Use 'e' instead of 'i' constraints with 64-bit atomic operations on amd64.

The ADD, AND, OR, and SUB instructions take at most a 32-bit
sign-extended immediate operand.  64-bit constants that do not fit into
that constraint need to be loaded into a register.  The 'i' constraint
tells the compiler it can pass any integer constant to the assembler,
whereas the 'e' constrain only permits constants that fit into a 32-bit
sign-extended value.  This fixes using
atomic_add/clear/set/subtract_long/64 with constants that do not fit into
a 32-bit sign-extended immediate.

Reported by:	several folks
Tested by:	Pete Wright <pete@nomadlogic.org>
MFC after:	2 weeks
This commit is contained in:
John Baldwin 2018-07-03 22:03:28 +00:00
parent cd1acc77e6
commit 79ba91952d
Notes: svn2git 2020-12-20 02:59:44 +00:00
svn path=/head/; revision=335913

View File

@ -446,10 +446,10 @@ ATOMIC_ASM(clear, int, "andl %1,%0", "ir", ~v);
ATOMIC_ASM(add, int, "addl %1,%0", "ir", v);
ATOMIC_ASM(subtract, int, "subl %1,%0", "ir", v);
ATOMIC_ASM(set, long, "orq %1,%0", "ir", v);
ATOMIC_ASM(clear, long, "andq %1,%0", "ir", ~v);
ATOMIC_ASM(add, long, "addq %1,%0", "ir", v);
ATOMIC_ASM(subtract, long, "subq %1,%0", "ir", v);
ATOMIC_ASM(set, long, "orq %1,%0", "er", v);
ATOMIC_ASM(clear, long, "andq %1,%0", "er", ~v);
ATOMIC_ASM(add, long, "addq %1,%0", "er", v);
ATOMIC_ASM(subtract, long, "subq %1,%0", "er", v);
#define ATOMIC_LOADSTORE(TYPE) \
ATOMIC_LOAD(TYPE); \