LinuxPPC-Dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
* Re: [PATCH] powerpc/64s: power4 nap fixup in C
From: Nicholas Piggin @ 2021-03-16  8:11 UTC (permalink / raw)
  To: Christophe Leroy, linuxppc-dev
In-Reply-To: <6ab2073b-dd10-ef2c-375d-1300f071ae1a@csgroup.eu>

Excerpts from Christophe Leroy's message of March 16, 2021 5:16 pm:
> 
> 
> Le 12/03/2021 à 02:20, Nicholas Piggin a écrit :
>> There is no need for this to be in asm, use the new intrrupt entry wrapper.
>> 
>> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
>> ---
>> Hopefully this works on a real G5 now, but I couldn't reproduce the
>> problem with QEMU.
>> 
>> Thanks,
>> Nick
>> 
>>   arch/powerpc/include/asm/interrupt.h   | 19 +++++++++++
>>   arch/powerpc/include/asm/processor.h   |  1 +
>>   arch/powerpc/include/asm/thread_info.h |  6 ++++
>>   arch/powerpc/kernel/exceptions-64s.S   | 45 --------------------------
>>   arch/powerpc/kernel/idle_book3s.S      |  4 +++
>>   5 files changed, 30 insertions(+), 45 deletions(-)
>> 
>> diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
>> index aedfba29e43a..ef015d3b5e39 100644
>> --- a/arch/powerpc/include/asm/interrupt.h
>> +++ b/arch/powerpc/include/asm/interrupt.h
>> @@ -9,6 +9,17 @@
>>   #include <asm/kprobes.h>
>>   #include <asm/runlatch.h>
>>   
>> +static inline void nap_adjust_return(struct pt_regs *regs)
>> +{
>> +#ifdef CONFIG_PPC_970_NAP
>> +	if (unlikely(test_thread_local_flags(_TLF_NAPPING))) {
>> +		/* Can avoid a test-and-clear because NMIs do not call this */
>> +		clear_thread_local_flags(_TLF_NAPPING);
>> +		regs->nip = (unsigned long)power4_idle_nap_return;
> 
> Why don't you do regs->nip = regs->link like PPC32 instead of going via an intermediate symbol that 
> does nothing else than branching to LR ?

It is supposed to keep the return branch predictor balanced.

I don't know if these CPUs have one, if it gets lost during nap, or if 
nap latency is so high it really doesn't matter. But I think it is good
practice to make a habit of keeping things balanced.

Thanks,
Nick

^ permalink raw reply

* Re: [PATCH 12/14] powerpc/64s: system call avoid setting MSR[RI] until we set MSR[EE]
From: Nicholas Piggin @ 2021-03-16  8:13 UTC (permalink / raw)
  To: Christophe Leroy, linuxppc-dev
In-Reply-To: <722aeb8d-507f-6702-dd79-26242f987e3e@csgroup.eu>

Excerpts from Christophe Leroy's message of March 16, 2021 5:21 pm:
> 
> 
> Le 15/03/2021 à 23:04, Nicholas Piggin a écrit :
>> This extends the MSR[RI]=0 window a little further into the system
>> call in order to pair RI and EE enabling with a single mtmsrd.
> 
> Time ago, I proposed to delay that on PPC32 and Michael objected, see 
> https://patchwork.ozlabs.org/project/linuxppc-dev/patch/9f9dd859d571e324c7412ed9db9da8cfba678257.1548956511.git.christophe.leroy@c-s.fr/

Yeah, it is a concern. The speedup should be at least 5% I think on
64s (have not measured in isolation yet), so might be worth it.

We'll see.

Thanks,
Nick

^ permalink raw reply

* Re: [PATCH 03/10] powerpc/64e/interrupt: use new interrupt return
From: Nicholas Piggin @ 2021-03-16  8:14 UTC (permalink / raw)
  To: Christophe Leroy, linuxppc-dev; +Cc: Scott Wood
In-Reply-To: <f452729c-6290-17e9-eb96-4c1ebffd4c7e@csgroup.eu>

Excerpts from Christophe Leroy's message of March 16, 2021 5:29 pm:
> 
> 
> Le 16/03/2021 à 08:25, Nicholas Piggin a écrit :
>> Excerpts from Nicholas Piggin's message of March 16, 2021 5:04 pm:
>>> Excerpts from Christophe Leroy's message of March 15, 2021 11:30 pm:
>>>>
>>>>
>>>> Le 15/03/2021 à 04:17, Nicholas Piggin a écrit :
>>>>> Update the new C and asm interrupt return code to account for 64e
>>>>> specifics, switch over to use it.
>>>>>
>>>>> The now-unused old ret_from_except code, that was moved to 64e after the
>>>>> 64s conversion, is removed.
>>>>>
>>>>> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
>>>>> ---
>>>>>    arch/powerpc/include/asm/asm-prototypes.h |   2 -
>>>>>    arch/powerpc/kernel/entry_64.S            |   9 +-
>>>>>    arch/powerpc/kernel/exceptions-64e.S      | 321 ++--------------------
>>>>>    arch/powerpc/kernel/interrupt.c           |  27 +-
>>>>>    arch/powerpc/kernel/irq.c                 |  76 -----
>>>>>    5 files changed, 56 insertions(+), 379 deletions(-)
>>>>>
>>>>> diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
>>>>> index fbabb49888d3..ae7b058b2970 100644
>>>>> --- a/arch/powerpc/kernel/interrupt.c
>>>>> +++ b/arch/powerpc/kernel/interrupt.c
>>>>> @@ -235,6 +235,10 @@ static notrace void booke_load_dbcr0(void)
>>>>>    #endif
>>>>>    }
>>>>>    
>>>>> +/* temporary hack for context tracking, removed in later patch */
>>>>> +#include <linux/sched/debug.h>
>>>>> +asmlinkage __visible void __sched schedule_user(void);
>>>>> +
>>>>>    /*
>>>>>     * This should be called after a syscall returns, with r3 the return value
>>>>>     * from the syscall. If this function returns non-zero, the system call
>>>>> @@ -292,7 +296,11 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
>>>>>    	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
>>>>>    		local_irq_enable();
>>>>>    		if (ti_flags & _TIF_NEED_RESCHED) {
>>>>> +#ifdef CONFIG_PPC_BOOK3E_64
>>>>> +			schedule_user();
>>>>> +#else
>>>>>    			schedule();
>>>>> +#endif
>>>>>    		} else {
>>>>>    			/*
>>>>>    			 * SIGPENDING must restore signal handler function
>>>>> @@ -360,7 +368,6 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
>>>>>    	return ret;
>>>>>    }
>>>>>    
>>>>> -#ifndef CONFIG_PPC_BOOK3E_64 /* BOOK3E not yet using this */
>>>>>    notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned long msr)
>>>>>    {
>>>>>    	unsigned long ti_flags;
>>>>> @@ -372,7 +379,9 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
>>>>>    	BUG_ON(!(regs->msr & MSR_PR));
>>>>>    	BUG_ON(!FULL_REGS(regs));
>>>>>    	BUG_ON(arch_irq_disabled_regs(regs));
>>>>> +#ifdef CONFIG_PPC_BOOK3S_64
>>>>
>>>> Shouldn't this go away in patch 6 as well ?
>>>> Or is that needed at all ? In syscall_exit_prepare() it is not ifdefed .
>>>
>>> Hmm, not sure. I'll take a look. It probably shouldn't be ifdefed at all
>>> but definitely by the end it should run without warning.
>> 
>> Oh I got confused and thought that was the syscall exit. Interrupt exit
>> has to keep this until patch 6 because 64e context tracking does
>> everything in interrupt wrappers, so by the time we get here it will
>> already be set to CONTEXT_USER.
>> 
> 
> ok, but that it has to go in patch 6 ? At the time being the #ifdef is still there at the end of the 
> series
> 

Yes, that was an oversight. I'll remove it.

Thanks,
Nick

^ permalink raw reply

* RE: [PATCH] powerpc/vdso32: Add missing _restgpr_31_x to fix build failure
From: David Laight @ 2021-03-16  9:35 UTC (permalink / raw)
  To: 'Segher Boessenkool'
  Cc: Paul Mackerras, linuxppc-dev@lists.ozlabs.org,
	linux-kernel@vger.kernel.org, 'Rasmus Villemoes'
In-Reply-To: <20210315235947.GD16691@gate.crashing.org>

From: Segher Boessenkool
> Sent: 16 March 2021 00:00
...
> > Although you may need to disable loop unrolling (often dubious at best)
> > and either force or disable some function inlining.
> 
> The cases where GCC does loop unrolling at -O2 always help quite a lot.
> Or, do you have a counter-example?  We'd love to see one.

The real problem with loop unrolling is that quite often a modern
out-of-order superscaler processor actually has 'spare' execution
cycles where the loop control can be done 'for free'.
Sometimes you do need to unroll (or interleave) a couple of
times to get enough spare execution cycles.

But the unrolled loop has to read a lot more code into cache
- so unless the code is 'hot cache' (that is usually arranged
for benchmarking) those delays apply as well.
The larger code footprint also displaces other code.

My real annoyance with gcc is unrolling (and vectorizing)
loops that I know are never executed as many times as even one
copy of the unrolled loop.

As an example intel (ivy bridge onwards) cpu execute the
following code (the middle of the ip checksum) at 8 bytes/clock.
(Limited by the carry flag.)
It just doesn't need any further unrolling.

+               "10:    jecxz 20f\n"
+               "       adc   (%[buff], %[len]), %[sum_0]\n"
+               "       adc   8(%[buff], %[len]), %[sum_1]\n"
+               "       lea   32(%[len]), %[len_tmp]\n"
+               "       adc   16(%[buff], %[len]), %[sum_0]\n"
+               "       adc   24(%[buff], %[len]), %[sum_1]\n"
+               "       mov   %[len_tmp], %[len]\n"
+               "       jmp   10b\n"

Annoyingly that loop is slow on my 8-core atom. 
The existing code only does 4 bytes/clock on intel cpu prior
to either broadwell or haswell (forgotten which) in spite
of much more unroling.


> And yup, inlining is hard.  GCC's heuristics there are very good
> nowadays, but any single decision has big effects.  Doing the important
> spots manually (always_inline or noinline) has good payoff.

Latest inline gripe was a function replicated about 20 times
when the non-inline version was a register load and 'tail call'.
The inlining is just bloat.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)


^ permalink raw reply

* [PATCH] ASoC: fsl_sai: remove reset code from dai_probe
From: Shengjiu Wang @ 2021-03-16  9:27 UTC (permalink / raw)
  To: timur, nicoleotsuka, Xiubo.Lee, festevam, lgirdwood, broonie,
	perex, tiwai, alsa-devel, linuxppc-dev, linux-kernel

From: Viorel Suman <viorel.suman@nxp.com>

SAI software reset is done in runtime resume,
there is no need to do it in fsl_sai_dai_probe.

Signed-off-by: Viorel Suman <viorel.suman@nxp.com>
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
---
 sound/soc/fsl/fsl_sai.c | 7 -------
 1 file changed, 7 deletions(-)

diff --git a/sound/soc/fsl/fsl_sai.c b/sound/soc/fsl/fsl_sai.c
index 6ef2ce348232..9e7893f91882 100644
--- a/sound/soc/fsl/fsl_sai.c
+++ b/sound/soc/fsl/fsl_sai.c
@@ -711,13 +711,6 @@ static int fsl_sai_dai_probe(struct snd_soc_dai *cpu_dai)
 	struct fsl_sai *sai = dev_get_drvdata(cpu_dai->dev);
 	unsigned int ofs = sai->soc_data->reg_offset;
 
-	/* Software Reset for both Tx and Rx */
-	regmap_write(sai->regmap, FSL_SAI_TCSR(ofs), FSL_SAI_CSR_SR);
-	regmap_write(sai->regmap, FSL_SAI_RCSR(ofs), FSL_SAI_CSR_SR);
-	/* Clear SR bit to finish the reset */
-	regmap_write(sai->regmap, FSL_SAI_TCSR(ofs), 0);
-	regmap_write(sai->regmap, FSL_SAI_RCSR(ofs), 0);
-
 	regmap_update_bits(sai->regmap, FSL_SAI_TCR1(ofs),
 			   FSL_SAI_CR1_RFW_MASK(sai->soc_data->fifo_depth),
 			   sai->soc_data->fifo_depth - FSL_SAI_MAXBURST_TX);
-- 
2.27.0


^ permalink raw reply related

* [PATCH] ASoC: fsl_spdif: use snd_ctl_boolean_mono_info
From: Shengjiu Wang @ 2021-03-16  9:42 UTC (permalink / raw)
  To: timur, nicoleotsuka, Xiubo.Lee, festevam, lgirdwood, broonie,
	perex, tiwai, alsa-devel, linuxppc-dev, linux-kernel

From: Viorel Suman <viorel.suman@nxp.com>

Remove redundant code and use snd_ctl_boolean_mono_info
instead.

Signed-off-by: Viorel Suman <viorel.suman@nxp.com>
Signed-off-by: Shengjiu Wang <shengjiu.wang@nxp.com>
---
 sound/soc/fsl/fsl_spdif.c | 30 +++---------------------------
 1 file changed, 3 insertions(+), 27 deletions(-)

diff --git a/sound/soc/fsl/fsl_spdif.c b/sound/soc/fsl/fsl_spdif.c
index 174e558224d8..1cd3441d1c03 100644
--- a/sound/soc/fsl/fsl_spdif.c
+++ b/sound/soc/fsl/fsl_spdif.c
@@ -801,18 +801,6 @@ static int fsl_spdif_qget(struct snd_kcontrol *kcontrol,
 	return ret;
 }
 
-/* Valid bit information */
-static int fsl_spdif_vbit_info(struct snd_kcontrol *kcontrol,
-				struct snd_ctl_elem_info *uinfo)
-{
-	uinfo->type = SNDRV_CTL_ELEM_TYPE_BOOLEAN;
-	uinfo->count = 1;
-	uinfo->value.integer.min = 0;
-	uinfo->value.integer.max = 1;
-
-	return 0;
-}
-
 /* Get valid good bit from interrupt status register */
 static int fsl_spdif_rx_vbit_get(struct snd_kcontrol *kcontrol,
 				 struct snd_ctl_elem_value *ucontrol)
@@ -925,18 +913,6 @@ static int fsl_spdif_rxrate_get(struct snd_kcontrol *kcontrol,
 	return 0;
 }
 
-/* User bit sync mode info */
-static int fsl_spdif_usync_info(struct snd_kcontrol *kcontrol,
-				struct snd_ctl_elem_info *uinfo)
-{
-	uinfo->type = SNDRV_CTL_ELEM_TYPE_BOOLEAN;
-	uinfo->count = 1;
-	uinfo->value.integer.min = 0;
-	uinfo->value.integer.max = 1;
-
-	return 0;
-}
-
 /*
  * User bit sync mode:
  * 1 CD User channel subcode
@@ -1018,7 +994,7 @@ static struct snd_kcontrol_new fsl_spdif_ctrls[] = {
 		.name = "IEC958 RX V-Bit Errors",
 		.access = SNDRV_CTL_ELEM_ACCESS_READ |
 			SNDRV_CTL_ELEM_ACCESS_VOLATILE,
-		.info = fsl_spdif_vbit_info,
+		.info = snd_ctl_boolean_mono_info,
 		.get = fsl_spdif_rx_vbit_get,
 	},
 	{
@@ -1027,7 +1003,7 @@ static struct snd_kcontrol_new fsl_spdif_ctrls[] = {
 		.access = SNDRV_CTL_ELEM_ACCESS_READ |
 			SNDRV_CTL_ELEM_ACCESS_WRITE |
 			SNDRV_CTL_ELEM_ACCESS_VOLATILE,
-		.info = fsl_spdif_vbit_info,
+		.info = snd_ctl_boolean_mono_info,
 		.get = fsl_spdif_tx_vbit_get,
 		.put = fsl_spdif_tx_vbit_put,
 	},
@@ -1047,7 +1023,7 @@ static struct snd_kcontrol_new fsl_spdif_ctrls[] = {
 		.access = SNDRV_CTL_ELEM_ACCESS_READ |
 			SNDRV_CTL_ELEM_ACCESS_WRITE |
 			SNDRV_CTL_ELEM_ACCESS_VOLATILE,
-		.info = fsl_spdif_usync_info,
+		.info = snd_ctl_boolean_mono_info,
 		.get = fsl_spdif_usync_get,
 		.put = fsl_spdif_usync_put,
 	},
-- 
2.27.0


^ permalink raw reply related

* [PATCH v2 00/11] Move 64e to new interrupt return code
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin

Since RFC:
- Rebased on Christophe's v3 ppc32 conversion
- Fixed up small details, adds some powerpc-wide cleanups at the end.

