v0.0.0-5041-gf4e0f594b

IWYU: Fix missing includes

Addressing the top missing includes:
  * xls/ir/value.h
  * xls/ir/bits.h
  * xls/ir/node.h
  * xls/ir/source_location.h
  * absl/types/span.h

Should fix at least around 1600 misc-include-cleaner warnings.

PiperOrigin-RevId: 629910777

v0.0.0-5036-g52752f50b

Add an example for the intent of the scrub_crasher.cc

PiperOrigin-RevId: 629559227

v0.0.0-5025-g51222db0e

Remove unnecessary dependencies in //xls/common:init_xls

PiperOrigin-RevId: 629189443

v0.0.0-5018-ga3b4f4c6b

Avoid 'else' after return (clang-tidy readability-else-after-return)

(fixes 27 of our clang-tidy warnings).
While at it, fix a couple of includes.

PiperOrigin-RevId: 628510900

v0.0.0-5007-g16831e939

Fix typo in comment.

PiperOrigin-RevId: 628245992

v0.0.0-5005-gd8b09c68c

Fix bugs in `Interval::IsTrueWhenAndWith`

This function's definition didn't quite meet its specification; in particular, it used an invalid mask. The result was incorrect behavior of all types (false positives & false negatives).

This could have caused some incorrect optimization analyses for `PrioritySelect` in the RangeQueryEngine.

Making an efficient implementation of this required the new `interval_ops::CoversTernary` function, which tests whether a ternary pattern intersects an `Interval` in O(bits) time. Since this function turned out to be quite complicated, I ended up writing a corresponding proof in the inline comments, and added fuzz testing (which quickly caught several bugs in my first implementation). I also exhaustively tested it for all combinations of 7-bit `Interval`s and 7-bit ternary patterns. (The exhaustive test took several minutes to run on my local machine, so I haven't checked that part in.)

PiperOrigin-RevId: 627946453

v0.0.0-4996-ged8e7f1fe

Fix incorrect serialization of `bypass` on channels.

PiperOrigin-RevId: 627545541

v0.0.0-4987-g611f12b40

Don't extract stages in benchmark_main when `--compare_delay_to_synth…

v0.0.0-4980-ga8ff162fa

[DSLX:FE][cleanup] Switch from kChannel to kChan to reflect the keyword.

PiperOrigin-RevId: 626460923

v0.0.0-4973-g3a37270a5

[DSLX:TS] Use type zipper to print mismatches in types more nicely.

Remaining to be done: we should identify when tuples have "substring matches";
e.g. they won't structurally match but some elements inside of them will. This
will help make messages better for cases e.g. where we leave off a tuple
element by accident.

Towards google/xls#1379

PiperOrigin-RevId: 626218553