Monitor's EDID not recognised

Monitor's EDID not recognised

[Jeremy Murphy]

Apologies for what might be a frequently asked question, but I haven't found 
an answer on the web yet.

I have a Sony CPD-G520 (21" CRT monitor) for which I would like to be able to 
read the EDID.  However, sensors-detect finds SPD EEPROM where the EDID 
EEPROM probably, and typically, is.

This is an edit from sensors-detect output:

Next adapter: SMBus I801 adapter at 0400 (i2c-0)
Do you want to scan it? (YES/no/selectively):
Client found at address 0x44
Probing for `Maxim MAX6633/MAX6634/MAX6635'...              No
Client found at address 0x50
Probing for `Analog Devices ADM1033'...                     No
Probing for `Analog Devices ADM1034'...                     No
Probing for `SPD EEPROM'...                                 Yes
    (confidence 8, not a hardware monitoring chip)
Probing for `EDID EEPROM'...                                No

It then gives the same output for 0x50 at 0x51, 0x52 and 0x53.  When I load 
the eeprom module it says that it is handling it.  The problem is that there 
are no valid EDIDs at those addresses, as reported by decode-edid.pl (it's a 
64-bit system so no get-edid).  So I'm not sure whether a) the EDID is 
corrupt, or b) it's not actually accessing the EDID.  Any ideas?

The graphics card is an NVIDIA 7600GS, motherboard chipset is the Intel ICH7 
family.  Hardware monitoring of CPU and motherboard temperature via I2C works 
fine.  Thanks!

Jeremy

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jean Delvare]

Hi Jeremy,

On Fri, 7 Mar 2008 19:23:31 +1100, Jeremy Murphy wrote:
> 
> Apologies for what might be a frequently asked question, but I haven't found 
> an answer on the web yet.
> 
> I have a Sony CPD-G520 (21" CRT monitor) for which I would like to be able to 
> read the EDID.  However, sensors-detect finds SPD EEPROM where the EDID 
> EEPROM probably, and typically, is.
> 
> This is an edit from sensors-detect output:
> 
> Next adapter: SMBus I801 adapter at 0400 (i2c-0)
> Do you want to scan it? (YES/no/selectively):
> Client found at address 0x44
> Probing for `Maxim MAX6633/MAX6634/MAX6635'...              No
> Client found at address 0x50
> Probing for `Analog Devices ADM1033'...                     No
> Probing for `Analog Devices ADM1034'...                     No
> Probing for `SPD EEPROM'...                                 Yes
>     (confidence 8, not a hardware monitoring chip)
> Probing for `EDID EEPROM'...                                No
> 
> It then gives the same output for 0x50 at 0x51, 0x52 and 0x53.  When I load 
> the eeprom module it says that it is handling it.  The problem is that there 
> are no valid EDIDs at those addresses, as reported by decode-edid.pl (it's a 
> 64-bit system so no get-edid).  So I'm not sure whether a) the EDID is 
> corrupt, or b) it's not actually accessing the EDID.  Any ideas?

What does decode-dimms.pl say? If you have 4 memory modules on this
system then it is expected that sensors-detect (or i2cdetect) will see
4 SPD EEPROMs at 0x50, 0x51, 0x52 and 0x53.

> The graphics card is an NVIDIA 7600GS, motherboard chipset is the Intel ICH7 
> family.  Hardware monitoring of CPU and motherboard temperature via I2C works 
> fine.  Thanks!

I'm not sure why you would expect the EDID EEPROM on the ICH7 SMBus.
I've never seen an EDID EEPROM there. The EDID EEPROM is typically
accessed through an I2C bus on the graphics adapter. You will need a
driver for this. The binary nvidia X11 driver should work, the nvidiafb
framebuffer driver should work as well (that's what I am using myself).
After loading the driver, 2 or 3 additional I2C buses should be listed
by "i2cdetect -l" (don't forget to load i2c-dev beforehand), and
decode-edid.pl (or the more user-friendly ddcmon script) should find
the EDID on one of them.

-- 
Jean Delvare

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jeremy Murphy]

On Fri, 7 Mar 2008 09:06:10 pm Jean Delvare wrote:
>
> What does decode-dimms.pl say? If you have 4 memory modules on this
> system then it is expected that sensors-detect (or i2cdetect) will see
> 4 SPD EEPROMs at 0x50, 0x51, 0x52 and 0x53.

