erdoukki
commented
3 years ago It's me or the available documents on Marvell Extranet are full of missing data !
Open erdoukki opened 3 years ago
erdoukki
commented
3 years ago It's me or the available documents on Marvell Extranet are full of missing data !
robimarko
commented
3 years ago You can find the AVS bits for voltage in the Functional specifications, but yeah there is a lot of stuff missing in the documentation.
erdoukki
commented
2 years ago What can I test to help on this ?
pali
commented
2 years ago @kostapr: Please, could you really look at this? It is really serious issue which lacks any progress :-(
kostapr
commented
2 years ago @pali, as far as I know this thread was passed to our support team. I am part of the development group and simply do not have enough bandwidth for supporting SOCs that are not in active development stage.
erdoukki
commented
2 years ago @globalscaletechnologies this post answer from Wayne Yang has disappeared from the EspressoBin.net Forum subject Crash after booting for about a minute !
August 16, 2017 at 6:20 pm
932
Wayne Yang Keymaster
Please change uboot parameter below to avoid the issue on ubuntu:
Marvell>> setenv bootcmd ‘mmc dev 0; ext4load mmc 0:1 $kernel_addr $image_name;ext4load mmc 0:1 $fdt_addr $fdt_name; mw.l 0xd0011500 0x78e3ffff; setenv bootargs $console root=/dev/mmcblk0p1 rw rootwait net.ifnames=0 biosdevname=0; booti $kernel_addr – $fdt_addr’ Marvell>> saveenv
If I understand, the mw.l 0xd0011500 0x78e3ffff will set the AVS status to "Pause" and will set also the min/max voltage values to 0x23, "1.155V"...
Is it a clue to study ?
Looks like the issue was already known in early 2017 !?
erdoukki
commented
2 years ago Can we try one more: 0x5BAEFFFF To make sure: 1 GHz is stable?
@stefanchulski my apologize about the test made, but do you need this to be tested on 1.GHz A3720 ??? I have only tested it on the Ultra, which is 1.2GHz !
If the test needed was about 1 GHz only, I have to made it again to report correctly ! {0x2e, "1.283V"} Without DFS also ?
erdoukki
commented
2 years ago @stefanchulski
I not familiar with all these AVS configurations on A37XX. But hardcoded values look strange, should you take into account chip skew and calculate AVS from SVC?
as @pali said :
But hardcoded values look strange
Yes, but we have absolutely no idea what is happening here. And if you look at referenced changed from above commit dc33b62 those hardcoded values were done by Marvell developers...
should you take into account chip skew and calculate AVS from SVC?
Probably, but we have no idea how... There is missing documentation about this topic. I have not seen any SVC documentation. So this is something which is probably only internally in Marvell.
How can Marvell help on this ?
erdoukki
commented
2 years ago @pali
DFS (AVS) is on A3720 disabled when armada-37xx-cpufreq driver is not initialized. And for 1.2 GHz mode it is already disabled if you see that Unsupported CPU frequency message.
@robimarko
It's already disabled in the OpenWrt if you are running it as they have backported the 5.14 patch for it, but like @pali said just check the boot log for the print.
If I want to do some tests with my own build of the linux kernel, I just had to disable (not enable) : CONFIG_ARM_ARMADA_37XX_CPUFREQ ?
robimarko
commented
2 years ago No point in disabling the driver as 1.2GHz version is blacklisted in it anyway
erdoukki
commented
2 years ago Okay, sure... But it was for my better understanding. In case of tests with 1.2GHz un-blacklisted.
erdoukki
commented
2 years ago Can we try one more: 0x5BAEFFFF To make sure: 1 GHz is stable?
@stefanchulski my apologize about the test made, but do you need this to be tested on 1.GHz A3720 ??? I have only tested it on the Ultra, which is 1.2GHz !
If the test needed is about 1 GHz only, I have to made it again to report correctly ! {0x2e, "1.283V"} Without DFS also ?
I am starting a stress test on 1.0 GHz EspressoBin.
SVC REV: 5, CPU VDD voltage: 1.108V
=> md 0xd0011500
d0011500: 57dfffff 02000257 00008000 800001e1 ...WW........... => mw 0xd0011500 0x5BAEFFFF
=> md 0xd0011500
d0011500: 5baeffff 02000257 00008000 800001e1 ...[W........... root@BORNE-GRENIER:/# lscpu | grep CPU
CPU op-mode(s): 32-bit, 64-bit
CPU(s): 2
On-line CPU(s) list: 0,1
CPU max MHz: 1000.0000
CPU min MHz: 200.0000 root@BORNE-GRENIER:/# uname -ar
Linux BORNE-GRENIER 4.14.221 #0 SMP Mon Feb 15 15:22:37 2021 aarch64 GNU/Linux root@BORNE-GRENIER:/# cat /etc/openwrt_release
DISTRIB_ID='OpenWrt'
DISTRIB_RELEASE='19.07.7'
DISTRIB_REVISION='r11306-c4a6851c72'
DISTRIB_TARGET='mvebu/cortexa53'
DISTRIB_ARCH='aarch64_cortex-a53'
DISTRIB_DESCRIPTION='OpenWrt 19.07.7 r11306-c4a6851c72' root@BORNE-GRENIER:/# stress --cpu 2 --io 2 --timeout 1h
stress: info: [4911] dispatching hogs: 2 cpu, 2 io, 0 vm, 0 hdd Will update after stress test concluded !
UPDATE :
stress: info: [4911] successful run completed in 3600s No crash on 1 GHz !
erdoukki
commented
2 years ago Done a full stress test with disabling AVS !
SVC REV: 5, CPU VDD voltage: 1.225V
Marvell>> md 0xd0011500
d0011500: 5a69ffff 02000257 00008000 800001e1 ..iZW........... Marvell>> mw 0xd0011500 0x78e3ffff
Marvell>> md 0xd0011500
d0011500: 78e3ffff 02000257 00008000 800001e1 ...xW........... root@OpenWrt:/# dmesg | grep CPU
[ 2.173010] Unsupported CPU frequency 1200 MHz root@OpenWrt:/# uname -ar
Linux OpenWrt 5.4.143 #0 SMP Tue Aug 31 22:20:08 2021 aarch64 GNU/Linux root@OpenWrt:/# devmem 0xd0011500
0x78E3FFFF root@OpenWrt:/# stress --cpu 2 --io 2 --timeout 1h
stress: info: [3766] dispatching hogs: 2 cpu, 2 io, 0 vm, 0 hdd
stress: info: [3766] successful run completed in 3600s NO CRASH !
Will redo the same test with the same board, only modifying the 0xd0011500 to 0x58e3ffff, to confirm it is crashing...
erdoukki
commented
2 years ago Will redo the same test with the same board, only modifying the 0xd0011500 to 0x58e3ffff, to confirm it is crashing...
Just a boot and CRASH !
SVC REV: 5, CPU VDD voltage: 1.225V
Marvell>> md 0xd0011500
d0011500: 5a69ffff 02000257 00008000 800001e1 ..iZW........... Marvell>> mw 0xd0011500 0x58e3ffff
Marvell>> md 0xd0011500
d0011500: 58e3ffff 02000257 00008000 800001e1 ...XW........... boot...
