As mentioned in the previous commit, we currently prevent new modesets
on recently-removed MST connectors by returning no encoder from our
->best_encoder() callback once the MST port has disappeared. This is
wrong however, because it prevents legacy modesetting users from being
able to disable CRTCs on MST connectors after the connector's respective
topology has disappeared.
So, fix this by instead by just always returning a valid encoder.
Changes since v2:
- Remove usage of atomic MST helper for now, since that got replaced
with a much simpler solution
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Ben Skeggs <bskeggs@redhat.com>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20181008232437.5571-3-lyude@redhat.com
It appears when testing my previous fix for some of the legacy
modesetting issues with MST, I misattributed some kernel splats that
started appearing on my machine after a rebase as being from upstream.
But it appears they actually came from my patch series:
[ 2.980512] [drm:drm_atomic_helper_check_modeset [drm_kms_helper]] Updating routing for [CONNECTOR:65:eDP-1]
[ 2.980516] [drm:drm_atomic_helper_check_modeset [drm_kms_helper]] [CONNECTOR:65:eDP-1] is not registered
[ 2.980516] ------------[ cut here ]------------
[ 2.980519] Could not determine valid watermarks for inherited state
[ 2.980553] WARNING: CPU: 3 PID: 551 at drivers/gpu/drm/i915/intel_display.c:14983 intel_modeset_init+0x14d7/0x19f0 [i915]
[ 2.980556] Modules linked in: i915(O+) i2c_algo_bit drm_kms_helper(O) syscopyarea sysfillrect sysimgblt fb_sys_fops drm(O) intel_rapl x86_pkg_temp_thermal iTCO_wdt wmi_bmof coretemp crc32_pclmul psmouse i2c_i801 mei_me mei i2c_core lpc_ich mfd_core tpm_tis tpm_tis_core wmi tpm thinkpad_acpi pcc_cpufreq video ehci_pci crc32c_intel serio_raw ehci_hcd xhci_pci xhci_hcd
[ 2.980577] CPU: 3 PID: 551 Comm: systemd-udevd Tainted: G O 4.19.0-rc7Lyude-Test+ #1
[ 2.980579] Hardware name: LENOVO 20BWS1KY00/20BWS1KY00, BIOS JBET63WW (1.27 ) 11/10/2016
[ 2.980605] RIP: 0010:intel_modeset_init+0x14d7/0x19f0 [i915]
[ 2.980607] Code: 89 df e8 ec 27 02 00 e9 24 f2 ff ff be 03 00 00 00 48 89 df e8 da 27 02 00 e9 26 f2 ff ff 48 c7 c7 c8 d1 34 a0 e8 23 cf dc e0 <0f> 0b e9 7c fd ff ff f6 c4 04 0f 85 37 f7 ff ff 48 8b 83 60 08 00
[ 2.980611] RSP: 0018:ffffc90000287988 EFLAGS: 00010282
[ 2.980614] RAX: 0000000000000000 RBX: ffff88031b488000 RCX: 0000000000000006
[ 2.980617] RDX: 0000000000000007 RSI: 0000000000000086 RDI: ffff880321ad54d0
[ 2.980620] RBP: ffffc90000287a10 R08: 000000000000040a R09: 0000000000000065
[ 2.980623] R10: ffff88030ebb8f00 R11: ffffffff81416590 R12: ffff88031b488000
[ 2.980626] R13: ffff88031b4883a0 R14: ffffc900002879a8 R15: ffff880319099800
[ 2.980630] FS: 00007f475620d180(0000) GS:ffff880321ac0000(0000) knlGS:0000000000000000
[ 2.980633] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2.980636] CR2: 00007f9ef28018a0 CR3: 000000031b72c001 CR4: 00000000003606e0
[ 2.980639] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 2.980642] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 2.980645] Call Trace:
[ 2.980675] i915_driver_load+0xb0e/0xdc0 [i915]
[ 2.980681] ? kernfs_add_one+0xe7/0x130
[ 2.980709] i915_pci_probe+0x46/0x60 [i915]
[ 2.980715] pci_device_probe+0xd4/0x150
[ 2.980719] really_probe+0x243/0x3b0
[ 2.980722] driver_probe_device+0xba/0x100
[ 2.980726] __driver_attach+0xe4/0x110
[ 2.980729] ? driver_probe_device+0x100/0x100
[ 2.980733] bus_for_each_dev+0x74/0xb0
[ 2.980736] driver_attach+0x1e/0x20
[ 2.980739] bus_add_driver+0x159/0x230
[ 2.980743] ? 0xffffffffa0393000
[ 2.980746] driver_register+0x70/0xc0
[ 2.980749] ? 0xffffffffa0393000
[ 2.980753] __pci_register_driver+0x57/0x60
[ 2.980780] i915_init+0x55/0x58 [i915]
[ 2.980785] do_one_initcall+0x4a/0x1c4
[ 2.980789] ? do_init_module+0x27/0x210
[ 2.980793] ? kmem_cache_alloc_trace+0x131/0x190
[ 2.980797] do_init_module+0x60/0x210
[ 2.980800] load_module+0x2063/0x22e0
[ 2.980804] ? vfs_read+0x116/0x140
[ 2.980807] ? vfs_read+0x116/0x140
