Related
I've a rather special question: is it possible in C/++ (both because I am sure the question is the same in both languages) to specify a functions's location? Why? I have a very large list of function pointers, and I want to eliminate them.
(Currently) This looks like that(repeated over lika a million times, stored in the user's RAM):
struct {
int i;
void(* funptr)();
} test;
Because I know that in most assembly languages, functions are just "goto" directives, I had the following idea. Is it possible to optimize the above construct so that it looks like that?
struct {
int i;
// embed the assembler of the function here
// so that all the functions
// instructions are located here
// like this: mov rax, rbx
// jmp _start ; just demo code
} test2;
In the end, the thing should look like this in memory: An int holding any value, followed by the function's assembly code, referenced by test2. I should be able to call these functions like that: ((void(*)()) (&pointerToTheStruct + sizeof(int)))();
You might think that I'm insane to optimize the app that way, and I cannot disclose any more details on it's function, but if anyone has some pointers on how solve this problem, I would appreciate it.
I do not think that there is a standard way to this, so any hacky way to do this via inline assembler/other crazy things is also appreciated!
The only thing you really have to do is make the compiler aware of the (constant) value of the function pointer you want in the struct. The compiler will then (presumably/hopefully) inline that function call wherever it sees it called through that function pointer:
template<void(*FPtr)()>
struct function_struct {
int i;
static constexpr auto funptr = FPtr;
};
void testFunc()
{
volatile int x = 0;
}
using test = function_struct<testFunc>;
int main()
{
test::funptr();
}
Demo - no call or jmp after optimization.
It remains unclear what the point of the int i is. Note that the code is not technically "directly after the i" here, but it is even more unclear how you'd expect instances of the struct to look like (is the code in them or is it "static" in a way? I feel there is some misunderstanding here on your part what compilers actually produce...). But consider the ways that compiler inlining can help you and you might find the solution you need. If you're worried about executable size after inlining, tell the compiler and it will compromise between speed and size.
This sounds like a terrible idea for a lot of reasons that probably won't save memory, and will hurt performance by diluting L1I-cache with data and L1D-cache with code. And worse if you ever modify or copy objects: self-modifying code stalls.
But yes, this would be possible in C99/C11 with a flexible array member at the end of the struct, which you cast to a function pointer.
struct int_with_code {
int i;
char code[]; // C99 flexible array member. GNU extension in C++
// Store machine code here
// you can't get the compiler to do this for you. Good Luck!
};
void foo(struct int_with_code *p) {
// explicit C-style cast compiles as both C and C++
void (*funcp)(void) = ( void (*)(void) ) p->code;
funcp();
}
Compiler output from clang7.0, on the Godbolt compiler explorer is the same when compiled as either C or C++. This is targeting the x86-64 System V ABI, where the first function arg is passed in RDI.
# this is the code that *uses* such an object, not the code that goes in its code[]
# This proves that it compiles,
# without showing any way to get compiler-generated code into code[]
foo: # #foo
add rdi, 4 # move the pointer 4 bytes forward, to point at code[]
jmp rdi # TAILCALL
(If you leave out the (void) arg-type declaration in C, the compiler will zero AL first in the x86-64 SysV calling convention, in case its actually a variadic function, because it's passing no FP args in registers.)
You'd have to allocate your objects in memory that was executable (normally not done unless they're const with static storage), e.g. compile with gcc -zexecstack. Or use a custom mmap/mprotect or VirtualAlloc/VirtualProtect on POSIX or Windows.
Or if your objects are all statically allocated, it might be possible to massage compiler output to turn functions in the .text section into objects by adding an int member right before each one. Maybe with some .section and linker tricks, and maybe a linker script, you could even somehow automate it.
But unless they're all the same length (e.g. with padding like char code[60]), that won't form an array you can index, so you'll need some way of referencing all these variable-length object.
There are potentially huge performance downsides if you ever modify an object before calling its function: on x86 you'll get self-modifying-code pipeline nuke for executing code near a just-written memory location.
Or if you copied an object before calling its function: x86 pipeline flush, or on other ISAs you need to manually flush caches to get the I-cache in sync with D-cache (so the newly-written bytes can be executed). But you can't copy such objects because their size isn't stored anywhere. You can't search the machine code for a ret instruction, because a 0xc3 byte might appear somewhere that's not the start of an x86 instruction. Or on any ISA, the function might have multiple ret instructions (tail duplication optimization). Or end with a jmp instead of a ret (tailcall).
