You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It's perhaps one of the most critical bug ever noticed for this project.
I think it impacts results of 4.15, but it can be the case for 4.13.3 as well (for recent compilations)
In the last entries of the database I've observed that CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y is set instead of CONFIG_CC_OPTIMIZE_FOR_SIZE=y.
I don't know why!
This piece of code is basically rewriting the whole .config file with the original content that we read earlier. To be sure that is from here that the bug come, i suggest that instead of reading the content, we can copy the .config file in other place, do our stuff and copying it back. If the bug remains, it should not come from here.
It should restore the compiled kernel to its earlier states, since it have been changed to use a compression. So yes, it recompile, but just in the later stage, where the option for compression algorithm kick in.
It's perhaps one of the most critical bug ever noticed for this project.
I think it impacts results of 4.15, but it can be the case for 4.13.3 as well (for recent compilations)
In the last entries of the database I've observed that
CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y
is set instead ofCONFIG_CC_OPTIMIZE_FOR_SIZE=y
.I don't know why!
Basically, I have impression we are sending the wrong configuration file due to computation of several kernel sizes.
The code here is suspicious:
https://github.com/TuxML/ProjetIrma/blob/dev/compilation/compiler.py#L320-L331
and something is forcing the value of
CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE=y
(see also TuxML/ProjetIrma#169 and the possible options impacted)
The text was updated successfully, but these errors were encountered: