stable branch checksum missmatch for quilt-0.45

stable branch checksum missmatch for quilt-0.45

[E Robertson]

[-- Attachment #1: Type: text/plain, Size: 1268 bytes --]

building helloworld:

OE Build Configuration:
BB_VERSION     = "1.8.10"
OE_REVISION    = "a0a3cc0088bc51023a19b1d6942f7cc643e6bb52"
TARGET_ARCH    = "arm"
TARGET_OS      = "linux-uclibcgnueabi"
MACHINE        = "at91sam9263ek"
DISTRO         = "angstrom"
DISTRO_VERSION = "2008.1-test-20080331"
TARGET_FPU     = "soft"

...
NOTE: package quilt-native-0.45: started
NOTE: package quilt-native-0.45-r1: task do_fetch: started
sh: oe_sha256sum: command not found
NOTE: The SHA256 Sums do not match. 
Wanted: 'de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8' 
Got: ''
NOTE: Task failed: Checksum 
of 'http://download.savannah.gnu.org/releases/quilt/quilt-0.45.tar.gz' failed
NOTE: package quilt-native-0.45-r1: task do_fetch: failed
ERROR: TaskFailed event exception, aborting
NOTE: package quilt-native-0.45: failed
ERROR: Build 
of /OpenEmbedded_Stuff/org.openembedded.stable/packages/quilt/quilt-native_0.45.bb 
do_fetch failed



according to checksums.ini:

[http://download.savannah.gnu.org/releases/quilt/quilt-0.45.tar.gz]
md5=5f8dd5a041ca7c700b16620228f922aa
sha256=de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8

If I edit this file and later do a stable update, would it erase my changes?

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 196 bytes --]

Re: stable branch checksum missmatch for quilt-0.45

[E Robertson]

[-- Attachment #1: Type: text/plain, Size: 1627 bytes --]

On Monday 31 March 2008 02:38:05 pm E Robertson wrote:
> building helloworld:
>
> OE Build Configuration:
> BB_VERSION     = "1.8.10"
> OE_REVISION    = "a0a3cc0088bc51023a19b1d6942f7cc643e6bb52"
> TARGET_ARCH    = "arm"
> TARGET_OS      = "linux-uclibcgnueabi"
> MACHINE        = "at91sam9263ek"
> DISTRO         = "angstrom"
> DISTRO_VERSION = "2008.1-test-20080331"
> TARGET_FPU     = "soft"
>
> ...
> NOTE: package quilt-native-0.45: started
> NOTE: package quilt-native-0.45-r1: task do_fetch: started
> sh: oe_sha256sum: command not found

Sorry, I checked the sums (should have done this in the first place) are they 
are correct. The problem I see (remember I'm new here) is oe_sha256sum. Is 
this an error or should their be a macro using the host sha256sum?


> NOTE: The SHA256 Sums do not match.
> Wanted: 'de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8'
> Got: ''
> NOTE: Task failed: Checksum
> of 'http://download.savannah.gnu.org/releases/quilt/quilt-0.45.tar.gz'
> failed NOTE: package quilt-native-0.45-r1: task do_fetch: failed
> ERROR: TaskFailed event exception, aborting
> NOTE: package quilt-native-0.45: failed
> ERROR: Build
> of
> /OpenEmbedded_Stuff/org.openembedded.stable/packages/quilt/quilt-native_0.4
>5.bb do_fetch failed
>
>
>
> according to checksums.ini:
>
> [http://download.savannah.gnu.org/releases/quilt/quilt-0.45.tar.gz]
> md5=5f8dd5a041ca7c700b16620228f922aa
> sha256=de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8
>
> If I edit this file and later do a stable update, would it erase my
> changes?



