All of lore.kernel.org
 help / color / mirror / Atom feed
* Tioga Pass OpenBMC 2.8
@ 2020-08-22  1:36 Neil Bradley
  2020-08-24 17:07 ` Bruce Mitchell
  2020-08-24 21:05 ` Vijay Khemka
  0 siblings, 2 replies; 40+ messages in thread
From: Neil Bradley @ 2020-08-22  1:36 UTC (permalink / raw)
  To: openbmc; +Cc: Bruce Mitchell, Patrick Voelker

[-- Attachment #1: Type: text/plain, Size: 1047 bytes --]

Greetings fellow BMC folk - I'm Neil Bradley, Phoenix Technologies' BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we've encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We're coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 7889 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-22  1:36 Tioga Pass OpenBMC 2.8 Neil Bradley
@ 2020-08-24 17:07 ` Bruce Mitchell
  2020-08-24 22:52   ` Vijay Khemka
  2020-08-24 21:05 ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Bruce Mitchell @ 2020-08-24 17:07 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Patrick Voelker

In addition to what Neil has ask I am find this issue with NC-SI:

Yet Ben Wei's OSFC 2019 "NIC Management and Monitoring in OpenBMC" 
leads me to believe that NC-SI should be in my image and functioning.
Has something changed since Ben's paper?

root@tiogapass:~#
root@tiogapass:~# uname -a
Linux tiogapass 5.4.39-30079d6 #1 Fri Aug 21 17:14:22 UTC 2020 armv6l GNU/Linux
root@tiogapass:~# cat /proc/version
Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
root@tiogapass:~# cat /proc/sys/kernel/version
#1 Fri Aug 21 17:14:22 UTC 2020
root@tiogapass:~# cat /etc/os-release
ID=openbmc-phosphor
NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro)"
VERSION="v00.000.000-45"
VERSION_ID=v00.000.000-45-g645f5cc08
PRETTY_NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro) v00.000.000-45"
BUILD_ID="v00.000.000"
OPENBMC_TARGET_MACHINE="tiogapass"
root@tiogapass:~# dmesg | grep -i ncsi
[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface
[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
root@tiogapass:~#

From: Neil Bradley 
Sent: Friday, August 21, 2020 18:36
To: openbmc@lists.ozlabs.org
Cc: Bruce Mitchell; Patrick Voelker
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk ? I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:

1. Host will not power on via the front panel power button
2. BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:

1. Was this built/done correctly? In other words, did I do something wrong in this process?
2. Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

→Neil

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-22  1:36 Tioga Pass OpenBMC 2.8 Neil Bradley
  2020-08-24 17:07 ` Bruce Mitchell
@ 2020-08-24 21:05 ` Vijay Khemka
  2020-08-24 21:17   ` Neil Bradley
  1 sibling, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-24 21:05 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Patrick Voelker, Bruce Mitchell

[-- Attachment #1: Type: text/plain, Size: 1717 bytes --]

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org> on behalf of Neil Bradley <Neil_Bradley@phoenix.com>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com>, Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 8967 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-24 21:05 ` Vijay Khemka
@ 2020-08-24 21:17   ` Neil Bradley
  2020-08-24 21:32     ` Neil Bradley
  2020-08-24 22:58     ` Vijay Khemka
  0 siblings, 2 replies; 40+ messages in thread
From: Neil Bradley @ 2020-08-24 21:17 UTC (permalink / raw)
  To: openbmc

[-- Attachment #1: Type: text/plain, Size: 4659 bytes --]

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com>; Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 15000 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-24 21:17   ` Neil Bradley
@ 2020-08-24 21:32     ` Neil Bradley
  2020-08-24 23:02       ` Vijay Khemka
  2020-08-24 22:58     ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-24 21:32 UTC (permalink / raw)
  To: openbmc

[-- Attachment #1: Type: text/plain, Size: 20201 bytes --]

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 59915 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-24 17:07 ` Bruce Mitchell
@ 2020-08-24 22:52   ` Vijay Khemka
  0 siblings, 0 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-24 22:52 UTC (permalink / raw)
  To: Bruce Mitchell, Neil Bradley, openbmc; +Cc: Patrick Voelker

Hi Bruce,
What NCSI issues you are facing, kernel messages mentioned below are initial log when ncsi configuration completes but it doesn't create any issues in functioning of NCSI. Is your network working? Or are you facing anything with NCSI. You don't need anything for NCSI management if you are just using as network card and not doing any NCSI management from application. Network works perfectly for ipv4 and ipv6 seamlessly since 1 year.

Please let me know actual issues you are facing, I will try to fix it. 

Regards
-Vijay 

On 8/24/20, 2:17 PM, "openbmc on behalf of Bruce Mitchell" <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org on behalf of Bruce_Mitchell@phoenix.com> wrote:

    In addition to what Neil has ask I am find this issue with NC-SI:

    Yet Ben Wei's OSFC 2019 "NIC Management and Monitoring in OpenBMC" 
    leads me to believe that NC-SI should be in my image and functioning.
    Has something changed since Ben's paper?

    root@tiogapass:~#
    root@tiogapass:~# uname -a
    Linux tiogapass 5.4.39-30079d6 #1 Fri Aug 21 17:14:22 UTC 2020 armv6l GNU/Linux
    root@tiogapass:~# cat /proc/version
    Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
    root@tiogapass:~# cat /proc/sys/kernel/version
    #1 Fri Aug 21 17:14:22 UTC 2020
    root@tiogapass:~# cat /etc/os-release
    ID=openbmc-phosphor
    NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro)"
    VERSION="v00.000.000-45"
    VERSION_ID=v00.000.000-45-g645f5cc08
    PRETTY_NAME="Phosphor OpenBMC (Phosphor OpenBMC Project Reference Distro) v00.000.000-45"
    BUILD_ID="v00.000.000"
    OPENBMC_TARGET_MACHINE="tiogapass"
    root@tiogapass:~# dmesg | grep -i ncsi
    [    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface
    [   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
    [   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    [   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
    root@tiogapass:~#

    From: Neil Bradley 
    Sent: Friday, August 21, 2020 18:36
    To: openbmc@lists.ozlabs.org
    Cc: Bruce Mitchell; Patrick Voelker
    Subject: Tioga Pass OpenBMC 2.8

    Greetings fellow BMC folk ? I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

    We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

    mkdir tiogapass
    cd tiogapass
    git clone https://github.com/openbmc/openbmc.git
    cd openbmc
    git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
    export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
    source openbmc-env
    bitbake obmc-phosphor-image

    The BMC does boot fully, however there are two major problems we’ve encountered:

    1. Host will not power on via the front panel power button
    2. BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

    I have a couple of questions:

    1. Was this built/done correctly? In other words, did I do something wrong in this process?
    2. Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

    We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

    →Neil



^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-24 21:17   ` Neil Bradley
  2020-08-24 21:32     ` Neil Bradley
@ 2020-08-24 22:58     ` Vijay Khemka
  1 sibling, 0 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-24 22:58 UTC (permalink / raw)
  To: Neil Bradley, openbmc

[-- Attachment #1: Type: text/plain, Size: 5324 bytes --]

Please ignore below NCSI message, I am assuming that you have image boot up to login prompt. You can use power-util to power on or off host system. Yes you need ME support in your UEFI image, as when you power on system BMC reads and monitor ME sensor via IPMB but if ME is not available then ipmb will throw error.

