Bryan Drewery
Bryan Drewery
What's the sane way to reload the completions when replacing the `_poudriere` file?
The `-j` problem is on `bulk` too. In general the ports are expected to come last. The `-o` flag for `testport` is not even needed. It's an accident of history.
Like #838, `distclean` is intended to keep everything listed and their dependencies, and delete anything not among those lists. Most likely the reason it did not work reliably was fixed...
Oh I do see the issue being reported here now. Surprised this was missed for so long.
#710 and #487 will need to be fixed for this to properly work as well.
Ugh that entire page is very stale.
Assigning to myself to document all of the internals that have been rewritten since.
Is there any sane documentation about `mkimg -p`? @davidchisnall > This is caused by the gptboot parameter not having the correct value. I have this fix locally: > > ```diff...
``` + mkimg -s gpt -b /poudriere/jails/13amd64_kern/boot/pmbr -p 'efi:=/poudriere/jails/13amd64_kern/boot/boot1.efifat' -p 'freebsd-boot:=/poudriere/jails/13amd64_kern/boot/gptboot' -p 'freebsd-ufs:=/poudrier e/data/images/poudriereimage-KdLy/img.part' -o /tmp/img/poudriereimage.img mkimg: partition 1: No such file or directory ``` ``` ls: /poudriere/jails/13amd64_kern/boot/boot1.efifat: No such...
Yeah that and a few other commits fixes it. ``` Populating `/poudriere/data/images/poudriereimage-kyh6/img.part' Image `/poudriere/data/images/poudriereimage-kyh6/img.part' complete [00:08:34] Image available at: /tmp/img/poudriereimage.img ``` I will push it out to the release branch...