Re: PCU panic on V240
Matt Thomas <[email protected]>
| Newsgroups | gmane.os.netbsd.ports.sparc64 |
|---|---|
| Message-ID | <[email protected]> |
On Dec 5, 2013, at 7:57 AM, Martin Husemann <[email protected]> wrote: > On Thu, Dec 05, 2013 at 07:24:12AM -0800, Matt Thomas wrote: >> That almost makes sense since SOFTINT_SERIAL can preempt >> SOFTINT_CLOCK/SOFTINT_NET but servicing softints shouldn't >> increase interrupt depth. > > Is cpu_intr_p() documented anywhere? > I don't think we distinguish soft and hard ints for ci_idepths. cpu_softintr_p() is for softintrs cpu_intr_p() is for hardintrs.