WebView memory leaks on Samsung Galaxy S4
Reported by
peabeyna...@gmail.com,
May 24 2017
|
||||||||||||||||||
Issue description
Steps to reproduce the problem:
1. Create empty android project;
2. Add code in function onCreate():
WebView webview = new WebView(this);
setContentView(webview);
webview.getSettings().setJavaScriptEnabled(true);
webview.loadUrl("http://viber.absolutist.com/promo/index5.php?platform=android&bundle=com.absolutist.spacelegendsfree&language=en");
3. Run app on Samsung Galaxy S4
4. Wait for 1 minute and app crashes with
Abort message: '[FATAL:memory.cc(22)] Out of memory. size=589824
What is the expected behavior?
App won't crash
What went wrong?
App crashes
Did this work before? No
Chrome version: 58.0.3029.83 Channel: stable
OS Version: 5.0.1
Flash Version:
MainActivity.java, apk-file and logcat is attached
,
May 24 2017
satyavathir@, please assign and see if this is reproducible and if it happens on system webview version.
,
May 24 2017
,
May 25 2017
Downloading 1 files took 0.003017 second(s)
Operating system: Android
samsung/ja3gxx/ja3g:5.0.1/LRX22C/I9500XXUHPK1:user/release-keys
CPU: arm
4 CPUs
GPU: OpenGL ES 2.0 build 1.10@2359475
Imagination Technologies
PowerVR SGX 544MP
Crash reason:
Crash address: 0x0
Process uptime: not available
Thread 0 (crashed)
0 libc.so + 0x3a0ec
r0 = 0x00000000 r1 = 0x000075e7 r2 = 0x00000006 r3 = 0x00000000
r4 = 0x9d001db8 r5 = 0x00000006 r6 = 0x00000001 r7 = 0x0000010c
r8 = 0xb6f92df4 r9 = 0x9d0019a4 r10 = 0x00000000 r12 = 0x000075e7
fp = 0xa4e44ca2 sp = 0x9d0014d0 lr = 0xb6f3c551 pc = 0xb6f5f0ec
Found by: given as instruction pointer in context
1 libc.so + 0x18161
sp = 0x9d0014e8 pc = 0xb6f3d163
Found by: stack scanning
2 libc.so + 0x14aa7
sp = 0x9d0014f0 pc = 0xb6f39aa9
Found by: stack scanning
3 libwebviewchromium.so!RealCalloc [allocator_shim_default_dispatch_to_linker_wrapped_symbols.cc : 43 + 0x7]
sp = 0x9d0014fc pc = 0xa2fe65f9
Found by: stack scanning
4 libc.so + 0x12a6e
sp = 0x9d001518 pc = 0xb6f37a70
Found by: stack scanning
5 libwebviewchromium.so!base::debug::BreakDebugger() [debugger_posix.cc : 221 + 0x3]
sp = 0x9d001520 pc = 0xa2f8d501
Found by: stack scanning
6 libwebviewchromium.so!logging::LogMessage::~LogMessage() [logging.cc : 759 + 0x3]
r3 = 0x0defaced sp = 0x9d001528 pc = 0xa2f99d67
Found by: call frame info
7 libwebviewchromium.so!OnNoMemory [memory.cc : 22 + 0x5]
r4 = 0x0defaced r5 = 0x9d0019a0 r6 = 0x9d0019d8 r7 = 0x9d00199c
r8 = 0xa4e94288 r9 = 0x9d0019a4 r10 = 0x0defaced sp = 0x9d001980
pc = 0xa2fad755
Found by: call frame info
8 libwebviewchromium.so!gpu::GpuMemoryBufferImplSharedMemory::Map() [gpu_memory_buffer_impl_shared_memory.cc : 180 + 0x5]
r4 = 0x0defaced r5 = 0x0defaced r6 = 0x00000000 r7 = 0x0defaced
