[PATCH v2 1/2] powerpc: Add helpers for LPCR PECE1 operations

Subsystems: cpu idle time management framework, linux for powerpc (32-bit and 64-bit), the rest

STALE4210d

5 messages, 2 authors, 2015-01-23 · open the first message on its own page

[PATCH v2 1/2] powerpc: Add helpers for LPCR PECE1 operations

From: Shreyas B. Prabhu <hidden>
Date: 2015-01-19 08:05:52

PECE1 bit in LPCR is used to control whether decrementer can cause exit
from powersaving states. PECE1 bit is cleared before entering fastsleep
or deeper powersaving state and it is set on waking up. Since both
cpuidle and cpu offline operations use these powersaving states, add
helper functions to be used in both these places.

Signed-off-by: Shreyas B. Prabhu <redacted>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: linuxppc-dev@lists.ozlabs.org
---
 arch/powerpc/include/asm/reg.h       | 4 ++++
 arch/powerpc/platforms/powernv/smp.c | 4 ++--
 drivers/cpuidle/cpuidle-powernv.c    | 3 +--
 3 files changed, 7 insertions(+), 4 deletions(-)
diff --git a/arch/powerpc/include/asm/reg.h b/arch/powerpc/include/asm/reg.h
index c870e38..0847303 100644
--- a/arch/powerpc/include/asm/reg.h
+++ b/arch/powerpc/include/asm/reg.h
@@ -339,6 +339,10 @@
 #define   LPCR_LPES_SH	2
 #define   LPCR_RMI     0x00000002      /* real mode is cache inhibit */
 #define   LPCR_HDICE   0x00000001      /* Hyp Decr enable (HV,PR,EE) */
+/* LPCR PECE1 helpers. Used to disable/enable wake up due to decrementer */
+#define   LPCR_CLEAR_PECE1(old)	(old & ~(u64)LPCR_PECE1)
+#define   LPCR_SET_PECE1(old)	(old | (u64)LPCR_PECE1)
+
 #ifndef SPRN_LPID
 #define SPRN_LPID	0x13F	/* Logical Partition Identifier */
 #endif
diff --git a/arch/powerpc/platforms/powernv/smp.c b/arch/powerpc/platforms/powernv/smp.c
index 781ec45..ab61cb0 100644
--- a/arch/powerpc/platforms/powernv/smp.c
+++ b/arch/powerpc/platforms/powernv/smp.c
@@ -165,7 +165,7 @@ static void pnv_smp_cpu_kill_self(void)
 	/* We don't want to take decrementer interrupts while we are offline,
 	 * so clear LPCR:PECE1. We keep PECE2 enabled.
 	 */
-	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1);
+	mtspr(SPRN_LPCR, LPCR_CLEAR_PECE1(mfspr(SPRN_LPCR)));
 	while (!generic_check_cpu_restart(cpu)) {
 
 		ppc64_runlatch_off();
@@ -203,7 +203,7 @@ static void pnv_smp_cpu_kill_self(void)
 		if (!generic_check_cpu_restart(cpu))
 			DBG("CPU%d Unexpected exit while offline !\n", cpu);
 	}
-	mtspr(SPRN_LPCR, mfspr(SPRN_LPCR) | LPCR_PECE1);
+	mtspr(SPRN_LPCR, LPCR_SET_PECE1(mfspr(SPRN_LPCR)));
 	DBG("CPU%d coming online...\n", cpu);
 }
 
diff --git a/drivers/cpuidle/cpuidle-powernv.c b/drivers/cpuidle/cpuidle-powernv.c
index de61b9a..ed0be4c 100644
--- a/drivers/cpuidle/cpuidle-powernv.c
+++ b/drivers/cpuidle/cpuidle-powernv.c
@@ -69,11 +69,10 @@ static int fastsleep_loop(struct cpuidle_device *dev,
 	if (unlikely(system_state < SYSTEM_RUNNING))
 		return index;
 
-	new_lpcr = old_lpcr;
 	/* Do not exit powersave upon decrementer as we've setup the timer
 	 * offload.
 	 */
-	new_lpcr &= ~LPCR_PECE1;
+	new_lpcr = LPCR_CLEAR_PECE1(old_lpcr);
 
 	mtspr(SPRN_LPCR, new_lpcr);
 	power7_sleep();
-- 
1.9.3

[PATCH v2 2/2] powerpc: powernv: winkle: Restore LPCR with LPCR_PECE1 cleared

From: Shreyas B. Prabhu <hidden>
Date: 2015-01-19 08:05:58

LPCR_PECE1 bit controls whether decrementer interrupts are allowed to
cause exit from power-saving mode. While waking up from winkle, restoring
LPCR with LPCR_PECE1 set (i.e Decrementer interrupts allowed) can cause
issue in the following scenario:

- All the threads in a core are offlined. The core enters deep winkle.
- Spurious interrupt wakes up a thread in the core. Here LPCR is restored
  with LPCR_PECE1 bit set.
- Since it was a spurious interrupt on a offline thread, the thread clears
  the interrupt and goes back to winkle.
- Here before the thread executes winkle and puts the core into deep winkle,
  if a decrementer interrupt occurs on any of the sibling threads in the core
  that thread wakes up.
- Since in offline loop we are flushing interrupt only in case of external
  interrupt, the decrementer interrupt does not get flushed. So at this stage
  the thread is stuck in this is loop of waking up at 0x100 due to decrementer
  interrupt, not flushing the interrupt as only external interrupts get flushed,
  entering winkle, waking up at 0x100 again.

