Re: [PATCH] perf: fix dwarf unwind using libunwind.

[Jiri Olsa]

On Tue, Jan 13, 2015 at 04:36:53PM +0800, Wang Nan wrote:
> On 2015/1/13 15:10, Jiri Olsa wrote:
> > On Tue, Jan 13, 2015 at 10:54:37AM +0800, Wang Nan wrote:
> > 
> > SNIP
> > 
> >>  
> >> +static int elf_is_exec(int fd, const char *name)
> >> +{
> >> +	Elf *elf;
> >> +	GElf_Ehdr ehdr;
> >> +	int retval = 0;
> >> +
> >> +	elf = elf_begin(fd, PERF_ELF_C_READ_MMAP, NULL);
> >> +	if (elf == NULL)
> >> +		return 0;
> >> +	if (gelf_getehdr(elf, &ehdr) == NULL)
> >> +		goto out;
> >> +
> >> +	retval = (ehdr.e_type == ET_EXEC);
> >> +
> >> +out:
> >> +	elf_end(elf);
> >> +	pr_debug("unwind: elf_is_exec(%s): %d\n", name, retval);
> >> +	return retval;
> >> +
> >> +}
> > 
> > yuou need to add stub for 'NO_LIBUNWIND_DEBUG_FRAME=1', otherwise:
> > 
> >   CC       util/dwarf-aux.o
> >   CC       util/unwind-libunwind.o
> > util/unwind-libunwind.c:188:12: error: âelf_is_execâ defined but not used [-Werror=unused-function]
> >  static int elf_is_exec(int fd, const char *name)
> >             ^
> > cc1: all warnings being treated as errors
> > make[1]: *** [util/unwind-libunwind.o] Error 1
> > make[1]: *** Waiting for unfinished jobs....
> > make: *** [all] Error 2
> > 
> > jirka
> > 
> 
> Thanks. I haven't test NO_LIBUNWIND_DEBUG_FRAME=1 case.
> 
> I found that it is possible to utilize map->dso->adjust_symbols for this propose so
> we don't need to introduce another function to check ehdr.e_type. However
> map->dso->adjust_symbols checks not only ET_EXEC but also ET_REL. What do you think?
> 
> I'd like to check it and post another patch.

please also update tests/mate

thanks,
jirka
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/