• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1***
2
3add gcc builtins for alpha instructions
4
5
6***
7
8custom expand byteswap into nifty
9extract/insert/mask byte/word/longword/quadword low/high
10sequences
11
12***
13
14see if any of the extract/insert/mask operations can be added
15
16***
17
18match more interesting things for cmovlbc cmovlbs (move if low bit clear/set)
19
20***
21
22lower srem and urem
23
24remq(i,j):  i - (j * divq(i,j)) if j != 0
25remqu(i,j): i - (j * divqu(i,j)) if j != 0
26reml(i,j):  i - (j * divl(i,j)) if j != 0
27remlu(i,j): i - (j * divlu(i,j)) if j != 0
28
29***
30
31add crazy vector instructions (MVI):
32
33(MIN|MAX)(U|S)(B8|W4) min and max, signed and unsigned, byte and word
34PKWB, UNPKBW pack/unpack word to byte
35PKLB UNPKBL pack/unpack long to byte
36PERR pixel error (sum across bytes of bytewise abs(i8v8 a - i8v8 b))
37
38cmpbytes bytewise cmpeq of i8v8 a and i8v8 b (not part of MVI extensions)
39
40this has some good examples for other operations that can be synthesised well
41from these rather meager vector ops (such as saturating add).
42http://www.alphalinux.org/docs/MVI-full.html
43