[IE-MYRIAD] Chromium crashed when running the quantized models

[Christywl]

Test Env:
Chromium Version: nightly build 79.0.3917.0 (74a844d)
Platform: Linux

Expected Result:
No crash.

Actual Result:
Crash happened:

$ /usr/bin/chromium-browser-unstable --no-sandbox --use-inference-engine [15260:15260:0612/135430.472859:ERROR:buffer_manager.cc(488)] [.DisplayCompositor]GL ERROR :GL_INVALID_OPERATION : glBufferData: <- error from previous GL command Received signal 11 SEGV_MAPERR 7f1818f87000 #0 0x561161ffacf9 <unknown> #1 0x561161f5e373 <unknown> #2 0x561161ffa881 <unknown> #3 0x7f182e29a390 <unknown> #4 0x5611609df969 <unknown> #5 0x5611609deb28 <unknown> #6 0x5611609d8afd <unknown> #7 0x5611609d68eb <unknown> #8 0x5611609c8b83 <unknown> #9 0x5611609d06cd <unknown> #10 0x5611609cf706 <unknown> #11 0x561162120b8b <unknown> #12 0x56116212661e <unknown> #13 0x561162125e98 <unknown> #14 0x56116211e9a6 <unknown> #15 0x56116211f37e <unknown> #16 0x5611621348de <unknown> #17 0x561161fa95bf <unknown> #18 0x561161fc0f32 <unknown> #19 0x561161fc0a69 <unknown> #20 0x56116200a80b <unknown> #21 0x561161fc06be <unknown> #22 0x561161fca883 <unknown> #23 0x561161fca6f4 <unknown> #24 0x56116200afde <unknown> #25 0x7f182e2906ba start_thread #26 0x7f182820741d clone r8: 0000000000000000 r9: 0000000000000001 r10: 0000000000000001 r11: 0000000000000001 r12: 00000000000170f4 r13: 00000000000170f4 r14: 0000000000000001 r15: 0000000000000000 di: 00001f67fa413000 si: 0000000000000000 bp: 00007f18185d3500 bx: 00000000000170f4 dx: 0000000000000001 ax: 00000000000170f4 cx: 00007f1818f2ac30 sp: 00007f18185d3330 ip: 00005611609df969 efl: 0000000000010202 cgf: 002b000000000033 erf: 0000000000000004 trp: 000000000000000e msk: 0000000000000000 cr2: 00007f1818f87000 [end of stack trace] Calling _exit(1). Core file will not be generated. [15382:15382:0612/135434.392461:ERROR:buffer_manager.cc(488)] [.DisplayCompositor]GL ERROR :GL_INVALID_OPERATION : glBufferData: <- error from previous GL command

How to Reproduce:

1. Launch chromium with /usr/bin/chromium-browser-unstable --no-sandbox --use-inference-engine and visit https://intel.github.io/webml-polyfill/examples/image_classification

2. Select LOW_POWER and unselect WASM

3. Select one quantized model:

   • MobileNet v1 Quant(TFLite)
   • MobileNet v2 Quant(TFLite)

And similar issue for Object Detection:

• SSD MobileNet v1 Quant(TFLite)
• SSD MobileNet v2 Quant(TFLite)