r8 = 0x0defaced r9 = 0x0defaced r10 = 0x0defaced fp = 0xa4e44ca2
sp = 0x9d001a60 pc = 0xa3436d47
Found by: call frame info
9 libwebviewchromium.so!cc::OneCopyRasterBufferProvider::PlaybackToStagingBuffer(cc::StagingBuffer*, cc::Resource const*, cc::RasterSource const*, gfx::Rect const&, gfx::Rect const&, float, sk_sp<SkColorSpace>, cc::RasterSource::PlaybackSettings const&, unsigned long long, unsigned long long) [one_copy_raster_buffer_provider.cc : 293 + 0x7]
r4 = 0x0defaced r5 = 0x0defaced r6 = 0x9d001aa8 r7 = 0x0defaced
r8 = 0x0defaced r9 = 0x0defaced r10 = 0x0defaced fp = 0xa4e44ca2
sp = 0x9d001a70 pc = 0xa3030813
Found by: call frame info
10 libwebviewchromium.so!cc::OneCopyRasterBufferProvider::PlaybackAndCopyOnWorkerThread(cc::Resource const*, cc::ResourceProvider::ScopedWriteLockGL*, gpu::SyncToken const&, cc::RasterSource const*, gfx::Rect const&, gfx::Rect const&, float, cc::RasterSource::PlaybackSettings const&, unsigned long long, unsigned long long) [one_copy_raster_buffer_provider.cc : 240 + 0x3]
r4 = 0x0defaced r5 = 0x9d001b24 r6 = 0x0defaced r7 = 0x0defaced
r8 = 0x0defaced r9 = 0x0defaced r10 = 0x000009e7 fp = 0x00000000
sp = 0x9d001af8 pc = 0xa3030ce5
Found by: call frame info
11 libwebviewchromium.so!cc::OneCopyRasterBufferProvider::RasterBufferImpl::Playback(cc::RasterSource const*, gfx::Rect const&, gfx::Rect const&, unsigned long long, float, cc::RasterSource::PlaybackSettings const&) [one_copy_raster_buffer_provider.cc : 66 + 0x33]
r4 = 0x0defaced r5 = 0x0defaced r6 = 0x9d001c40 r7 = 0x0defaced
r8 = 0x0defaced r9 = 0x0defaced r10 = 0x0defaced fp = 0x00000000
sp = 0x9d001b60 pc = 0xa3030dd7
Found by: call frame info
12 libwebviewchromium.so!RunOnWorkerThread [tile_manager.cc : 131 + 0x1d]
r4 = 0x0defaced r5 = 0xa3030d75 r6 = 0x9d001c40 r7 = 0x0defaced
r8 = 0x9d001c60 r9 = 0x9d001c2c r10 = 0x000009f6 fp = 0x00000000
sp = 0x9d001bf0 pc = 0xa3058897
Found by: call frame info
13 libwebviewchromium.so!content::CategorizedWorkerPool::RunTaskInCategoryWithLockAcquired(cc::TaskCategory) [categorized_worker_pool.cc : 359 + 0x3]
r4 = 0x9d001cf0 r5 = 0x0defaced r6 = 0x0defaced r7 = 0x9d001ce4
r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d fp = 0x9d001db0
sp = 0x9d001cd8 pc = 0xa4608843
Found by: call frame info
14 libwebviewchromium.so!content::CategorizedWorkerPool::RunTaskWithLockAcquired(std::__ndk1::vector<cc::TaskCategory, std::__ndk1::allocator<cc::TaskCategory> > const&) [categorized_worker_pool.cc : 338 + 0x7]
r4 = 0x0defaced r5 = 0x00000001 r6 = 0x0defaced r7 = 0x0defaced
r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d fp = 0x9d001db0
sp = 0x9d001d28 pc = 0xa46088c1
Found by: call frame info
