Hi Tomi, > Am 20.04.2017 um 11:53 schrieb Tomi Valkeinen <tomi.valkeinen@xxxxxx>: > > On 20/04/17 12:00, H. Nikolaus Schaller wrote: > >> This fixes the issue equally well as reverting the offending patch on my GTA04A4 board with 4.11-rc7. >> >> But I still have some debugging messages when Xorg starts (omap-drm driver), that were not in kernels up to 4.10. >> I am not sure if they are related or something else: >> >> [ 52.767211] ------------[ cut here ]------------ >> [ 52.772521] WARNING: CPU: 0 PID: 2260 at drivers/gpu/drm/omapdrm/omap_gem.c:938 omap_gem_put_paddr+0x30/0x60 [omapdrm] > > There is, and afaics, has never been, a WARN in omap_gem_put_paddr. What > does the code look like? Which patch adds the WARN? I have looked for it and it was introduced to our code by http://git.goldelico.com/?p=gta04-kernel.git;a=commitdiff;h=6d7accfd86ce89f5c786cae5bc0005d8a7ae7e98 It replaces if (omap_obj->paddr_cnt == 0) { by if (!WARN_ON(omap_obj->paddr_cnt == 0)) { Unfortunately without any explanation why this is a warning or needed or a problem (and I must admit that I don't understand the code at all because I haven't spent time for it). So it may simple be something that does not need a warning at all. To be fair, Matthijs has declared this patch set as purely experimental, so it is my fault to use it. Especially on OMAP3 where it is not really designed for. I just forgot that this part is not upstream code. So this bug report is a false alarm for upstream. But the more general question is how to get his TILER patches [1] upstream (after fixing such issues). The TILER seems to work well now on OMAP5 and by Andreas' patch it does not harm on OMAP3. I haven't tried OMAP4. But the comments indicate that there is still some work needed to polish it for upstream. Who would be interested in helping such work? BR, Nikolaus [1]: http://git.goldelico.com/?p=gta04-kernel.git;a=shortlog;h=refs/heads/extern/mvduin/patch/tiler-fbdev-v2
Attachment:
signature.asc
Description: Message signed with OpenPGP using GPGMail