Storing a size would start to defeat the purpose of saving size, eating up at least an extra byte in each object.
Writing code to an object at runtime, then casting to a function pointer, is undefined behaviour in ISO C and C++. On GNU C/C++, make sure you call __builtin___clear_cache on it to sync caches or whatever else is necessary. Yes, this is needed even on x86 to disable dead-store elimination optimizations: see this test case. On x86 it's just a compile-time thing, no extra asm. It doesn't actually clear any caches.
If you do copy at runtime startup, maybe allocate a big chunk of memory and carve out variable-length chunks of it, while copying. If you malloc each separately, you're wasting memory-management overhead on it.
This idea will not save you memory unless you have about as many functions as you have objects
Normally you have a fairly limited number of actual functions, with many objects having copies of the same function pointer. (You've kind of hand-rolled C++ virtual functions, but with only one function you just have a function pointer directly instead of a vtable pointer to a table of pointers for that class type. One fewer levels of indirection, and apparently you're not passing the object's own address to the function.)
One of the several benefits of this level of indirection is that one pointer is usually significantly smaller than the entire code for a function. For that to not be the case, your functions would have to be tiny.
Example: with 10 different functions of 32 bytes each, and 1000 objects with function pointers, you have a total of 320 bytes of code (which will stay hot in I-cache), and 8000 bytes of function pointers. (And in your objects, another 4 bytes per object wasted on padding to align the pointer, making the total size 16 instead of 12 bytes per object.) Anyway, that's 16320 bytes total for entire structs + code. If you allocated each object separately, there's per-object bookkeeping.
With inlining machine code into each object, and no padding, that's 1000 * (4+32) = 36000 bytes, over twice the total size.
x86-64 is probably a best-case scenario, where a pointer is 8 bytes and x86-64 machine code uses a (famously complex) variable-length instruction encoding which allows for high code density in some cases, especially when optimizing for code-size. (e.g. code-golfing. https://codegolf.stackexchange.com/questions/132981/tips-for-golfing-in-x86-x64-machine-code). But unless your functions are mostly something trivial like lea eax, [rdi + rdi*2] (3 bytes=opcode + ModRM + SIB) / ret (1 byte), they're still going to take more than 8 bytes. (That's return x*3; for a function that takes a 32-bit integer x arg, in the x86-64 System V ABI.)
If they're wrappers for larger functions, a normal call rel32 instruction is 5 bytes. A load of static data is at least 6 bytes (opcode + modrm + rel32 for a RIP-relative addressing mode, or loading EAX specifically can use the special no-modrm encoding for an absolute address. But in x86-64 that's a 64-bit absolute unless you use an address-size prefix too, potentially causing an LCP stall in the decoders on Intel. mov eax, [32 bit absolute address] = addr32 (0x67) + opcode + abs32 = 6 bytes again, so this is worse for no benefit).
Your function-pointer type doesn't have any args (assuming this is C++ where foo() means foo(void) in a declaration, not like old C where an empty arg list is somewhat similar to (...)). Thus we can assume you're not passing args, so to do anything useful the functions are probably accessing some static data or making another call.
Ideas that make more sense:
Use an ILP32 ABI like Linux x32, where the CPU runs in 64-bit mode but your code uses 32-bit pointers. This would make each of your objects only 8 bytes instead of 16. Avoiding pointer-bloat is a classic use-case for x32 or ILP32 ABIs in general.
Or (yuck) compile your code as 32-bit. But then you have obsolete 32-bit calling conventions that pass args on the stack instead of registers, and less than half the registers, and much higher overhead for position-independent code. (No EIP/RIP-relative addressing.)
Store an unsigned int table index to a table of function pointers. If you have 100 functions but 10k objects, the table is only 100 pointers long. In asm you could index an array of code directly (computed goto style) if all the functions were padded to the same length, but in C++ you can't do that. An extra level of indirection with a table of function pointers is probably your best bet.
e.g.
void (*const fptrs[])(void) = {
func1, func2, func3, ...
