LWN: Comments on "Monitoring the internal kernel ABI" https://lwn.net/Articles/800452/ This is a special feed containing comments posted to the individual LWN article titled "Monitoring the internal kernel ABI". en-us Mon, 08 Sep 2025 02:07:34 +0000 Mon, 08 Sep 2025 02:07:34 +0000 https://www.rssboard.org/rss-specification lwn@lwn.net Monitoring the internal kernel ABI https://lwn.net/Articles/800599/ https://lwn.net/Articles/800599/ Conan_Kudo <p>The conversation about kABIs in this article is pretty reasonable, but one thing I worry about is that Google acquiescing to creating a stabilized kABI for Android systems will actually further <i>reduce</i> the incentive for the Android ecosystem to contribute to the mainline kernel.</p> <p>There isn't <i>that much</i> incentive as it stands, admittedly, but it would be nice to see actions taken to <i>increase</i> incentives (maybe a stronger carrot and stick approach?) to contribute to mainline with FOSS drivers...</p> Thu, 26 Sep 2019 04:35:06 +0000 Monitoring the internal kernel ABI https://lwn.net/Articles/800572/ https://lwn.net/Articles/800572/ gbalane <div class="FormattedComment"> Good to know there is automation. This is done manually today by Red Hat and Oracle (UEK) kernel who maintains stable KABI by using whitelists. <br> There is also padding done for certain Kernel structures to make room for keeping stable ABI for the life of the kernel based on one stable tree release. <br> As is evident, the stable KABI is required for any out of tree kernel drivers. This an be reduced if vendors contribute back to upstream on a timely manner. <br> </div> Wed, 25 Sep 2019 18:04:30 +0000