Search
Wednesday, 24th of October 2018, 22:47:44 UTC
23:27:37
Bike
beach: what i was thinking about yesterday is how calls have to have some information about unwinding. in sicl that's through the dynamic-environment argument, which can be set up dynamically by pushing to it when a CATCH instruction is hit and popping from it after the block/whatever. clasp has a different implementation strategy where this is known statically instead (i.e., which points in the same function
23:27:43
Bike
a given call could unwind to)
23:27:49
Bike
currently clasp kind of fakes it but it would be good to do it more exactly
5:05:56
beach
Good morning everyone!
6:16:38
no-defun-allowed
morning beach and drmeister (IRC)
8:22:26
no-defun-allowed
beach (IRC): there's new [radeon drivers on linux](https://www.amd.com/en/support/kb/release-notes/rn-prorad-lin-18-40)
8:23:15
shka_
ok, so what is this issue status anyway?
8:27:27
no-defun-allowed
although those drivers don't seem to work on a (non-default) 4.17 kernel, which is a bit sad
8:49:31
beach
no-defun-allowed: Thanks.
8:50:41
beach
no-defun-allowed: Do you think I should install it?
8:51:39
no-defun-allowed
i don't know yet
8:51:59
beach
Can you find out somehow?
8:52:01
no-defun-allowed
i can't remember what tricks i used to install the ubuntu amdgpu in the first place, hah
8:52:56
no-defun-allowed
i asked my friend who helped me install it in the first place, but she's probably asleep
8:56:55
shka_
is it confirmed that drivers are to blame?
9:00:40
no-defun-allowed
it's my best guess
9:01:51
shka_
https://www.youtube.com/watch?v=EigIujKxDcE
9:11:28
beach
shka_: No, nothing has been confirmed.
Thursday, 25th of October 2018, 10:47:44 UTC