[ 0.121024] Modules linked in:
[ 0.124163] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.4.143 #0
[ 0.130341] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 0.137511] pstate: a0400005 (NzCv daif +PAN -UAO)
[ 0.142448] pc : parameq+0x44/0x88
[ 0.145933] lr : parameq+0x18/0x88
[ 0.149425] sp : ffffffc01002bda0
[ 0.152830] x29: ffffffc01002bda0 x28: ffffffc0108adc58
[ 0.158295] x27: ffffffc01097047c x26: ffffffc010963a80
[ 0.163761] x25: 000000000000008f x24: ffffffc0109650d8
[ 0.169227] x23: ffffff803fdec248 x22: 0000000000000000
[ 0.174692] x21: ffffff803fdec240 x20: ffffffc0107e3a90
[ 0.180158] x19: ffffff803fdec240 x18: 0000000000000014
[ 0.185624] x17: 00000000d4a9c6cf x16: 000000009aa4c235
[ 0.191089] x15: 00000000825af236 x14: 000000004be7dfc1
[ 0.196554] x13: 0000000000000000 x12: 0000000000000038
[ 0.202020] x11: 0101010101010101 x10: 7f7f7f7f7f7f7f7f
[ 0.207486] x9 : 302b2f554c787373 x8 : 7f7f7f7f7f7f7f7f
[ 0.212951] x7 : ff646b6e726d6e62 x6 : 0000000000000080
[ 0.218417] x5 : 8000000000000000 x4 : 000000000000005f
[ 0.223882] x3 : 0000000000000063 x2 : 00656c6f736e6f63
[ 0.229348] x1 : 0000000000000000 x0 : 0000000000000007
[ 0.234814] Call trace:
[ 0.237326] parameq+0x44/0x88
[ 0.240461] parse_args+0xc4/0x2a8
[ 0.243959] kernel_init_freeable+0x1a0/0x270
[ 0.248439] kernel_init+0x10/0xfc
[ 0.251932] ret_from_fork+0x10/0x1c
[ 0.255609] Code: eb00003f 91000421 540001c0 38616a63 (38616a82)
[ 0.261877] ---[ end trace d78e3bf596a26259 ]---
[ 0.266624] Kernel panic - not syncing: Fatal exception
[ 0.272001] SMP: stopping secondary CPUs
[ 0.276035] Rebooting in 1 seconds.. ... and CRASH !
erdoukki
commented
2 years ago
Marvell>> mw 0xd0011500 0x78e3ffffMarvell>> md 0xd0011500 d0011500: 78e3ffff 02000257 00008000 800001e1 ...xW...........
Same tests done on another ULTRA which crash more often !
SVC REV: 5, CPU VDD voltage: 1.237VThis time the Board FROZE in the stress-ng...
So still not a solution...
Will redo more tests with a Power Cycle between each test ! Will use this very "crashy" board for new tests...
erdoukki
commented
2 years ago NOTES :
SVC REV: 5, CPU VDD voltage: 1.237V
U-Boot 2017.03-armada-18.09.1-g51aa6c4772 (Sep 18 2019 - 15:19:13 +0800)
[ 2.170034] Unsupported CPU frequency 1200 MHz
OpenWrt 21.02.0, r16279-5cc0535800root@OpenWrt:/# ./mhz/mhz
count=516515 us50=21529 us250=107716 diff=86187 cpu_MHz=1198.591root@OpenWrt:/# stress-ng --matrix 0 -t 10m
Marvell>> md 0xd0011500
d0011500: 5aaaffff 02000257 00008000 800001e1 ...ZW...........-> CRASH !
Marvell>> mw 0xd0011500 0x5CF3FFFF
Marvell>> md 0xd0011500
d0011500: 5cf3ffff 02000257 00008000 800001e1 ...\W...........-> CRASH !
Marvell>> mw 0xd0011500 0x4cf3ffff
Marvell>> md 0xd0011500
d0011500: 4cf3ffff 02000257 00008000 800001e1 ...LW...........-> CRASH !
Marvell>> mw 0xd0011500 0x7aaaffff
Marvell>> md 0xd0011500
d0011500: 7aaaffff 02000257 00008000 800001e1 ...LW...........-> FROZEN !
Marvell>> mw 0xd0011500 0x4aaaffff
Marvell>> md 0xd0011500
d0011500: 4aaaffff 02000257 00008000 800001e1 ...LW...........-> CRASH !
pali
commented
2 years ago
mv_ddr-devel-g80be893d2b-d DDR4 16b 1GB 1CS
It seems that above commit from log does not exist: https://github.com/MarvellEmbeddedProcessors/mv-ddr-marvell/commit/80be893d2b
Can you check that you have latest mv-ddr-marvell version? Because @heaterC was fixing crashes on espressobins with DDR4 memories, see pull request: https://github.com/MarvellEmbeddedProcessors/mv-ddr-marvell/pull/25
erdoukki
commented
2 years ago Thanks @Pali,
mv_ddr-devel-g80be893d2b-d DDR4 16b 1GB 1CS
I will check... It is from one of the Ultra board which have upgraded u-Boot...
It was also compiled with CLOCKSPRESET:=CPU_1000_DDR_800 and I will refresh it with CLOCKSPRESET:=CPU_1200_DDR_750 as said here : http://espressobin.net/espressobin-ultra-build-instruction/
It seems that above commit from log does not exist: MarvellEmbeddedProcessors/mv-ddr-marvell@80be893
It is from the WIP patch for OpenWrt U-Boot for Ultra which I have proposed from Luka Kovakic patchwork which is in standby... https://github.com/openwrt/openwrt/pull/4443
All 3 others boards have the GST factory u-Boot, but it look to not contain the fix from @heaterC !? I am right ?? https://github.com/globalscaletechnologies/u-boot-marvell
071594b82bc61455c74f80448d1718a3 cellular-cpe-bootloader-cpu-1000-ddr4-1cs-1g-atf-95ac2fcd-uboot-g057aa3fce1-utils-d5b360a-20200616-rel.bin
ed0deaa0166c61c817d68ed978ca38f0 cellular-cpe-bootloader-cpu-1200-ddr4-1cs-1g-atf-95ac2fcd-uboot-g057aa3fce1-utils-d5b360a-20200616-rel.bin
64809267a7688be4eb2738119133c9cb cellular-cpe-emmcloader-cpu-1000-ddr4-1cs-1g-atf-95ac2fcd-uboot-g057aa3fce1-utils-d5b360a-20200616-rel.bin
c8769b249df29be09da65ae899eadfbd cellular-cpe-emmcloader-cpu-1200-ddr4-1cs-1g-atf-95ac2fcd-uboot-g057aa3fce1-utils-d5b360a-20200616-rel.bin
66cc8dd89c3f35a573fe2aacac867b2a cellular-cpe-uartboot-cpu-1000-ddr4-1cs-1g-atf-95ac2fcd-uboot-057aa3fce1-utils-d5b360a-20200616-rel.tgz
75071ad860d6dc080c3833b4588e3b3c cellular-cpe-uartboot-cpu-1200-ddr4-1cs-1g-atf-95ac2fcd-uboot-057aa3fce1-utils-d5b360a-20200616-rel.tgzI will test again with the most buggy Ultra board.
