Merge tag 'powerpc-4.2-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mpe/linux

Pull powerpc updates from Michael Ellerman:

 - disable the 32-bit vdso when building LE, so we can build with a
   64-bit only toolchain.

 - EEH fixes from Gavin & Richard.

 - enable the sys_kcmp syscall from Laurent.

 - sysfs control for fastsleep workaround from Shreyas.

 - expose OPAL events as an irq chip by Alistair.

 - MSI ops moved to pci_controller_ops by Daniel.

 - fix for kernel to userspace backtraces for perf from Anton.

 - merge pseries and pseries_le defconfigs from Cyril.

 - CXL in-kernel API from Mikey.

 - OPAL prd driver from Jeremy.

 - fix for DSCR handling & tests from Anshuman.

 - Powernv flash mtd driver from Cyril.

 - dynamic DMA Window support on powernv from Alexey.

 - LLVM clang fixes & workarounds from Anton.

 - reworked version of the patch to abort syscalls when transactional.

 - fix the swap encoding to support 4TB, from Aneesh.

 - various fixes as usual.

 - Freescale updates from Scott: Highlights include more 8xx
   optimizations, an e6500 hugetlb optimization, QMan device tree nodes,
   t1024/t1023 support, and various fixes and cleanup.

* tag 'powerpc-4.2-1' of git://git.kernel.org/pub/scm/linux/kernel/git/mpe/linux: (180 commits)
  cxl: Fix typo in debug print
  cxl: Add CXL_KERNEL_API config option
  powerpc/powernv: Fix wrong IOMMU table in pnv_ioda_setup_bus_dma()
  powerpc/mm: Change the swap encoding in pte.
  powerpc/mm: PTE_RPN_MAX is not used, remove the same
  powerpc/tm: Abort syscalls in active transactions
  powerpc/iommu/ioda2: Enable compile with IOV=on and IOMMU_API=off
  powerpc/include: Add opal-prd to installed uapi headers
  powerpc/powernv: fix construction of opal PRD messages
  powerpc/powernv: Increase opal-irqchip initcall priority
  powerpc: Make doorbell check preemption safe
  powerpc/powernv: pnv_init_idle_states() should only run on powernv
  macintosh/nvram: Remove as unused
  powerpc: Don't use gcc specific options on clang
  powerpc: Don't use -mno-strict-align on clang
  powerpc: Only use -mtraceback=no, -mno-string and -msoft-float if toolchain supports it
  powerpc: Only use -mabi=altivec if toolchain supports it
  powerpc: Fix duplicate const clang warning in user access code
  vfio: powerpc/spapr: Support Dynamic DMA windows
  vfio: powerpc/spapr: Register memory and define IOMMU v2
  ...

1 files changed, 33 insertions, 29 deletions

diff --git a/arch/powerpc/platforms/powernv/eeh-powernv.c b/arch/powerpc/platforms/powernv/eeh-powernv.c
index ce738ab..5cf5e6e 100644
--- a/arch/powerpc/platforms/powernv/eeh-powernv.c
+++ b/arch/powerpc/platforms/powernv/eeh-powernv.c
@@ -16,6 +16,7 @@
 #include <linux/delay.h>
 #include <linux/export.h>
 #include <linux/init.h>
+#include <linux/interrupt.h>
 #include <linux/list.h>
 #include <linux/msi.h>
 #include <linux/of.h>
@@ -40,6 +41,7 @@
 #include "pci.h"
 
 static bool pnv_eeh_nb_init = false;
+static int eeh_event_irq = -EINVAL;
 
 /**
  * pnv_eeh_init - EEH platform dependent initialization
@@ -88,34 +90,22 @@ static int pnv_eeh_init(void)
 	return 0;
 }
 
