net: Correct comparison of addr count #15345

1. 
   dongcarl commented at 6:15 PM on February 4, 2019: member

   LOCAL_NONE is supposed to be an enum indicating the nScore of a LocalServiceInfo rather than the count of an addr in mapLocalHost.

2. net: Correct comparison of addr count

   LOCAL_NONE is supposed to be an enum indicating the score of a
   LocalServiceInfo rather than the count of an addr in mapLocalHost.

   107623c26c
3. in src/net.cpp:177 in 107623c26c

    173 | @@ -174,8 +174,7 @@ CAddress GetLocalAddress(const CNetAddr *paddrPeer, ServiceFlags nLocalServices)
    174 |  static int GetnScore(const CService& addr)
    175 |  {
    176 |      LOCK(cs_mapLocalHost);
    177 | -    if (mapLocalHost.count(addr) == LOCAL_NONE)
   

   ---

   ---

   

   laanwj commented at 6:19 PM on February 4, 2019:

   Huh, this is really strange. I guess it's correct because LOCAL_NONE effectively is 0 but it's certainly not right.

4. 
   laanwj commented at 6:21 PM on February 4, 2019: member

   utACK 107623c26c2113428446effaa24edb986a3e780c

5. laanwj added the label P2P on Feb 4, 2019
6. MarcoFalke added the label Refactoring on Feb 4, 2019
7. 
   MarcoFalke commented at 6:31 PM on February 4, 2019: member

   ACK 107623c26c (also checked that this commit doesn't change the binary produced by clang7 on my system, so tagged with "refactoring")

8. MarcoFalke added this to the milestone 0.18.0 on Feb 4, 2019
9. 
   promag commented at 6:58 PM on February 4, 2019: member

   utACK 107623c, although it could improve a bit by avoiding 2nd lookup.

   std::map<CNetAddr, LocalServiceInfo>::const_iterator it = mapLocalHost.find(addr);
   return it == mapLocalHost.end() ? 0 : it->second.nScore;
   

10. 
    jonasschnelli commented at 7:01 PM on February 4, 2019: contributor

    utACK 107623c26c2113428446effaa24edb986a3e780c

11. 
    laanwj commented at 7:04 PM on February 4, 2019: member

    utACK 107623c, although it could improve a bit by avoiding 2nd lookup.

    I thought about commenting that, but I don't think that's really much of an improvement in readability (and this isn't exactly a critical path for perfomance).

12. laanwj merged this on Feb 4, 2019
13. laanwj closed this on Feb 4, 2019

    ---

14. laanwj referenced this in commit bfbe42585e on Feb 4, 2019
15. deadalnix referenced this in commit 7df5863a23 on May 30, 2020
16. ftrader referenced this in commit a5226ac3c4 on Aug 17, 2020
17. PastaPastaPasta referenced this in commit 00afeff55d on Jun 25, 2021
18. PastaPastaPasta referenced this in commit ae6ce01be6 on Jun 25, 2021
19. PastaPastaPasta referenced this in commit 2796f46f3c on Jun 27, 2021
20. MarcoFalke locked this on Dec 16, 2021

Milestone
0.18.0