Re: [PATCH] improve warning suppression for inlined functions (PR 98465, 98512)

[Martin Sebor via Gcc-patches]

On 1/21/21 12:01 PM, Florian Weimer wrote:

* Martin Sebor:

On 1/21/21 10:34 AM, Florian Weimer wrote:

* Martin Sebor via Gcc-patches:

This patch depends on the fix for PR 98664 (already approved but
not yet checked in).  I've tested it on x86_64-linux.

To avoid fallout I tried to keep the changes to a minimum, and
so the design isn't as robust as I'd like it ultimately to be.
I plan to enhance it in stage 1.

I've tested this patch on top of 43705f3fa343e08b2fb030460f (so with
the
PR98664 fix, I think), and the reproducer from PR98512 now ICEs:

Thanks for giving it a try!  I saw a similar ICE during my testing
-- it's caused by a couple of uninitialized variables.  I fixed
it in my tree (see below) but the fix didn't make it into the patch.

Please give this a try and let me know if it doesn't help:

index abcd991b829..d82a7eb67e5 100644
--- a/gcc/diagnostic.c
+++ b/gcc/diagnostic.c
@@ -1426,7 +1426,7 @@ diagnostic_impl (rich_location *richloc, const
diagnostic_metadata *metadata,
                  int opt, const char *gmsgid,
                  va_list *ap, diagnostic_t kind)
  {
-  diagnostic_info diagnostic;
+  diagnostic_info diagnostic{ };
    if (kind == DK_PERMERROR)
      {
        diagnostic_set_info (&diagnostic, gmsgid, ap, richloc,
@@ -1452,7 +1452,7 @@ diagnostic_n_impl (rich_location *richloc, const
diagnostic_metadata *metadata,
                    const char *plural_gmsgid,
                    va_list *ap, diagnostic_t kind)
  {
-  diagnostic_info diagnostic;
+  diagnostic_info diagnostic{ };
    unsigned long gtn;

    if (sizeof n <= sizeof gtn)

This fixes the crash for me, and the warnings is gone as well.

Great, thanks for the confirmation!  I'll post the updated patch
shortly.

Martin

Thanks,
Florian