[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I am getting the error in the subject line.  I have tried this both from
a test branch in the git repository and by unpacking the 2010.05 tarball
with the same results.  Perhaps I am doing something stupid, but I
cannot imagine what it might be.  I searched the list for ANY "recipe
commences before first target" instances but it seems I am the first to
encounter it, or at least the first to report it here.

Does anyone know what might be causing this?


[jdines at hydra buildroot-2010.05 ] make menuconfig
make CC="/usr/bin/gcc" -C package/config mconf
make[1]: Entering directory
`/home/jdines/Build/buildroot-2010.05/package/config'
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2 -MM *.c > .depend 2>/dev/null || :
make[1]: Leaving directory
`/home/jdines/Build/buildroot-2010.05/package/config'
make[1]: Entering directory
`/home/jdines/Build/buildroot-2010.05/package/config'
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/checklist.c -o lxdialog/checklist.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/inputbox.c -o lxdialog/inputbox.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/menubox.c -o lxdialog/menubox.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/textbox.c -o lxdialog/textbox.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/util.c -o lxdialog/util.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c lxdialog/yesno.c -o lxdialog/yesno.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c mconf.c -o mconf.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  mconf.o zconf.tab.o lxdialog/checklist.o lxdialog/util.o
lxdialog/inputbox.o lxdialog/textbox.o lxdialog/yesno.o
lxdialog/menubox.o -lncurses -o mconf
make[1]: Leaving directory
`/home/jdines/Build/buildroot-2010.05/package/config'
#
# configuration written to .config
#


*** End of Buildroot configuration.
*** Execute 'make' to build Buildroot or try 'make help'.

#
# make dependencies written to .auto.deps
# ATTENTION buildroot devels!
# See top of this file before playing with this auto-preprequisites!
#
[jdines at hydra buildroot-2010.05 ] cp .config
configs-make-menuconfig-no-changes
[jdines at hydra buildroot-2010.05 ] rm .config
rm: remove regular file `.config'? y
[jdines at hydra buildroot-2010.05 ] make defconfig
#
# configuration written to .config
#
#
# make dependencies written to .auto.deps
# ATTENTION buildroot devels!
# See top of this file before playing with this auto-preprequisites!
#
[jdines at hydra buildroot-2010.05 ] diff .config
configs-make-menuconfig-no-changes
3c3
< # Mon Aug 30 07:39:30 2010
- ---
> # Mon Aug 30 07:38:50 2010
[jdines at hydra buildroot-2010.05 ] make
package/atk/atk.mk:43: *** recipe commences before first target.  Stop.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMe5nJAAoJECSpOhdQESq6gdgIAIcKzSD/71L9mKosYOGyFJTb
VIf3UyADZnsUQC5JxrFel4EQEZdOzOUckbBJ/E3rvOgNbbhtIJH30hTZnWgbe5/4
bD4rlG1oUJAtqu65dPyrG5eGPANhA7U9Yca6//tU4xdy2Ecova35Fblw54AkWA+P
WhDyObVEjU/MagXP9KVlUlUKhcLJaDNFODdeBRe5UY2BiUtURBRJjz31+1BcBPPv
4lUrfYz6DrOoLc1iwh9ni2aICBDGd16exDAnvdqRbMeHoHKhCenzQagWlV7fGzdi
D1oyEYiQHraNxmJymXdTSQaBFynMVImWYAtDVQkd6r2bOwKdP/jRKYKT9spL568=
=3F3s
-----END PGP SIGNATURE-----

[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[Yann E. MORIN]

James, All,

On Monday 30 August 2010 13:45:25 James J. Dines wrote:
> I am getting the error in the subject line.  I have tried this both from
[--SNIP--]
> Does anyone know what might be causing this?

Please see the attached patch.

