Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Bubble sort slower with -O3 than -O2 with GCC

I made a bubble sort implementation in C, and was testing its performance when I noticed that the -O3 flag made it run even slower than no flags at all! Meanwhile -O2 was making it run a lot faster as expected.

Without optimisations:

time ./sort 30000

./sort 30000  1.82s user 0.00s system 99% cpu 1.816 total

-O2:

time ./sort 30000

./sort 30000  1.00s user 0.00s system 99% cpu 1.005 total

-O3:

time ./sort 30000

./sort 30000  2.01s user 0.00s system 99% cpu 2.007 total

The code:

#include <stdio.h>
#include <stdlib.h>
#include <stdbool.h>
#include <time.h>

int n;

void bubblesort(int *buf)
{
    bool changed = true;
    for (int i = n; changed == true; i--) { /* will always move at least one element to its rightful place at the end, so can shorten the search by 1 each iteration */
        changed = false;

        for (int x = 0; x < i-1; x++) {
            if (buf[x] > buf[x+1]) {
                /* swap */
                int tmp = buf[x+1];
                buf[x+1] = buf[x];
                buf[x] = tmp;

                changed = true;
            }
        }
    }
}

int main(int argc, char *argv[])
{
    if (argc != 2) {
        fprintf(stderr, "Usage: %s <arraysize>\n", argv[0]);
        return EXIT_FAILURE;
    }

    n = atoi(argv[1]);
    if (n < 1) {
        fprintf(stderr, "Invalid array size.\n");
        return EXIT_FAILURE;
    }

    int *buf = malloc(sizeof(int) * n);

    /* init buffer with random values */
    srand(time(NULL));
    for (int i = 0; i < n; i++)
        buf[i] = rand() % n + 1;

    bubblesort(buf);

    return EXIT_SUCCESS;
}

The assembly language generated for -O2 (from godbolt.org):

bubblesort:
        mov     r9d, DWORD PTR n[rip]
        xor     edx, edx
        xor     r10d, r10d
.L2:
        lea     r8d, [r9-1]
        cmp     r8d, edx
        jle     .L13
.L5:
        movsx   rax, edx
        lea     rax, [rdi+rax*4]
.L4:
        mov     esi, DWORD PTR [rax]
        mov     ecx, DWORD PTR [rax+4]
        add     edx, 1
        cmp     esi, ecx
        jle     .L2
        mov     DWORD PTR [rax+4], esi
        mov     r10d, 1
        add     rax, 4
        mov     DWORD PTR [rax-4], ecx
        cmp     r8d, edx
        jg      .L4
        mov     r9d, r8d
        xor     edx, edx
        xor     r10d, r10d
        lea     r8d, [r9-1]
        cmp     r8d, edx
        jg      .L5
.L13:
        test    r10b, r10b
        jne     .L14
.L1:
        ret
.L14:
        lea     eax, [r9-2]
        cmp     r9d, 2
        jle     .L1
        mov     r9d, r8d
        xor     edx, edx
        mov     r8d, eax
        xor     r10d, r10d
        jmp     .L5

And the same for -O3:

bubblesort:
        mov     r9d, DWORD PTR n[rip]
        xor     edx, edx
        xor     r10d, r10d
.L2:
        lea     r8d, [r9-1]
        cmp     r8d, edx
        jle     .L13
.L5:
        movsx   rax, edx
        lea     rcx, [rdi+rax*4]
.L4:
        movq    xmm0, QWORD PTR [rcx]
        add     edx, 1
        pshufd  xmm2, xmm0, 0xe5
        movd    esi, xmm0
        movd    eax, xmm2
        pshufd  xmm1, xmm0, 225
        cmp     esi, eax
        jle     .L2
        movq    QWORD PTR [rcx], xmm1
        mov     r10d, 1
        add     rcx, 4
        cmp     r8d, edx
        jg      .L4
        mov     r9d, r8d
        xor     edx, edx
        xor     r10d, r10d
        lea     r8d, [r9-1]
        cmp     r8d, edx
        jg      .L5
.L13:
        test    r10b, r10b
        jne     .L14
.L1:
        ret
.L14:
        lea     eax, [r9-2]
        cmp     r9d, 2
        jle     .L1
        mov     r9d, r8d
        xor     edx, edx
        mov     r8d, eax
        xor     r10d, r10d
        jmp     .L5

It seems like the only significant difference to me is the apparent attempt to use SIMD, which seems like it should be a large improvement, but I also can't tell what on earth it's attempting with those pshufd instructions... is this just a failed attempt at SIMD? Or maybe the couple of extra instructions is just about edging out my instruction cache?