I will recompile a new u-boot with latest patch from Luka Kovakic from here : https://patchwork.ozlabs.org/project/uboot/list/?series=257886
Can you check that you have latest mv-ddr-marvell version? Because @heaterC was fixing crashes on espressobins with DDR4 memories, see pull request: MarvellEmbeddedProcessors/mv-ddr-marvell#25
It was "default" from the MVEBU-UBOOT package in OpenWrt, I will refresh all and made a full check and report here...
erdoukki
commented
2 years ago
mv_ddr-devel-g80be893d2b-d DDR4 16b 1GB 1CSIt seems that above commit from log does not exist: MarvellEmbeddedProcessors/mv-ddr-marvell@80be893
Can you check that you have latest mv-ddr-marvell version? Because @heaterC was fixing crashes on espressobins with DDR4 memories, see pull request: MarvellEmbeddedProcessors/mv-ddr-marvell#25
Look like the actual mvebu-uboot from openwrt contains now : https://github.com/MarvellEmbeddedProcessors/mv-ddr-marvell/commit/02e23dbc
I do not understand from where came this "strange" commit ! The compilation was made in august of 2021.
May be a bug already fixed or a pending issue... I will verify this after compilation complete !
UPDATE : Add GITHUB reference of WIP https://github.com/erdoukki/openwrt/tree/ebin-ultra-uboot-WIP
erdoukki
commented
2 years ago refresh rebuild done !
TIM-1.0
mv_ddr-devel-gc87315c6f5 DDR4 16b 1GB 1CS
WTMI-devel-18.12.1-2efdb10f
WTMI: system early-init
SVC REV: 5, CPU VDD voltage: 1.237V
Setting clocks: CPU 1200 MHz, DDR 750 MHz
CZ.NIC's Armada 3720 Secure Firmware v2021.09.07 (Oct 11 2021 16:00:41)
Running on ESPRESSObin Ultra
NOTICE: Booting Trusted Firmware
NOTICE: BL1: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL1: Built : 16:00:41, Oct 11 2021
NOTICE: BL1: Booting BL2
NOTICE: BL2: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL2: Built : 16:00:41, Oct 11 2021
NOTICE: BL1: Booting BL31
NOTICE: BL31: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL31: Built : 16:00:41, Oct 11 2021
U-Boot 2021.07 (Oct 11 2021 - 16:00:41 +0000)
DRAM: 1 GiB
WDT: Not starting
Comphy chip #0:
Comphy-0: USB3_HOST0 5 Gbps
Comphy-1: PEX0 2.5 Gbps
Comphy-2: SATA0 5 Gbps
Target spinup took 0 ms.
AHCI 0001.0300 32 slots 1 ports 6 Gbps 0x1 impl SATA mode
flags: ncq led only pmp fbss pio slum part sxs
PCIE-0: Link up
MMC: sdhci@d8000: 0
Loading Environment from SPIFlash... SF: Detected mx25u3235f with page size 256B
OK
Successfully imported the Marvell hw_info parameters.
Model: Globalscale Marvell ESPRESSOBin Ultra Board
Net: eth0: neta@30000 [PRIME]
Autoboot in 2 seconds, to stop use 's' key
=>
=> But still strange commit from mv_ddr-devel-gc87315c6f5 DDR4 16b 1GB 1CS which refer to nothing ?!
It should link to the commit from the Makefile : https://github.com/openwrt/openwrt/blob/af9d31aacc286786a8765a44c2000d2eba02e61c/package/boot/arm-trusted-firmware-mvebu/Makefile#L138
So, is it a bug in this https://github.com/openwrt/openwrt/blob/master/package/boot/arm-trusted-firmware-mvebu/patches-mv-ddr-marvell/001-version.patch
I will redo all my tests now...
This board was really buggy and is now updated to 202107 uboot with EspressoBin-Ultra support backported. I push on my fork of OpenWrt the code used : https://github.com/erdoukki/openwrt/tree/ebin-ultra-uboot-WIP
I have set back the clock to 1.2 GHz : Setting clocks: CPU 1200 MHz, DDR 750 MHz
erdoukki
commented
2 years ago Frozing or crashing very fast at OpenWrt boot with default 0xd0011500 value of 0x5aaaffff :
ERROR: Unhandled External Abort received on 0x80000001 at EL3!
ERROR: exception reason=0 syndrome=0xbf000001
PANIC at PC : 0x0000000004027368
[ 29.804946] rcu: INFO: rcu_sched detected stalls on CPUs/tasks:
[ 29.811132] (detected by 0, t=2102 jiffies, g=-327, q=143)
[ 29.816916] rcu: All QSes seen, last rcu_sched kthread activity 2102 (4294940
[ 29.829603] rcu: rcu_sched kthread starved for 2102 jiffies! g-327 f0x2 RCU_1
[ 29.840122] rcu: Unless rcu_sched kthread gets sufficient CPU time, OOM .
[ 29.849380] rcu: RCU grace-period kthread stack dump:
[ 29.854613] task:rcu_sched state:R stack: 0 pid: 11 ppid: 2 f8
[ 29.863274] Call trace:
[ 29.865844] __switch_to+0xec/0x100
[ 29.869483] __schedule+0x24c/0x544
[ 29.873113] schedule+0x48/0xd0
[ 29.876380] schedule_timeout+0x70/0xd0
[ 29.880366] rcu_gp_kthread+0x550/0xc34
[ 29.884355] kthread+0x124/0x12c
[ 29.887712] ret_from_fork+0x10/0x3c While setting the value to 0x7aaaffff just let the booting completely.
Will stress with this value to check stability...
Will also reflash a default 21.02.0 with 1.2 GHz blacklisted...
Actually I am with a custom master of OpenWrt with the 1.2 GHz un-blacklisted !
pali
commented
2 years ago But still strange commit from
mv_ddr-devel-gc87315c6f5 DDR4 16b 1GB 1CSwhich refer to nothing ?!
You have something broken. Compile firmware according to steps in ARM trusted firmware documentation at: https://trustedfirmware-a.readthedocs.io/en/latest/plat/marvell/armada/build.html Search for how to build production release of Marvell firmware image
erdoukki
commented
2 years ago It is a bug here :
So, is it a bug in this https://github.com/openwrt/openwrt/blob/master/package/boot/arm-trusted-firmware-mvebu/patches-mv-ddr-marvell/001-version.patch
It takes my local git commit id instead of the predefined one in the Makefile !
My local OpenWrt commit is : c87315c6f513352655eff2e1af3231cd71196199
I will try to fix it... Or open an issue at OpenWrt !
