Re: [RFC 2/2] x86_64: expand kernel stack to 16K

From: Rusty Russell
Date: Wed May 28 2014 - 22:48:59 EST


Minchan Kim <minchan@xxxxxxxxxx> writes:
> On Wed, May 28, 2014 at 12:04:09PM +0300, Michael S. Tsirkin wrote:
>> On Wed, May 28, 2014 at 03:53:59PM +0900, Minchan Kim wrote:
>> > [ 1065.604404] kworker/-5766 0d..2 1071625993us : stack_trace_call: 9) 6456 80 __kmalloc+0x1cb/0x200
>> > [ 1065.604404] kworker/-5766 0d..2 1071625993us : stack_trace_call: 10) 6376 376 vring_add_indirect+0x36/0x200
>> > [ 1065.604404] kworker/-5766 0d..2 1071625993us : stack_trace_call: 11) 6000 144 virtqueue_add_sgs+0x2e2/0x320

Hmm, we can actually skip the vring_add_indirect if we're hurting for
stack. It just means the request will try to fit linearly in the ring,
rather than using indirect.

diff --git a/drivers/virtio/virtio_ring.c b/drivers/virtio/virtio_ring.c
index 1e443629f76d..496e727cefc8 100644
--- a/drivers/virtio/virtio_ring.c
+++ b/drivers/virtio/virtio_ring.c
@@ -184,6 +184,13 @@ static inline int vring_add_indirect(struct vring_virtqueue *vq,
return head;
}

+/* The Morton Technique */
+static noinline bool stack_trouble(void)
+{
+ unsigned long sp = (unsigned long)&sp;
+ return sp - (sp & ~(THREAD_SIZE - 1)) < 3000;
+}
+
static inline int virtqueue_add(struct virtqueue *_vq,
struct scatterlist *sgs[],
struct scatterlist *(*next)
@@ -226,7 +233,7 @@ static inline int virtqueue_add(struct virtqueue *_vq,

/* If the host supports indirect descriptor tables, and we have multiple
* buffers, then go indirect. FIXME: tune this threshold */
- if (vq->indirect && total_sg > 1 && vq->vq.num_free) {
+ if (vq->indirect && total_sg > 1 && vq->vq.num_free && !stack_trouble()) {
head = vring_add_indirect(vq, sgs, next, total_sg, total_out,
total_in,
out_sgs, in_sgs, gfp);
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/