[flashrom] Re: probing logic in command line vs libflashrom

[Peter Marheine]

It looks to me like the library only looks for up to two matches because it
returns an error if there's more than one match. I agree that it would be
good to make it able to report which chips matched (making
flashrom_flash_probe() suitable for CLI use) in addition to returning an
error, but the API would need to be extended in order to support that
additional information.

On Sun, Jan 26, 2025 at 10:59 PM Anastasia Klimchuk 
wrote:
> Hello everyone,
>
> It turns out, flashrom has two different code paths for probing from
> command line vs probing with libflashrom.
>
> I think the logic should be the same, so as a goal I want to
> understand which approach is correct and use the same in both places.
> Have one piece of code and call it in two places.
>
> What is the difference:
>
> Command line keeps probing until the end of flashchips array, and as a
> result all matches will be found (no matter how many of them are).
> Currently code is on these lines:
> https://github.com/flashrom/flashrom/blob/main/cli_classic.c#L1211
>
> Libflashrom only does probing twice, so at most two matches will be found.
> Currently code is on these lines:
> https://github.com/flashrom/flashrom/blob/main/libflashrom.c#L250
>
> These two paths were initially introduced like this (>14 and >7yr ago).
>
> To me, command line's approach makes sense - because why stop at two
> matches? What if there are three matching chip definitions, and #3 is
> actually what the user has?
>
> So I am interested in what people think.
>
> A context how did this come up: I started working towards giving
> SFDP-detection more priority and this involves updating the probing
> logic, and that's how I came to this.