13 files changed, 32 insertions, 19 deletions
diff --git a/src/thread/aarch64/clone.s b/src/thread/aarch64/clone.s
index e3c83395..9ac272bd 100644
--- a/src/thread/aarch64/clone.s
+++ b/src/thread/aarch64/clone.s
@@ -24,7 +24,8 @@ __clone:
 	// parent
 	ret
 	// child
-1:	ldp x1,x0,[sp],#16
+1:	mov fp, 0
+	ldp x1,x0,[sp],#16
 	blr x1
 	mov x8,#93 // SYS_exit
 	svc #0
diff --git a/src/thread/arm/clone.s b/src/thread/arm/clone.s
index bb0965da..4ff0c0e8 100644
--- a/src/thread/arm/clone.s
+++ b/src/thread/arm/clone.s
@@ -19,7 +19,8 @@ __clone:
 	ldmfd sp!,{r4,r5,r6,r7}
 	bx lr
 
-1:	mov r0,r6
+1:	mov fp,#0
+	mov r0,r6
 	bl 3f
 2:	mov r7,#1
 	svc 0
diff --git a/src/thread/loongarch64/clone.s b/src/thread/loongarch64/clone.s
index a165b365..cb4aacfc 100644
--- a/src/thread/loongarch64/clone.s
+++ b/src/thread/loongarch64/clone.s
@@ -22,6 +22,7 @@ __clone:
 	beqz    $a0, 1f         # whether child process
 	jirl    $zero, $ra, 0   # parent process return
 1:
+	move    $fp, $zero
 	ld.d    $t8, $sp, 0     # function pointer
 	ld.d    $a0, $sp, 8     # argument pointer
 	jirl    $ra, $t8, 0     # call the user's function
diff --git a/src/thread/m68k/clone.s b/src/thread/m68k/clone.s
index f6dfa06f..0134cf4e 100644
--- a/src/thread/m68k/clone.s
+++ b/src/thread/m68k/clone.s
@@ -18,7 +18,8 @@ __clone:
 	beq 1f
 	movem.l (%sp)+,%d2-%d5
 	rts
-1:	move.l %a1,-(%sp)
+1:	suba.l %fp,%fp
+	move.l %a1,-(%sp)
 	jsr (%a0)
 	move.l #1,%d0
 	trap #0
diff --git a/src/thread/microblaze/clone.s b/src/thread/microblaze/clone.s
index b68cc5fc..64e3f074 100644
--- a/src/thread/microblaze/clone.s
+++ b/src/thread/microblaze/clone.s
@@ -22,7 +22,8 @@ __clone:
 	rtsd    r15, 8
 	nop
 
-1:	lwi     r3, r1, 0
+1:	add     r19, r0, r0
+	lwi     r3, r1, 0
 	lwi     r5, r1, 4
 	brald   r15, r3
 	nop
diff --git a/src/thread/mips/clone.s b/src/thread/mips/clone.s
index 04463385..229b987e 100644
--- a/src/thread/mips/clone.s
+++ b/src/thread/mips/clone.s
@@ -27,7 +27,8 @@ __clone:
 	addu $sp, $sp, 16
 	jr $ra
 	nop
-1:	lw $25, 0($sp)
+1:	move $fp, $0
+	lw $25, 0($sp)
 	lw $4, 4($sp)
 	jalr $25
 	nop
diff --git a/src/thread/mips64/clone.s b/src/thread/mips64/clone.s
index 2d86899a..8de3db6c 100644
--- a/src/thread/mips64/clone.s
+++ b/src/thread/mips64/clone.s
@@ -25,7 +25,8 @@ __clone:
 	nop
 	jr	$ra
 	nop
-1:	ld	$25, 0($sp)	# function pointer
+1:	move	$fp, $0
+	ld	$25, 0($sp)	# function pointer
 	ld	$4, 8($sp)	# argument pointer
 	jalr	$25		# call the user's function
 	nop
diff --git a/src/thread/mipsn32/clone.s b/src/thread/mipsn32/clone.s
index 4d3c8c7a..9571231a 100644
--- a/src/thread/mipsn32/clone.s
+++ b/src/thread/mipsn32/clone.s
@@ -25,7 +25,8 @@ __clone:
 	nop
 	jr	$ra
 	nop
-1:	lw	$25, 0($sp)	# function pointer
+1:	move	$fp, $0
+	lw	$25, 0($sp)	# function pointer
 	lw	$4, 4($sp)	# argument pointer
 	jalr	$25		# call the user's function
 	nop
diff --git a/src/thread/or1k/clone.s b/src/thread/or1k/clone.s
index 2473ac20..b41488a2 100644
--- a/src/thread/or1k/clone.s
+++ b/src/thread/or1k/clone.s
@@ -6,6 +6,8 @@
 .hidden __clone
 .type   __clone,@function
 __clone:
+	l.xori	r11, r0, -4
+	l.and	r4, r4, r11
 	l.addi	r4, r4, -8
 	l.sw	0(r4), r3
 	l.sw	4(r4), r6
@@ -23,7 +25,8 @@ __clone:
 	l.jr	r9
 	 l.nop
 
