LWN: Comments on "New stable kernels" https://lwn.net/Articles/677653/ This is a special feed containing comments posted to the individual LWN article titled "New stable kernels". en-us Mon, 08 Sep 2025 21:25:00 +0000 Mon, 08 Sep 2025 21:25:00 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net New stable kernels https://lwn.net/Articles/679035/ https://lwn.net/Articles/679035/ nix <div class="FormattedComment"> Confirmed. The problem is somewhere in the 16-bit setup code, I think, though I don't understand why 2.26's generated code is breaking things yet: it looks like an improvement all around, and looks LX-compatible (but, from the behaviour observed, pretty clearly isn't).<br> <p> Will figure it out and raise a binutils bug report :(<br> <p> </div> Sun, 06 Mar 2016 20:18:19 +0000 New stable kernels https://lwn.net/Articles/678958/ https://lwn.net/Articles/678958/ nix <div class="FormattedComment"> OK, perhaps not. Either there are half a dozen commits between 4.4.2 and 4.4.3 alone with the same "nothing shown at all, gotta power off" symptoms (which is what bisection says), *none* of which affect anything else even though most of them are in generic code fairly far removed from the serial console, or the recent upgrade here from binutils 2.25 to 2.26 broke something. Probably there are non-Geode-LX instructions finding their way into the instruction stream. I have working and non-working examples, so figuring out what's going on is only a matter of time...<br> <p> </div> Sat, 05 Mar 2016 20:42:19 +0000 New stable kernels but no more penguins https://lwn.net/Articles/678424/ https://lwn.net/Articles/678424/ sgilles I ran into this <a rel="nofollow" href="https://www.marc.info/?l=linux-kernel&m=142249878826406&w=4">as well</a>. If you're unconcerned about the potential memory safety issues, I can confirm that <pre> diff --git a/drivers/video/logo/logo.c b/drivers/video/logo/logo.c index 10fbfd8..5fec3f5 100644 --- a/drivers/video/logo/logo.c +++ b/drivers/video/logo/logo.c @@ -44,7 +44,7 @@ const struct linux_logo * __init_refok fb_find_logo(int depth) { const struct linux_logo *logo = NULL; - if (nologo || logos_freed) + if (nologo) return NULL; if (depth &gt;= 1) { </pre> will bring back the penguin. Wed, 02 Mar 2016 19:46:55 +0000 New stable kernels but no more penguins https://lwn.net/Articles/678054/ https://lwn.net/Articles/678054/ pr1268 <p>Along a similar vein, I upgraded from 3.14.x to 4.4.x and now the penguins disappeared. Not sure what happened, but I do have i915 chipset, 1600×900 resolution with the framebuffer code and the 224-color boot logo built-in (not modules). I tried in vain searching various Linux-specific forums but these all seem to be frozen in time at 2011 or thereabouts.</p> <p>Granted, the <tt>make menuconfig</tt> options under device drivers have changed slightly, 3.14 → 4.4 is a big jump, and seeing how I run X, this is a moot point. But, I miss my penguins. ;-(</p> <p>Any quick ideas? Thanks!</p> Tue, 01 Mar 2016 07:48:49 +0000 New stable kernels https://lwn.net/Articles/678013/ https://lwn.net/Articles/678013/ nix <div class="FormattedComment"> As an aside, it seems likely to me that 4.4.3 breaks the serial console. (That's the best explanation I have for the last thing I see on the serial console being BIOS messages, followed by nothing and no userspace ever starting. Oh look, TTY-layer locking changes...)<br> <p> I'll bisect this weekend.<br> </div> Mon, 29 Feb 2016 22:53:34 +0000