| Message ID | 4a0f9e7d-53f1-b77f-e8a9-a75483884a6f@suse.com (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | x86emul: avoid assembler warning about .type not taking effect in test harness | expand |
On 07.07.2020 11:35, Jan Beulich wrote: > gcc 9.3 started to re-order top level blocks by default when optimizing. > This re-ordering results in all our .type directives to get emitted to > the assembly file first, followed by gcc's. The assembler warns about > attempts to change the type of a symbol when it was already set (and > when there's no intervening setting to "notype"). Turns out this wasn't a gcc change - the problem had been there all the time, it just went through silently. It was the newer gas that I built gcc 9.3 with that caused to issue to become visible. I've slightly updated the description to account for this. Jan
--- a/tools/tests/x86_emulator/Makefile +++ b/tools/tests/x86_emulator/Makefile @@ -295,4 +295,9 @@ x86-emulate.o cpuid.o test_x86_emulator. x86-emulate.o: x86_emulate/x86_emulate.c x86-emulate.o: HOSTCFLAGS += -D__XEN_TOOLS__ +# In order for our custom .type assembler directives to reliably land after +# gcc's, we need to keep it from re-ordering top-level constructs. +$(call cc-option-add,HOSTCFLAGS-toplevel,HOSTCC,-fno-toplevel-reorder) +test_x86_emulator.o: HOSTCFLAGS += $(HOSTCFLAGS-toplevel) + test_x86_emulator.o: $(addsuffix .h,$(TESTCASES)) $(addsuffix -opmask.h,$(OPMASK))
gcc 9.3 started to re-order top level blocks by default when optimizing. This re-ordering results in all our .type directives to get emitted to the assembly file first, followed by gcc's. The assembler warns about attempts to change the type of a symbol when it was already set (and when there's no intervening setting to "notype"). Signed-off-by: Jan Beulich <jbeulich@suse.com>