Skip to content
  1. Mar 02, 2004
  3. Mar 01, 2004
    • Chris Lattner's avatar
      Add this back, as its absence introduces assertions, and it seems to work now · f96405a1
      Chris Lattner authored 
      that Instructions are annotable again

llvm-svn: 12045
      f96405a1
    • Tanya Lattner's avatar
      fix bug in previous checkin · e63e8ea8
      Tanya Lattner authored 
      llvm-svn: 12044
      e63e8ea8
    • Brian Gaeke's avatar
      TargetCacheInfo has been removed; its only uses were to propagate a constant · 427cec13
      Brian Gaeke authored 
      (16) into certain areas of the SPARC V9 back-end. I'm fairly sure the US IIIi's
dcache has 32-byte lines, so I'm not sure where the 16 came from. However, in
the interest of not breaking things any more than they already are, I'm going
to leave the constant alone.

llvm-svn: 12043
      427cec13
    • Chris Lattner's avatar
      Handle passing constant integers to functions much more efficiently. Instead · 1f4642c4
      Chris Lattner authored 
      of generating this code:

        mov %EAX, 4
        mov DWORD PTR [%ESP], %EAX
        mov %AX, 123
        movsx %EAX, %AX
        mov DWORD PTR [%ESP + 4], %EAX
        call Y

we now generate:
        mov DWORD PTR [%ESP], 4
        mov DWORD PTR [%ESP + 4], 123
        call Y

Which hurts the eyes less.  :)

Considering that register pressure around call sites is already high (with all
of the callee clobber registers n stuff), this may help a lot.

llvm-svn: 12028
      1f4642c4
    • Chris Lattner's avatar
      Fix a minor code-quality issue. When passing 8 and 16-bit integer constants · 5c7d3cda
      Chris Lattner authored 
      to function calls, we would emit dead code, like this:

int Y(int, short, double);
int X() {
  Y(4, 123, 4);
}

--- Old
X:
        sub %ESP, 20
        mov %EAX, 4
        mov DWORD PTR [%ESP], %EAX
***     mov %AX, 123
        mov %AX, 123
        movsx %EAX, %AX
        mov DWORD PTR [%ESP + 4], %EAX
        fld QWORD PTR [.CPIX_0]
        fstp QWORD PTR [%ESP + 8]
        call Y
        mov %EAX, 0
        # IMPLICIT_USE %EAX %ESP
        add %ESP, 20
        ret

Now we emit:
X:
        sub %ESP, 20
        mov %EAX, 4
        mov DWORD PTR [%ESP], %EAX
        mov %AX, 123
        movsx %EAX, %AX
        mov DWORD PTR [%ESP + 4], %EAX
        fld QWORD PTR [.CPIX_0]
        fstp QWORD PTR [%ESP + 8]
        call Y
        mov %EAX, 0
        # IMPLICIT_USE %EAX %ESP
        add %ESP, 20
        ret

Next up, eliminate the mov AX and movsx entirely!

llvm-svn: 12026
      5c7d3cda
  5. Feb 29, 2004
  7. Feb 28, 2004