-1:	l.lwz	r11, 0(r1)
+1:	l.ori	r2, r0, 0
+	l.lwz	r11, 0(r1)
 	l.jalr	r11
 	 l.lwz	r3, 4(r1)
 
diff --git a/src/thread/riscv32/clone.s b/src/thread/riscv32/clone.s
index 3102239d..e2116e33 100644
--- a/src/thread/riscv32/clone.s
+++ b/src/thread/riscv32/clone.s
@@ -8,6 +8,7 @@
 .type  __clone, %function
 __clone:
 	# Save func and arg to stack
+	andi a1, a1, -16
 	addi a1, a1, -16
 	sw a0, 0(a1)
 	sw a3, 4(a1)
diff --git a/src/thread/riscv64/clone.s b/src/thread/riscv64/clone.s
index db908248..0e6f41a8 100644
--- a/src/thread/riscv64/clone.s
+++ b/src/thread/riscv64/clone.s
@@ -8,6 +8,7 @@
 .type  __clone, %function
 __clone:
 	# Save func and arg to stack
+	andi a1, a1, -16
 	addi a1, a1, -16
 	sd a0, 0(a1)
 	sd a3, 8(a1)
diff --git a/src/thread/s390x/__tls_get_offset.s b/src/thread/s390x/__tls_get_offset.s
index 8ee92de8..405f118b 100644
--- a/src/thread/s390x/__tls_get_offset.s
+++ b/src/thread/s390x/__tls_get_offset.s
@@ -1,17 +1,17 @@
 	.global __tls_get_offset
 	.type __tls_get_offset,%function
 __tls_get_offset:
-	stmg  %r14, %r15, 112(%r15)
-	aghi  %r15, -160
+	ear   %r0, %a0
+	sllg  %r0, %r0, 32
+	ear   %r0, %a1
 
-	la    %r2, 0(%r2, %r12)
-	brasl %r14, __tls_get_addr
+	la    %r1, 0(%r2, %r12)
 
-	ear   %r1, %a0
-	sllg  %r1, %r1, 32
-	ear   %r1, %a1
+	lg    %r3, 0(%r1)
+	sllg  %r4, %r3, 3
+	lg    %r5, 8(%r0)
+	lg    %r2, 0(%r4, %r5)
+	ag    %r2, 8(%r1)
+	sgr   %r2, %r0
 
-	sgr   %r2, %r1
-
-	lmg   %r14, %r15, 272(%r15)
 	br    %r14
diff --git a/src/thread/sem_post.c b/src/thread/sem_post.c
index 5c2517f2..1c8f763c 100644
--- a/src/thread/sem_post.c
+++ b/src/thread/sem_post.c
@@ -16,6 +16,6 @@ int sem_post(sem_t *sem)
 		if (waiters <= 1)
 			new &= ~0x80000000;
 	} while (a_cas(sem->__val, val, new) != val);
-	if (val<0) __wake(sem->__val, waiters>1 ? 1 : -1, priv);
+	if (val<0 || waiters) __wake(sem->__val, waiters>1 ? 1 : -1, priv);
 	return 0;
 }