tools: fixup the CMake custom command for elf-cleaner to correctly find sources #54
Conversation
…nd sources Signed-off-by: Tom Turner <433367+tturner@users.noreply.github.com>
tturner
force-pushed
the
fix-elf-cleaner-build
branch
from
17312ab
to
3f53a0a
Compare
|
I haven't run into a build error yet but this is a good change even for consistency sake, merged as 1ba84e0 |
|
Wondering how to repro this. Do you run gradle from a different CWD or something? |
|
I didn't think I was doing anything particularly unusual. Fairly vanilla Debian 12 (no custom shell). Installed latest beta? of the android studio (Hedgehog) because that seemed to be needed for the latest commits on master (I got errors relating to the agp alpha version when I tried the normal android studio release). I had that download the android SDK and then used ANDROID_HOME env to point to the sdk directory. I then ran the command from the readme in the base directory of the cloned repo. I'm not sure why it is different from your experience but the general recommendation is to make cmake builds agnostic to the folder they are built in and the old version of that cmake command assumed that the custom command was running in the source folder which isn't guaranteed by cmake. |
|
Well, there are two other GLOBs in there that aren't doing this. So that's a bit puzzling. I think I'd like to see some analysis of what's happening here. AFAICT, what was there before should be correct already and seems standard practice. |
|
No reply from you or further research so I'm going to revert the commit until the right time. |
zx2c4-bot
force-pushed
the
master
branch
4 times, most recently
from
827495b
to
4ba8794
Compare
This is a tiny change to allow elf-cleaner to build correctly during assembleRelease.
Without this I got failures to find these files with the latest Android Studio (and on the command line) - the glob found no files so the build command was malformed.