On 27/03/2020 15:58, jameszxj wrote:
> Hi all,
> ?0?2 ?0?2 ?0?2 ?0?2 Since rtems-libbsd is too heavy for our project, so I
> tried to
> port?0?2Lightweight TCP/IP stacks.?0?2 First I tried lwip, it worked but
> crashed now and then.
> The PC pointed at function?0?2_Heap_Block_split()??I could not figure it
> out. So I tried cycloneTCP . It is more simple than lwip.?0?2 But I
> encountered the same problem.
> System still crashed at?0?2_Heap_Block_split()?0?2sometimes
> (random),?0?2?0?2It made
> me crazy :(
> ?0?2 ?0?2 ?0?2 ?0?2I tried on diffrent CPUs (zynq7020, sunxi-t3), the problem
> was
> still there.?0?2?0?2_Heap_Block_split() is a core function, It seems to have
> nothing to do with the stack.
> Attachments are error messages and?0?2disassembly code.?0?2Any advices?
> ?0?2 ?0?2 ?0?2 ?0?2Thanks.
>
>
>
Hello,
I'm quite sure that I've seen some mails regarding lwIP. So there should
be already users for that.
Without having a detailed look at the crash message: If there are
problems with the heap you might want to use
--enable-rtems-debug
during RTEMS configure and maybe add
#define CONFIGURE_STACK_CHECKER_ENABLED
to your configuration. Problems with the heap can be from some part of
the application writing over the end of a buffer or similar. If that
destroys heap structures, you might get problems as soon as RTEMS want's
to do something with the heap. The rtems_debug enables quite a lot of
run time checks like a heap protection. That can help finding the bug.
Note that you have to re-build all libraries after you enable or disable
rtems-debug because some data structures change with that.
Best regards
Christian
--
--------------------------------------------
embedded brains GmbH
Herr Christian Mauderer
Dornierstr. 4
D-82178 Puchheim
Germany
email: christian.maude...@embedded-brains.de
Phone: +49-89-18 94 741 - 18
Fax: +49-89-18 94 741 - 08
PGP: Public key available on request.
Diese Nachricht ist keine gesch?0?1ftliche Mitteilung im Sinne des EHUG.
_______________________________________________
devel mailing list
devel@rtems.org
http://lists.rtems.org/mailman/listinfo/devel
