http://qs1969.pair.com?node_id=1121309

In your diagram above, draw two vertical pencil-lines, 64 bits apart, with at least 64 bits to the left of the leftmost line. (In other words, this is the second quadword in the value being sought.)

You total, utter, complete and unbashed, ^%\$\$%\$^ TWONK!

There is no "second quadword in the value being sought", there are SIXTY FOUR second quadwords being sought; because the second quadword of the needle, could appear at any bit offset!

Thus, there is a 63 out of 64 chance, that no single quadword that appears in the needle, will appear anywhere -- 64-bit aligned -- in the haystack!

Even though the haystack might consist of millions of exact repetitions of the needle. They only need to be offset from the alignment by 1 bit (or 2 bits, or 3 bits or ... or 62-bits) and your REPNE SCASx instruction will simply never *&^&&%\$ see them!

Please stop talking bollocks, and STFU.

With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday'
Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.
"Science is about questioning the status quo. Questioning authority". I'm with torvalds on this
In the absence of evidence, opinion is indistinguishable from prejudice. Agile (and TDD) debunked
• Comment on Re^10: [OT] The interesting problem of comparing bit-strings.

Replies are listed 'Best First'.
Re^11: [OT] The interesting problem of comparing bit-strings.
by AppleFritter (Vicar) on Mar 25, 2015 at 20:06 UTC