int128 and uint128 support?

[Jason Sachs]

Are there plans to add int128 and uint128 support?

I'm doing some stuff with linear feedback shift registers (LFSRs) that in the 64-96 bit range. The operations I need are just <<, >>, and bitwise operations; I don't care if math operations like multiply aren't the fastest in the world.

[Stanley Seibert]

We had not considered it, but it might not be too hard.  I believe LLVM can generate 128-bit integer code (possibly needing some special routines from their compiler_rt library).  I've opened an issue to track it, and we can take a look for 0.38 (0.37 will be focused mostly on bug fixing):

https://github.com/numba/numba/issues/2600

On Thu, Nov 9, 2017 at 8:48 AM, Jason Sachs <jms...@gmail.com> wrote:

Are there plans to add int128 and uint128 support?

I'm doing some stuff with linear feedback shift registers (LFSRs) that in the 64-96 bit range. The operations I need are just <<, >>, and bitwise operations; I don't care if math operations like multiply aren't the fastest in the world.

--
You received this message because you are subscribed to the Google Groups "Numba Public Discussion - Public" group.
To unsubscribe from this group and stop receiving emails from it, send an email to numba-users+unsubscribe@continuum.io.
To post to this group, send email to numba...@continuum.io.
To view this discussion on the web visit https://groups.google.com/a/continuum.io/d/msgid/numba-users/CAOo6sOM75t3F7r5DeJn4LdkeiOsk4ORLn0Q1FFYQi9_TKWQS1g%40mail.gmail.com.
For more options, visit https://groups.google.com/a/continuum.io/d/optout.