I'd like to calculate FPS to detect performance issue of an application based on existing Android profiling tool .
I noted that on Systrace, it can record the length of performTraversals. As far as I know, performTraversals performs measure, layout and draw, which include most of jobs when updating a frame. So can performTraversals be representative enough to measure whether a frame will take 60 ms to update?
I also noted that Systrace record the time spending on SurfaceFlinger. I know SurfaceFlinger served for rendering purpose, but I don't know the exact beginning point and ending point of a frame. Should I also considering the time spent on SurfaceFlinger to the frame rate? (Though I do observe that SurfaceFlinger perform more frequently than performTraversals, which means SurfaceFlinger may not necessarily follow performTraversals. It will also be triggered in other scenarios.)
P.S. I'm aware of the sysdump gfxinfo, but it can only record 128 frames(~2 seconds), while what I want may last much longer.
Systrace is not useful for measuring FPS overall, but you can do that trivially with a frame counter and System.nanoTime(). If you're not hitting your target framerate, though, it can help you figure out why not.
The official docs provide some useful pointers, but there's a lot of information and the interactions can be complex. The key things to know are:
The device display panel generates a vsync signal. You can see that on the VSYNC line. Every time it transitions between 1 and 0 is a refresh.
The vsync wakes surfaceflinger, which gathers up the incoming buffers for the various windows and composites them (either itself using OpenGL ES, or through the Hardware Composer).
If your app was running faster than the panel refresh rate (usually 60fps), it will have blocked waiting for surfaceflinger (in, say, eglSwapBuffers()). Once surfaceflinger acquires the buffer, the app is free to continue and generate another frame.
Unless you're rendering offscreen, you can't go faster than surfaceflinger.
As of Android 4.3 (API 18) you can add your own events to the systrace output using the android.os.Trace class. Wrapping your draw method with trace markers can be extremely informative. You have to enable their tag with systrace to see them.
If you want to be running at 60fps, your rendering must finish in well under 16.7ms. If you see a single invocation of performTraversals taking longer than that, you're not going to hit maximum speed.
Related
I am attempting to determine (to within 1 ms) when particular screen flips happen on Android. Choreographer fires every time a frame flips, but gives no way of determining which frame is actually being displayed. According to https://source.android.com/devices/graphics/architecture.html, there are several layers in the process: the user land buffer, which flips to a triple-buffered queue, which flips to the surface flinger, which flips to the hardware. Each of these layers can potentially drop a frame, but at this point I have only determined how to to monitor the user land buffer. Is there a way to monitor the other buffers/flips (in real time, on a non-rooted, non-custom phone)?
I have observed unexpected frame delays on the HTC M8 (about 1 every 5 minutes), but the Nexus 7 does not appear to have this problem. I measure the delays by using a Cedrus StimTracker (http://cedrus.com/stimtracker/) with a photo sensor and the Lab Streaming Layer (https://github.com/sccn/labstreaminglayer). I have tried using eglPresentationTimeANDROID to control when screens are flipped, and that has not fixed the problem.
Note that I'm using the ndk, but I can usually use the JNI to get access to non-ndk features when I need to.
The reason I care is in order to use Android for psychological and neurological experiments, where 1 ms precision is highly desirable.
As far as accessible APIs go, it sounds like you've found the relevant bits and pieces. If you haven't yet, please read through this stackoverflow item.
Using Choreographer and extrapolation, you can guess at when the next display refresh will occur. Using eglPresentationTimeANDROID() on an Android 5.0+ device, you can tell SurfaceFlinger when you want a particular frame to be sent to the display. Assuming SurfaceFlinger is properly accounting for all latency (such as additional frames added by "smart" panels), that should get you reliable timing.
(Bear in mind that the timing is based on when the display latches the next frame, not when the next frame is fully visible on the display... the latency there will depend on the panel.)
Grafika's "scheduled swap" Activity uses this feature, but it sounds like you're already familiar.
The only way to get signaled by the display when it does the swap would be to dup() the display-retire fence fd from the previous frame, and wait on it. Some of the code in SurfaceFlinger does this, notably DispSync watches the retire fences to see if the software "VSYNC" is drifting. There is no public API for fences, and the user-space response time could certainly be more than 1ms anyway... it usually works out better to schedule ahead than it does to react. Your requirement for non-rooted non-custom devices makes this problematic.
If you're mostly seeing correct behavior, but occasionally seeing a miss, your best bet is to use systrace to track down the cause.
I'm trying to use the output of systrace to detect janky scrolling during automated tests: I want to notice it early, without having to sit there watching.
I spent some time trying to fathom the trace, and found this ebook very helpful: https://www.safaribooksonline.com/library/view/high-performance-android/9781491913994/ch04.html
The most promising hypothesis was checking whether VSYNC-sf ever stopped ticking on phones displaying VSYNC-sf.
