On Tuesday, August 24, 2010 14:36:43 David Jander wrote:
On Tuesday 24 August 2010 06:06:26 pm Mike Frysinger wrote:
On Tuesday, August 24, 2010 02:39:16 David Jander wrote:
On Monday 23 August 2010 06:31:26 pm Mike Frysinger wrote:
On Monday, August 23, 2010 09:12:16 David Jander wrote:
- {
.idcode1 = SPSN_ID_S25FL032A,.idcode2 = SPSN_EXT_ID_S25FL032P,.idmask2 = 0xff00,what does the idcode2 look like such that you need a mask ?
According to the datasheet the RDID command (0x9f) returns the following bytes:
byte 0: Manufacturer ID = 0x01 byte 1,2: Device ID = 0x02, 0x15 (same as S25FL032A) byte 3: Extended ID = 0x4d byte 4,5,6: Spansion reserved (do not use).
byte 4 is read as 0x00 on my part, but due to the above explaination, I cannot say for sure it is always the same, so I had to implement a mask to check for it.
i'd rather we delay adding code to support something that may never change. so drop the whole idmask2 stuff and wait for it to become an actual problem
I agree that chances this ever breaks might seem rather tiny, but if it ever does break, waiting for it to happen could trigger a much bigger problem than it is to add these few lines; in the worst case, in some distant future, some boards will just not work for no apparent reason (if spansion decided to do something with byte 4 without notifying), and nobody will remember this discussion anymore...
considering the problem is rather minute and easily detected, i dont think it'll be that big of a deal
OTOH, you decide. It's ok with me if you want to leave it out. Given the fact that you had already accepted this patch, should I send a new version (without the mask)?
it's in a branch of mine that i can throw away, so just send a new patch based on current mainline and i'll integrate it. thanks ! -mike