Since v1:
- Christophe's review comments accounted for:
- Split common code interrupt changes for 64e from 64e conversion.
- Removed unnecessary ifdef additions from common interrupt code.
- Keep interrupt return context tracking debug check for 64e when
  it switches to new context tracing.
- __bad_page_fault is made static in page fault cleanup.
- CPU time accounting asm and asm offsets can be removed with 64e
  conversion.

Tested on qemu only (QEMU e500), which is not ideal for interrupt
handling particularly the critical interrupts which I don't know
whether it can generate.

Thanks,
Nick

Nicholas Piggin (11):
  powerpc/syscall: switch user_exit_irqoff and trace_hardirqs_off order
  powerpc/64e/interrupt: always save nvgprs on interrupt
  powerpc/interrupt: update common interrupt code for 64e
  powerpc/64e/interrupt: use new interrupt return
  powerpc/64e/interrupt: NMI save irq soft-mask state in C
  powerpc/64e/interrupt: reconcile irq soft-mask state in C
  powerpc/64e/interrupt: Use new interrupt context tracking scheme
  powerpc/64e/interrupt: handle bad_page_fault in C
  powerpc: clean up do_page_fault
  powerpc: remove partial register save logic
  powerpc: move norestart trap flag to bit 0

 arch/powerpc/include/asm/asm-prototypes.h |   2 -
 arch/powerpc/include/asm/bug.h            |   5 +-
 arch/powerpc/include/asm/interrupt.h      |  66 ++--
 arch/powerpc/include/asm/ppc_asm.h        |  20 -
 arch/powerpc/include/asm/ptrace.h         |  39 +-
 arch/powerpc/kernel/align.c               |   6 -
 arch/powerpc/kernel/asm-offsets.c         |  10 -
 arch/powerpc/kernel/entry_64.S            |  40 +-
 arch/powerpc/kernel/exceptions-64e.S      | 425 ++--------------------
 arch/powerpc/kernel/interrupt.c           |  26 +-
 arch/powerpc/kernel/irq.c                 |  76 ----
 arch/powerpc/kernel/process.c             |  12 -
 arch/powerpc/kernel/ptrace/ptrace-view.c  |  21 --
 arch/powerpc/kernel/ptrace/ptrace.c       |   2 -
 arch/powerpc/kernel/ptrace/ptrace32.c     |   4 -
 arch/powerpc/kernel/signal_32.c           |   3 -
 arch/powerpc/kernel/signal_64.c           |   2 -
 arch/powerpc/kernel/traps.c               |  14 +-
 arch/powerpc/lib/sstep.c                  |   4 -
 arch/powerpc/mm/book3s64/hash_utils.c     |  16 +-
 arch/powerpc/mm/fault.c                   |  30 +-
 arch/powerpc/xmon/xmon.c                  |  23 +-
 22 files changed, 126 insertions(+), 720 deletions(-)

-- 
2.23.0


^ permalink raw reply

* [PATCH v2 01/11] powerpc/syscall: switch user_exit_irqoff and trace_hardirqs_off order
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

user_exit_irqoff() -> __context_tracking_exit -> vtime_user_exit
warns in __seqprop_assert due to lockdep thinking preemption is enabled
because trace_hardirqs_off() has not yet been called.

Switch the order of these two calls, which matches their ordering in
interrupt_enter_prepare.

Fixes: 5f0b6ac3905f ("powerpc/64/syscall: Reconcile interrupts")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/kernel/interrupt.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
index c4dd4b8f9cfa..fbabb49888d3 100644
--- a/arch/powerpc/kernel/interrupt.c
+++ b/arch/powerpc/kernel/interrupt.c
@@ -43,11 +43,11 @@ notrace long system_call_exception(long r3, long r4, long r5,
 	if (IS_ENABLED(CONFIG_PPC_IRQ_SOFT_MASK_DEBUG))
 		BUG_ON(irq_soft_mask_return() != IRQS_ALL_DISABLED);
 
+	trace_hardirqs_off(); /* finish reconciling */
+
 	CT_WARN_ON(ct_state() == CONTEXT_KERNEL);
 	user_exit_irqoff();
 
-	trace_hardirqs_off(); /* finish reconciling */
-
 	if (!IS_ENABLED(CONFIG_BOOKE) && !IS_ENABLED(CONFIG_40x))
 		BUG_ON(!(regs->msr & MSR_RI));
 	BUG_ON(!(regs->msr & MSR_PR));
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 02/11] powerpc/64e/interrupt: always save nvgprs on interrupt
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

In order to use the C interrupt return, nvgprs must always be saved.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/ptrace.h    |  9 +--------
 arch/powerpc/kernel/entry_64.S       | 13 -------------
 arch/powerpc/kernel/exceptions-64e.S | 27 +++------------------------
 3 files changed, 4 insertions(+), 45 deletions(-)

diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index 1aca5fe79285..c5b3669918f4 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -186,18 +186,11 @@ static inline void regs_set_return_value(struct pt_regs *regs, unsigned long rc)
 	((struct pt_regs *)((unsigned long)task_stack_page(current) + THREAD_SIZE) - 1)
 
 #ifdef __powerpc64__
-#ifdef CONFIG_PPC_BOOK3S
 #define TRAP_FLAGS_MASK		0x10
 #define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
 #define FULL_REGS(regs)		true
 #define SET_FULL_REGS(regs)	do { } while (0)
-#else
-#define TRAP_FLAGS_MASK		0x11
-#define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
-#define FULL_REGS(regs)		(((regs)->trap & 1) == 0)
-#define SET_FULL_REGS(regs)	((regs)->trap &= ~1)
-#endif
-#define CHECK_FULL_REGS(regs)	BUG_ON(!FULL_REGS(regs))
+#define CHECK_FULL_REGS(regs)	do { } while (0)
 #define NV_REG_POISON		0xdeadbeefdeadbeefUL
 #else
 /*
diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
index 6c4d9e276c4d..853534b2ae2e 100644
--- a/arch/powerpc/kernel/entry_64.S
+++ b/arch/powerpc/kernel/entry_64.S
@@ -417,19 +417,6 @@ _GLOBAL(ret_from_kernel_thread)
 	li	r3,0
 	b	.Lsyscall_exit
 
-#ifdef CONFIG_PPC_BOOK3E
-/* Save non-volatile GPRs, if not already saved. */
-_GLOBAL(save_nvgprs)
-	ld	r11,_TRAP(r1)
-	andi.	r0,r11,1
-	beqlr-
-	SAVE_NVGPRS(r1)
-	clrrdi	r0,r11,1
-	std	r0,_TRAP(r1)
-	blr
-_ASM_NOKPROBE_SYMBOL(save_nvgprs);
-#endif
-
 #ifdef CONFIG_PPC_BOOK3S_64
 
 #define FLUSH_COUNT_CACHE	\
diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index e8eb9992a270..a7d9ce9f7fdb 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -417,14 +417,15 @@ exc_##n##_common:							    \
 	std	r6,_LINK(r1);						    \
 	std	r7,_CTR(r1);						    \
 	std	r8,_XER(r1);						    \
-	li	r3,(n)+1;		/* indicate partial regs in trap */ \
+	li	r3,(n);			/* regs.trap vector */		    \
 	std	r9,0(r1);		/* store stack frame back link */   \
 	std	r10,_CCR(r1);		/* store orig CR in stackframe */   \
 	std	r9,GPR1(r1);		/* store stack frame back link */   \
 	std	r11,SOFTE(r1);		/* and save it to stackframe */     \
 	std	r12,STACK_FRAME_OVERHEAD-16(r1); /* mark the frame */	    \
 	std	r3,_TRAP(r1);		/* set trap number		*/  \
-	std	r0,RESULT(r1);		/* clear regs->result */
+	std	r0,RESULT(r1);		/* clear regs->result */	    \
+	SAVE_NVGPRS(r1);
 
 #define EXCEPTION_COMMON(n) \
 	EXCEPTION_COMMON_LVL(n, SPRN_SPRG_GEN_SCRATCH, PACA_EXGEN)
@@ -561,7 +562,6 @@ __end_interrupts:
 	CRIT_EXCEPTION_PROLOG(0x100, BOOKE_INTERRUPT_CRITICAL,
 			      PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON_CRIT(0x100)
-	bl	save_nvgprs
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
@@ -573,7 +573,6 @@ __end_interrupts:
 	MC_EXCEPTION_PROLOG(0x000, BOOKE_INTERRUPT_MACHINE_CHECK,
 			    PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON_MC(0x000)
-	bl	save_nvgprs
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
@@ -623,7 +622,6 @@ __end_interrupts:
 	std	r14,_DSISR(r1)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXGEN+EX_R14(r13)
-	bl	save_nvgprs
 	bl	program_check_exception
 	b	ret_from_except
 
@@ -639,7 +637,6 @@ __end_interrupts:
 	bl	load_up_fpu
 	b	fast_exception_return
 1:	INTS_DISABLE
-	bl	save_nvgprs
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	kernel_fp_unavailable_exception
 	b	ret_from_except
@@ -661,7 +658,6 @@ BEGIN_FTR_SECTION
 END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 #endif
 	INTS_DISABLE
-	bl	save_nvgprs
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	altivec_unavailable_exception
 	b	ret_from_except
@@ -673,7 +669,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 				PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x220)
 	INTS_DISABLE
-	bl	save_nvgprs
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 #ifdef CONFIG_ALTIVEC
 BEGIN_FTR_SECTION
@@ -698,7 +693,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	CRIT_EXCEPTION_PROLOG(0x9f0, BOOKE_INTERRUPT_WATCHDOG,
 			      PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON_CRIT(0x9f0)
-	bl	save_nvgprs
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
@@ -723,7 +717,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 				PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0xf20)
 	INTS_DISABLE
-	bl	save_nvgprs
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	unknown_exception
 	b	ret_from_except
@@ -792,7 +785,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXCRIT+EX_R14(r13)
 	ld	r15,PACA_EXCRIT+EX_R15(r13)
-	bl	save_nvgprs
 	bl	DebugException
 	b	ret_from_except
 
@@ -864,7 +856,6 @@ kernel_dbg_exc:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXDBG+EX_R14(r13)
 	ld	r15,PACA_EXDBG+EX_R15(r13)
-	bl	save_nvgprs
 	bl	DebugException
 	b	ret_from_except
 
@@ -887,7 +878,6 @@ kernel_dbg_exc:
 	CRIT_EXCEPTION_PROLOG(0x2a0, BOOKE_INTERRUPT_DOORBELL_CRITICAL,
 			      PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON_CRIT(0x2a0)
-	bl	save_nvgprs
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
@@ -903,7 +893,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x2c0)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	save_nvgprs
 	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	ret_from_except
@@ -913,7 +902,6 @@ kernel_dbg_exc:
 	CRIT_EXCEPTION_PROLOG(0x2e0, BOOKE_INTERRUPT_GUEST_DBELL_CRIT,
 			      PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON_CRIT(0x2e0)
-	bl	save_nvgprs
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
@@ -926,7 +914,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x310)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	save_nvgprs
 	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	ret_from_except
@@ -937,7 +924,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x320)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	save_nvgprs
 	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	ret_from_except
@@ -948,7 +934,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x340)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	save_nvgprs
 	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	ret_from_except
@@ -1014,7 +999,6 @@ storage_fault_common:
 	cmpdi	r3,0
 	bne-	1f
 	b	ret_from_except_lite
-1:	bl	save_nvgprs
 	mr	r4,r3
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	__bad_page_fault
@@ -1030,16 +1014,12 @@ alignment_more:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXGEN+EX_R14(r13)
 	ld	r15,PACA_EXGEN+EX_R15(r13)
-	bl	save_nvgprs
 	INTS_RESTORE_HARD
 	bl	alignment_exception
 	b	ret_from_except
 
 	.align	7
 _GLOBAL(ret_from_except)
-	ld	r11,_TRAP(r1)
-	andi.	r0,r11,1
-	bne	ret_from_except_lite
 	REST_NVGPRS(r1)
 
 _GLOBAL(ret_from_except_lite)
@@ -1080,7 +1060,6 @@ _GLOBAL(ret_from_except_lite)
 	SCHEDULE_USER
 	b	ret_from_except_lite
 2:
-	bl	save_nvgprs
 	/*
 	 * Use a non volatile GPR to save and restore our thread_info flags
 	 * across the call to restore_interrupts.
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 03/11] powerpc/interrupt: update common interrupt code for
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

This makes adjustments to 64-bit asm and common C interrupt return
code to be usable by the 64e subarchitecture.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/kernel/entry_64.S  |  9 +++++++--
 arch/powerpc/kernel/interrupt.c | 35 ++++++++++++++++++++-------------
 2 files changed, 28 insertions(+), 16 deletions(-)

diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
index 853534b2ae2e..555b3d0a3f38 100644
--- a/arch/powerpc/kernel/entry_64.S
+++ b/arch/powerpc/kernel/entry_64.S
@@ -632,7 +632,6 @@ END_FTR_SECTION_IFCLR(CPU_FTR_ARCH_207S)
 	addi	r1,r1,SWITCH_FRAME_SIZE
 	blr
 
-#ifdef CONFIG_PPC_BOOK3S
 	/*
 	 * If MSR EE/RI was never enabled, IRQs not reconciled, NVGPRs not
 	 * touched, no exit work created, then this can be used.
@@ -644,6 +643,7 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
 	kuap_check_amr r3, r4
 	ld	r5,_MSR(r1)
 	andi.	r0,r5,MSR_PR
+#ifdef CONFIG_PPC_BOOK3S
 	bne	.Lfast_user_interrupt_return_amr
 	kuap_kernel_restore r3, r4
 	andi.	r0,r5,MSR_RI
@@ -652,6 +652,10 @@ _ASM_NOKPROBE_SYMBOL(fast_interrupt_return)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	unrecoverable_exception
 	b	. /* should not get here */
+#else
+	bne	.Lfast_user_interrupt_return
+	b	.Lfast_kernel_interrupt_return
+#endif
 
 	.balign IFETCH_ALIGN_BYTES
 	.globl interrupt_return
@@ -665,8 +669,10 @@ _ASM_NOKPROBE_SYMBOL(interrupt_return)
 	cmpdi	r3,0
 	bne-	.Lrestore_nvgprs
 
+#ifdef CONFIG_PPC_BOOK3S
 .Lfast_user_interrupt_return_amr:
 	kuap_user_restore r3, r4
+#endif
 .Lfast_user_interrupt_return:
 	ld	r11,_NIP(r1)
 	ld	r12,_MSR(r1)
@@ -775,7 +781,6 @@ ALT_FTR_SECTION_END_IFCLR(CPU_FTR_STCX_CHECKS_ADDRESS)
 
 	RFI_TO_KERNEL
 	b	.	/* prevent speculative execution */
-#endif /* CONFIG_PPC_BOOK3S */
 
 #ifdef CONFIG_PPC_RTAS
 /*
diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
index fbabb49888d3..381a618b5b5b 100644
--- a/arch/powerpc/kernel/interrupt.c
+++ b/arch/powerpc/kernel/interrupt.c
@@ -235,6 +235,10 @@ static notrace void booke_load_dbcr0(void)
 #endif
 }
 
