03:45 <la-s> Does anyone else here experience what seems like compression when using a 4k screen at 60 Hz? AFAICT nothing is using 10 bpc, but the monitor does support it (though through FRC it seems), but my colors still get slightly fucked up at 60 Hz. The monitor does only support HDMI 2.0, but that should be enough, so I am wondering whether my monitor is buggy, or whether it's some problem on my side or something in meson.

03:46 <la-s> I am using a Philips 278E1A monitor FWIW

08:38 <narmstrong> la-s: can you elaborate ?

08:38 <narmstrong> 10bit output is not supported, yet

08:38 <narmstrong> is the colors the same at 30hz ?

09:49 <lvrp16> xdarklight: any link to the slc mode for mlc nand?

11:20 <narmstrong> https://lore.kernel.org/linux-mtd/20191230165129.11925-1-miquel.raynal@bootlin.com/ right ?

14:16 <damex> is there nothing we could do but use vendor version of bl2 and bl30 for gxl?

15:53 <la-s> narmstrong: At 30 hz it looks normal. It's clearly noticeable, and also goes away if I just go to 1080p. It seems exactly like what would happen if there isn't enough bandwidth. It happens both with a raw tty, wlroots, and weston. It is especially noticeable for purely blue colors, which get distorted really badly.

15:55 <narmstrong> la-s: can you dump the /sys/kernel/debug/clk/clk_summary when in 60hz ?

15:56 <narmstrong> To check if it’s yuv444 or yuv420

15:58 <la-s> narmstrong: It's on here (http://[240d:1a:164:8700:21e:6ff:fe42:3663]:8000/) if you want to look, but how do I tell?

15:59 <narmstrong> You need to look at the hdmi frequency

16:08 <narmstrong> damex: bl31 is upstream for gxl

16:08 <narmstrong> But bl2 is quite complex, and works well so far

16:09 <narmstrong> damex: as I already said, please write an open source version if you want

16:23 <damex> narmstrong: bl31 is an 'atf', yeah - i have seen that. how about bl30?

16:24 <damex> i am just collecting information, have some boards with amlogic and it is quite painful to have all this closed source bundles of bl2 / bl30

16:32 <xdarklight> lvrp16: https://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux.git/commit/?h=nand/next&id=9e3307a169537a6adc30b13bf9063e94990a5493 from https://git.kernel.org/pub/scm/linux/kernel/git/mtd/linux.git/log/?h=nand/next

16:34 <lvrp16> xdarklight: wow maybe someone will be able to write a software flash controller now.

16:35 <xdarklight> lvrp16: for making use of the whole space on MLC NAND or what do you have in mind?

16:37 <lvrp16> xdarklight: basic handle hardware controller functions in software. Tunables to emulate what a hardware controllers do in terms of balancing speed and reliability.

20:14 <narmstrong> damex: only bl31 is upstream

20:14 <damex> narmstrong: is it true for all available allwinner boards or some are exception?

20:14 <damex> s/allwinner/amlogic/

20:15 <narmstrong> bl31 is not board specific, it’s soc specific

20:15 <damex> yeah, but is there some amlogic boards that have earlier stages available upstream?

20:16 <narmstrong> No

20:19 <damex> got it, thanks :(