May be @dhewg can help on this ?
gerald@P3530:/media/gerald/EBINDEV/DEVEL/OWRT/openwrt/staging_dir/target-aarch64_cortex-a53_musl/image/mv-ddr-marvell$ cat mv_ddr_build_message.c
const char mv_ddr_build_message[] = "(Oct 11 2021 - 18:59:43)"; const char mv_ddr_version_string[] = "mv_ddr: devel-gc87315c6f5";Is it a bug upstream in mv-ddr-marvell ? else where ?
cat /media/gerald/EBINDEV/DEVEL/OWRT/openwrt/staging_dir/target-aarch64_cortex-a53_musl/image/mv-ddr-marvell/mv_ddr_build_message.c
const char mv_ddr_build_message[] = "(Oct 11 2021 - 17:22:42)"; const char mv_ddr_version_string[] = "mv_ddr: devel-g02e23dbc";Is it a bug upstream in mv-ddr-marvell ? else where ?
No. It is OpenWRT issue.
Use original upstream version and not patched which can just cause issues and like in this example just makes whole debugging problematic.
OpenWrt people know about this issue as they have introduced it and they just making fun about it.
erdoukki
commented
2 years ago Use original upstream version and not patched which can just cause issues and like in this example just makes whole debugging problematic.
Okay... sure... will follow the documentation you have given !
You have something broken. Compile firmware according to steps in ARM trusted firmware documentation at: https://trustedfirmware-a.readthedocs.io/en/latest/plat/marvell/armada/build.html Search for how to build production release of Marvell firmware image
Will report after...
And redo the tests ! to be sure...
dhewg
commented
2 years ago It takes my local git commit id instead of the predefined one in the Makefile ! My local OpenWrt commit is :
c87315c6f513352655eff2e1af3231cd71196199
That sounds like a bug indeed, and yes, specific to OpenWrt.
OpenWrt people know about this issue as they have introduced it and they just making fun about it.
That's not the whole story and you know it, but let's not reiterate that drama please.
About the issue at hand: It indeed is probably better to coordinate with manually building the firmware here. If and when the issue is resolved you can throw whatever needs fixing in an OpenWrt PR.
pali
commented
2 years ago @dhewg: shut up and go away, you OpenWRT guys are breaking everything which can be possible just because you can and to ensure that debugging is totally impossible and if you are asked to explain why you are doing it then you start talking about something totally irrelevant as you have no argument for it. Like in this your last comment.
Seems that knows everything and now should start care about all these issues.
So @erdoukki and @robimarko please forward this crash issue to @dhewg as he is the best person for fixing it as he knows everything about this issue.
Now I'm unsubscribing from this bug report and I'm not going to respond anymore.
dhewg
commented
2 years ago Right, how silly of me to even attempt any sort of coordination
robimarko
commented
2 years ago @dhewg @pali Come on guys, no point in doing this here. Manually building stuff should be the approach here, I suspect that OpenWrt does that passing of the hash due to them packaging everything that it downloads including git repos into XZ TAR archives. That is not really good as it breaks down traceability to the upstream but that doesn't help anybody with the voltage/clocking issues that are generic to the A3720.
So can we all just calm down and try to resolve this peacefully?
dhewg
commented
2 years ago Yeah, OpenWrt doesn't work with full fledged git clones for any packages AFAICT. Which is the reason it carries patches to keep said tracability. And that worked in the past, but there's a new spot which now requires a similar fix-up.
It's a bug that needs to be fixed. I don't know why this topic always gets answered in the unreasonable way we're seeing here.
erdoukki
commented
2 years ago Sorry all. I discovered this issue, that I may already know, but I have forget it. I also used OpenWrt for the u-boot of EspressoBin-Ultra because I think the works was simple and may be useful for future release. It helps to find this OpenWrt issue, but get the actual cpufreq tests a little more complicated !
I can come back to cpufreq tests now !
TIM-1.0
mv_ddr-devel-g02e23dbc-d DDR4 16b 1GB 1CS
WTMI-devel-18.12.1-2efdb10f
WTMI: system early-init
SVC REV: 5, CPU VDD voltage: 1.237V
Setting clocks: CPU 1200 MHz, DDR 750 MHz
CZ.NIC's Armada 3720 Secure Firmware v2021.09.07 (Oct 11 2021 16:00:41)
Running on ESPRESSObin Ultra
NOTICE: Booting Trusted Firmware
NOTICE: BL1: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL1: Built : 16:00:41, Oct 11 2021
NOTICE: BL1: Booting BL2
NOTICE: BL2: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL2: Built : 16:00:41, Oct 11 2021
NOTICE: BL1: Booting BL31
NOTICE: BL31: v2.5(release):OpenWrt v2.5-8 (espressobin-ultra)
NOTICE: BL31: Built : 16:00:41, Oct 11 2021
U-Boot 2021.07 (Oct 11 2021 - 16:00:41 +0000) Back again on test of CPU ! but... also... with a new question for you all : Can I had some DEBUG which may help about this issue analysis ?
erdoukki
commented
2 years ago Back again in CRASH !
Unsupported CPU frequency 1200 MHz
=> md 0xd0011500
d0011500: 5aaaffff 02000257 00008000 800001e1 ...ZW........... root@OpenWrt:/# stress-ng --matrix 0 -t 10m
stress-ng: info: [2107] dispatching hogs: 2 matrix [ 225.233642] Unable to handle kernel paging request at virtual address ffffff0
[ 225.241815] Mem abort info:
[ 225.244689] ESR = 0x96000047
[ 225.247834] EC = 0x25: DABT (current EL), IL = 32 bits
[ 225.253306] SET = 0, FnV = 0
[ 225.256449] EA = 0, S1PTW = 0
[ 225.259680] Data abort info:
[ 225.262644] ISV = 0, ISS = 0x00000047
[ 225.266593] CM = 0, WnR = 1
[ 225.269648] swapper pgtable: 4k pages, 39-bit VAs, pgdp=00000000078b8000
[ 225.276555] [ffffffc01092c2b0] pgd=000000003ffff003, pud=000000003ffff003, p0
[ 225.287501] Internal error: Oops: 96000047 [#1] SMP
[ 225.292525] Modules linked in: pppoe ppp_async iptable_nat xt_state xt_nat xg
[ 225.344846] CPU: 1 PID: 2109 Comm: stress-ng Not tainted 5.4.143 #0
[ 225.351304] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 225.358481] pstate: 80400085 (Nzcv daIf +PAN -UAO)
[ 225.363419] pc : 0xffffffc0100eacb8
[ 225.367009] lr : 0xffffffc0100d7df4
[ 225.370599] sp : ffffffc01000bcb0
[ 225.374010] x29: ffffffc01000bcc0 x28: 0000000000000060
[ 225.379484] x27: 0000000000000080 x26: 000000346f9a2970
[ 225.384958] x25: 000000000000b855 x24: 0000000047bb212c
[ 225.390431] x23: 0000000000000009 x22: ffffff803e06f140
[ 225.395904] x21: 0000003470f83be0 x20: ffffff803e085ac0
[ 225.401379] x19: ffffff803fde0700 x18: 0000000000000000
[ 225.406853] x17: 0000000000000000 x16: 0000000000000000
[ 225.412327] x15: 0000000000000000 x14: 0000000000000000
[ 225.417800] x13: 0000000000000000 x12: 0000000000000000
[ 225.423274] x11: 0000000000000000 x10: 0000000000000040
[ 225.428748] x9 : 0000000000000000 x8 : 0000000000080000
[ 225.434222] x7 : ffffffc0109388e0 x6 : ffffffc01092c2b0
[ 225.439696] x5 : ffffff803fde1280 x4 : ffffff803fde1280
[ 225.445169] x3 : 0000000000000000 x2 : ffffffc01092c280
[ 225.450643] x1 : ffffff803fde1288 x0 : ffffff803fde07c0
[ 225.456117] Call trace:
[ 225.458635] 0xffffffc0100eacb8
[ 225.461867] 0xffffffc0100d87f0
[ 225.465099] 0xffffffc0100cc698
[ 225.468331] 0xffffffc0100cca2c
[ 225.471563] 0xffffffc0100cd480
[ 225.474795] 0xffffffc0100cd55c
[ 225.478027] 0xffffffc01010de64
[ 225.481259] 0xffffffc01010e200
[ 225.484492] 0xffffffc01010ec58
[ 225.487724] 0xffffffc0105bb43c
[ 225.490955] 0xffffffc0100f8710
[ 225.494187] 0xffffffc0100f1e14
[ 225.497420] 0xffffffc0100f2638
[ 225.500652] 0xffffffc0100810ac
[ 225.503883] 0xffffffc010083328
[ 225.507123] Code: d37c0426 910020a1 8b060046 f863d8e3 (f82368c5)
[ 225.513404] ---[ end trace 33773a6e53afe8b6 ]---
[ 225.518161] Kernel panic - not syncing: Fatal exception in interrupt
[ 225.524713] SMP: stopping secondary CPUs
[ 225.528757] Kernel Offset: disabled
[ 225.532349] CPU features: 0x0002,00002008
[ 225.536477] Memory Limit: none
[ 225.539622] Rebooting in 3 seconds.. Unsupported CPU frequency 1200 MHz
=> mw 0xd0011500 4aaaffff
=> md 0xd0011500
d0011500: 4aaaffff 02000257 00008000 800001e1 ...JW........... stress-ng --matrix 0 -t 10m
[ 187.013477] ------------[ cut here ]------------
[ 187.018251] Unexpected interrupt received!
