Forums - qnn-model-lib-generator fails

Hi Qualcomm guys,

I am trying to quantize and convert 3B Llama model for mobile device with Snapdragon 8 Gen 3, but I ended up being caught by an error while I was running 'qnn-model-lib-generator' tool.

With onnx model that I got with AIMET's calibration process as input, I ran `qnn-onnx-converter`. As a result, I got a pair of .cpp, and .bin files, so far so good.

However, when I tried to run `qnn-model-lib-generator` with the .cpp, .bin files as input, it stops running with the following error:

/usr/bin/../lib/gcc/x86_64-linux-gnu/9/crtbeginS.o: in function `deregister_tm_clones':

crtstuff.c:(.text+0x3): relocation truncated to fit: R_X86_64_PC32 against `.tm_clone_table'

crtstuff.c:(.text+0xa): relocation truncated to fit: R_X86_64_PC32 against symbol `__TMC_END__' defined in .data section in libs/x86_64-linux-clang/lib3B-26L_1_of_1.so

/usr/bin/../lib/gcc/x86_64-linux-gnu/9/crtbeginS.o: in function `register_tm_clones':

crtstuff.c:(.text+0x33): relocation truncated to fit: R_X86_64_PC32 against `.tm_clone_table'

crtstuff.c:(.text+0x3a): relocation truncated to fit: R_X86_64_PC32 against symbol `__TMC_END__' defined in .data section in libs/x86_64-linux-clang/lib3B-26L_1_of_1.so

/usr/bin/../lib/gcc/x86_64-linux-gnu/9/crtbeginS.o: in function `__do_global_dtors_aux':

crtstuff.c:(.text+0x76): relocation truncated to fit: R_X86_64_PC32 against `.bss'

crtstuff.c:(.text+0x9e): relocation truncated to fit: R_X86_64_PC32 against `.bss'

clang: error: linker command failed with exit code 1 (use -v to see invocation)

make: *** [Makefile.linux-x86_64:104: libs/x86_64-linux-clang/lib3B-26L_1_of_1.so] Error 1

I googled with the error message, but with so few clues, I couldn't get a fine solution for this.

Can anybody help me deal with this error?