Not so much a funny commit, but one that is quite rude to myself for a very dumb bug that I spent a few months debugging or working around. (I've edited this version of the committ slightly, see it in its full glory
The bug this caused was the first (or second, can't remember which) thread to wait on a mutex would be lost and hence block indefinitely. This ended up being the shell in VT1, so I had to switch VTs to test any of the userland.
commit da67fd4fd32d16dcd6b4cb0b63497a9925a2ef35
Author: John Hodge <
[email protected]>
Date: Sat Oct 2 21:56:44 2010 +0800
Placed a comment to not the location of the blocker
diff --git a/Kernel/threads.c b/Kernel/threads.c
index 3b73111..b9d59f3 100644
--- a/Kernel/threads.c
+++ b/Kernel/threads.c
@@ -1044,6 +1044,9 @@ void Mutex_Release(tMutex *Mutex)
Mutex->Owner = Mutex->Waiting; // Set owner
Mutex->Waiting = Mutex->Waiting->Next; // Next!
// Reset ->LastWaiting to NULL if we have just removed the last waiting thread
+ // 2010-10-02 21:50 - Comemerating the death of the longest single
+ // blocker in the Acess2 history. REMEMBER TO
+ // ____ING MAINTAIN YOUR ____ING LISTS DIPWIT
if( Mutex->LastWaiting == Mutex->Owner )
Mutex->LastWaiting = NULL;