Hi Frank,
On Wed, 2019-10-30 at 14:11 -0400, Frank Ch. Eigler wrote:
> Revised code to look like this (libdwfl/find-debuginfo.c):
>
> 00404 /* NB: this is slightly thread-unsafe */
> 00405 static __typeof__ (debuginfod_find_debuginfo)
> *fp_debuginfod_find_debuginfo;
> 00406
> 00407 if (fp_debuginfod_find_debuginfo == NULL)
> 00408 {
> 00409 void *debuginfod_so = dlopen("libdebuginfod-" VERSION ".so",
> RTLD_LAZY);
> 00410 if (debuginfod_so == NULL)
> 00411 debuginfod_so = dlopen("libdebuginfod.so", RTLD_LAZY);
> 00412 if (debuginfod_so != NULL)
> 00413 fp_debuginfod_find_debuginfo = dlsym (debuginfod_so,
> "debuginfod_find_debuginfo");
> 00414 if (fp_debuginfod_find_debuginfo == NULL)
> 00415 fp_debuginfod_find_debuginfo = (void *) -1; /* never try
> again */
> 00416 }
> 00417
> 00418 if (fp_debuginfod_find_debuginfo != NULL &&
> fp_debuginfod_find_debuginfo != (void *) -1)
> 00419 {
> 00420 /* If all else fails and a build-id is available, query the
> 00421 debuginfo-server if enabled. */
> 00422 if (fd < 0 && bits_len > 0)
> 00423 fd = (*fp_debuginfod_find_debuginfo) (bits, bits_len, NULL);
> 00424 }
Thanks, that looks perfectly fine for now. This should also be not too
hard to make thread-safe when using atomics.
> > > We document returning standard errnos generally, and a few particular
> > > ones for network unreachability.
> >
> > But those aren't propagated to users for libdwfl find_elf or
> > find_debuginfo. I was really just wondering if you thought about that
> > in the scope of the libdwfl calls. Currently there is no real good way
> > to do any error reporting there. So it isn't a big concern. But if you
> > have any ideas for improvement that would be nice.
>
> Will think about it.
One idea is described here:
https://lists.fedorahosted.org/pipermail/elfutils-devel/2015-May/thread.html#4841
Cheers,
Mark
