On Thu, Jul 30, 2020 at 06:55:18PM +0200, Greg Kurz wrote: > On Wed, 29 Jul 2020 12:54:41 +1000 > David Gibson wrote: > > > On Tue, Jul 28, 2020 at 11:14:13AM +0200, Greg Kurz wrote: > > > Without this patch, the irq number gets converted uselessly from int > > > to int32_t, back and forth. > > > > > > This doesn't fix an actual issue, it's just to make the code neater. > > > > > > Suggested-by: Markus Armbruster > > > Signed-off-by: Greg Kurz > > > > Applied to ppc-for-5.2, thanks. > > > > Daniel reported a crash that happens systematically on some systems that > don't support KVM XIVE (aka. bostons) since the patch "spapr: Simplify > error handling in spapr_phb_realize()" landed in the ppc-for-5.2 tree. > > The patch is good but it uncovered an issue we have in the KVM XIVE code > in QEMU (basically we should ignore the absence of KVM XIVE device when > claiming IRQ numbers). > > The fix is trivial but to avoid breaking bisect, it should rather go > before the patch mentioned above. Also I want to consolidate the error > handling a bit more so, in the meantime, for others to be able to use > the ppc-for-5.2 branch, I suggest you simply drop: > > spapr: Simplify error handling in spapr_phb_realize() > > and the current patch as well since it's a follow-up. > > I'll send a new patchset later. Ok, done, I've removed both those patches from ppc-for-5.2, resend the new version whenever you're ready. -- David Gibson | I'll have my music baroque, and my code david AT gibson.dropbear.id.au | minimalist, thank you. NOT _the_ _other_ | _way_ _around_! http://www.ozlabs.org/~dgibson