15 libwebviewchromium.so!content::CategorizedWorkerPool::Run(std::__ndk1::vector<cc::TaskCategory, std::__ndk1::allocator<cc::TaskCategory> > const&, base::ConditionVariable*) [categorized_worker_pool.cc : 230 + 0x7]
r3 = 0x00000000 r4 = 0x0defaced r5 = 0x0defaced r6 = 0x0defaced
r7 = 0x00000078 r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d
fp = 0x9d001db0 sp = 0x9d001d40 pc = 0xa46088e5
Found by: call frame info
16 libwebviewchromium.so!base::SimpleThread::ThreadMain() [simple_thread.cc : 68 + 0x7]
r4 = 0x0defaced r5 = 0x9d001d5c r6 = 0x0defaced r7 = 0x00000078
r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d fp = 0x9d001db0
sp = 0x9d001d58 pc = 0xa2fc625b
Found by: call frame info
17 libwebviewchromium.so!ThreadFunc [platform_thread_posix.cc : 71 + 0x7]
r4 = 0x0defaced r5 = 0x0defaced r6 = 0x0defaced r7 = 0x00000078
r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d fp = 0x9d001db0
sp = 0x9d001d88 pc = 0xa2fc3151
Found by: call frame info
18 libc.so + 0x16d3b
r4 = 0x0defaced r5 = 0x0defaced r6 = 0x0defaced r7 = 0x00000078
r8 = 0x0defaced r9 = 0x0defaced r10 = 0xb6f3bd1d fp = 0x9d001db0
sp = 0x9d001d98 pc = 0xb6f3bd3d
Found by: call frame info
19 libc.so + 0x16d1b
sp = 0x9d001d9c pc = 0xb6f3bd1d
Found by: stack scanning
20 libc.so + 0x16d1b
sp = 0x9d001da4 pc = 0xb6f3bd1d
Found by: stack scanning
21 libc.so + 0x14e03
sp = 0x9d001da8 pc = 0xb6f39e05
Found by: stack scanning
Loaded modules:
0x75d6b000 - 0x77ff3fff boot.oat ???
0x9e3a3000 - 0x9e3e2fff libglslcompiler.so ???
0xa0673000 - 0xa067bfff gralloc.exynos5.so ???
0xa1ae2000 - 0xa1cfefff data@app@com.absolutist.testwebview-2@base.apk@classes.dex ???
0xa2d00000 - 0xa51dbfff libwebviewchromium.so ???
0xa922b000 - 0xa929afff libGLESv2_POWERVR_SGX544_115.so ???
0xa929e000 - 0xa93bbfff libusc.so ???
0xad1a9000 - 0xad302fff libstagefright.so ???
0xae738000 - 0xae73efff librs_jni.so ???
0xae741000 - 0xae744fff libaudioeffect_jni.so ???
0xae747000 - 0xae748fff libsoundpool.so ???
0xae74b000 - 0xae762fff libjavacrypto.so ???
0xae766000 - 0xae771fff libstagefright_amrnb_common.so ???
0xae774000 - 0xae78bfff libvorbisidec.so ???
0xae78f000 - 0xae7eafff libsavsff.so ???
0xae904000 - 0xae904fff libwebviewchromium_loader.so ???
0xae907000 - 0xae907fff libjnigraphics.so ???
0xae90a000 - 0xae90ffff libcompiler_rt.so ???
0xae912000 - 0xae91efff libandroid.so ???
0xae922000 - 0xae924fff libpvrANDROID_WSEGL.so ???
0xae927000 - 0xae9a5fff libGLESv1_CM_POWERVR_SGX544_115.so ???
0xae9a8000 - 0xae9e4fff libsrv_um.so ???
0xae9e8000 - 0xae9fafff libIMGegl.so ???
0xae9fd000 - 0xae9fefff libEGL_POWERVR_SGX544_115.so ???
0xaed03000 - 0xaed04fff libstagefright_yuv.so ???
0xaed07000 - 0xaed26fff libstagefright_omx.so ???
0xaed2b000 - 0xaed2ffff libstagefright_avc_common.so ???
