Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 11 Jun 2010 10:28:26 -0500
From:      Nathan Whitehorn <nwhitehorn@freebsd.org>
To:        Ken Smith <kensmith@buffalo.edu>
Cc:        svn-src-stable@freebsd.org, svn-src-all@freebsd.org, src-committers@freebsd.org, svn-src-stable-8@freebsd.org, Martin Matuska <mm@FreeBSD.org>
Subject:   Re: svn commit: r208258 - in stable/8: contrib/top etc/mtree lib lib/liblzma lib/libusb share/mk usr.bin usr.bin/less usr.bin/lzmainfo usr.bin/xz usr.bin/xzdec
Message-ID:  <4C12561A.4060707@freebsd.org>
In-Reply-To: <1276268668.89354.38.camel@bauer.cse.buffalo.edu>
References:  <201005180959.o4I9xAsu012265@svn.freebsd.org> <1276268668.89354.38.camel@bauer.cse.buffalo.edu>

next in thread | previous in thread | raw e-mail | index | archive | help
On 06/11/10 10:04, Ken Smith wrote:
> On Tue, 2010-05-18 at 09:59 +0000, Martin Matuska wrote:
>    
>> Author: mm
>> Date: Tue May 18 09:59:09 2010
>> New Revision: 208258
>> URL: http://svn.freebsd.org/changeset/base/208258
>>
>> Log:
>>    MFC r207842, r207844, r208099:
>>
>>    MFC r207842:
>>    Import of liblzma, xz, xzdec, lzmainfo from vendor branch
>>    Add support for xz and lzma to lesspipe.sh (xzless, lzless)
>>
>>    MFC r207844:
>>    Add two public headers missing in r207842
>>    Adjust CFLAGS for lzmainfo, xz, xzdec
>>
>>    MFC r208099:
>>    Add versioned symbols to liblzma
>>    Use default SHLIB_MAJOR.
>>
>>    Approved by:	delphij (mentor)
>>
>>      
> xz on sparc64 appears to be broken.  Some time between 8.1-BETA1
> and now the libpng (required for release builds because it's part
> of the "docproj" port set) shifted from fetching a .bz2 file to
> fetching a .xz file.  Uncompressing the file works fine on amd64
> and i386 architectures but fails on sparc64.
>
> I haven't had time to dig into why it's busted but I figured I
> should at least let others know about it right away.  Is it
> possible some knob related to big endian versus little endian
> isn't being handled quite right?
>    

This is at least also true on powerpc (both 32 and 64 bit), so it is 
almost certainly an endian issue.
-Nathan




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4C12561A.4060707>