[ 2.980811] __do_sys_finit_module+0xbd/0x120
[ 2.980814] ? __do_sys_finit_module+0xbd/0x120
[ 2.980818] __x64_sys_finit_module+0x1a/0x20
[ 2.980821] do_syscall_64+0x5a/0x110
[ 2.980824] entry_SYSCALL_64_after_hwframe+0x44/0xa9
[ 2.980826] RIP: 0033:0x7f4754e32879
[ 2.980828] Code: 00 c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 8b 0d f7 45 2c 00 f7 d8 64 89 01 48
[ 2.980831] RSP: 002b:00007fff43fd97d8 EFLAGS: 00000246 ORIG_RAX: 0000000000000139
[ 2.980834] RAX: ffffffffffffffda RBX: 0000559a44ca64f0 RCX: 00007f4754e32879
[ 2.980836] RDX: 0000000000000000 RSI: 00007f475599f4cd RDI: 0000000000000018
[ 2.980838] RBP: 00007f475599f4cd R08: 0000000000000000 R09: 0000000000000000
[ 2.980839] R10: 0000000000000018 R11: 0000000000000246 R12: 0000000000000000
[ 2.980841] R13: 0000559a44c92fd0 R14: 0000000000020000 R15: 0000000000000000
[ 2.980881] WARNING: CPU: 3 PID: 551 at drivers/gpu/drm/i915/intel_display.c:14983 intel_modeset_init+0x14d7/0x19f0 [i915]
[ 2.980884] ---[ end trace 5eb47a76277d4731 ]---
The cause of this appears to be due to the fact that if there's
pre-existing display state that was set by the BIOS when i915 loads, it
will attempt to perform a modeset before the driver is registered with
userspace. Since this happens before the driver's registered with
userspace, it's connectors are also unregistered and thus-states which
would turn on DPMS on a connector end up getting rejected since the
connector isn't registered.
These bugs managed to get past Intel's CI partially due to the fact it
never ran a full test on my patches for some reason, but also because
all of the tests unload the GPU once before running. Since this bug is
only really triggered when the drivers tries to perform a modeset before
it's been fully registered with userspace when coming from whatever
display configuration the firmware left us with, it likely would never
have been picked up by CI in the first place.
After some discussion with vsyrjala, we decided the best course of
action would be to just move the unregistered connector checks out of
update_connector_routing() and into drm_atomic_set_crtc_for_connector().
The reason for this being that legacy modesetting isn't going to be
expecting failures anywhere (at least this is the case with X), so
ideally we want to ensure that any DPMS changes will still work even on
unregistered connectors. Instead, we now only reject new modesets which
would change the current CRTC assigned to an unregistered connector
unless no new CRTC is being assigned to replace the connector's previous
one.
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reported-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Fixes: 4d80273976 ("drm/atomic_helper: Disallow new modesets on unregistered connectors")
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Ville Syrjälä <ville.syrjala@linux.intel.com>
Cc: stable@vger.kernel.org
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20181009204424.21462-1-lyude@redhat.com
Currently, i915 appears to rely on blocking modesets on
no-longer-present MSTB ports by simply returning NULL for
->best_encoder(), which in turn causes any new atomic commits that don't
disable the CRTC to fail. This is wrong however, since we still want to
allow userspace to disable CRTCs on no-longer-present MSTB ports by
changing the DPMS state to off and this still requires that we retrieve
an encoder.
So, fix this by always returning a valid encoder regardless of the state
of the MST port.
Changes since v1:
- Remove mst atomic helper, since this got replaced with a much simpler
solution
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20181008232437.5571-6-lyude@redhat.com
Since we need to be able to allow DPMS on->off prop changes after an MST
port has disappeared from the system, we need to be able to make sure we
can compute a config for the resulting atomic commit. Currently this is
impossible when the port has disappeared, since the VCPI slot searching
we try to do in intel_dp_mst_compute_config() will fail with -EINVAL.
