Bug 229475

Summary:	File::writable? gives incorrect result if process changes euid/egid
Product:	[Fedora] Fedora	Reporter:	Akira TAGOH <tagoh>
Component:	ruby	Assignee:	Akira TAGOH <tagoh>
Status:	CLOSED RAWHIDE	QA Contact:	Bill Huang <bhuang>
Severity:	medium	Docs Contact:
Priority:	medium
Version:	rawhide	CC:	eng-i18n-bugs
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2007-03-30 11:28:41 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:
Bug Depends On:	222177
Bug Blocks:

Description Akira TAGOH 2007-02-21 13:12:52 UTC

+++ This bug was initially created as a clone of Bug #222177 +++

When a ruby script running as root changes euid/egid, the results of
File::writable?(f) and File::stat(f).writable? will differ because one uses
glibc's eaccess while the other doesn't (and File::writable? is wrong). The
attached program demonstrates the problem.

This bug is ultimately caused by a bug in glibc's eaccess
(http://sourceware.org/bugzilla/show_bug.cgi?id=3846) but it's a separate
regression in ruby-1.8.5 (compared to earlier versions of the ruby RPM) since
the ruby-1.8.5 rpm's now use eaccess where older ones used ruby's own code to
simulate eaccess.

A simple fix would be to build ruby w/o HAVE_EACCESS as it has been build in the
past until the glibc bug is fixed.

-- Additional comment from dlutter on 2007-01-10 14:09 EST --
Created an attachment (id=145276)
Ruby script demonstrating the problem (run as root)

Comment 1 Akira TAGOH 2007-03-30 11:28:41 UTC

seems like got fixed in rawhide. it works fine now.