LWN: Comments on "Why printk() is so complicated (and how to fix it)" https://lwn.net/Articles/800946/ This is a special feed containing comments posted to the individual LWN article titled "Why printk() is so complicated (and how to fix it)". en-us Sun, 19 Oct 2025 11:17:43 +0000 Sun, 19 Oct 2025 11:17:43 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801579/ https://lwn.net/Articles/801579/ KaiRo <div class="FormattedComment"> You already suggest how the story will continue: one person with a clean soul will need to be found to carry the one ring (buffer) to the source, er, I mean, the big Vulcan of evil, to finally burn it fire and free the world...<br> </div> Tue, 08 Oct 2019 00:41:50 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801429/ https://lwn.net/Articles/801429/ bene42 <div class="FormattedComment"> I am not aware that there are more competing multiple reader/multiple writer ringbuffer implementations out there.<br> A link would be very helpful. Maybe there is a simpler solution, since the new one is quite complex and hard to understand...<br> </div> Mon, 07 Oct 2019 13:10:49 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801426/ https://lwn.net/Articles/801426/ ncultra <div class="FormattedComment"> The virtio ring buffer is most frequently used as a ring of pointers to other buffers. This wreaks havoc with performance as the virtual machine monitor is constantly having to (re) map the pointed-to buffer addresses to host-physical from guest-physical. Just something to consider. The newer virtio spec has a good solution to this (anti) feature, but it is only recently in the mainline kernel, so most distributions still have to deal with what virtio calls indirect buffers. It's easy to overlook the performance impact of "indirect" memory access and function calls.<br> </div> Mon, 07 Oct 2019 12:22:39 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801333/ https://lwn.net/Articles/801333/ valarauca <div class="FormattedComment"> Situation: There are 5 different competing ring buffers<br> <p> <font class="QuotedText">&gt; Maybe we need one ring (buffer) to rule them all?</font><br> <p> Situation: There are 6 different competing ring buffers<br> <p> credit: <a href="https://xkcd.com/927/">https://xkcd.com/927/</a><br> <p> <p> </div> Fri, 04 Oct 2019 19:55:10 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801325/ https://lwn.net/Articles/801325/ zlynx <div class="FormattedComment"> This One Ring Buffer would of course be a ring of pointers to every other ring buffer.<br> </div> Fri, 04 Oct 2019 16:37:46 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801290/ https://lwn.net/Articles/801290/ shalem <div class="FormattedComment"> Maybe we need one ring (buffer) to rule them all?<br> </div> Fri, 04 Oct 2019 14:19:50 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801283/ https://lwn.net/Articles/801283/ smurf <div class="FormattedComment"> Yay for even more ring buffers.<br> </div> Fri, 04 Oct 2019 11:38:41 +0000 Why printk() is so complicated (and how to fix it) https://lwn.net/Articles/801274/ https://lwn.net/Articles/801274/ pmladek <div class="FormattedComment"> Just for the record, the alternative implementation of the ring buffer has been heavily inspired by the original one. I just had one or two ideas how to make it easier and it seems that they worked.<br> </div> Fri, 04 Oct 2019 07:29:05 +0000