port for intel x86 android phones??

[[Unknown]]

You said "anymore", did it work before? Did jit work?

If it worked before on an Android phone with the official apk, it was running on ARM. Could've been an ARM emulator on your phone.

-[Unknown]

[fischfrosch]

Yes it works before with JIT.

Ridge Race 2 with about between 30-40 FPS

Every Intel Phone emulates ARM.

[[Unknown]]

Well, okay then. Is there a way to write a native x86 app?

x86 phone emulating ARM running PPSSPP emulating MIPS = not the fastest possible option.
x86 phone running PPSSPP emulating MIPS = there we go, that's what we want.

So the idea would be to get your phone not to emulate ARM, at least not for PPSSPP.

-[Unknown]

[fischfrosch]

I don´t know it, but i found this :

http://www.infoworld.com/d/mobile-techno...now-181105

Maybe it will help you

[Henrik]

[Unknown], yes, I'm now building the Android build for x86 too, it's included as an additional .so . However, it seems it crashes...

[fischfrosch]

Will test all the x86 builds, hope for the best.

The Intel CPU has a lot of power.

[techloomp]

Correct, still don't work for x86. When I try to run PPSSPP 0.7 on RazrI logcat shows:

Code:

04-01 15:05:36.784   502   502 I ActivityManager: START {act=android.intent.action.MAIN cat=[android.intent.category.LAUNCHER] flg=0x10000000 pkg=org.ppsspp.ppsspp cmp=org.ppsspp.ppsspp/.PpssppActivity u=0} from pid 2649
04-01 15:05:36.784   502   541 W ApplicationContext: Unable to create files directory /data/data/org.ppsspp.ppsspp/files
04-01 15:05:36.844  2871  2871 I dalvikvm: Turning on JNI app bug workarounds for target SDK version 8...
04-01 15:05:36.844   502   541 W IMGSRV  : :0: gralloc_device_alloc: ID: 74 handle: 0x59e2b250 size: 540 x 888 fmt: 2 usage: 0x933
04-01 15:05:36.844   502   742 I ActivityManager: Start proc org.ppsspp.ppsspp for activity org.ppsspp.ppsspp/.PpssppActivity: pid=2871 uid=10112 gids={1015, 1028, 3003}
04-01 15:05:36.884   502   747 D dalvikvm: WAIT_FOR_CONCURRENT_GC blocked 0ms
04-01 15:05:36.954   502   747 D dalvikvm: GC_EXPLICIT freed 2606K, 42% free 18013K/30672K, paused 3ms+5ms, total 76ms
04-01 15:05:36.974  1515  1530 V ap      : set to portrait
04-01 15:05:36.974  2871  2871 E Trace   : error opening trace file: No such file or directory (2)
04-01 15:05:37.004  2871  2871 D dalvikvm: Trying to load lib /data/data/org.ppsspp.ppsspp/lib/libppsspp_jni.so 0x4225c500
04-01 15:05:37.014  2871  2871 D houdini : [2871] Loading library(version: 3.1.0.42836 RELEASE)... successfully.
04-01 15:05:37.024   762   762 W IMGSRV  : :0: gralloc_unregister_buffer: ID: 71 handle: 0x583f9bc0 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.024   762   762 W IMGSRV  : :0: gralloc_unregister_buffer: ID: 72 handle: 0x583e3990 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.024   762   762 W IMGSRV  : :0: gralloc_unregister_buffer: ID: 73 handle: 0x5483c2e0 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.024   502   764 W IMGSRV  : :0: gralloc_device_free: ID: 72 handle: 0x5ce051c0 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.024   502   764 W IMGSRV  : :0: gralloc_device_free: ID: 73 handle: 0x5ce06f70 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.044  2871  2871 F libc    : Fatal signal 11 (SIGSEGV) at 0x00000000 (code=1), thread 2871 (g.ppsspp.ppsspp)
04-01 15:05:37.144   141   141 I DEBUG   : *** *** *** *** *** *** *** *** *** *** *** *** *** *** *** ***
04-01 15:05:37.144   141   141 I DEBUG   : Build fingerprint: 'motorola/XT890_rtcoreeu/smi:4.1.2/9.8.2I-50_SMI-26/1358465787:user/release-keys'
04-01 15:05:37.144   141   141 I DEBUG   : pid: 2871, tid: 2871, name: g.ppsspp.ppsspp  >>> org.ppsspp.ppsspp <<<
04-01 15:05:37.144   141   141 I DEBUG   : signal 11 (SIGSEGV), code 1 (SEGV_MAPERR), fault addr 00000000
04-01 15:05:37.314   502   514 W IMGSRV  : :0: gralloc_device_free: ID: 71 handle: 0x588b9e10 size: 540 x 888 fmt: 5 usage: 0xb00
04-01 15:05:37.314   502   567 I WifiHW  : CMD: SIGNAL_POLL
04-01 15:05:37.314   574   574 D wpa_supplicant: ctrl_iface: SIGNAL_POLL
04-01 15:05:37.314   574   574 D wpa_supplicant: nl80211: survey data missing!
04-01 15:05:37.314   502   567 I WifiHW  : REPLY: RSSI=-37
04-01 15:05:37.314   502   567 I WifiHW  : REPLY: LINKSPEED=54
04-01 15:05:37.314   502   567 I WifiHW  : REPLY: NOISE=-97
04-01 15:05:37.314   502   567 I WifiHW  : REPLY: FREQUENCY=2462
04-01 15:05:37.354   502   514 D hwcomposer: isOverlayLayer: got an overlay layer
04-01 15:05:37.354   502   514 D hwcomposer: int IntelDisplayPlaneManager::setZOrderConfig(int, int): set zorder: 1
04-01 15:05:37.364   502   514 D hwcomposer: void IntelHWComposerLayerList::clearWithOpenGL() const: clear fb here, size 540 x 960
04-01 15:05:37.374   502   514 D hwcomposer: isOverlayLayer: got an overlay layer
04-01 15:05:37.374   502   514 D hwcomposer: void IntelHWComposerLayerList::clearWithOpenGL() const: clear fb here, size 540 x 960
04-01 15:05:37.534   141   141 I DEBUG   :     eax 0000000e  ebx 32000020  ecx 00000001  edx 000000ca
04-01 15:05:37.534   141   141 I DEBUG   :     esi 00000000  edi ffffffff
04-01 15:05:37.534   141   141 I DEBUG   :     xcs 00000073  xds 0000007b  xes 0000007b  xfs 0000003b  xss 0000007b
04-01 15:05:37.534   141   141 I DEBUG   :     eip 53719b0e  ebp 3200ff6c  esp 3200fe74  flags 00210293
04-01 15:05:37.604   502   530 I BootReceiver: Copying /data/tombstones/tombstone_01 to DropBox (SYSTEM_TOMBSTONE)
04-01 15:05:37.614   142   142 D Zygote  : Process 2871 terminated by signal (11)
04-01 15:05:37.614   502   704 I ActivityManager: Process org.ppsspp.ppsspp (pid 2871) has died.
04-01 15:05:37.614   502   704 W ActivityManager: Force removing ActivityRecord{21225848 org.ppsspp.ppsspp/.PpssppActivity}: app died, no saved state

