Extract Monitor Serial Number / Manufacture Date Using EDID?

[ASillyQuestion]

Hello,

I can not for the life of me determine how to get the monitor serial number / manufacture date using Extended Display Identification Data (EDID). Does anyone know?

None of the following provide this info:

monitor-get-edid
monitor-edid --v
monitor-edid --MonitorsDB
monitor-get-edid | monitor-parse-edid
dmidecode -q

Thanks!

[sonoran]

The info is in the Xorg log, thus:

[    30.459] (II) MGA(0): I2C monitor info
[    30.459] (II) MGA(0): Manufacturer: VSC  Model: 7d20  Serial#: 16843009
[    30.459] (II) MGA(0): Year: 2010  Week: 20
[    30.459] (II) MGA(0): EDID Version: 1.3
[    30.459] (II) MGA(0): Digital Display Input
[    30.459] (II) MGA(0): Max Image Size [cm]: horiz.: 38  vert.: 30
[    30.459] (II) MGA(0): Gamma: 2.20

It appears X gets this info via the driver, in my case mga.

---------- Post added at 11:55 PM CDT ---------- Previous post was at 09:07 PM CDT ----------

I downloaded the read-edid package from the wikipedia link, compiled it, but get-edid went through a few iterations and then told me

EDID claims 10 more blocks left
EDID blocks left is wrong.
Your EDID is probably invalid.

I also downloaded the source code for the mga driver, which I always wanted an excuse to look at anyway. The driver communicates directly with the graphics card, and sends a message to the Xorg server with the edid monitor info, which the server then writes to the log.

If you just want the edid information, it's in the log. If you're writing code for this purpose, you could get the data from the driver the same way the server does. There is probably a way to query the server for the information as well.

[ASillyQuestion]

Thanks, sorry, I should have added that to my list of unsuccessfully attempts. I'm using the Nvidia driver and /var/log/Xorg.0.log lists the manufacturer, but not the serial number or manufacture date. :-/

---------- Post added at 02:06 AM CDT ---------- Previous post was at 02:02 AM CDT ----------

Also, I discovered in my trials that you shouldn't need to compile anything manually, I don't believe, everything should be available via yum.

As root:

yum install monitor-edid i2c-tools
modprobe i2c-dev (May or may not be needed)

Possible options:

monitor-get-edid
monitor-edid --v
monitor-edid --MonitorsDB
monitor-get-edid | monitor-parse-edid

[sonoran]

Okay, tried it the easy way, ,and you're right, none of those programs reports the serial # or date of manufacture. Odd, considering that the Xorg server manages to get both for me, so that info must be available.

Looking at the way that information is presented in my log, the "I2C monitor info" comes first, including serial # and date, and THEN it says "EDID Version:" with the rest of the data. Maybe these are two separate steps. The man pages for the i2c tools are full of dire warnings, though: "WARNING: This program can confuse your I2C bus, cause data loss and worse!" So I don't really feel like toying with them and finding out what they mean by "worse".

[ASillyQuestion]

Thanks for trying. That is weird that the Xorg server gives you this info, but it doesn't work for me. I'd be okay parsing it from that if I could, but it just isn't there.

I still haven't figured out how to do this yet. If anyone has any suggestions, I'd love to hear them, thanks!

[sonoran]

I was reading some more on this and found references to a package called xresprobe and a program called ddcprobe, but could not find a current rpm for either. http://www.cyberciti.biz/faq/howto-u...probe-command/

And the danger in messing around on the I2C bus is that you can scramble, clear, or reprogram eproms in your hardware unless you know what you're doing - which I certainly don't here.

Interesting problem, though.

[SiliconSlick]

FWIW, here's a couple pennies from me:

1) a quick search reveals ddcprobe was replaced with dmidecode (and similar) around Fedora 9 and the introduction of HAL

2) on my FC13 x86_64 server running the RPMFusion nvidia driver, /var/log/Xorg.0.log, the nvidia driver correctly detects the monitor model (Viewsonic G90mb), but doesn't give the serial number

3) on my FC13 x86_64 laptop running the nouveau driver and with the same monitor attached, the nouveau driver not only gets the model but also the serial number (Serial No: 319004300773), which happens to match the serial number on the back of the CRT (i.e. it is correct)

4) on that same laptop, the serial number also shows up in a hex dump of a file in /sys. e.g.

$ lspci | grep VGA
01:00.0 VGA compatible controller: nVidia Corporation NV17 [GeForce4 440 Go 64M] (rev a3)
$ xxd  /sys/devices/pci0000:00/0000:00:0b.0/0000:01:00.0/drm/card0/card0-VGA-1/edid
0000000: 00ff ffff ffff ff00 5a63 0213 0101 0101  ........Zc......
0000010: 2b0a 0103 1c25 1bb0 eb00 b8a0 5749 9b26  +....%......WI.&
0000020: 1048 4cff ff80 8199 8159 714f 6159 4559  .HL......YqOaYEY
0000030: 3159 a94f 0101 863d 00c0 5100 3040 40a0  1Y.O...=..Q.0@@.
0000040: 1300 680e 1100 001e 0000 00ff 0033 3139  ..h..........319
0000050: 3030 3433 3030 3737 330a 0000 00fd 0032  004300773......2
0000060: a01e 6114 000a 2020 2020 2020 0000 00fc  ..a...      ....
0000070: 0047 3930 6d62 0a20 2020 2020 2020 00ba  .G90mb.       ..

(bytes 0x4D-0x58)

There is no similar file anywhere in /sys on the server running the nvidia driver.

So, at the very least, you might want to try the nouveau drivers instead of the nvidia one since it seems to be a feature of the display driver

SS.

[sonoran]

which happens to match the serial number on the back of the CRT

Gasp! You looked at the back of the monitor? That seems unfair, somehow. But thanks for the info.

[ASillyQuestion]

LOL.

This also worked for me on a RedHat machine:

startx -- -logverbose 6