[PATCH 00/13] idle and soft-irq improvements and POWER9 idle optimisation

The first 9 patches are a bunch of independent small cleanups and improvements I've collected. The next 4 are improvements to POWER9 idle state entry and exit. They significantly change how we enter idle, now in virtual mode and without ptesync. EC=0 idle is significantly simplified from there,

[PATCH 05/13] powerpc/64s: irq replay merge HV and non-HV paths for doorbell replay

This results in smaller code, and fewer branches. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/entry_64.S | 6 +- arch/powerpc/kernel/exceptions-64s.S | 2 +- arch/powerpc/kernel/irq.c| 2 -- 3 files changed, 2 insertions(+), 8 deletions(-)

[PATCH 09/13] powerpc/64: runlatch CTRL[RUN] set optimisation

The CTRL register is read-only except bit 63 which is the run latch control. This means it can be updated with a mtspr rather than mfspr/mtspr. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/process.c | 35 +++ 1 file changed, 27

[PATCH 13/13] powerpc/64s: idle ESL=0 stop can avoid all save/restore overhead

When stop is executed with EC=ESL=0, it appears to execute like a normal instruction (resuming from NIP when woken by interrupt). So all the save/restore handling can be avoided completely. In particular NV GPRs do not have to be saved, and MSR does not have to be switched back to kernel MSR. So

[PATCH] powerpc: fix invalid use of register expressions

binutils >= 2.26 now warns about misuse of register expressions in assembler operands that are actually literals, for example: arch/powerpc/kernel/entry_64.S:535: Warning: invalid register expression Signed-off-by: Andreas Schwab --- arch/powerpc/include/asm/ppc_asm.h |

Re: [PATCH 12/13] powerpc/64s: idle POWER9 can execute stop in virtual mode

On Sun, 2017-08-06 at 03:02 +1000, Nicholas Piggin wrote: > The hardware can execute stop in any context, and KVM does not > require real mode. This saves a switch to real-mode when going > idle. +Paulus. > Signed-off-by: Nicholas Piggin > --- >

Re: [PATCH 08/13] powerpc/64s: irq replay remove spurious irq reason

On Sun, 2017-08-06 at 03:02 +1000, Nicholas Piggin wrote: > HVI interrupts have always used 0x500, so remove the dead branch. Maybe we should fix that and "catch" in incorrect entry via 0x500 which would mean the XIVE is trying to deliver guest irqs to the OS... That can happen if some LPCR bits

Re: [PATCH 08/13] powerpc/64s: irq replay remove spurious irq reason

On Sun, 06 Aug 2017 09:00:32 +1000 Benjamin Herrenschmidt wrote: > On Sun, 2017-08-06 at 03:02 +1000, Nicholas Piggin wrote: > > HVI interrupts have always used 0x500, so remove the dead branch. > > Maybe we should fix that and "catch" in incorrect entry via 0x500 > which

[PATCH 02/13] powerpc/64s: masked interrupt avoid instruction

EE is always enabled in SRR1 for masked interrupts, so clearing it can use xor. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/exceptions-64s.S | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/arch/powerpc/kernel/exceptions-64s.S

[PATCH 10/13] powerpc/64s: idle simplify KVM idle on POWER9

POWER9 CPUs have independent MMU contexts per thread so KVM does not have to bring sibling threads into real-mode when switching MMU mode to guest. This can simplify POWER9 sleep/wake paths and avoids hwsyncs. Signed-off-by: Nicholas Piggin ---

[PATCH 06/13] powerpc/64s: irq replay external use the HV handler in HV mode on POWER9

POWER9 host external interrupts use the h_virt_irq_common handler, so use that to replay them rather than using the hardware_interrupt_common handler. Both call do_IRQ, but using the correct handler reduces i-cache footprint. Signed-off-by: Nicholas Piggin ---

[PATCH 08/13] powerpc/64s: irq replay remove spurious irq reason

HVI interrupts have always used 0x500, so remove the dead branch. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/exceptions-64s.S | 2 -- 1 file changed, 2 deletions(-) diff --git a/arch/powerpc/kernel/exceptions-64s.S b/arch/powerpc/kernel/exceptions-64s.S index

[PATCH 11/13] powerpc/64s: idle POWER9 can execute stop without ptesync

Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/idle_book3s.S | 7 --- arch/powerpc/kvm/book3s_hv_rmhandlers.S | 24 2 files changed, 16 insertions(+), 15 deletions(-) diff --git a/arch/powerpc/kernel/idle_book3s.S

Re: [PATCH 13/13] powerpc/64s: idle ESL=0 stop can avoid all save/restore overhead

On Sun, 6 Aug 2017 03:02:41 +1000 Nicholas Piggin wrote: > When stop is executed with EC=ESL=0, it appears to execute like a > normal instruction (resuming from NIP when woken by interrupt). > So all the save/restore handling can be avoided completely. In > particular NV GPRs

[PATCH 03/13] powerpc/64s: masked interrupt returns to kernel so avoid r13 restore

Places in the kernel where r13 is not the PACA pointer must have maskable interrupts disabled, so r13 does not have to be restored when returning from a soft-masked interrupt. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/exceptions-64s.S | 2 +- 1 file changed, 1

[PATCH 04/13] powerpc/64: cleanup __check_irq_replay

Move the clearing of irq_happened bits into the condition where they were found to be set. This reduces instruction count slightly, and reduces stores into irq_happened. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/irq.c | 45

[PATCH 12/13] powerpc/64s: idle POWER9 can execute stop in virtual mode

The hardware can execute stop in any context, and KVM does not require real mode. This saves a switch to real-mode when going idle. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/idle_book3s.S | 9 + 1 file changed, 9 insertions(+) diff --git

[PATCH 01/13] powerpc/64s: masked interrupt avoid branch

Interrupts which do not require EE to be cleared can all be tested with a single bitwise test. Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/exceptions-64s.S | 6 ++ 1 file changed, 2 insertions(+), 4 deletions(-) diff --git

[PATCH 07/13] powerpc/64: remove redundant instruction in interrupt replay

Signed-off-by: Nicholas Piggin --- arch/powerpc/kernel/entry_64.S | 1 - 1 file changed, 1 deletion(-) diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S index ec67f67dafab..3f2666d24a7e 100644 --- a/arch/powerpc/kernel/entry_64.S +++

Re: [PATCH 2/3] Split VGA default device handler out of VGA arbiter

On Fri, Aug 04, 2017 at 08:20:32PM +1000, Daniel Axtens wrote: > A system without PCI legacy resources (e.g. ARM64, powerpc) may > find that no default/boot VGA device has been marked, because the > VGA arbiter checks for legacy resource decoding before marking a > card as default. Would it be

Re: 4.13-rc3: Unrecoverable exception 4100

On Fri, 04 Aug 2017 21:54:57 +0200 Andreas Schwab wrote: > No, this is really a 4.13-rc1 regression. > > Andreas. > SLB miss with MSR[RI]=0 on lbz r0,THREAD+THREAD_LOAD_FP(r7) Caused by bc4f65e4cf9d6cc43e0e9ba0b8648cf9201cd55f Hmm, I'll see if something can be