0xaed32000 - 0xaed3afff libsfextcp.so ???
0xaed3d000 - 0xaed44fff libsecuibc.so ???
0xaed47000 - 0xaed93fff libsavscmn.so ???
0xaed95000 - 0xaed96fff libpowermanager.so ???
0xaed9a000 - 0xaedd2fff libopus.so ???
0xaedd6000 - 0xaededfff libdrmframework.so ???
0xaedf3000 - 0xaedf5fff libvirtualdisplay.so ???
0xaee13000 - 0xaee6bfff libPVROCL.so ???
0xaf401000 - 0xaf401fff libstagefright_enc_common.so ???
0xaf404000 - 0xaf405fff libhwcutilsmodule.so ???
0xaf408000 - 0xaf408fff libmpp.so ???
0xaf40b000 - 0xaf40bfff libion.so ???
0xaf40e000 - 0xaf414fff libMcClient.so ???
0xaf417000 - 0xaf41efff libhwcutils.so ???
0xaf422000 - 0xaf426fff libexynosscaler.so ???
0xaf42a000 - 0xaf431fff libexynosgscaler.so ???
0xaf434000 - 0xaf439fff libdisplay.so ???
0xaf43d000 - 0xaf441fff libhdmi.so ???
0xaf444000 - 0xaf445fff libexynosutils.so ???
0xaf448000 - 0xaf44cfff libexynosv4l2.so ???
0xaf44f000 - 0xaf456fff libExynosHWCService.so ???
0xaf45b000 - 0xaf46dfff libmtp.so ???
0xaf705000 - 0xaf70dfff libjhead.so ???
0xaf711000 - 0xaf72ffff libexif.so ???
0xaf73e000 - 0xaf777fff libmedia_jni.so ???
0xaf7d6000 - 0xaf7d6fff memtrack.exynos5.so ???
0xafb42000 - 0xafcf9fff data@app@com.google.android.webview-2@base.apk@classes.dex ???
0xb0eec000 - 0xb0f20fff libjavacore.so ???
0xb3b27000 - 0xb3b2bfff libpvr2d.so ???
0xb3b50000 - 0xb3b51fff libwebviewchromium_plat_support.so ???
0xb49e8000 - 0xb4cf2fff libart.so ???
0xb4f29000 - 0xb4f2efff libbacktrace_libc++.so ???
0xb4f52000 - 0xb4f53fff libusbhost.so ???
0xb4f56000 - 0xb4f90fff libssl.so ???
0xb4f97000 - 0xb4fa0fff libsoundtrigger.so ???
0xb4fa6000 - 0xb4faefff libsensorhub.so ???
0xb4fb3000 - 0xb4fc2fff libselinux.so ???
0xb4fc5000 - 0xb4fc7fff libsecnativefeature.so ???
0xb4fca000 - 0xb4fcbfff libprocessgroup.so ???
0xb4fce000 - 0xb5417fff libpdfium.so ???
0xb542e000 - 0xb542ffff libnetd_client.so ???
0xb5432000 - 0xb5435fff libnativehelper.so ???
0xb5439000 - 0xb543afff libnativebridge.so ???
0xb543d000 - 0xb5445fff libminikin.so ???
0xb5449000 - 0xb5449fff libmemtrack.so ???
0xb544c000 - 0xb545efff libstagefright_foundation.so ???
0xb5462000 - 0xb5468fff libsec_km.so ???
0xb546c000 - 0xb546ffff libsdp_crypto.so ???
0xb5472000 - 0xb5473fff libpersona.so ???
0xb5477000 - 0xb54fffff libsqlite.so ???
0xb5506000 - 0xb569afff libomadrm.so ???
0xb56b0000 - 0xb56fffff libsonivox.so ???
0xb5707000 - 0xb5711fff libcommon_time_client.so ???
0xb5716000 - 0xb571dfff libnbaio.so ???
0xb5720000 - 0xb57b9fff libmedia.so ???
