Build my own Vix image

[twol]

Yes, exactly that, thank you. First I built the image, then when it finished I added DISTRO_FEED_URI and finally I built the feeds. Next time I'll run 'make image' so it builds image and feeds in one go.

OK thats the issue in that you added the override after building the image …. So ag that point /etc/opkg is built

 

[Huevos]

Only way around that change is clean the recipe and rebuild.

 

[Huevos]

Mutant HD51 and AX61 now have official builds.

 

[Mickkie]

Mutant HD51 and AX61 now have official builds.

Thanks Huevos, a couple of users sent PMs to me to try my image, but I hadn't noticed them until today. Apologies for any late responses. I would of course be more than happy to share my image, but there's no longer a need for it since there is now a formal build.

 

[twol]

Thanks Huevos, a couple of users sent PMs to me to try my image, but I hadn't noticed them until today. Apologies for any late responses. I would of course be more than happy to share my image, but there's no longer a need for it since there is now a formal build.

…but congratulations in your ability and motivation to produce an image

 

[Mickkie]

…but congratulations in your ability and motivation to produce an image

Why thank you, but after all the noise I generated in this thread I think you're being too kind!

 

[BrianG61UK]

Lately, whenever I start from a fresh copy of the build-environment and do a build for my Zgemma H7S I get two red Taskhash mismatch errors similar to these in the middle of the process:

ERROR: network-usb-drivers-meta-1.0-r1 do_package: Taskhash mismatch 109a036cbc5a1c4486731744195eaf21ec2da779c58ddbbe0136c82e80d8f53c versus bfc26dcbd496d6274d2d85ed188c08a4abb7b5eba2be4824266839135a0ee0e2 for /home/briang/oe-alliance/build-enviroment/meta-oe-alliance/meta-oe/recipes-oe-alliance/enigma2-network-drivers/network-usb-drivers-meta.bb:do_package
ERROR: Taskhash mismatch 109a036cbc5a1c4486731744195eaf21ec2da779c58ddbbe0136c82e80d8f53c versus bfc26dcbd496d6274d2d85ed188c08a4abb7b5eba2be4824266839135a0ee0e2 for /home/briang/oe-alliance/build-enviroment/meta-oe-alliance/meta-oe/recipes-oe-alliance/enigma2-network-drivers/network-usb-drivers-meta.bb:do_package

related to network-usb-drivers-meta.bb.

I'm used to Red errors stopping the build completely, but after these it carries on to the end, and everything appears, to me, to work fine.
However, since I am not using anything that would be built by network-usb-drivers-meta.bb, I do wonder if just that part failed?

I guess that, at least as far as I'm concerned, nothing needs to be done about it. I just thought I'd mention it.

 

[birdman]

Lately, whenever I start from a fresh copy of the build-environment and do a build for my Zgemma H7S I get two red Taskhash mismatch errors similar to these in the middle of the process:

Yes. I've mentioned this too.
In my view there is something wrong here, as :

• You shouldn't be getting a taskhash mismatch on a first build of anything.
• Errors should stop the build (but these ones do not)

 

[Huevos]

They don't stop the build because IGNORE_TASKHASH_MISMATCH is set.

 

[birdman]

They don't stop the build because IGNORE_TASKHASH_MISMATCH is set.

Thanks.
But I'd much prefer that whatever causes it were fixed, rather than just ignored.

 

[Huevos]

Probably date or time used in package version.

 

[lincsat]

I remember a similar Taskhash error some time ago. I believe it was down to a date or time irregularity but can't remember how I fixed it or if it just went away.

 

[birdman]

Probably date or time used in package version.

So a bug.

 

[el bandido]

You want to waste time trying to find this???

I have had similar happen in the build, panicked and deleted tmp only to find it reappear.
Consider how many packages are being built along with the massive amount of steps...Something will happen!
I think you may find, during the first build after a cleaning:
1. BitBake calculates the initial taskhash based on metadata
2. As dependencies build, variables resolve, and the environment stabilizes
3. BitBake recalculates and notices the hash changed
4. It rebuilds the task with the correct hash (that's why it shows in red but continues)
5. The corrected hash is now cached

The -meta suffix is the clue. This is a meta-package (or bundle recipe) that
doesn't build anything itself. It just pulls in other driver packages via RDEPENDS or RRECOMMENDS
Probably a convenient way to install multiple USB network drivers at once.

 

[Huevos]

So a bug.

So just use [vardepsexclude] to exclude the problematic variable from the taskhash. I hope you have plenty of spare time.

 

[birdman]

You want to waste time trying to find this???

Any error is reporting something that should not have happened.
So something is wrong.

Fixing it would not be a waste of time (unless you think that its OK to have errors in your code).
It's fairly new, so is caused by a recent(ish) change.

 

[el bandido]

The error probably exists in your build machine.

 

[birdman]

The error probably exists in your build machine.

No, since others get the same error reported.

 

[birdman]

They don't stop the build because IGNORE_TASKHASH_MISMATCH is set.

If you mean:

BB_HASH_IGNORE_MISMATCH = "1"​

then that applies to all recipes, which means that there is no point in ever checking taskhashes at all.
Rather defeats the point.

 

[Huevos]

Yes, it is global. I build/rebuild 50 boxes frequently and do not see these errors, so I suggest you and the others reporting this look at your build environment.