On Fri, Mar 04, 2016 at 04:17:07PM -0800, Paul Rogers wrote: > > On ATI hardware, the screen stays blank while firmware is loaded - on > > my Kaveri that takes a lot longer than on my old R600 and my past R200 > > (but then it loads a lot more firmware). > > Ah, that's what it is, firmware. This is a Nvidia GeForce 6200. >
Up until this week, I thought that nvidia GPUs did not use separate firmware. So please, don't ask me about *details* on nouveau. > > Look at dmesg to see what size of framebuffer it has switched to. It > > sounds as if you perhaps have one extra character of line width in the > > framebuffer. > > Undoubtedly, but it's hard to figure something returning "81", or that > the kernel initialization uses "79". More likely an off by one > subtraction error somewhere. Grrrrr > > Ummmmmm, I've seen that different panels, ostensibly the same size, have > different numbers of pixels. Maybe it IS possible division by 80 and > roundoff comes up different. Grrrrr > > > I normally load a console font, and that can change the screen size. > > So, again on my Kaveri, the '[OK]' messages do not all line up. > > I haven't messed with fonts (yet). Are you off consistently by one to > the right after udev? > I think it is a bit more than one in my case - on the Kaveri I have a 1600x900 screen, and I load a 12x22 font in userspace - not sure if the kernel starts with 12x22, nor what screen size it thinks it has before the framebuffer. And again, AFAICS this has zero to do with udev. But generally, a screen has the specified number of pixels. Oh, hang on, you said you were forcing 640x480 - I suspect your real screen is bigger than that ? If it isn't, I doubt that Xorg will be usable. I would have thought 800x600 or 1024x768 would be more likely to give a sensible compromise between text size and available rows and columns. > > Does it matter ? My screens get cleared when I login, so the variable > > layout is not important to me. > > I've noticed login clears the screen in the last couple versions. I'd > rather not have something that appears to be a glitch obvious during > boot, it may diminish confidence. > > > If it does matter to you, I suppose you could try adding a series of > > your own "debug" bootscripts to report information from stty or > > whatever, and run them before and after the shift. Perhaps write to a > > logfile. > > Yeah, been thinking along those lines myself. Just thought I'd ask... > > Thanks. If you can spare the time to investigate that, I think you are lucky you don't have more urgent interests to pursue ;-) Me, I already spend too much time on systems: trying to check that an -rc kernel boots on all my desktops with their different configs (at least my bisection this week was fairly quick and got a very quick fix - who would have thought my Kaveri had a nutmeg ? LOL), updating for vulnerabilities, trying to build BLFS -rc in multiple variants, and then ideally finding time to do things which are actually *interesting*. But don't let me put you off: your system, your bootscripts, your learning experience if you take it. ĸen -- This email was written using 100% recycled letters. -- http://lists.linuxfromscratch.org/listinfo/lfs-support FAQ: http://www.linuxfromscratch.org/blfs/faq.html Unsubscribe: See the above information page Do not top post on this list. A: Because it messes up the order in which people normally read text. Q: Why is top-posting such a bad thing? A: Top-posting. Q: What is the most annoying thing in e-mail? http://en.wikipedia.org/wiki/Posting_style
