Skip to contentSkip to conversation

kvm:ppc:booke-64: soft-disable interrupts

RFC PATCH In review Tiejun Chen 7 participants 34 messages May 6, 2013 – May 10, 2013

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 kvm:ppc:booke-64: soft-disable interrupts

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

---
 arch/powerpc/kvm/bookehv_interrupts.S |    9 +++++++++
 1 file changed, 9 insertions(+)
arch/powerpc/kvm/bookehv_interrupts.S +9 -0
index e8ed7d6..2fd62bf 100644
@@ -33,6 +33,8 @@
 
 #ifdef CONFIG_64BIT
 #include <asm/exception-64e.h>
+#include <asm/hw_irq.h>
+#include <asm/irqflags.h>
 #else
 #include "../kernel/head_booke.h" /* for THREAD_NORMSAVE() */
 #endif
@@ -469,6 +471,13 @@ _GLOBAL(kvmppc_resume_host)
 	PPC_LL	r3, HOST_RUN(r1)
 	mr	r5, r14 /* intno */
 	mr	r14, r4 /* Save vcpu pointer. */
+#ifdef CONFIG_64BIT
+	/* Should we soft-disable interrupts? */
+	andi.	r6, r5, BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL
+	beq	skip_soft_dis
+	SOFT_DISABLE_INTS(r7,r8)
+skip_soft_dis:
+#endif
 	bl	kvmppc_handle_exit
 
 	/* Restore vcpu pointer and the nonvolatiles we used. */
Signature · 1 line
1.7.9.5
#1 lore raw
Quoted 4 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

Sorry, miss to send Ben.

Tiejun

Quoted 34 lines
---
  arch/powerpc/kvm/bookehv_interrupts.S |    9 +++++++++
  1 file changed, 9 insertions(+)
arch/powerpc/kvm/bookehv_interrupts.S +9 -0
index e8ed7d6..2fd62bf 100644
@@ -33,6 +33,8 @@
 
  #ifdef CONFIG_64BIT
  #include <asm/exception-64e.h>
+#include <asm/hw_irq.h>
+#include <asm/irqflags.h>
  #else
  #include "../kernel/head_booke.h" /* for THREAD_NORMSAVE() */
  #endif
@@ -469,6 +471,13 @@ _GLOBAL(kvmppc_resume_host)
  	PPC_LL	r3, HOST_RUN(r1)
  	mr	r5, r14 /* intno */
  	mr	r14, r4 /* Save vcpu pointer. */
+#ifdef CONFIG_64BIT
+	/* Should we soft-disable interrupts? */
+	andi.	r6, r5, BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL
+	beq	skip_soft_dis
+	SOFT_DISABLE_INTS(r7,r8)
+skip_soft_dis:
+#endif
  	bl	kvmppc_handle_exit
 
  	/* Restore vcpu pointer and the nonvolatiles we used. */
#2 lore raw
Quoted 5 lines · tiejun.chen
Quoted 4 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

http://patchwork.ozlabs.org/patch/241344/
http://patchwork.ozlabs.org/patch/241412/

:-)

Quoted 22 lines
Quoted 22 lines
---
  arch/powerpc/kvm/bookehv_interrupts.S |    9 +++++++++
  1 file changed, 9 insertions(+)
arch/powerpc/kvm/bookehv_interrupts.S +2 -0
index e8ed7d6..2fd62bf 100644
@@ -33,6 +33,8 @@
 
  #ifdef CONFIG_64BIT
  #include <asm/exception-64e.h>
+#include <asm/hw_irq.h>
+#include <asm/irqflags.h>
  #else
  #include "../kernel/head_booke.h" /* for THREAD_NORMSAVE() */
  #endif
@@ -469,6 +471,13 @@ _GLOBAL(kvmppc_resume_host)
  	PPC_LL	r3, HOST_RUN(r1)
  	mr	r5, r14 /* intno */
  	mr	r14, r4 /* Save vcpu pointer. */
Quoted 8 lines

arch/powerpc/kvm/bookehv_interrupts.S _GLOBAL(kvmppc_resume_host) · line 480

Quoted 8 lines
arch/powerpc/kvm/bookehv_interrupts.S +3 -0
+#ifdef CONFIG_64BIT
+	/* Should we soft-disable interrupts? */
+	andi.	r6, r5, BOOKE_INTERRUPT_EXTERNAL |  

BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

arch/powerpc/kvm/bookehv_interrupts.S +4 -0
+	beq	skip_soft_dis
+	SOFT_DISABLE_INTS(r7,r8)
+skip_soft_dis:
+#endif

Why wouldn't we always disable them? kvmppc_handle_exit() will enable interrupts when it's ready.

-Scott

#3 lore raw
Quoted 11 lines · Scott Wood
Quoted 5 lines · tiejun.chen
Quoted 4 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

http://patchwork.ozlabs.org/patch/241344/
http://patchwork.ozlabs.org/patch/241412/

:-)

I'm observing the same behaviour as well:

	WARN_ON_ONCE(!irqs_disabled());
Quoted 27 lines
Quoted 27 lines
Quoted 27 lines
---
  arch/powerpc/kvm/bookehv_interrupts.S |    9 +++++++++
  1 file changed, 9 insertions(+)
arch/powerpc/kvm/bookehv_interrupts.S +5 -0
index e8ed7d6..2fd62bf 100644
@@ -33,6 +33,8 @@
 
  #ifdef CONFIG_64BIT
  #include <asm/exception-64e.h>
+#include <asm/hw_irq.h>
+#include <asm/irqflags.h>
  #else
  #include "../kernel/head_booke.h" /* for THREAD_NORMSAVE() */
  #endif
@@ -469,6 +471,13 @@ _GLOBAL(kvmppc_resume_host)
      PPC_LL    r3, HOST_RUN(r1)
      mr    r5, r14 /* intno */
      mr    r14, r4 /* Save vcpu pointer. */
+#ifdef CONFIG_64BIT
+    /* Should we soft-disable interrupts? */
+    andi.    r6, r5, BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER
Quoted 8 lines
Quoted 5 lines

arch/powerpc/kvm/bookehv_interrupts.S _GLOBAL(kvmppc_resume_host) · line 480

Quoted 5 lines

| BOOKE_INTERRUPT_DOORBELL

arch/powerpc/kvm/bookehv_interrupts.S +4 -0
+    beq    skip_soft_dis
+    SOFT_DISABLE_INTS(r7,r8)
+skip_soft_dis:
+#endif

Why wouldn't we always disable them? kvmppc_handle_exit() will enable interrupts when it's ready.

