Musa Mahmood
dd7cf83ca6
The workers ran gpu_workers (4) runs at once whatever the images: the server model, four large images at a time, took 15 GB of a 16 GB GPU and spilled into shared system memory, and a smaller GPU would fare far worse. modules/OCR/budget.jai: each run reserves what it is expected to need, and waits while the runs in flight or the GPU's free memory leave too little; one run always goes. The need is measured when the engine is made, from the arena's peak (AllocatorGetStats MaxInUse) over runs on two input sizes, so it holds for any model and GPU: the mobile detector 6 MB + 137 MB a megapixel, the server's 4 MB + 1015, against 511 and 522 MB measured for a 2560 x 1440 tile. The limit is the GPU's free memory less 10% (at least 512 MB), read at the start of each ocr_process call (cudaMemGetInfo, ort_cuda_memory); the free memory is also read as each run starts, since finished runs' blocks stay in the arena until the stage ends. gpu-bench, RTX 5080 laptop: the server model 335 ms a capture (854 before), peak 13.1 GB (15.4), passes within 4%; the mobile model unchanged (140 ms, 5.7 GB), as its runs fit.
72 lines
2.6 KiB
Plaintext
72 lines
2.6 KiB
Plaintext
// Type-checks src/App.jai for another OS without producing or linking a binary, to catch errors
|
|
// in code your machine does not build:
|
|
//
|
|
// jai check_targets.jai - linux (from any OS)
|
|
// jai check_targets.jai - macos (from a Mac only: the Objective_C module runs code at compile time)
|
|
//
|
|
// The constants below stand in for the ones first.jai generates; keep their names in step.
|
|
#run,stallable {
|
|
set_build_options_dc(.{do_output = false});
|
|
args := get_build_options().compile_time_command_line;
|
|
target := Operating_System_Tag.LINUX;
|
|
cpu := CPU_Tag.X64;
|
|
if array_find(args, "macos") { target = .MACOS; cpu = .ARM64; }
|
|
|
|
options := get_build_options();
|
|
options.os_target = target;
|
|
options.cpu_target = cpu;
|
|
options.output_type = .NO_OUTPUT;
|
|
options.backend = .LLVM;
|
|
options.output_executable_name = "check";
|
|
options.output_path = ".build";
|
|
|
|
// rpmalloc (modules/Default_Allocator) also runs at compile time, with the target's OS layer,
|
|
// which crashes on another OS. A check for another OS uses Jai's own allocator instead, through a
|
|
// module of the same name ahead of modules/ (with the thread_idle the app calls, doing nothing).
|
|
if target != OS {
|
|
jai_modules := trim_right(options.import_path[options.import_path.count - 1], "/");
|
|
shim := tprint("%/.build/check_modules", get_working_directory());
|
|
make_directory_if_it_does_not_exist(tprint("%/Default_Allocator", shim), recursive = true);
|
|
write_entire_file(tprint("%/Default_Allocator/module.jai", shim), tprint(#string SHIM
|
|
// Written by check_targets.jai: Jai's own allocator, for type-checking another OS.
|
|
#load "%/Default_Allocator/module.jai";
|
|
thread_idle :: () {}
|
|
SHIM, jai_modules));
|
|
paths: [..] string;
|
|
array_add(*paths, shim);
|
|
array_add(*paths, ..options.import_path);
|
|
options.import_path = paths;
|
|
}
|
|
|
|
w := compiler_create_workspace("check");
|
|
set_build_options(options, w);
|
|
compiler_begin_intercept(w);
|
|
add_build_file("src/App.jai", w);
|
|
add_build_string(#string S
|
|
APP_NAME :: "mcap";
|
|
APP_FOLDER_NAME :: "mcap";
|
|
GIT_BRANCH :: "check";
|
|
GIT_HASH :: "-";
|
|
VERSION :: "0.1.25";
|
|
BUILD_DATE :: "-";
|
|
DEBUG :: true;
|
|
COMPILE_TESTS :: true;
|
|
COMPILE_ARGS :: "";
|
|
NO_CRASH_DIALOG :: false;
|
|
S, w);
|
|
while true {
|
|
message := compiler_wait_for_message();
|
|
if message.kind == .COMPLETE {
|
|
m := cast(*Message_Complete) message;
|
|
log("% check: %", target, m.error_code);
|
|
break;
|
|
}
|
|
}
|
|
compiler_end_intercept(w);
|
|
}
|
|
|
|
#import "Basic";
|
|
#import "Compiler";
|
|
#import "File";
|
|
#import "String";
|