[PATCH 1/6] gold: Support 64-bit entry size in SHT_HASH.
Cary Coutant
ccoutant@gmail.com
Wed Oct 28 22:42:00 GMT 2015
> - if (parameters->target().is_big_endian())
> + bool big_endian = parameters->target().is_big_endian();
> + if (size == 32)
> {
> -#if defined(HAVE_TARGET_32_BIG) || defined(HAVE_TARGET_64_BIG)
> - Dynobj::sized_create_elf_hash_table<true>(bucket, chain, phash,
> - hashlen);
> + if (big_endian)
> + {
> +#ifdef HAVE_TARGET_32_BIG
> + Dynobj::sized_create_elf_hash_table<32, true>(bucket, chain, phash,
> + hashlen);
> +#else
> + gold_unreachable();
> +#endif
> + }
> + else
> + {
> +#ifdef HAVE_TARGET_32_LITTLE
> + Dynobj::sized_create_elf_hash_table<32, false>(bucket, chain, phash,
> + hashlen);
> #else
> - gold_unreachable();
> + gold_unreachable();
> #endif
> + }
This part won't work for 64-bit targets with 32-bit hash table
entries: since the size of the target and the size of the hash entry
are now independent, we can't assume that we don't need a 32-bit hash
entry if we haven't instantiated a 32-bit target.
> }
> - else
> + else if (size == 64)
> {
> -#if defined(HAVE_TARGET_32_LITTLE) || defined(HAVE_TARGET_64_LITTLE)
> - Dynobj::sized_create_elf_hash_table<false>(bucket, chain, phash,
> - hashlen);
> + if (big_endian)
> + {
> +#ifdef HAVE_TARGET_64_BIG
> + Dynobj::sized_create_elf_hash_table<64, true>(bucket, chain, phash,
> + hashlen);
> #else
> - gold_unreachable();
> + gold_unreachable();
> +#endif
> + }
> + else
> + {
> +#ifdef HAVE_TARGET_64_LITTLE
> + Dynobj::sized_create_elf_hash_table<64, false>(bucket, chain, phash,
> + hashlen);
> +#else
> + gold_unreachable();
> #endif
> + }
Likewise for a hypothetical 32-bit target with 64-bit hash table entries.
I will commit the attached patch on your behalf, along with this
ChangeLog entry:
2015-10-28 Marcin Kościelnicki <koriakin@0x04.net>
gold/
* dynobj.cc (Dynobj::create_elf_hash_table): Create hash table with
target-specific entry size.
(Dynobj::sized_create_elf_hash_table): Add size template parameter.
* dynobj.h (Dynobj::sized_create_elf_hash_table): Likewise.
* layout.cc (Layout::create_dynamic_symtab): Set entsize to
hash_entry_size.
* target.h (Target::hash_entry_size): New method.
(Target::Target_info::hash_entry_size): New data member.
* aarch64.cc (Target_aarch64::aarch64_info): Add hash_entry_size.
* arm.cc (Target_arm::arm_info): Likewise.
(Target_arm_nacl::arm_nacl_info): Likewise.
* i386.cc (Target_i386::i386_info): Likewise.
(Target_i386_nacl::i386_nacl_info): Likewise.
(Target_iamcu::iamcu_info): Likewise.
* mips.cc (Target_mips::mips_info): Likewise.
(Target_mips_nacl::mips_nacl_info): Likewise.
* powerpc.cc (Target_powerpc::powerpc_info): Likewise.
* sparc.cc (Target_sparc::sparc_info): Likewise.
* tilegx.cc (Target_tilegx::tilegx_info): Likewise.
* x86_64.cc (Target_x86_64::x86_64_info): Likewise.
(Target_x86_64_nacl::x86_64_nacl_info): Likewise.
* testsuite/testfile.cc (Target_test::test_target_info): Likewise.
(Normally, I'd have asked you to write the ChangeLog entry, but since
I took so long to review this patch set, I wrote it for you.)
-cary
-------------- next part --------------
A non-text attachment was scrubbed...
Name: s390-1-revised.patch
Type: application/octet-stream
Size: 17345 bytes
Desc: not available
URL: <https://sourceware.org/pipermail/binutils/attachments/20151028/8507ae15/attachment.obj>
More information about the Binutils
mailing list