Bug 48641 (splice_EAGAIN) - [TRIVIAL]splice can return EAGAIN
Summary: [TRIVIAL]splice can return EAGAIN
Status: RESOLVED CODE_FIX
Alias: splice_EAGAIN
Product: Documentation
Classification: Unclassified
Component: man-pages (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: documentation_man-pages@kernel-bugs.osdl.org
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2012-10-09 10:28 UTC by Daurnimator
Modified: 2014-06-18 14:03 UTC (History)
3 users (show)

See Also:
Kernel Version: n/a
Subsystem:
Regression: No
Bisected commit-id:


Attachments

Description Daurnimator 2012-10-09 10:28:06 UTC
`splice (2)` can return EAGAIN, this is not mentioned in the man pages.

Example strace output:

    pipe([4, 5])                            = 0
    splice(4, NULL, 1, NULL, 5, SPLICE_F_NONBLOCK) = -1 EAGAIN (Resource temporarily unavailable)


I run Arch Linux, the date at the bottom of the man page is: 2012-05-04
Comment 1 sakshi.april5 2014-05-28 04:51:53 UTC
Hello,

I am a beginner and would like to work on this bug.
Comment 2 Michael Kerrisk 2014-05-28 13:10:36 UTC
(In reply to sakshi.april5 from comment #1)
> Hello,
> 
> I am a beginner and would like to work on this bug.

It's still open.
Comment 3 sakshi.april5 2014-05-28 13:18:15 UTC
Yes, I would like to work on this bug. Since this is my first time fixing a bug in Linux Kernel it would be nice if someone can help me out.
Comment 4 Michael Kerrisk 2014-05-28 13:27:16 UTC
This is a bug in the man page, not in the kernel...
Comment 5 sakshi.april5 2014-06-02 07:08:01 UTC
I have found these files where the changes need to be made:
man2/splice.2
man2/vmsplice.2
man2/tee.2

I suppose change in any one file would reflect changes in the other?
How exactly do we need to make the chnage to get the above mentioned output?
Comment 6 Michael Kerrisk 2014-06-02 07:13:30 UTC
(In reply to sakshi.april5 from comment #5)
> I have found these files where the changes need to be made:
> man2/splice.2
> man2/vmsplice.2
> man2/tee.2

Yes, it looks like all three pages need the fix.

> I suppose change in any one file would reflect changes in the other?

Yes, that sounds reasonable.

> How exactly do we need to make the chnage to get the above mentioned output?

Not sure that I your question. Clone the man-pages git repo [1], and then send a patch.

[1] https://www.kernel.org/doc/man-pages/download.html
Comment 7 sakshi.april5 2014-06-02 09:01:57 UTC
What I meant to ask is - do we have to replace every occurence of splice (2) with splice (2) can return EAGAIN in all the three files?
Is that the only change?
Comment 8 sakshi.april5 2014-06-06 05:42:11 UTC
What I meant to ask is - do we have to replace every occurence of splice (2) with splice (2) can return EAGAIN in all the three files?
Is that the only change?
Comment 9 Michael Kerrisk 2014-06-18 14:03:20 UTC
I've added the following text to each of the three man pages splice.2, vmsplice.2, tee.2:
[[
.TP
.B EAGAIN
.B SPLICE_F_NONBLOCK
was specified in
.IR flags ,
and the operation would block.
]]

Note You need to log in before you can comment on or make changes to this bug.