Message164139
| Author |
neologix |
| Recipients |
giampaolo.rodola, nadeem.vawda, neologix, pitrou, rosslagerwall |
| Date |
2012年06月27日.10:05:07 |
| SpamBayes Score |
-1.0 |
| Marked as misclassified |
Yes |
| Message-id |
<CAH_1eM2PPOvawJPsESv4TDRnpF1XniDd7WGT4gdL8dXgx44M-Q@mail.gmail.com> |
| In-reply-to |
<1340780430.33.0.138506780382.issue13876@psf.upfronthosting.co.za> |
| Content |
Looks like a kernel bug.
The only thing "non-standard" the test does is let the client call
close() before the server has called accept(), but this shouldn't
result in ENOTCONN - and doesn't for TCP sockets. |
|