Support i386 TLS code sequences without PLT

We can generate i386 TLS code sequences for general and local dynamic
models without PLT, which uses indirect call via GOT:

call *___tls_get_addr@GOT(%reg)

where EBX register isn't required as GOT base, instead of direct call:

call ___tls_get_addr[@PLT]

which requires EBX register as GOT base.

Since direct call is 4-byte long and indirect call, is 5-byte long, the
extra one byte must be handled properly.

For general dynamic model, 7-byte lea instruction before call instruction
is replaced by 6-byte one to make room for indirect call.  For local
dynamic model, we simply use 5-byte indirect call.

TLS linker optimization is updated to recognize new instruction patterns.
For local dynamic model to local exec model transition, we generate
a 6-byte lea instruction as nop, instead of a 1-byte nop plus a 4-byte
lea instruction.  Since linker may convert

call ___tls_get_addr[@PLT]

to

addr32 call ____tls_get_addr

when producing static executable, both patterns are recognized.

bfd/

	* elf64-i386.c (elf_i386_link_hash_entry): Add tls_get_addr.
	(elf_i386_link_hash_newfunc): Initialize tls_get_addr to 2.
	(elf_i386_check_tls_transition): Check indirect call and direct
	call with the addr32 prefix for general and local dynamic models.
	Set the tls_get_addr feild.
	(elf_i386_convert_load_reloc): Always use addr32 prefix for
	indirect ___tls_get_addr call via GOT.
	(elf_i386_relocate_section): Handle GD->LE, GD->IE and LD->LE
	transitions with indirect call and direct call with the addr32
	prefix.

ld/

	* testsuite/ld-i386/i386.exp: Run libtlspic2.so, tlsbin2,
	tlsgd3, tlsld2, tlsgd4, tlspie3a, tlspie3b and tlspie3c.
	* testsuite/ld-i386/pass.out: New file.
	* testsuite/ld-i386/tls-def1.c: Likewise.
	* testsuite/ld-i386/tls-gd1.S: Likewise.
	* testsuite/ld-i386/tls-ld1.S: Likewise.
	* testsuite/ld-i386/tls-main1.c: Likewise.
	* testsuite/ld-i386/tls.exp: Likewise.
	* testsuite/ld-i386/tlsbin2-nacl.rd: Likewise.
	* testsuite/ld-i386/tlsbin2.dd: Likewise.
	* testsuite/ld-i386/tlsbin2.rd: Likewise.
	* testsuite/ld-i386/tlsbin2.sd: Likewise.
	* testsuite/ld-i386/tlsbin2.td: Likewise.
	* testsuite/ld-i386/tlsbinpic2.s: Likewise.
	* testsuite/ld-i386/tlsgd3.dd: Likewise.
	* testsuite/ld-i386/tlsgd3.s: Likewise.
	* testsuite/ld-i386/tlsgd4.d: Likewise.
	* testsuite/ld-i386/tlsgd4.s: Likewise.
	* testsuite/ld-i386/tlsld2.s: Likewise.
	* testsuite/ld-i386/tlspic2-nacl.rd: Likewise.
	* testsuite/ld-i386/tlspic2.dd: Likewise.
	* testsuite/ld-i386/tlspic2.rd: Likewise.
	* testsuite/ld-i386/tlspic2.sd: Likewise.
	* testsuite/ld-i386/tlspic2.td: Likewise.
	* testsuite/ld-i386/tlspic3.s: Likewise.
	* testsuite/ld-i386/tlspie3.s: Likewise.
	* testsuite/ld-i386/tlspie3a.d: Likewise.
	* testsuite/ld-i386/tlspie3b.d: Likewise.
	* testsuite/ld-i386/tlspie3c.d: Likewise.

1 files changed, 71 insertions, 0 deletions

diff --git a/ld/testsuite/ld-i386/tls-ld1.S b/ld/testsuite/ld-i386/tls-ld1.S
new file mode 100644
index 0000000..f1295cf
--- /dev/null
+++ b/ld/testsuite/ld-i386/tls-ld1.S
@@ -0,0 +1,71 @@
+	.text
+	.p2align 4,,15
+	.globl	get_ld
+	.type	get_ld, @function
+get_ld:
+	pushl	%ebx
+	call	__x86.get_pc_thunk.bx
+	addl	$_GLOBAL_OFFSET_TABLE_, %ebx
+	subl	$8, %esp
+	leal	ld@tlsldm(%ebx), %eax
+	call	___tls_get_addr@PLT
+	leal	ld@dtpoff(%eax), %eax
+	addl	$8, %esp
+	popl	%ebx
+	ret
+	.size	get_ld, .-get_ld
+	.p2align 4,,15
+	.globl	set_ld
+	.type	set_ld, @function
+set_ld:
+	pushl	%ebx
+	call	__x86.get_pc_thunk.bx
+	addl	$_GLOBAL_OFFSET_TABLE_, %ebx
+	subl	$8, %esp
+	leal	ld@tlsldm(%ebx), %eax
+	call	___tls_get_addr@PLT
+	movl	16(%esp), %edx
+	leal	ld@dtpoff(%eax), %eax
+	movl	%edx, (%eax)
+	addl	$8, %esp
+	popl	%ebx
+	ret
+	.size	set_ld, .-set_ld
+	.p2align 4,,15
+	.globl	test_ld
+	.type	test_ld, @function
+test_ld:
+	call	__x86.get_pc_thunk.cx
+	addl	$_GLOBAL_OFFSET_TABLE_, %ecx
+	subl	$12, %esp
+	leal	ld@tlsldm(%ecx), %eax
+	call	*___tls_get_addr@GOT(%ecx)
+	movl	16(%esp), %ecx
+	leal	ld@dtpoff(%eax), %eax
+	cmpl	%ecx, (%eax)
+	sete	%al
+	addl	$12, %esp
+	movzbl	%al, %eax
+	ret
+	.size	test_ld, .-test_ld
+	.section	.tbss,"awT",@nobits
+	.align 4
+	.type	ld, @object
+	.size	ld, 4
+ld:
+	.zero	4
+	.section	.text.__x86.get_pc_thunk.bx,"axG",@progbits,__x86.get_pc_thunk.bx,comdat
+	.globl	__x86.get_pc_thunk.bx
+	.hidden	__x86.get_pc_thunk.bx
+	.type	__x86.get_pc_thunk.bx, @function
+__x86.get_pc_thunk.bx:
+	movl	(%esp), %ebx
+	ret
+	.section	.text.__x86.get_pc_thunk.cx,"axG",@progbits,__x86.get_pc_thunk.cx,comdat
+	.globl	__x86.get_pc_thunk.cx
+	.hidden	__x86.get_pc_thunk.cx
+	.type	__x86.get_pc_thunk.cx, @function
+__x86.get_pc_thunk.cx:
+	movl	(%esp), %ecx
+	ret
+	.section	.note.GNU-stack,"",@progbits