Re: [lwip-users] Timeout wraparound bug

[Sylvain Rochet]

Hi Fotis,

On Sat, Feb 18, 2017 at 01:55:41PM +0200, Fotis Panagiotopoulos wrote:
> Hi,
> 
> To begin with, I do call sys_check_timeouts() very often. I also tried
> v2.0.1, where the problem vanishes. I believe it is fixed by the following
> addition:
> 
>         } else if (timeout->time > msecs) {
>           /* If this is the case, 'timeouts_last_time' and 'now' differs too 
> much.
>              This can be due to sys_check_timeouts() not being called at the 
> right
>              times, but also when stopping in a breakpoint. Anyway, let's 
> assume
>              this is not wanted, so add the first timer's time instead of 
> 'diff' */
>           timeout->time = msecs + next_timeout->time;
>         }
>         timeout->next = t->next;
> 
> which looks more like a work-around for me, rather an actual fix.

You can only enter that if your main timer continue to run while your 
system is suspended (eg. waiting on a breakpoint while debugging on a 
desktop computer), this is a work-around to help us using debuggers. 
Systems that suspend should call sys_restart_timeouts() on resume to 
prevent that.

Otherwise it means you do not call sys_check_timeouts() at an acceptable 
rate or your sys_now() function is buggy.

Sylvain

signature.asc
Description: Digital signature