Please let me know any other issues you are getting, I will try to resolve that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org> on behalf of Neil Bradley <Neil_Bradley@phoenix.com>
Date: Monday, August 24, 2020 at 2:19 PM
To: "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com>; Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 15817 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-24 21:32     ` Neil Bradley
@ 2020-08-24 23:02       ` Vijay Khemka
  2020-08-24 23:34         ` Neil Bradley
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-24 23:02 UTC (permalink / raw)
  To: Neil Bradley, openbmc

[-- Attachment #1: Type: text/plain, Size: 20702 bytes --]

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org> on behalf of Neil Bradley <Neil_Bradley@phoenix.com>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 60341 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-24 23:02       ` Vijay Khemka
@ 2020-08-24 23:34         ` Neil Bradley
  2020-08-25  0:00           ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-24 23:34 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Bruce Mitchell

[-- Attachment #1: Type: text/plain, Size: 22162 bytes --]

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 63292 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-24 23:34         ` Neil Bradley
@ 2020-08-25  0:00           ` Vijay Khemka
  2020-08-25  1:35             ` Neil Bradley
  2020-08-25 22:22             ` Neil Bradley
  0 siblings, 2 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-25  0:00 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Bruce Mitchell, Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 23956 bytes --]

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 69443 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-25  0:00           ` Vijay Khemka
@ 2020-08-25  1:35             ` Neil Bradley
  2020-08-25 14:25               ` Patrick Williams
  2020-08-25 22:22             ` Neil Bradley
  1 sibling, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-25  1:35 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Bruce Mitchell, Amithash Prasad, Sai Dasari


[-- Attachment #1.1: Type: text/plain, Size: 25292 bytes --]

Thanks Vijay! Take a look at the attached picture. It occurs to me that the RJ-45 jack on the left side with the purple cable plugged in may not actually be a network connector (we got these machines last Thursday and are just getting acquainted with them), so please forgive my ignorance in case this isn’t what it’s supposed to be. 😉 The working ethernet port is the leftmost SFP+ connector on the mezzanine card (below the sticker), and the rightmost SFP+ connector doesn’t provide a leach. My assumption, which may be incorrect, is the leftmost ethernet jack is RGMII2 and the NICs on the mezzanine is RGMII1 (at least according to the schematic which shows it as an OCP connection).

Will try the ME commands you suggested. Thank you very much!

-->Neil



From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>; Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #1.2: Type: text/html, Size: 72026 bytes --]

[-- Attachment #2: TiogaPass.png --]
[-- Type: image/png, Size: 5352845 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-25  1:35             ` Neil Bradley
@ 2020-08-25 14:25               ` Patrick Williams
  2020-08-26 21:38                 ` Neil Bradley
  0 siblings, 1 reply; 40+ messages in thread
From: Patrick Williams @ 2020-08-25 14:25 UTC (permalink / raw)
  To: Neil Bradley; +Cc: Vijay Khemka, openbmc, Bruce Mitchell, Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 1847 bytes --]

Hello Neil,

On Tue, Aug 25, 2020 at 01:35:12AM +0000, Neil Bradley wrote:
> Thanks Vijay! Take a look at the attached picture. It occurs to me that the RJ-45 jack on the left side with the purple cable plugged in may not actually be a network connector (we got these machines last Thursday and are just getting acquainted with them), so please forgive my ignorance in case this isn’t what it’s supposed to be. 😉 The working ethernet port is the leftmost SFP+ connector on the mezzanine card (below the sticker), and the rightmost SFP+ connector doesn’t provide a leach. My assumption, which may be incorrect, is the leftmost ethernet jack is RGMII2 and the NICs on the mezzanine is RGMII1 (at least according to the schematic which shows it as an OCP connection).

From
https://www.opencompute.org/documents/facebook-server-intel-motherboard-v40-spec:

10.4.2 Management Network
The motherboard has three options of management network interfaces for the BMC
connection. The management network shares the data network’s physical interface. The
management connection should be independent from data traffic and OS/driver
condition.
 SFP+ shared-NIC from Mezzanine 10G NIC or PCIe NIC, driven by BMC through
RMII/NC-SI or I2C. I2C being default
 SGMII/KX shared-NIC connected to midplane interface from Intel® I210-AS, driven by
BMC through RMII/NCSI
 10/100/1000 MDI shared-NIC connected to RJ45 from Intel® I210-AT(co-layout with
Intel® I210-AS), driven by BMC through RMII/NCSI 


---

So it seems like that is another RMII/NCSI attached network interface.
We always use the "SPF+ shared-NIC (10G)" though in our lab.  I don't
know that we've ever attempted to get the Intel chip working.  There is
probably some device tree content, at least, missing for enabling that.

-- 
Patrick Williams

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-25  0:00           ` Vijay Khemka
  2020-08-25  1:35             ` Neil Bradley
@ 2020-08-25 22:22             ` Neil Bradley
  2020-08-26 16:46               ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-25 22:22 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 25644 bytes --]

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>; Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 72907 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-25 22:22             ` Neil Bradley
@ 2020-08-26 16:46               ` Vijay Khemka
  2020-08-26 16:48                 ` Neil Bradley
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-26 16:46 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 26105 bytes --]

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>; Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 73551 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-26 16:46               ` Vijay Khemka
@ 2020-08-26 16:48                 ` Neil Bradley
  2020-08-26 16:53                   ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-26 16:48 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 27204 bytes --]

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 75688 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-26 16:48                 ` Neil Bradley
@ 2020-08-26 16:53                   ` Vijay Khemka
  2020-08-26 18:43                     ` Neil Bradley
  2020-08-26 21:52                     ` Neil Bradley
  0 siblings, 2 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-26 16:53 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 27742 bytes --]

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 76887 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-26 16:53                   ` Vijay Khemka
@ 2020-08-26 18:43                     ` Neil Bradley
  2020-08-26 21:52                     ` Neil Bradley
  1 sibling, 0 replies; 40+ messages in thread
From: Neil Bradley @ 2020-08-26 18:43 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 28621 bytes --]

We get “unspecified error” for any power control command or power status retrieval. 0xff if we do a raw IPMI command.

Do we need the IPMBbridged service? It looks like any power related command is talking to the ME, which isn’t possible since there isn’t an IPMB service running.

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 80413 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-25 14:25               ` Patrick Williams
@ 2020-08-26 21:38                 ` Neil Bradley
  2020-08-28 12:44                   ` Patrick Williams
  0 siblings, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-26 21:38 UTC (permalink / raw)
  To: Patrick Williams; +Cc: Amithash Prasad, openbmc, Vijay Khemka, Bruce Mitchell

Thank you Patrick! Yes, indeed on Tioga Pass the RGMII (second interface) goes to the i210. Odd that it'd get added but no one uses it? Anyway, appreciate the hint. I'll dig in and see if we can get it initialized in the device tree somehow. The other mystery is why the second interface on the mezzanine card isn't working...

-->Neil

-----Original Message-----
From: openbmc <openbmc-bounces+neil_bradley=phoenix.com@lists.ozlabs.org> On Behalf Of Patrick Williams
Sent: Tuesday, August 25, 2020 7:26 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>
Cc: Amithash Prasad <amithash@fb.com>; openbmc@lists.ozlabs.org; Vijay Khemka <vijaykhemka@fb.com>; Bruce Mitchell <Bruce_Mitchell@phoenix.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hello Neil,

On Tue, Aug 25, 2020 at 01:35:12AM +0000, Neil Bradley wrote:
> Thanks Vijay! Take a look at the attached picture. It occurs to me that the RJ-45 jack on the left side with the purple cable plugged in may not actually be a network connector (we got these machines last Thursday and are just getting acquainted with them), so please forgive my ignorance in case this isn’t what it’s supposed to be. 😉 The working ethernet port is the leftmost SFP+ connector on the mezzanine card (below the sticker), and the rightmost SFP+ connector doesn’t provide a leach. My assumption, which may be incorrect, is the leftmost ethernet jack is RGMII2 and the NICs on the mezzanine is RGMII1 (at least according to the schematic which shows it as an OCP connection).

From
https://www.opencompute.org/documents/facebook-server-intel-motherboard-v40-spec:

10.4.2 Management Network
The motherboard has three options of management network interfaces for the BMC connection. The management network shares the data network’s physical interface. The management connection should be independent from data traffic and OS/driver condition.
 SFP+ shared-NIC from Mezzanine 10G NIC or PCIe NIC, driven by BMC through RMII/NC-SI or I2C. I2C being default  SGMII/KX shared-NIC connected to midplane interface from Intel® I210-AS, driven by BMC through RMII/NCSI  10/100/1000 MDI shared-NIC connected to RJ45 from Intel® I210-AT(co-layout with Intel® I210-AS), driven by BMC through RMII/NCSI 


---

So it seems like that is another RMII/NCSI attached network interface.
We always use the "SPF+ shared-NIC (10G)" though in our lab.  I don't know that we've ever attempted to get the Intel chip working.  There is probably some device tree content, at least, missing for enabling that.

