From: stephen gooberman-hill <stevegoobermanhill@...>
Date: 2012-02-03T06:04:29+09:00
Subject: [ruby-core:42328] [ruby-trunk - Bug #5957] pthreads not working on ulibc (linuxthreads)


Issue #5957 has been updated by stephen gooberman-hill.


Hi Motohiro,
thanks for taking a look :-)

I've been doing some more digging and I think my initial thoughts were wrong. I am now pretty sure that this is actually an issue around the uClibc build on my ARM board - basically it has been built with pretty much every possible pre-processor directive switched off, so it is not supporting sufficient functionality to enable pthreads to run sucessfully ( pthread_getattr_np is defined in a #ifdef block but not built).

I'm in touch with the board manufacturers (Techbase - it is an NPE series industrial computer / gsm/gprs/edge modem). If I can get ruby running (with thread support - as I need TCPServer) then I am looking at bulk orders - so they are (at the moment) being very helpful.

Kind regards

Steve

----------------------------------------
Bug #5957: pthreads not working on ulibc (linuxthreads)
https://bugs.ruby-lang.org/issues/5957

Author: stephen gooberman-hill
Status: Rejected
Priority: Normal
Assignee: 
Category: 
Target version: 
ruby -v: ruby 1.9.2p290 (2011-07-09 revision 32553) [arm-linux]


Hi
I have an arm-linux crosscompile of ruby 1.9.2p290 and 1.9.3p0

Trying to implement a thread (via Thrad.new) causes an error to be thrown
"ruby engine can initialize only in the main thread"

and a stack trace is thrown.

Tracing this issue (in 1.9.2p290) leads to thread_pthread.c lines 350-388

static int
native_thread_init_stack(rb_thread_t *th)
{
    rb_thread_id_t curr = pthread_self();

    if (pthread_equal(curr, native_main_thread.id)) {
	th->machine_stack_start = native_main_thread.stack_start;
	th->machine_stack_maxsize = native_main_thread.stack_maxsize;
    }
    else {
#ifdef STACKADDR_AVAILABLE
	void *start;
	size_t size;

	if (get_stack(&start, &size) == 0) {
	    th->machine_stack_start = start;
	    th->machine_stack_maxsize = size;
	}
#else
	rb_raise(rb_eNotImpError, "ruby engine can initialize only in the main thread");
#endif
    }
#ifdef __ia64
    th->machine_register_stack_start = native_main_thread.register_stack_start;
    th->machine_stack_maxsize /= 2;
    th->machine_register_stack_maxsize = th->machine_stack_maxsize;
#endif
    return 0;
}

I have done some digging, comparing by build machine on which I run ruby1.9.2p290 built from source, and my arm-linux target machine

STACKADDR_AVAILABLE is defined based on a whole set of possible conditions higher in the code (lines 221-229)
#if defined HAVE_PTHREAD_GETATTR_NP || defined HAVE_PTHREAD_ATTR_GET_NP
#define STACKADDR_AVAILABLE 1
#elif defined HAVE_PTHREAD_GET_STACKADDR_NP && defined HAVE_PTHREAD_GET_STACKSIZE_NP
#define STACKADDR_AVAILABLE 1
#elif defined HAVE_THR_STKSEGMENT || defined HAVE_PTHREAD_STACKSEG_NP
#define STACKADDR_AVAILABLE 1
#elif defined HAVE_PTHREAD_GETTHRDS_NP
#define STACKADDR_AVAILABLE 1
#endif

I have done some greping around /usr/includes, and I can't find any of these definitions.

So I assume that on my i686 build system ruby is running on the native_main_thread, but on the arm system it is not.

Is this the case. Any ideas on how to persuade ruby to run on the native main thread. I can probably rewrite a few critical sections of my app to run of fibers if necessary....but I'd prefer to have thread support

Kind regards

Steve G-H







-- 
http://bugs.ruby-lang.org/