[ 187.022498] WARNING: CPU: 0 PID: 2107 at 0xffffffc01008119c
[ 187.028240] Modules linked in: pppoe ppp_async iptable_nat xt_state xt_nat xg
[ 187.080560] CPU: 0 PID: 2107 Comm: stress-ng Not tainted 5.4.143 #0
[ 187.087018] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 187.094194] pstate: 80400085 (Nzcv daIf +PAN -UAO)
[ 187.099129] pc : 0xffffffc01008119c
[ 187.102719] lr : 0xffffffc01008119c
[ 187.106309] sp : ffffffc010003fd0
[ 187.109720] x29: ffffffc010003fd0 x28: ffffff803c03c380
[ 187.115194] x27: 0000000000000000 x26: ffffffc010004000
[ 187.120668] x25: ffffffc010000000 x24: 0000000000000000
[ 187.126142] x23: 0000000020000000 x22: 000000000045f468
[ 187.131615] x21: ffffffc01095a000 x20: 000000000000001e
[ 187.137089] x19: 000000000000001e x18: 0000000000000000
[ 187.142562] x17: 0000000000000000 x16: 0000000000000000
[ 187.148036] x15: 0000000000000000 x14: ffffffc0109a2a10
[ 187.153509] x13: 0000000000000000 x12: ffffffc0109a2000
[ 187.158983] x11: ffffffc010946000 x10: 0000000000000010
[ 187.164457] x9 : 0000000000000000 x8 : 7265746e69206465
[ 187.169930] x7 : 7463657078656e55 x6 : 0000000000000001
[ 187.175404] x5 : 0000000000000000 x4 : 0000000000000001
[ 187.180877] x3 : 0000000000000007 x2 : 0000000000000006
[ 187.186350] x1 : 0000000000000007 x0 : 000000000000001e
[ 187.191825] Call trace:
[ 187.194342] 0xffffffc01008119c
[ 187.197573] 0xffffffc010083328
[ 187.200806] ---[ end trace 6ac2408c4ad89849 ]---Unsupported CPU frequency 1200 MHz
=> mw 0xd0011500 0x5CF3FFFF
=> md 0xd0011500
d0011500: 5cf3ffff 02000257 00008000 800001e1 ...\W...........root@OpenWrt:/# stress-ng --matrix 0 -t 10m
stress-ng: info: [2108] dispatching hogs: 2 matrix[ 53.892924] Unable to handle kernel write to read-only memory at virtual add1
[ 53.902080] Mem abort info:
[ 53.904954] ESR = 0x96000045
[ 53.908099] EC = 0x25: DABT (current EL), IL = 32 bits
[ 53.913571] SET = 0, FnV = 0
[ 53.916713] EA = 0, S1PTW = 0
[ 53.919945] Data abort info:
[ 53.922908] ISV = 0, ISS = 0x00000045
[ 53.926857] CM = 0, WnR = 1
[ 53.929912] user pgtable: 4k pages, 39-bit VAs, pgdp=000000003c9f0000
[ 53.936550] [00000000000009d1] pgd=000000003d5cb003, pud=000000003d5cb003, p0
[ 53.945524] Internal error: Oops: 96000045 [#1] SMP
[ 53.950550] Modules linked in: pppoe ppp_async iptable_nat xt_state xt_nat xg
[ 54.002869] CPU: 1 PID: 2110 Comm: stress-ng Not tainted 5.4.143 #0
[ 54.009327] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 54.016504] pstate: 20400085 (nzCv daIf +PAN -UAO)
[ 54.021440] pc : 0xffffffc0100e7a14
[ 54.025030] lr : 0xffffffc0100d3c94
[ 54.028620] sp : ffffffc01000bce0
[ 54.032031] x29: ffffffc01000bce0 x28: afb504000afb5041
[ 54.037505] x27: ffffffc010927568 x26: ffffff803c146540
[ 54.042979] x25: ffffffc0109a14c0 x24: 0000000000000001
[ 54.048453] x23: 000000000000000a x22: ffffff803c146540
[ 54.053927] x21: 0000000000000001 x20: ffffff803c1466c0
[ 54.059400] x19: ffffff803c146600 x18: 0000000000000000
[ 54.064874] x17: 0000000000000000 x16: 0000000000000000
[ 54.070348] x15: 0000000000000000 x14: 0000000000000000
[ 54.075822] x13: 0000000000000000 x12: 0000000000000000
[ 54.081296] x11: 0000000000000000 x10: 0000000000000040
[ 54.086770] x9 : ffffffc010946f40 x8 : ffffffc010946f38
[ 54.092244] x7 : ffffff803e400288 x6 : 000000009733e079
[ 54.097718] x5 : 0000000000002919 x4 : 0000000c8babc400
[ 54.103191] x3 : 0000000c8c446400 x2 : ffffff803c146600
[ 54.108665] x1 : 00000000000002f2 x0 : 0056e47be90db135
[ 54.114139] Call trace:
[ 54.116656] 0xffffffc0100e7a14
[ 54.119888] 0xffffffc0100d3c94
[ 54.123119] 0xffffffc0100ce648
[ 54.126351] 0xffffffc01010da4c
[ 54.129584] 0xffffffc01011d7c4
[ 54.132816] 0xffffffc01010e200
[ 54.136048] 0xffffffc01010ec58
[ 54.139280] 0xffffffc0105bb43c
[ 54.142512] 0xffffffc0100f8710
[ 54.145744] 0xffffffc0100f1e14
[ 54.148975] 0xffffffc0100f2638
[ 54.152208] 0xffffffc0100810ac
[ 54.155440] 0xffffffc010083328
[ 54.158679] Code: 2a0503e3 d2800039 b9001e85 f9400681 (8b204021)
[ 54.164960] ---[ end trace 083395d87b832c9f ]---
[ 54.169718] Kernel panic - not syncing: Fatal exception in interrupt
[ 54.176269] SMP: stopping secondary CPUs
[ 54.180314] Kernel Offset: disabled
[ 54.183906] CPU features: 0x0002,00002008
[ 54.188033] Memory Limit: none
[ 54.191178] Rebooting in 3 seconds..@erdoukki I dont see a point in further testing. This is something that Marvell needs to fix or release the docs.
