11:15 a.m.
----- Original Message -----
Aaron and Buland,
I was just about to check this in when I thought that maybe it could
be a bit more informative than just "dmi_ident[#]: <string>", given
that the output could also incorporate the enum name/index, as is done
with a few other crash commands:
crash> dmi_field
enum dmi_field {
DMI_NONE = 0
DMI_BIOS_VENDOR = 1
DMI_BIOS_VERSION = 2
DMI_BIOS_DATE = 3
DMI_SYS_VENDOR = 4
DMI_PRODUCT_NAME = 5
DMI_PRODUCT_VERSION = 6
DMI_PRODUCT_SERIAL = 7
DMI_PRODUCT_UUID = 8
DMI_BOARD_VENDOR = 9
DMI_BOARD_NAME = 10
DMI_BOARD_VERSION = 11
DMI_BOARD_SERIAL = 12
DMI_BOARD_ASSET_TAG = 13
DMI_CHASSIS_VENDOR = 14
DMI_CHASSIS_TYPE = 15
DMI_CHASSIS_VERSION = 16
DMI_CHASSIS_SERIAL = 17
DMI_CHASSIS_ASSET_TAG = 18
DMI_STRING_MAX = 19
};
Although, gdb doesn't have the dmi_field enum data available like the
above in really old kernels, so in those rare cases it could only be
done as you have. Maybe there could be an index_to_DMI_field() function
that returns the string above, or just a number string if it can't do
the translation?
Dave
The attached patch is what I'm thinking. On a Lenovo laptop it looks like:
crash> sys -i
DMI_BIOS_VENDOR: LENOVO
DMI_BIOS_VERSION: G4ET37WW (1.12 )
DMI_BIOS_DATE: 05/29/2012
DMI_SYS_VENDOR: LENOVO
DMI_PRODUCT_NAME: 2429BQ1
DMI_PRODUCT_VERSION: ThinkPad T530
DMI_PRODUCT_SERIAL: R9R91HZ
DMI_PRODUCT_UUID: 568DFA01-5180-11CB-B851-BD06085ADDB0
DMI_BOARD_VENDOR: LENOVO
DMI_BOARD_NAME: 2429BQ1
DMI_BOARD_VERSION: Not Available
DMI_BOARD_SERIAL: 1ZLV127F17M
DMI_BOARD_ASSET_TAG: Not Available
DMI_CHASSIS_VENDOR: LENOVO
DMI_CHASSIS_TYPE: 10
DMI_CHASSIS_VERSION: Not Available
DMI_CHASSIS_SERIAL: R9R91HZ
DMI_CHASSIS_ASSET_TAG: RH0004111
crash>
A recent HP kernel looks like this though:
crash> sys -i
DMI_BIOS_VENDOR: Hewlett-Packard
DMI_BIOS_VERSION: 786F7 v01.05
DMI_BIOS_DATE: 04/25/2008
DMI_SYS_VENDOR: Hewlett-Packard
DMI_PRODUCT_NAME: HP xw4550 Workstation
DMI_PRODUCT_VERSION:
DMI_PRODUCT_SERIAL: BAL1010051
DMI_PRODUCT_UUID: 308A59FA-9B77-DE11-BBDA-C486CC82001C
DMI_BOARD_VENDOR: Hewlett-Packard
DMI_BOARD_NAME: 0AC8h
DMI_BOARD_VERSION:
DMI_BOARD_SERIAL: BAL1010051
DMI_BOARD_ASSET_TAG:
DMI_CHASSIS_VENDOR: Hewlett-Packard
DMI_CHASSIS_TYPE: 6
DMI_CHASSIS_VERSION:
DMI_CHASSIS_SERIAL: BAL1010051
DMI_CHASSIS_ASSET_TAG:
crash>
when the "empty" entries are not NULL pointers, but rather pointers
to NULL strings or strings with spaces in them:
crash> p dmi_ident
dmi_ident = $1 =
{0x0, 0xffffffff8203b000 "Hewlett-Packard", 0xffffffff8203b010 "786F7
v01.05", 0xffffffff8203b020 "04/25/2008", 0xffffffff8203b02c
"Hewlett-Packard", 0xffffffff8203b03c "HP xw4550 Workstation",
0xffffffff8203b054 " ", 0xffffffff8203b058 "BAL1010051",
0xffffffff8203b064 "308A59FA-9B77-DE11-BBDA-C486CC82001C", 0xffffffff8203b08c
"Hewlett-Packard", 0xffffffff8203b09c "0AC8h", 0xffffffff8203b0a4
"", 0xffffffff8203b0a8 "BAL1010051", 0xffffffff8203b0b4 "",
0xffffffff8203b0b8 "Hewlett-Packard", 0xffffffff8203b0c8 "6",
0xffffffff8203b0cc "", 0xffffffff8203b0d0 "BAL1010051",
0xffffffff8203b0dc " "}
crash>
Do you think those should should be just dropped, or shown as above?