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

b. bl kvmppc_handle_exit

c. kvmppc_handle_exit()

{
        int r = RESUME_HOST;
        int s;

        /* update before a new last_exit_type is rewritten */
        kvmppc_update_timing_stats(vcpu);

        /* restart interrupts if they were meant for the host */
        kvmppc_restart_interrupt(vcpu, exit_nr);

        local_irq_enable();	==> Enable again.

....

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr, ack)           \
        START_EXCEPTION(label);                                         \
        NORMAL_EXCEPTION_PROLOG(trapnum, intnum, PROLOG_ADDITION_MASKABLE)\
        EXCEPTION_COMMON(trapnum, PACA_EXGEN, *INTS_DISABLE*)             \
	...

So I think this should be reasonable :)

Tiejun

#4 lore raw
Quoted 8 lines · tiejun.chen
Quoted 7 lines · Scott Wood
Quoted 5 lines · tiejun.chen
Quoted 4 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

Quoted 8 lines
Quoted 4 lines

http://patchwork.ozlabs.org/patch/241344/
http://patchwork.ozlabs.org/patch/241412/

:-)

I'm observing the same behaviour as well:

	WARN_ON_ONCE(!irqs_disabled());

So, could you explain the benefits of your approach over what's being discussed in those threads?

Quoted 6 lines
Quoted 2 lines

Why wouldn't we always disable them? kvmppc_handle_exit() will enable interrupts when it's ready.

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

And I'd rather see any fix for this problem stay out of the asm code.

Quoted 18 lines

b. bl kvmppc_handle_exit

c. kvmppc_handle_exit()

{
        int r = RESUME_HOST;
        int s;

        /* update before a new last_exit_type is rewritten */
        kvmppc_update_timing_stats(vcpu);

        /* restart interrupts if they were meant for the host */
        kvmppc_restart_interrupt(vcpu, exit_nr);

        local_irq_enable();	==> Enable again.

....

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

Quoted 8 lines

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr, ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

-Scott

#5 lore raw
Quoted 13 lines · Scott Wood
Quoted 12 lines · tiejun.chen
Quoted 11 lines · Scott Wood
Quoted 6 lines · tiejun.chen
Quoted 5 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

http://patchwork.ozlabs.org/patch/241344/
http://patchwork.ozlabs.org/patch/241412/

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

:-)

I'm observing the same behaviour as well:

    WARN_ON_ONCE(!irqs_disabled());

So, could you explain the benefits of your approach over what's being discussed in those threads?

They're a long thread so I think I need to take time to see :)

Quoted 4 lines
Quoted 4 lines
Quoted 3 lines

Why wouldn't we always disable them? kvmppc_handle_exit() will enable interrupts when it's ready.

Quoted 8 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

This is like host handler, so I'm just disabling soft interrupt during kvmppc_restart_interrupt() for Doorbell interrupt/Decrementer Interrupt/External Input Interrupt.

I don't see anything should be disabled for any TLB exception in host handler.

Quoted 2 lines

And I'd rather see any fix for this problem stay out of the asm code.

We already have an appropriate SOFT_DISABLE_INTS so I think we can take this easily :)

Quoted 19 lines
Quoted 19 lines

b. bl kvmppc_handle_exit

c. kvmppc_handle_exit()

{
        int r = RESUME_HOST;
        int s;

        /* update before a new last_exit_type is rewritten */
        kvmppc_update_timing_stats(vcpu);

        /* restart interrupts if they were meant for the host */
        kvmppc_restart_interrupt(vcpu, exit_nr);

        local_irq_enable();    ==> Enable again.

....

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

Quoted 7 lines
Quoted 5 lines
#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr, ack)           \
        START_EXCEPTION(label);                                         \
        NORMAL_EXCEPTION_PROLOG(trapnum, intnum, PROLOG_ADDITION_MASKABLE)\
        EXCEPTION_COMMON(trapnum, PACA_EXGEN, *INTS_DISABLE*)             \
    ...

Could you elaborate on what you mean?

In host handler, we always use MASKABLE_EXCEPTION() to define-to-handle some exceptions: Doorbell interrupt/Decrementer Interrupt/External Input Interrupt:

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr, ack)           \
        START_EXCEPTION(label);                                         \
        NORMAL_EXCEPTION_PROLOG(trapnum, intnum, PROLOG_ADDITION_MASKABLE)\
        EXCEPTION_COMMON(trapnum, PACA_EXGEN, *INTS_DISABLE*)             \

This would call INTS_DISABLE, which is equal to SOFT_DISABLE_INTS(), to disable soft interrupt before call all associated handlers: do_IRQ()/timer_interrupt()/doorbell_exception().

But DO_KVM hook always skips INTS_DISABLE.

So I think we also need to do INTS_DISABLE for kvmppc_restart_interrupt() since actually that restarts interrupts for the host with a similar way as they are called by host.

Tiejun

#6 lore raw
Quoted 35 lines · tiejun.chen
Quoted 13 lines · Scott Wood
Quoted 12 lines · tiejun.chen
Quoted 11 lines · Scott Wood
Quoted 6 lines · tiejun.chen
Quoted 5 lines · Tiejun Chen

For the external interrupt, the decrementer exception and the doorbell excpetion, we also need to soft-disable interrupts while doing as host interrupt handlers since the DO_KVM hook is always performed to skip EXCEPTION_COMMON then miss this original chance with the 'ints' (INTS_DISABLE).

http://patchwork.ozlabs.org/patch/241344/
http://patchwork.ozlabs.org/patch/241412/

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

:-)

I'm observing the same behaviour as well:

    WARN_ON_ONCE(!irqs_disabled());

So, could you explain the benefits of your approach over what's being discussed in those threads?