--
Patrick Williams

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-26 16:53                   ` Vijay Khemka
  2020-08-26 18:43                     ` Neil Bradley
@ 2020-08-26 21:52                     ` Neil Bradley
  2020-08-27  5:27                       ` Vijay Khemka
  2020-08-27  5:39                       ` Vijay Khemka
  1 sibling, 2 replies; 40+ messages in thread
From: Neil Bradley @ 2020-08-26 21:52 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 29065 bytes --]

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 79672 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-26 21:52                     ` Neil Bradley
@ 2020-08-27  5:27                       ` Vijay Khemka
  2020-09-09 23:12                         ` Patrick Voelker
  2020-08-27  5:39                       ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-27  5:27 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 29675 bytes --]

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay



From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 81074 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-26 21:52                     ` Neil Bradley
  2020-08-27  5:27                       ` Vijay Khemka
@ 2020-08-27  5:39                       ` Vijay Khemka
  2020-08-27  6:29                         ` Neil Bradley
  1 sibling, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-27  5:39 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 1357 bytes --]

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 7568 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-27  5:39                       ` Vijay Khemka
@ 2020-08-27  6:29                         ` Neil Bradley
  2020-08-27  6:44                           ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-27  6:29 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 1957 bytes --]

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 6094 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-27  6:29                         ` Neil Bradley
@ 2020-08-27  6:44                           ` Vijay Khemka
  2020-08-27 17:19                             ` Neil Bradley
  2020-08-27 21:59                             ` Neil Bradley
  0 siblings, 2 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-27  6:44 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 2339 bytes --]

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 6634 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-27  6:44                           ` Vijay Khemka
@ 2020-08-27 17:19                             ` Neil Bradley
  2020-08-27 18:42                               ` Vijay Khemka
  2020-08-27 21:59                             ` Neil Bradley
  1 sibling, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-27 17:19 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 3102 bytes --]

As an aside, what is the difference between OpenBMC proper and the repo at https://github.com/facebook/openbmc/ ?

Thanks again! We’ll run our tests on the latest today.

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 9524 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-27 17:19                             ` Neil Bradley
@ 2020-08-27 18:42                               ` Vijay Khemka
  0 siblings, 0 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-27 18:42 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 3564 bytes --]

https://github.com/facebook/openbmc/  is separate FB internal repo for FB platforms and it is not a part of LF openbmc.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org> on behalf of Neil Bradley <Neil_Bradley@phoenix.com>
Date: Thursday, August 27, 2020 at 10:22 AM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

As an aside, what is the difference between OpenBMC proper and the repo at https://github.com/facebook/openbmc/ ?

Thanks again! We’ll run our tests on the latest today.

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 10235 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-27  6:44                           ` Vijay Khemka
  2020-08-27 17:19                             ` Neil Bradley
@ 2020-08-27 21:59                             ` Neil Bradley
  2020-08-28 16:25                               ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-27 21:59 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 3262 bytes --]

Thanks Vijay – we tried it with the 2.9-dev release and the GPIO. We’ll go that route.

As an aside, is there a quick/easy way to enable the i210 NCSI connection on the baseboard and the second port on the mezzanine card? I’m new to OpenBMC so sorry if this is a basic/simple Linux question.

Appreciate the help very much!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 8613 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-26 21:38                 ` Neil Bradley
@ 2020-08-28 12:44                   ` Patrick Williams
  0 siblings, 0 replies; 40+ messages in thread
From: Patrick Williams @ 2020-08-28 12:44 UTC (permalink / raw)
  To: Neil Bradley; +Cc: Amithash Prasad, openbmc, Vijay Khemka, Bruce Mitchell

[-- Attachment #1: Type: text/plain, Size: 899 bytes --]

On Wed, Aug 26, 2020 at 09:38:55PM +0000, Neil Bradley wrote:
> Thank you Patrick! Yes, indeed on Tioga Pass the RGMII (second interface) goes to the i210. Odd that it'd get added but no one uses it? 

Facebook designed the Tioga Pass server and we use the NC-SI path on
that server.  Since the server was designed for OCP, we recognize NC-SI
isn't always the preferred path for everyone, so we included the other
interface even thought we don't use it ourselves.

There was work at one point to qualify the second interface, but that
work was done on an old kernel branch and never sent upstream.  Since the
faraday driver had a large rewrite, I would be surprised if this patch
applies cleanly to the 5.x series, but here it is:

https://github.com/facebook/openbmc-linux/commit/eae4bbfbcb8faa091c089f9a367964b0eacd9ba1#diff-157e77f59ffce03237d949d9318dbff4

-- 
Patrick Williams

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 833 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-27 21:59                             ` Neil Bradley
@ 2020-08-28 16:25                               ` Vijay Khemka
  2020-08-28 16:33                                 ` Neil Bradley
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-08-28 16:25 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 3924 bytes --]

Hi Neil,
You can just enable device in device tree and see if it works. Only concern I have here is where to get mac address from, device may get random address and then you can reprogram it from user space. You can check in Intel NCSi specification to understand if get mac address is supported then its support can be added to NCSI driver code.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Thursday, August 27, 2020 at 3:00 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – we tried it with the 2.9-dev release and the GPIO. We’ll go that route.

As an aside, is there a quick/easy way to enable the i210 NCSI connection on the baseboard and the second port on the mezzanine card? I’m new to OpenBMC so sorry if this is a basic/simple Linux question.

Appreciate the help very much!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 9585 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-28 16:25                               ` Vijay Khemka
@ 2020-08-28 16:33                                 ` Neil Bradley
  2020-08-28 20:43                                   ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Neil Bradley @ 2020-08-28 16:33 UTC (permalink / raw)
  To: Vijay Khemka, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 4665 bytes --]

Thanks Vijay! I stumbled across the aspeed-bmc-facebook-tiogapass.dts file yesterday and will be enabling the other interface and giving that a shot today.

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Friday, August 28, 2020 9:25 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
You can just enable device in device tree and see if it works. Only concern I have here is where to get mac address from, device may get random address and then you can reprogram it from user space. You can check in Intel NCSi specification to understand if get mac address is supported then its support can be added to NCSI driver code.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Thursday, August 27, 2020 at 3:00 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – we tried it with the 2.9-dev release and the GPIO. We’ll go that route.

As an aside, is there a quick/easy way to enable the i210 NCSI connection on the baseboard and the second port on the mezzanine card? I’m new to OpenBMC so sorry if this is a basic/simple Linux question.

Appreciate the help very much!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 11226 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-08-28 16:33                                 ` Neil Bradley
@ 2020-08-28 20:43                                   ` Vijay Khemka
  0 siblings, 0 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-08-28 20:43 UTC (permalink / raw)
  To: Neil Bradley, openbmc; +Cc: Amithash Prasad, Sai Dasari

[-- Attachment #1: Type: text/plain, Size: 5006 bytes --]

Cool, Please let us know if you hit any issues.

From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Friday, August 28, 2020 at 9:33 AM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay! I stumbled across the aspeed-bmc-facebook-tiogapass.dts file yesterday and will be enabling the other interface and giving that a shot today.

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Friday, August 28, 2020 9:25 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
You can just enable device in device tree and see if it works. Only concern I have here is where to get mac address from, device may get random address and then you can reprogram it from user space. You can check in Intel NCSi specification to understand if get mac address is supported then its support can be added to NCSI driver code.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Thursday, August 27, 2020 at 3:00 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – we tried it with the 2.9-dev release and the GPIO. We’ll go that route.

As an aside, is there a quick/easy way to enable the i210 NCSI connection on the baseboard and the second port on the mezzanine card? I’m new to OpenBMC so sorry if this is a basic/simple Linux question.

Appreciate the help very much!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 11:45 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

No, you can say that power control for tiogapass was available from kernel 5.7 only.

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 11:29 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Does this mean the official 2.8 release was broken in this regard? We aren’t using the latest & greatest, but it’s worth a shot. Will do tomorrow!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 10:40 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil


[-- Attachment #2: Type: text/html, Size: 11725 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-08-27  5:27                       ` Vijay Khemka
@ 2020-09-09 23:12                         ` Patrick Voelker
  2020-09-10 21:13                           ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Patrick Voelker @ 2020-09-09 23:12 UTC (permalink / raw)
  To: Vijay Khemka, Neil Bradley, openbmc; +Cc: Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 30471 bytes --]

Vijay,

We’re having success with power control once booted into Linux (with the Facebook OpenBMC 2.8 tiogapass build) but can’t seem to control the power when stopped at the u-boot prompt.  Is that something that should be working?

If not, can you help me identify the minimum critical differences besides enabling the GPIOE passthrough?  Power control in u-boot is an important goal for our development.  I tried using all my learnings from wolfpass to get it to work but have not been successful yet.

Regards, Patrick


From: openbmc [mailto:openbmc-bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of Vijay Khemka
Sent: Wednesday, August 26, 2020 10:28 PM
To: Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay



From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 81082 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-09-09 23:12                         ` Patrick Voelker
@ 2020-09-10 21:13                           ` Vijay Khemka
  2020-09-10 21:40                             ` Patrick Voelker
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-09-10 21:13 UTC (permalink / raw)
  To: Patrick Voelker, Neil Bradley, openbmc; +Cc: Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 30908 bytes --]

Hi Patrick,
What exactly you mean by stopped at u-boot prompt, do you mean bmc stopped at u-boot?

Regards
-Vijay

From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Wednesday, September 9, 2020 at 4:13 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Vijay,

We’re having success with power control once booted into Linux (with the Facebook OpenBMC 2.8 tiogapass build) but can’t seem to control the power when stopped at the u-boot prompt.  Is that something that should be working?

If not, can you help me identify the minimum critical differences besides enabling the GPIOE passthrough?  Power control in u-boot is an important goal for our development.  I tried using all my learnings from wolfpass to get it to work but have not been successful yet.

Regards, Patrick


From: openbmc [mailto:openbmc-bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of Vijay Khemka
Sent: Wednesday, August 26, 2020 10:28 PM
To: Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay



From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 83723 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-09-10 21:13                           ` Vijay Khemka
@ 2020-09-10 21:40                             ` Patrick Voelker
  2020-09-11 19:47                               ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Patrick Voelker @ 2020-09-10 21:40 UTC (permalink / raw)
  To: Vijay Khemka, Neil Bradley, openbmc; +Cc: Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 31268 bytes --]

I’m referring to the interactive u-boot prompt that you can access by pressing a key on the BMC serial console to stop the BMC boot to linux.

From: Vijay Khemka [mailto:vijaykhemka@fb.com]
Sent: Thursday, September 10, 2020 2:13 PM
To: Patrick Voelker; Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Patrick,
What exactly you mean by stopped at u-boot prompt, do you mean bmc stopped at u-boot?

Regards
-Vijay

From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Wednesday, September 9, 2020 at 4:13 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Vijay,

We’re having success with power control once booted into Linux (with the Facebook OpenBMC 2.8 tiogapass build) but can’t seem to control the power when stopped at the u-boot prompt.  Is that something that should be working?

If not, can you help me identify the minimum critical differences besides enabling the GPIOE passthrough?  Power control in u-boot is an important goal for our development.  I tried using all my learnings from wolfpass to get it to work but have not been successful yet.

Regards, Patrick


From: openbmc [mailto:openbmc-bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of Vijay Khemka
Sent: Wednesday, August 26, 2020 10:28 PM
To: Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay



From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 83524 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-09-10 21:40                             ` Patrick Voelker
@ 2020-09-11 19:47                               ` Vijay Khemka
  2020-09-11 19:54                                 ` Vijay Khemka
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-09-11 19:47 UTC (permalink / raw)
  To: Patrick Voelker, Neil Bradley, openbmc; +Cc: Amithash Prasad

[-- Attachment #1: Type: text/plain, Size: 31911 bytes --]

So if BMC is not running then you can control power manually by power button or reset button. As far as I know There is no application in uboot which can control power for host, It is done through BMC when it is booted to linux and power control is running.

Please let me know if I am missing anything here.

Regards
-Vijay

From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Thursday, September 10, 2020 at 2:40 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

I’m referring to the interactive u-boot prompt that you can access by pressing a key on the BMC serial console to stop the BMC boot to linux.

From: Vijay Khemka [mailto:vijaykhemka@fb.com]
Sent: Thursday, September 10, 2020 2:13 PM
To: Patrick Voelker; Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Patrick,
What exactly you mean by stopped at u-boot prompt, do you mean bmc stopped at u-boot?

Regards
-Vijay

From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Wednesday, September 9, 2020 at 4:13 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

Vijay,

We’re having success with power control once booted into Linux (with the Facebook OpenBMC 2.8 tiogapass build) but can’t seem to control the power when stopped at the u-boot prompt.  Is that something that should be working?

If not, can you help me identify the minimum critical differences besides enabling the GPIOE passthrough?  Power control in u-boot is an important goal for our development.  I tried using all my learnings from wolfpass to get it to work but have not been successful yet.

Regards, Patrick


From: openbmc [mailto:openbmc-bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of Vijay Khemka
Sent: Wednesday, August 26, 2020 10:28 PM
To: Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8

Neil,
This means that power control application is not running. An I know why because we need kernel device tree from 5.7+ kernel to support proper gpio pins addition in device  tree.

So please download latest openbmc image, your all problem for ipmb and power control will be resolved.

Regards
-Vijay



From: Neil Bradley <Neil_Bradley@phoenix.com>
Date: Wednesday, August 26, 2020 at 2:53 PM
To: Vijay Khemka <vijaykhemka@fb.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>, Sai Dasari <sdasari@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8

A bit more info – does this mean the dbus interface for chassis control is missing?

root@tiogapass:/dev# power-util mb status
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable off
root@tiogapass:/dev# power-util mb on
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
Powering on Server
Failed to set property RequestedPowerTransition on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev# power-util mb off
Failed to get property CurrentPowerState on interface xyz.openbmc_project.State.Chassis: The name is not activatable
root@tiogapass:/dev#

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com>
Sent: Wednesday, August 26, 2020 9:54 AM
To: Neil Bradley <Neil_Bradley@phoenix.com>; openbmc@lists.ozlabs.org
Cc: Amithash Prasad <amithash@fb.com>; Sai Dasari <sdasari@fb.com>
Subject: Re: Tioga Pass OpenBMC 2.8

Host power control should work regardless of kernel version you are using. Please run following command
power-util mb status – it should show current power status
power-util mb on
power-util mb off

Please let me know if it is working

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Wednesday, August 26, 2020 at 9:49 AM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks Vijay – one question, shouldn’t the build I have allow software power control? Ultimately IPMB might not be the issue but I’m puzzled as to why a simple function like power control is nonfunctional. Thoughts?

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Wednesday, August 26, 2020 9:46 AM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Can you please pull latest openbmc image, actually ipmb support for tiogapass is available in 5.7+ kernel only. You are using 5.4 kernel.

Regards
-Vijay

From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Tuesday, August 25, 2020 at 3:23 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>, Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay! Thanks for everything. We did a bit more investigation today on the power control on the Tioga Pass board.

When we do any chassis power control commands to power up or down, regardless of whether or not the host is powered or not, we get 0xff back as a return code.

With host power off or on, the busctl command always returns “Call failed: The name is not activatable”. Which I guess would make sense. With host power on, it appears the IPMB service isn’t available, as when we do a busctl | grep -I channel, we get these two:

xyz.openbmc_project.Ipmi.Channel.eth0                            (provided by netipmid)
xyz.openbmc_project.Ipmi.Channel.ipmi-kcs3                   (provided by kcsbridged)

ipmibridged (the service that provides the Ipmb channel) exists in the BMC but it’s not running.  “systemctl status ipmb” indicates that it stopped with errors a couple of times.

In the system log “journalctl | grep –i ipmb”, I can see these error messages about why it failed to start:
ipmbbridged[316]: ipmbChannelInit: error opening ipmbI2cSlave
ipmbbridged[316]: initializeChannels: channel initialization failed

Looks like this might be the root of the problem. Is there somewhere else we can look to see why the ipmb service isn’t started, and are we looking in the right place?

Thanks again for the assist!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 5:00 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>; Amithash Prasad <amithash@fb.com<mailto:amithash@fb.com>>; Sai Dasari <sdasari@fb.com<mailto:sdasari@fb.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
Where do you connect onboard phy to as I never came across using this, you should be able to enable this as well. In our infrastructure we use only one mezz card with NCSI. If you have another NIC card connected then you can enable this in device tree like one enabled in

arch/arm/boot/dts/aspeed-bmc-facebook-tiogapass.dts



&mac0 {

  status = "okay";



  pinctrl-names = "default";

  pinctrl-0 = <&pinctrl_rmii1_default>;

  clocks = <&syscon ASPEED_CLK_GATE_MAC1CLK>,

     <&syscon ASPEED_CLK_MAC1RCLK>;

  clock-names = "MACCLK", "RCLK";

  use-ncsi;

};



I have never verified this, you may have to debug this through. But you don’t need anything in NCSI patch or code.



Regards



Regarding UEFI, we also use AMI Bios as base and it should have ME binary as well. Please check if host is powered on by running following command from BMC

power-util mb status



If it shows host power is on, then please run below command to get ME details
ME device ID: busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x1 0
Reset ME:     busctl call xyz.openbmc_project.Ipmi.Channel.Ipmb /xyz/openbmc_project/Ipmi/Channel/Ipmb org.openbmc.Ipmb sendRequest yyyyay 1 6 0 0x2 0

Output should look like this
(iyyyyay) 0 7 0 1 0 15 80 1 4 4 2 33 87 1 0 10 11 4 57 48 1  -- for device id
(iyyyyay) 0 47 0 223 0 3 87 1 0  -- for ME reset.


