AF:
NF:0
PS:10
SRH:1
SFN:
DSR:
MID:
CFG:
PT:0
S:andy.sharp@onstor.com
RQ:
SSV:onstor-exch02.onstor.net
NSV:
SSH:
R:<larry.scheer@onstor.com>
MAID:1
X-Sylpheed-Privacy-System:
X-Sylpheed-Sign:0
SCF:#mh/Mailbox/sent
RMID:#imap/andys@onstor.net@onstor-exch02.onstor.net/INBOX	0	BB375AF679D4A34E9CA8DFA650E2B04E056C988D@onstor-exch02.onstor.net
X-Sylpheed-End-Special-Headers: 1
Date: Mon, 20 Oct 2008 13:56:14 -0700
From: Andrew Sharp <andy.sharp@onstor.com>
To: "Larry Scheer" <larry.scheer@onstor.com>
Subject: Re: Please review 30970 tool chain rules Makefile
Message-ID: <20081020135614.42e0ba0b@ripper.onstor.net>
In-Reply-To: <BB375AF679D4A34E9CA8DFA650E2B04E056C988D@onstor-exch02.onstor.net>
References: <20081017191226.74a38b40@ripper.onstor.net>
	<BB375AF679D4A34E9CA8DFA650E2B04E056C988D@onstor-exch02.onstor.net>
Organization: Onstor
X-Mailer: Sylpheed-Claws 2.6.0 (GTK+ 2.8.20; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit

ok

On Mon, 20 Oct 2008 12:08:55 -0700 "Larry Scheer"
<larry.scheer@onstor.com> wrote:

> Response in-line below.
> 
> -----Original Message-----
> From: Andy Sharp 
> Sent: Friday, October 17, 2008 7:12 PM
> To: Larry Scheer
> Subject: Re: Please review 30970 tool chain rules Makefile
> 
> On Fri, 17 Oct 2008 18:51:27 -0700 "Larry Scheer"
> <larry.scheer@onstor.com> wrote:
> 
> > Andy,
> >    Please review. This is in my Mightydog workspace:
> > /homes/larrys/src/dev/nfx-tree
> > 
> > Change 30970 by larrys@larrys on 2008/10/17 18:35:01 *pending*
> > 
> >         Fixed some annoying problems and cleaned up installation
> > messages.
> >         -  Fixed failure when installing tool chain for the very
> > first time
> >            on a new box  /usr/mipsel-linux-gnu needed to be created
> > first
> >         -  Some packages such as tzdata were being re-installed
> > every time
> >            make install_toolchain was run due to rpm quirks --
> > fixed.
> >         -  The current-pkg-vers didn't show which compiler version
> > were installed.
> >         -  The remove_toolchain target wasn't removing all tool
> > chain packages
> >            that were installed.
> >         Reviewed by: AndyS
> > 
> > Affected files ...
> > 
> > ... //depot/dev/nfx-tree/Tools/tool-chain-pkg.mk#4 edit
> 
> 
> 
> 
> 
> = Change 30970 by larrys@larrys on 2008/10/17 18:35:01 *pending*
> = 
> = 	Fixed some annoying problems and cleaned up installation
> messages.
> = 	-  Fixed failure when installing tool chain for the very
> first time
> = 	   on a new box  /usr/mipsel-linux-gnu needed to be created
> first
> = 	-  Some packages such as tzdata were being re-installed
> every time
> = 	   make install_toolchain was run due to rpm quirks --
> fixed. = 	-  The current-pkg-vers didn't show which compiler
> version were installed.
> = 	-  The remove_toolchain target wasn't removing all tool
> chain packages
> = 	   that were installed.
> 
> It wasn't removing any packages, was it?
> LS -- Yes, it did and still does remove all packages installed in
> /usr/mipsel-linux-gnu. Now it also removes the support packages (i. e.
> quilt)
> 
> = 	Reviewed by: AndyS
> = 
> 
> nfx-tree/Tools/tool-chain-pkg.mk
> 
>      line 207, nodpes?
> LS -- Fixed
> 
>      line 212-217, it's looks like ROOTCMD is being used randomly and
>      not everywhere it should be?  certainly pkg_verify is used in
>      many places w/o ROOTCMD, but remove_pkg_verify is always used
>      with ROOTCMD everywhere else.
> LS -- Good catch. There was one instance of ROOTCMD before
> remove_pkg_ver 	that should have been removed because I changed
> remove_pkg_ver to 	use ROOTCMD. I only use ROOTCMD before
> pkg_verify when I am testing 	to see if a support package is
> installed. Maybe that is just paranoia.
> 	Otherwise the use of ROOTCMD is deliberate and necessary, not
> random.
> 	
>      line 293, the package itself should create this directory when
>      the package is installed, not some makefile, should it?
> LS -- This was needed because make install_toolchain would fail if it
> was 	being run for the very first time. I assumed the rpm
> process would 	create these directories but for some reason,
> perhaps due to rpm or 	packaging bugs the whole process runs
> better when I create these directories first.
> 
