How to measure the number of increments per second - c++

I want to measure the speed in which my PC can increment a counter N times (e.g., for N = 10^9).
I tried the following code:
using namespace std
auto start = chrono::steady_clock::now();
for (int i = 0; i < N; ++i)
{
}
auto end = chrono::steady_clock::now();
However, the compiler is smart enough to simply set i=N, and I get that start==end regardless of the value of N.
How can I change the code to measure the increment speed? (adding costly operations in the loop would dominate the runtime and would not allow the measurement to be correct).
I use Windows 10 and Visual Studio 15.9.7.
A bit of motivation: my code takes about 2 seconds for N=10^9. I'm wondering if there's any "meat" left in optimizing it further (e.g., could it possibly go down to 1 sec? or would the loop itself require more?)

This question doesn't really make sense in C or C++. The compiler aims to generate the fastest code that meets the constraints defined by your source code. In your question, you do not define a constraint that the compiler must do a loop at all. Because the loop has no effect, the optimizer will remove it.
Gabriel Staple's answer is probably the nearest thing you can get to a sensible answer to your question, but it is also not quite right because it defines too many constraints that limits the compiler's freedom to implement optimal code. Volatile often forces the compiler to write the result back to memory each time the variable is modified.
eg, this code:
void foo(int N) {
for (volatile int i = 0; i < N; ++i)
{
}
}
Becomes this assembly (on an x64 compiler I tried):
mov DWORD PTR [rsp-4], 0
mov eax, DWORD PTR [rsp-4]
cmp edi, eax
jle .L1
.L3:
mov eax, DWORD PTR [rsp-4] # Read i from mem
add eax, 1 # i++
mov DWORD PTR [rsp-4], eax # Write i to mem
mov eax, DWORD PTR [rsp-4] # Read it back again before
# evaluating the loop condition.
cmp eax, edi # Is i < N?
jl .L3 # Jump back to L3 if not.
.L1:
It sounds like your real question is more like how fast is:
L1: add eax, 1
jmp L1
Even the answer to that is complex and requires an understanding of the internals of your CPU's pipelines.
I recommend playing with Godbolt to understand more about what the compiler is doing. eg https://godbolt.org/z/59XUSu

You can directly measure the speed of the "empty loop", but it is not easy to convince a C++ compiler to emit it. GCC and Clang can be tricked with asm volatile("") but MSVC inline assembly has always been different and is disabled completely for 64bit programs.
It is possible to use MASM to side-step that restriction:
.MODEL FLAT
.CODE
_testfun PROC
sub ecx, 1
jnz _testfun
ret
_testfun ENDP
END
Import it into your code with extern "C" void testfun(unsigned N);.

Try volatile int i = 0 In your for loop. The volatile keyword tells the compiler this variable could change at any time, due to outside events or threads, and therefore it can't make the same assumptions about what the variable might be in the future.

Related

Why does MSVC generate nop instructions for atomic loads on x64?

If you compile code such as
#include <atomic>
int load(std::atomic<int> *p) {
return p->load(std::memory_order_acquire) + p->load(std::memory_order_acquire);
}
you see that MSVC generates NOP padding after each memory load:
int load(std::atomic<int> *) PROC
mov edx, DWORD PTR [rcx]
npad 1
mov eax, DWORD PTR [rcx]
npad 1
add eax, edx
ret 0
Why is this? Is there any way to avoid it without relaxing the memory order (which would affect the correctness of the code)?
p->load() may eventually use the _ReadWriteBarrier compiler intrinsic.
According to this: https://developercommunity.visualstudio.com/t/-readwritebarrier-intrinsic-emits-unnecessary-code/1538997
the nops get inserted because of the flag /volatileMetadata which is now on by default. You can return to the old behavior by adding /volatileMetadata-, but doing so will result in worse performance if your code is ever run emulated. It’ll still be emulated correctly, but the emulator will have to pessimistically assume every load/store needs a barrier.
And compiling with /volatileMetadata- does indeed remove the npad.

improve atomic read from InterlockedCompareExchange()

