relax | tiny release tool to support multi distribution | iOS library

This ptr.store(p, std::memory_order_release) (L1) guarantees that anything done prior to this line in this particular thread (T1) will be visible to other threads as long as those other threads are reading ptr in a correct fashion (in this case, using std::memory_order_acquire). This guarantee works only with this pair, alone this line guarantees nothing.

Now you have ptr.load(std::memory_order_acquire) (L2) on the other thread (T2) which, working with its pair from another thread, guarantees that as long as it read the value written in T1 you can see other values written prior to that line (in your case it is data). So because L1 synchronizes with L2, data = 42; happens before assert(data == 42).

Also there is a guarantee that ptr is written and read atomically, because, well, it is atomic. No other guarantees or promises are in that code.

Had the same issue. Solved it by increasing the memory dedicated to the GPU on the Raspberry Pi.

Edit your /boot/config.txt file (sudo required)

Scroll down to the [all] section, and edit the gpu_mem line as follows:

gpu_mem=256

Then reboot your pi.

Because you use relaxed ordering on a separate load & store in T2, the release sequence is broken and the second assert can trigger (although not on a TSO platform such as X86).
You can fix this by either using acq/rel ordering in thread T2 (as you suggested) or by modifying T2 to use an atomic read-modify-write operation (RMW), like this:

You may use this sed:

Yes, it's UB in ISO C++; value = data[oldReadPosition] in the C++ abstract machine involves reading the value of that object. (Usually that means lvalue to rvalue conversion, IIRC.)

But it's mostly harmless, probably only going to be a problem on machines with hardware race detection (not normal mainstream CPUs, but possibly on C implementations like clang with threadsanitizer).

Another use-case for non-atomic read and then checking for possible tearing is the SeqLock, where readers can prove no tearing by reading the same value from an atomic counter before and after the non-atomic read. It's UB in C++, even with volatile for the non-atomic data, although that may be helpful in making sure the compiler-generated asm is safe. (With memory barriers and current handling of atomics by existing compilers, even non-volatile makes working asm). See Optimal way to pass a few variables between 2 threads pinning different CPUs

atomic_thread_fence is still necessary for a SeqLock to be safe, and some of the necessary ordering of atomic loads wrt. non-atomic may be an implementation detail if it can't sync with something and create a happens-before.

People do use Seq Locks in real life, depending on the fact that real-life compilers de-facto define a bit more behaviour than ISO C++. Or another way to put it is that happen to work for now; if you're careful about what code you put around the non-atomic read it's unlikely for a compiler to be able to do anything problematic.

But you're definitely venturing out past the safe area of guaranteed behaviour, and probably need to understand how C++ compiles to asm, and how asm works on the target platforms you care about; see also Who's afraid of a big bad optimizing compiler? on LWN; it's aimed at Linux kernel code, which is the main user of hand-rolled atomics and stuff like that.

Relaxed order means that ordering of atomics and external operations only happens with regard to operations on the specific atomic object (and even then, the compiler is free to re-order them outside of the program-defined order). Thus, a relaxed store has no relationship to any state in external objects. So a relaxed load will not synchronize with your other mutexes.

The whole point of acquire/release semantics is to allow an atomic to control visibility of other memory. If you want an atomic load to mean that something is available, it must be an acquire and the value it acquired must have been released.

Relaxed is fine, you don't need any ordering wrt. access to any other elements during the process of updating. And for accesses to the same location, ISO C++ guarantees that a "modification order" exists for each location separately, and that even relaxed operations will only see the same or later values in the modification order of the location between loaded or RMWed.

You're just building an atomic fetch_max primitive out of a CAS retry loop. Since the other writers are doing the same thing, the value of each location is monotonically increasing. So it's totally safe to bail out any time you see a value greater than the new_value.

For the main thread to collect the results at the end, you do need release/acquire synchronization like thread.join or some kind of flag. (e.g. maybe fetch_sub(1, release) of a counter of how many threads still have work left to do, or an array of done flags so you can just do a pure store.)

BTW, this seems likely to be slow, with lots of time spent waiting for cache lines to bounce between cores. (Lots of false-sharing.) Ideally you you can efficiently change this to have each thread work on different parts of the array (e.g. computing multiple candidates for the same index so it doesn't need any atomic stuff).

I cannot guarantee that the computed indices do not overlap. In practice, the overlapping is usually small, but it cannot be eliminated.

So apparently that's a no. And if the indices touched by different threads are in different cache lines (chunk of 16 int32_t) then there won't be too much false sharing. (Also, if computation is expensive so you aren't producing values very fast, that's good so atomic updates aren't what your code is spending most of its time on.)

But if there is significant contention and the array isn't huge, you could give each thread its own output array, and collect the results at the end. e.g. have one thread do a[i] = max(a[i], b[i], c[i], d[i]) for 4 to 8 arrays per loop. (Not too many read streams at once, and not a variable number of inputs because that probably couldn't compile efficiently). This should benefit from SIMD, e.g. SSE4.1 pmaxsd doing 4 parallel max operations, so this should be limited mostly by L3 cache bandwidth.

Or divide the max work between threads as a second parallel phase, with each thread doing the above over part of the output array. Or have the thread_id % 4 == 0 reduce results from itself and the next 3 threads, so you have a tree of reductions if you have a system with many threads.

To answer the question in your title, there is no real relationship between atomics and sequence points.

The code as written does guarantee that the compiler must execute the atomic_fetch_sub before the atomic_load. But these functions are (in C's memory model) simply requests to the platform to perform certain actions on certain pieces of memory. Their effects, when they become visible to who, are specified by the memory model, and the ordering parameters. Thus even in the case when you know request A comes before request B, that does not mean the effects of request A are resolved before request B, unless you explicitly specify it to be.

No, the warning is useful. Fix your code.

C++20 relaxed the typename rules a little bit. But it's unrelated to this warning.

MSVC considers typename to be (almost?) completely optional, and is non-conforming in this regard. Clang apparently can do that too, for compatibility with MSVC. The warning says that your code is non-conforming, and might not work on other compilers (most notably on GCC).