Hi Daniel, I tested your branch " wip/2017-09/atomic-v12" on Intel Apollo Lake with: - Linux Kernel v4.9 - Libdrm 2.4.83 - Mesa 13.0.6
In the end, I could not test the atomic modesetting with this setup. I got this warning "DRM: does not support atomic modesetting". When I checked the source code, I saw that atomic modesetting depends on DRM_CAP_CRTC_IN_VBLANK_EVENT, which is introduced to the kernel with this commit "5db06a8a98f515f67446a69c57577c4c363ec65d" in 4.12 Linux kernel. I also saw a regression. After desktop-shell sets the display black, weston crashed with this callstack: Oct 04 02:25:39 gr-mrb-64 weston[1922]: weston: ../../../../../../../../weston/libweston/compositor.c:2474: weston_output_finish_frame: Assertion `output->repaint_status == REPAINT_AWAITING_COMPLETION' fa iled. Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] caught signal: 6 Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [0000000000405d2a] -- (/usr/bin/weston) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [0000000000405d97] -- (/usr/bin/weston) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [000000308aa330e0] -- (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [000000308aa3307f] gsignal (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [000000308aa344aa] abort (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [000000308aa2bed7] -- (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [000000308aa2bf82] -- (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [00007fe98acf909b] weston_output_finish_frame (/usr/lib64/libweston-3.so.0) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [00007fe98aaccaae] -- (/usr/lib64/libweston-3/drm-backend.so) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.784] [00007fe98aacd8e0] -- (/usr/lib64/libweston-3/drm-backend.so) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [00007fe98acf9241] -- (/usr/lib64/libweston-3.so.0) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [0000003093009849] -- (/usr/lib64/libwayland-server.so.0) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [0000003093009e02] wl_event_loop_dispatch (/usr/lib64/libwayland-server.so.0) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [00000030930087ba] wl_display_run (/usr/lib64/libwayland-server.so.0) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [0000000000405160] -- (/usr/bin/weston) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [000000308aa202c1] __libc_start_main (/lib64/libc.so.6) Oct 04 02:25:39 gr-mrb-64 weston[1922]: [02:25:39.785] [00000000004057aa] _start (/usr/bin/weston) Best regards Emre Ucan Engineering Software Base (ADITG/ESB) Tel. +49 5121 49 6937 > -----Original Message----- > From: wayland-devel [mailto:wayland-devel- > boun...@lists.freedesktop.org] On Behalf Of Daniel Stone > Sent: Dienstag, 26. September 2017 19:20 > To: wayland-devel@lists.freedesktop.org > Subject: [PATCH weston v12 00/40] Atomic modesetting > > Hi, > Here's a cleaned-up and fully-tested version of the atomic series. I'm > fairly happy with it now. There are a few things I'd want to change, but > figured I'd send it all out now rather than waiting another however > long: > * I'd rather track unused_{crtcs,connectors} by putting them all in > originally, and only removing them when an enabled output state > is committed from repaint_flush > * Adding unused_planes would also be smart (actually, I see it's > still #if 0'ed out in the atomic patch; for shame) > * The property handling for unused_* is not overly pretty, and I'd > happily take suggestions on how to fix it > > Compared to v10/v11, here are the major changes: > * A temporary commit allows disabling universal-plane/atomic support > through an environment variable, allowing easier testing of those > paths > * Cursor/scanout planes are now discovered at plane creation, rather > than enable, time > * dmabuf import now uses GBM_BO_IMPORT_FD_MODIFIER, rather than > using > EGLImage import, to avoid handle-uniqueness issues > * drmModeGetPlane2 is now spelled IN_FORMATS > * Bugfixes for viewported/scaled client buffers (tested with a cropped > and scaled simple-egl) > > Cheers, > Daniel > > _______________________________________________ > wayland-devel mailing list > wayland-devel@lists.freedesktop.org > https://lists.freedesktop.org/mailman/listinfo/wayland-devel _______________________________________________ wayland-devel mailing list wayland-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/wayland-devel