Merge pull request #5037 from dhiltgen/faster_win_build

More parallelism on windows generate

llm/generate/gen_windows.ps1

 #!powershell
 
-$ErrorActionPreference = "Stop"
-
 function amdGPUs {
     if ($env:AMDGPU_TARGETS) {
         return $env:AMDGPU_TARGETS
@@ -85,9 +83,9 @@ function init_vars {
 function git_module_setup {
     # TODO add flags to skip the init/patch logic to make it easier to mod llama.cpp code in-repo
     & git submodule init
-    if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
+    if ($LASTEXITCODE -ne 0) { throw($LASTEXITCODE)}
     & git submodule update --force "${script:llamacppDir}"
-    if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
+    if ($LASTEXITCODE -ne 0) { throw($LASTEXITCODE)}
 }
 
 function apply_patches {
@@ -121,10 +119,15 @@ function build {
     write-host "generating config with: cmake -S ${script:llamacppDir} -B $script:buildDir $script:cmakeDefs"
     & cmake --version
     & cmake -S "${script:llamacppDir}" -B $script:buildDir $script:cmakeDefs
-    if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
-    write-host "building with: cmake --build $script:buildDir --config $script:config $($script:cmakeTargets | ForEach-Object { `"--target`", $_ })"
-    & cmake --build $script:buildDir --config $script:config ($script:cmakeTargets | ForEach-Object { "--target", $_ })
-    if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
+    if ($LASTEXITCODE -ne 0) { throw($LASTEXITCODE)}
+    if ($cmakeDefs -contains "-G") {
+        $extra=@("-j8")
+    } else {
+        $extra= @("--", "/p:CL_MPcount=8")
+    }
+    write-host "building with: cmake --build $script:buildDir --config $script:config $($script:cmakeTargets | ForEach-Object { `"--target`", $_ }) $extra"
+    & cmake --build $script:buildDir --config $script:config ($script:cmakeTargets | ForEach-Object { "--target", $_ }) $extra
+    if ($LASTEXITCODE -ne 0) { write-host "cmake build exit status $LASTEXITCODE"; throw($LASTEXITCODE)}
     # Rearrange output to be consistent between different generators
     if ($null -ne ${script:config} -And (test-path -path "${script:buildDir}/bin/${script:config}" ) ) {
         mv -force "${script:buildDir}/bin/${script:config}/*" "${script:buildDir}/bin/"
@@ -138,7 +141,7 @@ function sign {
         foreach ($file in @(get-childitem "${script:buildDir}/bin/*.exe") + @(get-childitem "${script:buildDir}/bin/*.dll")){
             & "${script:SignTool}" sign /v /fd sha256 /t http://timestamp.digicert.com /f "${script:OLLAMA_CERT}" `
                 /csp "Google Cloud KMS Provider" /kc "${env:KEY_CONTAINER}" $file
-            if ($LASTEXITCODE -ne 0) { exit($LASTEXITCODE)}
+            if ($LASTEXITCODE -ne 0) { throw($LASTEXITCODE)}
         }
     }
 }
@@ -213,7 +216,13 @@ function build_static() {
     }
 }
 
-function build_cpu($gen_arch) {
+function build_cpu() {
+    if ($script:ARCH -eq "arm64") {
+        $gen_arch = "ARM64"
+    } else { # amd64
+        $gen_arch = "x64"
+    }
+
     if ((-not "${env:OLLAMA_SKIP_CPU_GENERATE}" ) -and ((-not "${env:OLLAMA_CPU_TARGET}") -or ("${env:OLLAMA_CPU_TARGET}" -eq "cpu"))) {
         # remaining llama.cpp builds use MSVC 
         init_vars
@@ -270,7 +279,15 @@ function build_cuda() {
         init_vars
         $script:buildDir="../build/windows/${script:ARCH}/cuda$script:CUDA_VARIANT"
         $script:distDir="$script:DIST_BASE\cuda$script:CUDA_VARIANT"
-        $script:cmakeDefs += @("-A", "x64", "-DLLAMA_CUDA=ON", "-DLLAMA_AVX=on", "-DLLAMA_AVX2=off", "-DCUDAToolkit_INCLUDE_DIR=$script:CUDA_INCLUDE_DIR", "-DCMAKE_CUDA_ARCHITECTURES=${script:CMAKE_CUDA_ARCHITECTURES}")
+        $script:cmakeDefs += @(
+            "-A", "x64",
+            "-DLLAMA_CUDA=ON",
+            "-DLLAMA_AVX=on",
+            "-DLLAMA_AVX2=off",
+            "-DCUDAToolkit_INCLUDE_DIR=$script:CUDA_INCLUDE_DIR",
+            "-DCMAKE_CUDA_FLAGS=-t8"
+            "-DCMAKE_CUDA_ARCHITECTURES=${script:CMAKE_CUDA_ARCHITECTURES}"
+            )
         if ($null -ne $env:OLLAMA_CUSTOM_CUDA_DEFS) {
             write-host "OLLAMA_CUSTOM_CUDA_DEFS=`"${env:OLLAMA_CUSTOM_CUDA_DEFS}`""
             $script:cmakeDefs +=@("${env:OLLAMA_CUSTOM_CUDA_DEFS}")
@@ -391,16 +408,29 @@ init_vars
 if ($($args.count) -eq 0) {
     git_module_setup
     apply_patches
-    build_static
-    if ($script:ARCH -eq "arm64") {
-        build_cpu("ARM64")
-    } else { # amd64
-        build_cpu("x64")
-        build_cpu_avx
-        build_cpu_avx2
-        build_cuda
-        build_oneapi
-        build_rocm
+
+    $tasks = @("build_static", "build_cpu")
+    $jobs = @()
+    if ($script:ARCH -ne "arm64") {
+        $tasks += $("build_cpu_avx", "build_cpu_avx2", "build_cuda", "build_oneapi", "build_rocm")
+    }
+    foreach ($t in $tasks) {
+        $jobs += @(Start-ThreadJob -ThrottleLimit 12 -FilePath .\gen_windows.ps1 -ArgumentList $t -Name $t)
+    }
+    get-job
+    foreach ($job in $jobs) {
+        write-host "----" $job.Name output follows
+        receive-job -wait -job $job
+        write-host "----" $job.Name $job.State
+        write-host ""
+        if ($job.State -contains 'Failed') {
+            cleanup
+            write-host "Terminating remaining jobs (this takes a while, you can ^C)"
+            # TODO find some way to kill the spawned cmake processes faster
+            remove-job -force -job $jobs
+            exit(-1)
+        }
+        get-job
     }
 
     cleanup