Sorry for the delay in replying, I had to find access to another S2915-e, the old one was gone.
I attach the two log files as requested.

The test machine had an Opteron 2220 and 4Gb ram.
OS was debian unstable with kernel 2.6.32.

HTH,
Alex

Carl-Daniel Hailfinger wrote: 
Hi Phil,
hi Alessandro,

you both have the same problem.

can you retry with latest flashrom from svn? While it might not work
either, it will give us better insights into what's wrong with your system.

One idea would be to run

flashrom -V -c SST49LF080A

directly after a cold boot (no other flashrom runs before, make sure it
is really a cold boot). Please also run
superiotool -deV
(latest superiotool from svn) after that.

Just to rule out any other strangeness, can you both tell me about how
much RAM the machine has, the OS/distribution and kernel version with
which you tested flashrom? You can get the kernel version with
uname -a


On 20.01.2010 18:03, i15@ornl.gov wrote:
  
Some addtional info. when I force the tool to read using the chip name ,
it will dump the flash (though warning it may be garbage) and some
string comparison with the "original V101.ROM shows it is at least not
random.
  
    

Yes, that's expected.


  
sapienza:~/Bios #  ./flashrom -V -f -c SST49LF080A -r dump.bin
flashrom v0.9.1-r710
No coreboot table found.
Found chipset "NVIDIA MCP55", enabling flash write... OK.
This chipset supports the following protocols: Non-SPI.
Calibrating delay loop... 648M loops per second, 100 myus = 191 us. OK.
Probing for SST SST49LF080A, 1024 KB: Chip lacks correct probe timing
information, using default 10mS/40uS. probe_jedec: id1 0xff, id2 0xff,
id1 parity violation, id1 is normal flash content, id2 is normal flash
content
No EEPROM/flash device found.
    
      

I really hope we can solve this mystery. Hm. Could be a screwed up MTRR
setting, or we hit a bug in libpci, or there is a new bug in the MCP55
code (which is rather unlikely because it works for others).


Regards,
Carl-Daniel