From diy-linux-dev@diy-linux.org  Sun Jan 14 01:20:32 2007
From: diy-linux-dev@diy-linux.org (Jon Grosshart)
Date: Sun, 14 Jan 2007 00:20:32 +0000
Subject: binutils-2.17 and -O3
Message-ID: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>

------=_Part_79102_17037792.1168734032605
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Greets all. What better time for a new build than winter? So I dove in using
4.1.1/2.4/2.17

binutils Pass 2 of temptools crapped out on the make with:


gcc -DHAVE_CONFIG_H -I. -I../../binutils-2.17/gas -I. -D_GNU_SOURCE -I.
-I../../binutils-2.17/gas -I../bfd -I../../binutils-2.17/gas/config
-I../../binutils-2.17/gas/../include -I../../binutils-2.17/gas/..
-I../../binutils-2.17/gas/../bfd -I../../binutils-2.17/gas/../intl -I../intl
-DLOCALEDIR="\"/temptools/share/locale\""   -W -Wall -Wstrict-prototypes
-Wmissing-prototypes -Werror -march=i686 -mtune=i686 -O3 -pipe  -c
../../binutils-2.17/gas/subsegs.c
cc1: warnings being treated as errors
../../binutils-2.17/gas/read.c: In function 'equals':
../../binutils-2.17/gas/read.c:5013: warning: 'stopc' may be used
uninitialized in this function
make[4]: *** [read.o] Error 1
make[4]: *** Waiting for unfinished jobs....

Took me 5 seconds of google till I found something. Seems -O3 is the culprit
somehow. Passing -O2 fixes it. You can also keep -O3 and use
--disable-werror but the test suite supposedly bombs horribly doing that.
Checked all but the test suite so I can confirm the resolution.

Perhaps it's time to rethink my indiscriminate/rampant use of CFLAGS... Not
a show-stopper but thought I'd report it all the same.

Jon

------=_Part_79102_17037792.1168734032605
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Greets all. What better time for a new build than winter? So I dove in using 4.1.1/2.4/2.17<br><br>binutils Pass 2 of temptools crapped out on the make with:<br><br><br>gcc -DHAVE_CONFIG_H -I. -I../../binutils-2.17/gas -I. -D_GNU_SOURCE -I. -I../../binutils-
2.17/gas -I../bfd -I../../binutils-2.17/gas/config -I../../binutils-2.17/gas/../include -I../../binutils-2.17/gas/.. -I../../binutils-2.17/gas/../bfd -I../../binutils-2.17/gas/../intl -I../intl -DLOCALEDIR=&quot;\&quot;/temptools/share/locale\&quot;&quot;&nbsp;&nbsp; -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Werror -march=i686 -mtune=i686 -O3 -pipe&nbsp; -c ../../binutils-
2.17/gas/subsegs.c<br>cc1: warnings being treated as errors<br>../../binutils-2.17/gas/read.c: In function &#39;equals&#39;:<br>../../binutils-2.17/gas/read.c:5013: warning: &#39;stopc&#39; may be used uninitialized in this function
<br>make[4]: *** [read.o] Error 1<br>make[4]: *** Waiting for unfinished jobs....<br><br>Took me 5 seconds of google till I found something. Seems -O3 is the culprit somehow. Passing -O2 fixes it. You can also keep -O3 and use --disable-werror but the test suite supposedly bombs horribly doing that. Checked all but the test suite so I can confirm the resolution.
<br><br>Perhaps it&#39;s time to rethink my indiscriminate/rampant use of CFLAGS... Not a show-stopper but thought I&#39;d report it all the same.<br><br>Jon<br><br>

------=_Part_79102_17037792.1168734032605--


From diy-linux-dev@diy-linux.org  Sun Jan 14 01:53:26 2007
From: diy-linux-dev@diy-linux.org (Jon Grosshart)
Date: Sun, 14 Jan 2007 00:53:26 +0000
Subject: binutils-2.17 and -O3
In-Reply-To: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>
References: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>
Message-ID: <9cf823c60701131653h755d449dlbe2542666a7bc41c@mail.gmail.com>

------=_Part_79294_10311093.1168736006787
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