0xb57da000 - 0xb5819fff libinputflinger.so ???
0xb581e000 - 0xb5835fff libinput.so ???
0xb583b000 - 0xb5844fff libimg_utils.so ???
0xb5848000 - 0xb5863fff libquramimagecodec.so ???
0xb5866000 - 0xb5894fff libjpeg.so ???
0xb5898000 - 0xb5898fff libfimg2d_board.so ???
0xb589b000 - 0xb589efff libfimg.so ???
0xb58a1000 - 0xb5b0afff libskia.so ???
0xb5b2c000 - 0xb5b3afff libsamsungeffect.so ???
0xb5b3d000 - 0xb5b57fff libRScpp.so ???
0xb5b5a000 - 0xb5b7cfff libpng.so ???
0xb5b82000 - 0xb5bd7fff libft2.so ???
0xb5bdc000 - 0xb5c16fff libbcinfo.so ???
0xb5c19000 - 0xb5c38fff libbcc.so ???
0xb5c5c000 - 0xb5ce9fff libc++.so ???
0xb5cf1000 - 0xb65dffff libLLVM.so ???
0xb6630000 - 0xb6665fff libRS.so ???
0xb666a000 - 0xb66bdfff libhwui.so ???
0xb66c3000 - 0xb67c6fff libicuuc.so ???
0xb67d7000 - 0xb67dafff libgabi++.so ???
0xb67dd000 - 0xb6937fff libicui18n.so ???
0xb6944000 - 0xb6989fff libharfbuzz_ng.so ???
0xb698c000 - 0xb698efff libwpa_client.so ???
0xb6991000 - 0xb6995fff libnetutils.so ???
0xb6999000 - 0xb699efff libhardware_legacy.so ???
0xb69a2000 - 0xb69b5fff libexpat.so ???
0xb69b9000 - 0xb6ae4fff libcrypto.so ???
0xb6afd000 - 0xb6afdfff libcc_manager.so ???
0xb6b00000 - 0xb6b00fff libsync.so ???
0xb6b03000 - 0xb6b03fff libhardware.so ???
0xb6b07000 - 0xb6b10fff libui.so ???
0xb6b13000 - 0xb6b1afff libremotedesktop_client.so ???
0xb6b1e000 - 0xb6b60fff libgui.so ???
0xb6b6e000 - 0xb6b73fff libcamera_metadata.so ???
0xb6b76000 - 0xb6baffff libcamera_client.so ???
0xb6bbb000 - 0xb6bbdfff libspeexresampler.so ???
0xb6bc1000 - 0xb6bc4fff libaudioutils.so ???
0xb6bc7000 - 0xb6bdefff libz.so ???
0xb6be1000 - 0xb6be1fff libion_exynos.so ???
0xb6be4000 - 0xb6c0cfff libbinder.so ???
0xb6c16000 - 0xb6c3afff libandroidfw.so ???
0xb6c3e000 - 0xb6c48fff libGLESv2.so ???
0xb6c4b000 - 0xb6c50fff libGLESv1_CM.so ???
0xb6c53000 - 0xb6c54fff libETC1.so ???
0xb6c57000 - 0xb6c58fff libunwind-ptrace.so ???
0xb6c5b000 - 0xb6c66fff libunwind.so ???
0xb6caf000 - 0xb6cb2fff libgccdemangle.so ???
0xb6cb8000 - 0xb6cbdfff libbacktrace.so ???
0xb6cc1000 - 0xb6cd5fff libutils.so ???
0xb6cd9000 - 0xb6d0ffff libstlport.so ???
0xb6d14000 - 0xb6d1efff libcutils.so ???
0xb6d22000 - 0xb6d8efff libGLES_trace.so ???
0xb6d93000 - 0xb6df0fff libEGL.so ???
0xb6e00000 - 0xb6ee9fff libandroid_runtime.so ???
0xb6f00000 - 0xb6f01fff libstdc++.so ???