};
struct int_with_func {
int i;
unsigned f;
};
void bar(struct int_with_func *p) {
fptrs[p->f] ();
}
clang/gcc -O3 output:
bar(int_with_func*):
mov eax, dword ptr [rdi + 4] # load p->f
jmp qword ptr [8*rax + fptrs] # TAILCALL # index the global table with it for a memory-indirect jmp
If you were compiling a shared library, PIE executable, or not targeting Linux, the compiler couldn't use a 32-bit absolute address to index a static array with one instruction. So there'd be a RIP-relative LEA in there and something like jmp [rcx+rax*8].
This is an extra level of indirection vs. storing a function pointer in each object, but it lets you shrink each object to 8 bytes, down from 16, like using 32-bit pointers. Or to 5 or 6 bytes, if you use an unsigned short or uint8_t and pack the structs with __attribute__((packed)) in GNU C.
No, not really.
The way to specify a function's location is to use a function pointer, which you're already doing.
You could make different types which have their own different member functions, but then you're back to the original problem.
I have in the past experimented with auto-generating (as a pre-build step, using Python) a function with a long switch statement that does the work of mapping int i to a normal function call. This gets rid of the function pointers, at the expense of branching. I don't remember whether it ended up being worthwhile in my case and, even if I did, that wouldn't tell us whether it's worthwhile in your case.
Because I know that in most assembly languages, functions are just "goto" directives
Well, it's perhaps a little more complicated than that…
You might think that I'm insane to optimize the app that way
Perhaps. Trying to eliminate indirection is not, in itself, a bad thing, so I don't think you're wrong to try to improve this. I just don't think that you necessarily can.
but if anyone has some pointers
lol
I don't understand the goal of this "optimization" is it about saving the memory?
I might be misunderstanding the question, but if you just replace your function pointer with a regular function, then you'll have your struct only containing the int as data and the function-pointer being inserted by the compiler when you take the address of it, instead of stored in memory.
So just do
struct {
int i;
void func();
} test;
Then sizeof(test)==sizeof(int) should hold true if you set alignment/packing to be tight.
I recall from Agner Fog's excellent guide that 64-bit Linux can pass 6 integer function parameters via registers:
http://www.agner.org/optimize/optimizing_cpp.pdf
(page 8)
I have the following function:
void x(signed int a, uint b, char c, unit d, uint e, signed short f);
and I need to pass an additional unsigned short parameter, which would make 7 in total. However, I can actually derive the value of the 7th from one of the existing 6.
So my question is which of the following is a better practice for performance:
Passing the already-calculated value as a 7th argument on 64-bit Linux
Not passing the already-calculated value, but calculating it again for a second time using one of the existing 6 arguments.
The operation in question is a simple bit-shift:
unsigned short g = c & 1;
Not fully understanding x86 assembler I am not too sure how precious registers are and whether it is better to recalculate a value as a local variable, than pass it through function calls as an argument?
My belief is that it would be better to calculate the value twice because it is such a simple 1 CPU cycle task.
EDIT I know I can just profile this- but I'd like to also understand what is happening under the hood with both approaches. Having a 7th argument does this mean cache/memory is involved, rather than registers?
The machine conventions to pass arguments is called the application binary interface (or ABI), and for Linux x86-64 is described in x86-64 ABI spec. See also x86 calling conventions wikipage.
In your case, it is probably not worthwhile to pass c & 1 as an additional parameter (since that 7th parameter is passed on stack).
Don't forget that current processor cores (on desktop or laptop computers) are often doing out-of-order execution and are superscalar, so the c & 1 operation could be done in parallel with other operations and might cost "nothing".
But leave such micro-optimizations to the compiler. If you care a lot about performance, use a recent GCC 4.8 compiler with gcc-4.8 -O3 -flto both for compiling and for linking (i.e. enable link-time optimization).
BTW, cache performance is much more relevant than such micro-optimizations. A single cache miss may take the same time (e.g. 250 nanoseconds) as hundreds of CPU machine instructions. Current CPUs are rumored to mostly wait for the caches. You might want to add a few explicit (and judicious) calls to __builtin_prefetch (see this question and this answer). But adding too much these prefetches would slow down your code.
At last, readability and maintainability of your code should matter much more than raw performance!
