<dzxzg>
but that makes me think that we should also be looping extra_txn first since it's so much smaller, in any case where we had to take from the extra pool we could still get the early exit from hashing the whole mempool
<dzxzg>
Unless there is some clever reason why txns_randomized was used :D
zeropoint has quit [Quit: leaving]
Cory49 has joined #bitcoin-core-dev
Cory54 has joined #bitcoin-core-dev
Cory59 has joined #bitcoin-core-dev
w0xlt has joined #bitcoin-core-dev
Cory24 has quit [Ping timeout: 250 seconds]
Cory49 has quit [Ping timeout: 250 seconds]
Cory54 has quit [Ping timeout: 250 seconds]
w0xlt has quit [Ping timeout: 248 seconds]
Cory16 has joined #bitcoin-core-dev
Cory59 has quit [Ping timeout: 250 seconds]
Cory16 has quit [Quit: Client closed]
Cory16 has joined #bitcoin-core-dev
Cory58 has joined #bitcoin-core-dev
Cory16 has quit [Ping timeout: 250 seconds]
Cory47 has joined #bitcoin-core-dev
Cory58 has quit [Ping timeout: 250 seconds]
Cory15 has joined #bitcoin-core-dev
Cory47 has quit [Ping timeout: 250 seconds]
Cory84 has joined #bitcoin-core-dev
Cory15 has quit [Ping timeout: 250 seconds]
bitdex has joined #bitcoin-core-dev
Cory99 has joined #bitcoin-core-dev
Cory84 has quit [Ping timeout: 250 seconds]
twistedline has quit []
twistedline has joined #bitcoin-core-dev
aleggg has quit [Ping timeout: 248 seconds]
w0xlt has joined #bitcoin-core-dev
w0xlt has quit [Ping timeout: 258 seconds]
Cory99 has quit [Quit: Client closed]
Cory99 has joined #bitcoin-core-dev
___nick___ has joined #bitcoin-core-dev
Cory63 has joined #bitcoin-core-dev
___nick___ has quit [Ping timeout: 248 seconds]
Cory61 has joined #bitcoin-core-dev
Cory99 has quit [Ping timeout: 250 seconds]
Cory15 has joined #bitcoin-core-dev
Cory85 has joined #bitcoin-core-dev
Cory63 has quit [Ping timeout: 250 seconds]
Cory61 has quit [Ping timeout: 250 seconds]
Cory15 has quit [Ping timeout: 250 seconds]
Cory37 has joined #bitcoin-core-dev
Cory85 has quit [Ping timeout: 250 seconds]
Cory67 has joined #bitcoin-core-dev
Cory28 has joined #bitcoin-core-dev
Cory37 has quit [Ping timeout: 250 seconds]
l0rinc has joined #bitcoin-core-dev
Cory67 has quit [Ping timeout: 250 seconds]
l0rinc_ has joined #bitcoin-core-dev
l0rinc has quit [Ping timeout: 256 seconds]
jonatack has quit [Ping timeout: 255 seconds]
Cory28 has quit [Quit: Client closed]
Cory28 has joined #bitcoin-core-dev
aleggg has joined #bitcoin-core-dev
cmirror has quit [Remote host closed the connection]
cmirror has joined #bitcoin-core-dev
robszarka has joined #bitcoin-core-dev
szarka has quit [Ping timeout: 256 seconds]
Cory28 has quit [Quit: Client closed]
Cory28 has joined #bitcoin-core-dev
l0rinc_ has quit [Quit: l0rinc_]
PaperSword has joined #bitcoin-core-dev
kevkevin has quit [Remote host closed the connection]
SpellChecker has quit [Remote host closed the connection]
kevkevin has quit [Ping timeout: 248 seconds]
kevkevin has joined #bitcoin-core-dev
kevkevin has quit [Ping timeout: 248 seconds]
sbddesign has joined #bitcoin-core-dev
sbddesign has quit [Ping timeout: 256 seconds]
<sipa>
dzxzg: the point of txns_randomized is used is because it only costs O(1) to maintain, and has perfect memory locality
<sipa>
using the live ancestor feerate-sorted index means 1000s of pointer indirections, and the cache locality effects that has
<sipa>
i haven't benchmarked it lately, it's possible this doesn't matter in practice
<sipa>
if need be, the effect can perhaps be replicated by having a number of txns_randomized vectors, e.g. one for everything that's in the top 1 MvB of the mempool, one for everything in the top 3 MvB, one for everything else?
<sipa>
(post-clustermempool this should not be too hard to do)
f321x has joined #bitcoin-core-dev
Saturday7 has joined #bitcoin-core-dev
Talkless has joined #bitcoin-core-dev
Saturday7 has quit [Read error: Connection reset by peer]
Saturday7 has joined #bitcoin-core-dev
Saturday7 has quit [Read error: Connection reset by peer]
Saturday7 has joined #bitcoin-core-dev
vincenzopalazzo has quit [Remote host closed the connection]
kevkevin has joined #bitcoin-core-dev
Saturday7 has quit [Read error: Connection reset by peer]
Saturday7 has quit [Read error: Connection reset by peer]
Saturday7 has joined #bitcoin-core-dev
kevkevin has joined #bitcoin-core-dev
<_aj_>
sipa: you're still deref'ing the CTransactionRef to get to the wtxid, which doesn't seem that great for cache locality? having a vector of wtxids directly might be worth looking at?
___nick___ has joined #bitcoin-core-dev
___nick___ has quit [Client Quit]
Saturday7 has quit [Ping timeout: 258 seconds]
___nick___ has joined #bitcoin-core-dev
w0xlt has joined #bitcoin-core-dev
w0xlt has quit [Ping timeout: 248 seconds]
koolazer has quit [Server closed connection]
koolazer has joined #bitcoin-core-dev
f321x has quit [Quit: f321x]
___nick___ has quit [Ping timeout: 260 seconds]
kevkevin has quit [Ping timeout: 260 seconds]
iSiUp has quit [Quit: WeeChat 4.7.1]
sbddesign has joined #bitcoin-core-dev
sbddesign has quit [Ping timeout: 248 seconds]
Lightsword has quit [Server closed connection]
Lightsword has joined #bitcoin-core-dev
w0xlt has joined #bitcoin-core-dev
w0xlt has quit [Ping timeout: 256 seconds]
iSiUp has joined #bitcoin-core-dev
kevkevin has joined #bitcoin-core-dev
theStack has joined #bitcoin-core-dev
<sipa>
_aj_: huh! i was sure it contained the wtxids directly!
<sipa>
it was changed fairly recently it seems, in #29752