[v2,1/3] rwlock: reimplement with __atomic builtins

Message ID 1547557979-153169-2-git-send-email-joyce.kong@arm.com
State Superseded, archived
Delegated to: Thomas Monjalon
Headers show
Series
  • reimplement rwlock and add relevant perf test case
Related show

Checks

Context Check Description
ci/intel-Performance-Testing success Performance Testing PASS
ci/mellanox-Performance-Testing success Performance Testing PASS
ci/Intel-compilation success Compilation OK
ci/checkpatch success coding style OK

Commit Message

Joyce Kong Jan. 15, 2019, 1:12 p.m.
From: Gavin Hu <gavin.hu@arm.com>

The __sync builtin based implementation generates full memory barriers
('dmb ish') on Arm platforms. Using C11 atomic builtins to generate one
way barriers.

Here is the assembly code of __sync_compare_and_swap builtin.
__sync_bool_compare_and_swap(dst, exp, src);
   0x000000000090f1b0 <+16>:    e0 07 40 f9 ldr x0, [sp, #8]
   0x000000000090f1b4 <+20>:    e1 0f 40 79 ldrh    w1, [sp, #6]
   0x000000000090f1b8 <+24>:    e2 0b 40 79 ldrh    w2, [sp, #4]
   0x000000000090f1bc <+28>:    21 3c 00 12 and w1, w1, #0xffff
   0x000000000090f1c0 <+32>:    03 7c 5f 48 ldxrh   w3, [x0]
   0x000000000090f1c4 <+36>:    7f 00 01 6b cmp w3, w1
   0x000000000090f1c8 <+40>:    61 00 00 54 b.ne    0x90f1d4
<rte_atomic16_cmpset+52>  // b.any
   0x000000000090f1cc <+44>:    02 fc 04 48 stlxrh  w4, w2, [x0]
   0x000000000090f1d0 <+48>:    84 ff ff 35 cbnz    w4, 0x90f1c0
<rte_atomic16_cmpset+32>
   0x000000000090f1d4 <+52>:    bf 3b 03 d5 dmb ish
   0x000000000090f1d8 <+56>:    e0 17 9f 1a cset    w0, eq  // eq = none

Fixes: af75078fece3 ("first public release")
Cc: stable@dpdk.org

Signed-off-by: Gavin Hu <gavin.hu@arm.com>
Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
Reviewed-by: Ola Liljedahl <ola.liljedahl@arm.com>
Reviewed-by: Joyce Kong <joyce.kong@arm.com>
Tested-by: Joyce Kong <joyce.kong@arm.com>
---
 lib/librte_eal/common/include/generic/rte_rwlock.h | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

Comments

Jerin Jacob Kollanukkaran March 12, 2019, 6:08 p.m. | #1
On Tue, 2019-01-15 at 21:12 +0800, Joyce Kong wrote:
> From: Gavin Hu <gavin.hu@arm.com>
> 
> The __sync builtin based implementation generates full memory
> barriers
> ('dmb ish') on Arm platforms. Using C11 atomic builtins to generate
> one
> way barriers.
> 
> Here is the assembly code of __sync_compare_and_swap builtin.
> __sync_bool_compare_and_swap(dst, exp, src);
>    0x000000000090f1b0 <+16>:    e0 07 40 f9 ldr x0, [sp, #8]
>    0x000000000090f1b4 <+20>:    e1 0f 40 79 ldrh    w1, [sp, #6]
>    0x000000000090f1b8 <+24>:    e2 0b 40 79 ldrh    w2, [sp, #4]
>    0x000000000090f1bc <+28>:    21 3c 00 12 and w1, w1, #0xffff
>    0x000000000090f1c0 <+32>:    03 7c 5f 48 ldxrh   w3, [x0]
>    0x000000000090f1c4 <+36>:    7f 00 01 6b cmp w3, w1
>    0x000000000090f1c8 <+40>:    61 00 00 54 b.ne    0x90f1d4
> <rte_atomic16_cmpset+52>  // b.any
>    0x000000000090f1cc <+44>:    02 fc 04 48 stlxrh  w4, w2, [x0]
>    0x000000000090f1d0 <+48>:    84 ff ff 35 cbnz    w4, 0x90f1c0
> <rte_atomic16_cmpset+32>
>    0x000000000090f1d4 <+52>:    bf 3b 03 d5 dmb ish
>    0x000000000090f1d8 <+56>:    e0 17 9f 1a cset    w0, eq  // eq =
> none
> 
> Fixes: af75078fece3 ("first public release")
> Cc: stable@dpdk.org
> 
> Signed-off-by: Gavin Hu <gavin.hu@arm.com>
> Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> Reviewed-by: Ola Liljedahl <ola.liljedahl@arm.com>
> Reviewed-by: Joyce Kong <joyce.kong@arm.com>
> Tested-by: Joyce Kong <joyce.kong@arm.com>
> ---

Nit:
Fix following git-checklog.sh issue

Wrong headline format:
	rwlock: reimplement with __atomic builtins


With above fix:
Acked-by: Jerin Jacob <jerinj@marvell.com>

Performance on octeontx2: +88% improvement with this patch.
Joyce Kong March 14, 2019, 1:17 p.m. | #2
> -----Original Message-----
> From: Jerin Jacob Kollanukkaran <jerinj@marvell.com>
> Sent: Wednesday, March 13, 2019 2:09 AM
> To: Joyce Kong (Arm Technology China) <Joyce.Kong@arm.com>;
> dev@dpdk.org
> Cc: chaozhu@linux.vnet.ibm.com; nd <nd@arm.com>;
> bruce.richardson@intel.com; thomas@monjalon.net;
> hemant.agrawal@nxp.com; stable@dpdk.org; Honnappa Nagarahalli
> <Honnappa.Nagarahalli@arm.com>; Gavin Hu (Arm Technology China)
> <Gavin.Hu@arm.com>
> Subject: Re: [EXT] [PATCH v2 1/3] rwlock: reimplement with __atomic builtins
> 
> On Tue, 2019-01-15 at 21:12 +0800, Joyce Kong wrote:
> > From: Gavin Hu <gavin.hu@arm.com>
> >
> > The __sync builtin based implementation generates full memory barriers
> > ('dmb ish') on Arm platforms. Using C11 atomic builtins to generate
> > one way barriers.
> >
> > Here is the assembly code of __sync_compare_and_swap builtin.
> > __sync_bool_compare_and_swap(dst, exp, src);
> >    0x000000000090f1b0 <+16>:    e0 07 40 f9 ldr x0, [sp, #8]
> >    0x000000000090f1b4 <+20>:    e1 0f 40 79 ldrh    w1, [sp, #6]
> >    0x000000000090f1b8 <+24>:    e2 0b 40 79 ldrh    w2, [sp, #4]
> >    0x000000000090f1bc <+28>:    21 3c 00 12 and w1, w1, #0xffff
> >    0x000000000090f1c0 <+32>:    03 7c 5f 48 ldxrh   w3, [x0]
> >    0x000000000090f1c4 <+36>:    7f 00 01 6b cmp w3, w1
> >    0x000000000090f1c8 <+40>:    61 00 00 54 b.ne    0x90f1d4
> > <rte_atomic16_cmpset+52>  // b.any
> >    0x000000000090f1cc <+44>:    02 fc 04 48 stlxrh  w4, w2, [x0]
> >    0x000000000090f1d0 <+48>:    84 ff ff 35 cbnz    w4, 0x90f1c0
> > <rte_atomic16_cmpset+32>
> >    0x000000000090f1d4 <+52>:    bf 3b 03 d5 dmb ish
> >    0x000000000090f1d8 <+56>:    e0 17 9f 1a cset    w0, eq  // eq =
> > none
> >
> > Fixes: af75078fece3 ("first public release")
> > Cc: stable@dpdk.org
> >
> > Signed-off-by: Gavin Hu <gavin.hu@arm.com>
> > Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> > Reviewed-by: Ola Liljedahl <ola.liljedahl@arm.com>
> > Reviewed-by: Joyce Kong <joyce.kong@arm.com>
> > Tested-by: Joyce Kong <joyce.kong@arm.com>
> > ---
> 
> Nit:
> Fix following git-checklog.sh issue
> 
> Wrong headline format:
> 	rwlock: reimplement with __atomic builtins
> 
Has fixed format error in v3.

> With above fix:
> Acked-by: Jerin Jacob <jerinj@marvell.com>
> 
> Performance on octeontx2: +88% improvement with this patch.
Gavin Hu March 15, 2019, 3:40 a.m. | #3
Hi Jerin,

> -----Original Message-----
> From: Jerin Jacob Kollanukkaran <jerinj@marvell.com>
> Sent: Wednesday, March 13, 2019 2:09 AM
> To: Joyce Kong (Arm Technology China) <Joyce.Kong@arm.com>;
> dev@dpdk.org
> Cc: chaozhu@linux.vnet.ibm.com; nd <nd@arm.com>;
> bruce.richardson@intel.com; thomas@monjalon.net;
> hemant.agrawal@nxp.com; stable@dpdk.org; Honnappa Nagarahalli
> <Honnappa.Nagarahalli@arm.com>; Gavin Hu (Arm Technology China)
> <Gavin.Hu@arm.com>
> Subject: Re: [EXT] [PATCH v2 1/3] rwlock: reimplement with __atomic
> builtins
> 
> On Tue, 2019-01-15 at 21:12 +0800, Joyce Kong wrote:
> > From: Gavin Hu <gavin.hu@arm.com>
> >
> > The __sync builtin based implementation generates full memory
> > barriers
> > ('dmb ish') on Arm platforms. Using C11 atomic builtins to generate
> > one
> > way barriers.
> >
> > Here is the assembly code of __sync_compare_and_swap builtin.
> > __sync_bool_compare_and_swap(dst, exp, src);
> >    0x000000000090f1b0 <+16>:    e0 07 40 f9 ldr x0, [sp, #8]
> >    0x000000000090f1b4 <+20>:    e1 0f 40 79 ldrh    w1, [sp, #6]
> >    0x000000000090f1b8 <+24>:    e2 0b 40 79 ldrh    w2, [sp, #4]
> >    0x000000000090f1bc <+28>:    21 3c 00 12 and w1, w1, #0xffff
> >    0x000000000090f1c0 <+32>:    03 7c 5f 48 ldxrh   w3, [x0]
> >    0x000000000090f1c4 <+36>:    7f 00 01 6b cmp w3, w1
> >    0x000000000090f1c8 <+40>:    61 00 00 54 b.ne    0x90f1d4
> > <rte_atomic16_cmpset+52>  // b.any
> >    0x000000000090f1cc <+44>:    02 fc 04 48 stlxrh  w4, w2, [x0]
> >    0x000000000090f1d0 <+48>:    84 ff ff 35 cbnz    w4, 0x90f1c0
> > <rte_atomic16_cmpset+32>
> >    0x000000000090f1d4 <+52>:    bf 3b 03 d5 dmb ish
> >    0x000000000090f1d8 <+56>:    e0 17 9f 1a cset    w0, eq  // eq =
> > none
> >
> > Fixes: af75078fece3 ("first public release")
> > Cc: stable@dpdk.org
> >
> > Signed-off-by: Gavin Hu <gavin.hu@arm.com>
> > Reviewed-by: Honnappa Nagarahalli <honnappa.nagarahalli@arm.com>
> > Reviewed-by: Ola Liljedahl <ola.liljedahl@arm.com>
> > Reviewed-by: Joyce Kong <joyce.kong@arm.com>
> > Tested-by: Joyce Kong <joyce.kong@arm.com>
> > ---
> 
> Nit:
> Fix following git-checklog.sh issue
> 
> Wrong headline format:
> 	rwlock: reimplement with __atomic builtins
> 
> 
> With above fix:
> Acked-by: Jerin Jacob <jerinj@marvell.com>
> 
> Performance on octeontx2: +88% improvement with this patch.
> 
With your test result, I assume you applied all the three patches, not this one only.
Would you review/ack the other two patches? 
>

Patch

diff --git a/lib/librte_eal/common/include/generic/rte_rwlock.h b/lib/librte_eal/common/include/generic/rte_rwlock.h
index b05d85a..3317a21 100644
--- a/lib/librte_eal/common/include/generic/rte_rwlock.h
+++ b/lib/librte_eal/common/include/generic/rte_rwlock.h
@@ -64,14 +64,14 @@  rte_rwlock_read_lock(rte_rwlock_t *rwl)
 	int success = 0;
 
 	while (success == 0) {
-		x = rwl->cnt;
+		x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED);
 		/* write lock is held */
 		if (x < 0) {
 			rte_pause();
 			continue;
 		}
-		success = rte_atomic32_cmpset((volatile uint32_t *)&rwl->cnt,
-					      (uint32_t)x, (uint32_t)(x + 1));
+		success = __atomic_compare_exchange_n(&rwl->cnt, &x, x+1, 1,
+					__ATOMIC_ACQUIRE, __ATOMIC_RELAXED);
 	}
 }
 
@@ -114,7 +114,7 @@  rte_rwlock_read_trylock(rte_rwlock_t *rwl)
 static inline void
 rte_rwlock_read_unlock(rte_rwlock_t *rwl)
 {
-	rte_atomic32_dec((rte_atomic32_t *)(intptr_t)&rwl->cnt);
+	__atomic_fetch_sub(&rwl->cnt, 1, __ATOMIC_RELEASE);
 }
 
 /**
@@ -156,14 +156,14 @@  rte_rwlock_write_lock(rte_rwlock_t *rwl)
 	int success = 0;
 
 	while (success == 0) {
-		x = rwl->cnt;
+		x = __atomic_load_n(&rwl->cnt, __ATOMIC_RELAXED);
 		/* a lock is held */
 		if (x != 0) {
 			rte_pause();
 			continue;
 		}
-		success = rte_atomic32_cmpset((volatile uint32_t *)&rwl->cnt,
-					      0, (uint32_t)-1);
+		success = __atomic_compare_exchange_n(&rwl->cnt, &x, -1, 1,
+					__ATOMIC_ACQUIRE, __ATOMIC_RELAXED);
 	}
 }
 
@@ -176,7 +176,7 @@  rte_rwlock_write_lock(rte_rwlock_t *rwl)
 static inline void
 rte_rwlock_write_unlock(rte_rwlock_t *rwl)
 {
-	rte_atomic32_inc((rte_atomic32_t *)(intptr_t)&rwl->cnt);
+	__atomic_store_n(&rwl->cnt, 0, __ATOMIC_RELEASE);
 }
 
 /**