Dave
6:37 a.m.
New subject: [PATCH] sys: Introduce -i option to dump dmi_ident data
On Tue 2015-12-08 12:15 -0500, Dave Anderson wrote:
Hi Dave,
Sorry for the delay. Please find a v2 patch attached with a similar
approach to your attempt to add support for 'dmi_field' when available.
The attached patch is what I'm thinking. On a Lenovo laptop it
looks
like:
crash> sys -i
DMI_BIOS_VENDOR: LENOVO
DMI_BIOS_VERSION: G4ET37WW (1.12 )
DMI_BIOS_DATE: 05/29/2012
DMI_SYS_VENDOR: LENOVO
DMI_PRODUCT_NAME: 2429BQ1
DMI_PRODUCT_VERSION: ThinkPad T530
DMI_PRODUCT_SERIAL: R9R91HZ
DMI_PRODUCT_UUID: 568DFA01-5180-11CB-B851-BD06085ADDB0
DMI_BOARD_VENDOR: LENOVO
DMI_BOARD_NAME: 2429BQ1
DMI_BOARD_VERSION: Not Available
DMI_BOARD_SERIAL: 1ZLV127F17M
DMI_BOARD_ASSET_TAG: Not Available
DMI_CHASSIS_VENDOR: LENOVO
DMI_CHASSIS_TYPE: 10
DMI_CHASSIS_VERSION: Not Available
DMI_CHASSIS_SERIAL: R9R91HZ
DMI_CHASSIS_ASSET_TAG: RH0004111
crash>
A recent HP kernel looks like this though:
crash> sys -i
DMI_BIOS_VENDOR: Hewlett-Packard
DMI_BIOS_VERSION: 786F7 v01.05
DMI_BIOS_DATE: 04/25/2008
DMI_SYS_VENDOR: Hewlett-Packard
DMI_PRODUCT_NAME: HP xw4550 Workstation
DMI_PRODUCT_VERSION:
DMI_PRODUCT_SERIAL: BAL1010051
DMI_PRODUCT_UUID: 308A59FA-9B77-DE11-BBDA-C486CC82001C
DMI_BOARD_VENDOR: Hewlett-Packard
DMI_BOARD_NAME: 0AC8h
DMI_BOARD_VERSION:
DMI_BOARD_SERIAL: BAL1010051
DMI_BOARD_ASSET_TAG:
DMI_CHASSIS_VENDOR: Hewlett-Packard
DMI_CHASSIS_TYPE: 6
DMI_CHASSIS_VERSION:
DMI_CHASSIS_SERIAL: BAL1010051
DMI_CHASSIS_ASSET_TAG:
crash>
when the "empty" entries are not NULL pointers, but rather pointers to
NULL strings or strings with spaces in them:
crash> p dmi_ident
dmi_ident = $1 =
{0x0, 0xffffffff8203b000 "Hewlett-Packard", 0xffffffff8203b010 "786F7
v01.05", 0xffffffff8203b020 "04/25/2008", 0xffffffff8203b02c
"Hewlett-Packard", 0xffffffff8203b03c "HP xw4550 Workstation",
0xffffffff8203b054 " ", 0xffffffff8203b058 "BAL1010051",
0xffffffff8203b064 "308A59FA-9B77-DE11-BBDA-C486CC82001C", 0xffffffff8203b08c
"Hewlett-Packard", 0xffffffff8203b09c "0AC8h", 0xffffffff8203b0a4
"", 0xffffffff8203b0a8 "BAL1010051", 0xffffffff8203b0b4 "",
0xffffffff8203b0b8 "Hewlett-Packard", 0xffffffff8203b0c8 "6",
0xffffffff8203b0cc "", 0xffffffff8203b0d0 "BAL1010051",
0xffffffff8203b0dc " "}
crash>
Do you think those should should be just dropped, or shown as above?
What does 'dmidecode' show? If dmidecode shows the same, I'd suggest that
perhaps we should follow suit.
--
Aaron Tomlin
8:33 a.m.
New subject: [PATCH] sys: Introduce -i option to dump dmi_ident data
----- Original Message -----
On Tue 2015-12-08 12:15 -0500, Dave Anderson wrote:
Hi Dave,
Sorry for the delay. Please find a v2 patch attached with a similar
approach to your attempt to add support for 'dmi_field' when available.
