-
-
Notifications
You must be signed in to change notification settings - Fork 249
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
infinite loop trying to insert a value into a hash set #250
Comments
Hum, unfortunately there were a couple of small changes in the |
the crash existed within 1.3.5 as well. I guess I just want assurances that whatever was causing the crash was fixed between then and now. the crash seems to behave identically with both versions, so I don't think its an issue with the load/dump functionality, seems more like 1.3.5 saved a broken hash map? (edit: ok the dumped array here was created with current version, let me try to dump one from 1.3.5) |
yeah its the same bug regardless, looks pretty much the same when stepping through in the debugger no matter which versions I'm using
|
I don't know what datadump's data come from, so I don't see how this demonstrates a phmap problem (unless I'm missing something, if so please let me know). If you can provide an example of populating a hash map normally, dumping it, and being unable to load it, I'll certainly look into it. |
datadump was a flat_hash_set that was saved with phmap_dump in version 1.3.5. I have no way to reconstruct the sequence of adds/removes that created that initial hash set, this is the only way I can reproduce the bug. I just converted it to bytes so I could store it in the reproduction like that. The hash set was only ever created by adding and removing int64_t's to the set (single threaded), and eventually one user reported a crash here, and I converted their saved hash set into that byte sequence for the reproduction. I dont think the dump/load is a relevant part of the bug here, its only there cause I cant reconstruct this set otherwise |
You are probably hitting this issue which us now fixed (since 1.3.12) |
ok I think this is actually what the issue was, and it looks like it was fixed at some point between 1.3.5 and current version,
|
Awesome, give me a sec. |
following reproduction infinite loops in _find_key when trying to insert "349" into the hash set
the data was saved with phmap_dump in an older version (I believe 1.35?). I dont have the exact sequence of inserts/removes that created this data unfortunately. All I know is this crashes when trying to insert 349. I updated to the current revision and the crash still exists.
infinite loop happens with MSVC when compiling in 64 bit mode
The text was updated successfully, but these errors were encountered: