https://sourceware.org/bugzilla/show_bug.cgi?id=25694
Nelson Chu <nelsonc1225 at sourceware dot org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |nelsonc1225 at sourceware dot
org
--- Comment #4 from Nelson Chu <nelsonc1225 at sourceware dot org> ---
Haven't test these changes for the riscv-gnu-toolchain regressions, and not yet
test for other project. However, at least TLS LE testcase mentioned here, and
the pr22263-1 testcase looks good after applying the following changes.
For the R_RISCV_TPREL_HI20 in check_reloc, not really sure why we used to goto
static_reloc when pie/pde. Refer to spare and loongarch, it's just pass and
won't reserve local dynamic reloc for TLS LE in pie/pde, and similar for other
targets. So we probably too conservative before.
As for the TLS GD/IE in riscv_elf_relocate_section, I refer to what Maciej did
for pr22570, commits 9143e72c6d4d and 1cb83cac9a89, seems also the right way to
do the same thing for risc-v.
diff --git a/bfd/elfnn-riscv.c b/bfd/elfnn-riscv.c
index 77a732b6a08..bbb5d0cd5ff 100644
--- a/bfd/elfnn-riscv.c
+++ b/bfd/elfnn-riscv.c
@@ -824,7 +824,7 @@ riscv_elf_check_relocs (bfd *abfd, struct bfd_link_info
*info,
break;
case R_RISCV_TLS_GOT_HI20:
- if (bfd_link_pic (info))
+ if (bfd_link_dll (info))
info->flags |= DF_STATIC_TLS;
if (!riscv_elf_record_got_reference (abfd, info, h, r_symndx)
|| !riscv_elf_record_tls_type (abfd, h, r_symndx, GOT_TLS_IE))
@@ -920,11 +920,12 @@ riscv_elf_check_relocs (bfd *abfd, struct bfd_link_info
*info,
goto static_reloc;
case R_RISCV_TPREL_HI20:
+ /* This is not allowed in the pic, but okay in pie. */
if (!bfd_link_executable (info))
return bad_static_reloc (abfd, r_type, h);
if (h != NULL)
riscv_elf_record_tls_type (abfd, h, r_symndx, GOT_TLS_LE);
- goto static_reloc;
+ break;
case R_RISCV_HI20:
if (bfd_link_pic (info))
@@ -2797,24 +2798,20 @@ riscv_elf_relocate_section (bfd *output_bfd,
if (htab->elf.srelgot == NULL)
abort ();
- if (h != NULL)
- {
- bool dyn, pic;
- dyn = htab->elf.dynamic_sections_created;
- pic = bfd_link_pic (info);
-
- if (WILL_CALL_FINISH_DYNAMIC_SYMBOL (dyn, pic, h)
- && (!pic || !SYMBOL_REFERENCES_LOCAL (info, h)))
- indx = h->dynindx;
- }
+ bool dyn = elf_hash_table (info)->dynamic_sections_created;
+ if (h != NULL
+ && h->dynindx != -1
+ && WILL_CALL_FINISH_DYNAMIC_SYMBOL (dyn, bfd_link_pic (info),
h)
+ && (bfd_link_dll (info) || !SYMBOL_REFERENCES_LOCAL (info,
h)))
+ indx = h->dynindx;
/* The GOT entries have not been initialized yet. Do it
now, and emit any relocations. */
- if ((bfd_link_pic (info) || indx != 0)
+ if ((bfd_link_dll (info) || indx != 0)
&& (h == NULL
|| ELF_ST_VISIBILITY (h->other) == STV_DEFAULT
|| h->root.type != bfd_link_hash_undefweak))
- need_relocs = true;
+ need_relocs = true;
if (tls_type & GOT_TLS_GD)
--
You are receiving this mail because:
You are on the CC list for the bug.
