Skip to contentSkip to conversation

bitbake -c clean <package> fails

DISCUSSION Ulf Samuelsson 5 participants 8 messages Jan 29, 2012 – Feb 8, 2012

lore
Share & tools
lore
#0 lore raw

Tried to test downloading "module-init-tools" from a new KERNELORG_MIRROR

First I attempted to just remove the source tarball from the downloads directory. No download, but build completed.

The I tried to clean out the "module-init-tools" directory:

$ bitbake -c clean module-init-tools

and run bitbake console-image.
The build completed with no source tarball, after the clean.

Only when I cleaned out the build/sstate-cache directory, was the file downloaded and built from a new tarball

Should not "bitbake -c clean" remove files in the cache?

If not, Is there another bitbake command which cleans out the cache?

Signature · 3 lines
Best Regards
Ulf Samuelsson
eMagii
#1 lore raw

Op 29 jan. 2012, om 14:56 heeft Ulf Samuelsson het volgende geschreven:

Quoted 12 lines

Tried to test downloading "module-init-tools" from a new KERNELORG_MIRROR

First I attempted to just remove the source tarball from the downloads directory. No download, but build completed.

The I tried to clean out the "module-init-tools" directory:

$ bitbake -c clean module-init-tools

and run bitbake console-image.
The build completed with no source tarball, after the clean.

Quoted 6 lines

Only when I cleaned out the build/sstate-cache directory, was the file downloaded and built from a new tarball

Should not "bitbake -c clean" remove files in the cache?

If not, Is there another bitbake command which cleans out the cache?

-c cleansstate

#2 lore raw
In reply to Ulf Samuelsson
Quoted 4 lines · Ulf Samuelsson

First I attempted to just remove the source tarball from the downloads directory.
No download, but build completed.

if you want a fresh fetch.
you can just do bitbake -ccleanall module-init-tools that will remove the sstate as well as the sources

#3 lore raw
Quoted 5 lines · Khem Raj
Quoted 3 lines · Ulf Samuelsson

First I attempted to just remove the source tarball from the downloads directory.
No download, but build completed.

Quoted 8 lines

if you want a fresh fetch.
you can just do bitbake -ccleanall module-init-tools that will remove the sstate as well as the sources

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Clean it's just a "clean". As in "make clean". So after a clean you will typically have configure, compile etc. "cleanall" = remove unpacked + remove source. So after a clean you will tipically have fetch.

#4 lore raw

2012-01-30 10:48, Andrei Gherzan skrev:

Quoted 10 lines · Khem Raj
Quoted 9 lines
Quoted 3 lines · Ulf Samuelsson

First I attempted to just remove the source tarball from the downloads directory.
No download, but build completed.

if you want a fresh fetch.
you can just do bitbake -ccleanall module-init-tools that will remove the sstate as well as the sources

Quoted 7 lines
Quoted 4 lines

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Clean it's just a "clean". As in "make clean". So after a clean you will typically have configure, compile etc. "cleanall" = remove unpacked + remove source. So after a clean you will tipically have fetch.

I have tried clean, cleansstate and cleanall.
All of them remove the source directory.

If I try to use XorA's method of testing out the kernel
http://www.xora.org.uk/2009/12/10/openembeddedangstrom-kernel-workflow/
then it fails since a recompile of the source will fail to update dependencies. If i bitbake -c compile virtual/kernel -f a
and then
bitbake virtual/kernel
then the previous build will be used.

Is it at all possible now to recompile, without having to clean out everything, and start from scratch?

Quoted 5 lines

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

#5 lore raw
Quoted 6 lines
Quoted 4 lines

Clean it's just a "clean". As in "make clean". So after a clean you will typically have configure, compile etc. "cleanall" = remove unpacked + remove source. So after a clean you will tipically have fetch.

I have tried clean, cleansstate and cleanall.
All of them remove the source directory.

if you are talking about ${S} then yes but relevant content in ${DL_DIR} are only removed with cleanall

Quoted 25 lines

If I try to use XorA's method of testing out the kernel
http://www.xora.org.uk/2009/12/10/openembeddedangstrom-kernel-workflow/
then it fails since a recompile of the source will fail to update dependencies.
If i bitbake -c compile virtual/kernel -f a
and then
bitbake virtual/kernel
then the previous build will be used.

Is it at all possible now to recompile, without having to clean out everything,
and start from scratch?

Quoted 5 lines

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Signature · 1 line
-Khem
#6 lore raw
Quoted 9 lines · Khem Raj
Quoted 6 lines
Quoted 4 lines

Clean it's just a "clean". As in "make clean". So after a clean you will typically have configure, compile etc. "cleanall" = remove unpacked + remove source. So after a clean you will tipically have fetch.

I have tried clean, cleansstate and cleanall.
All of them remove the source directory.

if you are talking about ${S} then yes but relevant content in ${DL_DIR} are only removed with cleanall

Yes, when I work on the kernel, I do not want to restart from unpack. I would like to edit the existing ${S}and then restart a compile. Once the compile succeeds, I want to deploy.
Once the kernel does what it is supposed to, I generate a patch which is then added to the recipe.

At that time, I want to cleansstate and rebuild/retest.
To come around this problem right now, my kernel recipe generates a script in ${S}, which will compile the kernel image and then copies it directly to my tftp directory.

Quoted 22 lines
Quoted 22 lines

If I try to use XorA's method of testing out the kernel
http://www.xora.org.uk/2009/12/10/openembeddedangstrom-kernel-workflow/
then it fails since a recompile of the source will fail to update dependencies.
If i bitbake -c compile virtual/kernel -f a
and then
bitbake virtual/kernel
then the previous build will be used.

Is it at all possible now to recompile, without having to clean out everything,
and start from scratch?

Quoted 4 lines

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

_______________________________________________
Openembedded-core mailing list
[email protected]
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

#7 lore raw
Quoted 17 lines · Ulf Samuelsson
Quoted 9 lines · Khem Raj
Quoted 6 lines
Quoted 4 lines

Clean it's just a "clean". As in "make clean". So after a clean you will typically have configure, compile etc. "cleanall" = remove unpacked + remove source. So after a clean you will tipically have fetch.

I have tried clean, cleansstate and cleanall.
All of them remove the source directory.

if you are talking about ${S} then yes but relevant content in ${DL_DIR} are only removed with cleanall

Yes, when I work on the kernel, I do not want to restart from unpack. I would like to edit the existing ${S}and then restart a compile. Once the compile succeeds, I want to deploy.
Once the kernel does what it is supposed to, I generate a patch which is then added to the recipe.

Quoted 4 lines

At that time, I want to cleansstate and rebuild/retest.
To come around this problem right now, my kernel recipe generates a script in ${S}, which will compile the kernel image and then copies it directly to my tftp directory.

Why don't you play along with sstate files? Detele what you need and it will be forced to re<stage>.

@g