You're absolutely right, they are my DIMMS.  I assumed that was the EDID and 
its shadows because they started at the same address and I'd seen the same 
output on some other nvidia users' reports.  Now I know better, thanks!


> I'm not sure why you would expect the EDID EEPROM on the ICH7 SMBus.
> I've never seen an EDID EEPROM there. The EDID EEPROM is typically
> accessed through an I2C bus on the graphics adapter. You will need a
> driver for this. The binary nvidia X11 driver should work, the nvidiafb
> framebuffer driver should work as well (that's what I am using myself).
> After loading the driver, 2 or 3 additional I2C buses should be listed
> by "i2cdetect -l" (don't forget to load i2c-dev beforehand), and
> decode-edid.pl (or the more user-friendly ddcmon script) should find
> the EDID on one of them.

I'm using nvidia's binary driver v169.07, and yes, the nvidia I2C buses are 
there.  i2cdetect -l:

i2c-3   i2c             NVIDIA i2c adapter                      I2C adapter
i2c-2   i2c             NVIDIA i2c adapter                      I2C adapter
i2c-1   i2c             NVIDIA i2c adapter                      I2C adapter
i2c-0   smbus           SMBus I801 adapter at 0400              SMBus adapter

However, decode-edid.pl and ddcmon both report that an EDID EEPROM is not 
found.

Jeremy

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jean Delvare]

On Fri, 7 Mar 2008 22:37:36 +1100, Jeremy Murphy wrote:
> > I'm not sure why you would expect the EDID EEPROM on the ICH7 SMBus.
> > I've never seen an EDID EEPROM there. The EDID EEPROM is typically
> > accessed through an I2C bus on the graphics adapter. You will need a
> > driver for this. The binary nvidia X11 driver should work, the nvidiafb
> > framebuffer driver should work as well (that's what I am using myself).
> > After loading the driver, 2 or 3 additional I2C buses should be listed
> > by "i2cdetect -l" (don't forget to load i2c-dev beforehand), and
> > decode-edid.pl (or the more user-friendly ddcmon script) should find
> > the EDID on one of them.
> 
> I'm using nvidia's binary driver v169.07, and yes, the nvidia I2C buses are 
> there.  i2cdetect -l:
> 
> i2c-3   i2c             NVIDIA i2c adapter                      I2C adapter
> i2c-2   i2c             NVIDIA i2c adapter                      I2C adapter
> i2c-1   i2c             NVIDIA i2c adapter                      I2C adapter
> i2c-0   smbus           SMBus I801 adapter at 0400              SMBus adapter
> 
> However, decode-edid.pl and ddcmon both report that an EDID EEPROM is not 
> found.

Please provide the output of:
i2cdetect 1
i2cdetect 2
i2cdetect 3

Either there is no EDID EEPROM connected at all (some monitors don't
have them, in particular older models) or it has a format different
from what the scripts expect.

If anything shows up at 0x50 with i2cdetect, please provide a dump
using i2cdump (rmmod eeprom first.)

-- 
Jean Delvare

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jeremy Murphy]

On Fri, 7 Mar 2008 11:11:20 pm Jean Delvare wrote:
> On Fri, 7 Mar 2008 22:37:36 +1100, Jeremy Murphy wrote:
> >
> > I'm using nvidia's binary driver v169.07, and yes, the nvidia I2C buses
> > are there.  i2cdetect -l:
> >
> > i2c-3   i2c             NVIDIA i2c adapter                      I2C
> > adapter i2c-2   i2c             NVIDIA i2c adapter                     
> > I2C adapter i2c-1   i2c             NVIDIA i2c adapter                   
> >   I2C adapter i2c-0   smbus           SMBus I801 adapter at 0400         
> >     SMBus adapter
> >
> > However, decode-edid.pl and ddcmon both report that an EDID EEPROM is not
> > found.
>
> Please provide the output of:

I've just learnt that a number of people could no longer access their EDID 
after upgrading their nvidia driver to the 100.* series.  I downgraded to 
96.43.01 and now their is _something_ on the I2C nvidia bus 1, but it's not 
pretty.

These following results are inconsistent, which I'm guessing is not a good 
sign.  They vary somewhat between each call but they always have a similar 
pattern (row+col).


