struggling with travis

[gcolvin]

another try at getting this past travis

[holiman]

Bad automerger!

[holiman]

@gcolvin Why did you revert this? I had updated these testcases

[holiman]

@gcolvin please make sure you make your changes based off the most recent master version. You just removed all the updated testcases I had fixed yesterday, plus some clarifications from #2656

[matkt]

When you say "JUMPSUB be low, and RETURNSUB be verylow" Do you mean JUMPSUB = 5 and RETURNSUB = 3? Because I see that you have changed JUMPSUB but RETURNSUB is equal to 2 on the tests

[holiman]

This PR made some random changes here and there, but the PR description does not even mention what the intended changes were -- I don't know which ones were intentional and which ones were somehow just accidents during merge :/

[holiman]

From the YP:

Gbase 2    Amount of gas to pay for operations of the setWbase.
Gverylow 3    Amount of gas to pay for operations of the setWverylow.
Glow 5    Amount of gas to pay for operations of the setWlow.
Gmid 8    Amount of gas to pay for operations of the setWmid.

So jumpsub should be 5, according to that sentence. IMO it should be at MID=8, since that's what a JUMP is at. A JUMPSUB is more resource intensive than a JUMP since it requires an internal stack push, which JUMP doesn't.

[matkt]

From the YP:

Gbase 2    Amount of gas to pay for operations of the setWbase.
Gverylow 3    Amount of gas to pay for operations of the setWverylow.
Glow 5    Amount of gas to pay for operations of the setWlow.
Gmid 8    Amount of gas to pay for operations of the setWmid.

So jumpsub should be 5, according to that sentence. IMO it should be at MID=8, since that's what a JUMP is at. A JUMPSUB is more resource intensive than a JUMP since it requires an internal stack push, which JUMP doesn't.

That makes sense