Peter, I'll send a proper patch shortly if you want.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
-------------- next part --------------
A non-text attachment was scrubbed...
Name: br.atk.patch
Type: text/x-diff
Size: 389 bytes
Desc: not available
URL: <http://lists.busybox.net/pipermail/buildroot/attachments/20100830/6e38a516/attachment.bin>

[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[Peter Korsgaard]

>>>>> "Yann" == Yann E MORIN <yann.morin.1998@anciens.enib.fr> writes:

 Yann> James, All,
 Yann> On Monday 30 August 2010 13:45:25 James J. Dines wrote:
 >> I am getting the error in the subject line.  I have tried this both from
 Yann> [--SNIP--]
 >> Does anyone know what might be causing this?

 Yann> Please see the attached patch.

 Yann> Peter, I'll send a proper patch shortly if you want.

Committed, thanks!

-- 
Bye, Peter Korsgaard

[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 09:37 AM, Peter Korsgaard wrote:
>>>>>> "Yann" == Yann E MORIN <yann.morin.1998@anciens.enib.fr> writes:
> 
>  Yann> James, All,
>  Yann> On Monday 30 August 2010 13:45:25 James J. Dines wrote:
>  >> I am getting the error in the subject line.  I have tried this both from
>  Yann> [--SNIP--]
>  >> Does anyone know what might be causing this?
> 
>  Yann> Please see the attached patch.
> 
>  Yann> Peter, I'll send a proper patch shortly if you want.
> 
> Committed, thanks!
> 

That patch indeed gets things rolling.  For further edification it
appears that there are some semantic changes in the parser for make 3.82
as the same tree builds fine with make 3.81.  It seems like it is
similar to this problem that Sam Ravnborg uncovered and patched (kernel
commit in the kernel Makefile 31110ebbec8688c6e9597b641101afc94e1c762a
)
 (See http://permalink.gmane.org/gmane.linux.kernel/1026412 for details)

I'm not sure how you guys handle regressions, but it seems like we make
3.82 users may look forward to experiencing other related problems so it
would be best to understand it fully.  Can anyone offer a good/thorough
explanation of what is going on by any chance, or do I have to figure it
out the old fashioned way?

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMe8Q2AAoJECSpOhdQESq6LkwH/3MM1pQSXcGrEhRujzOL6Oy0
m+PRc+113B8UCcj8Cdf3iMpCD+b47gyB4cDf/nhpmXJxh7V6id3kD4mt2mk9dwR4
UIW6VTVNJZhQiCUyI4C5r7zvbtiyb2v+eFYPxll+VqQuvhmLo5DJla/klYJuC/MX
yfhxmEzoiUxX2FaLagp6xT14WlY4KkXmGfKqZsv2oIv6fOZZwbYjuMTM5dA6OimA
xeyA7O71VjXr3X7iAuRttkuggPrktIJa/cl7Tk2AD4jWGXd5kekmunx6MNuj04VH
RWyHeP0aX7EEu0XVNXAVHSUECXTmdlZD9WVTpHZoEjwDvKq0flNVFNOZ4ufUZ4Y=
=X5Gx
-----END PGP SIGNATURE-----

[Buildroot] package/atk/atk.mk:43: *** recipe commences before first target. Stop.

[Yann E. MORIN]

James, All,

On Monday 30 August 2010 16:46:25 James J. Dines wrote:
> That patch indeed gets things rolling.  For further edification it
[--SNIP--]
> Can anyone offer a good/thorough
> explanation of what is going on by any chance, or do I have to figure it
> out the old fashioned way?

Well, in this case, it had nothing to do with make-3.82. It was 'just' a
variable definition that was spanning multiple lines, and did not have a
back-slash at the end of one of those lines, so the next line was interpreted
by make as a recipe without rule.

If you're suggesting that make-3.81 was not choking on this line, then this
would have been a bug in make (which I doubt...).

Or I mis-interpreted your message... :-/

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 223 225 172 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 11:01 AM, Yann E. MORIN wrote:
> James, All,
> 
> On Monday 30 August 2010 16:46:25 James J. Dines wrote:
>> That patch indeed gets things rolling.  For further edification it
> [--SNIP--]
>> Can anyone offer a good/thorough
>> explanation of what is going on by any chance, or do I have to figure it
>> out the old fashioned way?
> 
> Well, in this case, it had nothing to do with make-3.82. It was 'just' a
> variable definition that was spanning multiple lines, and did not have a
> back-slash at the end of one of those lines, so the next line was interpreted
> by make as a recipe without rule.