On other machines, SurfaceFlinger seems to be started by either HW_SYNC_0 or VSYNC (sometimes one or both of those VSYNCs stop) but SurfaceFlinger also seems to be involved with VsyncOn, which sometimes appears to keep track of whether there are activity buffers outstanding, and sometimes whether there are input events that need delivering. Confusingly, sometimes input events are delivered during half-second pauses when there's no surface flinger activity, no application drawing, and when even the VSYNC and HW_VSYNC signals decide to pause.
Does anyone know what's going on there?
Should I simply expect to see Surface Flinger always busy - not alternately busy and idle with each tick - and always aligned with one or other of the VSYNCs?
I also sometimes see SurfaceFlinger taking longer than a tick to complete its processing - is that the application's fault for having a very complicated display, or is it just something that happens because some queue isn't empty enough?
I'd prefer to miss a possible jank than claim to have found one which isn't there.
Thanks!
Testing Display Performance Lists how to use the new framestats command from dumpsys to get this type of information. It will provide information on what frames you've missed, and how many of them you've missed.
It's also worth noting that SurfaceFlinger isn't always busy. It's only active when part of the screen needs to be updated. If nothing on the screen needs updating, then no new rendering occurs, and such, SurfaceFlinger should be idle.
You can get a bigger-picture view of the Android rendering pipeline with the Rendering Performance 101 video from Android Performance Patterns.
I'm trying to get a better understanding of the Android display subsystem, but one item that's still confusing to me is how VSYNC signals are handled, and why so many exist in the first place.
Android is designed to use VSYNC at its core, but there are multiple VSYNC signals that it employs. Via https://source.android.com/devices/graphics/implement.html in the "VSYNC Offset" section, there is a flow diagram which diagrams three VSYNC signals: HW_VSYNC_0, VSYNC, and SF-VSYNC. I understand that HW_VSYNC is used to update the timing in DispSync, and that VSYNC and SF-VSYNC are used by the apps and surfaceflinger, but why are these individual signals necessary at all? Furthermore, how do the offsets impact these signals? Is there a timing diagram available anywhere which better explains this?
Thanks for any help you can offer.
To understand this stuff, it's best to start with the System-Level Graphics Architecture document, taking particular note of The Need for Triple-Buffering section and the associated diagram (which ideally would be an animated GIF). The sentence that begins, "If the app starts rendering halfway between VSYNC signals" is talking specifically about DispSync. Once you've read that, hopefully the DispSync section of the device graphics doc makes more sense.
Most devices don't have DispSync offsets configured, so there is really only one VSYNC signal. In what follows I'm assuming DispSync is enabled.
The hardware only provides one VSYNC signal, corresponding to the primary display refresh. The others are generated in software by the SurfaceFlinger DispSync code, firing at fixed offsets from the actual VSYNC. Some clever software is used to keep the timings from slipping out of phase.
The signals are used to trigger SurfaceFlinger composition and app rendering. If you follow the section in the architecture document, you can see that this establishes two frames of latency between when the app renders its content, and when the content appears on the screen. Think of it like this: given three occurrences of VSYNC, the app draws at V0, the system does composition at V1, and the composed frame is sent to the display at V2.
If you're trying to track touch input, perhaps moving a map around under the user's finger, any latency will be seen by the user as sluggish touch response. The goal is to minimize the latency to improve the user experience. Suppose we delayed the events slightly, so the app draws at V0.5, we composite at V1.2, and then swap to the display at V2. By offsetting the app and SF activity we reduce the total latency from 2 frames to 1.5, as shown below.
That's what DispSync is for. In the feedback diagram on the page you linked, HW_VSYNC_0 is the hardware refresh for the physical display, VSYNC causes the app to render, and SF_VSYNC causes SurfaceFlinger to perform composition. Referring to them as "VSYNC" is a bit of a misnomer, but on an LCD panel referring to anything as "VSYNC" is probably a misnomer.
The "retire fence timestamps" noted in the feedback loop diagram refers to a clever optimization. Since we're not doing any work on the actual hardware VSYNC, we can be slightly more efficient if we turn the refresh signal off. The DispSync code will instead use the timestamps from retire fences (which is a whole other discussion) to see if it is falling out of sync, and will temporarily re-enable the hardware signal until it's back on track.
Edit: you can see how the values are configured in the Nexus 5 boardconfig. Note the settings for VSYNC_EVENT_PHASE_OFFSET_NS and SF_VSYNC_EVENT_PHASE_OFFSET_NS.
I am building a game like application using android NDK and openGL ES 2.0
So far I understand the concept of vertices and shaders and programs.
The main game loop would be a loop in a single thread as follows
step 1. Read all the user input
step 2. Update game objects (if needed) based on the input
step 3. make draw calls for all the objects
step 4. call glSwapBuffers
then loop back to step 1
But I ran into various confusions regarding sync and threading so I'm listing all the question together.