They're a long thread so I think I need to take time to see :)

Quoted 10 lines
Quoted 9 lines
Quoted 3 lines

Why wouldn't we always disable them? kvmppc_handle_exit() will enable interrupts when it's ready.

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Quoted 6 lines
Quoted 2 lines

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

This is like host handler, so I'm just disabling soft interrupt during kvmppc_restart_interrupt() for Doorbell interrupt/Decrementer Interrupt/External Input Interrupt.

I don't see anything should be disabled for any TLB exception in host handler.

Every exception type needs consistent lazy EE state once we hit the local_irq_enable() (or any other C code that cares).

Plus, if you're going to add code to make something conditional, you should have a good reason for making it conditional. Being more like the 64-bit host handler for its own sake isn't good enough, especially if you introduce differences between 32 and 64 bit in the process.

Quoted 3 lines
Quoted 1 line

And I'd rather see any fix for this problem stay out of the asm code.

We already have an appropriate SOFT_DISABLE_INTS so I think we can take this easily :)

An unconditional call to hard_irq_disable() at the beginning of kvmppc_handle_exit() should suffice. I already have this in the next version of my patch that I'll be posting shortly.

Note that we need this on 32-bit as well, so that trace_hardirqs_off() gets called.

-Scott

#7 lore raw
In reply to Scott Wood
Quoted 15 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 8 lines
Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca->irq_happened.

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss). Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

#8 lore raw
Quoted 34 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of Caraman Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 8 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca->irq_happened.

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss).

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list of masked interruts.

-Bharat

Quoted 11 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#9 lore raw
Quoted 36 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of Caraman Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 8 lines
Quoted 6 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca->irq_happened.

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss).

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list of masked interruts.

Are you saying perfmon? If so, its also in that list:

        START_EXCEPTION(perfmon);
        NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                PROLOG_ADDITION_NONE)
        EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

#10 lore raw
Quoted 59 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 41 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 5 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

Quoted 6 lines

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss).

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

Quoted 8 lines

of masked interruts.

Are you saying perfmon? If so, its also in that list:

         START_EXCEPTION(perfmon);
         NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                 PROLOG_ADDITION_NONE)
         EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Quoted 13 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

#11 lore raw
Quoted 61 lines · On 05/09/2013 03:51 PM, Bhushan Bharat-R65777 wrote:
Quoted 61 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 41 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 5 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

Quoted 6 lines

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss).

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Quoted 8 lines
Quoted 6 lines

Are you saying perfmon? If so, its also in that list:

          START_EXCEPTION(perfmon);
          NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                  PROLOG_ADDITION_NONE)
          EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

In stead of PROLOG_ADDITION_MASKABLE, actually we have nothing to do with PROLOG_ADDITION_NONE for perfmon, so perfmon can be executed without lazy state.

Tiejun

Quoted 15 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

#12 lore raw
In reply to Bhushan Bharat-R65777
Quoted 61 lines · On Thu, May 09, 2013 at 07:51:09AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 61 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 41 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 5 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

Quoted 6 lines

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss).

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Quoted 8 lines
Quoted 6 lines

Are you saying perfmon? If so, its also in that list:

         START_EXCEPTION(perfmon);
         NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                 PROLOG_ADDITION_NONE)
         EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Actually we don't want replay the perfmon interrupt later. We would run it even soft irq is disabled and just treat it as NMI. Please see the following function quoted from arch/powerpc/perf/core-fsl-emb.c:
/*
* If interrupts were soft-disabled when a PMU interrupt occurs, treat
* it as an NMI.
*/
static inline int perf_intr_is_nmi(struct pt_regs *regs)
{

  #ifdef __powerpc64__                                                                                                                                                         
          return !regs->softe;
  #else
          return 0;

#endif
}

Thanks,
Kevin

Quoted 19 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#13 lore raw
Quoted 94 lines

-----Original Message-----
From: Kevin Hao [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:38 PM
To: Bhushan Bharat-R65777
Cc: tiejun.chen; Caraman Mihai Claudiu-B02008; [email protected]; Wood Scott- B07421; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 51 lines · On Thu, May 09, 2013 at 07:51:09AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 51 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 38 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 38 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf bounces+Of Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Quoted 8 lines
Quoted 6 lines

Are you saying perfmon? If so, its also in that list:

         START_EXCEPTION(perfmon);
         NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                 PROLOG_ADDITION_NONE)
         EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Actually we don't want replay the perfmon interrupt later. We would run it even soft irq is disabled and just treat it as NMI. Please see the following function quoted from arch/powerpc/perf/core-fsl-emb.c:
/*
* If interrupts were soft-disabled when a PMU interrupt occurs, treat
* it as an NMI.
*/
static inline int perf_intr_is_nmi(struct pt_regs *regs)
{

Quoted 6 lines
  #ifdef __powerpc64__
          return !regs->softe;
  #else
          return 0;

#endif
}

Is it because that we cannot afford to lose perfmon interrupt for more accurate capturing of data ?

-Bharat

Quoted 23 lines

Thanks,
Kevin

Quoted 19 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#14 lore raw
Quoted 101 lines · On 05/09/2013 04:12 PM, Bhushan Bharat-R65777 wrote:
Quoted 88 lines

-----Original Message-----
From: Kevin Hao [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:38 PM
To: Bhushan Bharat-R65777
Cc: tiejun.chen; Caraman Mihai Claudiu-B02008; [email protected]; Wood Scott- B07421; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 51 lines · On Thu, May 09, 2013 at 07:51:09AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 51 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 38 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 38 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf bounces+Of Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Quoted 8 lines
Quoted 6 lines

Are you saying perfmon? If so, its also in that list:

          START_EXCEPTION(perfmon);
          NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                  PROLOG_ADDITION_NONE)
          EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Actually we don't want replay the perfmon interrupt later. We would run it even soft irq is disabled and just treat it as NMI. Please see the following function quoted from arch/powerpc/perf/core-fsl-emb.c:

Quoted 12 lines
   /*
    * If interrupts were soft-disabled when a PMU interrupt occurs, treat
    * it as an NMI.
    */
   static inline int perf_intr_is_nmi(struct pt_regs *regs)
   {
   #ifdef __powerpc64__
           return !regs->softe;
   #else
           return 0;
   #endif
   }
Quoted 1 line

Is it because that we cannot afford to lose perfmon interrupt for more accurate capturing of data ?

powerpc/perf: e500 support

This implements perf_event support for the Freescale embedded performance monitor, based on the existing perf_event.c that supports server/classic chips.

