Re: [flashrom] Winbond W25Q64FVSIG - "ERASE FAILED!"

[Nico Huber]

On 17.09.2017 03:51, junkheap(a)tiscali.co.uk wrote:
  Thank you for the help Nico.
 
 Your advice was successful. After connecting /HOLD & /WP to 3.3V,
 Flashrom wrote & verified the W25Q64SIG without any error messages.
Nice!

  
 After previously checking the datasheet I thought that it was necessary
 to do that. But I have little experience of electronics and thought it
 wise to seek advice before trying, as all of the guides to programming
 EEPROMs with Raspberry Pi I found on the web had always left those pins
 unused.
Most guides probably assume that the chip is still soldered to a board.
In that case the circuit on the board pulls /HOLD and /WP already.

Nico

PS. Please always keep the mailing list in CC when you respond. Some
    email clients have an extra button (e.g. "Reply all") for that.

  
 
 On 16/09/2017 13:57, Nico Huber wrote:
  Hi,

 On 15.09.2017 18:56, junkheap--- via flashrom wrote:
  I am trying to write to a Winbond W25Q64FVSIG
EEPROM (SOIC-8) using
 Flashrom and a Raspberry Pi 3 (running up to date Raspbian). But I
 receive the message "ERASE FAILED!".

 I am using the shortest wires which are practical to connect to the
 Raspberry Pi (~10CM) and I have checked many times that I have connected
 to the correct GPIO pins. To make sure that there is a good connection,
 I abandoned the clip I was using and have soldered the EEPROM to the
 pads on a daughterboard (designed for the purpose). The wires I am using
 have factory-fitted female terminals on both ends and connect tightly to
 the GPIO pins on the RPI and pin head on the daughterboard.

  From what I have found some people have flashed this model of EEPROM
 successfully using Flashrom, but others have encountered the same
 problem. I have not been able to find a solution. If anyone can offer
 assistance or advice I would me most grateful. Thank you.
 I see two possible
causes:

 1. You probably didn't connect all pins of the flash chip. You have to
     pull /HOLD and /WP high (connect them to 3V3, preferably through some
     resistor) to succeed.

 2. You run the SPI interface at a too high clock. As you didn't supply
     a `spispeed` argument to the linux_spi programmer, it will at run
     at some (to me unknown) default speed.

 Nico