Assuming architecture is ARM64 or x86-64.
I want to make sure if these two are equivalent:
a = _InterlockedCompareExchange64((__int64*)p, 0, 0);
MyBarrier(); a = *(volatile __int64*)p; MyBarrier();
Where MyBarrier() is a memory barrier (hint) of compiler level, like __asm__ __volatile__ ("" ::: "memory").
So method 2 is supposed to be faster than method 1.
I heard that _Interlocked() functions would also imply memory barrier of both compiler and hardware level.
I heard that read (proper-aligned) intrinsic data is atomic on these architectures, but I am not sure if method 2 could be widely used?
(ps. because I think CPU will handle data dependency automatically so hardware barrier is not much considered here.)
Thank you for any advise/correction on this.
Here is some benchmarks on Ivy Bridge (i5 laptop).
(1E+006 loops: 27ms):
; __int64 a = _InterlockedCompareExchange64((__int64*)p, 0, 0);
xor eax, eax
lock cmpxchg QWORD PTR val$[rsp], rbx
(1E+006 loops: 27ms):
; __faststorefence(); __int64 a = *(volatile __int64*)p;
lock or DWORD PTR [rsp], 0
mov rcx, QWORD PTR val$[rsp]
(1E+006 loops: 7ms):
; _mm_sfence(); __int64 a = *(volatile __int64*)p;
sfence
mov rcx, QWORD PTR val$[rsp]
(1E+006 loops: 1.26ms, not synchronized?):
; __int64 a = *(volatile __int64*)p;
mov rcx, QWORD PTR val$[rsp]
For the second version to be functionally equivalent, you obviously need atomic 64-bit reads, which is true on your platform.
However, _MemoryBarrier() is not a "hint to the compiler". _MemoryBarrier() on x86 prevents compiler and CPU reordering, and also ensures global visibility after the write. You also probably only need the first _MemoryBarrier(), the second one could be replaced with a _ReadWriteBarrier() unless a is also a shared variable - but you don't even need that since you are reading through a volatile pointer, which will prevent any compiler reordering in MSVC.
When you create this replacement, you basically end up with pretty much the same result:
// a = _InterlockedCompareExchange64((__int64*)&val, 0, 0);
xor eax, eax
lock cmpxchg QWORD PTR __int64 val, r8 ; val
// _MemoryBarrier(); a = *(volatile __int64*)&val;
lock or DWORD PTR [rsp], r8d
mov rax, QWORD PTR __int64 val ; val
Running these two in a loop, on my i7 Ivy Bridge laptop, gives equal results, within 2-3%.
However, with two memory barriers, the "optimized version" is actually around 2x slower.
So the better question is: Why are you using _InterlockedCompareExchange64 at all? If you need atomic access to a variable, use std::atomic, and an optimizing compiler should compile it to the most optimized version for your architecture, and add all the necessary barriers to prevent reordering and ensure cache coherency.

Optimisation of a write out of a loop

Moving a member variable to a local variable reduces the number of writes in this loop despite the presence of the __restrict keyword. This is using GCC -O3. Clang and MSVC optimise the writes in both cases. [Note that since this question was posted we observed that adding __restrict to the calling function caused GCC to also move the store out of the loop. See the godbolt link below and the comments]
class X
{
public:
void process(float * __restrict d, int size)
{
for (int i = 0; i < size; ++i)
{
d[i] = v * c + d[i];
v = d[i];
}
}
void processFaster(float * __restrict d, int size)
{
float lv = v;
for (int i = 0; i < size; ++i)
{
d[i] = lv * c + d[i];
lv = d[i];
}
v = lv;
}
float c{0.0f};
float v{0.0f};
};
With gcc -O3 the first one has an inner loop that looks like:
.L3:
mulss xmm0, xmm1
add rdi, 4
addss xmm0, DWORD PTR [rdi-4]
movss DWORD PTR [rdi-4], xmm0
cmp rax, rdi
movss DWORD PTR x[rip+4], xmm0 ;<<< the extra store
jne .L3
.L1:
rep ret
The second here:
.L8:
mulss xmm0, xmm1
add rdi, 4
addss xmm0, DWORD PTR [rdi-4]
movss DWORD PTR [rdi-4], xmm0
cmp rdi, rax
jne .L8
.L7:
movss DWORD PTR x[rip+4], xmm0
ret
See https://godbolt.org/g/a9nCP2 for the complete code.
Why does the compiler not perform the lv optimisation here?
I'm assuming the 3 memory accesses per loop are worse than the 2 (assuming size is not a small number), though I've not measured this yet.
Am I right to make that assumption?
I think the observable behaviour should be the same in both cases.
This seems to be caused by the missing __restrict qualifier on the f_original function. __restrict is a GCC extension; it is not quite clear how it is expected to behave in C++. Maybe it is a compiler bug (missed optimization) that it appears to disappear after inlining.
The two methods are not identical. In the first, the value of v is updated multiple times during the execution. That may be or may not be what you want, but it is not the same as the second method, so it is not something the compiler can decide for itself as a possible optimization.
The restrict keyword says there is no aliasing with anything else, in effect same as if the value had been local (and no local had any references to it).
In the second case there is no external visible effect of v so it doesn't need to store it.
In the first case there is a potential that some external might see it, the compiler doesn't at this time know that there will be no threads that could change it, but it knows that it doesn't have to read it as its neither atomic nor volatile. And the change of d[] another externally visible variable make the storing necessary.
If the compiler writers reasoning, well neither d nor v are volatile nor atomic so we can just do it all using 'as-if', then the compiler has to be sure no one can touch v at all. I'm pretty sure this will come in one of the new version as there is no synchronisation before the return and this will be the case in 99+% of all cases anyway. Programmers will then have to put either volatile or atomic on variables that are changed, which I think I could live with.

