Scrcpy: How to compile with -Dskip_frames=false?
I saw on https://github.com/Genymobile/scrcpy/issues/244 there is a flag to disable the skipping of frames. However, when I use the command meson x -Dskip_frames=false and then compile, the resulting exe still skips frames. I also tried cd x and then ninja as well as meson x --reconfigure to no avail. Am I missing something obvious or does that no longer work? I am trying to compile the latest dev build for the lost focus bug fix if that makes a difference.
BaconCatBug
All 24 comments
the resulting exe still skips frames
How do you see that? If you show fps with Ctrl+i, does it still print "skipped frames" (sometimes)?
rom1v
on 29 May 2019
Yes, it still shows up as (+2 frames skipped). I assume I am missing something obvious and misconfiguring before compilation.
BaconCatBug
on 29 May 2019
Do I need to do anything specific after running meson x -Dskip_frames=false
BaconCatBug
on 3 Jun 2019
@BaconCatBug
Just ninja -Cx then ./run x to execute.
rom1v
on 3 Jun 2019
Ok I just compiled the latest dev build and I am still getting the +x Frames Skipped message when I press Control+i
I am running on windows and cross compiling on my Ubuntu laptop, both the 32 and 64 bit ones have the skipped frames.
BaconCatBug
on 3 Jun 2019
This is very weird, I cannot reproduce.
If I pass -Dskip_frames=true or nothing, I have (+x frames skipped). If I pass -Dskip_frames=false, I don't.
What is the result of meson configure x?
rom1v
on 3 Jun 2019
bcb@BCB-Laptop:~/Desktop/dev2$ meson x -Dskip_frames=false
The Meson build system
Version: 0.50.1
Source dir: /home/bcb/Desktop/dev2
Build dir: /home/bcb/Desktop/dev2/x
Build type: native build
Project name: scrcpy
Project version: 1.8
Native C compiler: cc (gcc 7.4.0 "cc (Ubuntu 7.4.0-1ubuntu1~18.04) 7.4.0")
Build machine cpu family: x86_64
Build machine cpu: x86_64
Found pkg-config: /usr/bin/pkg-config (0.29.1)
Dependency libavformat found: YES 57.83.100
Dependency libavcodec found: YES 57.107.100
Dependency libavutil found: YES 55.78.100
Dependency sdl2 found: YES 2.0.8
Configuring config.h using configuration
Program ./scripts/build-wrapper.sh found: YES (/home/bcb/Desktop/dev2/server/./scripts/build-wrapper.sh)
DEPRECATION: build_always is deprecated. Combine build_by_default and build_always_stale instead.
Build targets in project: 7
Found ninja-1.8.2 at /usr/bin/ninja
bcb@BCB-Laptop:~/Desktop/dev2$ meson configure x
Core properties:
Source dir /home/bcb/Desktop/dev2
Build dir /home/bcb/Desktop/dev2/x
Core options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
auto_features auto [enabled, disabled, auto] Override value of all 'auto' features
backend ninja [ninja, vs, vs2010, vs2015, vs2017, vs2019, xcode] Backend to use
buildtype debug [plain, debug, debugoptimized, release, minsize, custom] Build type to use
debug true [true, false] Debug
default_library shared [shared, static, both] Default library type
install_umask 0022 [preserve, 0000-0777] Default umask to apply on permissions of installed files
layout mirror [mirror, flat] Build directory layout
optimization 0 [0, g, 1, 2, 3, s] Optimization level
strip false [true, false] Strip targets on install
unity off [on, off, subprojects] Unity build
warning_level 1 [0, 1, 2, 3] Compiler warning level to use
werror false [true, false] Treat warnings as errors
wrap_mode default [default, nofallback, nodownload, forcefallback] Wrap mode
Backend options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
backend_max_links 0 >=0 Maximum number of linker processes to run or 0 for no limit
Base options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
b_asneeded true [true, false] Use -Wl,--as-needed when linking
b_colorout always [auto, always, never] Use colored output
b_coverage false [true, false] Enable coverage tracking.
b_lto false [true, false] Use link time optimization
b_lundef true [true, false] Use -Wl,--no-undefined when linking
b_ndebug false [true, false, if-release] Disable asserts
b_pch true [true, false] Use precompiled headers
b_pgo off [off, generate, use] Use profile guided optimization
b_pie false [true, false] Build executables as position independent
b_sanitize none [none, address, thread, undefined, memory, address,undefined] Code sanitizer to use
b_staticpic true [true, false] Build static libraries as position independent
Compiler options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
c_args [] Extra arguments passed to the C compiler
c_link_args [] Extra arguments passed to the C linker
c_std c11 [none, c89, c99, c11, gnu89, gnu99, gnu11] C language standard to use
Directories:
Option Current Value Description
------ ------------- -----------
bindir bin Executable directory
datadir share Data file directory
includedir include Header file directory
infodir share/info Info page directory
libdir lib/x86_64-linux-gnu Library directory
libexecdir libexec Library executable directory
localedir share/locale Locale data directory
localstatedir /var/local Localstate data directory
mandir share/man Manual page directory
prefix /usr/local Installation prefix
sbindir sbin System executable directory
sharedstatedir /var/local/lib Architecture-independent data directory
sysconfdir etc Sysconf data directory
Project options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
build_app true [true, false] Build the client
build_server true [true, false] Build the server
crossbuild_windows false [true, false] Build for Windows from Linux
hidpi_support true [true, false] Enable High DPI support
override_server_path Hardcoded path to find the server at runtime
prebuilt_server Path of the prebuilt server
skip_frames false [true, false] Always display the most recent frame
windows_noconsole false [true, false] Disable console on Windows (pass -mwindows flag)
Testing options:
Option Current Value Possible Values Description
------ ------------- --------------- -----------
errorlogs true [true, false] Whether to print the logs from failing tests
stdsplit true [true, false] Split stdout and stderr in test logs
bcb@BCB-Laptop:~/Desktop/dev2$
What is the result of:
ninja -Cx
and
cat x/app/config.h
?