Some limitations:
- Performance monitor interrupts are regular EE interrupts, and thus you can't profile places with interrupts disabled. We may want to implement soft IRQ-disabling, with perfmon interrupts exempted and treated as NMIs.

Tiejun

Quoted 29 lines

-Bharat

Quoted 23 lines

Thanks,
Kevin

Quoted 19 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#15 lore raw
Quoted 121 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:48 PM
To: Bhushan Bharat-R65777
Cc: Kevin Hao; Caraman Mihai Claudiu-B02008; [email protected]; Wood Scott- B07421; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 75 lines · On 05/09/2013 04:12 PM, Bhushan Bharat-R65777 wrote:
Quoted 75 lines

-----Original Message-----
From: Kevin Hao [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:38 PM
To: Bhushan Bharat-R65777
Cc: tiejun.chen; Caraman Mihai Claudiu-B02008; [email protected]; Wood Scott- B07421; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 55 lines · On Thu, May 09, 2013 at 07:51:09AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 55 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 42 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 42 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf bounces+Of Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 15 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 8 lines
Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

Quoted 5 lines
Quoted 5 lines
Quoted 5 lines

irq_happened.

Quoted 4 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 1 line
Quoted 1 line
Quoted 1 line
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 2 lines

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Are you saying perfmon? If so, its also in that list:

          START_EXCEPTION(perfmon);
          NORMAL_EXCEPTION_PROLOG(0x260,

BOOKE_INTERRUPT_PERFORMANCE_MONITOR,

Quoted 21 lines
Quoted 8 lines
Quoted 4 lines
Quoted 2 lines
                                  PROLOG_ADDITION_NONE)
          EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Actually we don't want replay the perfmon interrupt later. We would run it even soft irq is disabled and just treat it as NMI. Please see the following function quoted from arch/powerpc/perf/core-fsl-emb.c:

Quoted 12 lines
   /*
    * If interrupts were soft-disabled when a PMU interrupt occurs, treat
    * it as an NMI.
    */
   static inline int perf_intr_is_nmi(struct pt_regs *regs)
   {
   #ifdef __powerpc64__
           return !regs->softe;
   #else
           return 0;
   #endif
   }
Quoted 1 line

Is it because that we cannot afford to lose perfmon interrupt for more

accurate capturing of data ?

powerpc/perf: e500 support

Quoted 8 lines

This implements perf_event support for the Freescale embedded performance monitor, based on the existing perf_event.c that supports server/classic chips.

Some limitations:
- Performance monitor interrupts are regular EE interrupts, and thus you can't profile places with interrupts disabled. We may want to implement soft IRQ-disabling, with perfmon interrupts exempted and treated as NMIs.

Ahh, that gives the answer and same as I expected :)

-Bharat

Quoted 33 lines

Tiejun

Quoted 29 lines

-Bharat

Quoted 23 lines

Thanks,
Kevin

Quoted 19 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#16 lore raw
In reply to Bhushan Bharat-R65777
Quoted 101 lines · On Thu, May 09, 2013 at 08:12:51AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 94 lines

-----Original Message-----
From: Kevin Hao [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:38 PM
To: Bhushan Bharat-R65777
Cc: tiejun.chen; Caraman Mihai Claudiu-B02008; [email protected]; Wood Scott- B07421; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 51 lines · On Thu, May 09, 2013 at 07:51:09AM +0000, Bhushan Bharat-R65777 wrote:
Quoted 51 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 1:18 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 38 lines · On 05/09/2013 03:33 PM, Bhushan Bharat-R65777 wrote:
Quoted 38 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf bounces+Of Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 5 lines
Quoted 5 lines
Quoted 3 lines

Embedded Perfmon interrupt is also asynchronous, Why that is not in the list

of masked interruts.

Quoted 8 lines
Quoted 6 lines

Are you saying perfmon? If so, its also in that list:

         START_EXCEPTION(perfmon);
         NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
                                 PROLOG_ADDITION_NONE)
         EXCEPTION_COMMON(0x260, PACA_EXGEN, INTS_DISABLE)

Where it is recorded in paca->irq_happned to be replayed later ?

Actually we don't want replay the perfmon interrupt later. We would run it even soft irq is disabled and just treat it as NMI. Please see the following function quoted from arch/powerpc/perf/core-fsl-emb.c:
/*
* If interrupts were soft-disabled when a PMU interrupt occurs, treat
* it as an NMI.
*/
static inline int perf_intr_is_nmi(struct pt_regs *regs)
{

Quoted 6 lines
  #ifdef __powerpc64__
          return !regs->softe;
  #else
          return 0;

#endif
}

Quoted 1 line

Is it because that we cannot afford to lose perfmon interrupt for more accurate capturing of data ?

Yes, I think this will definitely improve the perf sample quality.

Thanks,
Kevin

Quoted 27 lines

-Bharat

Quoted 23 lines

Thanks,
Kevin

Quoted 19 lines
Quoted 14 lines

Tiejun

Quoted 9 lines

-Bharat

Quoted 6 lines

Now on KVM book3e we
don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

-Mike

_______________________________________________
Linuxppc-dev mailing list
[email protected]
https://lists.ozlabs.org/listinfo/linuxppc-dev

#17 lore raw
Quoted 4 lines · Kevin Hao
Quoted 1 line

Is it because that we cannot afford to lose perfmon interrupt for

more accurate capturing of data ?

Yes, I think this will definitely improve the perf sample quality.

This is one of the primary reason why we implemented lazy disabling in the first place and why I recently reworked it to decrease the periods where we are hard disabled.

