[RFC PATCH 3/3] cpuidle/powernv: Conditionally save-restore sprs using opal

From: Akshay Adiga
Date: Thu Aug 02 2018 - 00:52:56 EST

From: Abhishek Goel <huntbag@xxxxxxxxxxxxxxxxxx>

If a state has "opal-supported" compat flag in device-tree, an opal call
needs to be made during the entry and exit of the stop state. This patch
passes a hint to the power9_idle_stop and power9_offline_stop.

This patch moves the saving and restoring of sprs for P9 cpuidle
from kernel to opal. This patch still uses existing code to detect
first thread in core.
In an attempt to make the powernv idle code backward compatible,
and to some extent forward compatible, add support for pre-stop entry
and post-stop exit actions in OPAL. If a kernel knows about this
opal call, then just a firmware supporting newer hardware is required,
instead of waiting for kernel updates.

Signed-off-by: Abhishek Goel <huntbag@xxxxxxxxxxxxxxxxxx>
Signed-off-by: Akshay Adiga <akshay.adiga@xxxxxxxxxxxxxxxxxx>
---
arch/powerpc/include/asm/cpuidle.h | 1 +
arch/powerpc/include/asm/opal-api.h | 4 +-
arch/powerpc/include/asm/opal.h | 3 +
arch/powerpc/include/asm/paca.h | 5 +-
arch/powerpc/include/asm/processor.h | 6 +-
arch/powerpc/kernel/asm-offsets.c | 3 +
arch/powerpc/kernel/idle_book3s.S | 55 ++++++++++++++++++-
arch/powerpc/platforms/powernv/idle.c | 41 ++++++++++----
.../powerpc/platforms/powernv/opal-wrappers.S | 2 +
arch/powerpc/xmon/xmon.c | 14 ++---
10 files changed, 109 insertions(+), 25 deletions(-)

diff --git a/arch/powerpc/include/asm/cpuidle.h b/arch/powerpc/include/asm/cpuidle.h
index b965066560cc..2fb2324d15fc 100644
--- a/arch/powerpc/include/asm/cpuidle.h
+++ b/arch/powerpc/include/asm/cpuidle.h
@@ -96,6 +96,7 @@ struct pnv_idle_states_t {
u64 psscr_val;
u64 psscr_mask;
u32 flags;
+ bool req_opal_call;
enum idle_state_type_t type;
bool valid;
};
diff --git a/arch/powerpc/include/asm/opal-api.h b/arch/powerpc/include/asm/opal-api.h
index 3bab299eda49..6792a737bc9a 100644
--- a/arch/powerpc/include/asm/opal-api.h
+++ b/arch/powerpc/include/asm/opal-api.h
@@ -208,7 +208,9 @@
#define OPAL_SENSOR_READ_U64 162
#define OPAL_PCI_GET_PBCQ_TUNNEL_BAR 164
#define OPAL_PCI_SET_PBCQ_TUNNEL_BAR 165
-#define OPAL_LAST 165
+#define OPAL_IDLE_SAVE 168
+#define OPAL_IDLE_RESTORE 169
+#define OPAL_LAST 169

#define QUIESCE_HOLD 1 /* Spin all calls at entry */
#define QUIESCE_REJECT 2 /* Fail all calls with OPAL_BUSY */
diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
index e1b2910c6e81..12d57aeacde2 100644
--- a/arch/powerpc/include/asm/opal.h
+++ b/arch/powerpc/include/asm/opal.h
@@ -356,6 +356,9 @@ extern void opal_kmsg_init(void);

extern int opal_event_request(unsigned int opal_event_nr);

+extern int opal_cpuidle_save(u64 *stop_sprs, int scope, u64 psscr);
+extern int opal_cpuidle_restore(u64 *stop_sprs, int scope, u64 psscr, u64 srr1);
+
struct opal_sg_list *opal_vmalloc_to_sg_list(void *vmalloc_addr,
unsigned long vmalloc_size);
void opal_free_sg_list(struct opal_sg_list *sg);
diff --git a/arch/powerpc/include/asm/paca.h b/arch/powerpc/include/asm/paca.h
index 6d34bd71139d..586059594443 100644
--- a/arch/powerpc/include/asm/paca.h
+++ b/arch/powerpc/include/asm/paca.h
@@ -195,11 +195,14 @@ struct paca_struct {
/* The PSSCR value that the kernel requested before going to stop */
u64 requested_psscr;

