Bug 1506 - Support for JIT circular shifts
Summary: Support for JIT circular shifts
Alias: None
Product: Runtime
Classification: Mono
Component: JIT ()
Version: unspecified
Hardware: PC All
: --- enhancement
Target Milestone: ---
Assignee: Bugzilla
Depends on:
Reported: 2011-10-14 21:42 UTC by Stephen Kou
Modified: 2017-07-11 21:12 UTC (History)
2 users (show)

Is this bug a regression?: ---
Last known good build:

Notice (2018-05-24): bugzilla.xamarin.com is now in read-only mode.

Please join us on Visual Studio Developer Community and in the Xamarin and Mono organizations on GitHub to continue tracking issues. Bugzilla will remain available for reference in read-only mode. We will continue to work on open Bugzilla bugs, copy them to the new locations as needed for follow-up, and add the new items under Related Links.

Our sincere thanks to everyone who has contributed on this bug tracker over the years. Thanks also for your understanding as we make these adjustments and improvements for the future.

Please create a new report on GitHub or Developer Community with your current version information, steps to reproduce, and relevant error messages or log files if you are hitting an issue that looks similar to this resolved bug and you do not yet see a matching new report.

Related Links:

Description Stephen Kou 2011-10-14 21:42:31 UTC
This isn't really a bug, but it's a pretty awesome thing to have.  Many modern crypto hash and encryption algorithms make use of rotl and rotr (circular shift) operations.  MurmurHash3 is one example.  In C/C++ compilers, the expression:

(x >> r) | (x << (32 - r)) 

is recognized and is compiled to a single x86 ROR/ROL instruction.  While CIL has no opcode for rotation, it has a distinct signature which the JIT should be able to recognize:

(rotate right)
ldloc x
ldloc r
ldloc x
ldc_i4 32
ldloc r

Rotate left is pretty similar, with the shr.un / shl flipped.  

X86, ARM, PowerPC, and Itanium all have a rotate instruction, but SPARC does not.
Comment 1 Rodrigo Kumpera 2017-07-11 21:12:04 UTC
Thanks for the feedback. Right now this sort of optimization is left for LLVM to do.