Is it the case that comparisons imply a branch?

I was reading the wikipedia page on optimization:
http://en.wikibooks.org/wiki/Optimizing_C%2B%2B/Code_optimization/Pipeline
and I came across the line:
For pipelined processors, comparisons are slower than differences, because they imply a branch.
Why is it the case that comparisons imply a branch?
For example if:
int i = 2;
int x = i<5;
Is there a branch in this? It makes sense to me to branch for if statements with conditionals but I don't understand why a comparison alone causes a branch.
Preamble: Modern compilers are capable of eliminating branches in various ways. Thus, none of the examples necessarily results a branch in the final (assembler or machine) code.
So why does the logic basically imply branches?
The code
bool check_interval_branch(int const i, int const min_i, int const max_i)
{
return min_i <= i && i <= max_i;
}
can be logically rewritten to be:
bool check_interval_branch(int const i, int const min_i, int const max_i)
{
if (min_i <= i)
{
if (i < max_i) return true;
}
return false;
}
Here you obviously have two branches (where the second one is only executed if the first one is true - short circuit) which can be mispredicted by the branch predictor which in turn leads to a reroll of the pipeline.
Visual Studio 2013 (with optimization turned one) generates the following assembly containing two branches for check_interval_branch:
push ebp
mov ebp, esp
mov eax, DWORD PTR _i$[ebp]
cmp DWORD PTR _min_i$[ebp], eax // comparison
jg SHORT $LN3#check_inte // conditional jump
cmp eax, DWORD PTR _max_i$[ebp] // comparison
jg SHORT $LN3#check_inte // conditional jump
mov al, 1
pop ebp
ret 0
$LN3#check_inte:
xor al, al
pop ebp
ret 0
The code
bool check_interval_diff(int const i, int const min_i, int const max_i)
{
return unsigned(i - min_i) <= unsigned(max_i - min_i);
}
is logically identical to
bool check_interval_diff(int const i, int const min_i, int const max_i)
{
if (unsigned(i – min_i) <= unsigned(max_i – min_i)) { return true; }
return false;
}
which contains only a single branch but executes two differences.
The generated code for check_interval_diff of Visual Studio 2013 doesn't even contain a conditional jump.
push ebp
mov ebp, esp
mov edx, DWORD PTR _i$[ebp]
mov eax, DWORD PTR _max_i$[ebp]
sub eax, DWORD PTR _min_i$[ebp]
sub edx, DWORD PTR _min_i$[ebp]
cmp eax, edx // comparison
sbb eax, eax
inc eax
pop ebp
ret 0
(The trick here is that the subtraction done by sbb is different by 1 according to the carry flag which in turn is set to 1 or 0 by the cmp instruction.)
In fact you see three differences here (2x sub, 1x sbb).
It probably depends on your data / the use case which one is faster.
See Mysticals answer here about branch prediction.
Your code int x = i<5; is logically identical to
int x = false;
if (i < 5)
{
x = true;
}
which again contains a branch (x = true only executed if i < 5.)
This involves only a single branch:
unsigned(i – min_i) <= unsigned(max_i – min_i)
While this involves two:
min_i <= i && i <= max_i
When a CPU encounters a branch, it consults its predictor and follows the most likely path. If the prediction is correct, the branch is essentially free in terms of performance. If the prediction is wrong, the CPU needs to flush the pipeline and start all over.
This kind of optimization is a double-edged sword --- if your branches are highly predictable, the first might actually run slower than the second. It entirely depends on how much you know about your data.
While the answers given here are good, not all comparisons are translated into branch instructions (they do introduce data dependencies which may also cost you some performance).
For example, the following C code
int main()
{
volatile int i;
int x = i<5;
return x;
}
Is compiled by gcc (x86-64, Optimizations enabled) into:
movl -4(%rbp), %eax
cmpl $5, %eax
setl %al
movzbl %al, %eax
The setl instruction sets the value of AL according to the result of the comparison instruction preceding it.
Of course, this is a very simple example -- and the cmp/setl combination probably introduces dependencies that prevent the processor for executing them in parallel or may even cost you a few cycles.
Still, on a modern processor, not all comparisons are turned into branch instructions.
Who ever wrote that page is not competent as a programmer. First,
comparisons don't necessarily imply a branch; it depends on what you
do with them. And whether that implies a branch or not depends on the
processor and the compiler. An if generally requires a branch, but
even then, a good optimizer can sometimes avoid it. A while or a
for will generally require a branch, unless the compiler can unwind
the loop, but that branch is highly predictable, so even when branch
prediction is an issue, it may not matter.
More generally, if you worry about anything at this level when writing
your code, you're wasting your time, and making maintenance far more
difficult. The only time you should be concerned is once you have a
performance problem, and the profiler shows that this is the spot where
you're loosing performance. At that point, you can experiment with
several different ways of writing the code, to determine which one will
result in faster code for your combination of compiler and hardware.
(Change the compiler or the hardware, and it might not be the same one.)

