Hello, Camaleón and Salvatore. I have a similar problem with the TTYs and am very glad to have found this message. Hopefully, my reply can make it to the list. . The offending commit was fairly easy to find as will be detailed later. The git bisect on the linux-stable repo found the following commit as the problem: . ------------ First commit to cause the problem ------------ commit 0998a6cb232674408a03e8561dc15aa266b2f53b Author: Junjie Cao <[email protected]> Date: Mon Oct 20 21:47:01 2025 +0800
fbdev: bitblit: bound-check glyph index in bit_putcs* commit 18c4ef4e765a798b47980555ed665d78b71aeadf upstream. ------------ First commit to cause the problem ------------ . I leave it as an exercise to the reader to solve the problem created by this commit. Hopefully, soon. . The procedure to identify this commit was as follows. First the problem occurred only after installing the 5.10.247 kernel from the mentioned repo. The 5.10.246 release did not have the problem. So the 5.10.246 release is a known good commit. It was also noticed that there were several commits labelled as fbdev soon after the latter release. It was verified that commit ef226cacf564 had the problem, see below. The bisect then only had to work through about a dozen commits. Far easier than between 5.10.244 and 5.10.247. . It should be possible to verify that this commit is the issue with just two builds, the discovered commit ( 0998a6cb232674 ) which has the issue and the immediate prior commit, which should not have the issue. . Hopefully, this problem can be resolved easily. I am not familiar with this area of the code and font handling, so no suggestion. Sorry. . ----------- One of several commits labelled fbdev with the problem ------------- commit ef226cacf564f06e52a3b2a04b4924de5e22bca6 (HEAD) Author: Miaoqian Lin <[email protected]> Date: Mon Oct 27 16:43:37 2025 +0800 fbdev: valkyriefb: Fix reference count leak in valkyriefb_init ------------------------ . . Thanks for your time and Happy Holidays, Willam Burrow. . ---------------Previous message -------------------- > Got it. > > At the moment I am very busy at work, so I cannot have the time to test them > all. > > After Dec 23th I will do the tests on my home machine. Basically it's the > same OS and kernel and same hardware, but a little older. > > AFAIK the issue is in the production machine, that's a different hardware at > all (it's a MiniPC against a Dell Precision 3630 Workstation) > > Maybe both have an integrated Intel Graphic Card. > > I will check later. Ack, for now we put ourself then in wait position unless someone else affected can do the same bisect procedure and narrow down the problematic commit earlier. Thank you, Regards, Salvatore
