Re: Fix for config/sim.exp

[Steve Ellcey]

On Sat, 2014-02-15 at 10:01 +0000, Richard Sandiford wrote:


> Maybe a safer fix would be to provide sim_file instead.  Untested patch below.
> 
> Of course, that leaves remote_exec broken, but if it gets called and fails
> in other contexts then maybe it'd be clearer then whether to use host
> execution or to split out most of sim_load into sim_exec.
> 
> Thanks,
> Richard

I tested your sim_file and it worked fine for me.  Since we don't really
know what sim_exec should be doing maybe we should define one that
aborts.  I am not sure if this is completely correct since sim_exec
never seems to get called when running the GCC testsuite (given that
sim_file exists) but here is what I tested.  I am not sure if both the
perror and the verbose is needed in sim_exec.

Steve Ellcey
address@hidden

2014-02-21  Steve Ellcey  <address@hidden>
            Richard Sandiford  <address@hidden>

        * config/sim.exp (sim_exec): New.
        (sim_file): New.


diff --git a/config/sim.exp b/config/sim.exp
index ab83835..87bd403 100644
--- a/config/sim.exp
+++ b/config/sim.exp
@@ -133,6 +133,16 @@ proc sim_upload { dest srcfile args } {
     return [remote_upload host $srcfile $args]
 }
 
+proc sim_exec { dest srcfile args } {
+    perror "Remote execution for simulators not implemented."
+    verbose -log "Remote execution for simulators not implemented."
+    return -1
+}
+
+proc sim_file { board args } {
+    return [eval [list remote_file host] $args]
+}
+
 set_board_info protocol  "sim"
 
 # By default, assume the simulator is slow.  This causes some tests