Larry Ruane

@jamesob: > ... reindexing to 450,000 isn't showing much difference on my end between the HEAD of this branch and the mergebase That's strange. What does reindexing to 450,000 mean?...

@hebasto, thank you for doing all those tests. I did more testing (that's why I changed this PR to a Draft), and I'm now convinced there's no functionally important difference...

> a block file has random/broken bytes but _all block are correct_ within it. Can you explain more? Where exactly is the corruption?

#### TL;DR I think the PR is good as it is can and should be merged. @hebasto, I understand exactly what's going on, and having that corrupted file really helped,...

I benchmarked this latest version, and it's only 3% slower than the PR version. This latest version has the faster `FindByte()` and also resynchronizes with the byte stream in the...

@hebasto, that makes sense, I took your suggestion. Force-pushed ([diff](https://github.com/bitcoin/bitcoin/compare/ab68fdfae59dea806493f8542486bbb4ae07a3cb..ad386f705cd2879dccfb3e1ce0dbcf5b6f8e1ad6)). The only slight change I made from your suggestion is this: ``` if (!CheckProofOfWork(hash, header.nBits, chainparams.GetConsensus())) continue; { LOCK(cs_main); (...)...

> Two last commits with the same messages are confusing Yes, that was a mistake, thanks for catching that. Force-pushed ([diff](https://github.com/bitcoin/bitcoin/compare/ab68fdfae59dea806493f8542486bbb4ae07a3cb..e5724b5fde855b67ff500e21f11274f12b5c893a)).

@hebasto, you're too quick! Sorry, I confused myself with my lack of git skills... I just make a force-push to do a small cleanup ([diff](https://github.com/bitcoin/bitcoin/compare/e5724b5fde855b67ff500e21f11274f12b5c893a..84b995a9ef468e0badf7bc0378851ea8c698dfc7)) and ended up doing some...

@pinheadmz, @hebasto, I just added another commit, please review. I discovered an additional minor performance improvement in the case where the block is already present in the block index (the...

> The latest commit [969366d](https://github.com/bitcoin/bitcoin/commit/969366d07c8cfdf8ea56cdb2ae11f5c648152345) looks good, but it adds _another_ behavior change into this PR. I think it deserves its own PR to get the measurable effects from each...