Adjust language around what a device is#60
Conversation
|
ping? |
robherring
left a comment
There was a problem hiding this comment.
I feel this is documenting the exceptions as the norm which is not something I think we should encourage.
| probing. | ||
| A device in this context may be an actual hardware device, such as a UART. It | ||
| may be part of a hardware device, such as the random-number generator in a TPM. | ||
| It may also be a purely software concept, such as an bridge providing access to |
There was a problem hiding this comment.
'purely software concept' is vague and leaves the door open to anything.
I am assuming what you mean here is interfaces implemented firmware which provide access to h/w? I'd prefer something like 'A device may include functions implemented by firmware running in higher privilege levels or remote processors'.
There was a problem hiding this comment.
Yes that is a bit vague. See what you think of this version
| may be part of a hardware device, such as the random-number generator in a TPM. | ||
| It may also be a purely software concept, such as an bridge providing access to | ||
| an I2C device attached to a remote CPU. There is no requirement that a node in | ||
| a device tree be a physical hardware device. |
There was a problem hiding this comment.
No requirement as there are some exceptions, but generally nodes have some correlation to physical h/w devices.
Bindings which are designed 1:1 nodes to $OS drivers are a constant headache. We can't design bindings which cater to just $OS's abstraction (at the current time).
There was a problem hiding this comment.
OK I tried with some different words
The current definition is causing people to decide that only 'real' hardware devices should be in the device tree. Rewrite two paragraphs to adjust this. Signed-off-by: Simon Glass <sjg@chromium.org>
The current definition is causing people to decide that only 'real' hardware devices should be in the device tree. Rewrite two paragraphs to adjust this.
Signed-off-by: Simon Glass sjg@chromium.org