Basile's answer is good, I'll just point out another thing to keep in mind:
a) The stack is very likely to be in L1 cache, so passing arguments on the stack should not take more than ~3 cycles extra.
b) The ABI (x86-64 System V, in this case) requires clobbered registers to be restored. Some are saved by the caller, others by the callee. Obviously, the registers used to pass arguments must be saved by the caller if the original contents were needed again. But when your function uses more registers than the caller saved, any additional temporary results the function needs to calculate must go into a callee-saved register. So the function ends up spilling a register on the stack, reusing the register for your temporary variable, and then pops the original value back.
The only way you can avoid accessing memory is by using a smaller, simpler function that needs fewer temporary variables.
I'm copying elements from one array to another in C++. I found the rep movs instruction in x86 that seems to copy an array at ESI to an array at EDI of size ECX. However, neither the for nor while loops I tried compiled to a rep movs instruction in VS 2008 (on an Intel Xeon x64 processor). How can I write code that will get compiled to this instruction?
Honestly, you shouldn't. REP is sort of an obsolete holdover in the instruction set, and actually pretty slow since it has to call a microcoded subroutine inside the CPU, which has a ROM lookup latency and is nonpipelined as well.
In almost every implementation, you will find that the memcpy() compiler intrinsic both is easier to use and runs faster.
Under MSVC there are the __movsxxx & __stosxxx intrinsics that will generate a REP prefixed instruction.
there is also a 'hack' to force intrinsic memset aka REP STOS under vc9+, as the intrinsic no longer exits, due to the sse2 branching in the crt. this is better that __stosxxx due to the fact the compiler can optimize it for constants and order it correctly.
#define memset(mem,fill,size) memset((DWORD*)mem,((fill) << 24|(fill) << 16|(fill) << 8|(fill)),size)
__forceinline void memset(DWORD* pStart, unsigned long dwFill, size_t nSize)
{
//credits to Nepharius for finding this
DWORD* pLast = pStart + (nSize >> 2);
while(pStart < pLast)
*pStart++ = dwFill;
if((nSize &= 3) == 0)
return;
if(nSize == 3)
{
(((WORD*)pStart))[0] = WORD(dwFill);
(((BYTE*)pStart))[2] = BYTE(dwFill);
}
else if(nSize == 2)
(((WORD*)pStart))[0] = WORD(dwFill);
else
(((BYTE*)pStart))[0] = BYTE(dwFill);
}
of course REP isn't always the best thing to use, imo your way better off using memcpy, it'll branch to either sse2 or REPS MOV based on your system (under msvc), unless you feeling like writing custom assembly for 'hot' areas...
If you need exactly that instruction - use built-in assembler and write that instruction manually. You can't rely on the compiler to produce any specific machine code - even if it emits it in one compilation it can decide to emit some other equivalent during next compilation.
REP and friends was nice once upon a time, when the x86 CPU was a single-pipeline industrial CISC-processor.
But that has changed. Nowadays when the processor encounters any instruction, the first it does is translating it into an easier format (VLIW-like micro-ops) and schedules it for future execution (this is part of out-of-order-execution, part of scheduling between different logical CPU cores, it can be used to simplifying write-after-write-sequences into single-writes, et.c.). This machinery works well for instructions that translates into a few VLIW-like opcodes, but not machine-code that translates into loops. Loop-translated machine code will probably cause the execution pipeline to stall.
Rather than spending hundreds of thousands of transistors into building CPU-circuitry for handling looping portions of the micro-ops in the execution pipeline, they just handle it in some sort of crappy legacy-mode that stutterly stalls the pipeline, and ask modern programmers to write your own damn loops!
Therefore it is seldom used when machines write code. If you encounter REP in a binary executable, its probably a human assembly-muppet who didn't know better, or a cracker that really needed the few bytes it saved to use it instead of an actual loop, that wrote it.
(However. Take everything I just wrote with a grain of salt. Maybe this is not true anymore. I am not 100% up to date with the internals of x86 CPUs anymore, I got into other hobbies..)
I use the rep* prefix variants with cmps*, movs*, scas* and stos* instruction variants to generate inline code which minimizes the code size, avoids unnecessary calls/jumps and thereby keeps down the work done by the caches. The alternative is to set up parameters and call a memset or memcpy somewhere else which may overall be faster if I want to copy a hundred bytes or more but if it's just a matter of 10-20 bytes using rep is faster (or at least was the last time I measured).
