00:42fdobridge: <gfxstrand> IDK what's going wrong yet. I'm getting an unaligned address error which is really weird because nothing looks even slightly unaligned.
01:04fdobridge: <gfxstrand> It did turn out to be sched info. 🙂
01:04fdobridge: <karolherbst🐧🦀> classic
01:04fdobridge: <gfxstrand> The hardware hard-codes that you **must** set the wait mask to `0x3f`
01:05fdobridge: <gfxstrand> Which, fine, but also, it'd be nice if it gave a nicer error. 🙃
01:05fdobridge: <karolherbst🐧🦀> nah
01:05fdobridge: <karolherbst🐧🦀> the hardware doesn't have any validation on it anymore
01:05fdobridge: <karolherbst🐧🦀> it all got thrown out
01:05fdobridge: <karolherbst🐧🦀> unless there is a flag to turn on back on, but...
01:06fdobridge: <gfxstrand> I mean, maybe not? Or maybe my flow control just didn't work?
01:06fdobridge: <karolherbst🐧🦀> though I do have docs on the sched info stuff, I could try to figure out if there is some rule we have to follow
01:06fdobridge: <gfxstrand> IDK. I was waiting on all the things that were interesting.
01:06fdobridge:<gfxstrand> needs to get her hands on these mythical docs
01:06fdobridge: <karolherbst🐧🦀> but yeah.. cfg is a bit annoying with scheds
01:08fdobridge: <mhenning> where was the error message from?
01:09fdobridge: <karolherbst🐧🦀> but anyway.. sched wrong => undefined behavior
01:09fdobridge: <karolherbst🐧🦀> @gfxstrand you really really want to add an env var to be able to turn that sched stuff to this magic "wait on all and maximum time" value 😄
01:09fdobridge: <gfxstrand> I have one of those. 🙂
01:09fdobridge: <gfxstrand> It just didn't quite work.
01:10fdobridge: <gfxstrand> It just didn't quite work.🙃 (edited)
01:11fdobridge: <karolherbst🐧🦀> yeah.. that's the sched stuff already
01:11fdobridge: <karolherbst🐧🦀> it's a pain and it's often the best thing to try first
04:11fdobridge: <gfxstrand> The new compiler I'm writing
04:24fdobridge: <gfxstrand> `Pass: 300350, Fail: 6094, Crash: 6825, Skip: 1673047, Timeout: 32, Flake: 63, Duration: 1:19:01`
04:25fdobridge: <gfxstrand> I fixed half the crashes today. 😄
09:53fdobridge: <Mohamexiety> nice! that's awesome
12:14xiphmont: Hello all. I know ~11 Quadro FX2800 were ever deployed in the wild, but there are a few bugs since the big 11/07 drop that keep the card from booting. I have a possibly misguided patch that lets them come up, sent to the list. Should I file a bug as well?
12:15xiphmont: [it also has later on DMA management lockup bugs, but they're non-deterministic. That's for later]
12:15xiphmont: Also, to whom\ever added all the trace and debug logging over the past few months: THANK YOU
12:16xiphmont: You're a deity amongst humankind and I owe you sacrifices.
17:52fdobridge: <gfxstrand> `Pass: 300614, Fail: 6101, Crash: 6727, Skip: 1673845, Timeout: 32, Flake: 56, Duration: 1:19:05`
17:52fdobridge: <gfxstrand> No NAK changes, just a CTS update with a bunch of CTS fixes.
23:04fdobridge: <gfxstrand> `Pass: 302271, Fail: 7661, Crash: 3483, Skip: 1673845, Timeout: 32, Flake: 83, Duration: 1:25:15`