On 2016-12-11 17:27, Ming Lei wrote:
I didn't see your issue, and with the workaound in the following link:

      https://github.com/iovisor/bcc/issues/492

bcc can be built successfully on ubuntu 16.04 both arm64 and x86.

Thanks for the suggestion, but I got cmake to work just fine. Thus I'm not having the problem described at the github discussion. Rather, it's 'make' itself that fails, as outlined below:

alison@bonnet:~/bcc/build$ make
Scanning dependencies of target clang_frontend
[  1%] Building CXX object
src/cc/frontends/clang/CMakeFiles/clang_frontend.dir/loader.cc.o
[  3%] Building CXX object
src/cc/frontends/clang/CMakeFiles/clang_frontend.dir/b_frontend_action.cc.o

[ . . . ]

Linking CXX static library libbcc.a
[ 52%] Built target bcc-static
Scanning dependencies of target CPUDistribution
[ 54%] Building CXX object
examples/cpp/CMakeFiles/CPUDistribution.dir/CPUDistribution.cc.o
Linking CXX executable CPUDistribution
../../src/cc/libbcc.a(bpf_module.cc.o): In function
`ebpf::BPFModule::dump_ir(llvm::Module&)':
bpf_module.cc:(.text+0x4600): undefined reference to
`llvm::createPrintModulePass(llvm::raw_ostream&, std::string const&, bool)'

My plan is to try 'git log --src/cc/frontends/clang/CMakeFiles/clang_frontend.dir/" and to check out the last tagged release that had an earlier version of the files there. Debian Unstable right now has llvm-3.8; perhaps I should try llvm-3.9 or llvm-4.0 instead?

Per my experience, 1G should be enough for building anything almost on
arm64, even though on arm, but sometimes it can be caused by slow CPUs
or your storage.

Nice to say, but as soon as I starting building bcc natively, kswapd became the most CPU-intensive process, and then the OOM Killer came a-calling.

-- Alison

---
Alison Chaiken                      ali...@she-devel.com, 650-279-5600
http://{ she-devel.com, exerciseforthereader.org }
"You shouldn't try to optimize things you can't measure." -- Elecia White,
http://embedded.fm/episodes/170

_______________________________________________
iovisor-dev mailing list
iovisor-dev@lists.iovisor.org
https://lists.iovisor.org/mailman/listinfo/iovisor-dev

Reply via email to