Changes to CMakePresets.json to add ninja clang target on windows

[Srihari-mcw]

This PR adds changes to CMakePresets.json to add a ninja clang target on windows. This will help users to build llama.cpp with clang and ninja dependencies, with relative ease. The additional dependencies are listed in build.md

Build commands :

    cmake --preset x64-windows-ninja-release
    cmake --build build-x64-windows-ninja-release

Experiments were conducted to test llama.cpp with MSVC (Visual Studio Generator), Clang (Ninja) and Clang-Cl (Visual Studio Generator) and to compare the performance

Best performance was seen among the three for clang with ninja generator

Ninja Generator, Clang.exe/Clang++.exe (17.0.3)

• Clang.exe and clang++.exe are obtained from the same path as clang-cl - C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\Llvm\x64\bin
• CMAKE Command : cmake -G "Ninja" -S . -B Windows-build/ -DCMAKE_BUILD_TYPE=Release -DCMAKE_C_COMPILER=clang.exe -DCMAKE_CXX_COMPILER=clang++.exe
• Ninja path : C:\Program Files\Microsoft Visual Studio\2022\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\Ninja\ninja.exe and version (1.11.0)
• Additional library depedencies are resolved with : set LIB=C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\um\x64;C:\Program Files\Microsoft Visual Studio\2022\Community\VC\Tools\MSVC\14.41.34120\lib\x64\uwp;C:\Program Files (x86)\Windows Kits\10\Lib\10.0.22621.0\ucrt\x64 before cmake and build

Both the clang builds are done without the openmp dependencies

The tests were conducted in AMD Raphael 7600X which supports the following flags (The same were kept on in our windows tests) :
AVX = 1 | AVX_VNNI = 0 | AVX2 = 1 | AVX512 = 1 | AVX512_VBMI = 1 | AVX512_VNNI = 1 | AVX512_BF16 = 1 | FMA = 1 | NEON = 0 | SVE = 0 | ARM_FMA = 0 | F16C = 1 | FP16_VA = 0 | WASM_SIMD = 0 | BLAS = 0 | SSE3 = 1 | SSSE3 = 1 | VSX = 0 | MATMUL_INT8 = 0 | LLAMAFILE = 1|

Prompt Processing

model	Compiler/Generator	size	params	backend	threads	test	t/s	speedup	Commit id
llama 7B Q4_0	MSVC Compiler/ VS Generator	3.56 GiB	6.74 B	CPU	6	pp 512	48.14 ± 0.67		231f936
llama 7B Q4_0	clang-cl.exe/ VS Generator	3.56 GiB	6.74 B	CPU	6	pp 512	50.23 ± 0.91	4.33%	231f936
llama 7B Q4_0	clang.exe/ Ninja Generator	3.56 GiB	6.74 B	CPU	6	pp 512	55.69 ± 0.71	15.68%	231f936

Text Generation

model	Compiler/Generator	size	params	backend	threads	test	t/s	speedup	Commit id
llama 7B Q4_0	MSVC Compiler/ VS Generator	3.56 GiB	6.74 B	CPU	6	tg 128	14.8 ± 0.01		231f936
llama 7B Q4_0	clang-cl.exe/ VS Generator	3.56 GiB	6.74 B	CPU	6	tg 128	14.95 ± 0.02	1.00%	231f936
llama 7B Q4_0	clang.exe/ Ninja Generator	3.56 GiB	6.74 B	CPU	6	tg 128	14.95 ± 0.02	1.00%	231f936

OS : Windows

Tests done with Meta Llama2 7B model

[max-krasnyansky]

Do you want to maybe just match what I did for the Windows on ARM64 arm64-windows-llvm ?
i.e add CMake toolchain file that uses llvm/clang and call the targets x64-windows-llvm that use Ninja.
The main performance difference is llvm/clang not Ninja.
We should just use Ninja as the default anyway. It's faster in general and cleaner output.

[Srihari-mcw]

Hi @max-krasnyansky , the preset target names were changed as per the above comment. The aim is to facilitate users to build with clang on x64/x86 Platforms on windows. Thanks

[max-krasnyansky]

@Srihari-mcw
Sorry. I approved first but then realized that there is still no toolchain file.
Just copy cmake/arm64-windows-llvm.cmake to cmake/x64-windows-llvm.cmake and update CFLAGS and things.
We probably don't need to use .exe in the compiler name either.

[Srihari-mcw]

HI @max-krasnyansky ,

updates were made to use .cmake file. Thanks