Web lists-archives.com

Re: [Mingw-users] GCC-4.9.3 is now our current release




Damn it runs right to the end and then fails on objc here

thread_id = (objc_thread_t) new_thread_handle; cannot convert to a 
referral type

with an error about it not being able to convert it.

objc_thread_t thread_id; = scalar
pthread_t new_thread_handle; =  struct

this is from gthr-default.h

guess its back to the drawing board.

Den 16-03-2016 kl. 23:08 skrev K. Frank:
> Hi Ralph!
>
> On Wed, Mar 16, 2016 at 5:11 PM, ralph engels <ralphengels@xxxxxxxxx> wrote:
>> Agree your version seems more logic to me also.
>> the x from the ptw32_handle_t struct pointer is as far as i could glean
>> used for thread counters, if say we need a thread to run a specified
>> time, and it could collide with some pointers in std, i tried your
>> version and gcc is bootstrapping, ill know if it failed in some hours :)
>> but the error from thread is atleast gone.
> Yeah, I wish I understood better what that x member is for.  I'd say that
> my implementation of operator< is a legitimate educated guess, but since
> I don't really understand what is going on, I don't have a lot of confidence
> that it's right.
>
>> I think i found the culprit of the thread lock, i was compiling
>> pthreads-w32 with a posix thread compiler based on winpthreads which
>> seems to have introduced a race condition, after recompiling it with a
>> non posix thread built mingw it runs and i have not  seen any problems
>> so far.
> That's good news.  Those lockup /. race conditions can be annoying to
> find and debug.
>
>> Its getting there :)
> I am eager to hear how this all works out.
>
> By the way, I'm a little unclear on your motivation.  Are you trying to compile
> gcc so it can use threads when it runs?  (Maybe a dumb question.  I don't
> know anything about gcc's internals.)  Or are your trying to get things working
> so that an end user of gcc can build applications that use pthreads and / or
> std::thread?
>
> If the latter, I seem to recall that the gcc code base had  (back in the day)
> some unit tests for threading -- maybe std::thread specific.  I didn't have the
> strength to set up their testing framework and run them all automatically, but
> I did run a few manually.  Perhaps you could run the threading tests on your
> build of gcc to shake things down a little.
>
>
> Good luck!
>
>
> K. Frank
>
> ------------------------------------------------------------------------------
> Transform Data into Opportunity.
> Accelerate data analysis in your applications with
> Intel Data Analytics Acceleration Library.
> Click to learn more.
> http://pubads.g.doubleclick.net/gampad/clk?id=278785231&iu=/4140
> _______________________________________________
> MinGW-users mailing list
> MinGW-users@xxxxxxxxxxxxxxxxxxxxx
>
> This list observes the Etiquette found at
> http://www.mingw.org/Mailing_Lists.
> We ask that you be polite and do the same.  Disregard for the list etiquette may cause your account to be moderated.
>
> _______________________________________________
> You may change your MinGW Account Options or unsubscribe at:
> https://lists.sourceforge.net/lists/listinfo/mingw-users
> Also: mailto:mingw-users-request@xxxxxxxxxxxxxxxxxxxxx?subject=unsubscribe


------------------------------------------------------------------------------
Transform Data into Opportunity.
Accelerate data analysis in your applications with
Intel Data Analytics Acceleration Library.
Click to learn more.
http://pubads.g.doubleclick.net/gampad/clk?id=278785231&iu=/4140
_______________________________________________
MinGW-users mailing list
MinGW-users@xxxxxxxxxxxxxxxxxxxxx

This list observes the Etiquette found at 
http://www.mingw.org/Mailing_Lists.
We ask that you be polite and do the same.  Disregard for the list etiquette may cause your account to be moderated.

_______________________________________________
You may change your MinGW Account Options or unsubscribe at:
https://lists.sourceforge.net/lists/listinfo/mingw-users
Also: mailto:mingw-users-request@xxxxxxxxxxxxxxxxxxxxx?subject=unsubscribe