On 1/14/07, Jon Grosshart <jgrosshart@gmail.com> wrote:
>
> Greets all. What better time for a new build than winter? So I dove in
> using 4.1.1/2.4/2.17
>
> binutils Pass 2 of temptools crapped out on the make with:
>
>
> gcc -DHAVE_CONFIG_H -I. -I../../binutils-2.17/gas -I. -D_GNU_SOURCE -I.
> -I../../binutils- 2.17/gas -I../bfd -I../../binutils-2.17/gas/config
> -I../../binutils-2.17/gas/../include -I../../binutils-2.17/gas/..
> -I../../binutils-2.17/gas/../bfd -I../../binutils-2.17/gas/../intl
> -I../intl -DLOCALEDIR="\"/temptools/share/locale\""   -W -Wall
> -Wstrict-prototypes -Wmissing-prototypes -Werror -march=i686 -mtune=i686 -O3
> -pipe  -c ../../binutils- 2.17/gas/subsegs.c
> cc1: warnings being treated as errors
> ../../binutils-2.17/gas/read.c: In function 'equals':
> ../../binutils-2.17/gas/read.c:5013: warning: 'stopc' may be used
> uninitialized in this function
> make[4]: *** [read.o] Error 1
> make[4]: *** Waiting for unfinished jobs....
>
> Took me 5 seconds of google till I found something. Seems -O3 is the
> culprit somehow. Passing -O2 fixes it. You can also keep -O3 and use
> --disable-werror but the test suite supposedly bombs horribly doing that.
> Checked all but the test suite so I can confirm the resolution.
>
> Perhaps it's time to rethink my indiscriminate/rampant use of CFLAGS...
> Not a show-stopper but thought I'd report it all the same.
>
> Jon
>
>
Actually, upon further thought, my optimizations above are hardly over
zealous IMO. Seems the proper course of action for something like this would
be to handle it on a per compile basis I suppose. A tool chain with -O2
alone doesn't really matter since we are trashing it anyway, but I do like
-O3 for certain things like glibc.

Now I'm kinda curious about --disable-werror on 1st pass but not 2nd... :-)

------=_Part_79294_10311093.1168736006787
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

<br><div><span class="gmail_quote">On 1/14/07, <b class="gmail_sendername">Jon Grosshart</b> &lt;<a href="mailto:jgrosshart@gmail.com">jgrosshart@gmail.com</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Greets all. What better time for a new build than winter? So I dove in using 4.1.1/2.4/2.17<br><br>binutils Pass 2 of temptools crapped out on the make with:<br><br><br>gcc -DHAVE_CONFIG_H -I. -I../../binutils-2.17/gas -I. -D_GNU_SOURCE -I. -I../../binutils-
2.17/gas -I../bfd -I../../binutils-2.17/gas/config -I../../binutils-2.17/gas/../include -I../../binutils-2.17/gas/.. -I../../binutils-2.17/gas/../bfd -I../../binutils-2.17/gas/../intl -I../intl -DLOCALEDIR=&quot;\&quot;/temptools/share/locale\&quot;&quot;&nbsp;&nbsp; -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Werror -march=i686 -mtune=i686 -O3 -pipe&nbsp; -c ../../binutils-
2.17/gas/subsegs.c<br>cc1: warnings being treated as errors<br>../../binutils-2.17/gas/read.c: In function &#39;equals&#39;:<br>../../binutils-2.17/gas/read.c:5013: warning: &#39;stopc&#39; may be used uninitialized in this function
<br>make[4]: *** [read.o] Error 1<br>make[4]: *** Waiting for unfinished jobs....<br><br>Took me 5 seconds of google till I found something. Seems -O3 is the culprit somehow. Passing -O2 fixes it. You can also keep -O3 and use --disable-werror but the test suite supposedly bombs horribly doing that. Checked all but the test suite so I can confirm the resolution.
<br><br>Perhaps it&#39;s time to rethink my indiscriminate/rampant use of CFLAGS... Not a show-stopper but thought I&#39;d report it all the same.<br><span class="sg"><br>Jon<br><br>

</span></blockquote></div><br>Actually, upon further thought, my optimizations above are hardly over zealous IMO. Seems the proper course of action for something like this would be to handle it on a per compile basis I suppose. A tool chain with -O2 alone doesn&#39;t really matter since we are trashing it anyway, but I do like -O3 for certain things like glibc.
<br><br>Now I&#39;m kinda curious about --disable-werror on 1st pass but not 2nd... :-)<br>

------=_Part_79294_10311093.1168736006787--


From diy-linux-dev@diy-linux.org  Sun Jan 14 02:06:43 2007
From: diy-linux-dev@diy-linux.org (Greg Schafer)
Date: Sun, 14 Jan 2007 12:06:43 +1100
Subject: binutils-2.17 and -O3
In-Reply-To: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>
References: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>
Message-ID: <20070114010643.GA4826@eyo32.local>

On Sun, Jan 14, 2007 at 12:20:32AM +0000, Jon Grosshart wrote:

> Greets all. What better time for a new build than winter?

