20:37 <mntmn> oh, two new imx8s

20:37 <mntmn> one of them with GPU... the i.MX 8ULP

20:38 <mntmn> > RISC-V powered Power Management Subsystem (µpower)

20:43 <mntmn> also there's an i.MX9 page now, but no info except smart, smart, smart

20:43 <marex> mntmn: it would be much nicer if NXP fixed the 250kLoC diff of their downstream U-Boot against mainline and 1 MLoc diff of downstream Linux against mainline first, before rolling out any more new SoCs

20:44 <mntmn> marex: not gonna happen

20:44 <marex> mntmn: embedded world week, so no surprise

20:44 <marex> mntmn: I like your positive attitude :(

20:44 <mntmn> marex: i'm just being realistic, because SoCs are rolled out all the time it seems ;)

20:48 <marex> mntmn: the software seems to lack behind though, esp. with NXP :(

20:48 <mntmn> yeah, lets hope they'll improve this

20:48 <marex> mntmn: is the market-proven MX8MP datasheet now available to public or not yet ?

20:49 <marex> mntmn: the downstream kernel has a copy of everything for mx6, and then copy of everything for mx8m, I wonder what will happen with mx9, but I do see some pattern

20:50 <mntmn> let me check

20:50 <marex> if fine-grained power control works as well as SCFW (closed source blob) or GPCv2 (buggy and poorly documented), it will be so sad

20:51 <mntmn> marex: hey, it's replaced with RISC-V now, so it has to be awesome

20:51 <mntmn> marex: i still don't see any imx8mp manuals, except strange collateral like "i.MX 8M Plus NPU Warmup Time"

20:53 <marex> mntmn: the vacuum tubes take time to warm up, that's normal

20:59 <mntmn> hehe