rom1v
on 3 Jun 2019
bcb@BCB-Laptop:~/Desktop/dev2$ ninja -Cx
ninja: Entering directory `x'
[1/1] Generating scrcpy-server with a custom command.
> Task :server:preBuild UP-TO-DATE
> Task :server:preDebugBuild
> Task :server:compileDebugAidl
> Task :server:compileDebugRenderscript
> Task :server:checkDebugManifest
> Task :server:generateDebugBuildConfig
> Task :server:prepareLintJar UP-TO-DATE
> Task :server:generateDebugSources
> Task :server:javaPreCompileDebug
> Task :server:mainApkListPersistenceDebug
> Task :server:generateDebugResValues
> Task :server:generateDebugResources
> Task :server:mergeDebugResources
> Task :server:createDebugCompatibleScreenManifests
> Task :server:processDebugManifest
> Task :server:processDebugResources
> Task :server:compileDebugJavaWithJavac
> Task :server:compileDebugNdk NO-SOURCE
> Task :server:compileDebugSources
> Task :server:mergeDebugShaders
> Task :server:compileDebugShaders
> Task :server:generateDebugAssets
> Task :server:mergeDebugAssets
> Task :server:mergeExtDexDebug
> Task :server:mergeLibDexDebug
> Task :server:transformClassesWithDexBuilderForDebug
> Task :server:mergeProjectDexDebug
> Task :server:validateSigningDebug
> Task :server:signingConfigWriterDebug
> Task :server:mergeDebugJniLibFolders
> Task :server:transformNativeLibsWithMergeJniLibsForDebug
> Task :server:processDebugJavaRes NO-SOURCE
> Task :server:transformResourcesWithMergeJavaResForDebug
> Task :server:packageDebug
> Task :server:assembleDebug
BUILD SUCCESSFUL in 2s
27 actionable tasks: 26 executed, 1 up-to-date
bcb@BCB-Laptop:~/Desktop/dev2$ cat x/app/config.h
/*
* Autogenerated by the Meson build system.
* Do not edit, your changes will be lost.
*/
#pragma once
#define BUILD_DEBUG
#define DEFAULT_BIT_RATE 8000000
#define DEFAULT_LOCAL_PORT 27183
#define DEFAULT_MAX_SIZE 0
#define HIDPI_SUPPORT
#undef OVERRIDE_SERVER_PATH
#define PREFIX "/usr/local"
#define PREFIXED_SERVER_PATH "/share/scrcpy/scrcpy-server.jar"
#define SCRCPY_VERSION "1.8"
#undef SKIP_FRAMES
#undef WINDOWS_NOCONSOLE
bcb@BCB-Laptop:~/Desktop/dev2$
#undef SKIP_FRAMES
config.h is correctly included in fps_counter.h:
https://github.com/Genymobile/scrcpy/blob/8c8649cfcd710859ce18eab557ed2af8cedb9a42/app/src/fps_counter.h#L7
The code which prints (+x frames skipped) is in a #ifdef SKIP_FRAMES block:
https://github.com/Genymobile/scrcpy/blob/8c8649cfcd710859ce18eab557ed2af8cedb9a42/app/src/fps_counter.c#L31-L36
So it should not be executed (or even compiled) in your binary. The only reason I can think of is that you don't execute the correct binary. What command do you type to execute?
rom1v
on 3 Jun 2019
I run scrcpy -c 1080:1920:0:120 -m 1024 -b 20M on Windows 8.1 after compiling with make -f Makefile.CrossWindows
BaconCatBug
on 3 Jun 2019
after compiling with
make -f Makefile.CrossWindows
This generates the Windows release in dist/.
rom1v
on 3 Jun 2019
Yes, I know. I then transfer the zip to my windows machine and extract it, then run scrcpy.exe I am not running scrcpy on my Ubuntu laptop, although I will test it now actually to see if the linux version skips frames or not.
BaconCatBug
on 3 Jun 2019
From the right directory? Do you have another scrcpy.exe somewhere accessible from your PATH?
What if you execute .\scrcpy.exe from your release directory?
rom1v
on 3 Jun 2019
I do not have any other scrcpy.exe in my path (scrcpy is not recognised).
Using the exe from the scrcpy-win64 folder instead of the zip is the same result, skipped frames.