Heh, tis mid summer here in .au :-/

> So I dove in using 4.1.1/2.4/2.17
> 
> binutils Pass 2 of temptools crapped out on the make with:

<snip>

> Seems -O3 is the culprit somehow. Passing -O2 fixes it.

Yep. Came up again recently. Here's some comment from a Binutils dev:

http://lists.gnu.org/archive/html/bug-binutils/2007-01/msg00034.html

> Perhaps it's time to rethink my indiscriminate/rampant use of CFLAGS...

Indeed. My general rule of thumb for optimizing toolchain components is to
do only what the pros do.

Regards
Greg


From diy-linux-dev@diy-linux.org  Sun Jan 14 02:14:53 2007
From: diy-linux-dev@diy-linux.org (Greg Schafer)
Date: Sun, 14 Jan 2007 12:14:53 +1100
Subject: binutils-2.17 and -O3
In-Reply-To: <9cf823c60701131653h755d449dlbe2542666a7bc41c@mail.gmail.com>
References: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com> <9cf823c60701131653h755d449dlbe2542666a7bc41c@mail.gmail.com>
Message-ID: <20070114011453.GA4904@eyo32.local>

On Sun, Jan 14, 2007 at 12:53:26AM +0000, Jon Grosshart wrote:

> Now I'm kinda curious about --disable-werror on 1st pass but not 2nd... :-)

It's for robustness reasons ie: maximum buildability. Bootstrapping from
some old hosts (eg: RH62) results in compiler warnings which break the
build. By the time we get to pass 2, we are compiling with a modern GCC and
therefore no warnings.

Regards
Greg


From diy-linux-dev@diy-linux.org  Sun Jan 14 02:31:00 2007
From: diy-linux-dev@diy-linux.org (Jon Grosshart)
Date: Sun, 14 Jan 2007 01:31:00 +0000
Subject: binutils-2.17 and -O3
In-Reply-To: <20070114010643.GA4826@eyo32.local>
References: <9cf823c60701131620i3d47c6b4kc32b5c08d7808f44@mail.gmail.com>
 <20070114010643.GA4826@eyo32.local>
Message-ID: <9cf823c60701131731p7763bb22k9158eb6b0cfb118f@mail.gmail.com>

------=_Part_79570_17018482.1168738260430
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

On 1/14/07, Greg Schafer <gschafer@zip.com.au> wrote:

>
> Yep. Came up again recently. Here's some comment from a Binutils dev:
>
> http://lists.gnu.org/archive/html/bug-binutils/2007-01/msg00034.html
>
> > Perhaps it's time to rethink my indiscriminate/rampant use of CFLAGS...
>
> Indeed. My general rule of thumb for optimizing toolchain components is to
> do only what the pros do.
>
> Regards
> Greg



Thanks for the info. Think I will tone down the global flags and only apply
O3 from within a couple scripts as we do with glibc. Years of using those
flags from my first post, without problems, has made me somewhat complacent
I suppose.

Thanks again.

Jon

------=_Part_79570_17018482.1168738260430
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

<br><br><div><span class="gmail_quote">On 1/14/07, <b class="gmail_sendername">Greg Schafer</b> &lt;<a href="mailto:gschafer@zip.com.au">gschafer@zip.com.au</a>&gt; wrote:</span><br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
<br>Yep. Came up again recently. Here&#39;s some comment from a Binutils dev:<br><br><a href="http://lists.gnu.org/archive/html/bug-binutils/2007-01/msg00034.html">http://lists.gnu.org/archive/html/bug-binutils/2007-01/msg00034.html
</a><br><br>&gt; Perhaps it&#39;s time to rethink my indiscriminate/rampant use of CFLAGS...<br><br>Indeed. My general rule of thumb for optimizing toolchain components is to<br>do only what the pros do.<br><br>Regards<br>
Greg</blockquote><div><br><br>Thanks for the info. Think I will tone down the global flags and only apply O3 from within a couple scripts as we do with glibc. Years of using those flags from my first post, without problems, has made me somewhat complacent I suppose.
<br><br>Thanks again.<br><br>Jon<br></div><br></div><br>

