| To: | qmail@list.cr.yp.to |
|---|---|
| Subject: | Re: Is saferead not really safe on Solaris 8? |
| From: | Henning Brauer <lists-qmail@bsws.de> |
| Date: | Sat, 25 Nov 2006 21:22:35 +0100 |
| Delivered-to: | sp-com-lists@consult.net |
| Delivered-to: | gmail-qmail@securepoint.com |
| Delivered-to: | sp.com.list@gmail.com |
| Delivered-to: | mailing list qmail@list.cr.yp.to |
| In-reply-to: | <20061124.191156.02469879.hanche@math.ntnu.no> |
| Mail-followup-to: | qmail@list.cr.yp.to |
| Mailing-list: | contact qmail-help@list.cr.yp.to; run by ezmlm |
| References: | <20061124.191156.02469879.hanche@math.ntnu.no> |
| User-agent: | Mutt/1.5.12-2006-07-14 |
* Harald Hanche-Olsen <hanche@math.ntnu.no> [2006-11-24 19:14]: > Now this should not be possible, because qmail-smtpd uses saferead() on > all network reads, and saferead first calls select() and then read() > only if select() indicates that there is data to be read. Surely, if > select() said there is data, then read() cannot hang? AFAIR there is no guarantee that read() cannot block after select() anounced readability. The only way to make sure read() won't block is using nonblocking sockets. however, I have no case at hand where select/poll would indicate readability and a read() thereafter would block. -- Henning Brauer, hb@bsws.de, henning@openbsd.org BS Web Services, http://bsws.de Full-Service ISP - Secure Hosting, Mail and DNS Services Dedicated Servers, Rootservers, Application Hosting - Hamburg & Amsterdam |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: Question about Qmail relating to MX retry processing, Harald Hanche-Olsen |
|---|---|
| Next by Date: | Re: Is saferead not really safe on Solaris 8?, Harald Hanche-Olsen |
| Previous by Thread: | Re: Is saferead not really safe on Solaris 8?, Jorge Valdes |
| Next by Thread: | Re: Is saferead not really safe on Solaris 8?, Harald Hanche-Olsen |
| Indexes: | [Date] [Thread] [Top] [All Lists] |