Hi,

On Tue, Feb 27, 2024 at 12:33:15PM +0000, Richard Purdie wrote:
> On Tue, 2024-02-27 at 14:23 +0200, Mikko Rapeli wrote:
> > On Tue, Feb 27, 2024 at 12:06:41PM +0000, Richard Purdie wrote:
> > > Should we break it for others so it works for you?
> > > 
> > > If I merge this patch and then I get other reports in of issues,
> > > I'm
> > > going to be left to somehow fix those other problems. Since your
> > > case
> > > will work you will object if I revert this but you likely won't
> > > have
> > > time to work on the resulting other issues, or be able to reproduce
> > > them since you're building on zfs.
> > > 
> > > So you can see my dilemma? :/
> > > 
> > > Most people don't build on zfs which is probably what I need to
> > > balance
> > > this on.
> > 
> > What if I make this optional for zfs? If build directory type is zfs,
> > then --apparent-size is added, else not.
> 
> Wouldn't that cause some really interesting bug reports? 

Well, if anyone builds on zfs they will hit bugs like this.
zfs usage can be detected and build can fail nicely or apply
the workaround/fix in wic.

> Asking the user to report their host system filesystem type to debug
> anything isn't going to be great.
> 
> I don't like the idea of the non-determinism that introduces but I
> agree we do already have a determinism issue.
> 
> What we probably need is the absolute files size, plus a number of
> files and a number of directories, then you can round up to the worst
> cases for the block size.
> 
> Currently that size number is just including the block sizes.

FWIW, I tried to move this workaround into image recipe with

IMAGE_CMD:wic:prepend() {
    alias df="df --apparent-size"
}

but wic python code doesn't seem to take this into account. Tried also
changing the du calls to have exec_cmd(..., as_shell=True) but this didn't
help either. Something in wic causes the use of another shell or different 
default
environment for processes.

So I'm stuck with this change in poky to build on a machine with zfs.

Any hints how to apply this simple workaround are welcome.

Cheers,

-Mikko
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#196417): 
https://lists.openembedded.org/g/openembedded-core/message/196417
Mute This Topic: https://lists.openembedded.org/mt/104600633/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub 
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-

Reply via email to