New fru bus interfaces

[Andrei Kartashev]

Hi Ed,

I don't think this commit related to this issue. On our system we have
similar /xyz/openbmc_project/FruDevice/X_YY entries, but I didn't
bother about. My work was about real FRUs with broken content.
This entries are created by FruDevice for addresses is can't detect as
EEPROM:

# busctl tree xyz.openbmc_project.FruDevice
└─/xyz
  └─/xyz/openbmc_project
    └─/xyz/openbmc_project/FruDevice
      ├─/xyz/openbmc_project/FruDevice/12_81
      ├─/xyz/openbmc_project/FruDevice/12_82
      ├─/xyz/openbmc_project/FruDevice/12_83
      ├─/xyz/openbmc_project/FruDevice/12_84
[...]
# i2cdetect -y 12
     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- -- -- -- -- -- 
10: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
40: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
50: -- 51 52 53 54 -- -- -- -- -- -- -- -- -- -- -- 
60: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 
70: -- -- -- -- -- -- -- --                         

In this case, there are pmbus VRs on that addresses.
Actually, now I see, it creates such nodes for some EEPROM nodes too:

# busctl tree xyz.openbmc_project.FruDevice
[...]
      ├─/xyz/openbmc_project/FruDevice/7_80
      ├─/xyz/openbmc_project/FruDevice/7_82
      ├─/xyz/openbmc_project/FruDevice/7_88
      ├─/xyz/openbmc_project/FruDevice/7_90
      ├─/xyz/openbmc_project/FruDevice/S_1100ADU00_201
[...]

# busctl introspect xyz.openbmc_project.FruDevice
/xyz/openbmc_project/FruDevice/S_1100ADU00_201
NAME                                TYPE      SIGNATURE RESULT/VALUE      FLAGS
[...]
xyz.openbmc_project.FruDevice       interface -         -                 -
.ADDRESS                            property  u         82                emits-change
.BUS                                property  u         7                 emits-change
[...]

# busctl introspect xyz.openbmc_project.FruDevice /xyz/openbmc_project/FruDevice/7_82           
NAME                                         TYPE      SIGNATURE RESULT/VALUE FLAGS
org.freedesktop.DBus.Introspectable          interface -         -            -
.Introspect                                  method    -         s            -
org.freedesktop.DBus.Peer                    interface -         -            -
.GetMachineId                                method    -         s            -
.Ping                                        method    -         -            -
org.freedesktop.DBus.Properties              interface -         -            -
.Get                                         method    ss        v            -
.GetAll                                      method    s         a{sv}        -
.Set                                         method    ssv       -            -
.PropertiesChanged                           signal    sa{sv}as  -            -
xyz.openbmc_project.Inventory.Item.I2CDevice interface -         -            -
.Address                                     property  u         82           emits-change
.Bus                                         property  u         7            emits-change

But I didn't investigate this topic before.


On Wed, 2020-10-28 at 15:26 -0700, Ed Tanous wrote:
> On Wed, Oct 28, 2020 at 2:43 PM Vijay Khemka <vijaykhemka at fb.com>
> wrote:
> > Team,
> > 
> > I am seeing now new interfaces added to dbus for FRU devices which
> > has no fru device. I am using EM/FruDevice application and it scans
> > dynamically all i2c busses for Fru devices. And it used to list
> > only devices with Fru data only to dbus but current image adds all
> > i2c bus scanned as dbus interfaces. Please see below
> > 
> > 
> > 
> > root at tiogapass:~# busctl tree xyz.openbmc_project.FruDevice
> > 
> > └─/xyz
> > 
> >   └─/xyz/openbmc_project
> > 
> >     └─/xyz/openbmc_project/FruDevice
> > 
> >       ├─/xyz/openbmc_project/FruDevice/0_80
> > 
> >       ├─/xyz/openbmc_project/FruDevice/0_81
> > 
> >       ├─/xyz/openbmc_project/FruDevice/2_109
> > 
> >       ├─/xyz/openbmc_project/FruDevice/2_112
> > 
> >       ├─/xyz/openbmc_project/FruDevice/2_16
> > 
> >       ├─/xyz/openbmc_project/FruDevice/2_22
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_104
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_108
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_113
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_36
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_68
> > 
> >       ├─/xyz/openbmc_project/FruDevice/3_81
> > 
> >       ├─/xyz/openbmc_project/FruDevice/4_112
> > 
> >       ├─/xyz/openbmc_project/FruDevice/4_22
> > 
> >       ├─/xyz/openbmc_project/FruDevice/6_33
> > 
> >       ├─/xyz/openbmc_project/FruDevice/6_80
> > 
> >       ├─/xyz/openbmc_project/FruDevice/7_69
> > 
> >       ├─/xyz/openbmc_project/FruDevice/9_39
> > 
> >       ├─/xyz/openbmc_project/FruDevice/9_84
> > 
> >       ├─/xyz/openbmc_project/FruDevice/Ava_M_2_SSD_Adapter
> > 
> >       └─/xyz/openbmc_project/FruDevice/Tioga_Pass_Single_Side
> 
> Is that related to this commit?
> 
> https://github.com/openbmc/entity-manager/commit/d7b6659944c3527622d454526509a0a8697ada0f
> 
> > 
> > 
> > In above tree there are only last 2 interface represent real Fru
> > device, rest others have no Fru data. I am not sure why this was
> > added.
> > 
> > 
> > 
> > Please help me understand this feature if anything new added to the
> > application.
> > 
> > 
> > 
> > Regards
> > 
> > -Vijay
-- 
Best regards,
Andrei Kartashev