+/* temporary hack for context tracking, removed in later patch */
+#include <linux/sched/debug.h>
+asmlinkage __visible void __sched schedule_user(void);
+
 /*
  * This should be called after a syscall returns, with r3 the return value
  * from the syscall. If this function returns non-zero, the system call
@@ -292,7 +296,11 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
 	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
 		local_irq_enable();
 		if (ti_flags & _TIF_NEED_RESCHED) {
+#ifdef CONFIG_PPC_BOOK3E_64
+			schedule_user();
+#else
 			schedule();
+#endif
 		} else {
 			/*
 			 * SIGPENDING must restore signal handler function
@@ -349,18 +357,13 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
 
 	account_cpu_user_exit();
 
-#ifndef CONFIG_PPC_BOOK3E_64 /* BOOK3E not using this */
-	/*
-	 * We do this at the end so that we do context switch with KERNEL AMR
-	 */
+	/* Restore user access locks last */
 	kuap_user_restore(regs);
-#endif
 	kuep_unlock();
 
 	return ret;
 }
 
-#ifndef CONFIG_PPC_BOOK3E_64 /* BOOK3E not yet using this */
 notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned long msr)
 {
 	unsigned long ti_flags;
@@ -372,7 +375,9 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 	BUG_ON(!(regs->msr & MSR_PR));
 	BUG_ON(!FULL_REGS(regs));
 	BUG_ON(arch_irq_disabled_regs(regs));
+#ifdef CONFIG_PPC_BOOK3S_64
 	CT_WARN_ON(ct_state() == CONTEXT_USER);
+#endif
 
 	/*
 	 * We don't need to restore AMR on the way back to userspace for KUAP.
@@ -387,7 +392,11 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
 		local_irq_enable(); /* returning to user: may enable */
 		if (ti_flags & _TIF_NEED_RESCHED) {
+#ifdef CONFIG_PPC_BOOK3E_64
+			schedule_user();
+#else
 			schedule();
+#endif
 		} else {
 			if (ti_flags & _TIF_SIGPENDING)
 				ret |= _TIF_RESTOREALL;
@@ -432,10 +441,9 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 
 	account_cpu_user_exit();
 
-	/*
-	 * We do this at the end so that we do context switch with KERNEL AMR
-	 */
+	/* Restore user access locks last */
 	kuap_user_restore(regs);
+
 	return ret;
 }
 
@@ -456,7 +464,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 	 * CT_WARN_ON comes here via program_check_exception,
 	 * so avoid recursion.
 	 */
-	if (TRAP(regs) != 0x700)
+	if (IS_ENABLED(CONFIG_BOOKS) && TRAP(regs) != 0x700)
 		CT_WARN_ON(ct_state() == CONTEXT_USER);
 
 	kuap = kuap_get_and_assert_locked();
@@ -497,12 +505,11 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 #endif
 
 	/*
-	 * Don't want to mfspr(SPRN_AMR) here, because this comes after mtmsr,
-	 * which would cause Read-After-Write stalls. Hence, we take the AMR
-	 * value from the check above.
+	 * 64s does not want to mfspr(SPRN_AMR) here, because this comes after
+	 * mtmsr, which would cause Read-After-Write stalls. Hence, take the
+	 * AMR value from the check above.
 	 */
 	kuap_kernel_restore(regs, kuap);
 
 	return ret;
 }
-#endif
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 04/11] powerpc/64e/interrupt: use new interrupt return
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

Update the new C and asm interrupt return code to account for 64e
specifics, switch over to use it.

The now-unused old ret_from_except code, that was moved to 64e after the
64s conversion, is removed.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/asm-prototypes.h |   2 -
 arch/powerpc/include/asm/ppc_asm.h        |  20 --
 arch/powerpc/kernel/asm-offsets.c         |  10 -
 arch/powerpc/kernel/exceptions-64e.S      | 321 ++--------------------
 arch/powerpc/kernel/irq.c                 |  76 -----
 5 files changed, 25 insertions(+), 404 deletions(-)

diff --git a/arch/powerpc/include/asm/asm-prototypes.h b/arch/powerpc/include/asm/asm-prototypes.h
index 939f3c94c8f3..1c7b75834e04 100644
--- a/arch/powerpc/include/asm/asm-prototypes.h
+++ b/arch/powerpc/include/asm/asm-prototypes.h
@@ -77,8 +77,6 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 long ppc_fadvise64_64(int fd, int advice, u32 offset_high, u32 offset_low,
 		      u32 len_high, u32 len_low);
 long sys_switch_endian(void);
-notrace unsigned int __check_irq_replay(void);
-void notrace restore_interrupts(void);
 
 /* prom_init (OpenFirmware) */
 unsigned long __init prom_init(unsigned long r3, unsigned long r4,
diff --git a/arch/powerpc/include/asm/ppc_asm.h b/arch/powerpc/include/asm/ppc_asm.h
index 8998122fc7e2..d6739d700f0a 100644
--- a/arch/powerpc/include/asm/ppc_asm.h
+++ b/arch/powerpc/include/asm/ppc_asm.h
@@ -15,26 +15,6 @@
 
 #define SZL			(BITS_PER_LONG/8)
 
-/*
- * Stuff for accurate CPU time accounting.
- * These macros handle transitions between user and system state
- * in exception entry and exit and accumulate time to the
- * user_time and system_time fields in the paca.
- */
-
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-#define ACCOUNT_CPU_USER_EXIT(ptr, ra, rb)
-#else
-#define ACCOUNT_CPU_USER_EXIT(ptr, ra, rb)				\
-	MFTB(ra);			/* get timebase */		\
-	PPC_LL	rb, ACCOUNT_STARTTIME(ptr);				\
-	PPC_STL	ra, ACCOUNT_STARTTIME_USER(ptr);			\
-	subf	rb,rb,ra;		/* subtract start value */	\
-	PPC_LL	ra, ACCOUNT_SYSTEM_TIME(ptr);				\
-	add	ra,ra,rb;		/* add on to system time */	\
-	PPC_STL	ra, ACCOUNT_SYSTEM_TIME(ptr)
-#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
-
 /*
  * Macros for storing registers into and loading registers from
  * exception frames.
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 85ba2b0bc8d8..c880ad18b851 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -282,21 +282,11 @@ int main(void)
 	OFFSET(PACAHWCPUID, paca_struct, hw_cpu_id);
 	OFFSET(PACAKEXECSTATE, paca_struct, kexec_state);
 	OFFSET(PACA_DSCR_DEFAULT, paca_struct, dscr_default);
-	OFFSET(ACCOUNT_STARTTIME, paca_struct, accounting.starttime);
-	OFFSET(ACCOUNT_STARTTIME_USER, paca_struct, accounting.starttime_user);
-	OFFSET(ACCOUNT_USER_TIME, paca_struct, accounting.utime);
-	OFFSET(ACCOUNT_SYSTEM_TIME, paca_struct, accounting.stime);
 #ifdef CONFIG_PPC_BOOK3E
 	OFFSET(PACA_TRAP_SAVE, paca_struct, trap_save);
 #endif
 	OFFSET(PACA_SPRG_VDSO, paca_struct, sprg_vdso);
 #else /* CONFIG_PPC64 */
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-	OFFSET(ACCOUNT_STARTTIME, thread_info, accounting.starttime);
-	OFFSET(ACCOUNT_STARTTIME_USER, thread_info, accounting.starttime_user);
-	OFFSET(ACCOUNT_USER_TIME, thread_info, accounting.utime);
-	OFFSET(ACCOUNT_SYSTEM_TIME, thread_info, accounting.stime);
-#endif
 #endif /* CONFIG_PPC64 */
 
 	/* RTAS */
diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index a7d9ce9f7fdb..7a2e7be33092 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -139,7 +139,8 @@ ret_from_level_except:
 	ld	r3,_MSR(r1)
 	andi.	r3,r3,MSR_PR
 	beq	1f
-	b	ret_from_except
+	REST_NVGPRS(r1)
+	b	interrupt_return
 1:
 
 	LOAD_REG_ADDR(r11,extlb_level_exc)
@@ -208,7 +209,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_EMB_HV)
 	/*
 	 * Restore PACAIRQHAPPENED rather than setting it based on
 	 * the return MSR[EE], since we could have interrupted
-	 * __check_irq_replay() or other inconsistent transitory
+	 * interrupt replay or other inconsistent transitory
 	 * states that must remain that way.
 	 */
 	SPECIAL_EXC_LOAD(r10,IRQHAPPENED)
@@ -511,7 +512,7 @@ exc_##n##_bad_stack:							    \
 	CHECK_NAPPING();						\
 	addi	r3,r1,STACK_FRAME_OVERHEAD;				\
 	bl	hdlr;							\
-	b	ret_from_except_lite;
+	b	interrupt_return
 
 /* This value is used to mark exception frames on the stack. */
 	.section	".toc","aw"
@@ -623,7 +624,8 @@ __end_interrupts:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXGEN+EX_R14(r13)
 	bl	program_check_exception
-	b	ret_from_except
+	REST_NVGPRS(r1)
+	b	interrupt_return
 
 /* Floating Point Unavailable Interrupt */
 	START_EXCEPTION(fp_unavailable);
@@ -635,11 +637,11 @@ __end_interrupts:
 	andi.	r0,r12,MSR_PR;
 	beq-	1f
 	bl	load_up_fpu
-	b	fast_exception_return
+	b	fast_interrupt_return
 1:	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	kernel_fp_unavailable_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* Altivec Unavailable Interrupt */
 	START_EXCEPTION(altivec_unavailable);
@@ -653,14 +655,14 @@ BEGIN_FTR_SECTION
 	andi.	r0,r12,MSR_PR;
 	beq-	1f
 	bl	load_up_altivec
-	b	fast_exception_return
+	b	fast_interrupt_return
 1:
 END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 #endif
 	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	altivec_unavailable_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* AltiVec Assist */
 	START_EXCEPTION(altivec_assist);
@@ -674,10 +676,11 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 BEGIN_FTR_SECTION
 	bl	altivec_assist_exception
 END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
+	REST_NVGPRS(r1)
 #else
 	bl	unknown_exception
 #endif
-	b	ret_from_except
+	b	interrupt_return
 
 
 /* Decrementer Interrupt */
@@ -719,7 +722,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	unknown_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* Debug exception as a critical interrupt*/
 	START_EXCEPTION(debug_crit);
@@ -786,7 +789,8 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	ld	r14,PACA_EXCRIT+EX_R14(r13)
 	ld	r15,PACA_EXCRIT+EX_R15(r13)
 	bl	DebugException
-	b	ret_from_except
+	REST_NVGPRS(r1)
+	b	interrupt_return
 
 kernel_dbg_exc:
 	b	.	/* NYI */
@@ -857,7 +861,8 @@ kernel_dbg_exc:
 	ld	r14,PACA_EXDBG+EX_R14(r13)
 	ld	r15,PACA_EXDBG+EX_R15(r13)
 	bl	DebugException
-	b	ret_from_except
+	REST_NVGPRS(r1)
+	b	interrupt_return
 
 	START_EXCEPTION(perfmon);
 	NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
@@ -867,7 +872,7 @@ kernel_dbg_exc:
 	CHECK_NAPPING()
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	performance_monitor_exception
-	b	ret_from_except_lite
+	b	interrupt_return
 
 /* Doorbell interrupt */
 	MASKABLE_EXCEPTION(0x280, BOOKE_INTERRUPT_DOORBELL,
@@ -895,7 +900,7 @@ kernel_dbg_exc:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	INTS_RESTORE_HARD
 	bl	unknown_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* Guest Doorbell critical Interrupt */
 	START_EXCEPTION(guest_doorbell_crit);
@@ -916,7 +921,7 @@ kernel_dbg_exc:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	INTS_RESTORE_HARD
 	bl	unknown_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* Embedded Hypervisor priviledged  */
 	START_EXCEPTION(ehpriv);
@@ -926,7 +931,7 @@ kernel_dbg_exc:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	INTS_RESTORE_HARD
 	bl	unknown_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /* LRAT Error interrupt */
 	START_EXCEPTION(lrat_error);
@@ -936,7 +941,7 @@ kernel_dbg_exc:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	INTS_RESTORE_HARD
 	bl	unknown_exception
-	b	ret_from_except
+	b	interrupt_return
 
 /*
  * An interrupt came in while soft-disabled; We mark paca->irq_happened
@@ -998,11 +1003,11 @@ storage_fault_common:
 	bl	do_page_fault
 	cmpdi	r3,0
 	bne-	1f
-	b	ret_from_except_lite
+	b	interrupt_return
 	mr	r4,r3
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	__bad_page_fault
-	b	ret_from_except
+	b	interrupt_return
 
