Thread (12 messages) 12 messages, 2 authors, 2015-03-06

[PATCH v3 5/5] irqchip: gicv3-its: support safe initialization

From: Yun Wu Abel <hidden>
Date: 2015-03-06 01:35:25
Also in: lkml

On 2015/3/5 20:05, Marc Zyngier wrote:
On 04/03/15 03:18, Yun Wu wrote:
quoted
It's unsafe to change the configurations of an activated ITS directly
since this will lead to unpredictable results. This patch guarantees
the ITSes being initialized are quiescent.

Signed-off-by: Yun Wu <redacted>
---
 drivers/irqchip/irq-gic-v3-its.c | 35 +++++++++++++++++++++++++++++++++++
 1 file changed, 35 insertions(+)
diff --git a/drivers/irqchip/irq-gic-v3-its.c b/drivers/irqchip/irq-gic-v3-its.c
index d13c24e..9e09aa0 100644
--- a/drivers/irqchip/irq-gic-v3-its.c
+++ b/drivers/irqchip/irq-gic-v3-its.c
@@ -1320,6 +1320,34 @@ static const struct irq_domain_ops its_domain_ops = {
 	.deactivate		= its_irq_domain_deactivate,
 };

+static int its_check_quiesced(void __iomem *base)
Another nitpick: Rather than "its_check_quiesced", how about
"its_force_quiescent" instead? Because this does a lot more than just
checking.
Yes, indeed.
quoted
+{
+	u32 count = 1000000;	/* 1s */
+	u32 val;
+
+	val = readl_relaxed(base + GITS_CTLR);
+	if (val & GITS_CTLR_QUIESCENT)
+		return 0;
+
+	/* Disable the generation of all interrupts to this ITS */
+	val &= ~GITS_CTLR_ENABLE;
+	writel_relaxed(val, base + GITS_CTLR);
+
+	/* Poll GITS_CTLR and wait until ITS becomes quiescent */
+	while (1) {
+		val = readl_relaxed(base + GITS_CTLR);
+		if (val & GITS_CTLR_QUIESCENT)
+			return 0;
+
+		count--;
+		if (!count)
+			return -EBUSY;
+
+		cpu_relax();
+		udelay(1);
+	}
+}
+
I still dislike this repeated pattern, but I don't have a good solution
so far.
Me too.
quoted
 static int its_probe(struct device_node *node, struct irq_domain *parent)
 {
 	struct resource res;
@@ -1348,6 +1376,13 @@ static int its_probe(struct device_node *node, struct irq_domain *parent)
 		goto out_unmap;
 	}

+	err = its_check_quiesced(its_base);
+	if (err) {
+		pr_warn("%s: failed to quiesce, giving up\n",
+			node->full_name);
+		goto out_unmap;
+	}
+
 	pr_info("ITS: %s\n", node->full_name);

 	its = kzalloc(sizeof(*its), GFP_KERNEL);
--
1.8.0

Assuming you fix the above nitpick:

Acked-by: Marc Zyngier <redacted>
Thanks,
	Abel
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help