-static int pnv_eeh_event(struct notifier_block *nb,
-			 unsigned long events, void *change)
+static irqreturn_t pnv_eeh_event(int irq, void *data)
 {
-	uint64_t changed_evts = (uint64_t)change;
-
 	/*
-	 * We simply send special EEH event if EEH has
-	 * been enabled, or clear pending events in
-	 * case that we enable EEH soon
+	 * We simply send a special EEH event if EEH has been
+	 * enabled. We don't care about EEH events until we've
+	 * finished processing the outstanding ones. Event processing
+	 * gets unmasked in next_error() if EEH is enabled.
 	 */
-	if (!(changed_evts & OPAL_EVENT_PCI_ERROR) ||
-	    !(events & OPAL_EVENT_PCI_ERROR))
-		return 0;
+	disable_irq_nosync(irq);
 
 	if (eeh_enabled())
 		eeh_send_failure_event(NULL);
-	else
-		opal_notifier_update_evt(OPAL_EVENT_PCI_ERROR, 0x0ul);
 
-	return 0;
+	return IRQ_HANDLED;
 }
 
-static struct notifier_block pnv_eeh_nb = {
-	.notifier_call	= pnv_eeh_event,
-	.next		= NULL,
-	.priority	= 0
-};
-
 #ifdef CONFIG_DEBUG_FS
 static ssize_t pnv_eeh_ei_write(struct file *filp,
 				const char __user *user_buf,
@@ -237,16 +227,28 @@ static int pnv_eeh_post_init(void)
 
 	/* Register OPAL event notifier */
 	if (!pnv_eeh_nb_init) {
-		ret = opal_notifier_register(&pnv_eeh_nb);
-		if (ret) {
-			pr_warn("%s: Can't register OPAL event notifier (%d)\n",
-				__func__, ret);
+		eeh_event_irq = opal_event_request(ilog2(OPAL_EVENT_PCI_ERROR));
+		if (eeh_event_irq < 0) {
+			pr_err("%s: Can't register OPAL event interrupt (%d)\n",
+			       __func__, eeh_event_irq);
+			return eeh_event_irq;
+		}
+
+		ret = request_irq(eeh_event_irq, pnv_eeh_event,
+				IRQ_TYPE_LEVEL_HIGH, "opal-eeh", NULL);
+		if (ret < 0) {
+			irq_dispose_mapping(eeh_event_irq);
+			pr_err("%s: Can't request OPAL event interrupt (%d)\n",
+			       __func__, eeh_event_irq);
 			return ret;
 		}
 
 		pnv_eeh_nb_init = true;
 	}
 
+	if (!eeh_enabled())
+		disable_irq(eeh_event_irq);
+
 	list_for_each_entry(hose, &hose_list, list_node) {
 		phb = hose->private_data;
 
@@ -979,7 +981,7 @@ static int pnv_eeh_reset(struct eeh_pe *pe, int option)
 /**
  * pnv_eeh_wait_state - Wait for PE state
  * @pe: EEH PE
- * @max_wait: maximal period in microsecond
+ * @max_wait: maximal period in millisecond
  *
  * Wait for the state of associated PE. It might take some time
  * to retrieve the PE's state.
@@ -1000,13 +1002,13 @@ static int pnv_eeh_wait_state(struct eeh_pe *pe, int max_wait)
 		if (ret != EEH_STATE_UNAVAILABLE)
 			return ret;
 
-		max_wait -= mwait;
 		if (max_wait <= 0) {
 			pr_warn("%s: Timeout getting PE#%x's state (%d)\n",
 				__func__, pe->addr, max_wait);
 			return EEH_STATE_NOT_SUPPORT;
 		}
 
+		max_wait -= mwait;
 		msleep(mwait);
 	}
 
@@ -1303,12 +1305,10 @@ static int pnv_eeh_next_error(struct eeh_pe **pe)
 	int state, ret = EEH_NEXT_ERR_NONE;
 
 	/*
-	 * While running here, it's safe to purge the event queue.
-	 * And we should keep the cached OPAL notifier event sychronized
-	 * between the kernel and firmware.
+	 * While running here, it's safe to purge the event queue. The
+	 * event should still be masked.
 	 */
 	eeh_remove_event(NULL, false);
-	opal_notifier_update_evt(OPAL_EVENT_PCI_ERROR, 0x0ul);
 
 	list_for_each_entry(hose, &hose_list, list_node) {
 		/*
@@ -1477,6 +1477,10 @@ static int pnv_eeh_next_error(struct eeh_pe **pe)
 			break;
 	}
 
+	/* Unmask the event */
+	if (eeh_enabled())
+		enable_irq(eeh_event_irq);
+
 	return ret;
 }