modesetting: Add option for non-vsynced flips for "secondary" outputs.

Whenever an unredirected fullscreen window uses pageflipping for a
DRI3/Present PresentPixmap() operation and the X-Screen has more than
one active output, multiple crtc's need to execute pageflips. Only
after the last flip has completed can the PresentPixmap operation
as a whole complete.

If a sync_flip is requested for the present, then the current
implementation will synchronize each pageflip to the vblank of
its associated crtc. This provides tear-free image presentation
across all outputs, but introduces a different artifact, if not
all outputs run at the same refresh rate with perfect synchrony:
The slowest output throttles the presentation rate, and present
completion is delayed to flip completion of the "latest" output
to complete. This means degraded performance, e.g., a dual-display
setup with a 144 Hz monitor and a 60 Hz monitor will always be
throttled to at most 60 fps. It also means non-constant present
rate if refresh cycles drift against each other, creating complex
"beat patterns", tremors, stutters and periodic slowdowns - quite
irritating!

Such a scenario will be especially annoying if one uses multiple
outputs in "mirror mode" aka "clone mode". One output will usually
be the "production output" with the highest quality and fastest
display attached, whereas a secondary mirror output just has a
cheaper display for monitoring attached. Users care about perfect
and perfectly timed tear-free presentation on the "production output",
but cares less about quality on the secondary "mirror output". They
are willing to trade quality on secondary outputs away in exchange
for better presentation timing on the "production output".

One example use case for such production + monitoring displays are
neuroscience / medical science applications where one high quality
display device is used to present visual animations to test subjects
or patients in a fMRI scanner room (production display), whereas
an operator monitors the same visual animations from a control room
on a lower quality display. Presentation timing needs to be perfect,
and animations high-speed and tear-free for the production display,
whereas quality and timing don't matter for the monitoring display.

This commit gives users the option to choose such a trade-off as
opt-in:

It adds a new boolean option "AsyncFlipSecondaries" to the device section
of xorg.conf. If this option is specified as true, then DRI3 pageflip
behaviour changes as follows:

1. The "reference crtc" for a windows PresentPixmap operation does a
   vblank synced flip, or a DRM_MODE_PAGE_FLIP_ASYNC non-synchronized
   flip, as requested by the caller, just as in the past. Typically
   flips will be requested to be vblank synchronized for tear-free
   presentation. The "reference crtc" is the one chosen by the caller
   to drive presentation timing (as specified by PresentPixmap()'s
   "target_msc", "divisor", "remainder" parameters and implemented by
   vblank events) and to deliver Present completion timestamps (msc
   and ust) extracted from its pageflip completion event.

2. All other crtc's, which also page-flip in a multi-display configuration,
   will try to flip with DRM_MODE_PAGE_FLIP_ASYNC, ie. immediately and
   not synchronized to vblank. This allows the PresentPixmap operation
   to complete with little delay compared to a single-display present,
   especially if the different crtc's run at different video refresh
   rates or their refresh cycles are not perfectly synchronized, but
   drift against each other. The downside is potential tearing artifacts
   on all outputs apart from the one of the "reference crtc".

Successfully tested on a AMD gpu with single-display, dual-display and
triple-display setups, and with single-X-Screen as well as dual-X-Screen
"ZaphodHeads" configurations.

Please consider merging this commit for the upcoming server 1.21 branch.

Signed-off-by: Mario Kleiner <mario.kleiner.de@gmail.com>

hw/xfree86/drivers/modesetting/driver.c

@@ -144,6 +144,7 @@ static const OptionInfoRec Options[] = {
     {OPTION_ATOMIC, "Atomic", OPTV_BOOLEAN, {0}, FALSE},
     {OPTION_VARIABLE_REFRESH, "VariableRefresh", OPTV_BOOLEAN, {0}, FALSE},
     {OPTION_USE_GAMMA_LUT, "UseGammaLUT", OPTV_BOOLEAN, {0}, FALSE},
+    {OPTION_ASYNC_FLIP_SECONDARIES, "AsyncFlipSecondaries", OPTV_BOOLEAN, {0}, FALSE},
     {-1, NULL, OPTV_NONE, {0}, FALSE}
 };
 
@@ -1200,6 +1201,12 @@ PreInit(ScrnInfoPtr pScrn, int flags)
                                                  &ms->vrr_support) ? X_CONFIG : X_DEFAULT;
             xf86DrvMsg(pScrn->scrnIndex, from, "VariableRefresh: %sabled\n",
                        ms->vrr_support ? "en" : "dis");
+
+            ms->drmmode.async_flip_secondaries = FALSE;
+            from = xf86GetOptValBool(ms->drmmode.Options, OPTION_ASYNC_FLIP_SECONDARIES,
+                                     &ms->drmmode.async_flip_secondaries) ? X_CONFIG : X_DEFAULT;
+            xf86DrvMsg(pScrn->scrnIndex, from, "AsyncFlipSecondaries: %sabled\n",
+                       ms->drmmode.async_flip_secondaries ? "en" : "dis");
         }
     }
 

hw/xfree86/drivers/modesetting/driver.h

@@ -60,6 +60,7 @@ typedef enum {
     OPTION_ATOMIC,
     OPTION_VARIABLE_REFRESH,
     OPTION_USE_GAMMA_LUT,
+    OPTION_ASYNC_FLIP_SECONDARIES,
 } modesettingOpts;
 
 typedef struct

hw/xfree86/drivers/modesetting/drmmode_display.h

@@ -131,6 +131,8 @@ typedef struct {
     Bool present_flipping;
     Bool flip_bo_import_failed;
 
