[PATCH] x86: support SALC

[Maciej W. Rozycki]

On Thu, 6 Nov 2025, Jan Beulich wrote:

> >> --- a/gas/testsuite/gas/i386/opcode.d
> >> +++ b/gas/testsuite/gas/i386/opcode.d
> >> @@ -587,7 +587,8 @@ Disassembly of section .text:
> >>   9f5:	85 c3 [ 	]*test   %eax,%ebx
> >>   9f7:	85 d8 [ 	]*test   %ebx,%eax
> >>   9f9:	85 18 [ 	]*test   %ebx,\(%eax\)
> >> - 9fb:	f1 [ 	]*int1
> >> +[ 	]*[a-f0-9]+:	f1[ 	]*int1
> >> +[ 	]*[a-f0-9]+:	d6[ 	]*salc
> >>  [ 	]*[a-f0-9]+:	0f 4a 90 90 90 90 90 	cmovp  -0x6f6f6f70\(%eax\),%edx
> >>  [ 	]*[a-f0-9]+:	0f 4b 90 90 90 90 90 	cmovnp -0x6f6f6f70\(%eax\),%edx
> >>  [ 	]*[a-f0-9]+:	66 0f 4a 90 90 90 90 90 	cmovp  -0x6f6f6f70\(%eax\),%dx
> > 
> >  Why the unrelated change to `int1' (and likewise throughout)?
> 
> Minimal style improvements on earlier additions, in immediately adjacent code.
> Really the whole file (and much of the testsuite) could do with such, but I
> think it's better to do so incrementally. (And just to mention - I really
> consider this an overly nit-picky question.)

 So I think it should be done with a separate no-semantics-change commit.  
Though for many test cases it doesn't really matter if we match addresses 
exactly or not unless you envision future instruction insertion in the 
middle of the file (in which case you can overhaul address matching just 
before the change required; I've done it with MIPS test cases myself on 
several occasions).  Otherwise it seems just distracting noise.  My view 
anyway, please feel free to disagree.

  Maciej