+ u64 wakeup_psscr;
+ u8 req_opal_call;
/*
- * Save area for additional SPRs that need to be
+ * Save area for SPRs that need to be
* saved/restored during cpuidle stop.
*/
struct stop_sprs stop_sprs;
+ u64 *opal_stop_sprs;
#endif

#ifdef CONFIG_PPC_BOOK3S_64
diff --git a/arch/powerpc/include/asm/processor.h b/arch/powerpc/include/asm/processor.h
index 34f572056add..9f9fb1f11dd6 100644
--- a/arch/powerpc/include/asm/processor.h
+++ b/arch/powerpc/include/asm/processor.h
@@ -513,8 +513,10 @@ enum idle_boot_override {IDLE_NO_OVERRIDE = 0, IDLE_POWERSAVE_OFF};
extern int powersave_nap; /* set if nap mode can be used in idle loop */
extern unsigned long power7_idle_insn(unsigned long type); /* PNV_THREAD_NAP/etc*/
extern void power7_idle_type(unsigned long type);
-extern unsigned long power9_idle_stop(unsigned long psscr_val);
-extern unsigned long power9_offline_stop(unsigned long psscr_val);
+extern unsigned long power9_idle_stop(unsigned long psscr_val,
+ bool opal_enabled);
+extern unsigned long power9_offline_stop(unsigned long psscr_val,
+ bool opal_enabled);
extern void power9_idle_type(int index);

extern void flush_instruction_cache(void);
diff --git a/arch/powerpc/kernel/asm-offsets.c b/arch/powerpc/kernel/asm-offsets.c
index 262c44a90ea1..740ae068ec74 100644
--- a/arch/powerpc/kernel/asm-offsets.c
+++ b/arch/powerpc/kernel/asm-offsets.c
@@ -768,6 +768,9 @@ int main(void)
OFFSET(PACA_SIBLING_PACA_PTRS, paca_struct, thread_sibling_pacas);
OFFSET(PACA_REQ_PSSCR, paca_struct, requested_psscr);
OFFSET(PACA_DONT_STOP, paca_struct, dont_stop);
+ OFFSET(PACA_WAKEUP_PSSCR, paca_struct, wakeup_psscr);
+ OFFSET(PACA_REQ_OPAL_CALL, paca_struct, req_opal_call);
+ OFFSET(STOP_SPRS, paca_struct, opal_stop_sprs);
#define STOP_SPR(x, f) OFFSET(x, paca_struct, stop_sprs.f)
STOP_SPR(STOP_PID, pid);
STOP_SPR(STOP_LDBAR, ldbar);
diff --git a/arch/powerpc/kernel/idle_book3s.S b/arch/powerpc/kernel/idle_book3s.S
index e734f6e45abc..b5f5245011a5 100644
--- a/arch/powerpc/kernel/idle_book3s.S
+++ b/arch/powerpc/kernel/idle_book3s.S
@@ -45,6 +45,9 @@

#define PSSCR_EC_ESL_MASK_SHIFTED (PSSCR_EC | PSSCR_ESL) >> 16

+#define SCOPE_CORE 0
+#define SCOPE_THREAD 1
+
.text

/*
@@ -388,7 +391,18 @@ lwarx_loop_stop:
bne- lwarx_loop_stop
isync

+ ld r6,PACA_REQ_OPAL_CALL(r13)
+ cmpwi r6,1
+ beq opal_save
bl save_sprs_to_stack
+ PPC_STOP
+ b .
+
+opal_save:
+ ld r3,STOP_SPRS(r13)
+ li r4,SCOPE_CORE
+ ld r5,PACA_REQ_PSSCR(r13)
+ bl opal_cpuidle_save

PPC_STOP /* Does not return (system reset interrupt) */

@@ -435,13 +449,14 @@ _GLOBAL(power9_offline_stop)
* between threads, but in that case KVM has a barrier sync in real
* mode before and after switching between radix and hash.
*/
- li r4,KVM_HWTHREAD_IN_IDLE
- stb r4,HSTATE_HWTHREAD_STATE(r13)
+ li r5,KVM_HWTHREAD_IN_IDLE
+ stb r5,HSTATE_HWTHREAD_STATE(r13)
#endif
/* fall through */

_GLOBAL(power9_idle_stop)
std r3, PACA_REQ_PSSCR(r13)
+ std r4, PACA_REQ_OPAL_CALL(r13)
#ifdef CONFIG_KVM_BOOK3S_HV_POSSIBLE
BEGIN_FTR_SECTION
sync
@@ -566,6 +581,8 @@ ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_300)
#endif

