Issues with ASYNC_pause_job() wake up

["Valerio Di Gregorio \(vadigreg\) via openssl-users" <openssl-users@xxxxxxxxxxx>]

Hello,

This is my first post here. I need help with ASYNC_pause_job(). I’m writing an async engine to delegate certificate validation to a different process. Validation happens asynchronously through IPCs. To explain what I’m doing I’ll use some “pseudo” code:

 

// this happens in process #1

ctx = OPENSSL_malloc(sizeof(async_engine_ctx_t));

ctx->job = ASYNC_get_current_job()

ctx->wait_ctx = ASYNC_get_wait_ctx(job)

pipe(ctx->fds);

ASYNC_WAIT_CTX_set_wait_fd(ctx->wait_ctx, ctx, ctx->fds[0], ctx, async_engine_cleanup_cb);

 

async_validation_request(ctx->fds[1], cert, cert_len); // this starts async validation via IPC

 

printf(“pause”)

ASYNC_pause_job();

printf(“resume”)

 

read(ctx->fds[0], &validation_result, sizeof(validation_result));

 

For simplicity I omitted error checks (in my runs I have no error in any of the mentioned functions). When async response comes in, via IPC, what I do is:

 

// this happens in process #1

write(write_fd, &status, sizeof(status)); // I verified write_fd == ctx->fds[1]

 

I see 2 different issues with this:

1. ASYNC_pause_job() can wake up before write(). It will then block on the read(), which is too bad in my single-threaded code.
2. I fixed case 1 by making read() non-blocking, I then run ASYNC_pause_job() again and again until write() is actually performed. So now I hit another issue. Time between printf(“pause”) and write() is ~100/200 milliseconds. However time between write() and printf(“resume”) is usually ~4.5 seconds and this way too much.

 

Is there a reliable way (only when data is really ready) and a prompt way (in order of micro/milli-seconds) to wake up from ASYNC_pause_job()?

 

Many thanks, any help would be much appreciated.

 

Kind Regards,

Valerio Di Gregorio