I hope this will help.



-Vijay


From: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 4:34 PM
To: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>, "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hi Vijay!

As you probably know, Tioga Pass has an onboard PHY connected to RGMII2 on the BMC, and also an option for a mezzanine card with two NICs, connected to BMC’s RGMII1. With the Tioga pass build we’ve built, the onboard PHY is inactive, and only the first port on the mezzanine is functional, so Bruce is trying to get the other two interfaces going. Is there a simple configuration that we can change to get the second mezzanine interface and the onboard connection going?

We do get this error message:

[    1.846718] ftgmac100 1e660000.ethernet: Generated random MAC address 62:df:22:c8:1d:b0

Since we’re getting a MAC address for the mezzanine network card, we’re suspecting (guessing, really) that this is the result of the BMC not being able to obtain the MAC address over the PCH/Springville  RGMII interface. Does this make sense or are we missing something?

As an aside, what UEFI image are you using at Facebook? Our current UEFI image is an AMI BIOS – same image that we got with the system.

Appreciate the assistance!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 4:02 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: Re: Tioga Pass OpenBMC 2.8

Thanks for the log, I don’t see any issues. Everything is perfect. Baud rate for initial early kernel console 115200 and then linux boots with 57600. I am not aware about RGMII port, where is this connected to. We are not using that.

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Monday, August 24, 2020 at 2:33 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Subject: RE: Tioga Pass OpenBMC 2.8

Hey Vijay, this is the result of dmesg. Is this sufficient for what you were asking for? We can’t get the full console debug output because there’s a baud rate mismatch until the login prompt comes up, so it’s just garbage up to that point.

Additionally, it looks as if the Tioga Pass image only enables one of the ethernet connections on the mezzanine but not the onboard RGMII at all. Is that expected?

Thank you!

-->Neil

root@tiogapass:~# dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Linux version 5.4.39-30079d6 (oe-user@oe-host) (gcc version 9.3.0 (GCC)) #1 Fri Aug 21 17:14:22 UTC 2020
[    0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
[    0.000000] OF: fdt: Machine model: Facebook TiogaPass BMC
[    0.000000] Memory policy: Data cache writeback
[    0.000000] cma: Reserved 16 MiB at 0x9d800000
[    0.000000] On node 0 totalpages: 126976
[    0.000000]   Normal zone: 992 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 126976 pages, LIFO batch:31
[    0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
[    0.000000] pcpu-alloc: [0] 0
[    0.000000] Built 1 zonelists, mobility grouping on.  Total pages: 125984
[    0.000000] Kernel command line: console=ttyS4,115200n8 root=/dev/ram rw
[    0.000000] Dentry cache hash table entries: 65536 (order: 6, 262144 bytes, linear)
[    0.000000] Inode-cache hash table entries: 32768 (order: 5, 131072 bytes, linear)
[    0.000000] mem auto-init: stack:off, heap alloc:off, heap free:off
[    0.000000] Memory: 476068K/507904K available (6144K kernel code, 398K rwdata, 1456K rodata, 1024K init, 146K bss, 15452K reserved, 16384K cma-reserved)
[    0.000000] random: get_random_u32 called from cache_alloc_refill+0x428/0x964 with crng_init=0
[    0.000000] ftrace: allocating 21599 entries in 43 pages
[    0.000000] NR_IRQS: 16, nr_irqs: 16, preallocated irqs: 16
[    0.000000] i2c controller registered, irq 17
[    0.000000] clocksource: FTTMR010-TIMER2: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 77222644334 ns
[    0.000020] sched_clock: 32 bits at 24MHz, resolution 40ns, wraps every 86767015915ns
[    0.000091] Switching to timer-based delay loop, resolution 40ns
[    0.001397] Calibrating delay loop (skipped), value calculated using timer frequency.. 49.50 BogoMIPS (lpj=247500)
[    0.001454] pid_max: default: 32768 minimum: 301
[    0.002205] Mount-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.002254] Mountpoint-cache hash table entries: 1024 (order: 0, 4096 bytes, linear)
[    0.004248] CPU: Testing write buffer coherency: ok
[    0.006176] Setting up static identity map for 0x80100000 - 0x80100038
[    0.009638] devtmpfs: initialized
[    0.029450] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.029519] futex hash table entries: 256 (order: -1, 3072 bytes, linear)
[    0.032921] pinctrl core: initialized pinctrl subsystem
[    0.034050] NET: Registered protocol family 16
[    0.036390] DMA: preallocated 256 KiB pool for atomic coherent allocations
[    0.038876] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
[    0.038911] hw-breakpoint: maximum watchpoint size is 4 bytes.
[    0.084182] videodev: Linux video capture interface: v2.00
[    0.087772] clocksource: Switched to clocksource FTTMR010-TIMER2
[    0.150999] NET: Registered protocol family 2
[    0.152301] tcp_listen_portaddr_hash hash table entries: 512 (order: 0, 4096 bytes, linear)
[    0.152399] TCP established hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152491] TCP bind hash table entries: 4096 (order: 2, 16384 bytes, linear)
[    0.152591] TCP: Hash tables configured (established 4096 bind 4096)
[    0.152841] UDP hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.152915] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes, linear)
[    0.153488] NET: Registered protocol family 1
[    0.154441] Unpacking initramfs...
[    1.102186] Freeing initrd memory: 1088K
[    1.105176] workingset: timestamp_bits=30 max_order=17 bucket_order=0
[    1.105749] squashfs: version 4.0 (2009/01/31) Phillip Lougher
[    1.105789] jffs2: version 2.2. (SUMMARY)  © 2001-2006 Red Hat, Inc.
[    1.106560] NET: Registered protocol family 38
[    1.109833] aspeed-kcs-bmc: channel=2 addr=0xca8 idr=0x28 odr=0x34 str=0x40
[    1.110704] aspeed-kcs-bmc: channel=3 addr=0xca2 idr=0x2c odr=0x38 str=0x44
[    1.117654] Serial: 8250/16550 driver, 6 ports, IRQ sharing enabled
[    1.122199] 1e787000.serial: ttyS5 at MMIO 0x1e787000 (irq = 31, base_baud = 1546875) is a 16550A
[    1.124410] 1e783000.serial: ttyS0 at MMIO 0x1e783000 (irq = 29, base_baud = 1500000) is a 16550A
[    1.125923] 1e784000.serial: ttyS4 at MMIO 0x1e784000 (irq = 30, base_baud = 1500000) is a 16550A
[    1.500174] printk: console [ttyS4] enabled
[    1.506089] 1e78d000.serial: ttyS1 at MMIO 0x1e78d000 (irq = 32, base_baud = 1500000) is a 16550A
[    1.516573] 1e78e000.serial: ttyS2 at MMIO 0x1e78e000 (irq = 33, base_baud = 1500000) is a 16550A
[    1.527484] timeriomem_rng 1e6e2078.hwrng: 32bits from 0x(ptrval) @ 1us
[    1.536213] random: fast init done
[    1.548499] random: crng init done
[    1.567470] loop: module loaded
[    1.613884] aspeed-smc 1e620000.spi: Using 50 MHz SPI frequency
[    1.620094] aspeed-smc 1e620000.spi: w25q256 (32768 Kbytes)
[    1.625704] aspeed-smc 1e620000.spi: CE0 window [ 0x20000000 - 0x22000000 ] 32MB
[    1.633228] aspeed-smc 1e620000.spi: CE1 window [ 0x22000000 - 0x2a000000 ] 128MB
[    1.640809] aspeed-smc 1e620000.spi: read control register: 203b0641
[    1.785843] 5 fixed-partitions partitions found on MTD device 1e620000.spi
[    1.792857] Creating 5 MTD partitions on "1e620000.spi":
[    1.798309] 0x000000000000-0x000000060000 : "u-boot"
[    1.805858] 0x000000060000-0x000000080000 : "u-boot-env"
[    1.813936] 0x000000080000-0x0000004c0000 : "kernel"
[    1.821716] 0x0000004c0000-0x000001c00000 : "rofs"
[    1.829241] 0x000001c00000-0x000002000000 : "rwfs"
[    1.838478] aspeed-smc 1e630000.spi: Using 50 MHz SPI frequency
[    1.844533] aspeed-smc 1e630000.spi: unrecognized JEDEC id bytes: 00 00 00 00 00 00
[    1.852321] aspeed-smc 1e630000.spi: Aspeed SMC probe failed -2
[    1.862969] aspeed-smc: probe of 1e630000.spi failed with error -2
[    1.871399] libphy: Fixed MDIO Bus: probed
[    1.876920] ftgmac100 1e660000.ethernet: Read MAC address 1c:34:da:7b:40:e7 from chip
[    1.884902] ftgmac100 1e660000.ethernet: Using NCSI interface
[    1.892025] ftgmac100 1e660000.ethernet eth0: irq 20, mapped at 34f8ec25
[    1.899832] udc-core: couldn't find an available UDC - added [g_mass_storage] to list of pending drivers
[    1.910003] i2c /dev entries driver
[    1.914828] aspeed-i2c-bus 1e78a040.i2c-bus: i2c bus 0 registered, irq 34
[    1.923341] aspeed-i2c-bus 1e78a080.i2c-bus: i2c bus 1 registered, irq 35
[    1.931604] aspeed-i2c-bus 1e78a0c0.i2c-bus: i2c bus 2 registered, irq 36
[    1.939867] aspeed-i2c-bus 1e78a100.i2c-bus: i2c bus 3 registered, irq 37
[    1.948049] aspeed-i2c-bus 1e78a140.i2c-bus: i2c bus 4 registered, irq 38
[    1.959801] aspeed-i2c-bus 1e78a180.i2c-bus: i2c bus 5 registered, irq 39
[    2.725060] tpm_i2c_infineon 6-0020: could not request locality
[    2.733572] at24 6-0054: 8192 byte 24c64 EEPROM, writable, 32 bytes/write
[    2.740642] aspeed-i2c-bus 1e78a1c0.i2c-bus: i2c bus 6 registered, irq 40
[    2.748857] aspeed-i2c-bus 1e78a300.i2c-bus: i2c bus 7 registered, irq 41
[    2.757424] aspeed-i2c-bus 1e78a340.i2c-bus: i2c bus 8 registered, irq 42
[    2.765666] aspeed-i2c-bus 1e78a380.i2c-bus: i2c bus 9 registered, irq 43
[    2.773566] pca954x 1-0071: probe failed
[    2.778295] Driver for 1-wire Dallas network protocol.
[    2.801682] NET: Registered protocol family 10
[    2.808778] Segment Routing with IPv6
[    2.813220] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    2.820755] NET: Registered protocol family 17
[    2.825277] 8021q: 802.1Q VLAN Support v1.8
[    2.841568] printk: console [netcon0] enabled
[    2.845947] netconsole: network logging started
[    2.850919] hctosys: unable to open rtc device (rtc0)
[    2.862757] Freeing unused kernel memory: 1024K
[    2.870834] Checked W+X mappings: passed, no W+X pages found
[    2.876521] Run /init as init process
[    3.715741] jffs2: notice: (77) jffs2_build_xattr_subsystem: complete building xattr subsystem, 42 of xdatum (42 unchecked, 0 orphan) and 108 of xref (0 dead, 0 orphan) found.
[    3.782648] overlayfs: upper fs does not support tmpfile.
[    6.284828] systemd[1]: Failed to lookup module alias 'autofs4': Function not implemented
[    6.373603] systemd[1]: systemd 244.3+ running in system mode. (+PAM -AUDIT -SELINUX -IMA -APPARMOR -SMACK +SYSVINIT -UTMP -LIBCRYPTSETUP -GCRYPT -GNUTLS -ACL +XZ -LZ4 -SECCOMP +BLKID -ELFUTILS +KMOD -IDN2 -IDN -PCRE2 default-hierarchy=hybrid)
[    6.397338] systemd[1]: Detected architecture arm.
[    6.558718] systemd[1]: Set hostname to <tiogapass>.
[    8.522924] systemd[1]: /lib/systemd/system/phosphor-pid-control.service:7: Neither a valid executable name nor an absolute path: {bindir}/swampd
[    8.536320] systemd[1]: phosphor-pid-control.service: Unit configuration has fatal error, unit will not be started.
[    9.018415] systemd[1]: /lib/systemd/system/phosphor-ipmi-net@.socket:3<mailto:/lib/systemd/system/phosphor-ipmi-net@.socket:3>: Invalid interface name, ignoring: sys-subsystem-net-devices-%i.device
[    9.124171] systemd[1]: phosphor-pid-control.service: Cannot add dependency job, ignoring: Unit phosphor-pid-control.service has a bad unit file setting.
[    9.193233] systemd[1]: Created slice system-getty.slice.
[    9.231544] systemd[1]: Created slice system-mapper\x2dwait.slice.
[    9.271483] systemd[1]: Created slice system-obmc\x2dconsole.slice.
[    9.311612] systemd[1]: Created slice system-obmc\x2dled\x2dgroup\x2dstart.slice.
[    9.351503] systemd[1]: Created slice system-phosphor\x2dcertificate\x2dmanager.slice.
[    9.391513] systemd[1]: Created slice system-phosphor\x2ddiscover\x2dsystem\x2dstate.slice.
[    9.431363] systemd[1]: Created slice system-phosphor\x2dipmi\x2dkcs.slice.
[    9.471394] systemd[1]: Created slice system-phosphor\x2dipmi\x2dnet.slice.
[    9.511412] systemd[1]: Created slice system-serial\x2dgetty.slice.
[    9.549839] systemd[1]: Started Dispatch Password Requests to Console Directory Watch.
[    9.589722] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    9.630579] systemd[1]: Reached target Paths.
[    9.668965] systemd[1]: Reached target Remote File Systems.
[    9.708737] systemd[1]: Reached target Slices.
[    9.749062] systemd[1]: Reached target Swap.
[    9.792899] systemd[1]: Listening on Syslog Socket.
[    9.860252] systemd[1]: Listening on Process Core Dump Socket.
[    9.899969] systemd[1]: Listening on initctl Compatibility Named Pipe.
[    9.940519] systemd[1]: Condition check resulted in Journal Audit Socket being skipped.
[    9.951840] systemd[1]: Listening on Journal Socket (/dev/log).
[    9.991612] systemd[1]: Listening on Journal Socket.
[   10.032322] systemd[1]: Listening on Network Service Netlink Socket.
[   10.071319] systemd[1]: Listening on udev Control Socket.
[   10.110578] systemd[1]: Listening on udev Kernel Socket.
[   10.130583] systemd[1]: Condition check resulted in Huge Pages File System being skipped.
[   10.141345] systemd[1]: Condition check resulted in POSIX Message Queue File System being skipped.
[   10.162566] systemd[1]: Mounting Kernel Debug File System...
[   10.229312] systemd[1]: Mounting Temporary Directory (/tmp)...
[   10.245137] systemd[1]: Condition check resulted in Create list of static device nodes for the current kernel being skipped.
[   10.294121] systemd[1]: Condition check resulted in File System Check on Root Device being skipped.
[   10.321301] systemd[1]: Starting Journal Service...
[   10.406066] systemd[1]: Condition check resulted in Load Kernel Modules being skipped.
[   10.418296] systemd[1]: Condition check resulted in FUSE Control File System being skipped.
[   10.453150] systemd[1]: Mounting Kernel Configuration File System...
[   10.510636] systemd[1]: Starting Remount Root and Kernel File Systems...
[   10.619338] systemd[1]: Starting Apply Kernel Variables...
[   10.720466] systemd[1]: Starting udev Coldplug all Devices...
[   11.064565] systemd[1]: Started Hardware RNG Entropy Gatherer Daemon.
[   11.235291] systemd[1]: Mounted Kernel Debug File System.
[   11.315996] systemd[1]: Mounted Temporary Directory (/tmp).
[   11.370581] systemd[1]: Mounted Kernel Configuration File System.
[   11.433764] systemd[1]: Started Remount Root and Kernel File Systems.
[   11.518510] systemd[1]: Started Apply Kernel Variables.
[   11.558596] systemd[1]: Condition check resulted in Rebuild Hardware Database being skipped.
[   11.610138] systemd[1]: Condition check resulted in Create System Users being skipped.
[   11.650212] systemd[1]: Starting Create Static Device Nodes in /dev...
[   11.681287] systemd[1]: Started Journal Service.
[   12.032812] systemd-journald[99]: Received client request to flush runtime journal.
[   18.520429] 8021q: adding VLAN 0 to HW filter on device eth0
[   18.526634] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19
[   34.249707] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.258113] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.266384] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.274795] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.283212] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.291570] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.300049] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.308421] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.343667] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.358280] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.366552] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.374965] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.383368] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.391828] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   34.400206] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b
[   51.960204] i2c i2c-4: new_device: Instantiated device ipmb-dev at 0x10
[   53.512805] i2c i2c-4: Failed to register i2c client ipmb-dev at 0x10 (-16)


From: Neil Bradley
Sent: Monday, August 24, 2020 2:18 PM
To: openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Subject: RE: Tioga Pass OpenBMC 2.8

Thanks for the reply, Vijay! The part was programed directly with a dediprog. Hey Bruce, can you post the full boot log of that TP system? Vijay, Bruce (in another thread) did see this on the boot log, but I’ll work to get you the full boot console:


[    1.854939] ftgmac100 1e660000.ethernet: Using NCSI interface

[   17.831827] ftgmac100 1e660000.ethernet eth0: NCSI: Handler for packet type 0x82 returned -19

[   33.526950] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.535564] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.543813] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.554345] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.562598] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.570958] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.612936] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.629318] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.638683] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.647978] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.657291] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.665690] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.673956] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.682381] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

[   33.690792] ftgmac100 1e660000.ethernet eth0: NCSI: 'bad' packet ignored for type 0x8b

This is the image we used:

-rw-r--r-- 2 bruce bruce 33554432 Aug 21 17:50 obmc-phosphor-image-tiogapass-20200821170132.static.mtd

We are also wondering about power control. We get an IPMB error whenever we do any power control via IPMI instructions (power button works however) and it looks like it’s talking to the ME. Are you familiar with the ME’s role in power control/status? I couldn’t find any mention of the ME in the Tioga Pass spec, so perhaps I missed it. I’m suspecting that our UEFI image doesn’t have the right ME image to mate up with Tioga Pass.

Thank you!

-->Neil

From: Vijay Khemka <vijaykhemka@fb.com<mailto:vijaykhemka@fb.com>>
Sent: Monday, August 24, 2020 2:05 PM
To: Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>; openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>; Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Re: Tioga Pass OpenBMC 2.8

Hi Neil,
How did you copy image to target, I mean what file name you have copied and what was size.
There is no issue in available image for tiogapass build and it should work seamlessly.

Please share console boot log if you have, do you know what network card you system have.

Regards
-Vijay

From: openbmc <openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org<mailto:openbmc-bounces+vijaykhemka=fb.com@lists.ozlabs.org>> on behalf of Neil Bradley <Neil_Bradley@phoenix.com<mailto:Neil_Bradley@phoenix.com>>
Date: Friday, August 21, 2020 at 10:40 PM
To: "openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>" <openbmc@lists.ozlabs.org<mailto:openbmc@lists.ozlabs.org>>
Cc: Patrick Voelker <Patrick_Voelker@phoenix.com<mailto:Patrick_Voelker@phoenix.com>>, Bruce Mitchell <Bruce_Mitchell@phoenix.com<mailto:Bruce_Mitchell@phoenix.com>>
Subject: Tioga Pass OpenBMC 2.8

Greetings fellow BMC folk – I’m Neil Bradley, Phoenix Technologies’ BMC architect and have a question for the group.

We are attempting to load OpenBMC 2.8 on a Tioga Pass. This is what we did:

mkdir tiogapass
cd tiogapass
git clone https://github.com/openbmc/openbmc.git
cd openbmc
git checkout 35a774200999ac2fca48693c1c169bf99d2f63ea
export TEMPLATECONF=meta-facebook/meta-tiogapass/conf
source openbmc-env
bitbake obmc-phosphor-image

The BMC does boot fully, however there are two major problems we’ve encountered:


  1.  Host will not power on via the front panel power button
  2.  BMC Will not obtain or try DHCP even though its set enabled for the primary LAN channel

I have a couple of questions:


  1.  Was this built/done correctly? In other words, did I do something wrong in this process?
  2.  Does anyone know the current state of Tioga Pass support in 2.8 or the general health therein?

We’re coming up to speed on this quickly, so apologies if these are stupid questions. And greetings!

-->Neil

[-- Attachment #2: Type: text/html, Size: 86221 bytes --]

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-09-11 19:47                               ` Vijay Khemka
@ 2020-09-11 19:54                                 ` Vijay Khemka
  2020-09-14 17:30                                   ` Patrick Voelker
  0 siblings, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-09-11 19:54 UTC (permalink / raw)
  To: Patrick Voelker, Neil Bradley, openbmc; +Cc: Amithash Prasad

So if BMC is not running then you can control power manually by power button or reset button. As far as I know There is no application in uboot which can control power for host, It is done through BMC when it is booted to linux and power control is running.
 
Please let me know if I am missing anything here.
 
Regards
-Vijay
 
From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Thursday, September 10, 2020 at 2:40 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8
 
I’m referring to the interactive u-boot prompt that you can access by pressing a key on the BMC serial console to stop the BMC boot to linux.
 
From: Vijay Khemka [mailto:vijaykhemka@fb.com] 
Sent: Thursday, September 10, 2020 2:13 PM
To: Patrick Voelker; Neil Bradley; openbmc@lists.ozlabs.org
Cc: Amithash Prasad
Subject: Re: Tioga Pass OpenBMC 2.8
 
Hi Patrick,
What exactly you mean by stopped at u-boot prompt, do you mean bmc stopped at u-boot?
 
Regards
-Vijay
 
From: Patrick Voelker <Patrick_Voelker@phoenix.com>
Date: Wednesday, September 9, 2020 at 4:13 PM
To: Vijay Khemka <vijaykhemka@fb.com>, Neil Bradley <Neil_Bradley@phoenix.com>, "openbmc@lists.ozlabs.org" <openbmc@lists.ozlabs.org>
Cc: Amithash Prasad <amithash@fb.com>
Subject: RE: Tioga Pass OpenBMC 2.8
 
Vijay,
 
We’re having success with power control once booted into Linux (with the Facebook OpenBMC 2.8 tiogapass build) but can’t seem to control the power when stopped at the u-boot prompt.  Is that something that should be working?
 
If not, can you help me identify the minimum critical differences besides enabling the GPIOE passthrough?  Power control in u-boot is an important goal for our development.  I tried using all my learnings from wolfpass to get it to work but have not been successful yet.
 
Regards, Patrick


^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-09-11 19:54                                 ` Vijay Khemka
@ 2020-09-14 17:30                                   ` Patrick Voelker
  2020-09-15 17:19                                     ` Patrick Voelker
  2020-09-15 19:07                                     ` Vijay Khemka
  0 siblings, 2 replies; 40+ messages in thread
From: Patrick Voelker @ 2020-09-14 17:30 UTC (permalink / raw)
  To: Vijay Khemka, Neil Bradley, openbmc; +Cc: Amithash Prasad

What would an application need to do to control power for the host?  Power control while in u-boot _should_ be as simple as enabling the power button passthrough.

> -----Original Message-----
> From: Vijay Khemka [mailto:vijaykhemka@fb.com]
> Subject: Re: Tioga Pass OpenBMC 2.8
> 
> So if BMC is not running then you can control power manually by power
> button or reset button. As far as I know There is no application in uboot
> which can control power for host, It is done through BMC when it is booted
> to linux and power control is running.
> 
> Please let me know if I am missing anything here.
> 
> Regards
> -Vijay

^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-09-14 17:30                                   ` Patrick Voelker
@ 2020-09-15 17:19                                     ` Patrick Voelker
  2020-09-15 19:12                                       ` Vijay Khemka
  2020-09-15 19:07                                     ` Vijay Khemka
  1 sibling, 1 reply; 40+ messages in thread
From: Patrick Voelker @ 2020-09-15 17:19 UTC (permalink / raw)
  To: Patrick Voelker, Vijay Khemka, Neil Bradley, openbmc; +Cc: Amithash Prasad

The main issue I'm having is understanding the bare minimum requirements for BMC involvement in Tioga Pass power on.  I've found the recipe fb-powerctrl which sets a couple of GPIOs but beyond that I can't figure out how the BMC is enabling power on.  I'm missing something.

Is this something you can help with?

