Re: [PATCH] perf tools: Fix compilation on powerpc

[Arnaldo Carvalho de Melo]

Em Tue, Sep 28, 2021 at 10:39:05PM +0200, Mark Wielaard escreveu:
> Hi,
> 
> On Tue, Sep 28, 2021 at 09:52:53PM +0200, Jiri Olsa wrote:
> > Got following build fail on powerpc:
> > 
> >     CC      arch/powerpc/util/skip-callchain-idx.o
> >   In function ‘check_return_reg’,
> >       inlined from ‘check_return_addr’ at arch/powerpc/util/skip-callchain-idx.c:213:7,
> >       inlined from ‘arch_skip_callchain_idx’ at arch/powerpc/util/skip-callchain-idx.c:265:7:
> >   arch/powerpc/util/skip-callchain-idx.c:54:18: error: ‘dwarf_frame_register’ accessing 96 bytes \
> >   in a region of size 64 [-Werror=stringop-overflow=]
> >      54 |         result = dwarf_frame_register(frame, ra_regno, ops_mem, &ops, &nops);
> >         |                  ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> >   arch/powerpc/util/skip-callchain-idx.c: In function ‘arch_skip_callchain_idx’:
> >   arch/powerpc/util/skip-callchain-idx.c:54:18: note: referencing argument 3 of type ‘Dwarf_Op *’
> >   In file included from /usr/include/elfutils/libdwfl.h:32,
> >                    from arch/powerpc/util/skip-callchain-idx.c:10:
> >   /usr/include/elfutils/libdw.h:1069:12: note: in a call to function ‘dwarf_frame_register’
> >    1069 | extern int dwarf_frame_register (Dwarf_Frame *frame, int regno,
> >         |            ^~~~~~~~~~~~~~~~~~~~
> >   cc1: all warnings being treated as errors
> > 
> > The dwarf_frame_register args changed with [1],
> > Updating ops_mem accordingly.
> > 
> > [1] https://sourceware.org/git/?p=elfutils.git;a=commit;h=5621fe5443da23112170235dd5cac161e5c75e65
> 
> The warning (new with GCC11) and the fix are correct.

I'm taking this as an Acked-by, ok?

- Arnaldo
 
> The code probably worked fine before because the next var on the stack
> was the dummy Dwarf_Op which would have been as if the ops_mem[2] was
> actually ops_mem[3] already. You don't need the dummy var now and
> Dwarf_Op *ops doesn't need to be initialized (the dwarf_frame_register
> will initialize it).
> 
> Note that [1] didn't change the args of the dwarf_frame_register, but
> did fix a similar bug in one of the elfutils own tests (and updated
> the comment on the function to hopefully be more clear).
> 
> Cheers,
> 
> Mark
> 
> > Cc: Mark Wielaard <mjw@xxxxxxxxxx>
> > Cc: Sukadev Bhattiprolu <sukadev@xxxxxxxxxxxxxxxxxx>
> > Signed-off-by: Jiri Olsa <jolsa@xxxxxxxxxx>
> > ---
> >  tools/perf/arch/powerpc/util/skip-callchain-idx.c | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/tools/perf/arch/powerpc/util/skip-callchain-idx.c b/tools/perf/arch/powerpc/util/skip-callchain-idx.c
> > index 3018a054526a..20cd6244863b 100644
> > --- a/tools/perf/arch/powerpc/util/skip-callchain-idx.c
> > +++ b/tools/perf/arch/powerpc/util/skip-callchain-idx.c
> > @@ -45,7 +45,7 @@ static const Dwfl_Callbacks offline_callbacks = {
> >   */
> >  static int check_return_reg(int ra_regno, Dwarf_Frame *frame)
> >  {
> > -	Dwarf_Op ops_mem[2];
> > +	Dwarf_Op ops_mem[3];
> >  	Dwarf_Op dummy;
> >  	Dwarf_Op *ops = &dummy;
> >  	size_t nops;
> > -- 
> > 2.31.1
> > 

-- 

- Arnaldo