The linux version works, no skipped frames.
BaconCatBug
on 3 Jun 2019
Oh, ok, I get it. meson x/ninja -Cx is just for your local build.
Makefile.CrossWindows generates the release with its own build dir and configuration.
diff --git a/Makefile.CrossWindows b/Makefile.CrossWindows
index 0f7b14a..1e59969 100644
--- a/Makefile.CrossWindows
+++ b/Makefile.CrossWindows
@@ -57,6 +57,7 @@ build-win32: prepare-deps-win32
--buildtype release --strip -Db_lto=true \
-Dcrossbuild_windows=true \
-Dbuild_server=false \
+ -Dskip_frames=false \
-Doverride_server_path=scrcpy-server.jar )
ninja -C "$(WIN32_BUILD_DIR)"
@@ -67,6 +68,7 @@ build-win32-noconsole: prepare-deps-win32
--buildtype release --strip -Db_lto=true \
-Dcrossbuild_windows=true \
-Dbuild_server=false \
+ -Dskip_frames=false \
-Dwindows_noconsole=true \
-Doverride_server_path=scrcpy-server.jar )
ninja -C "$(WIN32_NOCONSOLE_BUILD_DIR)"
@@ -81,6 +83,7 @@ build-win64: prepare-deps-win64
--buildtype release --strip -Db_lto=true \
-Dcrossbuild_windows=true \
-Dbuild_server=false \
+ -Dskip_frames=false \
-Doverride_server_path=scrcpy-server.jar )
ninja -C "$(WIN64_BUILD_DIR)"
@@ -91,6 +94,7 @@ build-win64-noconsole: prepare-deps-win64
--buildtype release --strip -Db_lto=true \
-Dcrossbuild_windows=true \
-Dbuild_server=false \
+ -Dskip_frames=false \
-Dwindows_noconsole=true \
-Doverride_server_path=scrcpy-server.jar )
ninja -C "$(WIN64_NOCONSOLE_BUILD_DIR)"
Ah, thank you,
Adding those four -Dskip_frames=false \ to the Makefile.CrossWindows file has fixed the issue, it is now no longer skipping frames on Windows.
Thank you for your help and apologies for taking up your time!
BaconCatBug
on 3 Jun 2019
You're welcome.
Do you see any difference in practice (except in the console)?
rom1v
on 3 Jun 2019
I do but it could be placebo for all I know. I'd need to do a proper side
by side with recording to be sure. Even so it's still a little choppier
than when I view the app on the phone but that might be down to my pretty
old CPU/GPU combo. (Which happens with or without the skipframes).
On Mon, 3 Jun 2019 at 16:27, Romain Vimont notifications@github.com wrote:
You're welcome.
Do you see any difference in practice (except in the console)?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/Genymobile/scrcpy/issues/563?email_source=notifications&email_token=ABFLTCZIX4PPACNUKJJN7WTPYUZ5PA5CNFSM4HQOI4TKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGODWZYPOA#issuecomment-498304952,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ABFLTC5X4FNK3ZNPI23FON3PYUZ5PANCNFSM4HQOI4TA
.
BaconCatBug
on 3 Jun 2019
How many frames were skipped each second with skip_frames enabled?
rom1v
on 3 Jun 2019
Between 1 and 5, with the occasional 6 or 7.
On Mon, 3 Jun 2019 at 16:31, Romain Vimont notifications@github.com wrote:
>
How many frames were skipped each second with skip_frames enabled?
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.
BaconCatBug
on 3 Jun 2019
On dev branch, I replaced the compilation flag skip_frames by a runtime command-line parameter:
# equivalent to the old compilation option skip_frames=false
scrcpy --render-expired-frames
ebccb9f6cc111e8acfbe10d656cac5c1f1b744a0
rom1v
on 5 Jun 2019
Apologies for posting on a closed issue, but by replaced, you mean that if we're to run scrcpy without frame skips, we cannot compile it with the previous -Dskip_frames=false flag, so that it runs by default without frame skipping, and we absolutely have to use the new runtime command-line parameter?
I noticed that -Dskip_frames=false is a flag that's passed to mason, so I just edited the makefile to include -Dskip-frames=false on the build-win32/64 rules, including the no console ones, for a total of 4 times, to no avail, and then I found this thread. It would appear from the posts above that you had done something similar in an older version of scrcpy. I have pasted the makefile in a pastebin, if it helps: https://pastebin.com/KrHZpfSt
If we can only stop the frame skipping using that parameter, is it possible to make it so that it can read some sort of configuration file so that I don't have to run it from the command line every time?
Thank you for your time.
Xarizzar
on 20 Feb 2020
Just make a batch file or shell file to run the command with the parameters you want.
BaconCatBug
on 20 Feb 2020
Related issues
SuryaElavazhagan
·
4Comments
cordac1
·
4Comments
BloodPHamtOm
·
3Comments
behzadpooldar
·
4Comments
targor
·
3Comments
Most helpful comment
Between 1 and 5, with the occasional 6 or 7.
On Mon, 3 Jun 2019 at 16:31, Romain Vimont notifications@github.com wrote:
>