Open GHF opened 5 years ago
Hmmm, interesting —I wonder if this same issue might be why I can't seem to get the sanitisers provided by this package to fail my CI builds? https://github.com/saxbophone/arby/pull/153
I am also using target_sources()
, although IIRC I'm not getting a build warning or error, and when compiling it does appear to be using libasan/ubsan, but I can't seem to deliberately fail the build by introducing undefined behaviour into my code:
int k = 0x7fffffff;
k += 1; // cause integer overflow
which makes this pretty useless for actually detecting any UB...
With CMake 3.13, I've been writing a bunch of targets like this, where the
SOURCES
property of a target can be added by subdirectories and its evaluation is deferred to "generation time."Unfortunately, this makes these targets incompatible with
add_sanitizers(...)
because the scripts are not able to find any source files in the target:I'm pretty new to "modern" CMake and I don't know how to get around this. Maybe the compiler flags added by
sanitizer_add_flags
need to be generator expressions that get evaluated at generation time?