Re: new branch in gm2 git repro for 10.0.2

[Gaius Mulley]

Dragiša Durić <dragisha@m3w.org> writes:

> Hi Gaius,
>
> Built from Savannah... It looks like not only RTentity's thread-safety
> is problematic. This is what happens mst often.
>
> (gdb) r
> The program being debugged has been started already.
> Start it from the beginning? (y or n) y
> Starting program: /home/dragisha/bluewd-m2/gm2.lib0/a.out 
> [Thread debugging using libthread_db enabled]
> Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
> [New Thread 0x7ffff5ce5700 (LWP 16752)]
>
> Thread 1 "a.out" received signal SIGSEGV, Segmentation fault.
> _int_malloc (av=av@entry=0x7ffff6d24c40, bytes=bytes@entry=16385) at 
> malloc.c:4100
> 4100  malloc.c: No such file or directory.
> (gdb) bt
> #0  _int_malloc (av=av@entry=0x7ffff6d24c40, bytes=bytes@entry=16385) at 
> malloc.c:4100
> #1  0x00007ffff69d035d in __GI___libc_malloc (bytes=16385) at malloc.c:3065
> #2  0x0000000000475974 in Storage_ALLOCATE (addr=@0x7fffffffde08: 
> 0x7fff00000029, amount=16385)
>     at 
> ../../../../gcc-versionno/libgm2/libiso/../../gcc/gm2/gm2-libs-iso/Storage.mod:36
> #3  0x000000000041ecd5 in DoFormat (res=..., args=0x725410) at hStrIO.mod:205
> #4  0x000000000041e70c in hStrIO_PrintF (res=..., args=0x725410) at 
> hStrIO.mod:476
> #5  0x000000000046ad2f in AdHoc_Init (getyear=1, top=1) at AdHoc.mod:84
> #6  0x0000000000472144 in _M2_test5_init () at test5.mod:41
> #7  0x0000000000405ff1 in init (argc=1, argv=0x7fffffffe148) at 
> test5_m2.cpp:282
> #8  0x0000000000406249 in main (argc=1, argv=0x7fffffffe148) at 
> test5_m2.cpp:392
>
>
> Regards,
> Dragiša
>
> On Sat, 2020-08-08 at 13:18 +0100, Gaius Mulley wrote:
>> Dragiša Durić <dragisha@m3w.org> writes:
>> 
>> > Hi Gaius,
>> > 
>> > I was off for some time (other obligations) and just jumped into
>> > this
>> > again. Am I correct here - thredsfe chnges are not yet in gm2-
>> > master? 
>> > 
>> > How can I build with them?
>> > 
>> > Thanks,
>> > Dragisa
>> 
>> Hi Dragisa,
>> 
>> I've written a new regression test:
>> 
>>   gcc/testsuite/gm2/iso/run/pass/concurrentstore.mod
>> 
>> which exposed a few bugs in Processes.mod.  I've now fixed these bugs
>> and applied the fixes to the master and gcc_10_2_0_gm2 branch on
>> savannah.
>> 
>> regards,
>> Gaius

Hi Dragisa,

many thanks for the stack trace - I'll examine RTentity now.  I've
overhauled the underlying ISO process/coroutines code and added some
extra concurrency regression tests.  I'm assuming you still get problems
with RTentity with the latest commits?

I've applied the latest changes to the master and gcc-10 branch,

regards,
Gaius