Hello all,
First thanks to Uwe for nic3com support and to Mats Erik (IIRC) for helping with (or atleast pointing out the lack of) 3C905B support - i just tested a SST 29EE010 on an 3C905B-TXNM that had been useless for a while - works :)
Anyways, does anyone know how big flash chips these 3com's support? The SST i tested is 128k - i can propably (given time) test 256k also, but just curious.
And also, if i plan on using this for "hotflashing" a chip from another box to do my various hacks, it would be very much indeed too easy to accidentally write flashrom -w <file> instead of flashrom -p nic3com -w <file> . I'm not 100% sure whether this is a good idea (some might find it annoying), but i suggest that in case of "flashrom -w" or "flashrom -E" it would ask for sure, eg. "This will erase/write your INTERNAL flash chip. Are you sure (y/n)?". If given flashrom -p internal -w (or -E) it would not ask (for scripting).
Urja Rannikko wrote:
Anyways, does anyone know how big flash chips these 3com's support? The SST i tested is 128k - i can propably (given time) test 256k also, but just curious.
The theoretical limit is 512k, which would use all of the 32 PLCC leads.
But it really depends on how many address lines are connected from the NIC to the flash chip.
And also, if i plan on using this for "hotflashing" a chip from another box to do my various hacks, it would be very much indeed too easy to accidentally write flashrom -w <file> instead of flashrom -p nic3com -w <file> . I'm not 100% sure whether this is a good idea (some might find it annoying), but i suggest that in case of "flashrom -w" or "flashrom -E" it would ask for sure, eg. "This will erase/write your INTERNAL flash chip. Are you sure (y/n)?". If given flashrom -p internal -w (or -E) it would not ask (for scripting).
Interesting usability problem. Maybe an environment variable could be an alternative way of choosing what will be used to do flashing?
//Peter
On Fri, May 22, 2009 at 03:45:02PM +0200, Peter Stuge wrote:
Urja Rannikko wrote:
Anyways, does anyone know how big flash chips these 3com's support? The SST i tested is 128k - i can propably (given time) test 256k also, but just curious.
The theoretical limit is 512k, which would use all of the 32 PLCC leads.
But it really depends on how many address lines are connected from the NIC to the flash chip.
Yep. As per datasheet, you have 17 address lines for the B and also C variants of the cards, so a max. of 128KB chips.
Uwe.
2009/5/22 Urja Rannikko urjaman@gmail.com
Hello all,
First thanks to Uwe for nic3com support and to Mats Erik (IIRC) for helping with (or atleast pointing out the lack of) 3C905B support - i just tested a SST 29EE010 on an 3C905B-TXNM that had been useless for a while - works :)
Anyways, does anyone know how big flash chips these 3com's support? The SST i tested is 128k - i can propably (given time) test 256k also, but just curious.
Not that I have a 905B, but: http://www.hollants.com/pxerom_3c905.html http://www.infracity.de/pxe/ http://www.shikadi.net/hwwiki/Flash_chips
And also, if i plan on using this for "hotflashing" a chip from another box to do my various hacks, it would be very much indeed too easy to accidentally write flashrom -w <file> instead of flashrom -p nic3com -w <file> . I'm not 100% sure whether this is a good idea (some might find it annoying), but i suggest that in case of "flashrom -w" or "flashrom -E" it would ask for sure, eg. "This will erase/write your INTERNAL flash chip. Are you sure (y/n)?". If given flashrom -p internal -w (or -E) it would not ask (for scripting).
-- urjaman
-- coreboot mailing list: coreboot@coreboot.org http://www.coreboot.org/mailman/listinfo/coreboot
Idwer Vollering wrote:
This links to http://ctflasher.sourceforge.net/ which has drivers for programming flash chips on Intel, Realtek and VIA NICs as well.
//Peter
On Fri, May 22, 2009 at 03:51:45PM +0200, Peter Stuge wrote:
Idwer Vollering wrote:
This links to http://ctflasher.sourceforge.net/ which has drivers for programming flash chips on Intel, Realtek and VIA NICs as well.
Indeed, and I have patches pending for Realtek, DEC, and VIA NICs already, but for some reasons they don't seem to work yet (in contrast to the 3com code), I'm in the process of finding out if it's a hardware issue (wrong type of chip, missing NIC functionality) or me doing some stupid things in the code.
Will send patches soonish.
Uwe.
Urja Rannikko wrote:
Hello all,
First thanks to Uwe for nic3com support and to Mats Erik (IIRC) for helping with (or atleast pointing out the lack of) 3C905B support - i just tested a SST 29EE010 on an 3C905B-TXNM that had been useless for a while - works :)
Anyways, does anyone know how big flash chips these 3com's support? The SST i tested is 128k - i can propably (given time) test 256k also, but just curious.
And also, if i plan on using this for "hotflashing" a chip from another box to do my various hacks, it would be very much indeed too easy to accidentally write flashrom -w <file> instead of flashrom -p nic3com -w <file> . I'm not 100% sure whether this is a good idea (some might find it annoying), but i suggest that in case of "flashrom -w" or "flashrom -E" it would ask for sure, eg. "This will erase/write your INTERNAL flash chip. Are you sure (y/n)?". If given flashrom -p internal -w (or -E) it would not ask (for scripting).
Yes I'm for this too. Moreover I would like to have Are you sure[y/N] dialog if -p internal is not specified...
Otherwise it would mean someone will flash the PXE as main BIOS (yes its unlikely because of different size of flashes but it may happen one day)
Rudolf
-- urjaman
-
Idwer Vollering -
Peter Stuge -
Rudolf Marek -
Urja Rannikko -
Uwe Hermann