Web lists-archives.com

Re: [PATCH 5/5] Makefile: disable unaligned loads with UBSan




On Sat, Jul 15, 2017 at 07:18:56PM +0200, René Scharfe wrote:

> -- >8 --
> Subject: [PATCH] Makefile: allow combining UBSan with other sanitizers
> 
> Multiple sanitizers can be specified as a comma-separated list.  Set
> the flag NO_UNALIGNED_LOADS even if UndefinedBehaviorSanitizer is not
> the only sanitizer to build with.

Nice, I didn't know you could do comma-separation here. ;)

I always just built the various sanitizers separately since I was
testing whether each one worked. But if we can get UBSan to build
cleanly, I agree that "make SANITIZE=address,undefined test" would be a
nice thing to run periodically.

> diff --git a/Makefile b/Makefile
> index ba4359ef8d..9b98535a04 100644
> --- a/Makefile
> +++ b/Makefile
> @@ -1022,10 +1022,15 @@ ifdef DEVELOPER
>  CFLAGS += $(DEVELOPER_CFLAGS)
>  endif
>  
> +comma := ,
> +empty :=
> +space := $(empty) $(empty)
> +
>  ifdef SANITIZE
> +SANITIZERS := $(foreach flag,$(subst $(comma),$(space),$(SANITIZE)),$(flag))
>  BASIC_CFLAGS += -fsanitize=$(SANITIZE) -fno-sanitize-recover=$(SANITIZE)
>  BASIC_CFLAGS += -fno-omit-frame-pointer
> -ifeq ($(SANITIZE),undefined)
> +ifneq ($(filter undefined,$(SANITIZERS)),)

The implementation is convoluted in the way that "make" usually forces
us into. ;)

I would have written it as:

  ifeq ($findstring $(SANITIZERS), undefined), undefined)

but I think your version:

  1. Isn't fooled by superstrings like a potential undefined-foo
     sanitizer.

  2. Handles SANITIZE=undefined,undefined correctly.

Neither are expected, but it doesn't hurt to be a bit more robust.

-Peff