Re: BUG: soft lockup in ieee80211_tasklet_handler

On Wed, Mar 3, 2021 at 9:59 AM Hillf Danton [off-list ref] wrote:

On Tue, 02 Mar 2021 15:18:16 +0100  Johannes Berg wrote:

quoted

On Wed, 2021-02-24 at 10:30 +0800, Hillf Danton wrote:

quoted

Add budget for the 80211 softint handler - it's feasible not to try to
build the giant pyramid in a week.

--- x/net/mac80211/main.c
+++ y/net/mac80211/main.c

@@ -224,9 +224,15 @@ static void ieee80211_tasklet_handler(un
 {
    struct ieee80211_local *local = (struct ieee80211_local *) data;
    struct sk_buff *skb;
+   int i = 0;
+
+   while (i++ < 64) {
+           skb = skb_dequeue(&local->skb_queue);
+           if (!skb)
+                   skb = skb_dequeue(&local->skb_queue_unreliable);
+           if (!skb)
+                   return;

I guess that's not such a bad idea, but I do wonder how we get here,
userspace can submit packets faster than we can process?

I wonder why syzbot did not make other handlers stand out than
ieee80211_tasklet_handler.

syzbot has no relation to this whatsoever. It's just a proxy between
the kernel and you. Ask the kernel ;)

quoted

It feels like a simulation-only case, tbh, since over the air you have
limits how much bandwidth you can get ... unless you have a very slow
CPU?

Even with a slower CPU I want to run a FIFO task every tick - it can bear
latencies like two seconds.

quoted

In any case, if you want anything merged you're going to have to submit
a proper patch with a real commit message and Signed-off-by, etc.

johannes