Register keyword in C++

What is difference between
int x=7;
and
register int x=7;
?
I am using C++.
register is a hint to the compiler, advising it to store that variable in a processor register instead of memory (for example, instead of the stack).
The compiler may or may not follow that hint.
According to Herb Sutter in "Keywords That Aren't (or, Comments by Another Name)":
A register specifier has the same
semantics as an auto specifier...
According to Herb Sutter, register is "exactly as meaningful as whitespace" and has no effect on the semantics of a C++ program.
In C++ as it existed in 2010, any program which is valid that uses the keywords "auto" or "register" will be semantically identical to one with those keywords removed (unless they appear in stringized macros or other similar contexts). In that sense the keywords are useless for properly-compiling programs. On the other hand, the keywords might be useful in certain macro contexts to ensure that improper usage of a macro will cause a compile-time error rather than producing bogus code.
In C++11 and later versions of the language, the auto keyword was re-purposed to act as a pseudo-type for objects which are initialized, which a compiler will automatically replace with the type of the initializing expression. Thus, in C++03, the declaration: auto int i=(unsigned char)5; was equivalent to int i=5; when used within a block context, and auto i=(unsigned char)5; was a constraint violation. In C++11, auto int i=(unsigned char)5; became a constraint violation while auto i=(unsigned char)5; became equivalent to auto unsigned char i=5;.
With today's compilers, probably nothing. Is was orginally a hint to place a variable in a register for faster access, but most compilers today ignore that hint and decide for themselves.
register is deprecated in C++11. It is unused and reserved in C++17.
Source: http://en.cppreference.com/w/cpp/keyword/register
Almost certainly nothing.
register is a hint to the compiler that you plan on using x a lot, and that you think it should be placed in a register.
However, compilers are now far better at determining what values should be placed in registers than the average (or even expert) programmer is, so compilers just ignore the keyword, and do what they wants.
The register keyword was useful for:
Inline assembly.
Expert C/C++ programming.
Cacheable variables declaration.
An example of a productive system, where the register keyword was required:
typedef unsigned long long Out;
volatile Out out,tmp;
Out register rax asm("rax");
asm volatile("rdtsc":"=A"(rax));
out=out*tmp+rax;
It has been deprecated since C++11 and is unused and reserved in C++17.
As of gcc 9.3, compiling using -std=c++2a, register produces a compiler warning, but it still has the desired effect and behaves identically to C's register when compiling without -O1–-Ofast optimisation flags in the respect of this answer. Using clang++-7 causes a compiler error however. So yes, register optimisations only make a difference on standard compilation with no optimisation -O flags, but they're basic optimisations that the compiler would figure out even with -O1.
The only difference is that in C++, you are allowed to take the address of the register variable which means that the optimisation only occurs if you don't take the address of the variable or its aliases (to create a pointer) or take a reference of it in the code (only on - O0, because a reference also has an address, because it's a const pointer on the stack, which, like a pointer can be optimised off the stack if compiling using -Ofast, except they will never appear on the stack using -Ofast, because unlike a pointer, they cannot be made volatile and their addresses cannot be taken), otherwise it will behave like you hadn't used register, and the value will be stored on the stack.