erdoukki
commented
2 years ago Yes, I agree...
erdoukki
commented
2 years ago I read in Hardware Specifications
AVS is mandatory. VDD must be controlled by AVS circuitry.
and also this :
NOTE: If AVS is not applied, the maximum CPU frequency is limited to 600 MHz.
Do I understand well if I say that CPU frequency must stay limited to 600 MHz and cannot be more without AVS mode ? What is DFS behavior in this AVS control mode ? May be required also ?
Is it possible that VDD are correct but divider are incorrect ?
I have this feeling after a successful test few days ago with one of the two crashing EBIN-Ultra. It was done with a modified value on 0xd0011500, un-blacklisting of the 1.2 CPU frequency and divider modified.
I am trying to reproduce this on the second, more unstable EBIN-Ultra. May be I am on a bad analysis, but I share it "as-is", in case...
erdoukki
commented
2 years ago I am trying to reproduce this on the second, more unstable ESPRESSOBIN-Ultra.
Never get able to reproduce ! It was a strange but temporary success...
Need more patience and hope from a Marvell super-hero ? May be a Marvell Developer will be more realistic, or may be not.
I get some partial feedback from Marvell Support. It was confirmed of a known issue. But no more support and EOL SOC.
erdoukki
commented
2 years ago It was said by the Marvell Support Team that the latest SDK10 may already have fixed this issue !
robimarko
commented
2 years ago Well, if only their SDK was public
erdoukki
commented
2 years ago I was able to get access to the latest SDK10. I will try to get some free time and experiment the linux patches of the SDK10 on GlobalScaleTechnologies Ultra 1.2GHz… Will report results here when it will be done
robimarko
commented
2 years ago @erdoukki Any news? I still dont have access to SDK10 at all.
erdoukki
commented
2 years ago @robimarko Still no news... I have no free time for now... I plan to compile the SDK10 kernel to stress test the bugyg ULTRA !
erdoukki
commented
2 years ago I get the latest SDK10 and compile a full image for device : a37xx_espressobin_1000_800
It use a linux kernel version 4.14.207
I will try the system on an espressobin and will also try later on my buggy ultra !
erdoukki
commented
2 years ago close by error ! sorry
erdoukki
commented
2 years ago @pali @robimarko any advice to help on tests with the official SDK10 is welcome...
erdoukki
commented
2 years ago Héllo all, Have an nice and happy new year...
I get some new issues around the CPU bug on the 37xx. One of my EspressoBin-Ultra, which is mostly stable, get some reboot, on heavy CPU and network load.
I may look at it deeper if needed, because it is one of my working ULTRA, which get reboot only one CPU load... I use it as a 4G-Mobile routeur/gateway.
Add: I have to get in the testing of the official Marvell SDK, but it was postpone for now... STAY TUNED
erdoukki
commented
2 years ago FROM SDK10 (SDK-10.3.9.0) and OpenWrt 21.02.0, r16279-5cc0535800
compile the release version with:
compile.sh a37xx_espressobin_1000_800 -r SDK-10.3.9.0Booting an ESPRESSObin-ULTRA (one of my mostly unstable... checked before tests and confirmed to still CRASHING few seconds only after boot in OpenWrt with default kernel)
[ 0.000000] Linux version 5.4.143 (builder@buildhost) (gcc version 8.4.0 (OpenWrt GCC 8.4.0 r16279-5cc0535800)) #0 SMP Tue Aug 31 22:20:08 2021
[ 0.000000] Machine model: Globalscale Marvell ESPRESSOBin Ultra Board root@OpenWrt:/# dmesg | grep CPU
[ 0.000000] Booting Linux on physical CPU 0x0000000000 [0x410fd034]
[ 0.000000] Detected VIPT I-cache on CPU0
[ 0.000000] CPU features: detected: GIC system register CPU interface
[ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=2, Nodes=1
[ 0.000000] rcu: RCU restricting CPUs from NR_CPUS=4 to nr_cpu_ids=2.
[ 0.000000] GICv3: CPU0: found redistributor 0 region 0:0x00000000d1d40000
[ 0.051306] smp: Bringing up secondary CPUs ...
[ 0.056287] Detected VIPT I-cache on CPU1
[ 0.056315] GICv3: CPU1: found redistributor 1 region 0:0x00000000d1d60000
[ 0.056348] CPU1: Booted secondary processor 0x0000000001 [0x410fd034]
[ 0.056427] smp: Brought up 1 node, 2 CPUs
[ 0.083316] CPU features: detected: 32-bit EL0 Support
[ 0.088607] CPU features: detected: CRC32 instructions
[ 0.093923] CPU features: emulated: Privileged Access Never (PAN) using TTBR0_EL1 switching
[ 0.102496] CPU: All CPU(s) started at EL2
[ 0.889208] cacheinfo: Unable to detect cache hierarchy for CPU 0
[ 2.117397] Unsupported CPU frequency 1200 MHz
Then after few seconds...
