04:17 dcbaker: DavidHeidelberg: that was a port of something in the auto tools build, I don’t even remember who added it initially, but I remember reading the gcc log for why it was removed, and they basically found that outside a few synthetic benchmarks it either hurt or didn’t help performance, so I’d be inclined to mark the option as deprecated and delete the check
04:17 dcbaker: Which I can send a patch for Monday
05:30 DavidHeidelberg: dcbaker: muy bien, gracias! :)
06:20 dcbaker: DavidHeidelberg: de nada
08:33 colinmarc: something I've wondered about for a while: how does the vsync checkbox in games correspond to actual present modes? Is it FIFO?
08:36 K900: Depends on the game really
08:37 K900: But in most cases it's immediate for vsync off and FIFO/mailbox for vsync on
08:55 Lynne: so any way to debug CS rejection on radv?
08:55 Lynne: I'm not even running a shader anywhere
08:56 Lynne: only a single pipeline barrier
09:05 Lynne: oh no, if I disable host mapped image uploads, its fine
09:06 Lynne: but the code runs on other vendors, validation layers don't complain, and the math all checks out, no overreads are possibly happening
18:44 airlied: Lynne: anything in dmesg?
19:16 caughtfromtheact: So what are powers in that sense 1+3 4+0 3+1 0+4 2+2 and 0+4 all result to 4, 0+2 2+0 1+1 result to 2, 0+3 3+0 1+2 2+1 result to 3 and 0+1 yield to 1, so how many combinations one could have with 3bits 1 2 4? so 3+4=7 1+6=7 2+5 and vice versa ones so how many combinations? it is max 7 so you get 000 010 110 011 111 001 100 which are unique so such combinations you say in 32bits is
19:16 caughtfromtheact: 4.2xxxBillion , it's plausibly right but if you were really intelligent you would say 3+3, and storage is needed as to separate the states uniquely and carry capability for compressed intrinsic which is exposed shifting the operands and eliminating 0s that are represented with higher power or however liked, i represent them in polderan. It's because you did not store the programming
19:16 caughtfromtheact: operands for a program, a material did, You can not count to 4 already due to lack of the sense of reality and coordination and only cause youa re devils entering the planet with only birth injuries to start with, but material is not wrong as with absurd that you should doubt that material stored is wrong, so it could take back the shift somehow cause it lied or material did, and
19:16 caughtfromtheact: also because the shift would be performed for answer sets too
21:55 reportingall: Tuul lükkab tagasi! Samm edasi kaugemale viib , aga tuul lükkab tagasi, ning kõigi ees püüan ikka veel ja ma ei saa ikka edasi, ja ma magada ei saa, aina mõtteid keerutan! Samm edasi kaugemale viib aga tuul lükkab tagasi, tuul lükkab tagasi, tuul lükkab tagasi, edasi tagasi, tuul lükkab tagasi.. https://www.youtube.com/watch?v=cKgxUVo5uL0&list=RDcKgxUVo5uL0&start_radio=1 .
22:13 reportingall: how many typos in the language -- precisely with commas? Another math question, if Sarah married to Joe is 23, and Joe is 67 , then how much money should Joe have? And if you do not know the answer, you do not know shit about maths, hence never bother me again!
22:25 Ermine: how do you submit docs updates? Is patch 00 of N is suitable for that?
22:29 heat: wdym
22:33 Ermine: I've heard somewhere that one doesn't just send patches for typos or something like that, and one should send them along with some actual patches
22:54 heat: docs typos i'm sure are fine
22:55 heat: when sending out a patch set, "patch 0" is never an actual patch but the cover letter