/* Return SRR1 from power7_nap() */
+ rlwinm r11,r12,47-31,30,31
+ cmpwi cr3,r11,2
blt cr3,pnv_wakeup_noloss
b pnv_wakeup_loss

@@ -576,6 +593,8 @@ ALT_FTR_SECTION_END_IFSET(CPU_FTR_ARCH_300)
* cr3 - set to gt if waking up with partial/complete hypervisor state loss
*/
pnv_restore_hyp_resource_arch300:
+ mfspr r5, SPRN_PSSCR
+ std r5, PACA_WAKEUP_PSSCR(r13)
/*
* Workaround for POWER9, if we lost resources, the ERAT
* might have been mixed up and needs flushing. We also need
@@ -831,7 +850,24 @@ subcore_state_restored:
bne cr2,clear_lock

first_thread_in_core:
+ ld r6,PACA_REQ_OPAL_CALL(r13)
+ cmpwi r6,1
+ beq opal_restore_core
+ b kernel_restore_core
+
+opal_restore_core:
+ ld r3,STOP_SPRS(r13)
+ li r4,SCOPE_CORE
+ ld r5,PACA_WAKEUP_PSSCR(r13)
+ mr r6,r19 /*r19 contains SRR1*/
+ bl opal_cpuidle_restore
+ ld r1,PACAR1(r13)
+ xoris r15,r15,PNV_CORE_IDLE_LOCK_BIT@h
+ lwsync
+ stw r15,0(r14)
+ b hypervisor_state_restored

+kernel_restore_core:
/*
* First thread in the core waking up from any state which can cause
* partial or complete hypervisor state loss. It needs to
@@ -885,6 +921,21 @@ clear_lock:
stw r15,0(r14)

common_exit:
+ ld r6,PACA_REQ_OPAL_CALL(r13)
+ cmpwi r6,1
+ beq opal_restore_thread
+ b kernel_restore_thread
+
+opal_restore_thread:
+ ld r3,STOP_SPRS(r13)
+ li r4,SCOPE_THREAD
+ ld r5,PACA_WAKEUP_PSSCR(r13)
+ mr r6,r19 /*r19 contains SRR1*/
+ bl opal_cpuidle_restore
+ ld r1,PACAR1(r13)
+ b hypervisor_state_restored
+
+kernel_restore_thread:
/*
* Common to all threads.
*
diff --git a/arch/powerpc/platforms/powernv/idle.c b/arch/powerpc/platforms/powernv/idle.c
index a6ef9b68e27b..e5d38524aec3 100644
--- a/arch/powerpc/platforms/powernv/idle.c
+++ b/arch/powerpc/platforms/powernv/idle.c
@@ -147,6 +147,7 @@ static int pnv_save_sprs_for_deep_states(void)
return 0;
}

+#define MAX_STOP_SPRS_COUNT 25
static void pnv_alloc_idle_core_states(void)
{
int i, j;
@@ -186,6 +187,9 @@ static void pnv_alloc_idle_core_states(void)
for (j = 0; j < threads_per_core; j++) {
int cpu = first_cpu + j;

+ paca_ptrs[cpu]->opal_stop_sprs = kmalloc_node(
+ MAX_STOP_SPRS_COUNT * sizeof(u64),
+ GFP_KERNEL, node);
paca_ptrs[cpu]->core_idle_state_ptr = core_idle_state;
paca_ptrs[cpu]->thread_idle_state = PNV_THREAD_RUNNING;
paca_ptrs[cpu]->thread_mask = 1 << j;
@@ -372,7 +376,7 @@ static unsigned long __power9_idle_type(struct pnv_idle_states_t *state)
psscr = mfspr(SPRN_PSSCR);
psscr = (psscr & ~stop_psscr_mask) | stop_psscr_val;
__ppc64_runlatch_off();
- srr1 = power9_idle_stop(psscr, state->opal_supported);
+ srr1 = power9_idle_stop(psscr, state->req_opal_call);
__ppc64_runlatch_on();

fini_irq_for_idle_irqsoff();
@@ -518,7 +522,7 @@ unsigned long pnv_cpu_offline(unsigned int cpu)

psscr = mfspr(SPRN_PSSCR);
psscr = (psscr & ~state->psscr_mask) | state->psscr_val;
- srr1 = power9_offline_stop(psscr, state->opal_supported);
+ srr1 = power9_offline_stop(psscr, state->req_opal_call);

