coreutils
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: coreutils-8.10 released [stable]


From: Dmitry V. Levin
Subject: Re: coreutils-8.10 released [stable]
Date: Sat, 5 Feb 2011 03:28:01 +0300

On Sat, Feb 05, 2011 at 03:11:12AM +0300, Dmitry V. Levin wrote:
> On Fri, Feb 04, 2011 at 08:20:39PM +0100, Jim Meyering wrote:
> > This is to announce coreutils-8.10, a stable release.
> > 
> > There have been some minor bug fixes, along with two new features.  The
> > join feature is enabled via a new option, "-o auto".  The cp feature makes
> > copying sparse files much more efficient on several common file systems.
> > It takes advantage of a feature that was introduced in linux-2.6.27.
> > The improvement affects the default code path, so if you're looking
> > for risk potential, this is it.  It uses the feature if available,
> > and otherwise resorts to using the old, less-efficient copying code.
> 
> tests/cp/fiemap-perf fails with EFBIG on tmpfs:
> $ truncate -s1T f
> truncate: failed to truncate `f' at 1099511627776 bytes: File too large
> 
> Reducing the size from 1T to 256G makes the test pass.

Actually reducing the size just makes the truncate test pass.
Unfortunately, fiemap_capable_() fails, too:

+ df -T -t btrfs -t xfs -t ext4 -t ocfs2 -t gfs2 .
df: Warning: cannot read table of mounted file systems
Filesystem    Type   1K-blocks      Used Available Use% Mounted on
-                -    10485760    793504   9692256   8% /

Looks like df agrees on the list of allowed fs types when it fails to find
out the fs type of the given filesystem, e.g. when /etc/mtab is a symlink
to /proc/mounts where given filesystem is not listed.

As result, the test fails on a non-capable filesystem.


-- 
ldv

Attachment: pgpMTtnY2jQab.pgp
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]