[Henrik]

Okay it blows up pretty much immediately. To do some testing, I'll see if I can get the Android development emulator working with x86 and GPU, apparently it's possible nowadays.

[techloomp]

(04-01-2013 05:38 PM)Henrik Wrote:  Okay it blows up pretty much immediately. To do some testing, I'll see if I can get the Android development emulator working with x86 and GPU, apparently it's possible nowadays.

You should try running emulator with Atom CPU and with HAX enabled in fast virt mode (with virtualization) - you'll be amazed how fast is this setup compared to ARM emulation.

[VIRGIN KLM]

(04-02-2013 08:07 PM)techloomp Wrote:  You should try running emulator with Atom CPU and with HAX enabled in fast virt mode (with virtualization) - you'll be amazed how fast is this setup compared to ARM emulation.

That's because you do a simple recompilation, x86 to x86/x64!

[techloomp]

(04-03-2013 02:13 AM)VIRGIN KLM Wrote:  

(04-02-2013 08:07 PM)techloomp Wrote:  You should try running emulator with Atom CPU and with HAX enabled in fast virt mode (with virtualization) - you'll be amazed how fast is this setup compared to ARM emulation.

That's because you do a simple recompilation, x86 to x86/x64!

As I understand HAX there is no recompilation, android x86 code runs directly on host (x86) machine.

[fischfrosch]

Something new on the x86 frontline ? ;D

[fischfrosch]

0.7.5 from Google Play works fine ;D

[[Unknown]]

Yeah, it's back to slowmo mode. How is the speed?

-[Unknown]

[fischfrosch]

The speed is almost the same.

Menuscreens runs with 60 FPS in Monster Hunter United and Star Wars Battlefront 2 and than Ingame around 15 FPS

God of War is all the time 6 FPS

End of April i will buy a new smartphone maybe the Galaxy S4 or the Asus Padfone Infinity