[willtim]

C actually gives one rather limited control over modern hardware with it's memory hierarchies and superscaler CPUs. Programming language research has also moved on a lot since the 70's, which is why we should be considering less dangerous languages (e.g. better type systems and less undefined behaviour). Languages like ATS and Rust also support explicit memory management, whilst being a whole lot safer.

[vardump]

C alone doesn't provide the control directly, but you as a programmer can absolutely leverage C to take control of the memory hierarchies by controlling your data access patterns. IOW, high locality of reference.

Good C-compilers will most of the time take care of the superscalar CPU friendliness. When they don't, you can always drop down to the assembler level, and it'll mesh well with C.

[willtim]

High-locality of reference can be achieved in any language that supports unboxed types, it doesn't require C (even a very high-level language like Haskell has support for this). But this is a long way from having complete control how each memory heirarchy is used.

Likewise most static languages defer to the compiler for CPU-specific performance optimisations and will permit foreign native calls into C or ASM where necessary. So I don't see how this is an argument in C's favour.

[vardump]

> High-locality of reference can be achieved in any language that supports unboxed types, it doesn't require C (even a very high-level language like Haskell supports this).

You often also need correct alignment. Cache-line or page. Your unboxed access across two pages can cause two TLB misses, L1 misses etc. Not to mention two page faults.

Sometimes you need to ensure two (or more) buffers are NOT aligned in a particular way to avoid interfering with CPU caching mechanisms.

[willtim]

The only support C is giving you for this is that it has sized unboxed types (and raw pointer access). Even then, you'd have to trust the compiler and take measurements to be sure.

[unwind]

That's not true, since C11 we have:

    #include <stdlib.h>

    void *aligned_alloc(size_t alignment, size_t size);

which works like malloc() but lets you specify the required alignment.

[willtim]

Could I not just call such a function from my alternative language?

[rurban]

Only for sizes <4KB

[vardump]

That's not true. You can also control data alignment in various ways. For example, you don't need to write to the beginning of an allocated buffer, but skip to the point where low order address bits are what you want.

Something like this for example:

  char* aligned_buf;
  char* buf;
  size_t max_align_offset = (1<<align) - 1;
  buf = malloc(length_needed + max_align_offset);
  aligned_buf = (buf + max_align_offset) & ~max_align_offset;

In the example, if align==8, you have 256 byte alignment. If it's 12, 4kB alignment.

[pjmlp]

Good luck ensuring that doesn't trigger UB across all target architectures and compilers being used.

[willtim]

This example looks like it should be turned into a malloc variant, i.e. needs only to be written once. Raw pointer access is also available in other languages of course, albeit it is usually made much more difficult.

[pjmlp]

Even in the 70's there was NEWP, PL/I, PL/S, PL/8, Concurrent Pascal, Mesa, BLISS, Modula-2, ....

C wins them all in implicit conversions and opportunities for memory corruption.

Their major sin was to be tied to commercial OSes, instead of one with source code available for a symbolic price to universities.

[millstone]

Are you suggesting that other languages provide more control over modern hardware?

[willtim]

Yes. Currently access to modern hardware features are either via cumbersome APIs (e.g. NUMA, AVX intrinsics), handled via the OS (e.g. paging, scheduling), or handled via the hardware itself (cache memory hierarchy). The problem will get worse as modern CPUs and machines continue to diverge from those originally targetted by C in the 1970s.