> i2cdetect 1

This is with eeprom loaded:

     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f
10: 10 11 12 13 14 15 16 17 18 19 1a 1b 1c -- -- --
20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 3f
40: 40 41 42 43 44 45 46 47 48 49 4a 4b 4c 4d 4e 4f
50: 50 51 52 53 54 55 56 57 58 59 5a 5b 5c 5d 5e 5f
60: 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f
70: 70 71 72 73 74 75 76 77

This is without:

     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
00:          -- -- -- -- -- -- -- -- 0b 0c 0d 0e 0f
10: 10 11 12 13 14 15 16 -- -- -- -- -- -- -- -- --
20: -- -- -- -- 24 25 26 27 28 29 2a 2b 2c 2d 2e 2f
30: 30 31 32 33 34 35 36 37 38 39 3a 3b 3c 3d 3e 3f
40: 40 41 42 43 44 -- -- -- -- -- -- -- -- -- -- --
50: -- -- -- -- -- -- -- -- -- -- 5a 5b 5c 5d 5e 5f
60: 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f
70: 70 71 72 73 74 75 76 77

At a guess, I'd say this is garbage?


> i2cdetect 2
> i2cdetect 3

Both empty.


> Either there is no EDID EEPROM connected at all (some monitors don't
> have them, in particular older models) or it has a format different
> from what the scripts expect.
>
> If anything shows up at 0x50 with i2cdetect, please provide a dump
> using i2cdump (rmmod eeprom first.)

This is the output of i2cdump 1 0x50:

     0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
00: 3f XX XX XX XX XX XX XX XX XX XX 00 00 00 00 00    ?XXXXXXXXXX.....
10: 00 00 00 00 XX XX XX XX XX XX XX XX XX XX 00 00    ....XXXXXXXXXX..
20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
30: 00 00 00 00 00 XX XX XX XX XX XX XX XX XX 00 00    .....XXXXXXXXX..
40: 00 00 00 00 00 00 00 00 XX XX XX XX XX 00 00 00    ........XXXXX...
50: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
60: 00 00 00 00 00 00 00 00 00 00 00 XX XX XX XX XX    ...........XXXXX
70: XX XX XX XX XX XX 00 00 00 00 00 00 00 00 00 00    XXXXXX..........
80: 00 00 XX XX XX XX XX XX XX XX 00 00 00 00 00 00    ..XXXXXXXX......
90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
a0: XX 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    X...............
b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
c0: 00 00 00 00 XX XX XX XX XX XX XX XX XX XX XX 00    ....XXXXXXXXXXX.
d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
f0: 00 00 00 00 00 00 XX XX XX XX XX XX XX XX XX XX    ......XXXXXXXXXX

I presume the monitor has an EDID because the brochure says that it supports 
DDC-2B and DDC-2Bi, and it's relatively recent as far as CRTs go.  Thanks for 
your help, cheers.

Jeremy

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jean Delvare]

Hi Jeremy,

On Sat, 8 Mar 2008 00:39:55 +1100, Jeremy Murphy wrote:
> I've just learnt that a number of people could no longer access their EDID 
> after upgrading their nvidia driver to the 100.* series.  I downgraded to 
> 96.43.01 and now their is _something_ on the I2C nvidia bus 1, but it's not 
> pretty.
> 
> These following results are inconsistent, which I'm guessing is not a good 
> sign.  They vary somewhat between each call but they always have a similar 
> pattern (row+col).
> 
> 
> > i2cdetect 1
> 
> This is with eeprom loaded:
> 
>      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
> 00:          03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f
> 10: 10 11 12 13 14 15 16 17 18 19 1a 1b 1c -- -- --
> 20: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- --
> 30: -- -- -- -- -- -- -- -- -- -- -- -- -- -- -- 3f
> 40: 40 41 42 43 44 45 46 47 48 49 4a 4b 4c 4d 4e 4f
> 50: 50 51 52 53 54 55 56 57 58 59 5a 5b 5c 5d 5e 5f
> 60: 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f
> 70: 70 71 72 73 74 75 76 77
> 
> This is without:
> 
>      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f
> 00:          -- -- -- -- -- -- -- -- 0b 0c 0d 0e 0f
> 10: 10 11 12 13 14 15 16 -- -- -- -- -- -- -- -- --
> 20: -- -- -- -- 24 25 26 27 28 29 2a 2b 2c 2d 2e 2f
> 30: 30 31 32 33 34 35 36 37 38 39 3a 3b 3c 3d 3e 3f
> 40: 40 41 42 43 44 -- -- -- -- -- -- -- -- -- -- --
> 50: -- -- -- -- -- -- -- -- -- -- 5a 5b 5c 5d 5e 5f
> 60: 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f
> 70: 70 71 72 73 74 75 76 77
> 
> At a guess, I'd say this is garbage?