The other reasons are that storing bytes to the PACA is faster than manipulating EE on many processors.

Cheers,
Ben.

#18 lore raw
In reply to Caraman Mihai Claudiu-B02008
Quoted 40 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of Caraman Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 15 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 8 lines
Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca->irq_happened.

Quoted 5 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss). Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Another Question:

The case is:

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right? -> local_irq_enable() - This checks that irq_happened is set, and replays

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in irq_happened, But do not clear EE in SRR1 and rfi. So interrupts are not hard disabled. -> Now say EE interrupt happens, there we set PACA_IRQ_EE in irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. -> local_irq_enable() - This checks that irq_happened is set.
IIUC, it replays only one interrupt? is not it?

-Bharat

#19 lore raw
Quoted 47 lines · On 05/09/2013 04:23 PM, Bhushan Bharat-R65777 wrote:
Quoted 40 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of Caraman Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 15 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 8 lines
Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca->irq_happened.

Quoted 5 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss). Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Quoted 4 lines

Another Question:

The case is:

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I recall.

Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

I don't understand "the interrupt condition on device is cleared" here.

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Quoted 1 line

-> local_irq_enable() - This checks that irq_happened is set, and replays

ret_from_except also check to replay.

Quoted 8 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in irq_happened, But do not clear EE in SRR1 and rfi. So interrupts are not hard disabled. -> Now say EE interrupt happens, there we set PACA_IRQ_EE in irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. -> local_irq_enable() - This checks that irq_happened is set.
IIUC, it replays only one interrupt? is not it?

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard interrupt there:

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Additionally, ret_from_except probably check to replay all.

Tiejun

#20 lore raw
Quoted 77 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:15 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 50 lines · On 05/09/2013 04:23 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 5 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 6 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss). Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Quoted 4 lines

Another Question:

The case is:

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I recall.

Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

Quoted 4 lines

I don't understand "the interrupt condition on device is cleared" here.

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Once yes, but I think to avoid flood of device interrupt we disable MSR.EE when soft-disabled.

Quoted 17 lines
Quoted 3 lines

-> local_irq_enable() - This checks that irq_happened is set, and
replays

ret_from_except also check to replay.

Quoted 5 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in irq_happened, But do

not clear EE in SRR1 and rfi. So interrupts are not hard disabled.

Quoted 1 line

-> Now say EE interrupt happens, there we set PACA_IRQ_EE in irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled.

Quoted 2 lines

-> local_irq_enable() - This checks that irq_happened is set.
IIUC, it replays only one interrupt? is not it?

Quoted 7 lines

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard interrupt there:

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Do you mean that the interrupt should fire again?

Quoted 2 lines

Additionally, ret_from_except probably check to replay all.

Local_irq_enable() will not take us to ret_from_except.

-Bharat

Quoted 2 lines

Tiejun

#21 lore raw
Quoted 77 lines · On 05/09/2013 06:00 PM, Bhushan Bharat-R65777 wrote:
Quoted 77 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:15 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 50 lines · On 05/09/2013 04:23 PM, Bhushan Bharat-R65777 wrote:
Quoted 36 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 22 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 12 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...
Quoted 5 lines
Quoted 1 line

Could you elaborate on what you mean?

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 6 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB miss). Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Quoted 4 lines

Another Question:

The case is:

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I recall.

Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

Quoted 6 lines
Quoted 4 lines

I don't understand "the interrupt condition on device is cleared" here.

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Once yes, but I think to avoid flood of device interrupt we disable MSR.EE when soft-disabled.

But we neither ACK nor send EOI to that irq in the interrupt controller, so that should be in pending state.

Quoted 21 lines
Quoted 21 lines
Quoted 4 lines

-> local_irq_enable() - This checks that irq_happened is set, and
replays

ret_from_except also check to replay.

Quoted 5 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in irq_happened, But do

not clear EE in SRR1 and rfi. So interrupts are not hard disabled.

Quoted 1 line

-> Now say EE interrupt happens, there we set PACA_IRQ_EE in irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled.

Quoted 2 lines
   -> local_irq_enable() - This checks that irq_happened is set.

IIUC, it replays only one interrupt? is not it?

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard interrupt there:

Quoted 6 lines
Quoted 4 lines

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Do you mean that the interrupt should fire again?

I means the pending exception including external interrupt, the decrementer exception and the doorbell exception, can trap CPU once EE=1 with __hard_irq_enable() here. Then the kernel can handle those exception since soft enable is also 1 now.

Quoted 5 lines
Quoted 3 lines

Additionally, ret_from_except probably check to replay all.

Local_irq_enable() will not take us to ret_from_except.

Yes. I just say ret_from_except can provide an approach to replay all :)

Tiejun

#22 lore raw
Quoted 141 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:48 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 51 lines · On 05/09/2013 06:00 PM, Bhushan Bharat-R65777 wrote:
Quoted 51 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:15 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 38 lines · On 05/09/2013 04:23 PM, Bhushan Bharat-R65777 wrote:
Quoted 38 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 11 lines
Quoted 11 lines
Quoted 9 lines
Quoted 3 lines

Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Another Question:

The case is:

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I

recall.

Quoted 10 lines
Quoted 10 lines
Quoted 5 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in
irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

Quoted 6 lines
Quoted 4 lines

I don't understand "the interrupt condition on device is cleared" here.

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Once yes, but I think to avoid flood of device interrupt we disable MSR.EE

when soft-disabled.

But we neither ACK nor send EOI to that irq in the interrupt controller, so that should be in pending state.

Quoted 23 lines
Quoted 23 lines
Quoted 4 lines

-> local_irq_enable() - This checks that irq_happened is set, and
replays

ret_from_except also check to replay.

Quoted 6 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in
irq_happened, But do

not clear EE in SRR1 and rfi. So interrupts are not hard disabled.

Quoted 2 lines

-> Now say EE interrupt happens, there we set PACA_IRQ_EE in
irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled.

Quoted 2 lines
   -> local_irq_enable() - This checks that irq_happened is set.

IIUC, it replays only one interrupt? is not it?

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard interrupt there:

Quoted 6 lines
Quoted 4 lines

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Do you mean that the interrupt should fire again?