Since my compiler allows specification and use of inline assembly functions and includes their register usage/modification in the optimization activities it is possible for me to use them when the circumstances are right.
On a historic note - not having any insight into the manufacturer's strategies - there was a time when the "rep movs*" (etc) instructions were very slow. I think it was around the time of the Pentium/Pentium MMX. A colleague of mine (who had more insight than I) said that the manufacturers had decreased the chip area (<=> fewer transistors/more microcode) allocated to the rep handling and used it to make other, more used instructions faster.
In the fifteen years or so since rep has become relatively speaking faster again which would suggest more transistors/less microcode.
Given 2 functions, which should be faster, if there is any difference at all? Assume that the input data is very large
void iterate1(const char* pIn, int Size)
{
for ( int offset = 0; offset < Size; ++offset )
{
doSomething( pIn[offset] );
}
}
vs
void iterate2(const char* pIn, int Size)
{
const char* pEnd = pIn+Size;
while(pIn != pEnd)
{
doSomething( *pIn++ );
}
}
Are there other issues to be considered with either approach?
Chances are, your compiler's optimizer will create a loop induction variable for the first case to turn it into the second. I'd expect no difference after optimizations so I tend to prefer the first style because I find it clearer to read.
Boojum is correct - IF your compiler has a good optimizer and you have it enabled. If that's not the case, or your use of arrays isn't sequential and liable to optimization, using array offsets can be far, far slower.
Here's an example. Back about 1988, we were implementing a window with a simple teletype interface on a Mac II. This consisted of 24 lines of 80 characters. When you got a new line in from the ticker, you scrolled up the top 23 lines and displayed the new one on the bottom. When there was something on the teletype, which wasn't all the time, it came in at 300 baud, which with the serial protocol overhead was about 30 characters per second. So we're not talking something that should have taxed a 16 MHz 68020 at all!
But the guy who wrote this did it like:
char screen[24][80];
and used 2-D array offsets to scroll the characters like this:
int i, j;
for (i = 0; i < 23; i++)
for (j = 0; j < 80; j++)
screen[i][j] = screen[i+1][j];
Six windows like this brought the machine to its knees!
Why? Because compilers were stupid in those days, so in machine language, every instance of the inner loop assignment, screen[i][j] = screen[i+1][j], looked kind of like this (Ax and Dx are CPU registers);
Fetch the base address of screen from memory into the A1 register
Fetch i from stack memory into the D1 register
Multiply D1 by a constant 80
Fetch j from stack memory and add it to D1
Add D1 to A1
Fetch the base address of screen from memory into the A2 register
Fetch i from stack memory into the D1 register
Add 1 to D1
Multiply D1 by a constant 80
Fetch j from stack memory and add it to D1
Add D1 to A2
Fetch the value from the memory address pointed to by A2 into D1
Store the value in D1 into the memory address pointed to by A1
So we're talking 13 machine language instructions for each of the 23x80=1840 inner loop iterations, for a total of 23920 instructions, including 3680 CPU-intensive integer multiplies.
We made a few changes to the C source code, so then it looked like this:
int i, j;
register char *a, *b;
for (i = 0; i < 22; i++)
{
a = screen[i];
b = screen[i+1];
for (j = 0; j < 80; j++)
*a++ = *b++;
}
There are still two machine-language multiplies, but they're in the outer loop, so there are only 46 integer multiplies instead of 3680. And the inner loop *a++ = *b++ statement only consisted of two machine-language operations.
Fetch the value from the memory address pointed to by A2 into D1, and post-increment A2
Store the value in D1 into the memory address pointed to by A1, and post-increment A1.
Given there are 1840 inner loop iterations, that's a total of 3680 CPU-cheap instructions - 6.5 times fewer - and NO integer multiplies. After this, instead of dying at six teletype windows, we never were able to pull up enough to bog the machine down - we ran out of teletype data sources first. And there are ways to optimize this much, much further, as well.
Now, modern compilers will do that kind of optimization for you - IF you ask them to do it, and IF your code is structured in a way that permits it.
But there are still circumstances where compilers can't do that for you - for instance, if you're doing non-sequential operations in the array.
So I've found it's served me well to use pointers instead of array references whenever possible. The performance is certainly never worse, and frequently much, much better.