1.Since open GL draw calls are asynchronous the draw calls and glSwapBuffers may get called many times before the gpu has even rendered actually a single frame from calls from last iteration of loop. Will this be problematic? buffer overflow or tearing ?
2.Assuming VSYNC is enabled then does point 1 still causes problem?
3.Since all calls are async how do I measure the time spent rendering each frame? glSwapBuffers would return immediately so how can I know when was the frame actually done?
4.loading textures will occupy space in the ram is checking free memory before loading texture standard way or I should keep loading textures until I reach OUT_OF_MEMORY_ERROR?
5.If I switch to multithreaded approach calling just glswapbuffers at a fixed 60 times per second without any regard to the thread which is processing input and giving out draw calls then what is supposed to happen?
Also how do I control the fps in game loop? I know the exact fps depends on a large no of factors but how can you go close to that
The SwapBuffers() will not be executed out of order. Issuing it after all of the draw commands for the frame is fine. The driver will take care about it, you don't need to sync anything. You can only screw this about by using multiple threads or multiple contexts, but even that would take a lot of effort.
There is no problem with 1, and VSYNC does not directly change anything here.
The calls might be asynchronous, but the driver will not queue up an unlimit amount of work. Sooner or later, it will have to block, if you try to issue too many calls in advance. When vsync is on, the typicial behavior is that the driver will queue up at most a few frames (or just one, depending on the driver settings), and SwapBuffers() will block when that limit is reached. So the timing statistics you get there are accurate, after the first few frames. Note that this is still much better than completely flushing the queue, as the driver unblocks as soon as the first pending buffer swap was carried out.
This is a totally new topic, which probably belongs into another question. However: It is very unlikely that you get any of the current desktop GL implementations to ever generate GL_OUT_OF_MEMORY. The driver will automatically page textures (and other objects) between VRAM and system RAM (and the OS might even page that to disk). The GL also provides no means to query the available memory.
In that scenario, you will need to synchronize manually. That approach does not make the slightest sense and seems like trying to solve a problem which does not exist. If you want your game to use multithreading, still put all the gl rendering (and swapbuffers) into the same thread. You can use different threads for input processing, sound, physics, update of the scene, general game logic and whatever. But you should just use a single thread/single context approach for the GL. That way, it also won't hurt you when SwapBuffers() blocks your render thread, as your game logic and input handling is still done, and the render thread will just render new frames with the newest available data in the frequency the display needs (with vsync on) or as fast as the CPU and GPU can work (if vsync is off).
I'm developing a GL live wallpaper that uses very little CPU and only modest GPU. On my older test phone, it can run at a full 58fps or so most of the time. But occasionally the effects ramp up, and then the render times jitter between 16ms and 50ms per frame. For example, it'll render several frames at 16ms, slide up to 50ms over a dozen frames or so, render several more frames at 50ms, then slide back down to 16ms and repeat. I discovered that if I set the CPU governor to "performance" (or "conservative", curiously enough) instead of the default "ondemand" it'll render with full effects at full speed. Alternatively, if I leave the governor alone and insert a busy loop in my code (increment a variable 100,000 times per frame) that bumps my CPU usage up enough to transition to a higher clock rate and render smoothly as well.
So it seems on this phone my app is bottlenecked by the GPU, but only when it throttles down. Now, I wouldn't mind if the GLSurfaceView rendered at a slower rate according to the GPU clock, but my problem here is that I'm getting the bursts of alternating high and low frame rates which makes my animation look fluid/frameskippy/fluid/frameskippy/etc. several times per second. It seems like the GPU clock is ramping up and down like crazy?
I got a visible improvement by using RENDERMODE_WHEN_DIRTY and calling requestRender() on a strictly timed thread, but the darn GPU keeps ramping up and down. Why won't it either render as fast as it can at the slower clock, or just jump to and STAY AT the higher clock?
The best solution I've come up with so far is using a sliding window to detect the average frame update time, then applying the difference from the target frame time until the two values converge. The time between render updates is slower but at least it's roughly constant. So that works in theory, but it takes several seconds to reach a steady state and it looks bad in the meantime.
I think a third option might be to cannibalize the GLSurfaceView source and make a custom version. From what I understand, the blocking GL calls are made in there, so it would be much easier for me to time render calls and react accordingly. I'm not very comfortable attempting that though because there's a lot of code in there that I'd have to spend a lot of time understanding before I could even begin to mess with it. Plus I'd then have to worry about how well version X of GLSurfaceView plays with any version Y of Android.
So, with all that said, do I have any other options here? Is there an easier fix to this?
try fixing the frame rate by pausing the thread (thread sleep) for the remaining time to reach a constant frame rate.