 /*
  * Alignment exception doesn't fit entirely in the 0x100 bytes so it
@@ -1016,284 +1021,8 @@ alignment_more:
 	ld	r15,PACA_EXGEN+EX_R15(r13)
 	INTS_RESTORE_HARD
 	bl	alignment_exception
-	b	ret_from_except
-
-	.align	7
-_GLOBAL(ret_from_except)
 	REST_NVGPRS(r1)
-
-_GLOBAL(ret_from_except_lite)
-	/*
-	 * Disable interrupts so that current_thread_info()->flags
-	 * can't change between when we test it and when we return
-	 * from the interrupt.
-	 */
-	wrteei	0
-
-	ld	r9, PACA_THREAD_INFO(r13)
-	ld	r3,_MSR(r1)
-	ld	r10,PACACURRENT(r13)
-	ld	r4,TI_FLAGS(r9)
-	andi.	r3,r3,MSR_PR
-	beq	resume_kernel
-	lwz	r3,(THREAD+THREAD_DBCR0)(r10)
-
-	/* Check current_thread_info()->flags */
-	andi.	r0,r4,_TIF_USER_WORK_MASK
-	bne	1f
-	/*
-	 * Check to see if the dbcr0 register is set up to debug.
-	 * Use the internal debug mode bit to do this.
-	 */
-	andis.	r0,r3,DBCR0_IDM@h
-	beq	restore
-	mfmsr	r0
-	rlwinm	r0,r0,0,~MSR_DE	/* Clear MSR.DE */
-	mtmsr	r0
-	mtspr	SPRN_DBCR0,r3
-	li	r10, -1
-	mtspr	SPRN_DBSR,r10
-	b	restore
-1:	andi.	r0,r4,_TIF_NEED_RESCHED
-	beq	2f
-	bl	restore_interrupts
-	SCHEDULE_USER
-	b	ret_from_except_lite
-2:
-	/*
-	 * Use a non volatile GPR to save and restore our thread_info flags
-	 * across the call to restore_interrupts.
-	 */
-	mr	r30,r4
-	bl	restore_interrupts
-	mr	r4,r30
-	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	do_notify_resume
-	b	ret_from_except
-
-resume_kernel:
-	/* check current_thread_info, _TIF_EMULATE_STACK_STORE */
-	andis.	r8,r4,_TIF_EMULATE_STACK_STORE@h
-	beq+	1f
-
-	addi	r8,r1,INT_FRAME_SIZE	/* Get the kprobed function entry */
-
-	ld	r3,GPR1(r1)
-	subi	r3,r3,INT_FRAME_SIZE	/* dst: Allocate a trampoline exception frame */
-	mr	r4,r1			/* src:  current exception frame */
-	mr	r1,r3			/* Reroute the trampoline frame to r1 */
-
-	/* Copy from the original to the trampoline. */
-	li	r5,INT_FRAME_SIZE/8	/* size: INT_FRAME_SIZE */
-	li	r6,0			/* start offset: 0 */
-	mtctr	r5
-2:	ldx	r0,r6,r4
-	stdx	r0,r6,r3
-	addi	r6,r6,8
-	bdnz	2b
-
-	/* Do real store operation to complete stdu */
-	ld	r5,GPR1(r1)
-	std	r8,0(r5)
-
-	/* Clear _TIF_EMULATE_STACK_STORE flag */
-	lis	r11,_TIF_EMULATE_STACK_STORE@h
-	addi	r5,r9,TI_FLAGS
-0:	ldarx	r4,0,r5
-	andc	r4,r4,r11
-	stdcx.	r4,0,r5
-	bne-	0b
-1:
-
-#ifdef CONFIG_PREEMPT
-	/* Check if we need to preempt */
-	andi.	r0,r4,_TIF_NEED_RESCHED
-	beq+	restore
-	/* Check that preempt_count() == 0 and interrupts are enabled */
-	lwz	r8,TI_PREEMPT(r9)
-	cmpwi	cr0,r8,0
-	bne	restore
-	ld	r0,SOFTE(r1)
-	andi.	r0,r0,IRQS_DISABLED
-	bne	restore
-
-	/*
-	 * Here we are preempting the current task. We want to make
-	 * sure we are soft-disabled first and reconcile irq state.
-	 */
-	RECONCILE_IRQ_STATE(r3,r4)
-	bl	preempt_schedule_irq
-
-	/*
-	 * arch_local_irq_restore() from preempt_schedule_irq above may
-	 * enable hard interrupt but we really should disable interrupts
-	 * when we return from the interrupt, and so that we don't get
-	 * interrupted after loading SRR0/1.
-	 */
-	wrteei	0
-#endif /* CONFIG_PREEMPT */
-
-restore:
-	/*
-	 * This is the main kernel exit path. First we check if we
-	 * are about to re-enable interrupts
-	 */
-	ld	r5,SOFTE(r1)
-	lbz	r6,PACAIRQSOFTMASK(r13)
-	andi.	r5,r5,IRQS_DISABLED
-	bne	.Lrestore_irq_off
-
-	/* We are enabling, were we already enabled ? Yes, just return */
-	andi.	r6,r6,IRQS_DISABLED
-	beq	cr0,fast_exception_return
-
-	/*
-	 * We are about to soft-enable interrupts (we are hard disabled
-	 * at this point). We check if there's anything that needs to
-	 * be replayed first.
-	 */
-	lbz	r0,PACAIRQHAPPENED(r13)
-	cmpwi	cr0,r0,0
-	bne-	.Lrestore_check_irq_replay
-
-	/*
-	 * Get here when nothing happened while soft-disabled, just
-	 * soft-enable and move-on. We will hard-enable as a side
-	 * effect of rfi
-	 */
-.Lrestore_no_replay:
-	TRACE_ENABLE_INTS
-	li	r0,IRQS_ENABLED
-	stb	r0,PACAIRQSOFTMASK(r13);
-
-/* This is the return from load_up_fpu fast path which could do with
- * less GPR restores in fact, but for now we have a single return path
- */
-fast_exception_return:
-	wrteei	0
-1:	mr	r0,r13
-	ld	r10,_MSR(r1)
-	REST_4GPRS(2, r1)
-	andi.	r6,r10,MSR_PR
-	REST_2GPRS(6, r1)
-	beq	1f
-	ACCOUNT_CPU_USER_EXIT(r13, r10, r11)
-	ld	r0,GPR13(r1)
-
-1:	stdcx.	r0,0,r1		/* to clear the reservation */
-
-	ld	r8,_CCR(r1)
-	ld	r9,_LINK(r1)
-	ld	r10,_CTR(r1)
-	ld	r11,_XER(r1)
-	mtcr	r8
-	mtlr	r9
-	mtctr	r10
-	mtxer	r11
-	REST_2GPRS(8, r1)
-	ld	r10,GPR10(r1)
-	ld	r11,GPR11(r1)
-	ld	r12,GPR12(r1)
-	mtspr	SPRN_SPRG_GEN_SCRATCH,r0
-
-	std	r10,PACA_EXGEN+EX_R10(r13);
-	std	r11,PACA_EXGEN+EX_R11(r13);
-	ld	r10,_NIP(r1)
-	ld	r11,_MSR(r1)
-	ld	r0,GPR0(r1)
-	ld	r1,GPR1(r1)
-	mtspr	SPRN_SRR0,r10
-	mtspr	SPRN_SRR1,r11
-	ld	r10,PACA_EXGEN+EX_R10(r13)
-	ld	r11,PACA_EXGEN+EX_R11(r13)
-	mfspr	r13,SPRN_SPRG_GEN_SCRATCH
-	rfi
-
-	/*
-	 * We are returning to a context with interrupts soft disabled.
-	 *
-	 * However, we may also about to hard enable, so we need to
-	 * make sure that in this case, we also clear PACA_IRQ_HARD_DIS
-	 * or that bit can get out of sync and bad things will happen
-	 */
-.Lrestore_irq_off:
-	ld	r3,_MSR(r1)
-	lbz	r7,PACAIRQHAPPENED(r13)
-	andi.	r0,r3,MSR_EE
-	beq	1f
-	rlwinm	r7,r7,0,~PACA_IRQ_HARD_DIS
-	stb	r7,PACAIRQHAPPENED(r13)
-1:
-#if defined(CONFIG_PPC_IRQ_SOFT_MASK_DEBUG) && defined(CONFIG_BUG)
-	/* The interrupt should not have soft enabled. */
-	lbz	r7,PACAIRQSOFTMASK(r13)
-1:	tdeqi	r7,IRQS_ENABLED
-	EMIT_BUG_ENTRY 1b,__FILE__,__LINE__,BUGFLAG_WARNING
-#endif
-	b	fast_exception_return
-
-	/*
-	 * Something did happen, check if a re-emit is needed
-	 * (this also clears paca->irq_happened)
-	 */
-.Lrestore_check_irq_replay:
-	/* XXX: We could implement a fast path here where we check
-	 * for irq_happened being just 0x01, in which case we can
-	 * clear it and return. That means that we would potentially
-	 * miss a decrementer having wrapped all the way around.
-	 *
-	 * Still, this might be useful for things like hash_page
-	 */
-	bl	__check_irq_replay
-	cmpwi	cr0,r3,0
-	beq	.Lrestore_no_replay
-
-	/*
-	 * We need to re-emit an interrupt. We do so by re-using our
-	 * existing exception frame. We first change the trap value,
-	 * but we need to ensure we preserve the low nibble of it
-	 */
-	ld	r4,_TRAP(r1)
-	clrldi	r4,r4,60
-	or	r4,r4,r3
-	std	r4,_TRAP(r1)
-
-	/*
-	 * PACA_IRQ_HARD_DIS won't always be set here, so set it now
-	 * to reconcile the IRQ state. Tracing is already accounted for.
-	 */
-	lbz	r4,PACAIRQHAPPENED(r13)
-	ori	r4,r4,PACA_IRQ_HARD_DIS
-	stb	r4,PACAIRQHAPPENED(r13)
-
-	/*
-	 * Then find the right handler and call it. Interrupts are
-	 * still soft-disabled and we keep them that way.
-	*/
-	cmpwi	cr0,r3,0x500
-	bne	1f
-	addi	r3,r1,STACK_FRAME_OVERHEAD;
-	bl	do_IRQ
-	b	ret_from_except
-1:	cmpwi	cr0,r3,0x900
-	bne	1f
-	addi	r3,r1,STACK_FRAME_OVERHEAD;
-	bl	timer_interrupt
-	b	ret_from_except
-#ifdef CONFIG_PPC_DOORBELL
-1:
-	cmpwi	cr0,r3,0x280
-	bne	1f
-	addi	r3,r1,STACK_FRAME_OVERHEAD;
-	bl	doorbell_exception
-#endif /* CONFIG_PPC_DOORBELL */
-1:	b	ret_from_except /* What else to do here ? */
-
-_ASM_NOKPROBE_SYMBOL(ret_from_except);
-_ASM_NOKPROBE_SYMBOL(ret_from_except_lite);
-_ASM_NOKPROBE_SYMBOL(resume_kernel);
-_ASM_NOKPROBE_SYMBOL(restore);
-_ASM_NOKPROBE_SYMBOL(fast_exception_return);
+	b	interrupt_return
 
 /*
  * Trampolines used when spotting a bad kernel stack pointer in
diff --git a/arch/powerpc/kernel/irq.c b/arch/powerpc/kernel/irq.c
index 5b72abbff96c..08a747b92735 100644
--- a/arch/powerpc/kernel/irq.c
+++ b/arch/powerpc/kernel/irq.c
@@ -104,82 +104,6 @@ static inline notrace unsigned long get_irq_happened(void)
 	return happened;
 }
 
-#ifdef CONFIG_PPC_BOOK3E
-
-/* This is called whenever we are re-enabling interrupts
- * and returns either 0 (nothing to do) or 500/900/280 if
- * there's an EE, DEC or DBELL to generate.
- *
- * This is called in two contexts: From arch_local_irq_restore()
- * before soft-enabling interrupts, and from the exception exit
- * path when returning from an interrupt from a soft-disabled to
- * a soft enabled context. In both case we have interrupts hard
- * disabled.
- *
- * We take care of only clearing the bits we handled in the
- * PACA irq_happened field since we can only re-emit one at a
- * time and we don't want to "lose" one.
- */
-notrace unsigned int __check_irq_replay(void)
-{
-	/*
-	 * We use local_paca rather than get_paca() to avoid all
-	 * the debug_smp_processor_id() business in this low level
-	 * function
-	 */
-	unsigned char happened = local_paca->irq_happened;
-
-	/*
-	 * We are responding to the next interrupt, so interrupt-off
-	 * latencies should be reset here.
-	 */
-	trace_hardirqs_on();
-	trace_hardirqs_off();
-
-	if (happened & PACA_IRQ_DEC) {
-		local_paca->irq_happened &= ~PACA_IRQ_DEC;
-		return 0x900;
-	}
-
-	if (happened & PACA_IRQ_EE) {
-		local_paca->irq_happened &= ~PACA_IRQ_EE;
-		return 0x500;
-	}
-
-	if (happened & PACA_IRQ_DBELL) {
-		local_paca->irq_happened &= ~PACA_IRQ_DBELL;
-		return 0x280;
-	}
-
-	if (happened & PACA_IRQ_HARD_DIS)
-		local_paca->irq_happened &= ~PACA_IRQ_HARD_DIS;
-
-	/* There should be nothing left ! */
-	BUG_ON(local_paca->irq_happened != 0);
-
-	return 0;
-}
-
-/*
- * This is specifically called by assembly code to re-enable interrupts
- * if they are currently disabled. This is typically called before
- * schedule() or do_signal() when returning to userspace. We do it
- * in C to avoid the burden of dealing with lockdep etc...
- *
- * NOTE: This is called with interrupts hard disabled but not marked
- * as such in paca->irq_happened, so we need to resync this.
- */
-void notrace restore_interrupts(void)
-{
-	if (irqs_disabled()) {
-		local_paca->irq_happened |= PACA_IRQ_HARD_DIS;
-		local_irq_enable();
-	} else
-		__hard_irq_enable();
-}
-
-#endif /* CONFIG_PPC_BOOK3E */
-
 void replay_soft_interrupts(void)
 {
 	struct pt_regs regs;
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 05/11] powerpc/64e/interrupt: NMI save irq soft-mask state in C
From: Nicholas Piggin @ 2021-03-16 10:41 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

64e non-maskable interrupts save the state of the irq soft-mask in
asm. This can be done in C in interrupt wrappers as 64s does.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>

I haven't been able to test this with qemu because it doesn't seem
to cause FSL bookE WDT interrupts.

This makes WatchdogException an NMI interrupt, which affects 32-bit
as well (okay, or create a new handler?)
---
 arch/powerpc/include/asm/interrupt.h | 32 +++++++++++++++++--------
 arch/powerpc/kernel/exceptions-64e.S | 36 ++++------------------------
 arch/powerpc/kernel/traps.c          | 13 +++++++++-
 3 files changed, 38 insertions(+), 43 deletions(-)

diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
index 7c633896d758..305d7c17a4cf 100644
--- a/arch/powerpc/include/asm/interrupt.h
+++ b/arch/powerpc/include/asm/interrupt.h
@@ -130,18 +130,32 @@ static inline void interrupt_async_exit_prepare(struct pt_regs *regs, struct int
 
 struct interrupt_nmi_state {
 #ifdef CONFIG_PPC64
-#ifdef CONFIG_PPC_BOOK3S_64
 	u8 irq_soft_mask;
 	u8 irq_happened;
-#endif
 	u8 ftrace_enabled;
 #endif
 };
 
+static inline bool nmi_disables_ftrace(struct pt_regs *regs)
+{
+	/* Allow DEC and PMI to be traced when they are soft-NMI */
+	if (IS_ENABLED(CONFIG_PPC_BOOK3S_64)) {
+		if (TRAP(regs) == 0x900)
+		       return false;
+		if (TRAP(regs) == 0xf00)
+		       return false;
+	}
+	if (IS_ENABLED(CONFIG_PPC_BOOK3E)) {
+		if (TRAP(regs) == 0x260)
+			return false;
+	}
+
+	return true;
+}
+
 static inline void interrupt_nmi_enter_prepare(struct pt_regs *regs, struct interrupt_nmi_state *state)
 {
 #ifdef CONFIG_PPC64
-#ifdef CONFIG_PPC_BOOK3S_64
 	state->irq_soft_mask = local_paca->irq_soft_mask;
 	state->irq_happened = local_paca->irq_happened;
 
@@ -154,9 +168,8 @@ static inline void interrupt_nmi_enter_prepare(struct pt_regs *regs, struct inte
 	local_paca->irq_happened |= PACA_IRQ_HARD_DIS;
 
 	/* Don't do any per-CPU operations until interrupt state is fixed */
-#endif
-	/* Allow DEC and PMI to be traced when they are soft-NMI */
-	if (TRAP(regs) != 0x900 && TRAP(regs) != 0xf00 && TRAP(regs) != 0x260) {
+
+	if (nmi_disables_ftrace(regs)) {
 		state->ftrace_enabled = this_cpu_get_ftrace_enabled();
 		this_cpu_set_ftrace_enabled(0);
 	}
@@ -180,16 +193,14 @@ static inline void interrupt_nmi_exit_prepare(struct pt_regs *regs, struct inter
 		nmi_exit();
 
 #ifdef CONFIG_PPC64
-	if (TRAP(regs) != 0x900 && TRAP(regs) != 0xf00 && TRAP(regs) != 0x260)
+	if (nmi_disables_ftrace(regs))
 		this_cpu_set_ftrace_enabled(state->ftrace_enabled);
 
-#ifdef CONFIG_PPC_BOOK3S_64
 	/* Check we didn't change the pending interrupt mask. */
 	WARN_ON_ONCE((state->irq_happened | PACA_IRQ_HARD_DIS) != local_paca->irq_happened);
 	local_paca->irq_happened = state->irq_happened;
 	local_paca->irq_soft_mask = state->irq_soft_mask;
 #endif
-#endif
 }
 
 /*
@@ -402,6 +413,7 @@ DECLARE_INTERRUPT_HANDLER(SMIException);
 DECLARE_INTERRUPT_HANDLER(handle_hmi_exception);
 DECLARE_INTERRUPT_HANDLER(unknown_exception);
 DECLARE_INTERRUPT_HANDLER_ASYNC(unknown_async_exception);
+DECLARE_INTERRUPT_HANDLER_NMI(unknown_nmi_exception);
 DECLARE_INTERRUPT_HANDLER(instruction_breakpoint_exception);
 DECLARE_INTERRUPT_HANDLER(RunModeException);
 DECLARE_INTERRUPT_HANDLER(single_step_exception);
@@ -425,7 +437,7 @@ DECLARE_INTERRUPT_HANDLER(altivec_assist_exception);
 DECLARE_INTERRUPT_HANDLER(CacheLockingException);
 DECLARE_INTERRUPT_HANDLER(SPEFloatingPointException);
 DECLARE_INTERRUPT_HANDLER(SPEFloatingPointRoundException);
-DECLARE_INTERRUPT_HANDLER(WatchdogException);
+DECLARE_INTERRUPT_HANDLER_NMI(WatchdogException);
 DECLARE_INTERRUPT_HANDLER(kernel_bad_stack);
 
