RFR(S): 8207844: [lworld] Crash during method resolution because value type holder is not linked

Tobias Hartmann tobias.hartmann at oracle.com
Fri Jul 20 09:15:04 UTC 2018

Hi John,

thanks for the review.

On 19.07.2018 22:39, John Rose wrote:
> Fix is good, but seems a little more cut-n-pasty than I like.
> Suggest a cleanup which pull all of the conditional logic,
> plus the deopt, into emit_guard_for_new.

Yes, that's better. I've pushed a cleanup:


More information about the valhalla-dev mailing list