Messages in this thread | | | Date | Tue, 08 Apr 2014 17:42:00 +0200 | From | Levente Kurusa <> | Subject | Re: [RFC] QR encoding for Oops messages |
| |
Hi,
On 04/07/2014 05:20 PM, Jason Cooper wrote: > On Sat, Apr 05, 2014 at 11:11:02AM +0200, Levente Kurusa wrote: >> Or, we could use core_param and simply have 'oops_qr' or >> 'qr_oops'. In my humble opinion the latter sounds better. > > Ack. My original suggestion was focused on 0=disable, >0 is scale. I > literally pulled the name from my nether-regions. :-)
Pushed to Teodora. Hopefully she will pull it soon.
> >> Oh and another suggestion, I think placing it in the bottom-right >> corner would be better since then we wouldn't overwrite some of >> the timestamps and messages. > > The real text is still sent to the (hopefully written to disk) logs. If > a user (or distro) builds with this feature, I would think centered and > scaled for ease of scanning would be highest priority.
Yup, I'll be traveling on the train a lot this week, so I'll have plenty of time to implement scaling and centering. Maybe we could also implement this:
qr_oops=center (center the QR code with scale 1) qr_oops=center,3 (center the QR code with scale 3)
'center' could also be 'topleft', 'bottomright', etc. Or just remain at the KISS rule? (keep it simple)
Any objections?
> > I don't think there is a 'safe' part of the framebuffer real estate > where the QR could be written for all scenarios. Best to make it easy > to scan.
Yea we also need to prevent it from happening on panics. Currently on panics, (i.e. exit when init=/bin/sh) will cause half of the QR code not rendered on screen due to some reason. It looks like it is due to scrolling, but I am not sure then why doesn't happen when I do 'echo c > /proc/sysrq-trigger' which as well causes a panic.
Any ideas why this happens?
-- Regards, Levente Kurusa PGP: 4EF5D641
| |