Birch-san

ah, is this what you were looking for? the contents of `CMakeFiles/bpftrace.dir/link.txt`, as mentioned in that Makefile: ```bash /usr/bin/g++-10 -g CMakeFiles/bpftrace.dir/main.cpp.o -o bpftrace -Wl,-rpath,/usr/lib/llvm-8/lib:/usr/lib/llvm-12/lib: libbpftrace.a /usr/lib/x86_64-linux-gnu/libbfd.so /usr/lib/x86_64-linux-gnu/libopcodes.so ast/libast.a arch/libarch.a /usr/lib/llvm-8/lib/libclang-8.so.1...

okay lol yeah, I was trying that as you typed it. it works! 😄 `./tests/CMakeFiles/bpftrace_test.dir/link.txt` too. cool; definitely agree that a solution like the one in your initial comment is...

```python $x = join("hello", "world"); ``` Not sure whether assignment expressions have values in bpftrace… can you do `if($x = 123)` or `$x = $y = 123;`? But in this...

Looks like this is a duplicate of https://github.com/iovisor/bpftrace/issues/26

> Being able to read in loooooooong strings would be great but I don't think we should change the current behaviour of str(), as a construct like @[str(arg0)]=arg1 should continue...

> Also please clang-format and get the builds to pass @danobi I think I'm not quite ready to write tests for this, since I have to resolve the TODOs around...

> Also please clang-format and get the builds to pass @danobi I've had a stab at that, but maybe I'm doing something wrong? ```bash git --no-pager diff --name-only master |...

Update: I've now added handling for the case where str() or printf() fails to acquire a map element. I figured it made sense to re-use the existing helper_error infrastructure, except...

> I'm wondering now what happens if you do something like: > > ``` > $x = str(one); > $y = str(two); > printf("%s\n", $x); > ``` > > Does...

> I'd tend to agree but even harsher still. I'm looking at the kernel implementation ( https://github.com/torvalds/linux/blob/master/kernel/bpf/arraymap.c#L240-L250 ) and I don't really see how a lookup on a percpu map...