0xb6f04000 - 0xb6f1afff libm.so ???
0xb6f1e000 - 0xb6f21fff liblog.so ???
0xb6f25000 - 0xb6f8afff libc.so ??? (WARNING: No symbols, libc.so, C32E8C7F492AEB87E030F7B32C1C52BF0)
0xb6f9c000 - 0xb6f9cfff libsigchain.so ???
0xb6fa3000 - 0xb6faffff linker ???
0xb6fb3000 - 0xb6fb5fff app_process32 ???
,
May 25 2017
can't reproduce on Nexus 6P
,
May 25 2017
From the microdump address space stats: by the time the microdump was taken there actually should have been (just about) enough address space to satisfy this allocation (there's just one single 64MB chunk available, and only 149MB free total). But, there's loads of failures from the GPU driver between the initial OOM and the microdump - it's likely that we actually were out of memory at that point, but while the GPU driver was failing it unmapped some stuff. Unfortunately we don't dump the full address space map in a microdump so it's hard to see what was taking up all the space (minidumps do have this). The page doesn't appear to use an unusual amount of memory on my devices, but it also doesn't appear to load properly (it renders a logo, background, and a weird empty frame, and spews a bunch of JS errors) so it's hard to tell whether something's going wrong or not. It's possible we're tickling a GPU driver bug here that's making it misbehave and map too much stuff, but that's just a guess. We either need to be able to reproduce this (probably on the same device/build if it's a driver bug), or to get more data about what's mapped at the time of the crash (from a minidump, if there's a way to get it to generate that on a production build, or by doing a special build for the reporter that dumps it some other way?)
,
May 26 2017
satyavathir@ would you like to verify if this issue could be reproduced?
,
May 26 2017
,
May 30 2017
I have checked on Samsung S4 Verizon(SCH-I545)/LRX22C with WebView :43.0.2357.121 (system), 58..0.3029.83 (Stable) and 59.0.3071.71(Beta). I didn't see any crashes on app launch. I have also checked few things: Kept the app in #0 idle for 3 to 5 minutes - no crashes Back grounded the app, launched other WebView apps, launched app #0 from recent apps - no crashes Clicked the links(like Facebook, G+, twitter etc) within app - no crashes Clicked on the deck of apps- > it takes to Play store - no crashes May be we should find out more info, from the user? Thanks!
,
May 31 2017
Does the device you tested this on have the same build fingerprint: samsung/ja3gxx/ja3g:5.0.1/LRX22C/I9500XXUHPK1:user/release-keys
,
Jun 7 2017
I suppose, you can't reproduce because different GPU: Samsung Galaxy S4 I9500XXUHPK1 has PowerVR SGX 544MP, but Samsung S4 Verizon SCH-I545 has Adreno 320GPU.
,
Jun 7 2017
Thanks for confirming that. Satya, do we have one of the S4 variants with the Exynos SoC and Adreno GPU? I think those might not be sold in the US. The one you tested is the Snapdragon SoC with PVR GPU, so if it's a driver issue as I somewhat suspect then it won't repro there.
,
Jun 15 2017
torne@ - we have only S4 Verizon/L for WebView (we have many S4's, mostly < L versions). I will recheck all remaining S4's to see, if any of these matches to the exact GPU & update the bug. Thanks!
,
Jun 20 2017
satyavathir@, the folks in India (ET) may have a GT-I9500 with L. battun@, can you work with them to repro?
,
Jun 21 2017
I have checked on Samsung Galaxy S4 Verizon(GT-I9500 PowerVR SGX 544MP)/LRX22C with WebView :43.0.2357.121 (system), 58..0.3029.83 (Stable) and 59.0.3071.71(Beta). I didn't see any crashes on app launch. I have also tried by keeping the app back ground, idle for 5 to 10 minutes - we don't see any crash. Thanks!
,
Jun 22 2017
Device fingerprint: samsung/ja3gub/ja3g:5.0.1/LRX22C/I9500UBUHOK2:user/release-keys Thanks!
