Re: [PATCH] [v2] arm64: fix unreachable code issue with cmpxchg

[Nathan Chancellor]

On Wed, Sep 18, 2019 at 10:02:41AM +0100, Will Deacon wrote:
> On Tue, Sep 17, 2019 at 01:34:25PM -0700, Nathan Chancellor wrote:
> > On Tue, Sep 10, 2019 at 01:56:22PM +0200, Arnd Bergmann wrote:
> > > On arm64 build with clang, sometimes the __cmpxchg_mb is not inlined
> > > when CONFIG_OPTIMIZE_INLINING is set.
> > > Clang then fails a compile-time assertion, because it cannot tell at
> > > compile time what the size of the argument is:
> > > 
> > > mm/memcontrol.o: In function `__cmpxchg_mb':
> > > memcontrol.c:(.text+0x1a4c): undefined reference to `__compiletime_assert_175'
> > > memcontrol.c:(.text+0x1a4c): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `__compiletime_assert_175'
> > > 
> > > Mark all of the cmpxchg() style functions as __always_inline to
> > > ensure that the compiler can see the result.
> > > 
> > > Acked-by: Nick Desaulniers <ndesaulniers@xxxxxxxxxx>
> > > Reported-by: Nathan Chancellor <natechancellor@xxxxxxxxx>
> > > Link: https://github.com/ClangBuiltLinux/linux/issues/648
> > > Reviewed-by: Nathan Chancellor <natechancellor@xxxxxxxxx>
> > > Tested-by: Nathan Chancellor <natechancellor@xxxxxxxxx>
> > > Reviewed-by: Andrew Murray <andrew.murray@xxxxxxx>
> > > Tested-by: Andrew Murray <andrew.murray@xxxxxxx>
> > > Signed-off-by: Arnd Bergmann <arnd@xxxxxxxx>
> > > ---
> > > v2: skip unneeded changes, as suggested by Andrew Murray
> > > ---
> > >  arch/arm64/include/asm/cmpxchg.h | 6 +++---
> > >  1 file changed, 3 insertions(+), 3 deletions(-)
> > > 
> > > diff --git a/arch/arm64/include/asm/cmpxchg.h b/arch/arm64/include/asm/cmpxchg.h
> > > index a1398f2f9994..f9bef42c1411 100644
> > > --- a/arch/arm64/include/asm/cmpxchg.h
> > > +++ b/arch/arm64/include/asm/cmpxchg.h
> > > @@ -62,7 +62,7 @@ __XCHG_CASE( ,  ,  mb_, 64, dmb ish, nop,  , a, l, "memory")
> > >  #undef __XCHG_CASE
> > >  
> > >  #define __XCHG_GEN(sfx)							\
> > > -static inline unsigned long __xchg##sfx(unsigned long x,		\
> > > +static __always_inline  unsigned long __xchg##sfx(unsigned long x,	\
> > >  					volatile void *ptr,		\
> > >  					int size)			\
> > >  {									\
> > > @@ -148,7 +148,7 @@ __CMPXCHG_DBL(_mb)
> > >  #undef __CMPXCHG_DBL
> > >  
> > >  #define __CMPXCHG_GEN(sfx)						\
> > > -static inline unsigned long __cmpxchg##sfx(volatile void *ptr,		\
> > > +static __always_inline unsigned long __cmpxchg##sfx(volatile void *ptr,	\
> > >  					   unsigned long old,		\
> > >  					   unsigned long new,		\
> > >  					   int size)			\
> > > @@ -255,7 +255,7 @@ __CMPWAIT_CASE( ,  , 64);
> > >  #undef __CMPWAIT_CASE
> > >  
> > >  #define __CMPWAIT_GEN(sfx)						\
> > > -static inline void __cmpwait##sfx(volatile void *ptr,			\
> > > +static __always_inline void __cmpwait##sfx(volatile void *ptr,		\
> > >  				  unsigned long val,			\
> > >  				  int size)				\
> > >  {									\
> > > -- 
> > > 2.20.0
> > > 
> > 
> > Looks like the arm64 pull request happened without this patch so clang
> > all{mod,yes}config builds are broken. Did the maintainers have any
> > further comments on it or could this make it in with the next one?
> 
> Fear not! I plan to send this with some other fixes we've got for -rc1.
> I just to get my CI scripts going again (new machine), but that shouldn't
> take long.
> 
> Will

Great, thank you!

Cheers,
Nathan