Re: [Axiom-developer] MACOSX

[Camm Maguire]

Greetings!

megaera:books camm$ diff -u bookvol7.pamphlet~ bookvol7.pamphlet
--- bookvol7.pamphlet~  2014-07-10 13:21:20.000000000 -0500
+++ bookvol7.pamphlet   2014-07-17 12:04:06.000000000 -0500
@@ -11251,8 +11251,8 @@
                     XNextEvent(gXDisplay, &event);
                     handleEvent(&event);
                 }
-                else if FD_ISSET
-                    (spadSocket->socket, &rd)
+                else if (FD_ISSET
+                    (spadSocket->socket, &rd))
                     /*
                      * Axiom Socket do what handleEvent does The 100 is
                      * $SpadStuff in hypertex.boot
@@ -23066,11 +23066,11 @@
          */
         bsdSignal(SIGUSR2, sigusr2Handler,RestartSystemCalls);
         bsdSignal(SIGUSR1, SIG_IGN,RestartSystemCalls);
-#if defined(BSDplatform) || defined(MACOSXplatform)
+/*#if defined(BSDplatform) || defined(MACOSXplatform)*/
         bsdSignal(SIGCHLD, sigcldHandler,RestartSystemCalls);
-#else
+/*#else
         bsdSignal(SIGCLD, sigcldHandler,RestartSystemCalls);
-#endif
+#endif*/
         bsdSignal(SIGINT, SIG_IGN,RestartSystemCalls);
         /*
          * Now go to the main event loop. I will never return, so just end

The second hunk is probably not what you want, but MACOSXplatform was
not defined for hypertex.c, and I could not see where it was supposed to
be defined.

With this, I'm in the middle of the regression tests on mac.

Take care,


u1204 <address@hidden> writes:

> Camm,
>
>>Greetings!  I get through the mac build up to this non-gcl failure:
>
>>viewman.c: In function 'rmViewMgr':
>>viewman.c:546: warning: implicit declaration of function 'close'
>
> Sigh. Another gcc change that will require me to figure out the
> correct set of switches. It is hard to write stable software on
> so many moving platforms.
>
>>This is with git tag cygwin at the moment.  I will verify 2_6_10
>>afterwards.
>
>>Separately, I usually run with TESTSET=regresstests.  A simple 'make'
>>with no arguments seems to be taking forever on x86 on richtrig800-899,
>>some huge bignum calculation perhaps.  trace is spooling gc call
>>notifications.  How long should this test take?
>
> A straight 'make' call will run all of the tests, including the 
> CATS (Computer Algebra Test Suite), which can take many hours.
> I need to change the default not to run that suite.
>
> Tim
>
>
> _______________________________________________
> Axiom-developer mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/axiom-developer
>
>
>
>

-- 
Camm Maguire                                        address@hidden
==========================================================================
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah