[ECOS] Re: potential eCos mbox issue on AT91

[David Ho]

> I suggest we take a serious look at this, as mboxt2 is made default
> and the cdl option to use mboxt is taken out of the CVS tree.

I have run mboxt overnight and it has run fine for the last 20 hours.
 I think there ought to be an interrim workaround before mboxt2 is
properly looked at.

For those who have similar issues with mbox in a recent CVS pull, you
can do the following.

David

---
diff --git a/packages/hal/arm/at91/satmodem/current/misc/satmodem.ecm
b/packages/hal/arm/at91/satmodem/current/misc/satmodem.ecm
index 9d5fe1c..f47d6d5 100644
--- a/packages/hal/arm/at91/satmodem/current/misc/satmodem.ecm
+++ b/packages/hal/arm/at91/satmodem/current/misc/satmodem.ecm
@@ -33,3 +33,7 @@ cdl_component CYGFUN_LWIP_LOOPIF {
 cdl_option CYGDBG_LWIP_DEBUG {
    user_value 1
 };
+
+cdl_option CYGIMP_MBOX_USE_MBOXT_PLAIN {
+   user_value 1
+};
diff --git a/packages/kernel/current/cdl/synch.cdl
b/packages/kernel/current/cdl/synch.cdl
index ab52d70..48b6fab 100644
--- a/packages/kernel/current/cdl/synch.cdl
+++ b/packages/kernel/current/cdl/synch.cdl
@@ -141,6 +141,14 @@ cdl_component CYGSEM_KERNEL_SYNCH_MUTEX_

 }

+cdl_option CYGIMP_MBOX_USE_MBOXT_PLAIN {
+    display       "Use mboxt_plain mbox implementation"
+    default_value 0
+    description   "
+       Use the plain mboxt implementation instead of the more recent
+       mboxt2 implementation."
+}
+

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss