On Thu, Jul 08, 2021 at 09:18:44PM +0200, Matteo Croce wrote:
> On Thu, 20 May 2021 09:47:58 +0100
> Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx> wrote:
>
> > make W=1 generates the following warning for mm/page_alloc.c
> >
> > mm/page_alloc.c:3651:15: warning: no previous prototype for
> > ???should_fail_alloc_page??? [-Wmissing-prototypes] noinline bool
> > should_fail_alloc_page(gfp_t gfp_mask, unsigned int order)
> > ^~~~~~~~~~~~~~~~~~~~~~
> >
> > This function is deliberately split out for BPF to allow errors to be
> > injected. The function is not used anywhere else so it is local to
> > the file. Make it static which should still allow error injection
> > to be used similar to how block/blk-core.c:should_fail_bio() works.
> >
> > Signed-off-by: Mel Gorman <mgorman@xxxxxxxxxxxxxxxxxxx>
> > ---
> > mm/page_alloc.c | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/mm/page_alloc.c b/mm/page_alloc.c
> > index aaa1655cf682..26cc1a4e639b 100644
> > --- a/mm/page_alloc.c
> > +++ b/mm/page_alloc.c
> > @@ -3648,7 +3648,7 @@ static inline bool
> > __should_fail_alloc_page(gfp_t gfp_mask, unsigned int order)
> > #endif /* CONFIG_FAIL_PAGE_ALLOC */
> >
> > -noinline bool should_fail_alloc_page(gfp_t gfp_mask, unsigned int
> > order) +static noinline bool should_fail_alloc_page(gfp_t gfp_mask,
> > unsigned int order) {
> > return __should_fail_alloc_page(gfp_mask, order);
> > }
>
>
> Hi Mel,
>
> It seems that this breaks builds with CONFIG_DEBUG_INFO_BTF=y.
> Maybe that warning was a false positive because
> should_fail_alloc_page() is referenced via a macro?
>
> I proposed to revert it, feel free to propose another fix.
>
The alternative fix of making the symbol global was rejected. eBPF needs
to figure out a way of instrumenting code that is is unused by the kernel
and not globally visible but I don't know how that might be achieved.
--
Mel Gorman
SUSE Labs
