| Summary: | runqemu whitelists images | ||
|---|---|---|---|
| Product: | [Build System, Metadata & Runtime] OE-Core | Reporter: | Ross Burton <ross.burton> |
| Component: | deployment | Assignee: | Scott Garman <scott.a.garman> |
| Status: | VERIFIED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | Medium | CC: | alexandrux.palalau, dvhart |
| Version: | 1.5 | ||
| Target Milestone: | 1.6 M2 | ||
| Hardware: | x86 | ||
| OS: | Multiple | ||
| Whiteboard: | |||
| OS type for building Yocto: | --- | Type of Regression: | --- |
| Verified: | Documentation change: | No (bug/feature does not impact docs) | |
No argument from me. These sorts of things were meant as sanity checks to help ensure it did sensible things, but for me they have become more of a pain than the script is worth, and I tend to just run qemu manually. So it sounds like the desired logic is to pick the most recently modified file which matches the MACHINE's default filesystem's filename extension? I haven't fixed an oe-core bug in a while, I wouldn't mind taking care of this. Yeah, that should work. Images are in tmp/deploy/[$MACHINE]/ now too which helps when filtering the images. Fixed as of: http://git.openembedded.org/openembedded-core/commit/?id=9f69e00200cdbd5ba2e46a54f33c29797816e43f Verified on master. Commit ID: 6ee5d95317de77c1cfb6db6f1ab2de77f5fa085e |
If I use runqemu with just a machine name ("runqemu qemux86-64") it will find the most recent image *that matches a whitelist*. This whitelist is rather small: case $name in *core-image-sato* | \ *core-image-lsb* | \ *core-image-basic* | \ *core-image-minimal* ) ROOTFS=$name I just spent ten minutes getting very confused as I thought I was booting my fresh core-image-base, not my week-old core-image-sato-base. Why does runqemu even need to filter the list of rootfs in the deploy directory? If it really does need to, can it use a blacklist instead of a whitelist?