On Thursday 2014-02-27 15:44 -0500, Zack Weinberg wrote: > This is exactly the same thing dbaron said the last time I brought > this up (quite some time ago - 2010, maybe?) I didn't buy it then and > I don't buy it now. I think it is far more likely that a > maybe-used-uninitialized true positive will *go unnoticed* because > we've trained ourselves to ignore those warnings, and I don't think > better precision in a valgrind run is worth the risk of letting a true > positive slip through in a release build. Those bugs tend to be > security critical.
I'm ok with it if you're initializing it to a value that's properly handled as a reasonable default by the later code. But if you're initializing it to something that's arbitrary or useless (like, say, initializing an enumerated type to 0 so that you'll fall into whichever case that happens to be, whether it involves assuming a tagged pointer is of a certain type or not), I'm still against, and I'd rather see the valgrind warning. But I'm ok with restructuring the code to avoid having either problem. > I also defy anyone to demonstrate a measurable performance impact from > the tiny amount of additional machine code that might be emitted if we > added initializations to squelch all those warnings. I'm not at all concerned about performance impact. -David -- 𝄞 L. David Baron http://dbaron.org/ 𝄂 𝄢 Mozilla https://www.mozilla.org/ 𝄂 Before I built a wall I'd ask to know What I was walling in or walling out, And to whom I was like to give offense. - Robert Frost, Mending Wall (1914)
signature.asc
Description: Digital signature
_______________________________________________ dev-platform mailing list dev-platform@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-platform
