https://bugs.kde.org/show_bug.cgi?id=468575
--- Comment #21 from Petr Pavlu <petr.pa...@dagobah.cz> --- (In reply to Paul Floyd from comment #20) > And regtest results > > == 737 tests, 6 stderr failures, 2 stdout failures, 0 stderrB failures, 1 > stdoutB failure, 0 post failures == > gdbserver_tests/hgtls (stdoutB) The hgtls test fails because GDB is confused by missing DW_AT_location attributes for TLS variables. It looks to be a problem with the RISC-V target in GCC not implementing TARGET_ASM_OUTPUT_DWARF_DTPREL. I have a preliminary patch for GCC which adds it. > [...] > none/tests/riscv64/compressed (stdout) > none/tests/riscv64/compressed (stderr) > none/tests/riscv64/integer (stdout) > none/tests/riscv64/integer (stderr) > > [...] > The two riscv64 fails look like a problem with the toolchain. I'll see if > trying a local build of the toolchain helps. The mentioned compilation failure of RISC-V tests is some problem related to them being compiled as position-independent, which is apparently the default on Debian. It is possible for me to reproduce it locally. I'll have a closer look. -- You are receiving this mail because: You are watching all bug changes.
