x86 is space efficient, you say? AArch64: sub x9,x10,x11: 4 bytes. x86-64: mov r9,r10; sub r9,r11: 6 bytes. (Can’t use LEA here.) AArch64: sub x9,x10,#1234: 4 bytes. x86-64: lea r9,[r10-1234]: 7 bytes.
-
-
Replying to @pcwalton
Well arm64 is pretty good as far as load/store isas go too
1 reply 0 retweets 0 likes -
And while x86 may have been space optimized once, decades of extensions and prefixes completely ruined it
2 replies 0 retweets 11 likes -
Replying to @slava_pestov @jckarter
Yep. I keep saying this, but it feels like nobody believes me :)
3 replies 0 retweets 2 likes -
Replying to @pcwalton @slava_pestov
Probably a legacy of the i386 vs PowerPC days. There are better RISC ISAs, and x86 has gotten heftier, since then
2 replies 0 retweets 0 likes -
Replying to @jckarter @slava_pestov
I think it’s because Linus had a rant back in the Pentium 4 days about how awesome and space efficient x86 was and people still have that impression
2 replies 0 retweets 10 likes -
I always found this kind of goofy. For example, a smaller register file might help code density, but could cause more spilling, which is so much worse for holistic memory consumption.
1 reply 0 retweets 0 likes -
Classic x86 could also make some claim to being “Huffman coded” by having one byte encodings and special instructions for many things. In time many of those special instructions went unused and important extensions couldn’t fit compactly in the remaining encoding space, of course
1 reply 0 retweets 2 likes -
Most VEXing
1 reply 0 retweets 1 like
Exceptionally VEXing (EVEXing for short)
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.