Segmentation fault on Linux monerod

[celavek]

I get a segmentation fault with the latest release - Monero 'Helium Hydra' (v0.11.0.0-release) on
Debian Linux. This is the first time I try to synchronize the blockchain and create a wallet. With the current state of the blockchain that I have I consitently get the crash. The logs before the crash look a bit strange to me with confusing output:

2017-09-12 20:46:16.601 [P2P7] INFO global src/cryptonote_protocol/cryptonote_protocol_handler.inl:305 [75.88.131.213:57070 INC] Sync data returned a new top block candidate: 1288623 -> 1322390 [Your node is 33767 blocks (46 days) behind]
SYNCHRONIZATION started
2017-09-12 20:46:16.602 [P2P7] INFO global src/cryptonote_protocol/cryptonote_protocol_handler.inl:305 [37.79.255.43:41287 INC] Sync data returned a new top block candidate: 1288623 -> 1397703 [Your node is 109080 blocks (151 days) behind]
SYNCHRONIZATION started

first it's 46 days behind then the next line reports 151 days behind?

From the core dump it seems to be crashing during some boost::asio operation.
The log is attached but the dump is too big to attach

monerod_crash.txt

the stack trace:

monerod_stacktrace.txt

I will try to reinitialize the whole thing to see if I still get the crash.

[moneromooo-monero]

Hard to see where it is crashing. Try a debug build, and run with --log-level 1,blockchain:3.

[dEBRUYNE-1]

For visibility, I noticed two other people reporting this:

No stack trace (yet).

I have the latest version downloaded. I resynced from scratch, but got stuck on block 1009840. Closed and re-opened it, and now I get "Segmentation fault (core dumped)".

I'm using Ubuntu 16.04.

Source: https://www.reddit.com/r/Monero/comments/707elm/i_cant_sync_monerod/

---

Stack trace (old bins): https://paste.fedoraproject.org/paste/WhTPoZoEQNPb2pdLMGABCw

Stack trace (new bins): https://paste.fedoraproject.org/paste/15-XU-Qa50bLaGUFpDWGRg

Ubuntu 14.04 64bit, monero-v0.11.0.0

It was dumping the core on first run, so I deleted blockchain and started again. Seemed to stall at 92% so at that point I stopped daemon and upon starting again back to core dumped.

Source: https://www.reddit.com/r/Monero/comments/6yprar/mandatory_upgrade_monero_01100_helium_hydra/dn0dr7f/?context=3

[celavek]

@moneromooo-monero I think the stack trace I attached gives a hint where the crash might be.
I did not build it myself. The binary is the "official release". I will try a debug build and see how that goes.

I initialized the blockchain in a new location from scratch and this time it did not crash.

[moneromooo-monero]

Probably fixed by #2492, as reported in the bug referenced just above.

[moneromooo-monero]

+resolved

[celavek]

I've seen your comment on the related thread ... and from the commit the fix seems to be related to the call to __memcmp_sse4_1 () which my backtrace also contained ...

[Tahutipai]

I am experiencing this issue with binaries downloaded from getmonero.org as of 13/oct/2017. Can anyone please advise how to resolve?

[zzium]

Experiencing the exact same issue here. Running same Ubuntu version (17.04) - Also got stuck at block 1288623

Edit: Also running the 0.11.0.0 binaries from getmonero.org downloaded only few days ago

[palexande]

Getting the same behaviour. This is after a hard shutdown of the node. OS: Linux Mint.

Update: Was able to fix with "monerod --db-salvage"

[muncherelli]

Also getting the same error at block 1288623.

v0.11.0.0 binaries from getmonero.org, running on a single core Ubuntu 16.04 VPS with 1GB RAM

**********************************************************************
You are now synchronized with the network. You may now start monero-wallet-cli.

Use the "help" command to see the list of available commands.
**********************************************************************
2017-11-16 21:08:31.004 [P2P6]  INFO    global  src/cryptonote_protocol/cryptonote_protocol_handler.inl:305     [84.114.235.113:18080 OUT] Sync data returned a new top block candidate: 1288623 -> 1444455 [Your node is 155832 blocks (216 days) behind]
SYNCHRONIZATION started
Segmentation fault (core dumped)

[juliavi]

Same error here. Tried restarting multiple times, and starting from scratch multiple times.
--db-salvage didn't help.
I've built the latest package from github, so the fix should have applied.

[andrewvaca]

Tahnks @palexande, I was getting Segmentation fault: 11 after a hard shutdown of the node on macOS, but "monerod --db-salvage" did the trick and it's working again.