Timings were done on an AMD Ryzen 5 3600.

like image 669
anon Avatar asked Oct 13 '22 20:10

anon


People also ask

Why is bubble sort slower than insertion?

On average, the bubble sort performs poorly compared to the insertion sort. Due to the high number of swaps, it's expected to generate twice as many write operations and twice as many cache misses. Therefore, we don't prefer this algorithm for an ordinary sorting job.

Is O log n )) sorting algorithms always faster than O n2 algorithms?

No, O(n log n) algorithms are not always better than O(n^2) ones. The Big-O notation describes an upper bound of the asymptotic behavior of an algorithm, i.e. for n that tends towards infinity.

Why is STD sort faster than qsort?

STL's sort ran faster than C's qsort, because C++'s templates generate optimized code for a particular data type and a particular comparison function. STL's sort runs 20% to 50% faster than the hand-coded quicksort and 250% to 1000% faster than the C qsort library function.

Is bubble sort the slowest algorithm?

In computer science, bogosort (also known as permutation sort, stupid sort, or slowsort) is a sorting algorithm based on the generate and test paradigm. The function successively generates permutations of its input until it finds one that is sorted.

Is bubble sort really that bad?

A Bubble Sort (with an early-out for a pass that didn't do any swaps) is also not horrible in some almost-sorted cases but is worse than Insertion Sort. But an element can only move toward the front of the list one step per pass, so if the smallest element is near the end but otherwise fully sorted, it still takes Bubble Sort O (N^2) work.

What is the time complexity of bubble sort?

Bubble sort runs in O (n^2) time complexity. Merge sort takes O (n*log (n)) time, while quick sort takes O (n*log (n)) time on average, thus performing better than bubble sort. Refer to this: complexity of bubble sort.

Can you explain the O(n^2) sort algorithm?

If you can explain by c programming or assembler commands it would be much appreciated. There are lots of O (n^2) sort algorithms (bubble sort, insertion sort, selection sort, etc.) that are mainly used for teaching purposes because they tend to be easier to understand than the better sort algorithms that people actually use.

How efficient is the standard library sort in JavaScript?

The standard library sort should (usually?) be well-tuned for the JS interpreter / JIT to handle efficiently, and use an efficient implementation of an efficient algorithm. The rest of this answer is assuming a use-case like sorting an array of integers in an ahead-of-time compiled language like C compiled to native asm.


Video Answer


1 Answers

This is a regression in GCC11/12.
GCC10 and earlier were doing separate dword loads, even if it merged for a qword store.

It looks like GCC's naïveté about store-forwarding stalls is hurting its auto-vectorization strategy here. See also Store forwarding by example for some practical benchmarks on Intel with hardware performance counters, and What are the costs of failed store-to-load forwarding on x86? Also Agner Fog's x86 optimization guides.

(gcc -O3 enables -ftree-vectorize and a few other options not included by -O2, e.g. if-conversion to branchless cmov, which is another way -O3 can hurt with data patterns GCC didn't expect. By comparison, Clang enables auto-vectorization even at -O2, although some of its optimizations are still only on at -O3.)

It's doing 64-bit loads (and branching to store or not) on pairs of ints. This means, if we swapped the last iteration, this load comes half from that store, half from fresh memory, so we get a store-forwarding stall after every swap. But bubble sort often has long chains of swapping every iteration as an element bubbles far, so this is really bad.