> -----Original Message-----
> From: openbmc [mailto:openbmc-
> bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of
> Patrick Voelker
> Sent: Monday, September 14, 2020 10:30 AM
> Subject: RE: Tioga Pass OpenBMC 2.8
> 
> What would an application need to do to control power for the host?  Power
> control while in u-boot _should_ be as simple as enabling the power button
> passthrough.
> 
> > -----Original Message-----
> > From: Vijay Khemka [mailto:vijaykhemka@fb.com]
> > Subject: Re: Tioga Pass OpenBMC 2.8
> >
> > So if BMC is not running then you can control power manually by power
> > button or reset button. As far as I know There is no application in uboot
> > which can control power for host, It is done through BMC when it is booted
> > to linux and power control is running.
> >
> > Please let me know if I am missing anything here.
> >
> > Regards
> > -Vijay

^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-09-14 17:30                                   ` Patrick Voelker
  2020-09-15 17:19                                     ` Patrick Voelker
@ 2020-09-15 19:07                                     ` Vijay Khemka
  2020-09-15 19:42                                       ` Patrick Voelker
  1 sibling, 1 reply; 40+ messages in thread
From: Vijay Khemka @ 2020-09-15 19:07 UTC (permalink / raw)
  To: Patrick Voelker, Neil Bradley, openbmc; +Cc: Amithash Prasad

Honestly, I am not understanding how you want to control host power when bmc booted to u-boot only. 
Please explain what is your requirement, do you want to change u-boot code?

Regards
-Vijay

On 9/14/20, 10:30 AM, "Patrick Voelker" <Patrick_Voelker@phoenix.com> wrote:

    What would an application need to do to control power for the host?  Power control while in u-boot _should_ be as simple as enabling the power button passthrough.

    > -----Original Message-----
    > From: Vijay Khemka [mailto:vijaykhemka@fb.com]
    > Subject: Re: Tioga Pass OpenBMC 2.8
    > 
    > So if BMC is not running then you can control power manually by power
    > button or reset button. As far as I know There is no application in uboot
    > which can control power for host, It is done through BMC when it is booted
    > to linux and power control is running.
    > 
    > Please let me know if I am missing anything here.
    > 
    > Regards
    > -Vijay


^ permalink raw reply	[flat|nested] 40+ messages in thread
* Re: Tioga Pass OpenBMC 2.8
  2020-09-15 17:19                                     ` Patrick Voelker
@ 2020-09-15 19:12                                       ` Vijay Khemka
  0 siblings, 0 replies; 40+ messages in thread
From: Vijay Khemka @ 2020-09-15 19:12 UTC (permalink / raw)
  To: Patrick Voelker, Neil Bradley, openbmc; +Cc: Amithash Prasad

There is BMC_READY gpio which needs to be set by BMC before it can communicate with host, until this gpio is not set Host bios will keep waiting and will not boot completely.

On 9/15/20, 10:19 AM, "Patrick Voelker" <Patrick_Voelker@phoenix.com> wrote:

    The main issue I'm having is understanding the bare minimum requirements for BMC involvement in Tioga Pass power on.  I've found the recipe fb-powerctrl which sets a couple of GPIOs but beyond that I can't figure out how the BMC is enabling power on.  I'm missing something.

    Is this something you can help with?

    > -----Original Message-----
    > From: openbmc [mailto:openbmc-
    > bounces+patrick_voelker=phoenix.com@lists.ozlabs.org] On Behalf Of
    > Patrick Voelker
    > Sent: Monday, September 14, 2020 10:30 AM
    > Subject: RE: Tioga Pass OpenBMC 2.8
    > 
    > What would an application need to do to control power for the host?  Power
    > control while in u-boot _should_ be as simple as enabling the power button
    > passthrough.
    > 
    > > -----Original Message-----
    > > From: Vijay Khemka [mailto:vijaykhemka@fb.com]
    > > Subject: Re: Tioga Pass OpenBMC 2.8
    > >
    > > So if BMC is not running then you can control power manually by power
    > > button or reset button. As far as I know There is no application in uboot
    > > which can control power for host, It is done through BMC when it is booted
    > > to linux and power control is running.
    > >
    > > Please let me know if I am missing anything here.
    > >
    > > Regards
    > > -Vijay


^ permalink raw reply	[flat|nested] 40+ messages in thread
* RE: Tioga Pass OpenBMC 2.8
  2020-09-15 19:07                                     ` Vijay Khemka
@ 2020-09-15 19:42                                       ` Patrick Voelker
  0 siblings, 0 replies; 40+ messages in thread
From: Patrick Voelker @ 2020-09-15 19:42 UTC (permalink / raw)
  To: Vijay Khemka, Neil Bradley, openbmc; +Cc: Amithash Prasad

Yes, when in u-boot, having prevented the boot to Linux or having a corrupt Linux image, I'd like to be able to press the power button and boot the host.

I'm prepared to modify u-boot code to make the necessary adjustments if GPIO settings are all that is required.  If it requires more than GPIO settings, I'd love to know what the basic requirements are.

> -----Original Message-----
> From: Vijay Khemka [mailto:vijaykhemka@fb.com]
> Sent: Tuesday, September 15, 2020 12:07 PM
> To: Patrick Voelker; Neil Bradley; openbmc@lists.ozlabs.org
> Cc: Amithash Prasad
> Subject: Re: Tioga Pass OpenBMC 2.8
> 
> Honestly, I am not understanding how you want to control host power when
> bmc booted to u-boot only.
> Please explain what is your requirement, do you want to change u-boot code?
> 
> Regards
> -Vijay
> 
> On 9/14/20, 10:30 AM, "Patrick Voelker" <Patrick_Voelker@phoenix.com>
> wrote:
> 
>     What would an application need to do to control power for the host?
> Power control while in u-boot _should_ be as simple as enabling the power
> button passthrough.
> 
>     > -----Original Message-----
>     > From: Vijay Khemka [mailto:vijaykhemka@fb.com]
>     > Subject: Re: Tioga Pass OpenBMC 2.8
>     >
>     > So if BMC is not running then you can control power manually by power
>     > button or reset button. As far as I know There is no application in uboot
>     > which can control power for host, It is done through BMC when it is
> booted
>     > to linux and power control is running.
>     >
>     > Please let me know if I am missing anything here.
>     >
>     > Regards
>     > -Vijay


^ permalink raw reply	[flat|nested] 40+ messages in thread
end of thread, other threads:[~2020-09-15 19:42 UTC | newest]

Thread overview: 40+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-08-22  1:36 Tioga Pass OpenBMC 2.8 Neil Bradley
2020-08-24 17:07 ` Bruce Mitchell
2020-08-24 22:52   ` Vijay Khemka
2020-08-24 21:05 ` Vijay Khemka
2020-08-24 21:17   ` Neil Bradley
2020-08-24 21:32     ` Neil Bradley
2020-08-24 23:02       ` Vijay Khemka
2020-08-24 23:34         ` Neil Bradley
2020-08-25  0:00           ` Vijay Khemka
2020-08-25  1:35             ` Neil Bradley
2020-08-25 14:25               ` Patrick Williams
2020-08-26 21:38                 ` Neil Bradley
2020-08-28 12:44                   ` Patrick Williams
2020-08-25 22:22             ` Neil Bradley
2020-08-26 16:46               ` Vijay Khemka
2020-08-26 16:48                 ` Neil Bradley
2020-08-26 16:53                   ` Vijay Khemka
2020-08-26 18:43                     ` Neil Bradley
2020-08-26 21:52                     ` Neil Bradley
2020-08-27  5:27                       ` Vijay Khemka
2020-09-09 23:12                         ` Patrick Voelker
2020-09-10 21:13                           ` Vijay Khemka
2020-09-10 21:40                             ` Patrick Voelker
2020-09-11 19:47                               ` Vijay Khemka
2020-09-11 19:54                                 ` Vijay Khemka
2020-09-14 17:30                                   ` Patrick Voelker
2020-09-15 17:19                                     ` Patrick Voelker
2020-09-15 19:12                                       ` Vijay Khemka
2020-09-15 19:07                                     ` Vijay Khemka
2020-09-15 19:42                                       ` Patrick Voelker
2020-08-27  5:39                       ` Vijay Khemka
2020-08-27  6:29                         ` Neil Bradley
2020-08-27  6:44                           ` Vijay Khemka
2020-08-27 17:19                             ` Neil Bradley
2020-08-27 18:42                               ` Vijay Khemka
2020-08-27 21:59                             ` Neil Bradley
2020-08-28 16:25                               ` Vijay Khemka
2020-08-28 16:33                                 ` Neil Bradley
2020-08-28 20:43                                   ` Vijay Khemka
2020-08-24 22:58     ` Vijay Khemka

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.