sys/luid: shuffle bytes

[benpicco]

Contribution description

The CPU ID can have large parts that are identical between devices.
To add some more entropy to the generated IDs, shuffle the bytes.

Testing procedure

#include <stdio.h>
#include "luid.h"

int main(void)
{
    eui64_t addr;
    luid_get_eui64(&addr);

    for (size_t i = 0; i < sizeof(addr); ++i) {
        printf("%02x ", addr.uint8[i]);
    }
    puts("");

    return 0;
}

On two weact-f411ce boards (same stm32 CPU) this generates

1a 14 15 22 72 17 78 10

and

12 21 15 23 53 29 72 1b

Issues/PRs references

#14634 (comment)

[fjmolinas]

This does fix the issue for me:

iotlab-m3 1:

2020-09-11 11:41:02,622 # ifconfig
2020-09-11 11:41:02,622 # Iface  6  HWaddr: 12:6B  Channel: 26  Page: 0  NID: 0x23  PHY: O-QPSK
2020-09-11 11:41:02,622 #
2020-09-11 11:41:02,623 #           Long HWaddr: B6:25:28:56:11:FF:12:6B
2020-09-11 11:41:02,634 #            TX-Power: 0dBm  State: IDLE  max. Retrans.: 3  CSMA Retries: 4
2020-09-11 11:41:02,637 #           AUTOACK  ACK_REQ  CSMA  L2-PDU:102  MTU:1280  HL:64  RTR
2020-09-11 11:41:02,638 #           6LO  IPHC
2020-09-11 11:41:02,639 #           Source address length: 8
2020-09-11 11:41:02,641 #           Link type: wireless
2020-09-11 11:41:02,643 #           inet6 addr: fe80::b425:2856:11ff:126b  scope: link  VAL
2020-09-11 11:41:02,644 #           inet6 group: ff02::2
2020-09-11 11:41:02,645 #           inet6 group: ff02::1
2020-09-11 11:41:02,647 #           inet6 group: ff02::1:ffff:126b
2020-09-11 11:41:02,647 #
2020-09-11 11:41:02,648 #           Statistics for Layer 2
2020-09-11 11:41:02,648 #             RX packets 0  bytes 0
2020-09-11 11:41:02,649 #             TX packets 2 (Multicast: 2)  bytes 86
2020-09-11 11:41:02,650 #             TX succeeded 2 errors 0
2020-09-11 11:41:02,650 #           Statistics for IPv6
2020-09-11 11:41:02,650 #             RX packets 0  bytes 0
2020-09-11 11:41:02,651 #             TX packets 2 (Multicast: 2)  bytes 128
2020-09-11 11:41:02,651 #             TX succeeded 2 errors 0
2020-09-11 11:41:02,651 #14030 

• iotlab-m3 2

2020-09-11 11:40:53,043 # ifconfig
2020-09-11 11:40:53,044 # Iface  6  HWaddr: 5E:11  Channel: 26  Page: 0  NID: 0x23  PHY: O-QPSK
2020-09-11 11:40:53,044 #
2020-09-11 11:40:53,046 #           Long HWaddr: 66:69:37:AB:15:10:DE:11
2020-09-11 11:40:53,047 #            TX-Power: 0dBm  State: IDLE  max. Retrans.: 3  CSMA Retries: 4
2020-09-11 11:40:53,047 #           AUTOACK  ACK_REQ  CSMA  L2-PDU:102  MTU:1280  HL:64  RTR
2020-09-11 11:40:53,047 #           6LO  IPHC
2020-09-11 11:40:53,048 #           Source address length: 8
2020-09-11 11:40:53,059 #           Link type: wireless
2020-09-11 11:40:53,060 #           inet6 addr: fe80::6469:37ab:1510:de11  scope: link  VAL
2020-09-11 11:40:53,061 #           inet6 group: ff02::2
2020-09-11 11:40:53,061 #           inet6 group: ff02::1
2020-09-11 11:40:53,062 #           inet6 group: ff02::1:ff10:de11
2020-09-11 11:40:53,062 #
2020-09-11 11:40:53,063 #           Statistics for Layer 2
2020-09-11 11:40:53,063 #             RX packets 0  bytes 0
2020-09-11 11:40:53,064 #             TX packets 4 (Multicast: 4)  bytes 172
2020-09-11 11:40:53,065 #             TX succeeded 4 errors 0
2020-09-11 11:40:53,065 #           Statistics for IPv6
2020-09-11 11:40:53,066 #             RX packets 0  bytes 0
2020-09-11 11:40:53,067 #             TX packets 4 (Multicast: 4)  bytes 256
2020-09-11 11:40:53,068 #             TX succeeded 4 errors 0
2020-09-11 11:40:53,068 #
> ping6  fe80::b425:2856:11ff:126b
2020-09-11 11:41:52,303 #  ping6  fe80::b425:2856:11ff:126b
2020-09-11 11:41:52,320 # 12 bytes from fe80::b425:2856:11ff:126b%6: icmp_seq=0 ttl=64 rssi=-43 dBm time=9.165 ms
2020-09-11 11:41:53,311 # 12 bytes from fe80::b425:2856:11ff:126b%6: icmp_seq=1 ttl=64 rssi=-43 dBm time=10.125 ms
2020-09-11 11:41:54,320 # 12 bytes from fe80::b425:2856:11ff:126b%6: icmp_seq=2 ttl=64 rssi=-44 dBm time=9.806 ms
2020-09-11 11:41:54,320 #
2020-09-11 11:41:54,321 # --- fe80::b425:2856:11ff:126b PING statistics ---
2020-09-11 11:41:54,323 # 3 packets transmitted, 3 packets received, 0% packet loss
2020-09-11 11:41:54,324 # round-trip min/avg/max = 9.165/9.698/10.125 ms

[fjmolinas]

Is this a common shuffling alorithm?

[benpicco]

It's basically Fisher–Yates shuffle

[fjmolinas]

Ah nice, mind adding a reference? You can squash right away!

[benpicco]

I added a comment.

[fjmolinas]

ACK.