@BRIAN_____ Looking at the generated assembly. “volatile” can make the difference between code that works as intended now and what we have.
-
-
Replying to @volatile_void
@spun_off You mean that `volatile` is more effective than `memset_s`?
2 replies 0 retweets 0 likes -
Replying to @BRIAN_____
@BRIAN_____ @spun_off I'd generally trust the pointer-to-volatile code, the GCC/LLVM folks have worked really hard to fix volatile bugs1 reply 0 retweets 0 likes -
Replying to @johnregehr
@johnregehr @spun_off AFAICT, the compiler is allowed to recognize when the pointed-at thing isn't actually volatile. That makes me nervous.2 replies 0 retweets 0 likes -
Replying to @BRIAN_____
@johnregehr @spun_off In particular, given: int x = 1; int volatile *p = &x; Couldn't (shouldn't?) the compiler just ignore the "volatile"?1 reply 0 retweets 0 likes -
Replying to @BRIAN_____
@BRIAN_____@johnregehr @spun_off Even if it shouldn't, sometimes it does: http://goo.gl/LDfPHG2 replies 1 retweet 5 likes -
Replying to @sevenps
@sevenps@BRIAN_____ @spun_off I'll just leave this here and run away http://goo.gl/aJERCS1 reply 0 retweets 2 likes -
Replying to @johnregehr1 reply 0 retweets 1 like
-
Replying to @BRIAN_____
@BRIAN_____ Sorry that it's so redundant but have you seen https://github.com/regehr/ub-canaries/issues/5 … ?1 reply 0 retweets 0 likes -
Replying to @volatile_void
@BRIAN_____ https://github.com/regehr/ub-canaries/issues/5#issuecomment-175182921 … makes a good point: if the array is in memory & gets overwritten there's already no guarantee about regs1 reply 0 retweets 0 likes
@spun_off Literally would be easier for me to code asm implementations for every platform I care about than to keep thinking about this. :)
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.