POSIX close(2) is broken

Re:POSIX close(2) is broken by nknight on Saturday December 17, 2011 @04:11AM (#3363868)

This gets even stranger than you might think. Consider fclose()[1], which says, after the ordinary disclaimer that POSIX defers to the C standard, "The fclose() function shall perform the equivalent of a close() on the file descriptor that is associated with the stream pointed to by stream.".

And as part of the C standard, we have "After the call to fclose(), any use of stream results in undefined behavior.".

Uh-oh. So does this imply that fclose(), at least, has to do the job right, or does it mean it inherits the close() flaw, and leaves no defined way to try and clean up the mess?

[1] http://pubs.opengroup.org/onlinepubs/9699919799/functions/fc...

[ Parent | on HN ]

Re:POSIX close(2) is broken by cperciva on Saturday December 17, 2011 @04:14AM (#3363872)

Given that the standard states

    [EINTR]
        The fclose() function was interrupted by a signal.

I think it's safe to say that fclose is allowed to fail with no defined way to clean up the mess.

[ Parent | on HN ]

Re:POSIX close(2) is broken by nknight on Saturday December 17, 2011 @04:18AM (#3363875)

Yeah, I was just looking at that after I posted my initial comment, and I think I agree with you. That's the only reading I can see that doesn't create more questions than it answers.

I think in this case, Occam's razor says "this part of the standard is not fully thought through".

[ Parent | on HN ]

Re:POSIX close(2) is broken by cperciva on Saturday December 17, 2011 @04:42AM (#3363910)

this part of the standard is not fully thought through

I agree... and this terrifies me, given that EINTR is mentioned in the man page for close(2) back in BSD Net/2, over 20 years ago.

[ Parent | on HN ]

Re:POSIX close(2) is broken by ajasmin on Saturday December 17, 2011 @08:36AM (#3364202)

this terrifies me

Now think about all the libraries and application that end up calling close() and fclose().

Add a few abstraction layers and good luck figuring out if closing a file is safe in a multithreaded Java app.

[ Parent | on HN ]

Re:POSIX close(2) is broken by wingo on Sunday December 18, 2011 @07:07AM (#3366482)

It is somewhat unrelated, but you might enjoy Jim Meyering's "Goodbye, world: the perils of relying on output streams in C":

https://www.irill.org/events/ghm-gnu-hackers-meeting/videos/...

[ Parent | on HN ]