You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
There is a longstanding bug with the ethons library such that select() intermittently returns 22 on select on FreeBSD, I have just tracked this down to the type definition in ffi being incorrect for (at least) time_t and suseconds_t in ffi you have them as ints.. However a printf of sizeof(...) on FreeBSD 11-amd64 indicates they are 8 bytes long (and therefore int64).. I updated the specification to long in the files and was just able to solve the errno22 issue.
I am not sure at what point FreeBSD cut over to 64 bits for those structures (it was a LONG time ago though), how you would version differences in ABI formats within a platform, what is the correct value (long vs int64), what the difference is for the 2 time_t and suseconds_t types, or what other values there are incorrect; but I hope this could be quickly updated.
The text was updated successfully, but these errors were encountered:
There is a longstanding bug with the ethons library such that select() intermittently returns 22 on select on FreeBSD, I have just tracked this down to the type definition in ffi being incorrect for (at least) time_t and suseconds_t in ffi you have them as ints.. However a printf of sizeof(...) on FreeBSD 11-amd64 indicates they are 8 bytes long (and therefore int64).. I updated the specification to long in the files and was just able to solve the errno22 issue.
I am not sure at what point FreeBSD cut over to 64 bits for those structures (it was a LONG time ago though), how you would version differences in ABI formats within a platform, what is the correct value (long vs int64), what the difference is for the 2 time_t and suseconds_t types, or what other values there are incorrect; but I hope this could be quickly updated.
The text was updated successfully, but these errors were encountered: