Bug 58824

Summary:	write(2) can block despite fd being non-blocking
Product:	[Retired] Red Hat Linux	Reporter:	peter.benie
Component:	kernel	Assignee:	Arjan van de Ven <arjanv>
Status:	CLOSED CURRENTRELEASE	QA Contact:	Brian Brock <bbrock>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	6.2	CC:	alan
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
URL:	http://www.chiark.greenend.org.uk/~peterb/linux/nonblock/
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2004-09-30 15:39:21 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description peter.benie 2002-01-25 10:13:42 UTC

Description of problem:
If there is a blocking write to a tty or pipe that is pending completion, then
*all* writes to the same tty or pipe will block. This is true even for fds that
are supposed to be non-blocking!

How reproducible:
For pipes and ttys  on 2.2 kernels: Always
For ttys on 2.4 and 2.5 kernels: Always

Steps to Reproduce:
Run the test programs from the above URL

Additional info:

The above URL contains a patch to fix this issue for 2.2 and 2.4 kernels. The
2.4 patch applies cleanly and works on 2.5.

Comment 1 Arjan van de Ven 2002-01-25 10:18:36 UTC

Patch looks fine except that the return value might need to be
-EWOULDBLOCK; I'll go hunt the appropriate standards...

Comment 2 peter.benie 2002-01-25 11:03:29 UTC

SUSv2 only lists EAGAIN for write();  network operations (send(2) etc)  may
return either EAGAIN or EWOULDBLOCK. 

SUSv2 also states that EWOULDBLOCK may be the same value as EAGAIN, and this is
the case under Linux for all architectures except PA-RISC, where you will find
the following comment in errno.h:
#define EWOULDBLOCK     246     /* Operation would block (Linux returns EAGAIN)
*/

Comment 3 Alan Cox 2002-01-25 13:42:24 UTC

The tty one is fixed in 2.4.18-ac and  pushed on to Marcelo

Comment 4 Bugzilla owner 2004-09-30 15:39:21 UTC

Thanks for the bug report. However, Red Hat no longer maintains this version of
the product. Please upgrade to the latest version and open a new bug if the problem
persists.

The Fedora Legacy project (http://fedoralegacy.org/) maintains some older releases, 
and if you believe this bug is interesting to them, please report the problem in
the bug tracker at: http://bugzilla.fedora.us/