| Summary: |
Fetcher for Git submodules ('gitsm') bypasses BB_NO_NETWORK checks |
| Product: |
[Build System, Metadata & Runtime] BitBake
|
Reporter: |
Matt Hoosier <matt.hoosier> |
| Component: |
bitbake | Assignee: |
Richard Purdie <richard.purdie> |
| Status: |
RESOLVED
FIXED
|
QA Contact: |
|
| Severity: |
normal
|
|
|
| Priority: |
Medium+
|
CC: |
JPEWhacker, poky.bs.watcher, poky.watcher, randy.macleod
|
| Version: |
unspecified | |
|
| Target Milestone: |
2.6 M1 | |
|
| Hardware: |
x86 | |
|
| OS: |
Multiple | |
|
| Whiteboard: |
Backport to YP 2.4 and 2.5 |
|
OS type for building Yocto:
|
---
|
Type of Regression:
|
---
|
|
Verified:
|
|
Documentation change:
|
No (bug/feature does not impact docs)
|
| Attachments: |
|
Created attachment 4298 [details] Remember to copy fetched .git/modules into working copy clone As a consequence of accidentally not propagating the .git/modules directory (the histories of the submodules) as downloaded during the do_fetch() task into the repo cloned into ${WORKDIR}, the do_unpack() task inadvertently performs a live fetch of each submodule listed in .gitmodules in order to satisfy the 'git submodule update --init --recursive' command executed during the unpack task. The attached patch contains a proposes to fix the issue by propagating the .git/modules to the working copy clone.