On 04/27/2010 03:38 PM, Øyvind Harboe wrote:
Could you attach a patch?
Of course, but as I said, I don't know for certain that this is a
correct fix or not. I also don't know if it is needed only for certain
revisions of the device (<= B1). It would be nice if someone with a C1
or C3 rev Tempest could test "reset halt". On the other hand, maybe the
software reset works just as fine a hardware reset, so it doesn't really
matter.
The comment says that nRST clears some of the debug state for some LM3S
devices. Would it be possible to test for that instead of revision numbers?
/Tobias
diff --git a/src/target/cortex_m3.c b/src/target/cortex_m3.c
index 7f6cbaf..44bf466 100644
--- a/src/target/cortex_m3.c
+++ b/src/target/cortex_m3.c
@@ -910,7 +910,7 @@ static int cortex_m3_assert_reset(struct target *target)
assert_srst = 0;
break;
case 4:
- /* Tempest should be fine. */
+ assert_srst = 0;
break;
}
}
_______________________________________________
Openocd-development mailing list
Openocd-development@lists.berlios.de
https://lists.berlios.de/mailman/listinfo/openocd-development
