.h b/drivers/staging/batman-adv/main.h index 5e3f516..14d567d 100644 --- a/drivers/staging/batman-adv/main.h +++ b/drivers/staging/batman-adv/main.h @@ -109,6 +109,7 @@ #include <linux/mutex.h> /* mutex */ #include <linux/module.h> /* needed by all modules */ #include <linux/netdevice.h> /* netdevice */ +#include <linux/etherdevice.h> #include <linux/if_ether.h> /* ethernet header */ #include <linux/poll.h> /* poll_table */ #include <linux/kthread.h> /* kernel threads */ @@ -138,8 +139,6 @@ void dec_module_count(void); int compare_orig(void *data1, void *data2); int choose_orig(void *data, int32_t size); int is_my_mac(uint8_t *addr); -int is_bcast(uint8_t *addr); -int is_mcast(uint8_t *addr); #ifdef CONFIG_BATMAN_ADV_DEBUG int debug_log(struct bat_priv *bat_priv, char *fmt, ...); diff --git a/drivers/staging/batman-adv/routing.c b/drivers/staging/batman-adv/routing.c index 9010263..d42c165 100644 --- a/drivers/staging/batman-adv/routing.c +++ b/drivers/staging/batman-adv/routing.c @@ -756,11 +756,11 @@ int recv_bat_packet(struct sk_buff *skb, struct batman_if *batman_if) ethhdr = (struct ethhdr *)skb_mac_header(skb); /* packet with broadcast indication but unicast recipient */ - if (!is_bcast(ethhdr->h_dest)) + if (!is_broadcast_ether_addr(ethhdr->h_dest)) return NET_RX_DROP; /* packet with broadcast sender address */ - if (is_bcast(ethhdr->h_source)) + if (is_broadcast_ether_addr(ethhdr->h_source)) return NET_RX_DROP; /* create a copy of the skb, if needed, to modify it. */ @@ -933,11 +933,11 @@ int recv_icmp_packet(struct sk_buff *skb, struct batman_if *recv_if) ethhdr = (struct ethhdr *)skb_mac_header(skb); /* packet with unicast indication but broadcast recipient */ - if (is_bcast(ethhdr->h_dest)) + if (is_broadcast_ether_addr(ethhdr->h_dest)) return NET_RX_DROP; /* packet with broadcast sender address */ - if (is_bcast(ethhdr->h_source)) + if (is_broadcast_ether_addr(ethhdr->h_source)) return NET_RX_DROP; /* not for me */ @@ -1107,11 +1107,11 @@ static int check_unicast_packet(struct sk_buff *skb, int hdr_size) ethhdr = (struct ethhdr *)skb_mac_header(skb); /* packet with unicast indication but broadcast recipient */ - if (is_bcast(ethhdr->h_dest)) + if (is_broadcast_ether_addr(ethhdr->h_dest)) return -1; /* packet with broadcast sender address */ - if (is_bcast(ethhdr->h_source)) + if (is_broadcast_ether_addr(ethhdr->h_source)) return -1; /* not for me */ @@ -1283,11 +1283,11 @@ int recv_bcast_packet(struct sk_buff *skb, struct batman_if *recv_if) ethhdr = (struct ethhdr *)skb_mac_header(skb); /* packet with broadcast indication but unicast recipient */ - if (!is_bcast(ethhdr->h_dest)) + if (!is_broadcast_ether_addr(ethhdr->h_dest)) return NET_RX_DROP; /* packet with broadcast sender address */ - if (is_bcast(ethhdr->h_source)) + if (is_broadcast_ether_addr(ethhdr->h_source)) return NET_RX_DROP; /* ignore broadcasts sent by myself */ diff --git a/drivers/staging/batman-adv/soft-interface.c b/drivers/staging/batman-adv/soft-interface.c index 3904db9..820e141 100644 --- a/drivers/staging/batman-adv/soft-interface.c +++ b/drivers/staging/batman-adv/soft-interface.c @@ -140,7 +140,7 @@ int interface_tx(struct sk_buff *skb, struct net_device *soft_iface) hna_local_add(soft_iface, ethhdr->h_source); /* ethernet packet should be broadcasted */ - if (is_bcast(ethhdr->h_dest) || is_mcast(ethhdr->h_dest)) { + if (is_multicast_ether_addr(ethhdr->h_dest)) { if (!bat_priv->primary_if) goto dropped; diff --git a/drivers/staging/batman-adv/vis.c b/drivers/staging/batman-adv/vis.c index 3d2c1bc..c98a065 100644 --- a/drivers/staging/batman-adv/vis.c +++ b/drivers/staging/batman-adv/vis.c @@ -470,7 +470,7 @@ void receive_client_update_packet(struct bat_priv *bat_priv, int are_target = 0; /* clients shall not broadcast. */ - if (is_bcast(vis_packet->target_orig)) + if (is_broadcast_ether_addr(vis_packet->target_orig)) return; /* Are we the target for this VIS packet? */ @@ -747,7 +747,7 @@ static void send_vis_packet(struct bat_priv *bat_priv, struct vis_info *info) ETH_ALEN); packet->ttl--; - if (is_bcast(packet->target_orig)) + if (is_broadcast_ether_addr(packet->target_orig)) broadcast_vis_packet(bat_priv, info); else unicast_vis_packet(bat_priv, info);
Replace custom ethernet address check functions by calls to the helpers in linux/etherdevice.h In one case where the address was tested for broadcast and multicast address, the broadcast address check can be omitted as broadcast is also a multicast address. The patch is only compile-tested. Cc: Marek Lindner <lindner_marek@yahoo.de> Cc: Simon Wunderlich <siwu@hrz.tu-chemnitz.de> Cc: Andrew Lunn <andrew@lunn.ch> Signed-off-by: Tobias Klauser <tklauser@distanz.ch> --- drivers/staging/batman-adv/main.c | 12 +----------- drivers/staging/batman-adv/main.h | 3 +-- drivers/staging/batman-adv/routing.c | 16 ++++++++-------- drivers/staging/batman-adv/soft-interface.c | 2 +- drivers/staging/batman-adv/vis.c | 4 ++-- 5 files changed, 13 insertions(+), 24 deletions(-)