On Sat, 25 Feb 2017, Linus Torvalds wrote: > On Sat, Feb 25, 2017 at 1:07 AM, Ingo Molnar <mingo@xxxxxxxxxx> wrote: > > > > So, should we revert the hw-retrigger change: > > > > a9b4f08770b4 x86/ioapic: Restore IO-APIC irq_chip retrigger callback > > > > ... until we managed to fix CONFIG_DEBUG_SHIRQ=y? If you'd like to revert it > > upstream straight away: > > > > Acked-by: Ingo Molnar <mingo@xxxxxxxxxx> > > So I'm in no huge hurry to revert that commit as long as we're still > in the merge window or early -rc's. > > From a debug standpoint, the spurious early interrupts are fine, and > hopefully will help us find more broken drivers. > > It's just that I'd like to revert it before the actual 4.11 release, > unless we can find a better solution. > > Because it really seems like the interrupt re-trigger is entirely > bogus. It's not an _actual_ "re-trigger the interrupt that may have > gotten lost", it's some code that ends up triggering it for no good > reason. > > So I'd actually hope that we could figure out why IRQS_PENDING got > set, and perhaps fix the underlying cause? > > There are several things that set IRQS_PENDING, ranging from "try to > test mis-routed interrupts while irqd was working", to "prepare for > suspend losing the irq for us", to "irq auto-probing uses it on > unassigned probable irqs". > > The *actual* reason to re-send, namely getting a nested irq that we > had to drop because we got a second one while still handling the first > (or because it was disabled), is just one case. > > Personally, I'd suspect some left-over state from auto-probing earlier > in the boot, but I don't know. Could we fix that underlying issue? I'm on it. Thanks, tglx
