README: improve existing file

[dougsland]

Add more description and steps to users.

[ericcurtin]

I will do a detailed review of this in the next few hours, but thanks so much for this, on first initial skim the majority looks great!

[ericcurtin]

One thing worth adding, is we would regard this in alpha or beta stage, we've only tested on three types of storage device so far:

• virtio
• NVMe
• Qualcomm UFS specifically (there's a qualcomm specific driver for this)

Although it may work on other kinds of storage devices not covered here, I can't guarantee it yet. Don't want to be on the hook for helping debug unbootable machines. It's often easy to fix, but prevention right now is better than curing!

[dougsland]

One thing worth adding, is we would regard this in alpha or beta stage, we've only tested on three types of storage device so far:

• virtio
• NVMe
• Qualcomm UFS specifically (there's a qualcomm specific driver for this)

Although it may work on other kinds of storage devices not covered here, I can't guarantee it yet. Don't want to be on the hook for helping debug unbootable machines. It's often easy to fix, but prevention right now is better than curing!

Sure thing @ericcurtin. I am just waiting all feedback from you and @rhatdan to update this PR.

[ericcurtin]

No more feedback on my end @dougsland :smile:

[rhatdan]

LGTM

[ericcurtin]

More contributors :champagne:

I'm gonna expand this even more sometime to explain the scalability, maybe it's just me but I think that is just as important as the raw boot performance. Before this, it was hard to even consider using meatier technologies this early in boot because of the scalability issue of initramfs.

[rhatdan]

I agree, that was some of the stuff I was thinking of, now the initramfs can be huge, stick python and friends in there, we don't care.

[dougsland]

Thanks for the merge, more patches to come.