Re: Implementing read and write locks on files

From: Andrei Voropaev (avorop_at_mail.ru)
Date: 10/10/05 

Date: Mon, 10 Oct 2005 08:02:23 +0000 (UTC)

In comp.os.linux.development.apps JosephKK <joseph_barrett@sbcglobal.net> wrote:
> Kasper Dupont wrote:
>
[...]
> I dug it up with info read. It will return 0 with EWOULDBLOCK or EAGAIN
> while a real eof would also set FEOF.

On which system? It's not true for Linux. 

-- 
Minds, like parachutes, function best when open

Relevant Pages

  • Re: Implementing read and write locks on files
    ... > Kasper Dupont wrote: ... > I dug it up with info read. ... It will return 0 with EWOULDBLOCK or EAGAIN ...
    (comp.unix.programmer)
  • Re: read with timeout ??
    ... Hash: SHA1 ... timeout returns with the error EWOULDBLOCK but readrecvsend ... and writeonly list EAGAIN in their ERRORS section. ... Carlos, lets not go off on the socket trip for me... ...
    (freebsd-hackers)
  • Re: read with timeout ??
    ... timeout returns with the error EWOULDBLOCK but readrecvsend ... and writeonly list EAGAIN in their ERRORS section. ... recvlists both EAGAIN and EWOULDBLOCK. ... standards conformance and porting issues. ...
    (freebsd-hackers)
  • EAGAIN vs EWOULDBLOCK
    ... Is there any subtle difference between EAGAIN and EWOULDBLOCK? ... The socket is marked non-blocking and the requested operation would ... HP-UX so non-standard that it's not worth worrying about at this stage? ...
    (comp.unix.programmer)
  • Re: EAGAIN vs EWOULDBLOCK
    ... >>Is there any subtle difference between EAGAIN and EWOULDBLOCK? ... Linux is non-standard in many ways, so don't quote a header of a Linux implementation ... AFAIK the opengroup does not require EAGAIN or EWOULDBLOCK to have ...
    (comp.unix.programmer)