I means the pending exception including external interrupt, the decrementer exception and the doorbell exception, can trap CPU once EE=1 with __hard_irq_enable() here. Then the kernel can handle those exception since soft enable is also 1 now.

Quoted 2 lines
Quoted 2 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

Additionally, ret_from_except probably check to replay all.

Local_irq_enable() will not take us to ret_from_except.

Yes. I just say ret_from_except can provide an approach to replay all :)

__replay_interrupt() from arch_local_irq_enable() will take us to ret_from_except/lite :) There all pending interrupts are replayed one by one before we hard-enable and soft-enable interrupts.

-Bharat

Quoted 2 lines

Tiejun

#23 lore raw
Quoted 141 lines · On 05/09/2013 07:21 PM, Bhushan Bharat-R65777 wrote:
Quoted 141 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:48 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 51 lines · On 05/09/2013 06:00 PM, Bhushan Bharat-R65777 wrote:
Quoted 51 lines

-----Original Message-----
From: tiejun.chen [mailto:[email protected]]
Sent: Thursday, May 09, 2013 3:15 PM
To: Bhushan Bharat-R65777
Cc: Caraman Mihai Claudiu-B02008; Wood Scott-B07421; linuxppc- [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 38 lines · On 05/09/2013 04:23 PM, Bhushan Bharat-R65777 wrote:
Quoted 38 lines

-----Original Message-----
From: Linuxppc-dev [mailto:linuxppc-dev- [email protected]] On Behalf Of bounces+Caraman
Mihai Claudiu-B02008
Sent: Wednesday, May 08, 2013 6:44 PM
To: Wood Scott-B07421; tiejun.chen
Cc: [email protected]; [email protected]; [email protected]; [email protected]
Subject: RE: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64: soft-disable interrupts

Quoted 23 lines
Quoted 5 lines

This only disable soft interrupt for kvmppc_restart_interrupt() that restarts interrupts if they were meant for the host:

a. SOFT_DISABLE_INTS() only for BOOKE_INTERRUPT_EXTERNAL | BOOKE_INTERRUPT_DECREMENTER | BOOKE_INTERRUPT_DOORBELL

Those aren't the only exceptions that can end up going to the host. We could get a TLB miss that results in a heavyweight MMIO exit, etc.

Quoted 6 lines

And shouldn't we handle kvmppc_restart_interrupt() like the original HOST flow?

#define MASKABLE_EXCEPTION(trapnum, intnum, label, hdlr,
ack) \

Quoted 6 lines

START_EXCEPTION(label); \
NORMAL_EXCEPTION_PROLOG(trapnum, intnum,
PROLOG_ADDITION_MASKABLE)\
EXCEPTION_COMMON(trapnum, PACA_EXGEN,

*INTS_DISABLE*)             \
	...

Could you elaborate on what you mean?

Quoted 7 lines
Quoted 7 lines
Quoted 7 lines
Quoted 3 lines

I think Tiejun was saying that host has flags and replays only EE/DEC/DBELL interrupts. There is special macro masked_interrupt_book3e in those exception handlers that sets paca-

irq_happened.

Quoted 3 lines

The list of replied interrupts is limited to asynchronous noncritical interrupts which can be masked by MSR[EE] (therefore no TLB

miss).

Quoted 11 lines
Quoted 11 lines
Quoted 9 lines
Quoted 3 lines

Now on KVM book3e we don't want to put them in the irq_happened lazy state but rather to execute them directly, so there is no reason for exception handling symmetry between host and guest.

Another Question:

The case is:

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I

recall.

Quoted 10 lines
Quoted 10 lines
Quoted 5 lines

Case 1)

    -> Local_irq_disable()  will set soft_enabled = 0
    -> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

Quoted 6 lines
Quoted 4 lines

I don't understand "the interrupt condition on device is cleared" here.

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Once yes, but I think to avoid flood of device interrupt we disable MSR.EE

when soft-disabled.

But we neither ACK nor send EOI to that irq in the interrupt controller, so that should be in pending state.

Quoted 23 lines
Quoted 23 lines
Quoted 4 lines

    -> local_irq_enable() - This checks that irq_happened is set, and

replays

ret_from_except also check to replay.

Quoted 6 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0

    -> Now DEC interrupt happens. We set PACA_IRQ_DEC in

irq_happened, But do

not clear EE in SRR1 and rfi. So interrupts are not hard disabled.

Quoted 2 lines
    -> Now say EE interrupt happens, there we set PACA_IRQ_EE in

irq_happened,

Also clears EE in SRR1 and rfi. So interrupts are hard disabled.

Quoted 2 lines
    -> local_irq_enable() - This checks that irq_happened is set.

IIUC, it replays only one interrupt? is not it?

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard interrupt there:

Quoted 6 lines
Quoted 4 lines

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Do you mean that the interrupt should fire again?

I means the pending exception including external interrupt, the decrementer exception and the doorbell exception, can trap CPU once EE=1 with __hard_irq_enable() here. Then the kernel can handle those exception since soft enable is also 1 now.

Quoted 2 lines
Quoted 2 lines
Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

Additionally, ret_from_except probably check to replay all.

Local_irq_enable() will not take us to ret_from_except.

Yes. I just say ret_from_except can provide an approach to replay all :)

__replay_interrupt() from arch_local_irq_enable() will take us to ret_from_except/lite :) There all pending interrupts are replayed one by one before we hard-enable and soft-enable interrupts.

Yes, but a point needs to be corrected,

_replay_interrupt() is following set_soft_enabled(1), so __replay_interrupt() can go the exception entry to call the handler.

Tiejun

#24 lore raw
In reply to tiejun.chen
Quoted 3 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur as I recall.

Only if directed to the hypervisor.

Quoted 7 lines
Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting interrupt till the interrupt condition on device is cleared, right?

The external interrupt line is level sensitive normally, so we have to mask MSR:EE in that case or the interrupt would keep re-occurring (note that FSL has this concept of auto-acked interrupts via the on die MPIC for which you can potentially use PACA_IRQ_EE_EDGE instead and avoid having to mask MSR:EE).

