Here's my first patch for Mach; I hope it will get reviewed. There was
a variable reply_port that was set in the if statement, but not used as
an argument of ds_device_write_trap. I commented out the declaration
(as there is indication in the comments that in the future it might be
used in the call of ds_device_write_trap) and rewrote the if statement.
--- ./gnumach/kern/ipc_mig.c 2013-06-01 20:41:44.180548551 +0200
+++ ./gnumach1/kern/ipc_mig.c 2013-06-03 19:42:29.108955104 +0200
@@ -928,7 +928,7 @@
vm_size_t data_count)
{
device_t dev;
- ipc_port_t reply_port;
+ /*ipc_port_t reply_port;*/
io_return_t res;
/*
@@ -949,9 +949,10 @@
/*
* Translate reply port.
*/
- if (reply_name == MACH_PORT_NULL)
- reply_port = IP_NULL;
- else {
+ /*if (reply_name == MACH_PORT_NULL)
+ reply_port = IP_NULL;
+ else {*/
+ if (reply_name != MACH_PORT_NULL) {
/* Homey don't play that. */
device_deallocate(dev);
return KERN_INVALID_RIGHT;
