On Jul 4, 2011, at 8:08 PM, Mike Frysinger wrote:

> in general, this seems to go in a good direction.  certainly better than 
> previous attempts.  there are some issues with this patch though.
> 
> can you split up independent issues into independent patches ?  

Yes, I should have done that to begin with.

> the 
> flashverify command stuff for example should be a patch by itself (and 
> further 
> discussed since i'm pretty sure it breaks existing behavior).  

As you said, this should have been done separately, however I will give a quick 
overview of what I did:
The "flashmem" command as it is in the codebase programs the flash and then 
verifies it.  I made the verify step optional using the command line parameter 
"noverify", and added an additional "flashverify" command. 
I don't think it breaks any existing behavior as everything behaves the same as 
the codebase if you don't add the "noverify" parameter to flashmem.

> then have one 
> patch which simply adds the new stub enable/disable/write_start/cls fields.  
> then one which adds the spi bus driver, and finally one which adds the spi 
> flash driver.
I assume I can just submit each of these patches to this same thread?

> 
> the style in a bunch of new code is slightly broken:
> - dont use tabs ... use 4 spaces to indent
> - no trailing whitespace on lines
> - no trailing new lines in files
> - func calls are like:
>       foo (arg, arg, arg);
> - dont use inline var decls like:
>       for (int i .......
> - use space in between operators:
>       i == 0
>       i >= 0
> you call printf() in a bunch of places.  these all need to be changed to 
> urj_log() and such.

No problem cleaning up the formatting.

> with the spi bus driver, if the 3 main pins arent specified, then just try to 
> grab pins with the relevant name.  so if people dont specify MOSI, then 
> default to the signal name "MOSI".
Good idea.  

> i'm thinking we should abstract out the cfi structures a bit rather than 
> trying to stick spi flash info into cfi structs.
Well, some SPI flashes can provide much of the same identification and geometry 
information as a CFI.  So, it seemed like a good idea to use the same common 
structs.  That said, I have not implemented any code to read that information 
because the specific device I was working with at the time did not provide this 
information.  It also just seemed easier as it didn't break any other code by 
using this struct.

> 
> i think you changed the binary verify code to only work with msbin's.  the 
> current logic treats files as raw binaries.
I don't think I broke anything here; however, I don't have any msbin files to 
test with.  Everything I work with is all raw binary.  What I did was pull all 
of the verification code out of urf_flashmsbin and created a function called 
urj_flashverifymsbin.  Then I pulled all of the verification code out of 
urj_flashmem and created a function called urj_flashverify.

> 
> and of course, by now, the code will need to be rebased onto the current code 
> base as a bunch of things have changed.
No problem.  I try to keep everything up to date anyway.

> -mike

Attachment: smime.p7s
Description: S/MIME cryptographic signature

------------------------------------------------------------------------------
All of the data generated in your IT infrastructure is seriously valuable.
Why? It contains a definitive record of application performance, security 
threats, fraudulent activity, and more. Splunk takes this data and makes 
sense of it. IT sense. And common sense.
http://p.sf.net/sfu/splunk-d2d-c2
_______________________________________________
UrJTAG-development mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/urjtag-development

Reply via email to