nir: Always lower i2b

There are a lot of optimizations in opt_algebraic that match ('ine', a,
0), but there are almost none that match i2b.  Instead of adding a huge
pile of additional patterns (including variation that include both ine
and i2b), always lower i2b to a != 0.

All of the Haswell and Ivy Bridge shaders that were helped for spills
and fills were vertex shaders in Civilization VI.  I suspect the vec4
backend is just missing some optimization that the scalar backend has,
but I did not investigate further.

The failing test on d3d12 is a pre-existing bug that is triggered by
this change.  I talked to Jesse about it, and, after some analysis, he
suggested just adding it to the list of known failures.

v2: Don't rematerialize i2b instructions in dxil_nir_lower_x2b.

v3: Don't rematerialize i2b instructions in zink_nir_algebraic.py.

v4: Fix zink-on-TGL CI failures by calling nir_opt_algebraic after
nir_lower_doubles makes progress.  The latter can generate b2i
instructions, but nir_lower_int64 can't handle them (anymore).

Tiger Lake, Ice Lake, Skylake, and Broadwell had similar results. (Ice Lake shown)
Ice Lake
total instructions in shared programs: 19953931 -> 19954111 (<.01%)
instructions in affected programs: 87317 -> 87497 (0.21%)
helped: 0 / HURT: 73

total cycles in shared programs: 857170168 -> 857181940 (<.01%)
cycles in affected programs: 12416948 -> 12428720 (0.09%)
helped: 11 / HURT: 83

Haswell and Ivy Bridge had similar results. (Haswell shown)
total instructions in shared programs: 16697842 -> 16697804 (<.01%)
instructions in affected programs: 105011 -> 104973 (-0.04%)
helped: 57 / HURT: 86

total cycles in shared programs: 882352387 -> 882363250 (<.01%)
cycles in affected programs: 12901276 -> 12912139 (0.08%)
helped: 79 / HURT: 107

total spills in shared programs: 14937 -> 14925 (-0.08%)
spills in affected programs: 12 -> 0
helped: 9 / HURT: 0

total fills in shared programs: 17572 -> 17560 (-0.07%)
fills in affected programs: 12 -> 0
helped: 9 / HURT: 0

Sandy Bridge
total instructions in shared programs: 13916303 -> 13916309 (<.01%)
instructions in affected programs: 7311 -> 7317 (0.08%)
helped: 3 / HURT: 8

total cycles in shared programs: 741167673 -> 741162267 (<.01%)
cycles in affected programs: 433466 -> 428060 (-1.25%)
helped: 21 / HURT: 8

LOST:   3
GAINED: 0

No shader-db changes on Iron Lake or GM45.

All Intel platforms had similar results. (Ice Lake shown)
Instructions in all programs: 141165875 -> 141165873 (-0.0%)
Instructions helped: 2

Cycles in all programs: 9098956382 -> 9098956350 (-0.0%)
Cycles helped: 2

Acked-by: Jesse Natalie <jenatali@microsoft.com>
Tested-by: Daniel Schürmann <daniel@schuermann.dev>
229 jobs for !15121 with review/always-lower-i2b in 9 seconds (queued for 8 seconds)
merge request