#
✕ nick:marcan
#asahi · 2021-09-21 16:10
marcanj_ey: is this the keyboard with the corellium driver?
#asahi · 2021-09-21 16:09
marcanheh
#asahi · 2021-09-21 16:07
marcanyou can try adding always-on; props
#asahi · 2021-09-21 16:07
marcanif you are chaining all the way up, note that linux will probably try to power down things it shouldn't
#asahi · 2021-09-21 16:06
marcanso we're treating all the dynamic clocks as static
#asahi · 2021-09-21 16:06
marcanit's dynamic but macos never touches it and we don't even know what the clock mux inputs are for that
#asahi · 2021-09-21 16:05
marcanjust make another fixed clock, call it spiclk or something
#asahi · 2021-09-21 16:04
marcan200MHz
#asahi · 2021-09-21 16:04
marcanthat's a dynamic clock
#asahi · 2021-09-21 16:04
marcanj_ey: spi is 200MHz as I said
#asahi · 2021-09-21 16:04
marcanthat's the only one in use (and it's refclk)
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/i2c0.clk[0]
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/uart0.clk[1]
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/wdt.clk[0]
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/i2c2.clk[0]
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/uart2.clk[1]
#asahi · 2021-09-21 16:03
marcan User: /device-tree/arm-io/i2c1.clk[0]
#asahi · 2021-09-21 16:03
marcan #4
#asahi · 2021-09-21 16:03
marcan=== Fixed clocks ===
#asahi · 2021-09-21 16:03
marcanjust added a thing to dump_pmgr to also dump the fixed/hardcoded clocks
#asahi · 2021-09-21 16:01
marcanyes, IIRC
#asahi · 2021-09-21 15:58
marcan(that might explain why refclk didn't work, if the target device didn't like a ~10x too fast clock)
#asahi · 2021-09-21 15:58
marcansays 200MHz
#asahi · 2021-09-21 15:58
marcan User: /device-tree/arm-io/spi1.clk[0]
#asahi · 2021-09-21 15:58
marcan#121: 200000000 2 0xa0/0x23b0400f8: regval: 0x85100000
#asahi · 2021-09-21 15:57
marcanoh sorry, clock, not pstate
#asahi · 2021-09-21 15:56
marcanI should write a script that just spits it all out in DT format :)
#asahi · 2021-09-21 15:56
marcansee dump_pmgr.py
#asahi · 2021-09-21 15:48
marcan(sorry :p)
#asahi · 2021-09-21 15:48
marcanI turned it on anyway, seems useful to catch OF refcounting bugs
#asahi · 2021-09-21 15:47
marcanj_ey: also just force-pushed with a better fix :)
#asahi · 2021-09-21 15:46
marcanseems CONFIG_OF_OVERLAY is probably how you ended up with that
#asahi · 2021-09-21 15:45
marcanwhich is why I didn't
#asahi · 2021-09-21 15:45
marcanI did not
#asahi · 2021-09-21 15:44
marcanj_ey: pushed to pmgr/dev
#asahi · 2021-09-21 15:43
marcanfixed it
#asahi · 2021-09-21 15:40
marcan(and it's a bug in my pmgr code)
#asahi · 2021-09-21 15:40
marcanCONFIG_OF_DYNAMIC is what causes it :)
#asahi · 2021-09-21 15:37
marcanuse keep_bootcon for pmgr stuff, otherwise you probably won't get any output during that critical period
#asahi · 2021-09-21 15:37
marcanthat will let you pass the brk through to the guest, so you see the prints
#asahi · 2021-09-21 15:37
marcanI just added a lower() command to the hv
#asahi · 2021-09-21 15:37
marcanj_ey: BUG() does that
#asahi · 2021-09-21 14:41
marcanj_ey: anyway, let me see if I can fix that brk business
#asahi · 2021-09-21 14:35
marcanhuh, why is that going into printk back into of...
#asahi · 2021-09-21 14:33
marcanwhat's above it?
#asahi · 2021-09-21 14:29
marcantry now
#asahi · 2021-09-21 14:28
marcanj_ey: done
#asahi · 2021-09-21 14:11
marcananyway, let me fix this
#asahi · 2021-09-21 14:11
marcanthat's why, I was on 4K
#asahi · 2021-09-21 14:09
marcanj_ey: ah, is this a 4K or 16K kernel?

plus ›