[Bug gas/976] Superfluous 0 displacment added in x86 code
These explicit zero displacements are littered through header/include files
for clarity purposes. In 20+ years of x86 assembler programming, I've never
seen anyone use this as a mechanism for self-modifying code...
If given a choice, I'd opt for emulating the encoding behaviors of MASM/TASM
(which eliminate the superfluous zero)