> 
> If you're suggesting that make-3.81 was not choking on this line, then this
> would have been a bug in make (which I doubt...).

Can you in fact unpack the 2010.05 tarball and successfully do a 'make
defconfig && make' using make 3.82?

I scp'ed the same exact tree that was failing to a box with make 3.81
without changing a thing and it built fine.  To be clear, I am talking
about stable released tarballs here that I am sure people have been
using regularly for quite some time, not the git repo.  If it is a bug
in make, then it has been there up until 3.82 ;-)  OTOH, I just bought
an 8 core laptop and installed the latest development version of Mandiva
Linux on the box where I get the failure, so maybe something else is amiss.

Still, that same new box builds various versions of vanilla and Ingo
realtime patched kernels without a problem.

If you look at this thread from the LKML:

http://comments.gmane.org/gmane.linux.kernel/1022928

... starting at the line "BTW seems like 2.6.27 no more combatible with
GNU Make 3.82:" you will see that there is definitely something
different between older makes and 3.82 (since I know 2.6.27 compiles
with many different make versions).  My guess is there is a bug in make
3.82, but this is only conjecture, since I cannot be sure the make team
didn't intend for the change (it may be a bugfix for a long standing bug)

In any case, users of make 3.82 can expect to grapple with this, thus
the desire to fully understand what is really gong on here.


Here is the proof that make 3.82 acts differently:

[jdines at hydra linux-2.6 (make-fix)] git checkout -b
make-3.82-shoukld-fail v2.6.26
Checking out files: 100% (21206/21206), done.
Switched to a new branch 'make-3.82-shoukld-fail'
[jdines at hydra linux-2.6 (make-3.82-shoukld-fail)] make clean && make
Makefile:1550: *** mixed implicit and normal rules.  Stop.

real    0m0.269s
user    0m0.101s
sys     0m0.097s
[jdines at hydra linux-2.6 (make-3.82-shoukld-fail)]



> 
> Or I mis-interpreted your message... :-/
> 
> Regards,
> Yann E. MORIN.
> 

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMe9HMAAoJECSpOhdQESq6brsH/itD8LTfB1xRjwR7H5qvrUBW
VLzt226bW5dRwMYkX5Gf0SkwlmN3X4YSihhedraALwgFic9E6pEH17fnW0MqO4dc
RRIPZUImOekiZVvxRFuf+qPuckThP34yhNyMd/EAV/C0WEJBELhSRfeA0hFsaXo8
zT9de+8jn7FR5MUVPLe0JV2/GAMuFA4/DEbe1X7rbGtMH21AsjUJ04d9fa+V/u8o
USfi3KQRiV2222OxLJKpY+OZFdz/LB/k3JpudN/msTPCCFm1FuTaIhuaOduevDst
WgKbCBgwKO2Rq6JBPJegy4hxnnT+vvCIFq/XRTpqLPFVCJYBhN7vdSHJPtoGfi0=
=45cr
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[Sam Ravnborg]

> 
> If you look at this thread from the LKML:
> 
> http://comments.gmane.org/gmane.linux.kernel/1022928
> 
> ... starting at the line "BTW seems like 2.6.27 no more combatible with
> GNU Make 3.82:" you will see that there is definitely something
> different between older makes and 3.82 (since I know 2.6.27 compiles
> with many different make versions).  My guess is there is a bug in make
> 3.82, but this is only conjecture, since I cannot be sure the make team
> didn't intend for the change (it may be a bugfix for a long standing bug)
> 
> In any case, users of make 3.82 can expect to grapple with this, thus
> the desire to fully understand what is really gong on here.
> 
> 
> Here is the proof that make 3.82 acts differently:
> 
> [jdines at hydra linux-2.6 (make-fix)] git checkout -b
> make-3.82-shoukld-fail v2.6.26
> Checking out files: 100% (21206/21206), done.
> Switched to a new branch 'make-3.82-shoukld-fail'
> [jdines at hydra linux-2.6 (make-3.82-shoukld-fail)] make clean && make
> Makefile:1550: *** mixed implicit and normal rules.  Stop.

make 3.81 accept the following:

foo foo%: bar
	command

With make 3.82 you need to split this up as:

foo: bar
	command

foo%: bar
	command

And we use the "foo foo%" construct in the top-level Makefile of the kernel.
The stable team have now backported a fix for this and it should apply
to 2.6.26 too - or at least simple to do manually.

