use "ipmitool fru list",Missing complete FRUID0 device information in the echo data

root@test-evb:~# ipmitool fru list
FRU Device Description : Builtin FRU Device (ID 0)
FRU Device Description : motherboard (ID 1)
Device not present (Requested sensor, data, or record not found)

root@test-evb:# ipmitool fru list 0
Chassis Type : Rack Mount Chassis
Board Mfg Date : Sun Feb 18 02:35:00 2024 UTC
Board Mfg : Biggest International Corp.
Board Product : Some Cool Product
Product Manufacturer : China OpenBmc
Product Name : test S10B316
Product Part Number : 12345789
Product Serial : 12345789
root@test-evb:# ipmitool fru list 1
Device not present (Requested sensor, data, or record not found)

This is exposed when building under GCC 5.2:

arm-openbmc-linux-gnueabi-g++  -march=armv5e -marm  -mthumb-interwork --sysroot=/scratch/joel/openbmc/build/tmp/sysroots/palmetto  -O2 -pipe -g -feliminate-unused-debug-types  -fpic -std=gnu++14 -fvisibility-inlines-hidden   -c -o writefrudata.o writefrudata.C
cc1: warning: command line option '-std=gnu++14' is valid for C++/ObjC++ but not for C
writefrudata.C: In function 'int ipmi_validate_fru_area(uint8_t, const char*, sd_bus*, bool)':
writefrudata.C:588:51: error: cannot convert 'std::ifstream {aka std::basic_ifstream<char>}' to 'bool' in initialization
         bool present = std::ifstream(fru_file_name);
                                                   ^
<builtin>: recipe for target 'writefrudata.o' failed
make: *** [writefrudata.o] Error 1

1. workspace\sources\phosphor-ipmi-host\dbus-sdr\storagecommands.cpp
   //
   ipmi::registerHandler(ipmi::prioOpenBmcBase, ipmi::netFnStorage,
   ipmi::storage::cmdWriteFruData,
   ipmi::Privilege::Operator, ipmiStorageWriteFruData);

2. workspace\sources\phosphor-ipmi-fru\strgfnhandler.cpp
   void register_netfn_storage_write_fru()
   {
   std::printf("Registering NetFn:[0x%X], Cmd:[0x%X]\n", NETFUN_STORAGE,
   IPMI_CMD_WRITE_FRU_DATA);

   ipmi_register_callback(NETFUN_STORAGE, IPMI_CMD_WRITE_FRU_DATA, NULL,
   ipmiStorageWriteFruData, SYSTEM_INTERFACE);
   }

Hi,

Found a copy-paste typo in readeeprom.cpp. The fruid_str parsed value is not actually checked for emptiness, the eeprom_str is checked instead. This may result in false-positive argument checks.

The attached patch fixes this typo.

Regards,
Dmitry
0001-read-eeprom-fix-fruid-str-emptines-check.patch.gz

Even though std::string is part of the supported types of Value std::variant, it's deceivingly hard to use.

I was running into this issue: https://stackoverflow.com/questions/44086269/why-does-my-variant-convert-a-stdstring-to-a-bool where:
"The type of "hello" is const char [6], which decays to const char *. The conversion from const char * to bool is a built-in conversion, while the conversion from const char * to std::string is a user-defined conversion, which means the former is performed."

The fix was to declare the string using a Folded Block Scalar > and then wrapping the string in std::string("hello") for example.

A comment / documentation somewhere of this behavior may be useful as this is not very obvious.

variable 'offset' in function 'ipmiStorageWriteFruData' is declared as uint8_t, but it should be uint16_t.

ipmi_ret_t ipmiStorageWriteFruData(ipmi_netfn_t netfn, ipmi_cmd_t cmd,
ipmi_request_t request,
ipmi_response_t response,
ipmi_data_len_t dataLen,
ipmi_context_t context)
{
...
uint16_t offset = 0; // offset should be declared as uint16_t
...
offset = ((uint16_t)reqptr->offsetms) << 8 | reqptr->offsetls; // otherwise, this operation shall give unexcepted result

My understanding is we are importing code from another open source package and thus this is going to be GPLv3. Need confirmation that I picked the correct license for the initial commit.

