RFC (M): Purge ShenandoahVerifyReadsToFromSpace

[Aleksey Shipilev]

On 04/20/2017 01:05 PM, Roman Kennke wrote:
> I am all for purging the VerifyReads part, it has never been very
> useful, for the reasons you mention.

Ok, I am purging VerifyReads then.

> While VerifyWrites is more useful, it is also very suspicious, because
> we need to unprotect from-space all the time during evac, only to poke
> in the new brooks ptrs. And since this is not possible to do
> thread-local (unless you've got special-feature-kernel stuff which,
> iirc, Intel did for one of their new processors a while ago), we open up
> the doors for plenty of from-space-writes to slip through. The
> compiler-generated verification code for stores is more helpful (but
> depends on us first having to find all the places where compilers
> generate stores, which sortof defeats the purpose too).

I guess faced with two inconvenient cases: one that loses writes because it
needs to unprotect for fwdptr writes, and another that loses writes because it
can miss the write path in the VM code -- it makes sense to have both
operational. This multiplies the probability that we catch the offending write
by at least one of the mechanisms. Two independent unreliable mechanisms seem
better than one :)

> Maybe wait for Christine to voice her opinions too.

Christine?

-Aleksey