 /* slb.c */
diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index 7a2e7be33092..18be576fc0b3 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -63,9 +63,6 @@
 	ld	reg, (SPECIAL_EXC_##name * 8 + SPECIAL_EXC_FRAME_OFFS)(r1)
 
 special_reg_save:
-	lbz	r9,PACAIRQHAPPENED(r13)
-	RECONCILE_IRQ_STATE(r3,r4)
-
 	/*
 	 * We only need (or have stack space) to save this stuff if
 	 * we interrupted the kernel.
@@ -119,15 +116,11 @@ BEGIN_FTR_SECTION
 	mtspr	SPRN_MAS5,r10
 	mtspr	SPRN_MAS8,r10
 END_FTR_SECTION_IFSET(CPU_FTR_EMB_HV)
-	SPECIAL_EXC_STORE(r9,IRQHAPPENED)
-
 	mfspr	r10,SPRN_DEAR
 	SPECIAL_EXC_STORE(r10,DEAR)
 	mfspr	r10,SPRN_ESR
 	SPECIAL_EXC_STORE(r10,ESR)
 
-	lbz	r10,PACAIRQSOFTMASK(r13)
-	SPECIAL_EXC_STORE(r10,SOFTE)
 	ld	r10,_NIP(r1)
 	SPECIAL_EXC_STORE(r10,CSRR0)
 	ld	r10,_MSR(r1)
@@ -194,27 +187,6 @@ BEGIN_FTR_SECTION
 	mtspr	SPRN_MAS8,r10
 END_FTR_SECTION_IFSET(CPU_FTR_EMB_HV)
 
-	lbz	r6,PACAIRQSOFTMASK(r13)
-	ld	r5,SOFTE(r1)
-
-	/* Interrupts had better not already be enabled... */
-	tweqi	r6,IRQS_ENABLED
-
-	andi.	r6,r5,IRQS_DISABLED
-	bne	1f
-
-	TRACE_ENABLE_INTS
-	stb	r5,PACAIRQSOFTMASK(r13)
-1:
-	/*
-	 * Restore PACAIRQHAPPENED rather than setting it based on
-	 * the return MSR[EE], since we could have interrupted
-	 * interrupt replay or other inconsistent transitory
-	 * states that must remain that way.
-	 */
-	SPECIAL_EXC_LOAD(r10,IRQHAPPENED)
-	stb	r10,PACAIRQHAPPENED(r13)
-
 	SPECIAL_EXC_LOAD(r10,DEAR)
 	mtspr	SPRN_DEAR,r10
 	SPECIAL_EXC_LOAD(r10,ESR)
@@ -566,7 +538,7 @@ __end_interrupts:
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	unknown_exception
+	bl	unknown_nmi_exception
 	b	ret_from_crit_except
 
 /* Machine Check Interrupt */
@@ -702,7 +674,7 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 #ifdef CONFIG_BOOKE_WDT
 	bl	WatchdogException
 #else
-	bl	unknown_exception
+	bl	unknown_nmi_exception
 #endif
 	b	ret_from_crit_except
 
@@ -886,7 +858,7 @@ kernel_dbg_exc:
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	unknown_exception
+	bl	unknown_nmi_exception
 	b	ret_from_crit_except
 
 /*
@@ -910,7 +882,7 @@ kernel_dbg_exc:
 	bl	special_reg_save
 	CHECK_NAPPING();
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	unknown_exception
+	bl	unknown_nmi_exception
 	b	ret_from_crit_except
 
 /* Hypervisor call */
diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
index c74e7727860a..97b5f3d83ff7 100644
--- a/arch/powerpc/kernel/traps.c
+++ b/arch/powerpc/kernel/traps.c
@@ -1079,6 +1079,16 @@ DEFINE_INTERRUPT_HANDLER_ASYNC(unknown_async_exception)
 	_exception(SIGTRAP, regs, TRAP_UNK, 0);
 }
 
+DEFINE_INTERRUPT_HANDLER_NMI(unknown_nmi_exception)
+{
+	printk("Bad trap at PC: %lx, SR: %lx, vector=%lx\n",
+	       regs->nip, regs->msr, regs->trap);
+
+	_exception(SIGTRAP, regs, TRAP_UNK, 0);
+
+	return 0;
+}
+
 DEFINE_INTERRUPT_HANDLER(instruction_breakpoint_exception)
 {
 	if (notify_die(DIE_IABR_MATCH, "iabr_match", regs, 5,
@@ -2183,10 +2193,11 @@ void __attribute__ ((weak)) WatchdogHandler(struct pt_regs *regs)
 	return;
 }
 
-DEFINE_INTERRUPT_HANDLER(WatchdogException) /* XXX NMI? async? */
+DEFINE_INTERRUPT_HANDLER_NMI(WatchdogException)
 {
 	printk (KERN_EMERG "PowerPC Book-E Watchdog Exception\n");
 	WatchdogHandler(regs);
+	return 0;
 }
 #endif
 
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 06/11] powerpc/64e/interrupt: reconcile irq soft-mask state in C
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

Use existing 64s interrupt entry wrapper code to reconcile irqs in C.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/interrupt.h |  8 +++---
 arch/powerpc/kernel/entry_64.S       | 18 ++++++-------
 arch/powerpc/kernel/exceptions-64e.S | 39 +---------------------------
 3 files changed, 13 insertions(+), 52 deletions(-)

diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
index 305d7c17a4cf..29b48d083156 100644
--- a/arch/powerpc/include/asm/interrupt.h
+++ b/arch/powerpc/include/asm/interrupt.h
@@ -40,14 +40,14 @@ static inline void interrupt_enter_prepare(struct pt_regs *regs, struct interrup
 		kuap_save_and_lock(regs);
 	}
 #endif
-	/*
-	 * Book3E reconciles irq soft mask in asm
-	 */
-#ifdef CONFIG_PPC_BOOK3S_64
+
+#ifdef CONFIG_PPC64
 	if (irq_soft_mask_set_return(IRQS_ALL_DISABLED) == IRQS_ENABLED)
 		trace_hardirqs_off();
 	local_paca->irq_happened |= PACA_IRQ_HARD_DIS;
+#endif
 
+#ifdef CONFIG_PPC_BOOK3S_64
 	if (user_mode(regs)) {
 		CT_WARN_ON(ct_state() != CONTEXT_USER);
 		user_exit_irqoff();
diff --git a/arch/powerpc/kernel/entry_64.S b/arch/powerpc/kernel/entry_64.S
index 555b3d0a3f38..03727308d8cc 100644
--- a/arch/powerpc/kernel/entry_64.S
+++ b/arch/powerpc/kernel/entry_64.S
@@ -117,13 +117,12 @@ BEGIN_FTR_SECTION
 END_FTR_SECTION_IFSET(CPU_FTR_HAS_PPR)
 
 	/*
-	 * RECONCILE_IRQ_STATE without calling trace_hardirqs_off(), which
-	 * would clobber syscall parameters. Also we always enter with IRQs
-	 * enabled and nothing pending. system_call_exception() will call
-	 * trace_hardirqs_off().
-	 *
-	 * scv enters with MSR[EE]=1, so don't set PACA_IRQ_HARD_DIS. The
-	 * entry vector already sets PACAIRQSOFTMASK to IRQS_ALL_DISABLED.
+	 * scv enters with MSR[EE]=1 and is immediately considered soft-masked.
+	 * The entry vector already sets PACAIRQSOFTMASK to IRQS_ALL_DISABLED,
+	 * and interrupts may be masked and pending already.
+	 * system_call_exception() will call trace_hardirqs_off() which means
+	 * interrupts could already have been blocked before trace_hardirqs_off,
+	 * but this is the best we can do.
 	 */
 
 	/* Calling convention has r9 = orig r0, r10 = regs */
@@ -288,9 +287,8 @@ END_BTB_FLUSH_SECTION
 	std	r11,-16(r10)		/* "regshere" marker */
 
 	/*
-	 * RECONCILE_IRQ_STATE without calling trace_hardirqs_off(), which
-	 * would clobber syscall parameters. Also we always enter with IRQs
-	 * enabled and nothing pending. system_call_exception() will call
+	 * We always enter kernel from userspace with irq soft-mask enabled and
+	 * nothing pending. system_call_exception() will call
 	 * trace_hardirqs_off().
 	 */
 	li	r11,IRQS_ALL_DISABLED
diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index 18be576fc0b3..3c222a97f023 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -409,28 +409,6 @@ exc_##n##_common:							    \
 #define EXCEPTION_COMMON_DBG(n) \
 	EXCEPTION_COMMON_LVL(n, SPRN_SPRG_DBG_SCRATCH, PACA_EXDBG)
 
-/*
- * This is meant for exceptions that don't immediately hard-enable.  We
- * set a bit in paca->irq_happened to ensure that a subsequent call to
- * arch_local_irq_restore() will properly hard-enable and avoid the
- * fast-path, and then reconcile irq state.
- */
-#define INTS_DISABLE	RECONCILE_IRQ_STATE(r3,r4)
-
-/*
- * This is called by exceptions that don't use INTS_DISABLE (that did not
- * touch irq indicators in the PACA).  This will restore MSR:EE to it's
- * previous value
- *
- * XXX In the long run, we may want to open-code it in order to separate the
- *     load from the wrtee, thus limiting the latency caused by the dependency
- *     but at this point, I'll favor code clarity until we have a near to final
- *     implementation
- */
-#define INTS_RESTORE_HARD						    \
-	ld	r11,_MSR(r1);						    \
-	wrtee	r11;
-
 /* XXX FIXME: Restore r14/r15 when necessary */
 #define BAD_STACK_TRAMPOLINE(n)						    \
 exc_##n##_bad_stack:							    \
@@ -479,7 +457,6 @@ exc_##n##_bad_stack:							    \
 	START_EXCEPTION(label);						\
 	NORMAL_EXCEPTION_PROLOG(trapnum, intnum, PROLOG_ADDITION_MASKABLE)\
 	EXCEPTION_COMMON(trapnum)					\
-	INTS_DISABLE;							\
 	ack(r8);							\
 	CHECK_NAPPING();						\
 	addi	r3,r1,STACK_FRAME_OVERHEAD;				\
@@ -559,7 +536,6 @@ __end_interrupts:
 	mfspr	r14,SPRN_DEAR
 	mfspr	r15,SPRN_ESR
 	EXCEPTION_COMMON(0x300)
-	INTS_DISABLE
 	b	storage_fault_common
 
 /* Instruction Storage Interrupt */
@@ -569,7 +545,6 @@ __end_interrupts:
 	li	r15,0
 	mr	r14,r10
 	EXCEPTION_COMMON(0x400)
-	INTS_DISABLE
 	b	storage_fault_common
 
 /* External Input Interrupt */
@@ -591,7 +566,6 @@ __end_interrupts:
 				PROLOG_ADDITION_1REG)
 	mfspr	r14,SPRN_ESR
 	EXCEPTION_COMMON(0x700)
-	INTS_DISABLE
 	std	r14,_DSISR(r1)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXGEN+EX_R14(r13)
@@ -610,8 +584,7 @@ __end_interrupts:
 	beq-	1f
 	bl	load_up_fpu
 	b	fast_interrupt_return
-1:	INTS_DISABLE
-	addi	r3,r1,STACK_FRAME_OVERHEAD
+1:	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	kernel_fp_unavailable_exception
 	b	interrupt_return
 
@@ -631,7 +604,6 @@ BEGIN_FTR_SECTION
 1:
 END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 #endif
-	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	altivec_unavailable_exception
 	b	interrupt_return
@@ -642,7 +614,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 				BOOKE_INTERRUPT_ALTIVEC_ASSIST,
 				PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x220)
-	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 #ifdef CONFIG_ALTIVEC
 BEGIN_FTR_SECTION
@@ -691,7 +662,6 @@ END_FTR_SECTION_IFSET(CPU_FTR_ALTIVEC)
 	NORMAL_EXCEPTION_PROLOG(0xf20, BOOKE_INTERRUPT_AP_UNAVAIL,
 				PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0xf20)
-	INTS_DISABLE
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	unknown_exception
 	b	interrupt_return
@@ -827,7 +797,6 @@ kernel_dbg_exc:
 	 */
 	mfspr	r14,SPRN_DBSR
 	EXCEPTION_COMMON_DBG(0xd08)
-	INTS_DISABLE
 	std	r14,_DSISR(r1)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXDBG+EX_R14(r13)
@@ -840,7 +809,6 @@ kernel_dbg_exc:
 	NORMAL_EXCEPTION_PROLOG(0x260, BOOKE_INTERRUPT_PERFORMANCE_MONITOR,
 				PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x260)
-	INTS_DISABLE
 	CHECK_NAPPING()
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	bl	performance_monitor_exception
@@ -870,7 +838,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x2c0)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	interrupt_return
 
@@ -891,7 +858,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x310)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	interrupt_return
 
@@ -901,7 +867,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x320)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	interrupt_return
 
@@ -911,7 +876,6 @@ kernel_dbg_exc:
 			        PROLOG_ADDITION_NONE)
 	EXCEPTION_COMMON(0x340)
 	addi	r3,r1,STACK_FRAME_OVERHEAD
-	INTS_RESTORE_HARD
 	bl	unknown_exception
 	b	interrupt_return
 
@@ -991,7 +955,6 @@ alignment_more:
 	addi	r3,r1,STACK_FRAME_OVERHEAD
 	ld	r14,PACA_EXGEN+EX_R14(r13)
 	ld	r15,PACA_EXGEN+EX_R15(r13)
-	INTS_RESTORE_HARD
 	bl	alignment_exception
 	REST_NVGPRS(r1)
 	b	interrupt_return
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 07/11] powerpc/64e/interrupt: Use new interrupt context tracking scheme
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

With the new interrupt exit code, context tracking can be managed
more precisely, so remove the last of the 64e workarounds and switch
to the new context tracking code already used by 64s.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/interrupt.h | 28 ----------------------------
 arch/powerpc/kernel/interrupt.c      | 16 +---------------
 2 files changed, 1 insertion(+), 43 deletions(-)

diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
index 29b48d083156..94fd8e1ff52c 100644
--- a/arch/powerpc/include/asm/interrupt.h
+++ b/arch/powerpc/include/asm/interrupt.h
@@ -10,9 +10,6 @@
 #include <asm/runlatch.h>
 
 struct interrupt_state {
-#ifdef CONFIG_PPC_BOOK3E_64
-	enum ctx_state ctx_state;
-#endif
 };
 
 static inline void booke_restore_dbcr0(void)
@@ -45,9 +42,7 @@ static inline void interrupt_enter_prepare(struct pt_regs *regs, struct interrup
 	if (irq_soft_mask_set_return(IRQS_ALL_DISABLED) == IRQS_ENABLED)
 		trace_hardirqs_off();
 	local_paca->irq_happened |= PACA_IRQ_HARD_DIS;
-#endif
 
-#ifdef CONFIG_PPC_BOOK3S_64
 	if (user_mode(regs)) {
 		CT_WARN_ON(ct_state() != CONTEXT_USER);
 		user_exit_irqoff();
@@ -64,12 +59,6 @@ static inline void interrupt_enter_prepare(struct pt_regs *regs, struct interrup
 	}
 #endif
 
-#ifdef CONFIG_PPC_BOOK3E_64
-	state->ctx_state = exception_enter();
-	if (user_mode(regs))
-		account_cpu_user_entry();
-#endif
-
 	booke_restore_dbcr0();
 }
 
