Thread (17 messages) 17 messages, 3 authors, 2015-03-04

[PATCH v2 3/6] irqchip: gicv3-its: limit order of DT size to MAX_ORDER

From: Marc Zyngier <hidden>
Date: 2015-02-17 09:20:21
Also in: lkml

On Sun, 15 Feb 2015 09:32:00 +0000
Yun Wu [off-list ref] wrote:
When required DT size is out of the kmalloc()'s capability, the whole
Nit: Using the DT acronym is very confusing here, as it means "Device
Tree" to most people, including me. Please use "Device Table" instead.

Also, the reference to kmalloc is wrong, as we're really using the page
allocator.
quoted hunk ↗ jump to hunk
ITS will fail in probing. This actually is not the hardware's problem
and is mainly a limitation of the kernel memory allocator. This patch
will keep ITS going on to the next initializaion step with an explicit
warning.

Signed-off-by: Yun Wu <redacted>
---
 drivers/irqchip/irq-gic-v3-its.c | 6 ++++++
 1 file changed, 6 insertions(+)
diff --git a/drivers/irqchip/irq-gic-v3-its.c
b/drivers/irqchip/irq-gic-v3-its.c index f5bfa42..de36606 100644
--- a/drivers/irqchip/irq-gic-v3-its.c
+++ b/drivers/irqchip/irq-gic-v3-its.c
@@ -828,6 +828,12 @@ static int its_alloc_tables(struct its_node *its)
 			u32 ids = GITS_TYPER_DEVBITS(typer);

 			order = get_order((1UL << ids) * entry_size);
+			if (order >= MAX_ORDER) {
+				pr_warn("%s: DT size too large,
reduce to %u pages\n",
+
its->msi_chip.of_node->full_name,
+					1 << order);
+				order = MAX_ORDER;
+			}
 		}
Something is wrong here. Either (order == MAX_ORDER) is acceptable, or
it is not. Also, your warning says that you're reducing the size to a
new value, but you're displaying the size you can't handle. That
message should be fixed.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help