00:15 <alfayt[m]> Btw what happened to framebuffer on sagami?

11:50 <Marijn[m]> alfayt: if you know...

11:51 <Marijn[m]> alfayt: Sorry I mixed up 5III and 5IV. 5III is known-working, 5IV has a panel driver that I haven't yet generated/fixed but it shouldn't have anything out of the ordinary.

11:51 <Marijn[m]> alfayt: 1III however has the exact same panel as 1IV, so the exact same ~problems~ TODO items.

11:52 <Marijn[m]> (For everyone else: 1IV/5IV is Sony's Nagara platform, 1III/5III is Sony's Sagami platform)

18:19 <mort_> okay soo https://elixir.bootlin.com/linux/v6.1.34/source/drivers/gpu/drm/drm_atomic_helper.c#L2241 this drm commit struct never gets freed

18:21 <mort_> every frame, that kzalloc leaks 256 bytes from the kmalloc-256 slab allocator

18:23 <mort_> I've added prints to the drm_crtc_commit_get/drm_crtc_commit_put functions, and for each allocated commit, the refcount jumps up and down a bit before reaching 1 and staying there

18:24 <mort_> https://p.mort.coffee/DVX here's the stack trace for every get and put for a particular commit

18:25 <mort_> I don't know that I'm capable of debugging this on my own...