https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116614
--- Comment #5 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I've been testing
2024-09-05 Jakub Jelinek <ja...@redhat.com>
PR lto/116614
* simple-object-elf.c (SHN_COMMON): Align comment with neighbouring
comments.
(SHN_HIRESERVE): Use uppercase hex digits instead of lowercase for
consistency.
(simple_object_elf_copy_lto_debug_section): Don't fail for new_i - 1
>= SHN_LORESERVE.
--- libiberty/simple-object-elf.c.jj 2024-01-03 12:07:48.461085637 +0100
+++ libiberty/simple-object-elf.c 2024-09-05 18:33:28.761156351 +0200
@@ -128,9 +128,9 @@ typedef struct {
#define SHN_UNDEF 0 /* Undefined section */
#define SHN_LORESERVE 0xFF00 /* Begin range of reserved indices */
-#define SHN_COMMON 0xFFF2 /* Associated symbol is in common */
+#define SHN_COMMON 0xFFF2 /* Associated symbol is in common */
#define SHN_XINDEX 0xFFFF /* Section index is held elsewhere */
-#define SHN_HIRESERVE 0xffff /* End of reserved indices */
+#define SHN_HIRESERVE 0xFFFF /* End of reserved indices */
/* 32-bit ELF program header. */
@@ -1291,11 +1291,6 @@ simple_object_elf_copy_lto_debug_section
else
sh_map[i] = new_i++;
}
- if (new_i - 1 >= SHN_LORESERVE)
- {
- *err = ENOTSUP;
- return "Too many copied sections";
- }
eow->shdrs = XNEWVEC (unsigned char, shdr_size * (new_i - 1));
/* Then perform the actual copying. */
overnight.
I think for PR93117 maybe what mattered is that PR104617 wasn't fixed at that
point yet. In any case, readelf seems to be happy with the resulting debug
temp file.
