May disk's driver type need more checks

[nevermore-muyi]

As the code https://github.com/jaypipes/ghw/blob/main/pkg/block/block_linux.go#L271 shows, when I create a device mapper device which starts with dm- prefix, and the function https://github.com/jaypipes/ghw/blob/main/pkg/block/block_linux.go#L316 thinks it's a unknown device, but (https://github.com/jaypipes/ghw/blob/main/pkg/block/block_linux.go#L274) may define it's a SSD.

So, in my option, driver type may need more checks.

[jaypipes]

@nevermore-muyi hi! thanks for your submission!

Are you suggesting that instead of returning an UNKNOWN disk type, we should check for whether the disk is rotational, and if not, return SSD disk type?

[nevermore-muyi]

@jaypipes Maybe a little different. I suggest only KNOWN disk type needs more check if it's rotational and define it's a SSD, but UNKNOWN disk type, just return it?

[ffromani]

I tend to agree the code should not set the drive type depending only on the value of rotational

[jaypipes]

@nevermore-muyi we are considering creating a separate package for devicemappings/devicemapper. See conversation here.

[nevermore-muyi]

@nevermore-muyi we are considering creating a separate package for devicemappings/devicemapper. See conversation here.

Got it. Looks very cool.