Re: [PATCH 05/11] kselftest/arm64: Exit with an error code on data mismatches in fp-stress

[Lorenzo Stoakes (ARM)]

On Tue, Sep 01, 2026 at 06:06:45PM +0100, Mark Brown wrote:
> Several of the fp-stress test loads have blocks at the end of their barf
> functions with a range of commented out options for how to exit the program
> after logging an error. Comments indicate that these were intended to
> interact with some long gone debugging code in the kernel. At the minute
> the option selected by all the programs is to delivera SIGABRT to
> themselves but there is no need to do this over a normal exit with a non
> zero status. In order to facilitate running as a KVM guest replace these

Same comment as before re: bare metal. Not sure on nomenclature however!

> Signed-off-by: Mark Brown <broonie@xxxxxxxxxx>

That was some hairy existing code :)

All seems sensible so:

Reviewed-by: Lorenzo Stoakes (ARM) <ljs@xxxxxxxxxx>

One nit - there's still a:

function barf
// fpsimd.c acitivty log dump hack
//	ldr	w0, =0xdeadc0de
//	mov	w8, #__NR_exit
//	svc	#0
// end hack

That sticks around in the files, should those be removed also?

--
Cheers, Lorenzo