| Summary: | recipe checksum bypass bug | ||
|---|---|---|---|
| Product: | [Build System, Metadata & Runtime] BitBake | Reporter: | akuster <akuster808> |
| Component: | bitbake | Assignee: | Richard Purdie <richard.purdie> |
| Status: | RESOLVED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | Medium | CC: | akuster808, dorindabassey, meta.mr.watcher, meta.watcher, poky.bs.watcher, poky.watcher, randy.macleod, weaverjs |
| Version: | 3.3 | ||
| Target Milestone: | 3.4 | ||
| Hardware: | x86 | ||
| OS: | Multiple | ||
| Whiteboard: | NEWCOMER | ||
| OS type for building Yocto: | --- | Type of Regression: | --- |
| Verified: | Documentation change: | Don't know | |
|
Description
akuster
2021-02-10 04:23:06 UTC
Dorinda, do you want to work on this, check for 'None' and error. Discuss with the list I guess. Yes, I would love to work on this thanks. I just reproduced this on a recipe, and from what I understood here a check in bitbake __init__.py might suffice, let me think of a solution. Hi All, I noticed that they're some inconsistencies when this variable is set to empty string or commented out in some recipes as such: SRC_URI[sha256sum] = "" SRC_URI[md5sum] = "" or #SRC_URI[sha256sum] = "" #SRC_URI[md5sum] = "" sometimes the expected error message is displayed after running the build, but when i run the build again the error disappears. i don't know what could be causing this inconsistency. Can you list the examples that you are concerned about? (In reply to comment #4) > Can you list the examples that you are concerned about? who is the question for? Okay, I tested this for rsync and elfutils recipes. The question was for Dorinda about her statement: "this variable is set to empty string or commented out in some recipes" but now I understand that she's saying that if she changes a recipe to have an empty SRC_URI[*sum] then she's getting different results on the first and second run of bitbake. Anyway, I think that for the sha256sum at least there are not any recipes where an empty string is valid and that should result in an error. Hello, I haven’t contributed before but I saw the newcomers post and thought I would take a look and see if I can give a little back to the project. If this issue is unresolved and unassigned then I believe I have a solution to it. I know that in previous comments Dorinda was going to look into this and if that’s still the case then I’ll look at the other issues. It's fine, dig-in your contributions will be much appreciated. Thanks, Dorinda. I've submitted a patch to the bitbake-devel list. |