Search
Thursday, 12th of October 2017, 0:40:35 UTC
8:41:47
scymtym
x86 and macos builds are broken (at least on our CI server): https://ci.cor-lab.org/job/sbcl-master/2533/
11:43:31
stassats
ALLOCATION-INFORMATION still fails, on x86
11:50:50
stassats
scymtym: the mac (not really) problems should go away
11:52:00
scymtym
stassats: great, i was starting to lose track of what broke when and where
11:52:26
stassats
can you switch on sb-thread for darwin?
11:52:55
stassats
or maybe, i should switch them on by default
12:01:52
scymtym
thanks. saves me a bit of work
12:02:17
stassats
that wasn't just for you
12:02:46
scymtym
i know, still better for me :)
12:02:58
stassats
but it'll be helpful now to have a -sb-thread test target
12:12:00
stassats
why are BIND and UNBOUND so eager to get assigned RDX as a temporary?
12:24:58
stassats
i think it's the old problem of different functional not communicating their register preference, but for unbind, the return wants RDX, why is UNBIND getting it?
12:25:12
stassats
did it get it first or something?
12:38:27
stassats
saving on constant loads fro BIND produces 49C784246823000017001020, a 12 byte instruction
12:38:36
stassats
saves only 1 byte compared to loading NIL into a register first
12:40:11
stassats
but saves a register, but who cares, it has other plans on using registers
Thursday, 12th of October 2017, 12:40:35 UTC