Why is vsldoi's SHB parameter signed?

[Gustavo Serra Scalet]

Great,

I'll do it then.


> -----Original Message-----
> From: Doerr, Martin [mailto:martin.doerr at sap.com]
> Sent: quarta-feira, 8 de fevereiro de 2017 08:27
> To: Gustavo Serra Scalet <gustavo.scalet at eldorado.org.br>; ppc-aix-port-
> dev at openjdk.java.net
> Subject: RE: Why is vsldoi's SHB parameter signed?
> 
> Hi Gustavo,
> 
> I also think that using unsigned would be more comprehensible. Feel free
> to change it to uimm with the next change which uses it. There are only
> a few places which will need to get updated.
> 
> Best regards,
> Martin
> 
> 
> -----Original Message-----
> From: ppc-aix-port-dev [mailto:ppc-aix-port-dev-
> bounces at openjdk.java.net] On Behalf Of Gustavo Serra Scalet
> Sent: Dienstag, 7. Februar 2017 19:11
> To: ppc-aix-port-dev at openjdk.java.net
> Subject: Why is vsldoi's SHB parameter signed?
> 
> Hi,
> 
> I noticed that vsldoi instruction parameter SHB is being checked as a
> signed argument of 4 bits:
> 
> assembler_ppc.inline.hpp:804:inline void Assembler::vsldoi(
> VectorRegister d, VectorRegister a, VectorRegister b, int si4)         {
> emit_int32( VSLDOI_OPCODE| vrt(d) | vra(a) | vrb(b) |
> vsldoi_shb(simm(si4,4))); }
> 
> I wonder why it was not defined as an unsigned parameter of 4 bits as
> described by ISA (attached). Is there a different way to look at it?!
> 
> Thanks