On Thu, Jan 04, 2024 at 11:03:42AM +0100, Alexander Gordeev wrote: > On Tue, Jan 02, 2024 at 04:05:31PM +0100, Heiko Carstens wrote: > Hi Heiko, > ... > > > @@ -253,9 +253,17 @@ static unsigned long setup_kernel_memory_layout(void) > > > MODULES_END = round_down(__abs_lowcore, _SEGMENT_SIZE); > > > MODULES_VADDR = MODULES_END - MODULES_LEN; > > > VMALLOC_END = MODULES_VADDR; > > > +#ifdef CONFIG_KMSAN > > > + VMALLOC_END -= MODULES_LEN * 2; > > > +#endif > > > > > > /* allow vmalloc area to occupy up to about 1/2 of the rest virtual space left */ > > > vmalloc_size = min(vmalloc_size, round_down(VMALLOC_END / 2, _REGION3_SIZE)); > > > +#ifdef CONFIG_KMSAN > > > + /* take 2/3 of vmalloc area for KMSAN shadow and origins */ > > > + vmalloc_size = round_down(vmalloc_size / 3, _REGION3_SIZE); > > > + VMALLOC_END -= vmalloc_size * 2; > > > +#endif > > > > Please use > > > > if (IS_ENABLED(CONFIG_KMSAN)) > > > > above, since this way we get more compile time checks. > > This way we will get a mixture of CONFIG_KASAN and CONFIG_KMSAN > #ifdef vs IS_ENABLED() checks within one function. I guess, we > would rather address it with a separate cleanup? I don't think so, since you can't convert the CONFIG_KASAN ifdef to IS_ENABLED() here: it won't compile. But IS_ENABLED(CONFIG_KMSAN) should work. I highly prefer IS_ENABLED() over ifdef since it allows for better compile time checks, and you won't be surprised by code that doesn't compile if you just change a config option. We've seen that way too often.
