aha/arch/sparc64/lib/atomic.S
David S. Miller b445e26cbf [SPARC64]: Avoid membar instructions in delay slots.
In particular, avoid membar instructions in the delay
slot of a jmpl instruction.

UltraSPARC-I, II, IIi, and IIe have a bug, documented in
the UltraSPARC-IIi User's Manual, Appendix K, Erratum 51

The long and short of it is that if the IMU unit misses
on a branch or jmpl, and there is a store buffer synchronizing
membar in the delay slot, the chip can stop fetching instructions.

If interrupts are enabled or some other trap is enabled, the
chip will unwedge itself, but performance will suffer.

We already had a workaround for this bug in a few spots, but
it's better to have the entire tree sanitized for this rule.

Signed-off-by: David S. Miller <davem@davemloft.net>
2005-06-27 15:42:04 -07:00

149 lines
3 KiB
ArmAsm

/* $Id: atomic.S,v 1.4 2001/11/18 00:12:56 davem Exp $
* atomic.S: These things are too big to do inline.
*
* Copyright (C) 1999 David S. Miller (davem@redhat.com)
*/
#include <linux/config.h>
#include <asm/asi.h>
.text
/* Two versions of the atomic routines, one that
* does not return a value and does not perform
* memory barriers, and a second which returns
* a value and does the barriers.
*/
.globl atomic_add
.type atomic_add,#function
atomic_add: /* %o0 = increment, %o1 = atomic_ptr */
1: lduw [%o1], %g1
add %g1, %o0, %g7
cas [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %icc, 1b
nop
retl
nop
.size atomic_add, .-atomic_add
.globl atomic_sub
.type atomic_sub,#function
atomic_sub: /* %o0 = decrement, %o1 = atomic_ptr */
1: lduw [%o1], %g1
sub %g1, %o0, %g7
cas [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %icc, 1b
nop
retl
nop
.size atomic_sub, .-atomic_sub
/* On SMP we need to use memory barriers to ensure
* correct memory operation ordering, nop these out
* for uniprocessor.
*/
#ifdef CONFIG_SMP
#define ATOMIC_PRE_BARRIER membar #StoreLoad | #LoadLoad;
#define ATOMIC_POST_BARRIER \
ba,pt %xcc, 80b; \
membar #StoreLoad | #StoreStore
80: retl
nop
#else
#define ATOMIC_PRE_BARRIER
#define ATOMIC_POST_BARRIER
#endif
.globl atomic_add_ret
.type atomic_add_ret,#function
atomic_add_ret: /* %o0 = increment, %o1 = atomic_ptr */
ATOMIC_PRE_BARRIER
1: lduw [%o1], %g1
add %g1, %o0, %g7
cas [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %icc, 1b
add %g7, %o0, %g7
sra %g7, 0, %o0
ATOMIC_POST_BARRIER
retl
nop
.size atomic_add_ret, .-atomic_add_ret
.globl atomic_sub_ret
.type atomic_sub_ret,#function
atomic_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */
ATOMIC_PRE_BARRIER
1: lduw [%o1], %g1
sub %g1, %o0, %g7
cas [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %icc, 1b
sub %g7, %o0, %g7
sra %g7, 0, %o0
ATOMIC_POST_BARRIER
retl
nop
.size atomic_sub_ret, .-atomic_sub_ret
.globl atomic64_add
.type atomic64_add,#function
atomic64_add: /* %o0 = increment, %o1 = atomic_ptr */
1: ldx [%o1], %g1
add %g1, %o0, %g7
casx [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %xcc, 1b
nop
retl
nop
.size atomic64_add, .-atomic64_add
.globl atomic64_sub
.type atomic64_sub,#function
atomic64_sub: /* %o0 = decrement, %o1 = atomic_ptr */
1: ldx [%o1], %g1
sub %g1, %o0, %g7
casx [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %xcc, 1b
nop
retl
nop
.size atomic64_sub, .-atomic64_sub
.globl atomic64_add_ret
.type atomic64_add_ret,#function
atomic64_add_ret: /* %o0 = increment, %o1 = atomic_ptr */
ATOMIC_PRE_BARRIER
1: ldx [%o1], %g1
add %g1, %o0, %g7
casx [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %xcc, 1b
add %g7, %o0, %g7
mov %g7, %o0
ATOMIC_POST_BARRIER
retl
nop
.size atomic64_add_ret, .-atomic64_add_ret
.globl atomic64_sub_ret
.type atomic64_sub_ret,#function
atomic64_sub_ret: /* %o0 = decrement, %o1 = atomic_ptr */
ATOMIC_PRE_BARRIER
1: ldx [%o1], %g1
sub %g1, %o0, %g7
casx [%o1], %g1, %g7
cmp %g1, %g7
bne,pn %xcc, 1b
sub %g7, %o0, %g7
mov %g7, %o0
ATOMIC_POST_BARRIER
retl
nop
.size atomic64_sub_ret, .-atomic64_sub_ret