On 20.12.2023 12:48, Julien Grall wrote: > On 20/12/2023 11:03, Federico Serafini wrote: >> --- a/xen/arch/arm/arm64/vsysreg.c >> +++ b/xen/arch/arm/arm64/vsysreg.c >> @@ -210,8 +210,8 @@ void do_sysreg(struct cpu_user_regs *regs, >> /* RO at EL0. RAZ/WI at EL1 */ >> if ( regs_mode_is_user(regs) ) >> return handle_ro_raz(regs, regidx, hsr.sysreg.read, hsr, 0); >> - else >> - return handle_raz_wi(regs, regidx, hsr.sysreg.read, hsr, 1); >> + >> + return handle_raz_wi(regs, regidx, hsr.sysreg.read, hsr, 1); > > I don't 100% like this change (mostly because I find if/else clearer > here).
While (it doesn't matter here) my view on this is different, I'm still puzzled why the tool would complain / why a change here is necessary. It is not _one_ return statement, but there's still (and obviously) no way of falling through. > But I have the feeling any other solution would probably be > worse. Use the conditional operator? Jan > So: > > Acked-by: Julien Grall <jgr...@amazon.com> > > Cheers, >
