[PATCH v3 03/14] KVM: selftests: Return a value from vcpu_get_reg() instead of using an out-param

Sean Christopherson seanjc at google.com
Fri Nov 1 09:31:05 PDT 2024


On Fri, Nov 01, 2024, Oliver Upton wrote:
> On Fri, Nov 01, 2024 at 09:16:42AM -0700, Sean Christopherson wrote:
> > One thing I'll add to my workflow would be to do a local merge (and smoke test)
> > of linux-next into kvm-x86 next before pushing it out.  This isn't the only snafu
> > this cycle where such a sanity check would have saved me and others a bit of pain.
> 
> Eh, shit happens, that's what -next is for :)

Heh, but I also don't actually test -next, which was another snafu (not my fault
this time!) from this cycle[*].  Testing 6.12-next prior to the merge window
wouldn't have made that any less painful to bisect, but I think it would at least
have allowed me to detect that the issue specifically came in from linux-next,
and the bug report would have gotten to PeterZ almost two months earlier.

https://lore.kernel.org/all/ZwdA0sbA2tJA3IKh@google.com



More information about the kvm-riscv mailing list