Few Minor per-utxo assert-semantics re-adds and tweak #10537

1. 
   TheBlueMatt commented at 0:23 am on June 6, 2017: member

   One small tweak which @sdaftuar commented and got missed, and two restore-previous-assert-semantics that I figured would be nice to re-add.
2. Remove useless mapNextTx lookup in CTxMemPool::TrimToSize.

   Prior to per-utxo CCoins, we checked that no other in-mempool tx
   spent any of the given transaction's outputs, as we don't want to
   uncache that entire tx in such a case. However, we now are checking
   only that there exists no other mempool spends of the same output,
   which should clearly be impossible after we removed the transaction
   which was spending said output (barring massive mempool
   inconsistency).
   
   Thanks to @sdaftuar for the suggestion.

   ec1271f2be
3. TheBlueMatt force-pushed on Jun 6, 2017
4. laanwj added the label UTXO Db and Indexes on Jun 6, 2017
5. 
   gmaxwell commented at 9:53 am on June 6, 2017: contributor

   utACK (will test)
6. 
   sipa commented at 0:37 am on June 9, 2017: member

   utACK 743bdb5d52f7f7a0bcddfd55bccdac287140b491
7. in src/validation.cpp:1121 in 30a2265ac7 outdated

   1117@@ -1118,7 +1118,8 @@ void UpdateCoins(const CTransaction& tx, CCoinsViewCache& inputs, CTxUndo &txund
   1118         txundo.vprevout.reserve(tx.vin.size());
   1119         BOOST_FOREACH(const CTxIn &txin, tx.vin) {
   1120             txundo.vprevout.emplace_back();
   1121-            inputs.SpendCoin(txin.prevout, &txundo.vprevout.back());
   1122+            bool fSpent = inputs.SpendCoin(txin.prevout, &txundo.vprevout.back());
   

   ---

   ---

   

   ryanofsky commented at 3:09 pm on June 9, 2017:

   In commit “Return a bool in SpendCoin to restore pre-per-utxo assert semantics”

   Maybe drop hungarian.

8. in src/consensus/tx_verify.cpp:126 in fc9fe49ba8 outdated

   125@@ -126,7 +126,9 @@ unsigned int GetP2SHSigOpCount(const CTransaction& tx, const CCoinsViewCache& in
   126     unsigned int nSigOps = 0;
   

   ---

   ---

   

   ryanofsky commented at 3:12 pm on June 9, 2017:

   In commit “Restore some assert semantics in sigop cost calculations”

   Maybe update commit message to suggest why policy/atmp asserts aren’t added back. Not possible there, or just not worth the effort?

   ---

   

   TheBlueMatt commented at 4:09 pm on June 9, 2017:

   I figured it wasn’t worth it - if we hit those errors in consensus code, we’d better crash cause we might fall out of consensus…if we accidentally add something to our mempool that should be out of policy…oh well, not a big deal. (I updated commit message).
9. 
   ryanofsky commented at 3:46 pm on June 9, 2017: member

   utACK 743bdb5d52f7f7a0bcddfd55bccdac287140b491

   It took me a little while to decipher first commit, but just to put it in my own words, the map mapNextTx.count(txin.prevout) check that this PR removes would always returns 0 because you can’t have two transactions in the mempool spending the same prevout, and the one transaction that had been spending it just got removed out a few lines previously in RemoveStaged.

10. TheBlueMatt force-pushed on Jun 9, 2017
11. Return a bool in SpendCoin to restore pre-per-utxo assert semantics

    Since its free to do so, assert that Spends succeeded when we expect
    them to.

    3533fb4d33
12. Restore some assert semantics in sigop cost calculations

    There are some similar asserts which are left removed in policy
    and ATMP (policy code being broken isn't a huge deal, but if we
    fail to verify some consensus rules, we should most definitely
    crash).

    f58349ca85
13. Be much more agressive in AccessCoin docs.

    While the current implementation is pretty free, there is a lot
    of possibility for this to blow up in our face with future changes,
    especially as the backing map gets tweaked.

    9417d7a336
14. TheBlueMatt force-pushed on Jun 9, 2017
15. 
    ryanofsky commented at 5:27 pm on June 13, 2017: member

    utACK 9417d7a336e71ee31ce673a6de5abb3d56941204. Has style guide fixes and the updated commit message.
16. 
    sipa commented at 1:17 am on June 21, 2017: member

    utACK 9417d7a336e71ee31ce673a6de5abb3d56941204
17. sipa merged this on Jun 21, 2017
18. sipa closed this on Jun 21, 2017

    ---

19. sipa referenced this in commit b3eb0d6485 on Jun 21, 2017
20. codablock referenced this in commit ab159b0c53 on Sep 27, 2017
21. codablock referenced this in commit 066ce91851 on Oct 12, 2017
22. codablock referenced this in commit b3f3191eba on Oct 26, 2017
23. codablock referenced this in commit 16989c3d60 on Oct 26, 2017
24. codablock referenced this in commit 4d742f5bee on Oct 26, 2017
25. codablock referenced this in commit fad5bdcfc0 on Oct 30, 2017
26. codablock referenced this in commit 0354804df2 on Oct 31, 2017
27. codablock referenced this in commit 44a2ccb733 on Oct 31, 2017
28. codablock referenced this in commit 151c552c71 on Oct 31, 2017
29. UdjinM6 referenced this in commit d5c53d5c43 on Nov 8, 2017
30. jasonbcox referenced this in commit 28c6e32e5d on Sep 13, 2019
31. jonspock referenced this in commit 7564d05192 on Dec 22, 2019
32. proteanx referenced this in commit 38f7d315f4 on Dec 23, 2019
33. DrahtBot locked this on Sep 8, 2021

Contributors
TheBlueMatt gmaxwell sipa ryanofsky

Labels
UTXO Db and Indexes