------=_Part_79570_17018482.1168738260430--


From diy-linux-dev@diy-linux.org  Sun Jan 14 20:46:16 2007
From: diy-linux-dev@diy-linux.org (Jon Grosshart)
Date: Sun, 14 Jan 2007 14:46:16 -0500
Subject: glibc and host kernel. revisited
Message-ID: <9cf823c60701141146kbc26b10s1fa4628a697f0e74@mail.gmail.com>

------=_Part_87408_8726138.1168803976171
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Sorry to spam the dev list and more importantly everyones in-box, but I
found something else worth mentioning (atleast in my book)

Ran into some chroot glibc check errors just now and traced them back to my
host kernel (2.6.18.6)... Suprise, suprise... :-)

make[2]: [/tmp/glibc-build/posix/annexc.out] Error 1 (ignored)
make[2]: *** [/tmp/glibc-build/nptl/tst-cancel24.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx4.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx5.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx16.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx20.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx21.out] Error 1
make[2]: *** [/tmp/glibc-build/nptl/tst-cleanupx4.out] Error 1
make[1]: *** [nptl/tests] Error 2
make: *** [check] Error 2

I disabled - Processor type and features  ---> Compat VDSO support

recompiled, reboot, rechroot and am now left with the familiar annexc
error...

Huzzah! :-) She's a finicky thing, no doubt.

------=_Part_87408_8726138.1168803976171
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Sorry to spam the dev list and more importantly everyones in-box, but I found something else worth mentioning (atleast in my book)<br><br>Ran into some chroot glibc check errors just now and traced them back to my host kernel (
<a href="http://2.6.18.6">2.6.18.6</a>)... Suprise, suprise... :-)<br><br>make[2]: [/tmp/glibc-build/posix/annexc.out] Error 1 (ignored)<br>make[2]: *** [/tmp/glibc-build/nptl/tst-cancel24.out] Error 1<br>make[2]: *** [/tmp/glibc-build/nptl/tst-
cancelx4.out] Error 1<br>make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx5.out] Error 1<br>make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx16.out] Error 1<br>make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx20.out] Error 1<br>
make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx21.out] Error 1<br>make[2]: *** [/tmp/glibc-build/nptl/tst-cleanupx4.out] Error 1<br>make[1]: *** [nptl/tests] Error 2<br>make: *** [check] Error 2<br><br>I disabled - Processor type and features&nbsp; ---&gt; Compat VDSO support
<br><br>recompiled, reboot, rechroot and am now left with the familiar annexc error...<br><br>Huzzah! :-) She&#39;s a finicky thing, no doubt.<br><br><br>

------=_Part_87408_8726138.1168803976171--


From diy-linux-dev@diy-linux.org  Mon Jan 15 00:02:10 2007
From: diy-linux-dev@diy-linux.org (Greg Schafer)
Date: Mon, 15 Jan 2007 10:02:10 +1100
Subject: glibc and host kernel. revisited
In-Reply-To: <9cf823c60701141146kbc26b10s1fa4628a697f0e74@mail.gmail.com>
References: <9cf823c60701141146kbc26b10s1fa4628a697f0e74@mail.gmail.com>
Message-ID: <20070114230210.GA13198@eyo32.local>

On Sun, Jan 14, 2007 at 02:46:16PM -0500, Jon Grosshart wrote:

> Sorry to spam the dev list and more importantly everyones in-box, but I
> found something else worth mentioning (atleast in my book)

Worth mentioning for sure. Thanks for reporting.

> Ran into some chroot glibc check errors just now and traced them back to my
> host kernel (2.6.18.6)... Suprise, suprise... :-)
> 
> make[2]: [/tmp/glibc-build/posix/annexc.out] Error 1 (ignored)
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancel24.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx4.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx5.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx16.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx20.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cancelx21.out] Error 1
> make[2]: *** [/tmp/glibc-build/nptl/tst-cleanupx4.out] Error 1
> make[1]: *** [nptl/tests] Error 2
> make: *** [check] Error 2
> 
> I disabled - Processor type and features  ---> Compat VDSO support

Yep. Came up recently here:

http://sourceware.org/ml/libc-alpha/2007-01/msg00004.html

Incidentally, a patch set was just sent to LKML which may or may not fix it:

http://lkml.org/lkml/2007/1/14/2

BTW, the help text in the kernel config is reasonably clear. But it seems
the option defaults to Y, which is unfortunate.

> recompiled, reboot, rechroot and am now left with the familiar annexc
> error...

