So, a fun problem. Compiling openocd 0.8.0, using libtool 2.4.2 in /usr/bin/libtool. On some machines, this worked fine. On others, even though as far as I could tell my tools were identical, not so much. (Fedora 20, on both machines; the problematic one was a VM.)
The error looked like this.
1234 | libtool: Version mismatch error. This is libtool 2.4 [snip], but the
libtool: definition of this LT_INIT comes from libtool 2.4.2.
libtool: You should recreate aclocal.m4 with macros from libtool 2.4 [snip]
libtool: and run autoconf again.
|
I recently had an interesting bit of code to work with - functions were aliased to a mangled function name. I had changed some code, and now the project refused to compile, saying that the name could not be found. I wondered how to solve this chicken and egg problem: the mangled name is output by the compiler, but it won't compile when there are errors.
Turns out you can do it manually!
1234567891011 |
#define MYFOOBARALIAS __attribute__ ((weak, alias ("_ZN3Foo3barEv")))
void Foo::bar() {
...
}
|
A fantastic article on biases, tendencies, and misjudgments. An abridged speech can be found
on youtube.
I've worked in different areas of semi for a couple years now. This article is exceptionally accurate, yet simple enough to understand with a minimum of physics, EE, and CE background.
I several times watched a talk by an Intel Fellow (internal Intel lecture) about similar topics, including the future of semiconductors. While I can't repeat what she said, this article is one of the best I've seen, similar to what she talked about, just without the incredible insight of someone working on future node development or the juicy inside details.