@jj git fetch/git reset?
Notices by Paul Khuong (pkhuong@discuss.systems)
-
Embed this notice
Paul Khuong (pkhuong@discuss.systems)'s status on Monday, 17-Mar-2025 17:25:46 JST Paul Khuong
-
Embed this notice
Paul Khuong (pkhuong@discuss.systems)'s status on Thursday, 23-Jan-2025 01:16:01 JST Paul Khuong
@mcc movzx/sx for widening. regular mov for narrowing, but be mindful of partial dependencies on 8 and 16 bit destinations (16b in particular).
The size suffix is an at&t special and not part of the ISA (which also has movd/movq…)
-
Embed this notice
Paul Khuong (pkhuong@discuss.systems)'s status on Thursday, 23-Jan-2025 01:15:59 JST Paul Khuong
@cr1901 @mcc 16 bit ops are extra slow nowadays (no one uses them).
-
Embed this notice
Paul Khuong (pkhuong@discuss.systems)'s status on Monday, 06-Jan-2025 12:02:47 JST Paul Khuong
So, the MOAR LAYERS folks like log-sum-exp… do we know if one tried to replace log with a bitcast of float to int, and exp with the reverse? In a float world, the cost is int -> float after bitcast, and float -> int before bitcast.
-
Embed this notice
Paul Khuong (pkhuong@discuss.systems)'s status on Tuesday, 17-Dec-2024 07:09:50 JST Paul Khuong
Let's say you know the input is an exact multiple of your integer divisor. Does that let you come up with a smaller div-by-mul constant?
EDIT: I don't mean a reciprocal inverse mod 2^w, but literally a fixed point div by mul (which also avoids the need to ensure the divisor is odd).