With modern compiler there shouldn't be any difference in performance between the two, especially in such simplistic easily recognizable examples. Moreover, even if the compiler does not recognize their equivalence, i.e. translates each code "literally", there still shouldn't be any noticeable performance difference on a typical modern hardware platform. (Of course, there might be more specialized platforms out there where the difference might be noticeable.)
As for other considerations... Conceptually, when you implement an algorithm using the index access you impose a random-access requirement on the underlying data structure. When you use a pointer ("iterator") access, you only impose a sequential-access requirement on the underlying data structure. Random-access is a stronger requirement than sequential-access. For this reason I, for one, in my code prefer to stick to pointer access whenever possible, and use index access only when necessary.
More generally, if an algorithm can be implemented efficiently through sequential access, it is better to do it that way, without involving the unnecessary stronger requirement of random-access. This might prove useful in the future, should a need arise to refactor the code or to change the algorithm.
They are almost identical. Both solutions involve a temporary variable, an increment of a word on your system (int or ptr), and a logical check which should take one assembly instruction.
The only difference I see is the array lookup
arr[idx]
might require pointer arithmetic then a fetch while the dereference:
*ptr
just requires a fetch
My advice is that if it really matters, implement both and see if there's any savings.
To be sure, you must profile in your intended target environment.
That said, my guess is that any modern compiler is going to optimize them both down to very similar (if not identical) code.
If you didn't have an optimizer, the second has a chance of being faster, because you aren't re-computing the pointer on every iteration. But unless Size is a VERY large number (or the routine is called quite often), the difference isn't going to matter to your program's overall execution speed.
The pointer op used to be much faster. Now it's a bit faster, but the compiler may optimize it for you
Historically it was much faster to iterate via *p++ than p[i]; that was part of the motivation for having pointers in the language.
Plus, p[i] often requires a slower multiply op or at least a shift, so the optimization of replacing multiplies in a loop with adds to a pointer was sufficiently important to have a specific name: strength reduction. The subscript also tended to produce bigger code.
However, two things have changed: one is that compilers are much more sophisticated and are generally capable of doing this optimization for you.
The other is that the relative difference between an op and a memory access has increased. When *p++ was invented memory and cpu op times were similar. Today, a random desktop machine can do 3 billion integer ops / second, but only about 10 or 20 million random DRAM reads. Cache accesses are faster, and the system will prefetch and stream sequential memory accesses as you step through an array, but it still costs a lot to hit memory, and a bit of subscript fiddling isn't such a big deal.
Several years ago I asked this exact question. Someone in an interview was failing a candidate for picking the array notation because it was supposedly obviously slower. At that point I compiled both versions and looked at the disassembly. There was one opcode extra in the array notation. This was with Visual C++ (.net?). Based on what I saw I concluded that there is no appreciable difference.
Doing this again, here is what I found:
iterate1(arr, 400); // array notation
011C1027 mov edi,dword ptr [__imp__printf (11C20A0h)]
011C102D add esp,0Ch
011C1030 xor esi,esi
011C1032 movsx ecx,byte ptr [esp+esi+8] <-- Loop starts here
011C1037 push ecx
011C1038 push offset string "%c" (11C20F4h)
011C103D call edi
011C103F inc esi
011C1040 add esp,8
011C1043 cmp esi,190h
011C1049 jl main+32h (11C1032h)
iterate2(arr, 400); // pointer offset notation
011C104B lea esi,[esp+8]
011C104F nop
011C1050 movsx edx,byte ptr [esi] <-- Loop starts here
011C1053 push edx
011C1054 push offset string "%c" (11C20F4h)
011C1059 call edi
011C105B inc esi
011C105C lea eax,[esp+1A0h]
011C1063 add esp,8
011C1066 cmp esi,eax
011C1068 jne main+50h (11C1050h)
Why don't you try both and time them? My guess would be that they are optimized by the compiler into basically the same code. Just remember to turn on optimizations when comparing (-O3).
In the "other considerations" column, I'd say approach one is more clear. That's just my opinion though.
You're asking the wrong question. Should a developer aim for readability or performance first?
The first version is idiomatic for processing array, and your intent will be clear to anyone who has worked with arrays before, whereas the second relies heavily on the equivalence between array names and pointers, forcing someone reading the code to switch metaphors several times.