@@ -89,25 +78,8 @@ static inline void interrupt_enter_prepare(struct pt_regs *regs, struct interrup
  */
 static inline void interrupt_exit_prepare(struct pt_regs *regs, struct interrupt_state *state)
 {
-#ifdef CONFIG_PPC_BOOK3E_64
-	exception_exit(state->ctx_state);
-#endif
-
 	if (user_mode(regs))
 		kuep_unlock();
-	/*
-	 * Book3S exits to user via interrupt_exit_user_prepare(), which does
-	 * context tracking, which is a cleaner way to handle PREEMPT=y
-	 * and avoid context entry/exit in e.g., preempt_schedule_irq()),
-	 * which is likely to be where the core code wants to end up.
-	 *
-	 * The above comment explains why we can't do the
-	 *
-	 *     if (user_mode(regs))
-	 *         user_exit_irqoff();
-	 *
-	 * sequence here.
-	 */
 }
 
 static inline void interrupt_async_enter_prepare(struct pt_regs *regs, struct interrupt_state *state)
diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
index 381a618b5b5b..1b0e1792ac37 100644
--- a/arch/powerpc/kernel/interrupt.c
+++ b/arch/powerpc/kernel/interrupt.c
@@ -235,10 +235,6 @@ static notrace void booke_load_dbcr0(void)
 #endif
 }
 
-/* temporary hack for context tracking, removed in later patch */
-#include <linux/sched/debug.h>
-asmlinkage __visible void __sched schedule_user(void);
-
 /*
  * This should be called after a syscall returns, with r3 the return value
  * from the syscall. If this function returns non-zero, the system call
@@ -296,11 +292,7 @@ notrace unsigned long syscall_exit_prepare(unsigned long r3,
 	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
 		local_irq_enable();
 		if (ti_flags & _TIF_NEED_RESCHED) {
-#ifdef CONFIG_PPC_BOOK3E_64
-			schedule_user();
-#else
 			schedule();
-#endif
 		} else {
 			/*
 			 * SIGPENDING must restore signal handler function
@@ -375,9 +367,7 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 	BUG_ON(!(regs->msr & MSR_PR));
 	BUG_ON(!FULL_REGS(regs));
 	BUG_ON(arch_irq_disabled_regs(regs));
-#ifdef CONFIG_PPC_BOOK3S_64
 	CT_WARN_ON(ct_state() == CONTEXT_USER);
-#endif
 
 	/*
 	 * We don't need to restore AMR on the way back to userspace for KUAP.
@@ -392,11 +382,7 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 	while (unlikely(ti_flags & (_TIF_USER_WORK_MASK & ~_TIF_RESTORE_TM))) {
 		local_irq_enable(); /* returning to user: may enable */
 		if (ti_flags & _TIF_NEED_RESCHED) {
-#ifdef CONFIG_PPC_BOOK3E_64
-			schedule_user();
-#else
 			schedule();
-#endif
 		} else {
 			if (ti_flags & _TIF_SIGPENDING)
 				ret |= _TIF_RESTOREALL;
@@ -464,7 +450,7 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 	 * CT_WARN_ON comes here via program_check_exception,
 	 * so avoid recursion.
 	 */
-	if (IS_ENABLED(CONFIG_BOOKS) && TRAP(regs) != 0x700)
+	if (TRAP(regs) != 0x700)
 		CT_WARN_ON(ct_state() == CONTEXT_USER);
 
 	kuap = kuap_get_and_assert_locked();
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 08/11] powerpc/64e/interrupt: handle bad_page_fault in C
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

With non-volatile registers saved on interrupt, bad_page_fault
can now be called by do_page_fault.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/kernel/exceptions-64e.S | 6 ------
 arch/powerpc/mm/fault.c              | 5 +----
 2 files changed, 1 insertion(+), 10 deletions(-)

diff --git a/arch/powerpc/kernel/exceptions-64e.S b/arch/powerpc/kernel/exceptions-64e.S
index 3c222a97f023..7c3654b0d0f4 100644
--- a/arch/powerpc/kernel/exceptions-64e.S
+++ b/arch/powerpc/kernel/exceptions-64e.S
@@ -937,12 +937,6 @@ storage_fault_common:
 	ld	r14,PACA_EXGEN+EX_R14(r13)
 	ld	r15,PACA_EXGEN+EX_R15(r13)
 	bl	do_page_fault
-	cmpdi	r3,0
-	bne-	1f
-	b	interrupt_return
-	mr	r4,r3
-	addi	r3,r1,STACK_FRAME_OVERHEAD
-	bl	__bad_page_fault
 	b	interrupt_return
 
 /*
diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
index 2e54bac99a22..44833660b21d 100644
--- a/arch/powerpc/mm/fault.c
+++ b/arch/powerpc/mm/fault.c
@@ -552,12 +552,9 @@ static long __do_page_fault(struct pt_regs *regs)
 	if (likely(entry)) {
 		instruction_pointer_set(regs, extable_fixup(entry));
 		return 0;
-	} else if (!IS_ENABLED(CONFIG_PPC_BOOK3E_64)) {
+	} else {
 		__bad_page_fault(regs, err);
 		return 0;
-	} else {
-		/* 32 and 64e handle the bad page fault in asm */
-		return err;
 	}
 }
 NOKPROBE_SYMBOL(__do_page_fault);
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 09/11] powerpc: clean up do_page_fault
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

search_exception_tables + __bad_page_fault can be substituted with
bad_page_fault, do_page_fault no longer needs to return a value
to asm for any sub-architecture, and __bad_page_fault can be static.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/bug.h        |  5 +----
 arch/powerpc/include/asm/interrupt.h  |  2 +-
 arch/powerpc/mm/book3s64/hash_utils.c | 16 +++++++---------
 arch/powerpc/mm/fault.c               | 27 ++++++++-------------------
 4 files changed, 17 insertions(+), 33 deletions(-)

diff --git a/arch/powerpc/include/asm/bug.h b/arch/powerpc/include/asm/bug.h
index d1635ffbb179..0b2162890d8b 100644
--- a/arch/powerpc/include/asm/bug.h
+++ b/arch/powerpc/include/asm/bug.h
@@ -111,11 +111,8 @@
 #ifndef __ASSEMBLY__
 
 struct pt_regs;
-long do_page_fault(struct pt_regs *);
-long hash__do_page_fault(struct pt_regs *);
+void hash__do_page_fault(struct pt_regs *);
 void bad_page_fault(struct pt_regs *, int);
-void __bad_page_fault(struct pt_regs *regs, int sig);
-void do_bad_page_fault_segv(struct pt_regs *regs);
 extern void _exception(int, struct pt_regs *, int, unsigned long);
 extern void _exception_pkey(struct pt_regs *, unsigned long, int);
 extern void die(const char *, struct pt_regs *, long);
diff --git a/arch/powerpc/include/asm/interrupt.h b/arch/powerpc/include/asm/interrupt.h
index 94fd8e1ff52c..bd0bd9430f78 100644
--- a/arch/powerpc/include/asm/interrupt.h
+++ b/arch/powerpc/include/asm/interrupt.h
@@ -420,7 +420,7 @@ DECLARE_INTERRUPT_HANDLER(do_bad_slb_fault);
 DECLARE_INTERRUPT_HANDLER_RAW(do_hash_fault);
 
 /* fault.c */
-DECLARE_INTERRUPT_HANDLER_RET(do_page_fault);
+DECLARE_INTERRUPT_HANDLER(do_page_fault);
 DECLARE_INTERRUPT_HANDLER(do_bad_page_fault_segv);
 
 /* process.c */
diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 581b20a2feaf..1c4b0a29f0f5 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -1572,10 +1572,11 @@ DEFINE_INTERRUPT_HANDLER_RET(__do_hash_fault)
 DEFINE_INTERRUPT_HANDLER_RAW(do_hash_fault)
 {
 	unsigned long dsisr = regs->dsisr;
-	long err;
 
-	if (unlikely(dsisr & (DSISR_BAD_FAULT_64S | DSISR_KEYFAULT)))
-		goto page_fault;
+	if (unlikely(dsisr & (DSISR_BAD_FAULT_64S | DSISR_KEYFAULT))) {
+		hash__do_page_fault(regs);
+		return 0;
+	}
 
 	/*
 	 * If we are in an "NMI" (e.g., an interrupt when soft-disabled), then
@@ -1595,13 +1596,10 @@ DEFINE_INTERRUPT_HANDLER_RAW(do_hash_fault)
 		return 0;
 	}
 
-	err = __do_hash_fault(regs);
-	if (err) {
-page_fault:
-		err = hash__do_page_fault(regs);
-	}
+	if (__do_hash_fault(regs))
+		hash__do_page_fault(regs);
 
-	return err;
+	return 0;
 }
 
 #ifdef CONFIG_PPC_MM_SLICES
diff --git a/arch/powerpc/mm/fault.c b/arch/powerpc/mm/fault.c
index 44833660b21d..2c6e9a6bb86f 100644
--- a/arch/powerpc/mm/fault.c
+++ b/arch/powerpc/mm/fault.c
@@ -539,36 +539,25 @@ static int ___do_page_fault(struct pt_regs *regs, unsigned long address,
 }
 NOKPROBE_SYMBOL(___do_page_fault);
 
-static long __do_page_fault(struct pt_regs *regs)
+static __always_inline void __do_page_fault(struct pt_regs *regs)
 {
-	const struct exception_table_entry *entry;
 	long err;
 
 	err = ___do_page_fault(regs, regs->dar, regs->dsisr);
-	if (likely(!err))
-		return err;
-
-	entry = search_exception_tables(regs->nip);
-	if (likely(entry)) {
-		instruction_pointer_set(regs, extable_fixup(entry));
-		return 0;
-	} else {
-		__bad_page_fault(regs, err);
-		return 0;
-	}
+	if (unlikely(err))
+		bad_page_fault(regs, err);
 }
-NOKPROBE_SYMBOL(__do_page_fault);
 
-DEFINE_INTERRUPT_HANDLER_RET(do_page_fault)
+DEFINE_INTERRUPT_HANDLER(do_page_fault)
 {
-	return __do_page_fault(regs);
+	__do_page_fault(regs);
 }
 
 #ifdef CONFIG_PPC_BOOK3S_64
 /* Same as do_page_fault but interrupt entry has already run in do_hash_fault */
-long hash__do_page_fault(struct pt_regs *regs)
+void hash__do_page_fault(struct pt_regs *regs)
 {
-	return __do_page_fault(regs);
+	__do_page_fault(regs);
 }
 NOKPROBE_SYMBOL(hash__do_page_fault);
 #endif
@@ -578,7 +567,7 @@ NOKPROBE_SYMBOL(hash__do_page_fault);
  * It is called from the DSI and ISI handlers in head.S and from some
  * of the procedures in traps.c.
  */
-void __bad_page_fault(struct pt_regs *regs, int sig)
+static void __bad_page_fault(struct pt_regs *regs, int sig)
 {
 	int is_write = page_fault_is_write(regs->dsisr);
 
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 10/11] powerpc: remove partial register save logic
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

All subarchitectures always save all GPRs to pt_regs interrupt frames
now. Remove FULL_REGS and associated bits.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/ptrace.h        | 17 ++---------------
 arch/powerpc/kernel/align.c              |  6 ------
 arch/powerpc/kernel/interrupt.c          |  3 ---
 arch/powerpc/kernel/process.c            | 12 ------------
 arch/powerpc/kernel/ptrace/ptrace-view.c | 21 ---------------------
 arch/powerpc/kernel/ptrace/ptrace.c      |  2 --
 arch/powerpc/kernel/ptrace/ptrace32.c    |  4 ----
 arch/powerpc/kernel/signal_32.c          |  3 ---
 arch/powerpc/kernel/signal_64.c          |  2 --
 arch/powerpc/kernel/traps.c              |  1 -
 arch/powerpc/lib/sstep.c                 |  4 ----
 arch/powerpc/xmon/xmon.c                 | 23 +++++++----------------
 12 files changed, 9 insertions(+), 89 deletions(-)

diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index c5b3669918f4..91194fdd5d01 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -188,29 +188,16 @@ static inline void regs_set_return_value(struct pt_regs *regs, unsigned long rc)
 #ifdef __powerpc64__
 #define TRAP_FLAGS_MASK		0x10
 #define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
-#define FULL_REGS(regs)		true
-#define SET_FULL_REGS(regs)	do { } while (0)
-#define CHECK_FULL_REGS(regs)	do { } while (0)
-#define NV_REG_POISON		0xdeadbeefdeadbeefUL
 #else
 /*
- * We use the least-significant bit of the trap field to indicate
- * whether we have saved the full set of registers, or only a
- * partial set.  A 1 there means the partial set.
- * On 4xx we use the next bit to indicate whether the exception
+ * On 4xx we use bit 1 in the trap word to indicate whether the exception
  * is a critical exception (1 means it is).
  */
-#define TRAP_FLAGS_MASK		0x1F
+#define TRAP_FLAGS_MASK		0x1E
 #define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
-#define FULL_REGS(regs)		true
-#define SET_FULL_REGS(regs)	do { } while (0)
 #define IS_CRITICAL_EXC(regs)	(((regs)->trap & 2) != 0)
 #define IS_MCHECK_EXC(regs)	(((regs)->trap & 4) != 0)
 #define IS_DEBUG_EXC(regs)	(((regs)->trap & 8) != 0)
-#define NV_REG_POISON		0xdeadbeef
-#define CHECK_FULL_REGS(regs)						      \
-do {									      \
-} while (0)
 #endif /* __powerpc64__ */
 
 static inline void set_trap(struct pt_regs *regs, unsigned long val)
diff --git a/arch/powerpc/kernel/align.c b/arch/powerpc/kernel/align.c
index c7797eb958c7..ae525397947e 100644
--- a/arch/powerpc/kernel/align.c
+++ b/arch/powerpc/kernel/align.c
@@ -299,12 +299,6 @@ int fix_alignment(struct pt_regs *regs)
 	struct instruction_op op;
 	int r, type;
 
-	/*
-	 * We require a complete register set, if not, then our assembly
-	 * is broken
-	 */
-	CHECK_FULL_REGS(regs);
-
 	if (unlikely(__get_user_instr(instr, (void __user *)regs->nip)))
 		return -EFAULT;
 	if ((regs->msr & MSR_LE) != (MSR_KERNEL & MSR_LE)) {
diff --git a/arch/powerpc/kernel/interrupt.c b/arch/powerpc/kernel/interrupt.c
index 1b0e1792ac37..b953bb5027e6 100644
--- a/arch/powerpc/kernel/interrupt.c
+++ b/arch/powerpc/kernel/interrupt.c
@@ -51,7 +51,6 @@ notrace long system_call_exception(long r3, long r4, long r5,
 	if (!IS_ENABLED(CONFIG_BOOKE) && !IS_ENABLED(CONFIG_40x))
 		BUG_ON(!(regs->msr & MSR_RI));
 	BUG_ON(!(regs->msr & MSR_PR));
-	BUG_ON(!FULL_REGS(regs));
 	BUG_ON(arch_irq_disabled_regs(regs));
 
 #ifdef CONFIG_PPC_PKEY
@@ -365,7 +364,6 @@ notrace unsigned long interrupt_exit_user_prepare(struct pt_regs *regs, unsigned
 	if (!IS_ENABLED(CONFIG_BOOKE) && !IS_ENABLED(CONFIG_40x))
 		BUG_ON(!(regs->msr & MSR_RI));
 	BUG_ON(!(regs->msr & MSR_PR));
-	BUG_ON(!FULL_REGS(regs));
 	BUG_ON(arch_irq_disabled_regs(regs));
 	CT_WARN_ON(ct_state() == CONTEXT_USER);
 
