I just removed binchecks restriction from your package. That was a *nasty* QA violation: if your package have textrels and other problems you _should_ declare them in the ebuild, not restrict the sanity checks! Also please verify that the RESTRICT=strip is warranted and you're not just trying to bail out of using QA_PRESTRIPPED.
I clean ed this up and removed the stirp RESTRICT.
@@ -32,17 +32,6 @@ eclipse? ( >=dev-util/eclipse-sdk-3.4 )" DESTINATION="opt/intel/Compiler/${RELEASE}/${BUILD}" -EDESTINATION="${EROOT#/}${DESTINATION}" - -QA_TEXTRELS="${EDESTINATION}/*" -QA_EXECSTACK="${EDESTINATION}/*" -QA_PRESTRIPPED="${EDESTINATION}/lib/*/.*libFNP.so ${EDESTINATION}/bin/*/.* ${EDESTINATION}/idb/*/*/.*" -QA_DT_HASH=" - ${EDESTINATION}/bin/*/.* - ${EDESTINATION}/lib/*/.* - ${EDESTINATION}/mkl/lib/*/.* - ${EDESTINATION}/mkl/benchmarks/mp_linpack/bin_intel/*/.* - ${EDESTINATION}/idb/*/*/.*" Which comes pretty close to RESTRICT=binchecks I am not quite convinced of the utility of the QA_* in the case of the Intel libs and compilers. Upstream ignores us completely and the QA problems changes every release. Also for binary only packages in the case of QA_PRESTRIPPED, calling it a QA problem is a bit stretched.
I'm actually quite positive toward having a QA_PREBUILT_WE_CANT_DO_FUCK_ABOUT but you should talk to Zac about that :)