[ 269.289957] ------------[ cut here ]------------
[ 269.294726] bdi-block not registered
[ 269.298425] WARNING: CPU: 1 PID: 937 at 0xffffffc010211bec
[ 269.304077] Modules linked in: pppoe ppp_async iptable_nat xt_state xt_nat xt_conntrack xt_REDIRECT xt_MASQUERADE xt_FLOWOFFLOAD xt_CT pppox ppp_generic nf_nat nf_flow_table_hw nf_flow_tableg
[ 269.356396] CPU: 1 PID: 937 Comm: ash Not tainted 5.4.143 #0
[ 269.362226] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 269.369403] pstate: 80400005 (Nzcv daif +PAN -UAO)
[ 269.374338] pc : 0xffffffc010211bec
[ 269.377929] lr : 0xffffffc010211bec
[ 269.381520] sp : ffffffc010e73d60
[ 269.384931] x29: ffffffc010e73d60 x28: ffffff803e0521c0
[ 269.390405] x27: 0000000000000000 x26: 0000000040000001
[ 269.395879] x25: 0000000000000001 x24: 0000000000000001
[ 269.401352] x23: 0000000000000000 x22: 0000000000000000
[ 269.406826] x21: 0000000000000001 x20: ffffff803cdb6878
[ 269.412300] x19: ffffff803e721d58 x18: 0000000000000000
[ 269.417773] x17: 0000000000000000 x16: 0000000000000000
[ 269.423247] x15: 0000000000000000 x14: ffffffc0109a2a10
[ 269.428721] x13: 0000000000000000 x12: ffffffc0109a2000
[ 269.434195] x11: ffffffc010946000 x10: 0000000000000010
[ 269.439668] x9 : 0000000000000000 x8 : 6465726574736967
[ 269.445142] x7 : 657220746f6e206b x6 : 0000000000000001
[ 269.450615] x5 : 0000000000000000 x4 : 0000000000000001
[ 269.456089] x3 : 0000000000000007 x2 : 0000000000000006
[ 269.461562] x1 : 0000000000000007 x0 : 0000000000000018
[ 269.467037] Call trace:
[ 269.469554] 0xffffffc010211bec
[ 269.472785] 0xffffffc0101fdbac
[ 269.476017] 0xffffffc010200bcc
[ 269.479249] 0xffffffc0101f4318
[ 269.482481] 0xffffffc0101f5a04
[ 269.485712] 0xffffffc0101f5c2c
[ 269.488945] 0xffffffc010094fac
[ 269.492178] 0xffffffc010083748
[ 269.495411] ---[ end trace de9ce09de484892c ]--- Now entering the SDK10 tests !
Just booting with SDK10 Image (and modules) in OpenWrt 21.02.0...
Marvell>> setenv bootargs $console root=/dev/mmcblk0p2 rw rootwait net.ifnames=0 biosdevname=0 $extra_params usb-storage.quirks=$usbstoragequirks
Marvell>> load usb 0 $kernel_addr_r ULTRA-SDK10/Image
Marvell>> booti $kernel_addr_r - $fdt_addr_r
## Flattened Device Tree blob at 06f00000
Booting using the fdt blob at 0x6f00000
Using Device Tree in place at 0000000006f00000, end 0000000006f05fbf root@OpenWrt:/# uname -ar
Linux OpenWrt 4.14.207-10.3.9.0-2 #1 SMP PREEMPT Wed Jan 12 15:53:40 CET 2022 aarch64 GNU/Linux root@OpenWrt:/# stress-ng --matrix 0 -t 10mroot@OpenWrt:/# dmesg | grep CPU
[ 0.000000] Booting Linux on physical CPU 0x0
[ 0.000000] Boot CPU: AArch64 Processor [410fd034]
[ 0.000000] Detected VIPT I-cache on CPU0
[ 0.000000] CPU features: enabling workaround for ARM erratum 845719
[ 0.000000] CPU features: kernel page table isolation disabled by kernel configuration
[ 0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=2, Nodes=1
[ 0.000000] RCU restricting CPUs from NR_CPUS=96 to nr_cpu_ids=2.
[ 0.000000] GICv3: CPU0: found redistributor 0 region 0:0x00000000d1d40000
[ 0.000000] NO_HZ: Full dynticks CPUs: 1.
[ 0.000000] Note: kernel parameter 'rcu_nocbs=' contains nonexistent CPUs.
[ 0.000000] Offload RCU callbacks from CPUs: 1.
[ 0.116792] smp: Bringing up secondary CPUs ...
[ 0.149654] Detected VIPT I-cache on CPU1
[ 0.149682] GICv3: CPU1: found redistributor 1 region 0:0x00000000d1d60000
[ 0.149714] CPU1: Booted secondary processor [410fd034]
[ 0.149830] smp: Brought up 1 node, 2 CPUs
[ 0.175386] CPU features: detected: GIC system register CPU interface
[ 0.182026] CPU features: detected: 32-bit EL0 Support
[ 0.187456] CPU: All CPU(s) started at EL2
[ 1.880786] kvm [1]: GIC system register CPU interface enabled
[ 2.032688] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.046744] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.14.207-10.3.9.0-2 #1
[ 2.245749] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.259806] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W 4.14.207-10.3.9.0-2 #1
[ 2.460047] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.474104] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W 4.14.207-10.3.9.0-2 #1
[ 3.110841] cacheinfo: Unable to detect cache hierarchy for CPU 0 stressed with:
crash (but not reset):
[ 2.027996] ------------[ cut here ]------------
[ 2.032688] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.043604] Modules linked in:
[ 2.046744] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.14.207-10.3.9.0-2 #1
[ 2.053997] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 2.061166] task: ffff80002e9b8000 task.stack: ffff80002e9b4000
[ 2.067260] pc : mvebu_comphy_probe+0x2f8/0x330
[ 2.071919] lr : mvebu_comphy_probe+0x2f8/0x330
[ 2.076576] sp : ffff80002e9b7be0 pstate : 60000045
[ 2.081594] x29: ffff80002e9b7be0 x28: ffff80002e005800
[ 2.087059] x27: ffff80002e0b9b80 x26: ffff000008ecf8a8
[ 2.092525] x25: 00000000014080c0 x24: ffff000008ecfbb0
[ 2.097990] x23: ffff80002e0b9680 x22: ffff80002eb51810
[ 2.103456] x21: ffff80002eb51800 x20: ffff0000091665a8
[ 2.108922] x19: ffff80002ffeb300 x18: 0000000000000010
[ 2.114387] x17: 0000000000000003 x16: 0000000000000000
[ 2.119853] x15: ffffffffffffffff x14: 0000000000000000
[ 2.125318] x13: 0000000000000000 x12: 0000000078696cc0
[ 2.130784] x11: 0000000000000000 x10: 00000000000009f0
[ 2.136250] x9 : ffff80002e9b7950 x8 : ffff80002e9b8a50
[ 2.141715] x7 : 0000000000000400 x6 : 0000000000000108
[ 2.147181] x5 : 0000000000000002 x4 : 0000000000000001
[ 2.152646] x3 : fffffffffffffffe x2 : ffff80002e9b7910
[ 2.158112] x1 : ffff000009468ae0 x0 : 0000000000000032
[ 2.163578] Call trace:
[ 2.166089] mvebu_comphy_probe+0x2f8/0x330
[ 2.170393] platform_drv_probe+0x58/0xc0
[ 2.174511] driver_probe_device+0x248/0x2e0
[ 2.178901] __driver_attach+0xbc/0xc0
