CanUSE_PTHREADS and MINIMAL_RUNTIME exist at the same time?

[tigercosmos]

CanUSE_PTHREADS and MINIMAL_RUNTIME exist at the same time?

I have a simple pthread code test.c.

I run em with the following arguments:

emcc test.c \
		-s USE_PTHREADS=1 \
		-s PTHREAD_POOL_SIZE=4 \
		-s ENVIRONMENT=web,worker \
		-s MINIMAL_RUNTIME \
		-O3 \
		-o test.html

Then I got an error:

The code works well when I don't set `MINIMAL_RUNTIME`.

I use version 2.0.7.

test.c looks like:

#include <pthread.h>
#include <stdio.h>
#include <stdlib.h>

#define NUMTHRDS 4
#define MAGNIFICATION 1e9

typedef struct
{
    double *sum;
} Arg;

pthread_t callThd[NUMTHRDS];

void *count_pi(void *arg)
{

    Arg *data = (Arg *)arg;
    double *sum = data->sum;
    *sum += 1;

    printf("sum=%f\n", *sum);

    pthread_exit((void *)0);
}

int main(int argc, char *argv[])
{
    pthread_attr_t attr;
    pthread_attr_init(&attr);
    pthread_attr_setdetachstate(&attr, PTHREAD_CREATE_JOINABLE);

    double *sum = malloc(sizeof(*sum));
    *sum = 0;

    Arg arg[NUMTHRDS];
    for (int i = 0; i < NUMTHRDS; i++)
    {
        arg[i].sum = sum;
        pthread_create(&callThd[i], &attr, count_pi, (void *)&arg[i]);
    }

    pthread_attr_destroy(&attr);

    void *status;
    for (int i = 0; i < NUMTHRDS; i++)
    {
        pthread_join(callThd[i], &status);
    }

    printf("Final Sum =  %f \n", *sum);

    free(sum);
}

[tigercosmos]

I found the problem:

it's because there is inrelevant code not been deleted:

function _emscripten_futex_wait(addr, val, timeout) {
    if (addr <= 0 || addr > HEAP8.length || addr & 3 != 0) return -28;
    if (!ENVIRONMENT_IS_WEB) {
        var ret = Atomics.wait(HEAP32, addr >> 2, val, timeout);
        if (ret === "timed-out") return -73;
        if (ret === "not-equal") return -6;
        if (ret === "ok") return 0;
        throw "Atomics.wait returned an unexpected value " + ret
    } else {

Same problem here, no need ENVIRONMENT_IS_NODE:

function _pthread_exit(status) {
    if (!ENVIRONMENT_IS_PTHREAD) _exit(status);
    else PThread.threadExit(status);
    if (ENVIRONMENT_IS_NODE) {
        process.exit(status)
    }
    throw "unwind"
}

I can help open a PR, but I would need some hints!

[kripken]

cc @juj

Yes, I think this is not done yet, and PRs would be great of course!

Is the issue that ENVIRONMENT_IS_WEB etc. are not defined? I'm not sure offhand how best to fix that. If it's a tiny ifdef maybe that's ok, but long-term we should be able to avoid all of this using #12203

[tigercosmos]

Yes, ENVIRONMENT_IS_WEB and ENVIRONMENT_IS_NODE are not defined in the output code, but they also should not exist at all.

I fix the issue by

	emcc test.c \
		-s USE_PTHREADS=1 \
		-s PTHREAD_POOL_SIZE=4 \
		-s ENVIRONMENT=web,worker \
		-s MINIMAL_RUNTIME \
		-O3 \
		-o test.html
	mv test.js tmp.js
	sed 's/ENVIRONMENT_IS_NODE/false/g' tmp.js | \
	sed 's/ENVIRONMENT_IS_WEB/true/g' > test.js

[stale]

This issue has been automatically marked as stale because there has been no activity in the past year. It will be closed automatically if no further activity occurs in the next 30 days. Feel free to re-open at any time if this issue is still relevant.

[tigercosmos]

@kripken I haven't checked this issue for a long time. Not sure if this issue still an issue.

[kripken]

I believe this should be fixed, as we have tests for it now.