@@ -445,7 +443,6 @@ notrace unsigned long interrupt_exit_kernel_prepare(struct pt_regs *regs, unsign
 	    unlikely(!(regs->msr & MSR_RI)))
 		unrecoverable_exception(regs);
 	BUG_ON(regs->msr & MSR_PR);
-	BUG_ON(!FULL_REGS(regs));
 	/*
 	 * CT_WARN_ON comes here via program_check_exception,
 	 * so avoid recursion.
diff --git a/arch/powerpc/kernel/process.c b/arch/powerpc/kernel/process.c
index 7989d9ce468b..1e62a70a29aa 100644
--- a/arch/powerpc/kernel/process.c
+++ b/arch/powerpc/kernel/process.c
@@ -1447,11 +1447,9 @@ static void print_msr_bits(unsigned long val)
 #ifdef CONFIG_PPC64
 #define REG		"%016lx"
 #define REGS_PER_LINE	4
-#define LAST_VOLATILE	13
 #else
 #define REG		"%08lx"
 #define REGS_PER_LINE	8
-#define LAST_VOLATILE	12
 #endif
 
 static void __show_regs(struct pt_regs *regs)
@@ -1487,8 +1485,6 @@ static void __show_regs(struct pt_regs *regs)
 		if ((i % REGS_PER_LINE) == 0)
 			pr_cont("\nGPR%02d: ", i);
 		pr_cont(REG " ", regs->gpr[i]);
-		if (i == LAST_VOLATILE && !FULL_REGS(regs))
-			break;
 	}
 	pr_cont("\n");
 	/*
@@ -1691,7 +1687,6 @@ int copy_thread(unsigned long clone_flags, unsigned long usp,
 	} else {
 		/* user thread */
 		struct pt_regs *regs = current_pt_regs();
-		CHECK_FULL_REGS(regs);
 		*childregs = *regs;
 		if (usp)
 			childregs->gpr[1] = usp;
@@ -1796,13 +1791,6 @@ void start_thread(struct pt_regs *regs, unsigned long start, unsigned long sp)
 	regs->ccr = 0;
 	regs->gpr[1] = sp;
 
-	/*
-	 * We have just cleared all the nonvolatile GPRs, so make
-	 * FULL_REGS(regs) return true.  This is necessary to allow
-	 * ptrace to examine the thread immediately after exec.
-	 */
-	SET_FULL_REGS(regs);
-
 #ifdef CONFIG_PPC32
 	regs->mq = 0;
 	regs->nip = start;
diff --git a/arch/powerpc/kernel/ptrace/ptrace-view.c b/arch/powerpc/kernel/ptrace/ptrace-view.c
index 2bad8068f598..8eb826aa2a10 100644
--- a/arch/powerpc/kernel/ptrace/ptrace-view.c
+++ b/arch/powerpc/kernel/ptrace/ptrace-view.c
@@ -221,17 +221,9 @@ static int gpr_get(struct task_struct *target, const struct user_regset *regset,
 #ifdef CONFIG_PPC64
 	struct membuf to_softe = membuf_at(&to, offsetof(struct pt_regs, softe));
 #endif
-	int i;
-
 	if (target->thread.regs == NULL)
 		return -EIO;
 
-	if (!FULL_REGS(target->thread.regs)) {
-		/* We have a partial register set.  Fill 14-31 with bogus values */
-		for (i = 14; i < 32; i++)
-			target->thread.regs->gpr[i] = NV_REG_POISON;
-	}
-
 	membuf_write(&to, target->thread.regs, sizeof(struct user_pt_regs));
 
 	membuf_store(&to_msr, get_user_msr(target));
@@ -252,8 +244,6 @@ static int gpr_set(struct task_struct *target, const struct user_regset *regset,
 	if (target->thread.regs == NULL)
 		return -EIO;
 
-	CHECK_FULL_REGS(target->thread.regs);
-
 	ret = user_regset_copyin(&pos, &count, &kbuf, &ubuf,
 				 target->thread.regs,
 				 0, PT_MSR * sizeof(reg));
@@ -723,19 +713,9 @@ static int gpr32_get(struct task_struct *target,
 		     const struct user_regset *regset,
 		     struct membuf to)
 {
-	int i;
-
 	if (target->thread.regs == NULL)
 		return -EIO;
 
-	if (!FULL_REGS(target->thread.regs)) {
-		/*
-		 * We have a partial register set.
-		 * Fill 14-31 with bogus values.
-		 */
-		for (i = 14; i < 32; i++)
-			target->thread.regs->gpr[i] = NV_REG_POISON;
-	}
 	return gpr32_get_common(target, regset, to,
 			&target->thread.regs->gpr[0]);
 }
@@ -748,7 +728,6 @@ static int gpr32_set(struct task_struct *target,
 	if (target->thread.regs == NULL)
 		return -EIO;
 
-	CHECK_FULL_REGS(target->thread.regs);
 	return gpr32_set_common(target, regset, pos, count, kbuf, ubuf,
 			&target->thread.regs->gpr[0]);
 }
diff --git a/arch/powerpc/kernel/ptrace/ptrace.c b/arch/powerpc/kernel/ptrace/ptrace.c
index 4f3d4ff3728c..f59883902b35 100644
--- a/arch/powerpc/kernel/ptrace/ptrace.c
+++ b/arch/powerpc/kernel/ptrace/ptrace.c
@@ -59,7 +59,6 @@ long arch_ptrace(struct task_struct *child, long request,
 		if ((addr & (sizeof(long) - 1)) || !child->thread.regs)
 			break;
 
-		CHECK_FULL_REGS(child->thread.regs);
 		if (index < PT_FPR0)
 			ret = ptrace_get_reg(child, (int) index, &tmp);
 		else
@@ -81,7 +80,6 @@ long arch_ptrace(struct task_struct *child, long request,
 		if ((addr & (sizeof(long) - 1)) || !child->thread.regs)
 			break;
 
-		CHECK_FULL_REGS(child->thread.regs);
 		if (index < PT_FPR0)
 			ret = ptrace_put_reg(child, index, data);
 		else
diff --git a/arch/powerpc/kernel/ptrace/ptrace32.c b/arch/powerpc/kernel/ptrace/ptrace32.c
index d30b9ad70edc..19c224808982 100644
--- a/arch/powerpc/kernel/ptrace/ptrace32.c
+++ b/arch/powerpc/kernel/ptrace/ptrace32.c
@@ -83,7 +83,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
 		if ((addr & 3) || (index > PT_FPSCR32))
 			break;
 
-		CHECK_FULL_REGS(child->thread.regs);
 		if (index < PT_FPR0) {
 			ret = ptrace_get_reg(child, index, &tmp);
 			if (ret)
@@ -133,7 +132,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
 		if ((addr & 3) || numReg > PT_FPSCR)
 			break;
 
-		CHECK_FULL_REGS(child->thread.regs);
 		if (numReg >= PT_FPR0) {
 			flush_fp_to_thread(child);
 			/* get 64 bit FPR */
@@ -187,7 +185,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
 		if ((addr & 3) || (index > PT_FPSCR32))
 			break;
 
-		CHECK_FULL_REGS(child->thread.regs);
 		if (index < PT_FPR0) {
 			ret = ptrace_put_reg(child, index, data);
 		} else {
@@ -226,7 +223,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
 		 */
 		if ((addr & 3) || (numReg > PT_FPSCR))
 			break;
-		CHECK_FULL_REGS(child->thread.regs);
 		if (numReg < PT_FPR0) {
 			unsigned long freg;
 			ret = ptrace_get_reg(child, numReg, &freg);
diff --git a/arch/powerpc/kernel/signal_32.c b/arch/powerpc/kernel/signal_32.c
index 75ee918a120a..73551237de9d 100644
--- a/arch/powerpc/kernel/signal_32.c
+++ b/arch/powerpc/kernel/signal_32.c
@@ -99,8 +99,6 @@ save_general_regs_unsafe(struct pt_regs *regs, struct mcontext __user *frame)
 	elf_greg_t64 *gregs = (elf_greg_t64 *)regs;
 	int val, i;
 
-	WARN_ON(!FULL_REGS(regs));
-
 	for (i = 0; i <= PT_RESULT; i ++) {
 		/* Force usr to alway see softe as 1 (interrupts enabled) */
 		if (i == PT_SOFTE)
@@ -153,7 +151,6 @@ static inline int get_sigset_t(sigset_t *set, const sigset_t __user *uset)
 static __always_inline int
 save_general_regs_unsafe(struct pt_regs *regs, struct mcontext __user *frame)
 {
-	WARN_ON(!FULL_REGS(regs));
 	unsafe_copy_to_user(&frame->mc_gregs, regs, GP_REGS_SIZE, failed);
 	return 0;
 
diff --git a/arch/powerpc/kernel/signal_64.c b/arch/powerpc/kernel/signal_64.c
index f9e4a1ac440f..0e3637722e97 100644
--- a/arch/powerpc/kernel/signal_64.c
+++ b/arch/powerpc/kernel/signal_64.c
@@ -160,7 +160,6 @@ static long setup_sigcontext(struct sigcontext __user *sc,
 	}
 #endif /* CONFIG_VSX */
 	err |= __put_user(&sc->gp_regs, &sc->regs);
-	WARN_ON(!FULL_REGS(regs));
 	err |= __copy_to_user(&sc->gp_regs, regs, GP_REGS_SIZE);
 	err |= __put_user(msr, &sc->gp_regs[PT_MSR]);
 	err |= __put_user(softe, &sc->gp_regs[PT_SOFTE]);
@@ -294,7 +293,6 @@ static long setup_tm_sigcontexts(struct sigcontext __user *sc,
 
 	err |= __put_user(&sc->gp_regs, &sc->regs);
 	err |= __put_user(&tm_sc->gp_regs, &tm_sc->regs);
-	WARN_ON(!FULL_REGS(regs));
 	err |= __copy_to_user(&tm_sc->gp_regs, regs, GP_REGS_SIZE);
 	err |= __copy_to_user(&sc->gp_regs,
 			      &tsk->thread.ckpt_regs, GP_REGS_SIZE);
diff --git a/arch/powerpc/kernel/traps.c b/arch/powerpc/kernel/traps.c
index 97b5f3d83ff7..6c62e4e87979 100644
--- a/arch/powerpc/kernel/traps.c
+++ b/arch/powerpc/kernel/traps.c
@@ -1319,7 +1319,6 @@ static int emulate_instruction(struct pt_regs *regs)
 
 	if (!user_mode(regs))
 		return -EINVAL;
-	CHECK_FULL_REGS(regs);
 
 	if (get_user(instword, (u32 __user *)(regs->nip)))
 		return -EFAULT;
diff --git a/arch/powerpc/lib/sstep.c b/arch/powerpc/lib/sstep.c
index 739ea6dc461c..45bda2520755 100644
--- a/arch/powerpc/lib/sstep.c
+++ b/arch/powerpc/lib/sstep.c
@@ -1401,10 +1401,6 @@ int analyse_instr(struct instruction_op *op, const struct pt_regs *regs,
 		break;
 	}
 
-	/* Following cases refer to regs->gpr[], so we need all regs */
-	if (!FULL_REGS(regs))
-		return -1;
-
 	rd = (word >> 21) & 0x1f;
 	ra = (word >> 16) & 0x1f;
 	rb = (word >> 11) & 0x1f;
diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c
index 3fe37495f63d..42a2c831d87a 100644
--- a/arch/powerpc/xmon/xmon.c
+++ b/arch/powerpc/xmon/xmon.c
@@ -1815,25 +1815,16 @@ static void prregs(struct pt_regs *fp)
 	}
 
 #ifdef CONFIG_PPC64
-	if (FULL_REGS(fp)) {
-		for (n = 0; n < 16; ++n)
-			printf("R%.2d = "REG"   R%.2d = "REG"\n",
-			       n, fp->gpr[n], n+16, fp->gpr[n+16]);
-	} else {
-		for (n = 0; n < 7; ++n)
-			printf("R%.2d = "REG"   R%.2d = "REG"\n",
-			       n, fp->gpr[n], n+7, fp->gpr[n+7]);
-	}
+#define R_PER_LINE 2
 #else
+#define R_PER_LINE 4
+#endif
+
 	for (n = 0; n < 32; ++n) {
-		printf("R%.2d = %.8lx%s", n, fp->gpr[n],
-		       (n & 3) == 3? "\n": "   ");
-		if (n == 12 && !FULL_REGS(fp)) {
-			printf("\n");
-			break;
-		}
+		printf("R%.2d = "REG"%s", n, fp->gpr[n],
+			(n % R_PER_LINE) == R_PER_LINE - 1 ? "\n" : "   ");
 	}
-#endif
+
 	printf("pc  = ");
 	xmon_print_symbol(fp->nip, " ", "\n");
 	if (!trap_is_syscall(fp) && cpu_has_feature(CPU_FTR_CFAR)) {
-- 
2.23.0


^ permalink raw reply related

* [PATCH v2 11/11] powerpc: move norestart trap flag to bit 0
From: Nicholas Piggin @ 2021-03-16 10:42 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Scott Wood, Nicholas Piggin
In-Reply-To: <20210316104206.407354-1-npiggin@gmail.com>

Compact the trap flags down to use the low 4 bits of regs.trap.

A few 64e interrupt trap numbers set bit 4. Although they tended to be
trivial so it wasn't a real problem[1], it is not the right thing to do,
and confusing.

[*] E.g., 0x310 hypercall goes to unknown_exception, which prints
    regs->trap directly so 0x310 will appear fine, and only the syscall
    interrupt will test norestart, so it won't be confused by 0x310.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/include/asm/ptrace.h | 17 +++++++++++------
 1 file changed, 11 insertions(+), 6 deletions(-)

diff --git a/arch/powerpc/include/asm/ptrace.h b/arch/powerpc/include/asm/ptrace.h
index 91194fdd5d01..1e34791b9f80 100644
--- a/arch/powerpc/include/asm/ptrace.h
+++ b/arch/powerpc/include/asm/ptrace.h
@@ -185,20 +185,25 @@ static inline void regs_set_return_value(struct pt_regs *regs, unsigned long rc)
 #define current_pt_regs() \
 	((struct pt_regs *)((unsigned long)task_stack_page(current) + THREAD_SIZE) - 1)
 
+/*
+ * The 4 low bits (0xf) are available as flags to overload the trap word,
+ * because interrupt vectors have minimum alignment of 0x10. TRAP_FLAGS_MASK
+ * must cover the bits used as flags, including bit 0 which is used as the
+ * "norestart" bit.
+ */
 #ifdef __powerpc64__
-#define TRAP_FLAGS_MASK		0x10
-#define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
+#define TRAP_FLAGS_MASK		0x1
 #else
 /*
  * On 4xx we use bit 1 in the trap word to indicate whether the exception
  * is a critical exception (1 means it is).
  */
-#define TRAP_FLAGS_MASK		0x1E
-#define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
+#define TRAP_FLAGS_MASK		0xf
 #define IS_CRITICAL_EXC(regs)	(((regs)->trap & 2) != 0)
 #define IS_MCHECK_EXC(regs)	(((regs)->trap & 4) != 0)
 #define IS_DEBUG_EXC(regs)	(((regs)->trap & 8) != 0)
 #endif /* __powerpc64__ */
+#define TRAP(regs)		((regs)->trap & ~TRAP_FLAGS_MASK)
 
 static inline void set_trap(struct pt_regs *regs, unsigned long val)
 {
@@ -222,12 +227,12 @@ static inline bool trap_is_syscall(struct pt_regs *regs)
 
 static inline bool trap_norestart(struct pt_regs *regs)
 {
-	return regs->trap & 0x10;
+	return regs->trap & 0x1;
 }
 
 static inline void set_trap_norestart(struct pt_regs *regs)
 {
-	regs->trap |= 0x10;
+	regs->trap |= 0x1;
 }
 
 #define arch_has_single_step()	(1)
-- 
2.23.0


^ permalink raw reply related

* Re: [PATCH v2 04/11] powerpc/64e/interrupt: use new interrupt return
From: Christophe Leroy @ 2021-03-16 10:49 UTC (permalink / raw)
  To: Nicholas Piggin, linuxppc-dev; +Cc: Scott Wood
In-Reply-To: <20210316104206.407354-5-npiggin@gmail.com>



