Lines Matching refs:Each
113 ;# Each step takes 8*2 = 16 instructions
147 ;# we need to convert 16 rows of 4 pels each into 4 registers of 16 pels
148 ;# each. Writing 0 ... 63 for the pixel indices, the desired result is:
173 ;# In other words, we transpose each of the four 4x4 submatrices.
240 ;# each iteration can be realized by (d=0, s=2):
372 ;# do abs(vala-valb) > limit. but no need to compare each
638 ;# Filtering a vertical mb. Each mb is aligned on a 16 byte boundary.
