This is the mail archive of the
ecos-discuss@sources.redhat.com
mailing list for the eCos project.
Re: problem about mutex lock/unlock in ecos 1.3.1
- From: Nick Garnett <nickg at ecoscentric dot com>
- To: cvigne at free dot fr
- Cc: ecos-discuss at sources dot redhat dot com,UNEXPECTED_DATA_AFTER_ADDRESS at dot SYNTAX-ERROR
- Date: 23 May 2003 10:58:33 +0100
- Subject: Re: [ECOS] problem about mutex lock/unlock in ecos 1.3.1
- References: <1053678813.3ecddcdd9acf2@imp.free.fr>
cvigne@free.fr writes:
> Hi,
>
> we have some problems when you use the mutex-lock: for instance, when we used
> the wait (in mutex.cxx in ecos 1.3.1), there are sometimes currentsched_lock ==
> 2 (whereas it must be 0 or 1!!): so just before to finish the wait, there is
> CYG_ASSERT(Scheduler::get_sched_lock() == 1, "...") !!
>
> Is this a resolved problems in the new version of ecos? or have you any ideas
> to resolve our problems?
>
> thanks for your answer
>
I'm not quite sure what you are asking here, but there has been a
change in all the synchronization code since 1.3.1. In 1.3.1 all
synchronization primitives had to be called with the scheduler lock
zero. In current versions this is not the case and synchronization can
be performed with the scheduler locked. This makes the scheduler
lock essentially a per-thread variable.
--
Nick Garnett eCos Kernel Architect
http://www.ecoscentric.com/ The eCos and RedBoot experts
--
Before posting, please read the FAQ: http://sources.redhat.com/fom/ecos
and search the list archive: http://sources.redhat.com/ml/ecos-discuss