Since the only commits we want to allow on no-longer-present MST ports
are ones that shut off display hardware, we already know that no VCPI
allocations are needed. So, hardcode the VCPI slot count to 0 when
intel_dp_mst_compute_config() is called on an MST port that's gone.
Changes since V4:
- Don't use mst_port_gone at all, just check whether or not the drm
connector is registered - Daniel Vetter
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20181008232437.5571-5-lyude@redhat.com
Currently we set intel_connector->mst_port to NULL to signify that the
MST port has been removed from the system so that we can prevent further
action on the port such as connector probes, mode probing, etc.
However, we're going to need access to intel_connector->mst_port in
order to fixup ->best_encoder() so that it can always return the correct
encoder for an MST port to prevent legacy DPMS prop changes from
failing. This should be safe, so instead keep intel_connector->mst_port
always set and instead just check the status of
drm_connector->regustered to signify whether or not the connector has
disappeared from the system.
Changes since v2:
- Add a comment to mst_port_gone (Jani Nikula)
- Change mst_port_gone to a u8 instead of a bool, per the kernel bot.
Apparently bool is discouraged in structs these days
Changes since v4:
- Don't use mst_port_gone at all! Just check if the connector is
registered or not - Daniel Vetter
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20181008232437.5571-4-lyude@redhat.com
With the exception of modesets which would switch the DPMS state of a
connector from on to off, we want to make sure that we disallow all
modesets which would result in enabling a new monitor or a new mode
configuration on a monitor if the connector for the display in question
is no longer registered. This allows us to stop userspace from trying to
enable new displays on connectors for an MST topology that were just
removed from the system, without preventing userspace from disabling
DPMS on those connectors.
Changes since v5:
- Fix typo in comment, nothing else
Signed-off-by: Lyude Paul <lyude@redhat.com>
Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: stable@vger.kernel.org
Link: https://patchwork.freedesktop.org/patch/msgid/20181008232437.5571-2-lyude@redhat.com
There's not much point in following the primary vs. sprite split
for the SKL+ universal plane init code. The only difference is
of our own doing in the form of the .check_plane(). Let's make
a small exception for that little detail and otherwise share
the same code to initialize all the universal planes.
Eventually we should eliminate the mess around .check_plane()
as well, but for now let's be happy with some code reduction.
v2: Remember to set up plane->has_fbc
Make skl_plane_has_ccs() static
v3: Rebase due to NV12, rename some variables
v4: Don't leave the color_encoding/range props behind
v5: Rebase dur to blend properties, skl_plane_max_stride() and
skl_plane_check()
v6: Make skl_update_plane() static
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com> #v4
Link: https://patchwork.freedesktop.org/patch/msgid/20181005125817.22576-10-ville.syrjala@linux.intel.com
We're currently not providing the possible_crtcs mask to
drm_universal_plane_init() for primary/cursor planes. While that does
work on account of drm_crtc_init_with_planes() filling those up
for us, it's inconsisten with what we're doing for sprite planes.
Let's just always pass the possible_crtcs bitmask to
drm_universal_plane_init(). This does assume that crtc->index
== pipe. But we're already making that assumption elsewhere so
it doesn't seem like a very big sin here.
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Stanislav Lisovskiy <stanislav.lisovskiy@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20181005125817.22576-3-ville.syrjala@linux.intel.com
Sprite enable on ILK-IVB may take two frames to complete
when the hardware is in big FIFO mode (LP1+). That is
not entirely great as it means the sprite enable may
actually happen one frame after we've already signalled
flip completion. At the very least crc checks may fail
due to the sprite not yet being visible when we expect it.
We already have code to deal with big FIFO mode when it
comes to the sprite scaling on IVB
(WaCxSRDisabledForSpriteScaling:ivb). Let's extend that
workaround to kick in whenever the sprite is in the process
of being enabled. Also ILK/SNB bspec has some notes to
indicate that we should most likely also do the sprite
scaling w/a on all three platforms, so let's do that as well.
Pretty easy to reproduce on SNB/IVB. ILK has proved more
elusive, but let's trust the spec and include it as well.
v2: Make sure the pipe is active before the vblank wait
Cc: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com>
Testcase: igt/kms_plane/pixel-format-pipe-*-planes
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=107749
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Reviewed-by: Juha-Pekka Heikkila <juhapekka.heikkila@gmail.com> #v1
Link: https://patchwork.freedesktop.org/patch/msgid/20181004121527.30249-1-ville.syrjala@linux.intel.com