Cue the comments saying that the second version is crystal clear to any developer worth his keybaord.
If you wrote your program, and it's running slow, and you have profiled to the point where you have identified this loop as the bottleneck, then it would make sense to pop the hood and look at which of these is faster. But get something clear up and running first using well-known idiomatic language constructs.
Performance questions aside, it strikes me that the while loop variant has potential maintainability issues, as a programmer coming along to add some new bells and whistles has to remember to put the array increment in the right place, whereas the for loop variant puts it safely out of the body of the loop.
I was reading some old game programming books and as some of you might know, back in that day it was usually faster to do bit hacks than do things the standard way. (Converting float to int, mask sign bit, convert back for absolute value, instead of just calling fabs(), for example)
Nowadays is almost always better to just use the standard library math functions, since these tiny things are hardly the cause of most bottlenecks anyway.
But I still want to do a comparison, just for curiosity's sake. So I want to make sure when I profile, I'm not getting skewed results. As such, I'd like to make sure the compiler does not optimize out statements that have no side effect, such as:
void float_to_int(float f)
{
int i = static_cast<int>(f); // has no side-effects
}
Is there a way to do this? As far as I can tell, doing something like i += 10 will still have no side-effect and as such won't solve the problem.
The only thing I can think of is having a global variable, int dummy;, and after the cast doing something like dummy += i, so the value of i is used. But I feel like this dummy operation will get in the way of the results I want.
I'm using Visual Studio 2008 / G++ (3.4.4).
Edit
To clarify, I would like to have all optimizations maxed out, to get good profile results. The problem is that with this the statements with no side-effect will be optimized out, hence the situation.
Edit Again
To clarify once more, read this: I'm not trying to micro-optimize this in some sort of production code.
We all know that the old tricks aren't very useful anymore, I'm merely curious how not useful they are. Just plain curiosity. Sure, life could go on without me knowing just how these old hacks perform against modern day CPU's, but it never hurts to know.
So telling me "these tricks aren't useful anymore, stop trying to micro-optimize blah blah" is an answer completely missing the point. I know they aren't useful, I don't use them.
Premature quoting of Knuth is the root of all annoyance.
Assignment to a volatile variable shold never be optimized away, so this might give you the result you want:
static volatile int i = 0;
void float_to_int(float f)
{
i = static_cast<int>(f); // has no side-effects
}
So I want to make sure when I profile, I'm not getting skewed results. As such, I'd like to make sure the compiler does not optimize out statements
You are by definition skewing the results.
Here's how to fix the problem of trying to profile "dummy" code that you wrote just to test: For profiling, save your results to a global/static array and print one member of the array to the output at the end of the program. The compiler will not be able to optimize out any of the computations that placed values in the array, but you'll still get any other optimizations it can put in to make the code fast.
In this case I suggest you make the function return the integer value:
int float_to_int(float f)
{
return static_cast<int>(f);
}
Your calling code can then exercise it with a printf to guarantee it won't optimize it out. Also make sure float_to_int is in a separate compilation unit so the compiler can't play any tricks.
extern int float_to_int(float f)
int sum = 0;
// start timing here
for (int i = 0; i < 1000000; i++)
{
sum += float_to_int(1.0f);
}
// end timing here
printf("sum=%d\n", sum);
Now compare this to an empty function like:
int take_float_return_int(float /* f */)
{
return 1;
}
Which should also be external.
The difference in times should give you an idea of the expense of what you're trying to measure.
What always worked on all compilers I used so far:
extern volatile int writeMe = 0;
void float_to_int(float f)
{
writeMe = static_cast<int>(f);
}
note that this skews results, boith methods should write to writeMe.
volatile tells the compiler "the value may be accessed without your notice", thus the compiler cannot omit the calculation and drop the result. To block propagiation of input constants, you might need to run them through an extern volatile, too:
extern volatile float readMe = 0;
extern volatile int writeMe = 0;
void float_to_int(float f)
{
writeMe = static_cast<int>(f);
}
int main()
{
readMe = 17;
float_to_int(readMe);
}
Still, all optimizations inbetween the read and the write can be applied "with full force". The read and write to the global variable are often good "fenceposts" when inspecting the generated assembly.