Fix this by programming PORE to restore LPCR with LPCR_PECE1 bit
cleared when waking up from winkle.

Signed-off-by: Shreyas B. Prabhu <redacted>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <redacted>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: linuxppc-dev@lists.ozlabs.org
---
Changes is v2:
==============
Using the helper function introduced in the previous patch.

 arch/powerpc/platforms/powernv/setup.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/arch/powerpc/platforms/powernv/setup.c b/arch/powerpc/platforms/powernv/setup.c
index ad0e32e..ded7fc8 100644
--- a/arch/powerpc/platforms/powernv/setup.c
+++ b/arch/powerpc/platforms/powernv/setup.c
@@ -298,7 +298,7 @@ int pnv_save_sprs_for_winkle(void)
 	 * all cpus at boot. Get these reg values of current cpu and use the
 	 * same accross all cpus.
 	 */
-	uint64_t lpcr_val = mfspr(SPRN_LPCR);
+	uint64_t lpcr_val = LPCR_CLEAR_PECE1(mfspr(SPRN_LPCR));
 	uint64_t hid0_val = mfspr(SPRN_HID0);
 	uint64_t hid1_val = mfspr(SPRN_HID1);
 	uint64_t hid4_val = mfspr(SPRN_HID4);
-- 
1.9.3

Re: [PATCH v2 1/2] powerpc: Add helpers for LPCR PECE1 operations

From: Michael Ellerman <mpe@ellerman.id.au>
Date: 2015-01-23 03:06:35

On Mon, 2015-01-19 at 13:35 +0530, Shreyas B. Prabhu wrote:
PECE1 bit in LPCR is used to control whether decrementer can cause exit
from powersaving states. PECE1 bit is cleared before entering fastsleep
or deeper powersaving state and it is set on waking up. Since both
cpuidle and cpu offline operations use these powersaving states, add
helper functions to be used in both these places.
Thanks.

That isn't really much clearer than the original, so in the end I just merged
your original fix.

I'll think if there's a bigger consolidation we can do that makes it clearer.

cheers

Re: [PATCH v2 1/2] powerpc: Add helpers for LPCR PECE1 operations

From: Shreyas B Prabhu <hidden>
Date: 2015-01-23 03:47:24


On Friday 23 January 2015 08:36 AM, Michael Ellerman wrote:
On Mon, 2015-01-19 at 13:35 +0530, Shreyas B. Prabhu wrote:
quoted
PECE1 bit in LPCR is used to control whether decrementer can cause exit
from powersaving states. PECE1 bit is cleared before entering fastsleep
or deeper powersaving state and it is set on waking up. Since both
cpuidle and cpu offline operations use these powersaving states, add
helper functions to be used in both these places.
Thanks.

That isn't really much clearer than the original, so in the end I just merged
your original fix.

I'll think if there's a bigger consolidation we can do that makes it clearer.

cheers
Helper could have been this :

#define   LPCR_CLEAR_PECE1	(mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1)

This perhaps would make it more clearer, but it will end up using additional mfspr here-

    static int fastsleep_loop(struct cpuidle_device *dev,
    				struct cpuidle_driver *drv,
    				int index)
    {
    	...
    
    	new_lpcr = old_lpcr;
    	/* Do not exit powersave upon decrementer as we've setup the timer
    	 * offload.
    	 */
    	new_lpcr &= ~LPCR_PECE1;
    
    	mtspr(SPRN_LPCR, new_lpcr);

Re: [PATCH v2 1/2] powerpc: Add helpers for LPCR PECE1 operations

From: Michael Ellerman <mpe@ellerman.id.au>
Date: 2015-01-23 04:20:27

On Fri, 2015-01-23 at 09:16 +0530, Shreyas B Prabhu wrote:
On Friday 23 January 2015 08:36 AM, Michael Ellerman wrote:
quoted
On Mon, 2015-01-19 at 13:35 +0530, Shreyas B. Prabhu wrote:
quoted
PECE1 bit in LPCR is used to control whether decrementer can cause exit
from powersaving states. PECE1 bit is cleared before entering fastsleep
or deeper powersaving state and it is set on waking up. Since both
cpuidle and cpu offline operations use these powersaving states, add
helper functions to be used in both these places.
Thanks.

That isn't really much clearer than the original, so in the end I just merged
your original fix.

I'll think if there's a bigger consolidation we can do that makes it clearer.
Helper could have been this :

#define   LPCR_CLEAR_PECE1	(mfspr(SPRN_LPCR) & ~(u64)LPCR_PECE1)

This perhaps would make it more clearer, but it will end up using additional mfspr here-
 
Yeah.

Maybe we just leave it open coded, I'll have a think about it.

cheers
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help