I'd say the same, yes. The bus driver sees seemingly random acks and
nacks, but it is impossible that the above map is anything real. This
looks more like a hardware problem than a software issue though, so I'm
surprised that downgrading the nvidia driver changed anything.

BTW, note that you should NOT access the nvidia I2C buses created by
the binary nvidia X11 driver while not under X.

> > i2cdetect 2
> > i2cdetect 3
> 
> Both empty.
> 
> 
> > Either there is no EDID EEPROM connected at all (some monitors don't
> > have them, in particular older models) or it has a format different
> > from what the scripts expect.
> >
> > If anything shows up at 0x50 with i2cdetect, please provide a dump
> > using i2cdump (rmmod eeprom first.)
> 
> This is the output of i2cdump 1 0x50:
> 
>      0  1  2  3  4  5  6  7  8  9  a  b  c  d  e  f    0123456789abcdef
> 00: 3f XX XX XX XX XX XX XX XX XX XX 00 00 00 00 00    ?XXXXXXXXXX.....
> 10: 00 00 00 00 XX XX XX XX XX XX XX XX XX XX 00 00    ....XXXXXXXXXX..
> 20: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> 30: 00 00 00 00 00 XX XX XX XX XX XX XX XX XX 00 00    .....XXXXXXXXX..
> 40: 00 00 00 00 00 00 00 00 XX XX XX XX XX 00 00 00    ........XXXXX...
> 50: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> 60: 00 00 00 00 00 00 00 00 00 00 00 XX XX XX XX XX    ...........XXXXX
> 70: XX XX XX XX XX XX 00 00 00 00 00 00 00 00 00 00    XXXXXX..........
> 80: 00 00 XX XX XX XX XX XX XX XX 00 00 00 00 00 00    ..XXXXXXXX......
> 90: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> a0: XX 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    X...............
> b0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> c0: 00 00 00 00 XX XX XX XX XX XX XX XX XX XX XX 00    ....XXXXXXXXXXX.
> d0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> e0: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00    ................
> f0: 00 00 00 00 00 00 XX XX XX XX XX XX XX XX XX XX    ......XXXXXXXXXX

Noise again...

> I presume the monitor has an EDID because the brochure says that it supports 
> DDC-2B and DDC-2Bi, and it's relatively recent as far as CRTs go.  Thanks for 
> your help, cheers.

OK. Two more things you can try:

* Get rid of the binary nvidia driver for a moment, and try the
nvidiafb driver. If it gets better, it suggests that the binary nvidia
driver is at fault. If you get the same noise, it has to be a hardware
issue.

* Did you try another video cable? The I2C bus signal is carried over 2
dedicated pins on the VGA connector (12 and 15), if your cable happens
to not have them wired properly for some reason, this could explain
your problem.

-- 
Jean Delvare

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jeremy Murphy]

On Sat, 8 Mar 2008 01:08:27 am Jean Delvare wrote:
>
> BTW, note that you should NOT access the nvidia I2C buses created by
> the binary nvidia X11 driver while not under X.

OK, thanks.  I did look for them outside of X, but they weren't there.


> OK. Two more things you can try:
>
> * Get rid of the binary nvidia driver for a moment, and try the
> nvidiafb driver. If it gets better, it suggests that the binary nvidia
> driver is at fault. If you get the same noise, it has to be a hardware
> issue.