Arghhh! Please don't mention annexc here! :-) It drives me crazy seeing
references to this "error" on the LFS lists all the time. The test is
*expected* to fail hence the (ignored) message. Instead of grepping for
"Error", folks should be instructed to grep the log for something like:

"\*\*\*.*Error"

Regards
Greg


From diy-linux-dev@diy-linux.org  Mon Jan 15 00:12:21 2007
From: diy-linux-dev@diy-linux.org (Jon Grosshart)
Date: Sun, 14 Jan 2007 18:12:21 -0500
Subject: glibc and host kernel. revisited
In-Reply-To: <20070114230210.GA13198@eyo32.local>
References: <9cf823c60701141146kbc26b10s1fa4628a697f0e74@mail.gmail.com>
 <20070114230210.GA13198@eyo32.local>
Message-ID: <9cf823c60701141512ga5b3bfah7f56babd24ed078a@mail.gmail.com>

------=_Part_89315_16409529.1168816341121
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

On 1/14/07, Greg Schafer <gschafer@zip.com.au> wrote:
>
> The test is
> *expected* to fail hence the (ignored) message.



I know.. That's why I was happy it was the only one showing.. ;-)

Greg, do you know of every single possible thing that has ever gone wrong
during a bootstrap? I'm starting to think so. Your like a walking
'calculator' with the bug report links... One of these days I might post
something your not aware of but I doubt it...

------=_Part_89315_16409529.1168816341121
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

<br><br><div><span class="gmail_quote">On 1/14/07, <b class="gmail_sendername">Greg Schafer</b> &lt;<a href="mailto:gschafer@zip.com.au">gschafer@zip.com.au</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
The test is<br>*expected* to fail hence the (ignored) message. </blockquote><div><br><br>I know.. That&#39;s why I was happy it was the only one showing.. ;-)<br><br>Greg, do you know of every single possible thing that has ever gone wrong during a bootstrap? I&#39;m starting to think so. Your like a walking &#39;calculator&#39; with the bug report links... One of these days I might post something your not aware of but I doubt it... 
<br></div><br></div><br>

------=_Part_89315_16409529.1168816341121--


From diy-linux-dev@diy-linux.org  Tue Jan 23 22:00:02 2007
From: diy-linux-dev@diy-linux.org (Greg Schafer)
Date: Wed, 24 Jan 2007 08:00:02 +1100
Subject: bzip2-1.0.4 Makefile bug
Message-ID: <20070123210002.GA8514@eyo32.local>

Hi Guys,

This latest version has introduced a bug for "staged" DESTDIR installs. eg:

make PREFIX=/tmp/staging install

Specifically, the symlinks for bzegrep, bzfgrep, bzless and bzcmp all end up
pointing to the staging area. eg:

$ ls -l /tmp/staging/usr/bin
total 488
-rwxr-xr-x 1 pkgmgr pkgmgr 149426 Jan 23 17:53 bunzip2
-rwxr-xr-x 1 pkgmgr pkgmgr 149426 Jan 23 17:53 bzcat
lrwxrwxrwx 1 pkgmgr pkgmgr     27 Jan 23 17:53 bzcmp -> /tmp/staging/usr/bin/bzdiff
-rwxr-xr-x 1 pkgmgr pkgmgr   2128 Jan 23 17:53 bzdiff
lrwxrwxrwx 1 pkgmgr pkgmgr     27 Jan 23 17:53 bzegrep -> /tmp/staging/usr/bin/bzgrep
lrwxrwxrwx 1 pkgmgr pkgmgr     27 Jan 23 17:53 bzfgrep -> /tmp/staging/usr/bin/bzgrep
-rwxr-xr-x 1 pkgmgr pkgmgr   1677 Jan 23 17:53 bzgrep
-rwxr-xr-x 1 pkgmgr pkgmgr 149426 Jan 23 17:53 bzip2
-rwxr-xr-x 1 pkgmgr pkgmgr  16788 Jan 23 17:53 bzip2recover
lrwxrwxrwx 1 pkgmgr pkgmgr     27 Jan 23 17:53 bzless -> /tmp/staging/usr/bin/bzmore
-rwxr-xr-x 1 pkgmgr pkgmgr   1259 Jan 23 17:53 bzmore

This is a regression from 1.0.3. I've sent a report and patch upstream which
has been acknowledged. For now, I'll add a sed similar to this:

sed -i.bak '/ln .*PREFIX/s,$(PREFIX)/bin/,,' Makefile

Regards
Greg