+    Bool can_async_flip;
+    Bool async_flip_secondaries;
     Bool dri2_enable;
     Bool present_enable;
 

hw/xfree86/drivers/modesetting/modesetting.man

@@ -71,6 +71,21 @@ One of \*qglamor\*q or \*qnone\*q.  Default: glamor.
 Enable DRI3 page flipping.  The default is
 .B on.
 .TP
+.BI "Option \*qAsyncFlipSecondaries\*q \*q" boolean \*q
+Use async flips for secondary video outputs on multi-display setups. If a screen
+has multiple displays attached and DRI3 page flipping is used, then only one of
+the displays will have its page flip synchronized to vblank for tear-free
+presentation. This is the display that is used for presentation timing and
+timestamping, usually the one covering the biggest pixel area of the screen.
+All other displays ("Secondaries") will not synchronize their flips. This may
+cause some tearing on these displays, but it prevents a permanent or periodic
+slowdown or irritating judder of animations if not all video outputs are running
+synchronized with each other and with the same refresh rate. There is no perfect
+solution apart from perfectly synchronized outputs, but this option may give
+preferrable results if the displays in a multi-display setup mirror or clone
+each other.  The default is
+.B off.
+.TP
 .BI "Option \*qZaphodHeads\*q \*q" string \*q
 Specify the RandR output(s) to use with zaphod mode for a particular driver
 instance.  If you use this option you must use this option for all instances

hw/xfree86/drivers/modesetting/pageflip.c

@@ -368,10 +368,6 @@ ms_do_pageflip(ScreenPtr screen,
             ms->drmmode.flip_bo_import_failed = FALSE;
     }
 
-    flags = DRM_MODE_PAGE_FLIP_EVENT;
-    if (async)
-        flags |= DRM_MODE_PAGE_FLIP_ASYNC;
-
     /* Queue flips on all enabled CRTCs.
      *
      * Note that if/when we get per-CRTC buffers, we'll have to update this.
@@ -384,10 +380,31 @@ ms_do_pageflip(ScreenPtr screen,
     for (i = 0; i < config->num_crtc; i++) {
         enum queue_flip_status flip_status;
         xf86CrtcPtr crtc = config->crtc[i];
+        drmmode_crtc_private_ptr drmmode_crtc = crtc->driver_private;
 
         if (!xf86_crtc_on(crtc))
             continue;
 
+        flags = DRM_MODE_PAGE_FLIP_EVENT;
+        if (ms->drmmode.can_async_flip && async)
+            flags |= DRM_MODE_PAGE_FLIP_ASYNC;
+
+        /*
+         * If this is not the reference crtc used for flip timing and flip event
+         * delivery and timestamping, ie. not the one whose presentation timing
+         * we do really care about, and async flips are possible, and requested
+         * by an xorg.conf option, then we flip this "secondary" crtc without
+         * sync to vblank. This may cause tearing on such "secondary" outputs,
+         * but it will prevent throttling of multi-display flips to the refresh
+         * cycle of any of the secondary crtcs, avoiding periodic slowdowns and
+         * judder caused by unsynchronized outputs. This is especially useful for
+         * outputs in a "clone-mode" or "mirror-mode" configuration.
+         */
+        if (ms->drmmode.can_async_flip && ms->drmmode.async_flip_secondaries &&
+            (drmmode_crtc->vblank_pipe != ref_crtc_vblank_pipe) &&
+            (ref_crtc_vblank_pipe >= 0))
+            flags |= DRM_MODE_PAGE_FLIP_ASYNC;
+
         flip_status = queue_flip_on_crtc(screen, crtc, flipdata,
                                          ref_crtc_vblank_pipe,
                                          flags);

hw/xfree86/drivers/modesetting/present.c

@@ -464,8 +464,10 @@ ms_present_screen_init(ScreenPtr screen)
     int ret;
 
     ret = drmGetCap(ms->fd, DRM_CAP_ASYNC_PAGE_FLIP, &value);
-    if (ret == 0 && value == 1)
+    if (ret == 0 && value == 1) {
         ms_present_screen_info.capabilities |= PresentCapabilityAsync;
+        ms->drmmode.can_async_flip = TRUE;
+    }
 
     return present_screen_init(screen, &ms_present_screen_info);
 }