There are no nVidia I2C buses visible with the nvidiafb driver, i2cdetect 
reports only the SMBus.  Here's a snippet from the X log, though:

(II) Loading sub module "i2c"
(II) LoadModule: "i2c"(II) Module already built-in
(II) Loading sub module "ddc"
(II) LoadModule: "ddc"(II) Module already built-in
(II) NV(0): I2C bus "DDC" initialized.
(II) NV(0): Probing for analog device on output A...
(--) NV(0):   ...found one
(II) NV(0): Probing for analog device on output B...
(--) NV(0):   ...can't find one
(II) NV(0): Probing for EDID on I2C bus A...
(II) NV(0): I2C device "DDC:ddc2" registered at address 0xA0.
(II) NV(0): I2C device "DDC:ddc2" removed.
(II) NV(0):   ... none found
(II) NV(0): Probing for EDID on I2C bus B...
(II) NV(0): I2C device "DDC:ddc2" registered at address 0xA0.
(II) NV(0): I2C device "DDC:ddc2" removed.
(II) NV(0):   ... none found
(--) NV(0): CRTC 0 appears to have a CRT attached
(II) NV(0): Using CRT on CRTC 0

Address 0xA0?  Should that be 0x50?  


> * Did you try another video cable? The I2C bus signal is carried over 2
> dedicated pins on the VGA connector (12 and 15), if your cable happens
> to not have them wired properly for some reason, this could explain
> your problem.

OK, I'll have to ask around for one.  Thanks for your help.

Jeremy

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c

Re: Monitor's EDID not recognised

[Jean Delvare]

On Sun, 9 Mar 2008 17:08:20 +1100, Jeremy Murphy wrote:
> On Sat, 8 Mar 2008 01:08:27 am Jean Delvare wrote:
> >
> > BTW, note that you should NOT access the nvidia I2C buses created by
> > the binary nvidia X11 driver while not under X.
> 
> OK, thanks.  I did look for them outside of X, but they weren't there.

Err, that's exactly what I said. With the binary nvidia X11 driver, you
must be in the X session when trying to access the I2C buses, otherwise
it doesn't work.

> > * Get rid of the binary nvidia driver for a moment, and try the
> > nvidiafb driver. If it gets better, it suggests that the binary nvidia
> > driver is at fault. If you get the same noise, it has to be a hardware
> > issue.
> 
> There are no nVidia I2C buses visible with the nvidiafb driver, i2cdetect 
> reports only the SMBus.

Strange. Do you have i2c-algo-bit's option bit_test enabled by any
chance? "/sbin/modprobe -c | grep i2c.algo.bit" should tell.

> Here's a snippet from the X log, though:
> 
> (II) Loading sub module "i2c"
> (II) LoadModule: "i2c"(II) Module already built-in
> (II) Loading sub module "ddc"
> (II) LoadModule: "ddc"(II) Module already built-in
> (II) NV(0): I2C bus "DDC" initialized.
> (II) NV(0): Probing for analog device on output A...
> (--) NV(0):   ...found one
> (II) NV(0): Probing for analog device on output B...
> (--) NV(0):   ...can't find one
> (II) NV(0): Probing for EDID on I2C bus A...
> (II) NV(0): I2C device "DDC:ddc2" registered at address 0xA0.
> (II) NV(0): I2C device "DDC:ddc2" removed.
> (II) NV(0):   ... none found
> (II) NV(0): Probing for EDID on I2C bus B...
> (II) NV(0): I2C device "DDC:ddc2" registered at address 0xA0.
> (II) NV(0): I2C device "DDC:ddc2" removed.
> (II) NV(0):   ... none found
> (--) NV(0): CRTC 0 appears to have a CRT attached
> (II) NV(0): Using CRT on CRTC 0
> 
> Address 0xA0?  Should that be 0x50?

0xA0 == 0x50 << 1. I2C addresses are 7-bit values, but sometimes
software or documentation shifts them by one bit to the left because
that's how they go on the wire.

The important thing in the log above is that the nv X driver doesn't
see any EDID EEPROM either.

> > * Did you try another video cable? The I2C bus signal is carried over 2
> > dedicated pins on the VGA connector (12 and 15), if your cable happens
> > to not have them wired properly for some reason, this could explain
> > your problem.  
> 
> OK, I'll have to ask around for one.  Thanks for your help.

If you have another computer at hand which is known to retrieve the
EDID properly, you might also try swapping the monitors. This would
tell you right away whether the graphics adapter (or its driver), the
VGA cable, or the monitor itself is responsible for the problem.

-- 
Jean Delvare

_______________________________________________
i2c mailing list
i2c-GZX6beZjE8VD60Wz+7aTrA@public.gmane.org
http://lists.lm-sensors.org/mailman/listinfo/i2c