pcp
[Top] [All Lists]

Re: [pcp] question on rpm builds

To: Nathan Scott <nathans@xxxxxxxxxx>, Ken McDonell <kenj@xxxxxxxxxxxxxxxx>
Subject: Re: [pcp] question on rpm builds
From: Mark Goodwin <mgoodwin@xxxxxxxxxx>
Date: Thu, 18 Feb 2016 15:55:05 +1100
Cc: PCP <pcp@xxxxxxxxxxx>
Delivered-to: pcp@xxxxxxxxxxx
In-reply-to: <1436029232.21802619.1455770764295.JavaMail.zimbra@xxxxxxxxxx>
References: <56C548F7.5020900@xxxxxxxxxxxxxxxx> <1436029232.21802619.1455770764295.JavaMail.zimbra@xxxxxxxxxx>
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:38.0) Gecko/20100101 Thunderbird/38.5.0
On 02/18/2016 03:46 PM, Nathan Scott wrote:


----- Original Message -----
In the process of debugging something else, I noticed this coming out of the
rpm builds ...

     Explicit %attr() mode not applicaple to symlink:
     
/home/kenj/src/pcp/pcp-3.11.1/BUILDROOT/pcp-3.11.1-8.x86_64/usr/lib64/libpcp.so

and similar lines that appear to be for every symlink in the PCP packages.

Is this expected?

Yeah, its been generating that warning for years.  Its most likely from:
$1 == "l" { print "%attr(0777,root,root)", $3 >> f }'
... in pcp.spec.in?  It may have once been accepted by an old version of
rpm/rpmbuild.

Is it something to worry about / fix?

Nope, but we could fix if its bothering.  Theres also the make -j1 warning,
and a handful of rpm "warning: File listed twice" warnings we could also go
fix - they just never seem to percolate up the to-do list here.


yep it's benign .. ever since we did the original rpm packaging.
The particular issue you've reported is because symlinks don't
have modes, so specifying one with the %attr generates a warning.

Cheers

<Prev in Thread] Current Thread [Next in Thread>