chromatic wrote:
This bug is the primary thing I've been working on the past couple of days, and I'll let you know when I check in a fix for you to test.
I've checked in an update that fixes two bugs. Jim, please let me know if these fixes resolved the persistent hang.
There are, unfortunately a couple more bugs in the current STM implementation, particularly related to garbage collection. Since I'll be working on the garbage collection PDD next week, and revamping the threading/STM implementation next month, I'm going to set this aside for now. (Especially since it isn't blocking the language implementers.)
Until that point, can we disable the test? I hate to make all of our lovely smokes hang.
It would be useful to see if it does make all the smokes hang. But if it does, then yes, let's disable this test for now.
Allison