In any recent kernels this is fixed long time ago (except for powerpc
where we fixed this only recently).

I hope this helps.

	Sam

[Buildroot] Make 3.82 does behave differently than make 3.81

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 02:50 PM, Sam Ravnborg wrote:
> On Mon, Aug 30, 2010 at 01:00:52PM -0400, James J. Dines wrote:
>> -----BEGIN PGP SIGNED MESSAGE-----
>> Hash: SHA1
>>
>> On 08/30/2010 12:31 PM, Sam Ravnborg wrote:
>>>>
>> <snip>
>>> And we use the "foo foo%" construct in the top-level Makefile of the kernel.
>>> The stable team have now backported a fix for this and it should apply
>>> to 2.6.26 too - or at least simple to do manually.
>>
>> It is a git tree.  With a tarball, you can unpack it, patch it, repack
>> it and put it up on the web with the same name thereby having two
>> identically named tarballs with different contents (though this is a
>> horrible idea of course.)
> 
> The mainstream kernel is the one that you are referring to.
> 
> The stable kernel serie I refer to can be found here:
> http://git.kernel.org/?p=linux/kernel/git/hpa/linux-2.6-allstable.git;a=summary
> 
> The branch linux-2.6.27.y contains all the fixes backported to 2.6.27.
> This includes the following commit:
> http://git.kernel.org/?p=linux/kernel/git/hpa/linux-2.6-allstable.git;a=commit;h=5981f194641d4bcbff7e80adf393035b113f5022
> 
> The referred commit it the one that fixes the make incompatibility in the 2.6.27 kernel.
> The more specific name for this kernel release is: 2.6.27.53 as this is the 53th time
> the stable team released a new stable kernel in the 2.6.27 serie.
> 
> PS. Did you drop the mailing list on purpose?
> If not then I do not mind if you forward this to the list
> as it may benefit others.
> 
> 	Sam

I did send to you directly because I am new here and didn't want to
generate too much noise in case I was completely off base.  I fear
people already think I am clueless complaining about make 3.82 as it is :-)

I was unaware if the linux-2.6-allstable tree and so I am replying to
the list because I agree that your feedback may well be helpful to
others.  Thanks for your time and effort.

I am looking more deeply into this issue, but it is really starting to
look like make 2.82 has a bug, as I am seeing many errors as I feared.
It could also be some other issue of course, but I don't think so at
this point.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMe//tAAoJECSpOhdQESq62msIAIibNg7ude+Pf8X8pkwuKzO0
2a9FrLPolzZjboKeBJg5U097ikZ1ywJE/TB5ZMt13KuiyKL/BPngdvKAnyfzOu5G
uzQNXO8/v13rgzkzj1A5YtoNiPeyzVw57e65EOZrxOM75Fd0K1IxacyLpfiI+eAK
ES0xZwn+YsCIJsoFEMTnFW8So0RsEkp+HESOdrxXUybC7zrPpVOEoBhD+0SMauxE
Qkl4Nrgk4mx9mfAjwMmdSyMoW+3TOlUiopYCG02Q9OsnJteDD8bepJxbSXu9cQMm
2p/Ny2PV9WCZrFguvd4KbaY7+eP55w/J9dvA4D2mjfxSLAA/CrgK/FdHoBTHW9k=
=f6S2
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[Peter Korsgaard]

>>>>> "Sam" == Sam Ravnborg <sam@ravnborg.org> writes:

Hi,

 Sam> make 3.81 accept the following:

 Sam> foo foo%: bar
 Sam> 	command

 Sam> With make 3.82 you need to split this up as:

 Sam> foo: bar
 Sam> 	command

 Sam> foo%: bar
 Sam> 	command

Yeah, saw the kernel changes passing by. I haven't had time to test
building with make 3.82 yet though (it's not in Debian yet), so we might
have something similar lingering around somewhere ..

-- 
Bye, Peter Korsgaard

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 04:00 PM, Peter Korsgaard wrote:
>>>>>> "Sam" == Sam Ravnborg <sam@ravnborg.org> writes:
> 
> Hi,
> 
>  Sam> make 3.81 accept the following:
> 
>  Sam> foo foo%: bar
>  Sam> 	command
> 
>  Sam> With make 3.82 you need to split this up as:
> 
>  Sam> foo: bar
>  Sam> 	command
> 
>  Sam> foo%: bar
>  Sam> 	command
> 
> Yeah, saw the kernel changes passing by. I haven't had time to test
> building with make 3.82 yet though (it's not in Debian yet), so we might
> have something similar lingering around somewhere ..
> 

I am really interested to hear your results, because I just bought this
new 8 core machine and set it up this weekend.  So far everything has
worked great, but there is always this nagging worry that I have
something set up wrong and I will have to post back to the list stating
that I was an idiot and missed something :-)

In any case, untarring buildroot-2010.05, doing a make defconfig
followed by a make clean seems like a set of steps that should go off
without a hitch (it certainly does on my other machine) but it fails
with make 3.82 for me anyway :-(



jdines at hydra buildroot-2010.05 ] ls
CHANGES    configs/  docs/  Makefile  scripts/  TODO
Config.in  COPYING   fs/    package/  target/   toolchain/
[jdines at hydra buildroot-2010.05 ] make defconfig
make CC="/usr/bin/gcc" -C package/config conf
make[1]: Entering directory
`/home/jdines/Build/buildroot-2010.05/package/config'
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2 -MM *.c > .depend 2>/dev/null || :
make[1]: Leaving directory
`/home/jdines/Build/buildroot-2010.05/package/config'
make[1]: Entering directory
`/home/jdines/Build/buildroot-2010.05/package/config'
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c conf.c -o conf.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  -c kxgettext.c -o kxgettext.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2 -I. -c zconf.tab.c -o zconf.tab.o
/usr/bin/gcc -I/usr/include/ncurses -DCURSES_LOC="<ncurses.h>" -DLOCALE
 -g -O2  conf.o  zconf.tab.o -lncurses -o conf
make[1]: Leaving directory
`/home/jdines/Build/buildroot-2010.05/package/config'
#
# configuration written to .config
#
#
# make dependencies written to .auto.deps
# ATTENTION buildroot devels!
# See top of this file before playing with this auto-preprequisites!
#
[jdines at hydra buildroot-2010.05 ] make clean
package/atk/atk.mk:43: *** recipe commences before first target.  Stop.
[jdines at hydra buildroot-2010.05 ] view Makefile
[jdines at hydra buildroot-2010.05 ] ls
CHANGES    configs/  docs/  Makefile  package/  target/  toolchain/
Config.in  COPYING   fs/    output/   scripts/  TODO
[jdines at hydra buildroot-2010.05 ] make clean
package/atk/atk.mk:43: *** recipe commences before first target.  Stop.

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMfBLYAAoJECSpOhdQESq66n4IAJ86+yxdo+MbQjsWAvnVuLwr
Dbd6DRg0SjOO/ScxxOm5cdRa501lDOFqwWh1Pz09NKGMbGz7flw7BagE6+/dLfCe
S+75uhXkbI4hTvvpspd7XIaqXUr6FQmlmWQvIerDyaTjRxwNHCjeoRaWXo2/AnpY
rfHcUSOYUfw+1akIPu9Z0/5wwMRB7DlVpecChQ88ePhiMWmQ6PFrKDiXZkCJJfDq
R0gxrWasjbVf0F3429d3NUneseJLjXxUqnhcZt1aNOri+oFR0OBlElPaHuCjTTPJ
Kml2qSrSkylGGEVieP93YCY3P4EalpnHwGWWBPQLBSjAlWs4/ZYVM6dzlSOpTew=
=KCGa
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[Peter Korsgaard]

>>>>> "James" == James J Dines <jdines@jdines.net> writes:

Hi,

 James> I am really interested to hear your results, because I just
 James> bought this new 8 core machine and set it up this weekend.  So
 James> far everything has worked great, but there is always this
 James> nagging worry that I have something set up wrong and I will have
 James> to post back to the list stating that I was an idiot and missed
 James> something :-)

 James> In any case, untarring buildroot-2010.05, doing a make defconfig
 James> followed by a make clean seems like a set of steps that should
 James> go off without a hitch (it certainly does on my other machine)
 James> but it fails with make 3.82 for me anyway :-(

 James> [jdines at hydra buildroot-2010.05 ] make clean
 James> package/atk/atk.mk:43: *** recipe commences before first target.  Stop.

Yeah, that's the bug in atk.mk Yann fixed earlier today. Could you give
git a try instead (either git clone or wait for tomorrows snapshot
tarball).

-- 
Bye, Peter Korsgaard

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 04:57 PM, Peter Korsgaard wrote:
>>>>>> "James" == James J Dines <jdines@jdines.net> writes:
> 
> Hi,
> 
>  James> I am really interested to hear your results, because I just
>  James> bought this new 8 core machine and set it up this weekend.  So
>  James> far everything has worked great, but there is always this
>  James> nagging worry that I have something set up wrong and I will have
>  James> to post back to the list stating that I was an idiot and missed
>  James> something :-)
> 
>  James> In any case, untarring buildroot-2010.05, doing a make defconfig
>  James> followed by a make clean seems like a set of steps that should
>  James> go off without a hitch (it certainly does on my other machine)
>  James> but it fails with make 3.82 for me anyway :-(
> 
>  James> [jdines at hydra buildroot-2010.05 ] make clean
>  James> package/atk/atk.mk:43: *** recipe commences before first target.  Stop.
> 
> Yeah, that's the bug in atk.mk Yann fixed earlier today. Could you give
> git a try instead (either git clone or wait for tomorrows snapshot
> tarball).
> 

I went for a third option (applying Yann's patch) and it indeed allows
the 'make clean' to work.  I remain confused why we did not need this
patch to do a complete build with make 3.81 however.  I just tried it
again, and I do not get the error message on my other machine with 3.81
even without the patch.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMfB4eAAoJECSpOhdQESq6WcYH/j1xrDLXel5j4J71Rqsv6gFr
5zy99A62YqgRjh4VrEZM07SnjZxAAimpEp5eUQCGGGbbxM3q0qFfZmGc6xPo+0lA
+bUP9fMdwjQiSJYZ42fmbSSNz3fvoQ6RhP85J3H5KvZEBD91H5jGtrNV7RK7Z+oI
PHy4aZroBlUP+TyQNRPB9Zd8ylC/sWghUya3cWs67h1rROGJKtOLeSaOWTSeRuKI
B7Lj7l3MDPB4rt1rC++vE7Sn17hiX1MCDoBijxuyR8jCmCwGUuiG/C+yUL+5oY2v
10uoQq0/wZ7P1uNiutWsoviLv8/KStQuOYg86ONUsJEVgDubHgLNF9rMXlQTopg=
=b3Zf
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[Peter Korsgaard]

>>>>> "James" == James J Dines <jdines@jdines.net> writes:

Hi,

 James> I went for a third option (applying Yann's patch) and it indeed
 James> allows the 'make clean' to work.  I remain confused why we did
 James> not need this patch to do a complete build with make 3.81
 James> however.  I just tried it again, and I do not get the error
 James> message on my other machine with 3.81 even without the patch.

No, it seems like make 3.81 ignores those invalid lines.

besides this issue, is everything working for you with ake 3.82?

-- 
Bye, Peter Korsgaard

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[Malte Starostik]

Am Montag, 30. August 2010, 23:10:00 schrieb James J. Dines:
> I went for a third option (applying Yann's patch) and it indeed allows
> the 'make clean' to work.  I remain confused why we did not need this
> patch to do a complete build with make 3.81 however.  I just tried it
> again, and I do not get the error message on my other machine with 3.81
> even without the patch.

From make 3.82's NEWS file:

* WARNING: Backward-incompatibility!
  As a result of parser enhancements, three backward-compatibility issues
  exist: first, a prerequisite containing an "=" cannot be escaped with a
  backslash any longer.  [...].  Second, variable names can no
  longer contain whitespace, unless you put the whitespace in a variable and
  use the variable.  Third, in previous versions of make it was sometimes
  not flagged as an error for explicit and pattern targets to appear in the
  same rule.  Now this is always reported as an error.


What you're referring to re the kernel Makefile is the third mentioned issue.  
What happens with atk.mk is the second one instead:

The incorrent Makefile snippet:

ATK_CONF_OPT =  --enable-shared \
                --enable-static
                --disable-glibtest --enable-explicit-deps=no \
                --disable-debug

assigns the vaue "--enable-shared --enable-static" to a variable named 
"ATK_CONF_OPT" and the value "no --disable-debug" to a variable named "--
disable-glibtest --enable-explicit-deps".

I didn't know variable names with spaces (and even leading hyphens) were 
allowed prior to 3.82, but that made make not barf on the above while of 
course depriving $(ATK_CONF_OPT) of some of the options.

Cheers
Malte

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 05:36 PM, Peter Korsgaard wrote:
>>>>>> "James" == James J Dines <jdines@jdines.net> writes:
> 
> Hi,
> 
>  James> I went for a third option (applying Yann's patch) and it indeed
>  James> allows the 'make clean' to work.  I remain confused why we did
>  James> not need this patch to do a complete build with make 3.81
>  James> however.  I just tried it again, and I do not get the error
>  James> message on my other machine with 3.81 even without the patch.
> 
> No, it seems like make 3.81 ignores those invalid lines.
> 
> besides this issue, is everything working for you with ake 3.82?
> 
I think things are working OK.  I am currently getting this error while
building oprofile 0.9.4 during the configure stage:

checking for bfd_openr in -lbfd... no
configure: error: bfd library not found
make: ***
[/home/jdines/Build/buildroot/output/build/oprofile-0.9.4/.stamp_configured]
Error 1

... but it seems like this is might be the issue:

http://old.nabble.com/configure-error:-bfd-library-not-found-td23863431.html

Excerpt:

So, do you have libbfd(.a or .so) installed in the usual place where
config would look for it?  If you do, then perhaps the problem is you
have a newer binutils where libbfd has a dependency on libz that
oprofile 0.9.4 did not handle, which resulted in this rather unhelpful
error message.  Pull the latest oprofile cvs code for a fix.


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMfD6AAAoJECSpOhdQESq6ESEH/j8zRLV8g3oejph8MFgci0Wi
82wE9f9MDw0mtoyiWWYD5z2n8VhkqgsnhCqbOjKHMwjGi9G7AuyTzuwqZxC50A6E
kuVajig90Hxz0Ae0kH97ntte7mtwS19I/yXw8cFVorvUV6n6iOlmtni+vi/4T6tt
72xRLm3MmG46EGrn0W8297e3a6O4eUAM3f+GXV/pG2RVb5YnE9WnYC4axAZpMiT3
dyARY+w6XkXdy2+JX6IbzlXxj9v7mAtSd/9puvkhxkGrVLd60lnw2CQ/U0h8kMZq
NuPJYF/umHTO7hnS5Vf9Sp2hHTCjjL+rzRCDBE0M7AbMwdfVnIAe1nweC4ji4K8=
=QQ8F
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 does behave differently than make 3.81 for sure, but ...

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 06:15 PM, Malte Starostik wrote:
> Am Montag, 30. August 2010, 23:10:00 schrieb James J. Dines:
<snip>
>>From make 3.82's NEWS file:
> 
> * WARNING: Backward-incompatibility!
>   As a result of parser enhancements, three backward-compatibility issues
>   exist: first, a prerequisite containing an "=" cannot be escaped with a
>   backslash any longer.  [...].  Second, variable names can no
>   longer contain whitespace, unless you put the whitespace in a variable and
>   use the variable.  Third, in previous versions of make it was sometimes
>   not flagged as an error for explicit and pattern targets to appear in the
>   same rule.  Now this is always reported as an error.
> 
> 
> What you're referring to re the kernel Makefile is the third mentioned issue.  
> What happens with atk.mk is the second one instead:
> 
> The incorrent Makefile snippet:
> 
> ATK_CONF_OPT =  --enable-shared \
>                 --enable-static
>                 --disable-glibtest --enable-explicit-deps=no \
>                 --disable-debug
> 
> assigns the vaue "--enable-shared --enable-static" to a variable named 
> "ATK_CONF_OPT" and the value "no --disable-debug" to a variable named "--
> disable-glibtest --enable-explicit-deps".
> 
> I didn't know variable names with spaces (and even leading hyphens) were 
> allowed prior to 3.82, but that made make not barf on the above while of 
> course depriving $(ATK_CONF_OPT) of some of the options.
> 
> Cheers
> Malte
> _______________________________________________
> buildroot mailing list
> buildroot at busybox.net
> http://lists.busybox.net/mailman/listinfo/buildroot

Thanks Malte!

In truth I would have NEVER looked in the NEWS file, as this seems like
it belongs in a file called IMPORTANT or something similar.

Now that I have this info I am armed to really take a look at what is
different about the build now that these errors will be properly caught.
 I have begun doing so and will report back on (some of?) my findings
later today.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMfPeoAAoJECSpOhdQESq6itwH/01ZhkzvL+g+KMnQCwdNwkxI
DvcKLwg17YDeRip7+xSewEMriTbw0S+0c6ya4Ck+4Qkf7kgXSETBuLShkDD0TJ46
rdCj5iJFnkWDRXxfltvF5AMYf9Cjvg1V3M+QciSyaYyCzvP5ayjG2JmtwjboCJhX
TcgcMsUiB1uYp8MzV7erdg2LC/LhQXuJztG8GXvvepPA7En/TVhKyfZNXX4w6Y9/
f6Cti1Zoe3JumIeT1a9dD/1u0Vxq4+biKWatJzXHT5BlfUCee0meCHZYmmRfx494
85iYHkjuL3GWyszqenL2NemUCdpIZYvPl2qMw619lIouoM5bBBkdfN66iOv+L2Q=
=W+BU
-----END PGP SIGNATURE-----

[Buildroot] Make 3.82 testing progress update

[James J. Dines]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 08/30/2010 05:36 PM, Peter Korsgaard wrote:
>>>>>> "James" == James J Dines <jdines@jdines.net> writes:
> 
> Hi,
> 
>  James> I went for a third option (applying Yann's patch) and it indeed
>  James> allows the 'make clean' to work.  I remain confused why we did
>  James> not need this patch to do a complete build with make 3.81
>  James> however.  I just tried it again, and I do not get the error
>  James> message on my other machine with 3.81 even without the patch.
> 
> No, it seems like make 3.81 ignores those invalid lines.
> 
> besides this issue, is everything working for you with ake 3.82?
> 

Other than the host-module-init-tools-3.11 issue already covered in
great detail today, which presumably needs to be fixed upstream
eventually, I was able to get everything to build but a kernel.

I haven't determined if this is something I am doing wrong or not yet.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)

iQEcBAEBAgAGBQJMfX+BAAoJECSpOhdQESq6PHAH/Rn7gPJxS8hKmImZpUzRDOXQ
+BREl/iizfrpz0zIwn7PYuYQG9yErWRBZxOntLVU7Qm54mdH24LJAqG+wR8V+T70
uzsh1iB8M1ZiXFPwa8wXoq5H4pzg3lrWqzAwEQ2HFtY5SViKpCNhyFUVA8z/Iqyh
iy8VPM7H+Evuhv+XDBvXsqX9M7Ls7xK2sQ17vsKfMMIishe9vB8GnqxzqJmrc+5G
bd98aENrw/0czP5brNUlPVrkBhkT7H+IX4FVMKS8DeRwQEiyUJxx9HEWo9J4TVRs
IHjNf5mFiEVOkWh1VdZXz/2itxbn8sMYzagmjrk6VosLwurgxSG7mj7V70S7tYk=
=EVoC
-----END PGP SIGNATURE-----