Le 16/03/2021 à 11:41, Nicholas Piggin a écrit :
> Update the new C and asm interrupt return code to account for 64e
> specifics, switch over to use it.
> 
> The now-unused old ret_from_except code, that was moved to 64e after the
> 64s conversion, is removed.
> 
> Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
> ---
>   arch/powerpc/include/asm/asm-prototypes.h |   2 -
>   arch/powerpc/include/asm/ppc_asm.h        |  20 --
>   arch/powerpc/kernel/asm-offsets.c         |  10 -
>   arch/powerpc/kernel/exceptions-64e.S      | 321 ++--------------------
>   arch/powerpc/kernel/irq.c                 |  76 -----
>   5 files changed, 25 insertions(+), 404 deletions(-)
> 

> diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
> index 85ba2b0bc8d8..c880ad18b851 100644
> --- a/arch/powerpc/kernel/asm-offsets.c
> +++ b/arch/powerpc/kernel/asm-offsets.c
> @@ -282,21 +282,11 @@ int main(void)
>   	OFFSET(PACAHWCPUID, paca_struct, hw_cpu_id);
>   	OFFSET(PACAKEXECSTATE, paca_struct, kexec_state);
>   	OFFSET(PACA_DSCR_DEFAULT, paca_struct, dscr_default);
> -	OFFSET(ACCOUNT_STARTTIME, paca_struct, accounting.starttime);
> -	OFFSET(ACCOUNT_STARTTIME_USER, paca_struct, accounting.starttime_user);
> -	OFFSET(ACCOUNT_USER_TIME, paca_struct, accounting.utime);
> -	OFFSET(ACCOUNT_SYSTEM_TIME, paca_struct, accounting.stime);
>   #ifdef CONFIG_PPC_BOOK3E
>   	OFFSET(PACA_TRAP_SAVE, paca_struct, trap_save);
>   #endif
>   	OFFSET(PACA_SPRG_VDSO, paca_struct, sprg_vdso);
>   #else /* CONFIG_PPC64 */

The #else is useless

> -#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
> -	OFFSET(ACCOUNT_STARTTIME, thread_info, accounting.starttime);
> -	OFFSET(ACCOUNT_STARTTIME_USER, thread_info, accounting.starttime_user);
> -	OFFSET(ACCOUNT_USER_TIME, thread_info, accounting.utime);
> -	OFFSET(ACCOUNT_SYSTEM_TIME, thread_info, accounting.stime);
> -#endif
>   #endif /* CONFIG_PPC64 */
>   
>   	/* RTAS */

^ permalink raw reply

* [PATCH] powerpc/64s: Fix hash fault to use TRAP accessor
From: Nicholas Piggin @ 2021-03-16 10:52 UTC (permalink / raw)
  To: linuxppc-dev; +Cc: Nicholas Piggin

Hash faults use the trap vector to decide whether this is an
instruction or data fault. This should use the TRAP accessor
rather than open access regs->trap.

This won't cause a problem at the moment because 64s only uses
trap flags for system call interrupts (the norestart flag), but
that could change if any other trap flags get used in future.

Fixes: a4922f5442e7e ("powerpc/64s: move the hash fault handling logic to C")
Suggested-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 arch/powerpc/mm/book3s64/hash_utils.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/arch/powerpc/mm/book3s64/hash_utils.c b/arch/powerpc/mm/book3s64/hash_utils.c
index 1c4b0a29f0f5..977da0dce80c 100644
--- a/arch/powerpc/mm/book3s64/hash_utils.c
+++ b/arch/powerpc/mm/book3s64/hash_utils.c
@@ -1545,10 +1545,10 @@ DEFINE_INTERRUPT_HANDLER_RET(__do_hash_fault)
 	if (user_mode(regs) || (region_id == USER_REGION_ID))
 		access &= ~_PAGE_PRIVILEGED;
 
-	if (regs->trap == 0x400)
+	if (TRAP(regs) == 0x400)
 		access |= _PAGE_EXEC;
 
-	err = hash_page_mm(mm, ea, access, regs->trap, flags);
+	err = hash_page_mm(mm, ea, access, TRAP(regs), flags);
 	if (unlikely(err < 0)) {
 		// failed to instert a hash PTE due to an hypervisor error
 		if (user_mode(regs)) {
-- 
2.23.0


^ permalink raw reply related

* Re: [PATCH] powerpc/numa: Fix topology_physical_package_id() on pSeries
From: Srikar Dronamraju @ 2021-03-16 11:28 UTC (permalink / raw)
  To: C?dric Le Goater
  Cc: Nathan Lynch, Daniel Henrique Barboza, Greg Kurz, Vasant Hegde,
	linuxppc-dev, David Gibson
In-Reply-To: <20210312143154.3181109-1-clg@kaod.org>

* C?dric Le Goater <clg@kaod.org> [2021-03-12 15:31:54]:

> Initial commit 15863ff3b8da ("powerpc: Make chip-id information
> available to userspace") introduce a cpu_to_chip_id() routine for the
> PowerNV platform using the "ibm,chip-id" property to query the chip id
> of a CPU. But PAPR does not specify such a property and the node id
> query is broken.
>
> Use cpu_to_node() instead which guarantees to have a correct value on
> all platforms, PowerNV an pSeries.
>
> Cc: Nathan Lynch <nathanl@linux.ibm.com>
> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
> Cc: Vasant Hegde <hegdevasant@linux.vnet.ibm.com>
> Signed-off-by: Cédric Le Goater <clg@kaod.org>
> ---
>  arch/powerpc/include/asm/topology.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>

(Sorry I somehow managed to mangle to-address. Hence resending this mail
again)

While this looks good to me, @mpe had reservations on using nid as chip-id.
https://lore.kernel.org/linuxppc-dev/87lfwhypv0.fsf@concordia.ellerman.id.au/t/#u
He may be okay with using nid as a "virtual" package id in a pseries
environment.

Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>

[---=| TOFU protection by t-prot: 24 lines snipped |=---]

--
Thanks and Regards
Srikar Dronamraju


> diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
> index 3beeb030cd78..887c42a4e43d 100644
> --- a/arch/powerpc/include/asm/topology.h
> +++ b/arch/powerpc/include/asm/topology.h
> @@ -123,7 +123,7 @@ static inline int cpu_to_coregroup_id(int cpu)
>  #ifdef CONFIG_PPC64
>  #include <asm/smp.h>
>
> -#define topology_physical_package_id(cpu)	(cpu_to_chip_id(cpu))
> +#define topology_physical_package_id(cpu)	(cpu_to_node(cpu))
>
>  #define topology_sibling_cpumask(cpu)	(per_cpu(cpu_sibling_map, cpu))
>  #define topology_core_cpumask(cpu)	(cpu_cpu_mask(cpu))
> --
> 2.26.2
>

--
Thanks and Regards
Srikar Dronamraju

^ permalink raw reply

* [PATCH] powerpc/numa: Fix topology_physical_package_id() on pSeries
From: Cédric Le Goater @ 2021-03-16 12:24 UTC (permalink / raw)
  To: linuxppc-dev
  Cc: Nathan Lynch, Srikar Dronamraju, Daniel Henrique Barboza,
	Greg Kurz, Vasant Hegde, Cédric Le Goater, David Gibson

The topology-id of a CPU in a pSeries machine can be queried from
sysfs but under PowerVM the value is always -1 even if NUMA nodes are
defined. This is because the topology_physical_package_id() routine is
using the "ibm,chip-id" property which is not specified in PAPR.

Under QEMU/KVM, things are different because QEMU populates the CPU DT
node with "ibm,chip-id" property. However, its value can be incorrect
for uncommon SMT configuration and expose a bogus topology-id value in
sysfs.

The use of cpu_to_node() guarantees to have a correct NUMA node id
under both environments QEMU/KVM and PowerVM. This introduces a slight
change for the QEMU/KVM guest, as the topology-id now matches the NUMA
node and not the socket-id as before. Since QEMU also needs to remove
"ibm,chip-id" property for the DT to follow the PAPR specs, both
hypervisor environments will be in sync.

On the PowerNV side, the NUMA node id returned by cpu_to_node() is
computed from the "ibm,associativity" property of the CPU. Its value
is built from the OPAL chip id and is equivalent to "ibm,chip-id".

Cc: Nathan Lynch <nathanl@linux.ibm.com>
Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Vasant Hegde <hegdevasant@linux.vnet.ibm.com>
Reviewed-by: Greg Kurz <groug@kaod.org>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Tested-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Signed-off-by: Cédric Le Goater <clg@kaod.org>
---
 arch/powerpc/include/asm/topology.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
index 3beeb030cd78..887c42a4e43d 100644
--- a/arch/powerpc/include/asm/topology.h
+++ b/arch/powerpc/include/asm/topology.h
@@ -123,7 +123,7 @@ static inline int cpu_to_coregroup_id(int cpu)
 #ifdef CONFIG_PPC64
 #include <asm/smp.h>
 
-#define topology_physical_package_id(cpu)	(cpu_to_chip_id(cpu))
+#define topology_physical_package_id(cpu)	(cpu_to_node(cpu))
 
 #define topology_sibling_cpumask(cpu)	(per_cpu(cpu_sibling_map, cpu))
 #define topology_core_cpumask(cpu)	(cpu_cpu_mask(cpu))
-- 
2.26.2


^ permalink raw reply related

* Re: [PATCH] powerpc/numa: Fix topology_physical_package_id() on pSeries
From: Cédric Le Goater @ 2021-03-16 12:28 UTC (permalink / raw)
  To: Srikar Dronamraju
  Cc: Nathan Lynch, Daniel Henrique Barboza, Greg Kurz, Vasant Hegde,
	linuxppc-dev, David Gibson
In-Reply-To: <20210316112811.GA2339179@linux.vnet.ibm.com>

On 3/16/21 12:28 PM, Srikar Dronamraju wrote:
> * C?dric Le Goater <clg@kaod.org> [2021-03-12 15:31:54]:
> 
>> Initial commit 15863ff3b8da ("powerpc: Make chip-id information
>> available to userspace") introduce a cpu_to_chip_id() routine for the
>> PowerNV platform using the "ibm,chip-id" property to query the chip id
>> of a CPU. But PAPR does not specify such a property and the node id
>> query is broken.
>>
>> Use cpu_to_node() instead which guarantees to have a correct value on
>> all platforms, PowerNV an pSeries.
>>
>> Cc: Nathan Lynch <nathanl@linux.ibm.com>
>> Cc: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
>> Cc: Vasant Hegde <hegdevasant@linux.vnet.ibm.com>
>> Signed-off-by: Cédric Le Goater <clg@kaod.org>
>> ---
>>  arch/powerpc/include/asm/topology.h | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
> 
> (Sorry I somehow managed to mangle to-address. Hence resending this mail
> again)
> 
> While this looks good to me, @mpe had reservations on using nid as chip-id.
> https://lore.kernel.org/linuxppc-dev/87lfwhypv0.fsf@concordia.ellerman.id.au/t/#u

that was a different approach.

> He may be okay with using nid as a "virtual" package id in a pseries
> environment.

ibm,chip-id is only used in a couple of places in low level PowerNV
drivers. The CPU doesn't use this property. We should be fine on
all platforms.

> Reviewed-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>

Thanks,

C.
 
> [---=| TOFU protection by t-prot: 24 lines snipped |=---]
> 
> --
> Thanks and Regards
> Srikar Dronamraju
> 
> 
>> diff --git a/arch/powerpc/include/asm/topology.h b/arch/powerpc/include/asm/topology.h
>> index 3beeb030cd78..887c42a4e43d 100644
>> --- a/arch/powerpc/include/asm/topology.h
>> +++ b/arch/powerpc/include/asm/topology.h
>> @@ -123,7 +123,7 @@ static inline int cpu_to_coregroup_id(int cpu)
>>  #ifdef CONFIG_PPC64
>>  #include <asm/smp.h>
>>
>> -#define topology_physical_package_id(cpu)	(cpu_to_chip_id(cpu))
>> +#define topology_physical_package_id(cpu)	(cpu_to_node(cpu))
>>
>>  #define topology_sibling_cpumask(cpu)	(per_cpu(cpu_sibling_map, cpu))
>>  #define topology_core_cpumask(cpu)	(cpu_cpu_mask(cpu))
>> --
>> 2.26.2
>>
> 
> --
> Thanks and Regards
> Srikar Dronamraju
> 


^ permalink raw reply

* Re: [PATCH 14/17] iommu: remove DOMAIN_ATTR_DMA_USE_FLUSH_QUEUE
From: Robin Murphy @ 2021-03-16 13:03 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: freedreno, kvm, Will Deacon, linux-arm-msm, dri-devel, Li Yang,
	iommu, netdev, virtualization, linuxppc-dev, David Woodhouse,
	linux-arm-kernel
In-Reply-To: <20210315083347.GA28445@lst.de>

On 2021-03-15 08:33, Christoph Hellwig wrote:
> On Fri, Mar 12, 2021 at 04:18:24PM +0000, Robin Murphy wrote:
>>> Let me know what you think of the version here:
>>>
>>> http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/iommu-cleanup
>>>
>>> I'll happily switch the patch to you as the author if you're fine with
>>> that as well.
>>
>> I still have reservations about removing the attribute API entirely and
>> pretending that io_pgtable_cfg is anything other than a SoC-specific
>> private interface,
> 
> I think a private inteface would make more sense.  For now I've just
> condensed it down to a generic set of quirk bits and dropped the
> attrs structure, which seems like an ok middle ground for now.  That
> being said I wonder why that quirk isn't simply set in the device
> tree?

Because it's a software policy decision rather than any inherent 
property of the platform, and the DT certainly doesn't know *when* any 
particular device might prefer its IOMMU to use cacheable pagetables to 
minimise TLB miss latency vs. saving the cache capacity for larger data 
buffers. It really is most logical to decide this at the driver level.

In truth the overall concept *is* relatively generic (a trend towards 
larger system caches and cleverer usage is about both raw performance 
and saving power on off-SoC DRAM traffic), it's just the particular 
implementation of using io-pgtable to set an outer-cacheable walk 
attribute in an SMMU TCR that's pretty much specific to Qualcomm SoCs. 
Hence why having a common abstraction at the iommu_domain level, but 
where the exact details are free to vary across different IOMMUs and 
their respective client drivers, is in many ways an ideal fit.

>> but the reworked patch on its own looks reasonable to
>> me, thanks! (I wasn't too convinced about the iommu_cmd_line wrappers
>> either...) Just iommu_get_dma_strict() needs an export since the SMMU
>> drivers can be modular - I consciously didn't add that myself since I was
>> mistakenly thinking only iommu-dma would call it.
> 
> Fixed.  Can I get your signoff for the patch?  Then I'll switch it to
> over to being attributed to you.

Sure - I would have thought that the one I originally posted still 
stands, but for the avoidance of doubt, for the parts of commit 
8b6d45c495bd in your tree that remain from what I wrote:

Signed-off-by: Robin Murphy <robin.murphy@arm.com>

Cheers,
Robin.

^ permalink raw reply

* Re: [PATCH] ASoC: fsl_sai: remove reset code from dai_probe
From: Mark Brown @ 2021-03-16 12:58 UTC (permalink / raw)
  To: Shengjiu Wang
  Cc: alsa-devel, timur, Xiubo.Lee, linuxppc-dev, tiwai, lgirdwood,
	perex, nicoleotsuka, festevam, linux-kernel
In-Reply-To: <1615886826-30844-1-git-send-email-shengjiu.wang@nxp.com>

[-- Attachment #1: Type: text/plain, Size: 422 bytes --]

On Tue, Mar 16, 2021 at 05:27:06PM +0800, Shengjiu Wang wrote:
> From: Viorel Suman <viorel.suman@nxp.com>
> 
> SAI software reset is done in runtime resume,
> there is no need to do it in fsl_sai_dai_probe.

People can disable runtime PM in their configurations - do you not still
need a reset on probe in case there's no runtime PM?  It'd probably make
sense to factor the rest code out itno a function though.

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 484 bytes --]

^ permalink raw reply


This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox