All of lore.kernel.org
 help / color / mirror / Atom feed
From: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
To: benh@kernel.crashing.org, mpe@ellerman.id.au
Cc: anton@samba.org, paulus@samba.org, npiggin@gmail.com,
	linuxppc-dev@lists.ozlabs.org,
	Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
Subject: [RFC PATCH v3 12/12] powerpc: rewrite local_t using soft_irq
Date: Thu, 25 Aug 2016 12:00:03 +0530	[thread overview]
Message-ID: <1472106603-23336-13-git-send-email-maddy@linux.vnet.ibm.com> (raw)
In-Reply-To: <1472106603-23336-1-git-send-email-maddy@linux.vnet.ibm.com>

Local atomic operations are fast and highly reentrant per CPU counters.
Used for percpu variable updates. Local atomic operations only guarantee
variable modification atomicity wrt the CPU which owns the data and
these needs to be executed in a preemption safe way.

Here is the design of this patch. Since local_* operations
are only need to be atomic to interrupts (IIUC), we have two options.
Either replay the "op" if interrupted or replay the interrupt after
the "op". Initial patchset posted was based on implementing local_* operation
based on CR5 which replay's the "op". Patchset had issues in case of
rewinding the address pointor from an array. This make the slow patch
really slow. Since CR5 based implementation proposed using __ex_table to find
the rewind addressr, this rasied concerns about size of __ex_table and vmlinux.

https://lists.ozlabs.org/pipermail/linuxppc-dev/2014-December/123115.html

But this patch uses, arch_local_irq_*() to soft_disable
interrupts (including PMIs). After finishing the "op", arch_local_irq_restore()
called and correspondingly interrupts are replayed if any occured.

patch re-write the current local_* functions to use arch_local_irq_disbale.
Base flow for each function is

{
	soft_irq_set_mask()
	load
	..
	store
	arch_local_irq_restore()
}

Reason for the approach is that, currently l[w/d]arx/st[w/d]cx.
instruction pair is used for local_* operations, which are heavy
on cycle count and they dont support a local variant. So to
see whether the new implementation helps, used a modified
version of Rusty's benchmark code on local_t.

https://lkml.org/lkml/2008/12/16/450

Modifications to Rusty's benchmark code:
- Executed only local_t test

Here are the values with the patch.

Time in ns per iteration

Local_t             Without Patch           With Patch

_inc                        28              8
_add                        28              8
_read                       3               3
_add_return         28              7

Currently only asm/local.h has been rewrite, and also
the entire change is tested only in PPC64 (pseries guest)
and PPC64 host (LE).

TODO:
	- local_cmpxchg and local_xchg needs modification.

Signed-off-by: Madhavan Srinivasan <maddy@linux.vnet.ibm.com>
---
 arch/powerpc/include/asm/local.h | 91 +++++++++++++++++++++++++++-------------
 1 file changed, 63 insertions(+), 28 deletions(-)

diff --git a/arch/powerpc/include/asm/local.h b/arch/powerpc/include/asm/local.h
index b8da91363864..4ab726e99fea 100644
--- a/arch/powerpc/include/asm/local.h
+++ b/arch/powerpc/include/asm/local.h
@@ -14,24 +14,50 @@ typedef struct
 #define local_read(l)	atomic_long_read(&(l)->a)
 #define local_set(l,i)	atomic_long_set(&(l)->a, (i))
 
-#define local_add(i,l)	atomic_long_add((i),(&(l)->a))
-#define local_sub(i,l)	atomic_long_sub((i),(&(l)->a))
-#define local_inc(l)	atomic_long_inc(&(l)->a)
-#define local_dec(l)	atomic_long_dec(&(l)->a)
+static __inline__ void local_add(long i, local_t *l)
+{
+	long t;
+	unsigned long flags;
+
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
+	__asm__ __volatile__(
+	PPC_LL" %0,0(%2)\n\
+	add     %0,%1,%0\n"
+	PPC_STL" %0,0(%2)\n"
+	: "=&r" (t)
+	: "r" (i), "r" (&(l->a.counter)));
+	arch_local_irq_restore(flags);
+}
+
+static __inline__ void local_sub(long i, local_t *l)
+{
+	long t;
+	unsigned long flags;
+
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
+	__asm__ __volatile__(
+	PPC_LL" %0,0(%2)\n\
+	subf    %0,%1,%0\n"
+	PPC_STL" %0,0(%2)\n"
+	: "=&r" (t)
+	: "r" (i), "r" (&(l->a.counter)));
+	arch_local_irq_restore(flags);
+}
 
 static __inline__ long local_add_return(long a, local_t *l)
 {
 	long t;
+	unsigned long flags;
 
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 	__asm__ __volatile__(
-"1:"	PPC_LLARX(%0,0,%2,0) "			# local_add_return\n\
+	PPC_LL" %0,0(%2)\n\
 	add	%0,%1,%0\n"
-	PPC405_ERR77(0,%2)
-	PPC_STLCX	"%0,0,%2 \n\
-	bne-	1b"
+	PPC_STL	"%0,0(%2)\n"
 	: "=&r" (t)
 	: "r" (a), "r" (&(l->a.counter))
 	: "cc", "memory");
+	arch_local_irq_restore(flags);
 
 	return t;
 }
@@ -41,16 +67,18 @@ static __inline__ long local_add_return(long a, local_t *l)
 static __inline__ long local_sub_return(long a, local_t *l)
 {
 	long t;
+	unsigned long flags;
+
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 
 	__asm__ __volatile__(
-"1:"	PPC_LLARX(%0,0,%2,0) "			# local_sub_return\n\
+"1:"	PPC_LL" %0,0(%2)\n\
 	subf	%0,%1,%0\n"
-	PPC405_ERR77(0,%2)
-	PPC_STLCX	"%0,0,%2 \n\
-	bne-	1b"
+	PPC_STL	"%0,0(%2)\n"
 	: "=&r" (t)
 	: "r" (a), "r" (&(l->a.counter))
 	: "cc", "memory");
+	arch_local_irq_restore(flags);
 
 	return t;
 }
@@ -58,16 +86,17 @@ static __inline__ long local_sub_return(long a, local_t *l)
 static __inline__ long local_inc_return(local_t *l)
 {
 	long t;
+	unsigned long flags;
 
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 	__asm__ __volatile__(
-"1:"	PPC_LLARX(%0,0,%1,0) "			# local_inc_return\n\
+"1:"	PPC_LL" %0,0(%1)\n\
 	addic	%0,%0,1\n"
-	PPC405_ERR77(0,%1)
-	PPC_STLCX	"%0,0,%1 \n\
-	bne-	1b"
+	PPC_STL "%0,0(%1)\n"
 	: "=&r" (t)
 	: "r" (&(l->a.counter))
 	: "cc", "xer", "memory");
+	arch_local_irq_restore(flags);
 
 	return t;
 }
@@ -85,20 +114,24 @@ static __inline__ long local_inc_return(local_t *l)
 static __inline__ long local_dec_return(local_t *l)
 {
 	long t;
+	unsigned long flags;
 
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 	__asm__ __volatile__(
-"1:"	PPC_LLARX(%0,0,%1,0) "			# local_dec_return\n\
+	PPC_LL" %0,0(%1)\n\
 	addic	%0,%0,-1\n"
-	PPC405_ERR77(0,%1)
-	PPC_STLCX	"%0,0,%1\n\
-	bne-	1b"
+	PPC_STL "%0,0(%1)\n"
 	: "=&r" (t)
 	: "r" (&(l->a.counter))
 	: "cc", "xer", "memory");
+	arch_local_irq_restore(flags);
 
 	return t;
 }
 
+#define local_inc(l)	local_inc_return(l)
+#define local_dec(l)	local_dec_return(l)
+
 #define local_cmpxchg(l, o, n) \
 	(cmpxchg_local(&((l)->a.counter), (o), (n)))
 #define local_xchg(l, n) (xchg_local(&((l)->a.counter), (n)))
@@ -115,20 +148,21 @@ static __inline__ long local_dec_return(local_t *l)
 static __inline__ int local_add_unless(local_t *l, long a, long u)
 {
 	long t;
+	unsigned long flags;
 
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 	__asm__ __volatile__ (
-"1:"	PPC_LLARX(%0,0,%1,0) "			# local_add_unless\n\
+	PPC_LL" %0,0(%1)\n\
 	cmpw	0,%0,%3 \n\
 	beq-	2f \n\
 	add	%0,%2,%0 \n"
-	PPC405_ERR77(0,%2)
-	PPC_STLCX	"%0,0,%1 \n\
-	bne-	1b \n"
+	PPC_STL" %0,0(%1) \n"
 "	subf	%0,%2,%0 \n\
 2:"
 	: "=&r" (t)
 	: "r" (&(l->a.counter)), "r" (a), "r" (u)
 	: "cc", "memory");
+	arch_local_irq_restore(flags);
 
 	return t != u;
 }
@@ -145,19 +179,20 @@ static __inline__ int local_add_unless(local_t *l, long a, long u)
 static __inline__ long local_dec_if_positive(local_t *l)
 {
 	long t;
+	unsigned long flags;
 
+	flags = soft_irq_set_mask(IRQ_DISABLE_MASK_PMU | IRQ_DISABLE_MASK_LINUX);
 	__asm__ __volatile__(
-"1:"	PPC_LLARX(%0,0,%1,0) "			# local_dec_if_positive\n\
+	PPC_LL" %0,0(%1)\n\
 	cmpwi	%0,1\n\
 	addi	%0,%0,-1\n\
 	blt-	2f\n"
-	PPC405_ERR77(0,%1)
-	PPC_STLCX	"%0,0,%1\n\
-	bne-	1b"
+	PPC_STL "%0,0(%1)\n"
 	"\n\
 2:"	: "=&b" (t)
 	: "r" (&(l->a.counter))
 	: "cc", "memory");
+	arch_local_irq_restore(flags);
 
 	return t;
 }
-- 
2.7.4

  parent reply	other threads:[~2016-08-25  6:30 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-08-25  6:29 [RFC PATCH v3 00/12] powerpc: "paca->soft_enabled" based local atomic operation implementation Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 01/12] powerpc: Add #defs for paca->soft_enabled flags Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 02/12] powerpc: Cleanup to use IRQ_DISABLE_MASK_* macros for paca->soft_enabled update Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 03/12] powerpc: move set_soft_enabled() Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 04/12] powerpc: Use set_soft_enabled api to update paca->soft_enabled Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 05/12] powerpc: reverse the soft_enable logic Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 06/12] powerpc: Avoid using EXCEPTION_PROLOG_1 macro in MASKABLE_* Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 07/12] powerpc: Add new _EXCEPTION_PROLOG_1 macro Madhavan Srinivasan
2016-08-25  6:29 ` [RFC PATCH v3 08/12] powerpc: Introduce new mask bit for soft_enabled Madhavan Srinivasan
2016-08-25  7:05   ` Nicholas Piggin
2016-08-26  1:27     ` Madhavan Srinivasan
2016-08-25  6:30 ` [RFC PATCH v3 09/12] powerpc: Add "bitmask" paramater to MASKABLE_* macros Madhavan Srinivasan
2016-08-25  6:30 ` [RFC PATCH v3 10/12] powerpc: Add support to mask perf interrupts Madhavan Srinivasan
2016-08-25  6:30 ` [RFC PATCH v3 11/12] powerpc: Support to replay PMIs Madhavan Srinivasan
2016-08-25  7:08   ` Nicholas Piggin
2016-08-26  1:29     ` Madhavan Srinivasan
2016-08-25  6:30 ` Madhavan Srinivasan [this message]
2016-08-25  7:15 ` [RFC PATCH v3 00/12] powerpc: "paca->soft_enabled" based local atomic operation implementation Nicholas Piggin
2016-08-26  1:31   ` Madhavan Srinivasan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1472106603-23336-13-git-send-email-maddy@linux.vnet.ibm.com \
    --to=maddy@linux.vnet.ibm.com \
    --cc=anton@samba.org \
    --cc=benh@kernel.crashing.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=mpe@ellerman.id.au \
    --cc=npiggin@gmail.com \
    --cc=paulus@samba.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.