Comment by canarypilot | Hacker News Reader

Comment by canarypilot | Hacker News Reader

Parent Full thread

GCC treats inline assembly as an opaque blob of text It knows nothing about, plus some register constraints it knows something about. It will “optimise” assembly if you’ve passed the right registers as input/output, and haven’t marked it volatile and memory clobbering. Simple things like eliminating it if it is dead code, but this probably isn’t the optimisation you were looking for!

Depending on who you ask, this is a feature, not a defect. Inline assembly is your last line of defence against a compiler which thinks it knows better than you.

Clang/llvm go a step further and do parse and optimize your inline assembly; that makes me uncomfortable, but I’ve always subscribed to the “don’t touch user asm” mindset for compiler development.

Your main benefit of builtins over assembly is that no matter how hard I tried I couldn’t get the suggested replacement given by the parent for __atomic_compare_exchange to assemble on my arm64 server!

I've never coded directly for ARM, but if it's anything like x86, then this example might help you (on GCC6+):

char DidIt; asm("lock cmpxchg" : "=@ccz"(DidIt), "+m"(IfThing), "+a"(IsEqualToMe) : "r"(ReplaceItWithMe) : "cc");

Thank you for assuming the best from me, and providing a genuine attempt at help in response to my much more bad-faith remark.

Two things complicate the AArch64 (Arm64) story for atomic builtins.

One is Arm’s weak memory model. This permits different, more efficient, implementations of compare exchange depending on which of the C++ ordering guarantees you need and ask for. A compiler will generate different code for the sequential consistent model than it will for acquire/release.

The other is that there are two ways to provide atomics on Arm, and which you have access to depends on the architecture revision implemented by your particular Arm hardware. If you have Armv8.1-A atomic operations available to you, you’ll use a similar interface to x86; a single instruction which looks like a cmpxchg. If you don’t have the atomic extension, you’ll instead be using a load-lock/store-conditional loop of ldrex/strex for all of your atomic operations.

Getting this exactly right is tricky, so where possible I’d encourage using the atomic builtins (or, of course, your language’s portable API, over assembly).

The biggest benefit is that if in future architecture technology moves forward again, the GCC and language built ins will let you move with it, while assembly leaves you relearning occasionally.

I think it's awesome that GCC provides builtins that help ARM devs generate the best atomic code possible, for ARM. As someone who only cares about her coding running natively on x86, I would obviously never use C/C++11 atomics, since it's kind of regressive for a high-level language to specify 42 different ways to say "MOV". How would I even know my code works? The way I see it, the "portability" here only flows in one direction: helping ARM/MIPS/etc. folks move to x86. The only thing I don't get, is why it's part of the language standard.

Reply on news.ycombinator.com