On Mon, 28 Jan 2019, Vineet Gupta wrote: > In the failing case, the dsos are NOT getting the DT_INIT/DT_FINI > dynamic tags And they shouldn't be (they should have DT_INIT_ARRAY / DT_FINI_ARRAY instead). > (1). gcc config test failing (false alaram) > > | host-gcc-final-ab544139bfee/build/gcc/config.log > | > |configure:22941: checking for .preinit_array/.init_array/.fini_array support > |configure:23106: checking cross compile... guessing > |configure:23111: result: no > > gcc was seemingly built w/o init_array support leading to inclusion of default > gcc/config/initfini-array.h > However doing a glibc aarch64 build I see the same output so perhaps it is not > relevant. This was discussed at length in the C-Sky discussion. In short: the proper fix is to change that configure test to apply for cross compilation as well, other than in the very limited case where it tries to run something (but doing so is not appropriate in the current regression-fixes-only GCC development stage). You can hack around it with gcc_cv_initfini_array=yes for the target in config.gcc, as done by a few architectures. Or you can define the relevant target macros in gcc/config/<arch>/<something>.h to force init_array / fini_array use independent of the configure test, as done for AArch64. -- Joseph S. Myers joseph@xxxxxxxxxxxxxxxx _______________________________________________ linux-snps-arc mailing list linux-snps-arc@xxxxxxxxxxxxxxxxxxx http://lists.infradead.org/mailman/listinfo/linux-snps-arc