> The attached patch is what I'm thinking. On a Lenovo laptop it looks
> like:
>
> crash> sys -i
> DMI_BIOS_VENDOR: LENOVO
> DMI_BIOS_VERSION: G4ET37WW (1.12 )
> DMI_BIOS_DATE: 05/29/2012
> DMI_SYS_VENDOR: LENOVO
> DMI_PRODUCT_NAME: 2429BQ1
> DMI_PRODUCT_VERSION: ThinkPad T530
> DMI_PRODUCT_SERIAL: R9R91HZ
> DMI_PRODUCT_UUID: 568DFA01-5180-11CB-B851-BD06085ADDB0
> DMI_BOARD_VENDOR: LENOVO
> DMI_BOARD_NAME: 2429BQ1
> DMI_BOARD_VERSION: Not Available
> DMI_BOARD_SERIAL: 1ZLV127F17M
> DMI_BOARD_ASSET_TAG: Not Available
> DMI_CHASSIS_VENDOR: LENOVO
> DMI_CHASSIS_TYPE: 10
> DMI_CHASSIS_VERSION: Not Available
> DMI_CHASSIS_SERIAL: R9R91HZ
> DMI_CHASSIS_ASSET_TAG: RH0004111
> crash>
>
> A recent HP kernel looks like this though:
>
> crash> sys -i
> DMI_BIOS_VENDOR: Hewlett-Packard
> DMI_BIOS_VERSION: 786F7 v01.05
> DMI_BIOS_DATE: 04/25/2008
> DMI_SYS_VENDOR: Hewlett-Packard
> DMI_PRODUCT_NAME: HP xw4550 Workstation
> DMI_PRODUCT_VERSION:
> DMI_PRODUCT_SERIAL: BAL1010051
> DMI_PRODUCT_UUID: 308A59FA-9B77-DE11-BBDA-C486CC82001C
> DMI_BOARD_VENDOR: Hewlett-Packard
> DMI_BOARD_NAME: 0AC8h
> DMI_BOARD_VERSION:
> DMI_BOARD_SERIAL: BAL1010051
> DMI_BOARD_ASSET_TAG:
> DMI_CHASSIS_VENDOR: Hewlett-Packard
> DMI_CHASSIS_TYPE: 6
> DMI_CHASSIS_VERSION:
> DMI_CHASSIS_SERIAL: BAL1010051
> DMI_CHASSIS_ASSET_TAG:
> crash>
>
> when the "empty" entries are not NULL pointers, but rather pointers to
> NULL strings or strings with spaces in them:
>
> crash> p dmi_ident
> dmi_ident = $1 =
> {0x0, 0xffffffff8203b000 "Hewlett-Packard", 0xffffffff8203b010
"786F7
> v01.05", 0xffffffff8203b020 "04/25/2008", 0xffffffff8203b02c
> "Hewlett-Packard", 0xffffffff8203b03c "HP xw4550 Workstation",
> 0xffffffff8203b054 " ", 0xffffffff8203b058 "BAL1010051",
> 0xffffffff8203b064 "308A59FA-9B77-DE11-BBDA-C486CC82001C",
> 0xffffffff8203b08c "Hewlett-Packard", 0xffffffff8203b09c
"0AC8h",
> 0xffffffff8203b0a4 "", 0xffffffff8203b0a8 "BAL1010051",
> 0xffffffff8203b0b4 "", 0xffffffff8203b0b8 "Hewlett-Packard",
> 0xffffffff8203b0c8 "6", 0xffffffff8203b0cc "",
0xffffffff8203b0d0
> "BAL1010051", 0xffffffff8203b0dc " "}
> crash>
>
> Do you think those should should be just dropped, or shown as above?
What does 'dmidecode' show? If dmidecode shows the same, I'd suggest that
perhaps we should follow suit.
It's difficult to tell with dmidecode. Plus there are a bunch of other
odd-ball settings like these examples:
DMI_BOARD_ASSET_TAG: -
DMI_CHASSIS_ASSET_TAG: ................
DMI_SYS_VENDOR: To Be Filled By O.E.M.
DMI_BOARD_VERSION: N/A
DMI_CHASSIS_SERIAL: Not Available
DMI_CHASSIS_ASSET_TAG: No Asset Information
DMI_PRODUCT_SERIAL: n/a
DMI_BOARD_VERSION: (none)
DMI_CHASSIS_VERSION: none
DMI_BOARD_SERIAL: None
DMI_CHASSIS_VENDOR: No Enclosure
DMI_CHASSIS_ASSET_TAG: No Asset Tag
In any case, if there is anything but a NULL pointer, then some kernel
entity has explicitly set the string, so I'm going to just show it.
Queued for crash-7.1.4:
https://github.com/crash-utility/crash/commit/9ba68adb0c2a0e6f4839fdc3c09...
Thanks,
Dave
3275
days inactive
3276
days old
devel@lists.crash-utility.osci.io
2 comments
2 participants
participants (2)
-
Aaron Tomlin -
Dave Anderson