On -O0, another difference is that const register on gcc C and gcc C++ do not behave the same. On gcc C, const register behaves like register, because block-scope consts are not optimised on gcc. On clang C, register does nothing and only const block-scope optimisations apply. On gcc C, register optimisations apply but const at block-scope has no optimisation. On gcc C++, both register and const block-scope optimisations combine.
#include <stdio.h> //yes it's C code on C++
int main(void) {
const register int i = 3;
printf("%d", i);
return 0;
}
int i = 3;:
.LC0:
.string "%d"
main:
push rbp
mov rbp, rsp
sub rsp, 16
mov DWORD PTR [rbp-4], 3
mov eax, DWORD PTR [rbp-4]
mov esi, eax
mov edi, OFFSET FLAT:.LC0
mov eax, 0
call printf
mov eax, 0
leave
ret
register int i = 3;:
.LC0:
.string "%d"
main:
push rbp
mov rbp, rsp
push rbx
sub rsp, 8
mov ebx, 3
mov esi, ebx
mov edi, OFFSET FLAT:.LC0
mov eax, 0
call printf
mov eax, 0
mov rbx, QWORD PTR [rbp-8] //callee restoration
leave
ret
const int i = 3;
.LC0:
.string "%d"
main:
push rbp
mov rbp, rsp
sub rsp, 16
mov DWORD PTR [rbp-4], 3 //still saves to stack
mov esi, 3 //immediate substitution
mov edi, OFFSET FLAT:.LC0
mov eax, 0
call printf
mov eax, 0
leave
ret
const register int i = 3;
.LC0:
.string "%d"
main:
push rbp
mov rbp, rsp
mov esi, 3 //loads straight into esi saving rbx push/pop and extra indirection (because C++ block-scope const is always substituted immediately into the instruction)
mov edi, OFFSET FLAT:.LC0 // can't optimise away because printf only takes const char*
mov eax, 0 //zeroed: https://stackoverflow.com/a/6212755/7194773
call printf
mov eax, 0 //default return value of main is 0
pop rbp //nothing else pushed to stack -- more efficient than leave (rsp == rbp already)
ret
register tells the compiler to 1)store a local variable in a callee saved register, in this case rbx, and 2)optimise out stack writes if address of variable is never taken. const tells the compiler to substitute the value immediately (instead of assigning it a register or loading it from memory) and write the local variable to the stack as default behaviour. const register is the combination of these emboldened optimisations. This is as slimline as it gets.
Also, on gcc C and C++, register on its own seems to create a random 16 byte gap on the stack for the first local on the stack, which doesn't happen with const register.
Compiling using -Ofast however; register has 0 optimisation effect because if it can be put in a register or made immediate, it always will be and if it can't it won't be; const still optimises out the load on C and C++ but at file scope only; volatile still forces the values to be stored and loaded from the stack.
.LC0:
.string "%d"
main:
//optimises out push and change of rbp
sub rsp, 8 //https://stackoverflow.com/a/40344912/7194773
mov esi, 3
mov edi, OFFSET FLAT:.LC0
xor eax, eax //xor 2 bytes vs 5 for mov eax, 0
call printf
xor eax, eax
add rsp, 8
ret
Consider a case when compiler's optimizer has two variables and is forced to spill one onto stack. It so happened that both variables have the same weight to the compiler. Given there is no difference, the compiler will arbitrarily spill one of the variables. On the other hand, the register keyword gives compiler a hint which variable will be accessed more frequently. It is similar to x86 prefetch instruction, but for compiler optimizer.
Obviously register hints are similar to user-provided branch probability hints, and can be inferred from these probability hints. If compiler knows that some branch is taken often, it will keep branch related variables in registers. So I suggest caring more about branch hints, and forgetting about register. Ideally your profiler should communicate somehow with the compiler and spare you from even thinking about such nuances.