Use “ipmitool fru edit”，want to modify the information of a certain domain；Then ipmitool fru list to check ,The data viewed using ipmitool is OK [it is the modified data];
But:

1. Abnormal l printing information has occurred here

Mar 09 16:06:01 test-evb ipmid[327]: FRU file is incomplete
Mar 09 16:06:01 test-evb ipmid[327]: Populating FRU areas failed
Mar 09 16:06:01 test-evb ipmid[327]: Incomplete FRU file
Mar 09 16:06:01 test-evb ipmid[327]: Populating FRU areas failed
Mar 09 16:06:01 test-evb ipmid[327]: Incomplete FRU file
Mar 09 16:06:01 test-evb ipmid[327]: Populating FRU areas failed
Mar 09 16:06:01 test-evb ipmid[327]: Incomplete FRU file
Mar 09 16:06:01 test-evb ipmid[327]: Populating FRU areas failed

2. eeprom EEPROM data not synchronized

test befoe:
root@test-evb:# ipmitool fru list
FRU Device Description : Builtin FRU Device (ID 0)
Board Mfg Date : Wed Jan 10 04:58:00 2024 UTC
Board Mfg : AnyCompany
Board Product : HangZhou
Board Serial : pn123456789987654321
Board Part Number : Cloudnineinfo123456789
root@test-evb:#
EEPROM Data:
root@test-evb:~# hexdump -Cv /sys/class/i2c-dev/i2c-28/device/28-0050/eeprom
00000000 01 00 00 01 00 00 00 fe 01 0a 19 aa ea e0 ca 41 |...............A|
00000010 6e 79 43 6f 6d 70 61 6e 79 c8 48 61 6e 67 5a 68 |nyCompany.HangZh|
00000020 6f 75 d4 70 6e 31 32 33 34 35 36 37 38 39 39 38 |ou.pn12345678998|
00000030 37 36 35 34 33 32 31 d6 43 6c 6f 75 64 6e 69 6e |7654321.Cloudnin|
00000040 65 69 6e 66 6f 31 32 33 34 35 36 37 38 39 c7 4e |einfo123456789.N|
00000050 6f 74 68 69 6e 67 c1 e8 ff ff ff ff ff ff ff ff |othing..........|
00000060 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000070 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000080 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000090 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|

test after:
root@test-evb:# ipmitool fru list
FRU Device Description : Builtin FRU Device (ID 0)
Board Mfg Date : Wed Jan 10 04:58:00 2024 UTC
Board Mfg : CloudNineInfo
Board Product : HangZhou
Board Serial : pn123456789987654321
Board Part Number : Cloudnineinfo123456789
root@test-evb:#
EEPROM Data:
root@test-evb:~# hexdump -Cv /sys/class/i2c-dev/i2c-28/device/28-0050/eeprom
00000000 01 00 00 01 00 00 00 fe 01 0a 19 aa ea e0 ca 41 |...............A|
00000010 6e 79 43 6f 6d 70 61 6e 79 c8 48 61 6e 67 5a 68 |nyCompany.HangZh|
00000020 6f 75 d4 70 6e 31 32 33 34 35 36 37 38 39 39 38 |ou.pn12345678998|
00000030 37 36 35 34 33 32 31 d6 43 6c 6f 75 64 6e 69 6e |7654321.Cloudnin|
00000040 65 69 6e 66 6f 31 32 33 34 35 36 37 38 39 c7 4e |einfo123456789.N|
00000050 6f 74 68 69 6e 67 c1 e8 ff ff ff ff ff ff ff ff |othing..........|
00000060 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000070 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000080 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|
00000090 ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff ff |................|

Reading vpd on IO board on PVT Barreleye seg faults and data is not added to the inventory.
System running 8.0.1 openbmc.

'''
root@barreleye:~# phosphor-read-eeprom --eeprom /sys/bus/i2c/devices/0-0050/eeprom --fruid 64
Segmentation fault
'''

What should be sent to dbus and in what format

MFG_Name = 0x9801
Model_Number = "9965439-092.A00LF "
Version = 0x0000

Serial_number = 0x1520e691

IPMI Incoming: Seq 0x1c, NetFn 0x28, CMD: 0x12
0x000000: 03 00 00 01 00 00 00 01 00 00 fe 01 05 00 02 98 ................
0x000010: 01 01 0b d2 39 39 36 35 34 33 39 2d 30 39 32 2e ....9965439-092.
0x000020: 41 30 30 4c 46 20 02 00 00 04 15 20 e6 91 00 00 A00LF ..... ....
0x000030: 00 c1 48 ..H

01 00 00 00 01 00 00 fe
^-- Product area

01 - version
05 - length = 5*8 = 40
00 - Language code
02 - 0b00 000010 = MFG Name - Binary for 2 bytes
98 01 - 0x9801 is the MFG name
01 - Product Name length - Binary for 1 byte
0b - Name is 0x0b
d2 - 0b11010010 Product Model Number - Ascii for 18 bytes
39 39 36 35 34 33 39 2d 30 39 32 2e 41 30 30 4c 46 20 - (9965439-092.A00LF )
02 - Product Version Length - binary for 2 bytes
00 00
04 - Product Serial number length
15 20 e6 91 - 0x1520e691 is the serial number
00 - Asset tag length
00 fru file id length
00 - Custom product info area length
C1 - Done
48 - CRC

During system boot time the mapper is under heavy load and could be flaky in returning the service name. Better to try call phosphor-inventory-manager directly if the mapper call fails.