SST Flash Support (SST39VF160)

SST Flash Support (SST39VF160)

[Stefan Roese]

Hi,

I am right now struggeling with mtd and sst flash chips (39VF160) support.
The amd devices (am29lv160) work without problems.

My question: The file "amd_flash.c" contains some defines for SST flash
chips, but they are never used. Is this because some problems occured, or
because nobody had the time to fully implement the sst support.

Best regards,
Stefan.

Re: SST Flash Support (SST39VF160)

[David Woodhouse]

stefan.roese@esd-electronics.com said:
>  I am right now struggeling with mtd and sst flash chips (39VF160)
> support. The amd devices (am29lv160) work without problems.

SST 39VF160 chips are CFI-compliant. You shouldn't need any special code 
for them -- just use cfi_probe.

--
dwmw2

Re: SST Flash Support (SST39VF160)

[Stefan Roese]

> SST 39VF160 chips are CFI-compliant. You shouldn't need any special code
> for them -- just use cfi_probe.

That's what I tried too. Here is what I get when I use the sst flash's (all
debug info turned on):

>>>
physmap flash device: 200000 at ffc00000
CFI: Found no Physically mapped flash device at location zero
<<<

On the amd flash board I get the following debug messages:

>>>
physmap flash device: 200000 at ffc00000
Number of erase regions: 4
Primary Vendor Command Set: 0002 (AMD/Fujitsu Standard)
Primary Algorithm Table at 0040
Alternative Vendor Command Set: 0000 (None)
No Alternate Algorithm Table
Vcc Minimum: 2.7 V
Vcc Maximum: 3.6 V
No Vpp line
Typical byte/word write timeout: 16 ?s
Maximum byte/word write timeout: 512 ?s
Full buffer write not supported
Typical block erase timeout: 1024 ?s
Maximum block erase timeout: 16384 ?s
Chip erase not supported
Device size: 0x200000 bytes (2 MiB)
Flash Device Interface description: 0x0002
  - supports x8 and x16 via BYTE# with asynchronous interface
Max. bytes in buffer write: 0x1
Number of Erase Block Regions: 4
  Erase Region #0: BlockSize 0x4000 bytes, 1 blocks
  Erase Region #1: BlockSize 0x2000 bytes, 2 blocks
  Erase Region #2: BlockSize 0x8000 bytes, 1 blocks
  Erase Region #3: BlockSize 0x10000 bytes, 31 blocks
kmod: failed to exec /sbin/modprobe -s -k cfi_cmdset_0002, errno = 2
Support for command set 0002 not present
cfi_probe: No supported Vendor Command Set found
<<<


Any thoughts/suggestions?

Thanks a lot,
Stefan.

Re: SST Flash Support (SST39VF160)

[David Woodhouse]

stefan.roese@esd-electronics.com said:
> physmap flash device: 200000 at ffc00000 CFI: Found no Physically
> mapped flash device at location zero

Weird. That chip is definitely supposed to do CFI. 
http://www.ssti.com/products/pdf/399-39LF-VF160-02.000-DS.pdf

Sure it's wired up the same and your configured buswidth is right?

stefan.roese@esd-electronics.com said:
> On the amd flash board I get the following debug messages: 
> kmod: failed to exec /sbin/modprobe -s -k cfi_cmdset_0002, errno = 2
> Support for command set 0002 not present
> cfi_probe: No supported Vendor Command Set found

CONFIG_CFI_AMDSTD=y

--
dwmw2

Re: SST Flash Support (SST39VF160)

[Stefan Roese]

> stefan.roese@esd-electronics.com said:
> > physmap flash device: 200000 at ffc00000 CFI: Found no Physically
> > mapped flash device at location zero
>
> Weird. That chip is definitely supposed to do CFI.
> http://www.ssti.com/products/pdf/399-39LF-VF160-02.000-DS.pdf
>
> Sure it's wired up the same and your configured buswidth is right?

Absolutely sure. It's the same board!

After digging a little deeper into the sst documentation, I found the
problem:

SST flash need a three byte write sequence (no one), to enter the cfi query
mode!

5555 <- aa
2aaa <- 55
5555 <- 98

This does not seem to be CFI compliant!!!

How can we support the sst chips? Extend the cfi code (special for sst), or
extend the amd code? In PPCBoot we use the amd mode and it works pretty
good.

Any ideas/comments?

Stefan.

Re: SST Flash Support (SST39VF160)

[David Woodhouse]

stefan.roese@esd-electronics.com said:
> SST flash need a three byte write sequence (no one), to enter the cfi
> query mode!

> 5555 <- aa 2aaa <- 55 5555 <- 98

 <insert relevant expletives here>

> This does not seem to be CFI compliant!!! 

OK. Add the device ID and size information to jedec_probe.c. That code 
should handle the 55/aa bits already.

--
dwmw2

Re: SST Flash Support (SST39VF160)

[Stefan Roese]

> > This does not seem to be CFI compliant!!!
>
> OK. Add the device ID and size information to jedec_probe.c. That code
> should handle the 55/aa bits already.

OK, I added the defines in jedec_probe.c, but I can't get linux to use the
"jedec_probe_chip()" routine. Even with further debug code, this routine
doesn't seem to get called :-(.

Any ideas?

Best regards,
Stefan.

Re: SST Flash Support (SST39VF160)

[David Woodhouse]

stefan.roese@esd-electronics.com said:
>  OK, I added the defines in jedec_probe.c, but I can't get linux to
> use the "jedec_probe_chip()" routine. Even with further debug code,
> this routine doesn't seem to get called :-(. 

Are you calling do_map_probe("jedec_probe",..) from your map driver?

--
dwmw2