FYI: Jump from -r340287 variant to -r341766 variant failed to boot Pine64+ 2GB

[Mark Millard]

On 2018-Dec-11, at 06:22, Emmanuel Vadot <manu at bidouilliste.com> wrote:

> On Tue, 11 Dec 2018 02:36:30 -0800
> Mark Millard via freebsd-arm <freebsd-arm at freebsd.org> wrote:
> 
>> FYI: The following does at this time still involve my historical
>> changes to allow:
>> 
>> mc0: Card at relative address 0x0002 added:
>> mmc0:  card: MMCHC DJNB4R 0.7 SN 9F43B2D6 MFG 06/2016 by 21 0x0000
>> mmc0:  quirks: 0
>> mmc0:  bus: 4bit, 52MHz (dual data rate timing)
>> mmc0:  memory: 244277248 blocks, erase sector 1024 blocks
>> 
>> as boot media on the Pine64+ 2GB. (This is via a microsd card
>> adapter.) So this is a report for a variant build instead
>> of an official ci.freebsd.org or other such official-source
>> build. I have not yet tried normal boot media. I'm not sure
>> when I'll get that done.
>> 
>> 
>> Showing the tail of boot -v before showing everything after 
>> Loading Kernel...
>> 
>> Release APs...done
>> Trying to mount root from ufs:/dev/label/PINE64P2Groot [rw,noatime]...
>> CPU  0: ARM Cortex-A53 r0p4 affinity:  0
>> Instruction Set Attributes 0 = <AES+PMULL,SHA1,SHA2,CRC32>
>> Instruction Set Attributes 1 = <>
>>         Processor Features 0 = <AdvSIMD,Float,EL3 32,EL2 32,EL1 32,EL0 32>
>> rtc0:          Processor Features 1 = <0>
>> providing initial system time
>>      Memory Model Features 0 = <4k Granule,64k Granule,MixedEndian,S/NS Mem,16bit ASID,1TB PA>
>> start_init: trying /sbin/init
>>      Memory Model Features 1 = <>
>>      Memory Model Features 2 = <32b CCIDX,48b VA>
>>             Debug Features 0 = <2 CTX Breakpoints,4 Watchpoints,6 Breakpoints,PMUv3,Debug v8>
>>             Debug Features 1 = <0>
>>         Auxiliary Features 0 = <0>
>>         Auxiliary Features 1 = <0>
>> CPU  1: ARM Cortex-A53 r0p4 affinity:  1
>> CPU  2: ARM Cortex-A53 r0p4 affinity:  2
>> CPU  3: ARM Cortex-A53 r0p4 affinity:  3
>> regulator: shutting down dcdc3
>> axp8xx_pmu0: Disable dcdc3 (dcdc3)
>> regulator: shutting down dcdc4
>> axp8xx_pmu0: Disable dcdc4 (dcdc4)
>> regulator: shutting down vcc-hdmi
>> axp8xx_pmu0: Disable vcc-hdmi (dldo1)
>> regulator: shutting down vcc-mipi
>> axp8xx_pmu0: Disable vcc-mipi (dldo2)
>> regulator: shutting down dldo3
>> axp8xx_pmu0: Disable dldo3 (dldo3)
>> regulator: shutting down vcc-wifi
>> axp8xx_pmu0: Disable vcc-wifi (dldo4)
>> regulator: shutting down aldo1
>> axp8xx_pmu0: Disable aldo1 (aldo1)
>> regulator: shutting down cpvdd
>> axp8xx_pmu0: Disable cpvdd (eldo1)
>> regulator: shutting down eldo2
>> axp8xx_pmu0: Disable eldo2 (eldo2)
>> regulator: shutting down eldo3
>> axp8xx_pmu0: Disable eldo3 (eldo3)
>> regulator: shutting down vcc-1v2-hsic
>> axp8xx_pmu0: Disable vcc-1v2-hsic (fldo1)
>> 
>> It never reports the next line (or later)
>> that my prior kernel ( -r340287 ) reported:
>> 
>> regulator: shutting down vcc-phy
>> 
>> The same sources built and installed for the OverDriver 1000
>> (cortexA57 vs. cortexA53) boot the OVerDrive just fine (no
>> such odd boot-media involved).
>> 
>> 
>> 
>> More completely the -r341766 boot -v shows:
>> 
>> Loading kernel...
> . . .
>> axp8xx_pmu0: Disable eldo3 (eldo3)
>> regulator: shutting down vcc-1v2-hsic
>> axp8xx_pmu0: Disable vcc-1v2-hsic (fldo1)
>> 
>> (And that is all that shows up.)
>> 
>> Ignoring minor ordering differences and other minor
>> differences, -r340287 also had in its output (or
>> missing, if so noted):
>> 
>> OK unload
>> OK load /boot/kerold/kernel
>> /boot/kerold/kernel text=0x8f880f data=0x17dc10+0x79ccbc syms=[0x8+0x1206d8+0x8+0x113e33]
>> 
>> 
>> [a 1 MB difference in one of its Exclude memory regions:]
>> 
>>  0xb1000000 - 0xb267cfff,    22 MB (   5757 pages) NoAlloc 
>> 
>> 
>> [There were no lines like:]
>> 
>> Preloaded boot_entropy_cache "/boot/entropy" at 0xffff0000014f7098.
>> Preloaded elf module "/boot/kernel/umodem.ko" at 0xffff0000014f70f0.
>> Preloaded elf module "/boot/kernel/ucom.ko" at 0xffff0000014f7908.
>> 
>> [And, of course, -r340287 had the line (and what followed):]
>> 
>> regulator: shutting down vcc-phy
>> 
> . . .
> That's weird, I'll have a look.
> 
> In the meantime you can set hw.regulator.disable_unused=0 in
> loader.conf as a hack.

I got a chance to try that and it did allow the -r341766 based boot to complete.
Thanks.

===
Mark Millard
marklmi at yahoo.com
( dsl-only.net <http://dsl-only.net/> went
away in early 2018-Mar)