unpin the yamux dependency

[Stebalien]

This has been pinned for a while under suspicion that it was causing graphsync stalls. However, the libp2p team has been unable to reliably reproduce the issue and believes that it's likely a graphsync issue.

Leaving it "pinned" indefinitely isn't going to fix the issue, so I'm filing this PR hoping we can find a solution.

My latest theory is that graphsync is deadlocking somewhere when sending a message blocks on the other side reading the message. However, to confirm this, we'd need goroutine dumps from both sides of a stall showing that:

1. One side is writing on the yamux stream.
2. The other side is reading.
3. It's actually stalled.

This could also be done with sufficient logging.

Libp2p issue: libp2p/go-yamux#61

💣 👈 👇 👉 ❓

[Stebalien]

cc @marten-seemann & @BigLep

[codecov]

Codecov Report

Merging #7532 (3b2e8c4) into master (98ff1c4) will decrease coverage by 0.11%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #7532      +/-   ##
==========================================
- Coverage   39.96%   39.85%   -0.12%     
==========================================
  Files         632      632              
  Lines       66926    66926              
==========================================
- Hits        26750    26675      -75     
- Misses      35560    35625      +65     
- Partials     4616     4626      +10     

Impacted Files	Coverage Δ
markets/loggers/loggers.go	89.28% <0.00%> (-10.72%)	⬇️
chain/stmgr/call.go	67.87% <0.00%> (-7.28%)	⬇️
chain/events/observer.go	71.64% <0.00%> (-6.72%)	⬇️
extern/sector-storage/manager_calltracker.go	57.70% <0.00%> (-4.85%)	⬇️
chain/actors/builtin/paych/paych.go	19.31% <0.00%> (-4.55%)	⬇️
chain/sub/incoming.go	54.43% <0.00%> (-3.38%)	⬇️
miner/miner.go	55.29% <0.00%> (-2.65%)	⬇️
chain/exchange/client.go	52.96% <0.00%> (-1.70%)	⬇️
itests/kit/blockminer.go	93.65% <0.00%> (-1.59%)	⬇️
chain/store/store.go	63.00% <0.00%> (-1.34%)	⬇️
... and 11 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 98ff1c4...3b2e8c4. Read the comment docs.

[Stebalien]

It also came up that there was an underflow issue in the graphsync allocation budgeting system that could cause stalls. Given that the repro for this always involved graphsync, I'm finding it more and more likely that changing how yamux receive buffers work triggered an underlying issue in graphsync.

[raulk]

That was probably the case, @Stebalien. Just to play it safe, can we ask Estuary (or some other very active deal maker) to upgrade to the latest yamux before we merge this in?

[Stebalien]

Once we figure out the stalls they're experiencing... yes.