,
Jun 22 2017
Thanks for trying. peabeynarovich@, since we can't reproduce this the only way we'll be able to figure out what's happening is if you can get us some more information. A memory-infra trace might help: https://chromium.googlesource.com/chromium/src/+/master/docs/memory-infra/README.md
,
Jun 29 2017
Hi How to start chromium memory trace on Android WebView app?
,
Jun 29 2017
Thank you for providing more feedback. Adding requester "torne@chromium.org" to the cc list and removing "Needs-Feedback" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jun 29 2017
Assuming that you can't find a userdebug ROM for the device, you would need to add the following to your app initialization code: WebView.setWebContentsDebuggingEnabled(true); Then, when the device is connected via usb to a machine running chrome, browse to: chrome://inspect?tracing You should see a list of WebViews in your application, along with a "trace" link. Clicking on the trace link and then on "Record" should give you a list of categories to trace. If you select the memory-infra category, record a trace, and then save it to disk, the file produced should be the information we require.
,
Jul 3 2017
Hi again Trace file is attached.
,
Jul 3 2017
Thank you for providing more feedback. Adding requester "tobiasjs@chromium.org" to the cc list and removing "Needs-Feedback" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jul 14 2017
all the leaks are mappings from /dev/pvrsrvkm, that's the powervr graphics driver. However there is no growth in chrome's internal tracking, meaning chrome/webview considers that it has freed all its allocated resources (or at least there is no growth). So this is likely a gpu driver bug. Can you open chrome browser on the device, go to chrome://gpu, and print that page as pdf and attach it here? Don't really have high hopes of actually having this fixed, since device has probably long past oem support, and it's unlikely to get driver fixes, and finding workaround is hard.
,
Jul 17 2017
Adding needs-feedback for the pdf of chrome://gpu
,
Jul 18 2017
I can't create pdf because I didn't find menu Share. I have made screenshots (attached)
,
Jul 18 2017
Thank you for providing more feedback. Adding requester "ntfschr@chromium.org" to the cc list and removing "Needs-Feedback" label. For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot
,
Jul 18 2017
we have one s4 with the same gpu, but it's stuck on kitkat..
,
Jul 18 2017
played with the kitkat device a bit test shell leaks plenty of memory, expanding the avoid_egl_image_target_texture_reuse workaround does not help, disabling texture reuse in makes the rate of leak much much faster. so this is driver leaking some memory each time is created and shared. And on kitkat at least, KHR_fence_sync is blacklisted, so technically, webview doesn't even have enough extensions to operate in GL mode (although the code that checks for that doesn't respect the blacklist) wonder if we should just blacklist hardware acceleration altogether on this config..
,
Nov 17 2017
is this still an issue? is WebView memory still leaking on Samsung Galaxy S4?
,
Nov 21 2017
Yes, WebView memory is still leaking on Samsung Galaxy S4 (Chrome version: 62.0.3202.84)
,
Dec 22 2017
It seems boliu@ had investigated this and there is no obvious way from WebView side to get this fixed. Potential workaround is blacklist hardware acceleration on this config per c#28. cc boliu@ in case he want to make decision. I'll change the status to available since the issue looks confirmed already.
,
Dec 22 2017
Changing to available + p3 ensures that this will never be looked at by another team member. Is this configuration popular? Does it merit real action from our team (i.e. blacklisting the hardware)?
,
Dec 22 2017
Hmm, S4 still shares reasonable amount of the market from the public resource (0.8% [1]), but with this GPU, maybe only Samsung knows. boliu@, assigning this to you, feel free to reassign though. [1] https://www.appbrain.com/stats/top-android-phones
,
Feb 8 2018
,
Sep 27
,
Sep 27
|
||||||||||||||||||
►
Sign in to add a comment |
||||||||||||||||||
Comment 1 by rsgav...@chromium.org
, May 24 2017