[ 2.182757] bus_for_each_dev+0x4c/0xa0
[ 2.186696] driver_attach+0x20/0x30
[ 2.190369] bus_add_driver+0x1b0/0x220
[ 2.194312] driver_register+0x60/0x100
[ 2.198255] __platform_driver_register+0x40/0x50
[ 2.203098] mvebu_comphy_driver_init+0x18/0x20
[ 2.207755] do_one_initcall+0x38/0x130
[ 2.211697] kernel_init_freeable+0x184/0x220
[ 2.216178] kernel_init+0x10/0x110
[ 2.219759] ret_from_fork+0x10/0x24
[ 2.223436] ---[ end trace 97c6934e1fd8503a ]---
[ 2.228520] mvebu-comphy d0018300.phy: RELYING ON BOTLOADER SETTINGS
[ 2.235043] mvebu-comphy d0018300.phy: firmware updated needed
[ 2.241036] ------------[ cut here ]------------
[ 2.245749] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.256665] Modules linked in:
[ 2.259806] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W 4.14.207-10.3.9.0-2 #1
[ 2.268313] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 2.275482] task: ffff80002e9b8000 task.stack: ffff80002e9b4000
[ 2.281576] pc : mvebu_comphy_probe+0x2f8/0x330
[ 2.286235] lr : mvebu_comphy_probe+0x2f8/0x330
[ 2.290892] sp : ffff80002e9b7be0 pstate : 40000045
[ 2.295910] x29: ffff80002e9b7be0 x28: ffff80002e005400
[ 2.301375] x27: ffff80002e0b9180 x26: ffff000008ecf8a8
[ 2.306841] x25: 00000000014080c0 x24: ffff000008ecfbb0
[ 2.312306] x23: ffff80002e0b9680 x22: ffff80002eb51810
[ 2.317772] x21: ffff80002eb51800 x20: ffff0000091665a8
[ 2.323238] x19: ffff80002ffeb580 x18: 0000000000000010
[ 2.328703] x17: 0000000000000004 x16: 0000000000000000
[ 2.334169] x15: ffffffffffffffff x14: ffff000089454587
[ 2.339634] x13: ffff000009454595 x12: ffff000009389000
[ 2.345100] x11: 0000000005f5e0ff x10: ffff80002e9b7910
[ 2.350566] x9 : ffff000008726f40 x8 : 000000000000000d
[ 2.356031] x7 : 776d726966203a79 x6 : 00000000000000e5
[ 2.361497] x5 : 0000000000000000 x4 : 0000000000000000
[ 2.366962] x3 : ffffffffffffffff x2 : ffff0000093897e0
[ 2.372428] x1 : ffff80002e9b8000 x0 : 0000000000000032
[ 2.377894] Call trace:
[ 2.380405] mvebu_comphy_probe+0x2f8/0x330
[ 2.384708] platform_drv_probe+0x58/0xc0
[ 2.388827] driver_probe_device+0x248/0x2e0
[ 2.393216] __driver_attach+0xbc/0xc0
[ 2.397073] bus_for_each_dev+0x4c/0xa0
[ 2.401012] driver_attach+0x20/0x30
[ 2.404685] bus_add_driver+0x1b0/0x220
[ 2.408628] driver_register+0x60/0x100
[ 2.412570] __platform_driver_register+0x40/0x50
[ 2.417412] mvebu_comphy_driver_init+0x18/0x20
[ 2.422070] do_one_initcall+0x38/0x130
[ 2.426013] kernel_init_freeable+0x184/0x220
[ 2.430492] kernel_init+0x10/0x110
[ 2.434075] ret_from_fork+0x10/0x24
[ 2.437747] ---[ end trace 97c6934e1fd8503b ]---
[ 2.442820] mvebu-comphy d0018300.phy: RELYING ON BOTLOADER SETTINGS
[ 2.449319] mvebu-comphy d0018300.phy: firmware updated needed
[ 2.455364] ------------[ cut here ]------------
[ 2.460047] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
[ 2.470964] Modules linked in:
[ 2.474104] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W 4.14.207-10.3.9.0-2 #1
[ 2.482612] Hardware name: Globalscale Marvell ESPRESSOBin Ultra Board (DT)
[ 2.489781] task: ffff80002e9b8000 task.stack: ffff80002e9b4000
[ 2.495875] pc : mvebu_comphy_probe+0x2f8/0x330
[ 2.500534] lr : mvebu_comphy_probe+0x2f8/0x330
[ 2.505191] sp : ffff80002e9b7be0 pstate : 60000045
[ 2.510208] x29: ffff80002e9b7be0 x28: ffff80002e005000
[ 2.515674] x27: ffff80002dcb4380 x26: ffff000008ecf8a8
[ 2.521139] x25: 00000000014080c0 x24: ffff000008ecfbb0
[ 2.526605] x23: ffff80002e0b9680 x22: ffff80002eb51810
[ 2.532071] x21: ffff80002eb51800 x20: ffff0000091665a8
[ 2.537536] x19: ffff80002ffeb800 x18: 0000000000000010
[ 2.543002] x17: 0000000000000003 x16: 0000000000000000
[ 2.548467] x15: ffffffffffffffff x14: 0000000000000000
[ 2.553933] x13: 0000000000000000 x12: 0000000091e01c30
[ 2.559398] x11: 0000000000000000 x10: 00000000000009f0
[ 2.564864] x9 : ffff80002e9b7950 x8 : ffff80002e9b8a50
[ 2.570330] x7 : 0000000000000400 x6 : 00000000000002b0
[ 2.575795] x5 : 0000000000000002 x4 : 0000000000000001
[ 2.581261] x3 : fffffffffffffffe x2 : ffff80002e9b7910
[ 2.586726] x1 : ffff000009468ae0 x0 : 0000000000000032
[ 2.592193] Call trace:
[ 2.594704] mvebu_comphy_probe+0x2f8/0x330
[ 2.599007] platform_drv_probe+0x58/0xc0
[ 2.603126] driver_probe_device+0x248/0x2e0
[ 2.607515] __driver_attach+0xbc/0xc0
[ 2.611372] bus_for_each_dev+0x4c/0xa0
[ 2.615310] driver_attach+0x20/0x30
[ 2.618984] bus_add_driver+0x1b0/0x220
[ 2.622926] driver_register+0x60/0x100
[ 2.626869] __platform_driver_register+0x40/0x50
[ 2.631711] mvebu_comphy_driver_init+0x18/0x20
[ 2.636369] do_one_initcall+0x38/0x130
[ 2.640311] kernel_init_freeable+0x184/0x220
[ 2.644792] kernel_init+0x10/0x110
[ 2.648374] ret_from_fork+0x10/0x24
[ 2.652046] ---[ end trace 97c6934e1fd8503c ]--- PANICs l ooks like to be from something else:
[ 2.460047] WARNING: CPU: 0 PID: 1 at drivers/phy/marvell/phy-mvebu-cp110-comphy.c:536 mvebu_comphy_probe+0x2f8/0x330
more information with the working kernel from SDK10
root@OpenWrt:~# lscpu
Architecture: aarch64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
CPU(s): 2
On-line CPU(s) list: 0,1
Thread(s) per core: 1
Core(s) per socket: 2
Socket(s): 1
NUMA node(s): 1
Vendor ID: ARM
Model: 4
Model name: Cortex-A53
Stepping: r0p4
CPU max MHz: 1200.0000
CPU min MHz: 200.0000
BogoMIPS: 25.00
NUMA node0 CPU(s): 0,1
Vulnerability Itlb multihit: Not affected
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Spec store bypass: Not affected
Vulnerability Spectre v1: Mitigation; __user pointer sanitization
Vulnerability Spectre v2: Not affected
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Not affected
Flags: fp asimd aes pmull sha1 sha2 crc32 cpuid
https://github.com/openwrt/openwrt/commit/f407b2f43c27a7b35c4f96d3efcee2cc440f8efe
How can we contact Marvell to have the needed information ?