heap_lang/lifting.v: Don't run auto on hopeless goals

See merge request iris/iris!369

Paolo G. Giarrusso authored 5 years ago and Robbert Krebbers committed 5 years ago

Those goals happen to be solvable by [done] as well, so use that.

I also dropped some inconsistent line breaks.

Added variant of big_sepL_lookup_acc which allows updating the value

See merge request iris/iris!365

Fix unexpected implicit binder warning

See merge request iris/iris!367

Tej Chajed authored 5 years ago and Ralf Jung committed 5 years ago

Coq master is stricter about checking for meaningless implicit binders;
see https://github.com/coq/coq/pull/10202.

Better handling of persistent results in `iDestruct`, `iPoseProof`, `iAssert` and friends

See merge request iris/iris!341

Make handling of `AddModal` in `tac_specialize_frame` consistent with that in `tac_specialize_assert`.

The result of `iSpecialize ("H" with "[H1 .. Hn]")` was always put in the
spatial context. This commit strenghtens this tactic by putting the result
in the intuitionistic context if the following conditions hold:

1. The hypothesis `H` is persistent.
2. All the hypotheses `H1 .. Hn` are intuitionistic (or similarly, in case
   `[-H1 .. Hn]` is used, if all remaining hypotheses are intuitionistic).
3. If the pattern `[> ..]` for adding a modality is not used.

Clarify use of Require

See merge request iris/iris!345

Disclaimer for STSs.

See merge request iris/iris!366

Add several lemmas about list singletons

See merge request iris/iris!364

A lemma that allows to relate a singleton with another list.

Add a stronger version of `list_core_id`.

See merge request !361

Add pair_op_1 and pair_op_2

See merge request !362

`list_lookup_singletonM_ne` is not sufficient when we need to
compare a singleton with another list, for example, to see if one
is included in the other.

A new lemma, `list_core_id'`, allows to infer that a list is
`CoreId` by only checking that all its elements are `CoreId`, as
opposed to the existing instance, `list_core_id`, that only works
when the list contains elements of the type where every element is
`CoreId`.

The two new lemmas allow splitting the resources in one component
of a pair when the other component has nothing. In combination
with `pair_split`, they allow to arbitrarily split the resource
`(a ⋅ a', b ⋅ b')`.

This is in line with `prod_local_update_1` and
`prod_local_update_2`, the lemmas that allow, in a sense, to only
consider one component of a pair.

Added Actris to Case Studies in README

See merge request iris/iris!360

More proper/non-expansiveness results for maps, lists, and big ops

See merge request iris/iris!359