(Bubble sort is bad in general, especially if implemented naively without keeping the previous iteration's second element around in a register. It can be interesting to analyze the asm details of exactly why it sucks, so it is fair enough for wanting to try.)

Anyway, this is pretty clearly an anti-optimization you should report on GCC Bugzilla with the "missed-optimization" keyword. Scalar loads are cheap, and store-forwarding stalls are costly. (Can modern x86 implementations store-forward from more than one prior store? no, nor can microarchitectures other than in-order Atom efficiently load when it partially overlaps with one previous store, and partially from data that has to come from the L1d cache.)

Even better would be to keep buf[x+1] in a register and use it as buf[x] in the next iteration, avoiding a store and load. (Like good hand-written asm bubble sort examples, a few of which exist on Stack Overflow.)

If it wasn't for the store-forwarding stalls (which AFAIK GCC doesn't know about in its cost model), this strategy might be about break-even. SSE 4.1 for a branchless pmind / pmaxd comparator might be interesting, but that would mean always storing and the C source doesn't do that.


If this strategy of double-width load had any merit, it would be better implemented with pure integer on a 64-bit machine like x86-64, where you can operate on just the low 32 bits with garbage (or valuable data) in the upper half. E.g.,

## What GCC should have done,
## if it was going to use this 64-bit load strategy at all

        movsx   rax, edx           # apparently it wasn't able to optimize away your half-width signed loop counter into pointer math
        lea     rcx, [rdi+rax*4]   # Usually not worth an extra instruction just to avoid an indexed load and indexed store, but let's keep it for easy comparison.
.L4:
        mov     rax, [rcx]       # into RAX instead of XMM0
        add     edx, 1
            #  pshufd  xmm2, xmm0, 0xe5
            #  movd    esi, xmm0
            #  movd    eax, xmm2
            #  pshufd  xmm1, xmm0, 225
        mov     rsi, rax
        rol     rax, 32   # swap halves, just like the pshufd
        cmp     esi, eax  # or eax, esi?  I didn't check which is which
        jle     .L2
        movq    QWORD PTR [rcx], rax   # conditionally store the swapped qword

(Or with BMI2 available from -march=native, rorx rsi, rax, 32 can copy-and-swap in one uop. Without BMI2, mov and swapping the original instead of the copy saves latency if running on a CPU without mov-elimination, such as Ice Lake with updated microcode.)

So total latency from load to compare is just integer load + one ALU operation (rotate). Vs. XMM load -> movd. And its fewer ALU uops. This does nothing to help with the store-forwarding stall problem, though, which is still a showstopper. This is just an integer SWAR implementation of the same strategy, replacing 2x pshufd and 2x movd r32, xmm with just mov + rol.

Actually, there's no reason to use 2x pshufd here. Even if using XMM registers, GCC could have done one shuffle that swapped the low two elements, setting up for both the store and movd. So even with XMM regs, this was sub-optimal. But clearly two different parts of GCC emitted those two pshufd instructions; one even printed the shuffle constant in hex while the other used decimal! I assume one swapping and the other just trying to get vec[1], the high element of the qword.


slower than no flags at all

The default is -O0, consistent-debugging mode that spills all variables to memory after every C statement, so it's pretty horrible and creates big store-forwarding latency bottlenecks. (Somewhat like if every variable was volatile.) But it's successful store forwarding, not stalls, so "only" ~5 cycles, but still much worse than 0 for registers. (A few modern microarchitectures including Zen 2 have some special cases that are lower latency). The extra store and load instructions that have to go through the pipeline don't help.

It's generally not interesting to benchmark -O0. -O1 or -Og should be your go-to baseline for the compiler to do the basic amount of optimization a normal person would expect, without anything fancy, but also not intentionally gimp the asm by skipping register allocation.


Semi-related: optimizing bubble sort for size instead of speed can involve memory-destination rotate (creating store-forwarding stalls for back-to-back swaps), or a memory-destination xchg (implicit lock prefix -> very slow). See this Code Golf answer.

like image 171
Peter Cordes Avatar answered Nov 07 '22 17:11

Peter Cordes