Without the extern the compiler may notice that a reference to the variable is never taken, and thus determine it can't be volatile. Technically, with Link Time Code Generation, it might not be enough, but I haven't found a compiler that agressive. (For a compiler that indeed removes the access, the reference would need to be passed to a function in a DLL loaded at runtime)
Compilers are unfortunately allowed to optimise as much as they like, even without any explicit switches, if the code behaves as if no optimisation takes place. However, you can often trick them into not doing so if you indicate that value might be used later, so I would change your code to:
int float_to_int(float f)
{
return static_cast<int>(f); // has no side-effects
}
As others have suggested, you will need to examine the assemnler output to check that this approach actually works.
You just need to skip to the part where you learn something and read the published Intel CPU optimisation manual.
These quite clearly state that casting between float and int is a really bad idea because it requires a store from the int register to memory followed by a load into a float register. These operations cause a bubble in the pipeline and waste many precious cycles.
a function call incurs quite a bit of overhead, so I would remove this anyway.
adding a dummy += i; is no problem, as long as you keep this same bit of code in the alternate profile too. (So the code you are comparing it against).
Last but not least: generate asm code. Even if you can not code in asm, the generated code is typically understandable since it will have labels and commented C code behind it. So you know (sortoff) what happens, and which bits are kept.
R
p.s. found this too:
inline float pslNegFabs32f(float x){
__asm{
fld x //Push 'x' into st(0) of FPU stack
fabs
fchs //change sign
fstp x //Pop from st(0) of FPU stack
}
return x;
}
supposedly also very fast. You might want to profile this too. (although it is hardly portable code)
Return the value?
int float_to_int(float f)
{
return static_cast<int>(f); // has no side-effects
}
and then at the call site, you can sum all the return values up, and print out the result when the benchmark is done. The usual way to do this is to somehow make sure you depend on the result.
You could use a global variable instead, but it seems like that'd generate more cache misses. Usually, simply returning the value to the caller (and making sure the caller actually does something with it) does the trick.
If you are using Microsoft's compiler - cl.exe, you can use the following statement to turn optimization on/off on a per-function level [link to doc].
#pragma optimize("" ,{ on |off })
Turn optimizations off for functions defined after the current line:
#pragma optimize("" ,off)
Turn optimizations back on:
#pragma optimize("" ,on)
For example, in the following image, you can notice 3 things.
Compiler optimizations flag is set - /O2, so code will get optimized.
Optimizations are turned off for first function - square(), and turned back on before square2() is defined.
Amount of assembly code generated for 1st function is higher. In second function there is no assembly code generated for int i = num; statement in code.
Thus while 1st function is not optimized, the second function is.
See https://godbolt.org/z/qJTBHg for link to this code on compiler explorer.
A similar directive exists for gcc too - https://gcc.gnu.org/onlinedocs/gcc/Function-Specific-Option-Pragmas.html
A micro-benchmark around this statement will not be representative of using this approach in a genuine scenerio; the surrounding instructions and their affect on the pipeline and cache are generally as important as any given statement in itself.
GCC 4 does a lot of micro-optimizations now, that GCC 3.4 has never done. GCC4 includes a tree vectorizer that turns out to do a very good job of taking advantage of SSE and MMX. It also uses the GMP and MPFR libraries to assist in optimizing calls to things like sin(), fabs(), etc., as well as optimizing such calls to their FPU, SSE or 3D Now! equivalents.
I know the Intel compiler is also extremely good at these kinds of optimizations.
My suggestion is to not worry about micro-optimizations like this - on relatively new hardware (anything built in the last 5 or 6 years), they're almost completely moot.
Edit: On recent CPUs, the FPU's fabs instruction is far faster than a cast to int and bit mask, and the fsin instruction is generally going to be faster than precalculating a table or extrapolating a Taylor series. A lot of the optimizations you would find in, for example, "Tricks of the Game Programming Gurus," are completely moot, and as pointed out in another answer, could potentially be slower than instructions on the FPU and in SSE.
All of this is due to the fact that newer CPUs are pipelined - instructions are decoded and dispatched to fast computation units. Instructions no longer run in terms of clock cycles, and are more sensitive to cache misses and inter-instruction dependencies.
Check the AMD and Intel processor programming manuals for all the gritty details.