Quoted 2 lines

I don't understand "the interrupt condition on device is cleared" here.

Well, the external interrupt line will remain asserted until the device (via the PIC) stops interrupting the processor :-) Either because we mask at the PIC level (or raise the priority) or because the condition goes away.

Quoted 2 lines

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Why ? Depends on your PIC actually but if it's a sane one, it won't "remember" a level interrupt that is gone. An edge interrupt is a different matter and is latched.

Some MPIC implementations tend to generate a spurrious IRQ in the case of level IRQs going away. IE. they still remember an event occurred and interrupt the processor, but on IACK return the spurious vector. However that isn't guaranteed to be the case and it is perfectly ok (and a good idea) for the PIC to just stop asserting the external interrupt line if the original device interrupt goes away (and it's configured as level sensitive).

You might notice that we try to re-hard-enable (while still soft disabled) as soon as we have gone get_irq in do_IRQ. This is because fetching the interrupt normally also masks it (either by masking at the source or by raisin the processor interrupt priority depending on the specific PIC) so we know we can re-hard enable.

Quoted 4 lines
Quoted 1 line

-> local_irq_enable() - This checks that irq_happened is set, and

replays

ret_from_except also check to replay.

ret_from_except checks because it essentially can act as an implicit local_irq_enable.

Quoted 6 lines
Quoted 4 lines

Now the case 2)
Case 2)
-> Local_irq_disable() will set soft_enabled = 0
-> Now DEC interrupt happens. We set PACA_IRQ_DEC in irq_happened,

But do not clear EE in SRR1 and rfi. So interrupts are not hard disabled.

Right. We move the decrementer to its maximum value however since on most implementations, it will continue interrupting the processor while it's top bit is set (and effectively act as a level sensitive interrupt).

Again, the goal here is to run as much as possible with interrupts hard enabled which allow better perf sampling.

Quoted 5 lines
Quoted 1 line

-> Now say EE interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled.

Quoted 2 lines

-> local_irq_enable() - This checks that irq_happened is set.
IIUC, it replays only one interrupt? is not it?

It replays one interrupt, but interrupt are still disabled during the replay. Upon return from the replayed interrupt, the ret_from_except will see the other one and replay it too.

Quoted 16 lines

After anyone is replayed in arch_local_irq_restore(), we will set soft/hard
interrupt there:

set_soft_enabled(1);
__hard_irq_enable();

Then any pending interrupt can be executed now.

Additionally, ret_from_except probably check to replay all.

Tiejun
--
To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#25 lore raw
Quoted 7 lines

Some MPIC implementations tend to generate a spurrious IRQ in the case of level IRQs going away. IE. they still remember an event occurred and interrupt the processor, but on IACK return the spurious vector. However that isn't guaranteed to be the case and it is perfectly ok (and a good idea) for the PIC to just stop asserting the external interrupt line if the original device interrupt goes away (and it's configured as level sensitive).

That will happen if the IRQ goes away while the cpu is performing the IACK sequence.
If the IRQ goes away while the cpu has interrupts masked then the cpu won't start the interrupt sequence and then try to read a vector when no interrupt is pending.

David

#26 lore raw
Quoted 5 lines · David Laight

That will happen if the IRQ goes away while the cpu is performing the IACK sequence.
If the IRQ goes away while the cpu has interrupts masked then the cpu won't start the interrupt sequence and then try to read a vector when no interrupt is pending.

Right, and get a spurrious vector which shouldn't be a big deal.

We tend to call that a "short interrupt". There have been many other cases of short interrupts in the past, for example on some MPICs, when distribution is enabled, they occasionally shoot an interrupt to more than one CPU at once :-)

Cheers,
Ben.

#27 lore raw
In reply to Benjamin Herrenschmidt
Quoted 10 lines

-----Original Message-----
From: Benjamin Herrenschmidt [mailto:[email protected]]
Sent: Thursday, May 09, 2013 8:38 PM
To: Chen, Tiejun
Cc: Bhushan Bharat-R65777; Caraman Mihai Claudiu-B02008; Wood Scott-B07421; [email protected]; [email protected]; [email protected]; [email protected]
Subject: Re: [RFC][KVM][PATCH 1/1] kvm:ppc:booke-64:
soft-disable interrupts

Quoted 7 lines
Quoted 2 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still

occur as I

Quoted 1 line

recall.

Only if directed to the hypervisor.

Yes, this is our current booehv design with EPCR[EXTGS] = 0.

Quoted 18 lines
Quoted 5 lines
Quoted 4 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts

are hard

Quoted 2 lines

disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting

interrupt

Quoted 1 line

till the interrupt condition on device is cleared, right?

The external interrupt line is level sensitive normally, so we have to mask MSR:EE in that case or the interrupt would keep re-occurring (note that FSL has this concept of auto-acked interrupts via the on die MPIC for which you can potentially use PACA_IRQ_EE_EDGE instead and avoid having to mask MSR:EE).

Quoted 5 lines
Quoted 2 lines

I don't understand "the interrupt condition on device is cleared" here.

Well, the external interrupt line will remain asserted until the device (via the PIC) stops interrupting the processor :-)

Yes, the device keeps on interrupting the interrupt until the device clear its interrupted condition.

Quoted 9 lines

Either because we mask at the PIC level (or raise the priority) or because the condition goes away.

Quoted 2 lines

I think regardless if you clear the device interrupt status, the system still receive a pending interrupt once EE or GS = 1.

Why ? Depends on your PIC actually but if it's a sane one, it won't "remember" a level interrupt that is gone. An edge interrupt is a different matter and is latched.

But the interrupt controller like MPIC should leave this irq as pending if we don't ACK/EOI this irq at all if we just clear the device, right?

Quoted 5 lines

Some MPIC implementations tend to generate a spurrious IRQ in the case of level IRQs going away. IE. they still remember an event occurred and interrupt the processor, but on IACK return the spurious vector. However that isn't guaranteed to

Yes, this is just what I mean. No matter if this vector is still valid, the external interrupt exception always occur once EE = 1 again.

