regehr

13 ms·

91.
by regehr 16y ago
In the meantime, here's one: regehr@home:~/volatile/bugs/tmp007$ cparser --version cparser (0.9.11) using libFirm (1.18) This is free software; see the source for copying conditions. There is NO warranty; not even for MERCHANTABI
92.
by regehr 16y ago
SWEET AND SOUR BITTER MELON!
93.
by regehr 16y ago
Joke from a lawyer I know: Family court == good people at their worst. Criminal court == bad people at their best.
94.
by regehr 16y ago
Hard to believe this article didn't contain the word "invariant."
95.
by regehr 16y ago
"exercise science"
96.
by regehr 16y ago
If you can forget the silly musical, _Wicked_ was a pretty awesome book along these same lines.
97.
by regehr 16y ago
Gas is too cheap here. This will fix itself sooner or later.
98.
by regehr 16y ago
Yep. As we tried to make clear, there's an implicit contract and both sides must take their responsibilities seriously.
99.
by regehr 16y ago
paf paf paf
100.
by regehr 16y ago
Another reason we reported more bugs to LLVM is that on average, they fixed bugs faster.
101.
by regehr 16y ago
Weighting our benchmark functions using profile data (as discussed in the conclusion of the blog post) should address this problem, at least partially.
102.
by regehr 16y ago
And also grad students. Everyone who stops by has a legitimate reason (need signature on some random form, question about a homework, etc.) but the cumulative effect is to prevent coherent thought.
103.
by regehr 16y ago
The source article is one of my absolute favorites. But CS department != industrial research lab...
104.
by regehr 16y ago
It's probably best to love the mediocre and stupid children as well.
105.
by regehr 16y ago
Argh, the "special edition" appears to no longer be available on Amazon!
106.
by regehr 16y ago
We can only hope that Google Millenium Edition lives up to the awesome legacy left behind by Windows ME.
107.
by regehr 16y ago
Everyone knows Airwolf is faster.
108.
by regehr 16y ago
Never fight a land war in Asia.
109.
Nine ways to break your systems code using volatile(blog.regehr.org)
2 points by regehr 17y ago|0 comments
110.
by regehr 17y ago
This reminds me a bit of Andy Tanenbaum's cookbook, which contains a breathtakingly bad lasagna recipe consisting of noodles and beef topped with ketchup: http://www.cs.vu.nl/~ast/home/how_to_prep.ps I first read this more than 10 years a
111.
by regehr 17y ago
It was really a bug, and they fixed it: https://bugs.launchpad.net/ubuntu/hardy/+source/gcc-4.2/+bug...
112.
by regehr 17y ago
Thanks :). I'll be teaching it in the Fall. It was actually this class that motivated the whole compiler bug-finding project. The quality of the average embedded compiler is appalling, students trip on codegen bugs all the time. Of course
113.
by regehr 17y ago
-O3 is indeed riskier. All experienced embedded systems developers know this: embedded compilers tend to be much buggier than compilers for desktop platforms. But desktop compilers are buggy too. Over the last couple years my group has r