} else if ((idle_states & OPAL_PM_WINKLE_ENABLED) &&
(idle_states & OPAL_PM_LOSE_FULL_CONTEXT)) {
@@ -815,6 +819,7 @@ static int pnv_parse_cpuidle_dt(void)
u32 *temp_u32;
u64 *temp_u64;
const char **temp_string;
+ bool fall_back_to_opal = false;

np = of_find_node_by_path("/ibm,opal/power-mgt");
if (!np) {
@@ -929,21 +934,33 @@ static int pnv_parse_cpuidle_dt(void)
/* Parse each child node with appropriate parser_fn */
for_each_child_of_node(np1, dt_node) {
bool found_known_version = false;
- /* we don't have state falling back to opal*/
- for (i = 0; i < nr_known_versions ; i++) {
- if (of_device_is_compatible(dt_node, known_versions[i].name)) {
- rc = known_versions[i].parser_fn(dt_node);
- if (rc) {
- pr_err("%s could not parse\n",known_versions[i].name);
- continue;
+ int idx = nr_pnv_idle_states;
+ if (!fall_back_to_opal) {
+ /* we don't have state falling back to opal*/
+ for (i = 0; i < nr_known_versions ; i++) {
+ if (of_device_is_compatible(dt_node, known_versions[i].name)) {
+ rc = known_versions[i].parser_fn(dt_node);
+ if (rc) {
+ pr_err("%s could not parse\n",known_versions[i].name);
+ continue;
+ }
+ found_known_version = true;
}
- found_known_version = true;
}
}
-
- if (!found_known_version) {
+ if (!found_known_version || fall_back_to_opal) {
+ if (of_device_is_compatible(dt_node, "opal-support")) {
+ rc = known_versions[0].parser_fn(dt_node);
+ if (rc) {
+ pr_err("%s could not parse\n", "opal-support");
+ continue;
+ }
+ pnv_idle_states[idx].req_opal_call = true;
+ fall_back_to_opal = true;
+ } else {
pr_info("Unsupported state, skipping all further state\n");
goto out;
+ }
}
nr_pnv_idle_states++;
}
diff --git a/arch/powerpc/platforms/powernv/opal-wrappers.S b/arch/powerpc/platforms/powernv/opal-wrappers.S
index a8d9b4089c31..b75c37d93efd 100644
--- a/arch/powerpc/platforms/powernv/opal-wrappers.S
+++ b/arch/powerpc/platforms/powernv/opal-wrappers.S
@@ -327,3 +327,5 @@ OPAL_CALL(opal_npu_tl_set, OPAL_NPU_TL_SET);
OPAL_CALL(opal_pci_get_pbcq_tunnel_bar, OPAL_PCI_GET_PBCQ_TUNNEL_BAR);
OPAL_CALL(opal_pci_set_pbcq_tunnel_bar, OPAL_PCI_SET_PBCQ_TUNNEL_BAR);
OPAL_CALL(opal_sensor_read_u64, OPAL_SENSOR_READ_U64);
+OPAL_CALL(opal_cpuidle_save, OPAL_IDLE_SAVE);
+OPAL_CALL(opal_cpuidle_restore, OPAL_IDLE_RESTORE);
diff --git a/arch/powerpc/xmon/xmon.c b/arch/powerpc/xmon/xmon.c
index 47166ad2a669..8ae34f37e60a 100644
--- a/arch/powerpc/xmon/xmon.c
+++ b/arch/powerpc/xmon/xmon.c
@@ -2431,14 +2431,14 @@ static void dump_one_paca(int cpu)
DUMP(p, subcore_sibling_mask, "%#-*x");
DUMP(p, thread_sibling_pacas, "%-*px");
DUMP(p, requested_psscr, "%#-*llx");
- DUMP(p, stop_sprs.pid, "%#-*llx");
- DUMP(p, stop_sprs.ldbar, "%#-*llx");
- DUMP(p, stop_sprs.fscr, "%#-*llx");
- DUMP(p, stop_sprs.hfscr, "%#-*llx");
- DUMP(p, stop_sprs.mmcr1, "%#-*llx");
- DUMP(p, stop_sprs.mmcr2, "%#-*llx");
- DUMP(p, stop_sprs.mmcra, "%#-*llx");
DUMP(p, dont_stop.counter, "%#-*x");
+
+ /*
+ * TODO Either kernel or opal has sprs stored. If opal stored it,
+ * we can find a way to make the indices available to kernel through
+ * paca.
+ */
+
#endif

DUMP(p, accounting.utime, "%#-*lx");
--
2.18.0.rc2.85.g1fb9df7