Quoted 4 lines

be the case and it is perfectly ok (and a good idea) for the PIC to just stop asserting the external interrupt line if the original device interrupt goes away (and it's configured as level sensitive).

I don't remember MPIC can work with this way. So I'd like to look the manual again :)

Tiejun

#28 lore raw
In reply to Benjamin Herrenschmidt
Quoted 7 lines · Benjamin Herrenschmidt
Quoted 2 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur

as I

Quoted 1 line

recall.

Only if directed to the hypervisor.

This is always the case with KVM, right? At least on booke...

Quoted 6 lines
Quoted 6 lines
Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting

Quoted 8 lines

interrupt

Quoted 1 line

till the interrupt condition on device is cleared, right?

The external interrupt line is level sensitive normally, so we have to mask MSR:EE in that case or the interrupt would keep re-occurring (note that FSL has this concept of auto-acked interrupts via the on die MPIC for which you can potentially use PACA_IRQ_EE_EDGE instead and avoid having to mask MSR:EE).

Note that if we do this, we can no longer leave the interrupt vector in EPR, and would have to track (potentially multiple different) pending external interrupts in software.

-Scott

#29 lore raw
Quoted 10 lines · Scott Wood
Quoted 7 lines · Benjamin Herrenschmidt
Quoted 2 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur

as I

Quoted 1 line

recall.

Only if directed to the hypervisor.

This is always the case with KVM, right? At least on booke...

Hrm, on A2 we could choose iirc. Well not DEC but EXT at least, I don't remember about DBELL.

Quoted 16 lines
Quoted 7 lines
Quoted 6 lines
Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting

interrupt

Quoted 8 lines
Quoted 1 line

till the interrupt condition on device is cleared, right?

The external interrupt line is level sensitive normally, so we have to mask MSR:EE in that case or the interrupt would keep re-occurring (note
that FSL has this concept of auto-acked interrupts via the on die MPIC for which you can potentially use PACA_IRQ_EE_EDGE instead and avoid having to mask MSR:EE).

Quoted 3 lines

Note that if we do this, we can no longer leave the interrupt vector in EPR, and would have to track (potentially multiple different) pending external interrupts in software.

Right, you can have a little queue in the PACA and leave EE disabled if it's full. You can probably get away with a queue of 1 though :-)

Cheers,
Ben.

#30 lore raw
Quoted 7 lines · Benjamin Herrenschmidt
Quoted 4 lines · Scott Wood
Quoted 3 lines · Benjamin Herrenschmidt
Quoted 2 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still

occur

Quoted 1 line
Quoted 1 line

as I

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

recall.

Only if directed to the hypervisor.

This is always the case with KVM, right? At least on booke...

Hrm, on A2 we could choose iirc. Well not DEC but EXT at least, I don't remember about DBELL.

As for as the hardware goes we can choose on e500mc as well -- we just don't support it in KVM, because we can't distinguish between guest and host interrupts without routing the latter to critical, which is way too much of a pain for Linux (we did it in our standalone hypervisor, though).

-Scott

#31 lore raw
In reply to Benjamin Herrenschmidt
Quoted 31 lines · Benjamin Herrenschmidt
Quoted 10 lines · Scott Wood
Quoted 7 lines · Benjamin Herrenschmidt
Quoted 2 lines · tiejun.chen

Actually in the case GS=1 even if EE=0, EXT/DEC/DBELL still occur

as I

Quoted 1 line

recall.

Only if directed to the hypervisor.

This is always the case with KVM, right? At least on booke...

Hrm, on A2 we could choose iirc. Well not DEC but EXT at least, I don't remember about DBELL.

Quoted 16 lines
Quoted 15 lines
Quoted 6 lines
Quoted 3 lines

Case 1)
-> Local_irq_disable() will set soft_enabled = 0
-> Now Externel interrupt happens, there we set PACA_IRQ_EE in

irq_happened, Also clears EE in SRR1 and rfi. So interrupts are hard disabled. No more other interrupt gated by MSR.EE can happen. Looks like the idea here is to not let a device keep on inserting

interrupt

Quoted 1 line

till the interrupt condition on device is cleared, right?

The external interrupt line is level sensitive normally, so we have to mask MSR:EE in that case or the interrupt would keep re-occurring (note
that FSL has this concept of auto-acked interrupts via the on die MPIC for which you can potentially use PACA_IRQ_EE_EDGE instead and avoid having to mask MSR:EE).

Quoted 6 lines
Quoted 3 lines

Note that if we do this, we can no longer leave the interrupt vector in EPR, and would have to track (potentially multiple different) pending external interrupts in software.

Right, you can have a little queue in the PACA and leave EE disabled if it's full. You can probably get away with a queue of 1 though :-)

So I would assume you will not pick up these two patches, right?
http://patchwork.ozlabs.org/patch/235530/
http://patchwork.ozlabs.org/patch/235532/

Anyway it is more easier to enable the external proxy by using this method. But if you insist, I can respin a patch to use the method you suggested since it will definitely reduce the window where the irq is hard disabled.

Thanks,
Kevin

Quoted 8 lines

Cheers,
Ben.

--
To unsubscribe from this list: send the line "unsubscribe kvm" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#32 lore raw
Quoted 7 lines · Kevin Hao

So I would assume you will not pick up these two patches, right?
http://patchwork.ozlabs.org/patch/235530/
http://patchwork.ozlabs.org/patch/235532/

Anyway it is more easier to enable the external proxy by using this method. But if you insist, I can respin a patch to use the method you suggested since it will definitely reduce the window where the irq is hard disabled.

I would keep the EE_EDGE bit definition. I have no objection to a gradual approach however for the other one where we apply it as is now to enable coreint while you do a rework to make it better :-)

Cheers,
Ben.

#33 lore raw
Quoted 3 lines · Benjamin Herrenschmidt

I would keep the EE_EDGE bit definition. I have no objection to a gradual approach however for the other one where we apply it as is now to enable coreint while you do a rework to make it better :-)

Note also that I generally don't apply FSL related patches directly, I rely on Kumar Gala picking them up so he's the one ultimately making that choice.

Cheers,
Ben.