[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 196 bytes --]

Re: stable branch checksum missmatch for quilt-0.45

[Holger Freyther]

On Monday 31 March 2008 21:38:05 E Robertson wrote:

> NOTE: package quilt-native-0.45: started
> NOTE: package quilt-native-0.45-r1: task do_fetch: started
> sh: oe_sha256sum: command not found
> NOTE: The SHA256 Sums do not match.
> Wanted: 'de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8'
> Got: ''

My prolog knowledge is pretty weak but let us analys this:

1. sh: oe_sha256sum: command not found

So executing a command called oe_sha256sum failed


2. > NOTE: The SHA256 Sums do not match.
 > Wanted: 'de7290f334baf631a14d031df304ee94cfb550aa69e7512aa589d21c2381c7d8'
 > Got: ''

So the SHA256 did not match? "" vs. long number (every sha256 hash has the 
same size)


Resolving facts:
	sha256 might be equal to SHA256

	oe_sha256sum might be an OpenEmbedded (oe) program to calculate a
	sha256. It was not able to execute it

	It was not able to execute oe_sha256sum, the stuff printed on the console
	was "".

	This could explain the mismatch?

Applying knowledge:
	You have a failure because oe_sha256sum failed to be executed.

Possible next steps:
	- Check if you have a oe_sha256sum
	- Check if your home partition allows execution of binaries
	- Check your build logs if oe_sha256sum was attempted to be built
	- Check the dep graph if there is a oe_sha256sum in it (bitbake -g and then 
	  read the files and search for sha256)

oe_sha256sum is supposed to be built before everything else, and normally is. 
So good luck and please provide more feedback (you seem to get started)

z.

Re: stable branch checksum missmatch for quilt-0.45

[E Robertson]

[-- Attachment #1: Type: text/plain, Size: 1610 bytes --]

Hi Z,
You must be the funny one.

> Resolving facts:
> 	sha256 might be equal to SHA256
>
> 	oe_sha256sum might be an OpenEmbedded (oe) program to calculate a
> 	sha256. It was not able to execute it
>
> 	It was not able to execute oe_sha256sum, the stuff printed on the console
> 	was "".
>
> 	This could explain the mismatch?
>
> Applying knowledge:
> 	You have a failure because oe_sha256sum failed to be executed.
>
> Possible next steps:
> 	- Check if you have a oe_sha256sum
> 	- Check if your home partition allows execution of binaries
> 	- Check your build logs if oe_sha256sum was attempted to be built
> 	- Check the dep graph if there is a oe_sha256sum in it (bitbake -g and
> then read the files and search for sha256)
>
> oe_sha256sum is supposed to be built before everything else, and normally
> is. So good luck and please provide more feedback (you seem to get started)

I had already checked if the oe_sha256sum executable was build and it was not. 
I also did a grep to see where the build was initiated.
Building from the dev branch did not give me this error because it was build - 
hence I specify the _stable_ branch:

./oetmp/work/i686-linux/shasum-native-1.0-r1/oe_sha256sum
./oetmp/staging/i686-linux/usr/bin/oe_sha256sum

Since it was already built by the dev branch and I use the same local file to 
build the stable branch, would it / should it skip the build and use the 
built binary? Or would it just rebuild the binary?
What is the command to a clean build?

In your last point above, what is a "dep graph"?

Thanks.









[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 196 bytes --]

Re: stable branch checksum missmatch for quilt-0.45

[E Robertson]

[-- Attachment #1: Type: text/plain, Size: 5099 bytes --]

On Tuesday 01 April 2008 08:07:34 am E Robertson wrote:
> Hi Z,
> You must be the funny one.
>
> > Resolving facts:
> > 	sha256 might be equal to SHA256
> >
> > 	oe_sha256sum might be an OpenEmbedded (oe) program to calculate a
> > 	sha256. It was not able to execute it
> >
> > 	It was not able to execute oe_sha256sum, the stuff printed on the
> > console was "".
> >
> > 	This could explain the mismatch?
> >
> > Applying knowledge:
> > 	You have a failure because oe_sha256sum failed to be executed.
> >
> > Possible next steps:
> > 	- Check if you have a oe_sha256sum
> > 	- Check if your home partition allows execution of binaries
> > 	- Check your build logs if oe_sha256sum was attempted to be built
> > 	- Check the dep graph if there is a oe_sha256sum in it (bitbake -g and
> > then read the files and search for sha256)
> >
> > oe_sha256sum is supposed to be built before everything else, and normally
> > is. So good luck and please provide more feedback (you seem to get
> > started)
>
> I had already checked if the oe_sha256sum executable was build and it was
> not. I also did a grep to see where the build was initiated.
> Building from the dev branch did not give me this error because it was
> build - hence I specify the _stable_ branch:
>
> ./oetmp/work/i686-linux/shasum-native-1.0-r1/oe_sha256sum
> ./oetmp/staging/i686-linux/usr/bin/oe_sha256sum
>
> Since it was already built by the dev branch and I use the same local file
> to build the stable branch, would it / should it skip the build and use the
> built binary? Or would it just rebuild the binary?
> What is the command to a clean build?
>
> In your last point above, what is a "dep graph"?

I did a clean build (?) (bitbake shasum-native -c clean && bitbake -b 
shasum-native.bb)  and it build suscessfully (at least it appears to be).
At what point is staging/i686-linux/usr/bin/oe_sha256sum copy updated?

rebuilding bitbake -b quilt_0.45.bb I got the error below but the quilt-native 
package build successfully. 

NOTE: package quilt-0.45: started
NOTE: package quilt-0.45-r1: task do_configure: started
ERROR: function do_configure failed
ERROR: log data follows 
(/OpenEmbedded_Stuff/oetmp/work/armv5te-angstrom-linux-uclibcgnueabi/quilt-0.45-r1/temp/log.do_configure.16497)
| automake (GNU automake) 1.10.1
| Copyright (C) 2008 Free Software Foundation, Inc.
| License GPLv2+: GNU GPL version 2 or later 
<http://gnu.org/licenses/gpl.html>
| This is free software: you are free to change and redistribute it.
| There is NO WARRANTY, to the extent permitted by law.
|
| Written by Tom Tromey <tromey@redhat.com>
|        and Alexandre Duret-Lutz <adl@gnu.org>.
| AUTOV is 1.10
| NOTE: Executing 
autoreconf --verbose --install --force --exclude=autopoint -I/OpenEmbedded_Stuff/oetmp/staging/arm-angstrom-linux-uclibcgnueabi/share/aclocal-1.10 -I /OpenEmbedded_Stuff/oetmp/staging/arm-angstrom-linux-uclibcgnueabi/share/aclocal
| /usr/bin/autoreconf: unrecognized option `--exclude=autopoint'
| Try `/usr/bin/autoreconf --help' for more information.
| FATAL: autoreconf execution failed.
NOTE: Task 
failed: /OpenEmbedded_Stuff/oetmp/work/armv5te-angstrom-linux-uclibcgnueabi/quilt-0.45-r1/temp/log.do_configure.16497
NOTE: package quilt-0.45-r1: task do_configure: failed
ERROR: TaskFailed event exception, aborting
NOTE: package quilt-0.45: failed
ERROR: Build of 'quilt' failed


Also:
bitbake helloworld
NOTE: Psyco JIT Compiler (http://psyco.sf.net) not available. Install it to 
increase performance.
NOTE: Handling BitBake files: - (4974/4974) [100 %]
NOTE: Parsing finished. 4718 cached, 0 parsed, 256 skipped, 0 masked.
NOTE: build 200804010853: started

OE Build Configuration:
BB_VERSION     = "1.8.10"
OE_REVISION    = "a0a3cc0088bc51023a19b1d6942f7cc643e6bb52"
TARGET_ARCH    = "arm"
TARGET_OS      = "linux-uclibcgnueabi"
MACHINE        = "at91sam9263ek"
DISTRO         = "angstrom"
DISTRO_VERSION = "2008.1-test-20080401"
TARGET_FPU     = "soft"

NOTE: Resolving any missing task queue dependencies
NOTE: multiple providers are available for 
virtual/arm-angstrom-linux-uclibcgnueabi-libc-for-gcc (uclibc, 
external-toolchain);
NOTE: consider defining 
PREFERRED_PROVIDER_virtual/arm-angstrom-linux-uclibcgnueabi-libc-for-gcc
NOTE: Preparing runqueue
NOTE: Executing runqueue
NOTE: Running task 1 of 290 (ID: 
22, /OpenEmbedded_Stuff/org.openembedded.stable/packages/shasum/shasum-native.bb, 
do_prepackaged_stage)
NOTE: package shasum-native-1.0: started
ERROR: InvalidTask event exception, aborting
NOTE: package shasum-native-1.0: failed
ERROR: Build 
of /OpenEmbedded_Stuff/org.openembedded.stable/packages/shasum/shasum-native.bb 
do_prepackaged_stage failed
ERROR: Task 22 
(/OpenEmbedded_Stuff/org.openembedded.stable/packages/shasum/shasum-native.bb, 
do_prepackaged_stage) failed
NOTE: Tasks Summary: Attempted 0 tasks of which 0 didn't need to be rerun and 
1 failed.
ERROR: '/OpenEmbedded_Stuff/org.openembedded.stable/packages/shasum/shasum-native.bb' 
failed


[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 196 bytes --]

Re: stable branch checksum missmatch for quilt-0.45

[Holger Freyther]

On Tuesday 01 April 2008 15:07:34 E Robertson wrote:
> You must be the funny one.

Sometimes, I have many faces and don't hide them :)




> Since it was already built by the dev branch and I use the same local file
> to build the stable branch, would it / should it skip the build and use the
> built binary? Or would it just rebuild the binary?
> What is the command to a clean build?

Ah you build stable and dev from the same directory and have a dev built 
there? No this does not work. We recently changed the layout of staging 
directory so dev and stable are not compatible. The only bug we have there is 
we warn about 'upgrading' but not about 'downgrading' of the ABI in the 
staging directory.



> In your last point above, what is a "dep graph"?

With OE/BitBake we have tasks (e.g. do bitbake -clisttasks nano). These are 
like fetch, unpack, configure, compile, install, stage, package. E.g. to 
configure nano you will need a C library for your target, this is where 
dependencies between tasks get added. This ends up in a dep(endency) graph.


>
> Thanks.

Re: stable branch checksum missmatch for quilt-0.45

[E Robertson]

[-- Attachment #1: Type: text/plain, Size: 1625 bytes --]

On Tuesday 01 April 2008 09:48:57 am Holger Freyther wrote:
> On Tuesday 01 April 2008 15:07:34 E Robertson wrote:
> > You must be the funny one.
>
> Sometimes, I have many faces and don't hide them :)
>
> > Since it was already built by the dev branch and I use the same local
> > file to build the stable branch, would it / should it skip the build and
> > use the built binary? Or would it just rebuild the binary?
> > What is the command to a clean build?
>
> Ah you build stable and dev from the same directory and have a dev built
> there? No this does not work. We recently changed the layout of staging
> directory so dev and stable are not compatible. The only bug we have there
> is we warn about 'upgrading' but not about 'downgrading' of the ABI in the
> staging directory.

Thanks. I created a new TMPDIR and rebuild successfully.

Unless their are any objections, I'll add this to the user manual along with a 
note that the best way to do a _clean_ build is to delete tmp dir 
contents(?).


> > In your last point above, what is a "dep graph"?
>
> With OE/BitBake we have tasks (e.g. do bitbake -clisttasks nano). These are
> like fetch, unpack, configure, compile, install, stage, package. E.g. to
> configure nano you will need a C library for your target, this is where
> dependencies between tasks get added. This ends up in a dep(endency) graph.
>
> > Thanks.
>
> _______________________________________________
> Openembedded-devel mailing list
> Openembedded-devel@lists.openembedded.org
> http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-devel



[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 196 bytes --]

Re: stable branch checksum missmatch for quilt-0.45

[Holger Freyther]

On Tuesday 01 April 2008 17:52:56 E Robertson wrote:

>
> Thanks. I created a new TMPDIR and rebuild successfully.
>
> Unless their are any objections, I'll add this to the user manual along
> with a note that the best way to do a _clean_ build is to delete tmp dir
> contents(?).

Sure, this would be appreciated. You might want to attempt to come up with a 
patch to check this condition (of downgrading). 

You might want to backport changes to classes/sanity.bbclass.


take care
	z.